Agarsenone, a cadinane sesquiterpenoid from Commiphora erythraea

Article abstract of DOI:10.1021/np400114b

Agarsenone (1), a new cadinane sesquiterpenoid, was isolated from the resin of Commiphora erythraea. The structures of 1 and its decomposition products agarsenolides (2a and 2b) and myrrhone (3) were established by extensive NMR spectroscopic analysis. The absolute configuration of 3 and the relative and absolute configurations of 1 were assigned by comparison of experimental and calculated optical rotatory dispersion and electronic circular dichroism spectra.

Full text of DOI:10.1021/np400114b

Article
pubs.acs.org/jnp
Agarsenone, a Cadinane Sesquiterpenoid from Commiphora
erythraea
Stefano Santoro,†,§ Stefano Superchi,*,‡ Federica Messina,† Ernesto Santoro,‡ Ornelio Rosati,†
Claudio Santi,† and M. Carla Marcotullio*,†
†Dipartimento di Chimica e Tecnologia del Farmaco-Sez. Chimica Organica, Universita
Italy
̀
di Perugia, Via del Liceo, 1-06123 Perugia,
‡Dipartimento di Scienze, Universita
̀
della Basilicata, Viale dell’Ateneo Lucano, 10- 85100 Potenza, Italy
S
* Supporting Information
ABSTRACT: Agarsenone (1), a new cadinane sesquiterpe-
noid, was isolated from the resin of Commiphora erythraea. The
structures of 1 and its decomposition products agarsenolides
(2a and 2b) and myrrhone (3) were established by extensive
NMR spectroscopic analysis. The absolute configuration of 3
and the relative and absolute configurations of 1 were assigned
by comparison of experimental and calculated optical rotatory
dispersion and electronic circular dichroism spectra.
he genus Commiphora comprises more than 150 species
properties has become an important tool for the assignment of
Tand is known for the production of resin commonly called absolute16 and relative configurations.
myrrh. Commiphora erythraea (Ehrenb.) Engl. (Burseraceae) is
RESULTS AND DISCUSSION
a small tree growing in the Arabian Peninsula and widespread
in Ethiopia, where its resin (called agarsu) is used to protect
livestock from ticks, to prevent colds and fever, to enhance
wound healing, as an antimalarial, against snake venom
poisoning, as an anti-inflammatory, as an antiseptic, and against
skin infections.1 The paucity of chemical data2,3 about C.
erythraea initiated our study of the composition of its resin.4
Furanosesquiterpenoids with eudesmane, germacrane, elemane,
or guaiane skeletons are the most abundant classes of
metabolites of oils and extracts prepared from myrrh.5,6
In our previous studies of the resin extracts we identified four
main known compounds, namely, 1,10(15)-furanogermacra-
dien-6-one (4), 1(10),4-furanodien-6-one (5), rel-3R-methoxy-
4S-furanogermacra-1E,10(15)-dien-6-one (6), and rel-2R-me-
thoxy-4R-furanogermacra-1(10)E-en-6-one (7)7,8 along with
dihydropyrocurzerenone (8),7 curzerenone (9),7 myrrhone
(3),9,10 and alismol (10)11 (Figure S1, Supporting Informa-
tion). Compounds 4−7 and myrrhone (3) showed antiradical,6
anti-inflammatory,12 and antiviral activity.13 A new investigation
of the resin permitted the isolation of two known
furanosesquiterpenoids (1114 and 1215) and a new furanoca-
dinan-2-one, named agarsenone (1).
■
Agarsenone (1) was isolated as a colorless oil from the MeOH
extract of the commercial resin. The HRMS of the quasi-
molecular [M + H]+ ion at m/z 231.1367 established the
molecular formula C15H18O2 for 1, indicating seven indices of
hydrogen deficiency, and the IR spectrum showed the presence
of a conjugated carbonyl group (1762 cm−1). 1H NMR analysis
in CDCl3 revealed the presence of a furanosesquiterpenoid
compound. During acquisition of NMR spectra, the compound
partially decomposed, and TLC analysis, further confirmed by
GC-MS, revealed the presence of a small amount of 1 along
with 3. After 48 h in CDCl3 solution, compound 1 was
completely transformed into 3.
The stability of 1 in several solvents, such as CHCl3, Et2O,
benzene, and n-hexane, was checked to find experimental
conditions suitable to prevent its decomposition and to
determine its structure. Compound 1 was found to be unstable
only in CHCl3 and in other chlorinated solvents.17 No acidic
degradation was observed on SiO2, unless chlorinated solvents
were used for the elution. NMR spectra were then recorded in
benzene-d6, allowing the determination of the agarsenone
structure (Table 1). However, some coupling constants were
taken from the 1H NMR data in CDCl3, which showed patterns
more consistent with first-order spin systems (see Experimental
Section).
Herein, we report the isolation and structure elucidation of
agarsenone (1) and its decomposition products, agarsenolides
(2a and 2b) and myrrhone (3). This work also illustrates a
successful application of computational analysis of optical
rotatory dispersion (ORD) and electronic circular dichroism
(ECD) in determining the relative and absolute configuration
of these noncrystalline natural products. In fact, in recent years,
the use of computational techniques to calculate chiroptical
The 13C NMR spectrum of 1, supported by a JMODXH
experiment,18 indicated one carbonyl, five nonprotonated sp2
Received: February 5, 2013
Published: July 11, 2013
© 2013 American Chemical Society and
American Society of Pharmacognosy
1254
dx.doi.org/10.1021/np400114b | J. Nat. Prod. 2013, 76, 1254−1259

Journal of Natural Products
Article
a
Table 1. NMR Data for Compounds 1 and 2
agarsenolides
agarsenone (1)
major (2a)
minor (2b)
b
b
c
c
c
c
position
δC type
δH (J in Hz)
δC type
δH (J in Hz)
δC type
δH (J in Hz)
1
131.4, C
195.5, C
46.0, CH2
144.9, C
144.7, C
198.5, C
d
e
2
198.5, C
3α
3β
4
2.48 d (12.9)
1.80−1.95, m
1.80−1.95, m
2.24, d (13.1)
1.80−1.95, m
48.9, CH2
2.69−2.77, m
2.27−2.38, m
2.69−2.77, m
2.85, dd (3.5, 16.0)
48.5, CH2
2.69−2.77, m
2.27−2.38, m
2.69−2.77, m
3.08, dd (3.1, 15.6)
30.3, CH
35.5, CH2
29.9, CH
36.4, CH2
29.2, CH
35.1, CH2
5α
5β
6
2.60, dd (10.6, 16.0)
2.52, dd (10.5, 15.6)
146.3, C
117.8, C
156.5, C
29.2, CH2
142.2, C
124.0, C
155.6, C
113.0, CH
141.6, C
124.1, C
155.9, C
112.9, CH
7
8
c
c
9α
9β
10
11
12
13
14
15
2.64, dd (8.1, 17.1)
2.34, dd (1.0, 17.1)
3.51, m (7.2)
6.89, s
6.91, s
26.9, CH
118.4, C
127.5, C
38.1, CH
127.7, C
38.0, CH
3.74, q (7.5)
3.71, q (7.6)
d
e
140.6, CH
10.2, CH3
21.3, CH3
18.7, CH3
6.81, brs
177.7, C
177.7, C
1.74, d (1.2)
0.69, d (5.6)
1.01, d (7.0)
16.1, CH3
21.7, CH3
24.3, CH3
1.61, d (7.5)
1.16, d (6.4)
2.68, s
16.0, CH3
20.8, CH3
24.2, CH3
1.64, d (7.6)
1.11, d (6.2)
2.67, s
a
Data were recorded on a Bruker Avance DPX-400 spectrometer. Chemical shifts: δ values are given in ppm with reference to the signal of CDCl3 (δ
b
7.26 ppm) or benzene-d6 (δ 7.16 ppm) and to the center peak of the signal of CDCl3 (δ 77.1 ppm) and of benzene-d6 (δ 128.0 ppm) for 13C. In
benzene-d6. In CDCl3. Overlapped signals.
c
d,e
carbons, one sp2 methine, two sp3 methines, three sp3
methylenes, and three methyls. COSY data allowed the
determination of the spin system sequence from H-3 to H-5
and H-14 and from H-9 to H-10 and H-15 and showed a long-
range coupling between H-13 and H-15. Further support for
this assignment was provided by HMBC correlations observed
between H-3α/C-2, H-14/C-3, H-14/C-5, H-9α/C-7, H-9β/C-
1, H-9β/C-8, and H-15/C-9 (Figure 1).
doublets centered at δ 2.75 and 2.52, respectively. H-3α showed
a geminal coupling constant (J = 17.5 Hz) and an axial−
equatorial coupling constant with H-4 (J = 4.5 Hz), while H-5α
showed a geminal coupling constant (J = 15.8 Hz) and an
axial−equatorial coupling constant with H-4 (J = 2.1 Hz). On
the other hand, H-3β (δ 2.38) resonated as a doublet of
doublets with a geminal constant (J = 17.5 Hz) and a trans-
diaxial coupling constant with H-4 (J = 10.0 Hz). These values
are compatible with an equatorial disposition of the 14-methyl
group.
No information about the relative configuration of C-4 and
C-10 could be obtained by NOESY experiments (Figure 2).
Therefore, a computational analysis of chiroptical properties,
i.e., ORD and ECD, of 1 and its degradation product 3 was
undertaken in order to ascertain their relative and absolute
configurations. Although 3 is a known compound,9 its absolute
configuration remains undefined.
Figure 1. Key COSY (thick bonds) and HMBC (arrows) correlations
for 1 and 2.
The conversion of 1 to 3 in chlorinated solvents provided
definitive proof of the furanocadinane skeleton. In 1 the
coupling constants between H-9 and H-10 (J9β,10 < 1 Hz; J9α,10
= 8.1 Hz) indicated a dihedral angle of ca. 90° between H-9β
and H-10 and close to 40° between H-9α and H-10, due to a
quasi-axial orientation of the 15-methyl. H-3α and H-5α
1
appeared in the H NMR spectrum (CDCl3) as a doublet of
Figure 2. Key NOESY correlations for agarsenone (1).
1255
dx.doi.org/10.1021/np400114b | J. Nat. Prod. 2013, 76, 1254−1259

Journal of Natural Products
Article
First, the absolute configuration at C-4 of compound (−)-3
was assigned by comparison between experimental chiroptical
data and TDDFT calculations for (4R)-3. Conformational
analysis for (4R)-3 was initially carried out by molecular
mechanics calculations to find two conformers having an
equatorial and axial C-4 methyl group, respectively. The
conformers were fully optimized in the gas phase at the DFT/
B3LYP/TZVP level of theory, obtaining relative free energies
and Boltzmann population percentages (Figure 3). The
and (4R,10R) diastereoisomers. To assign the C-10 config-
uration, the ECD spectra of (4R,10S)-1 and (4R,10R)-1 were
computed and compared with experimental data. Conforma-
tional analysis for (4R,10S)-1 and (4R,10R)-1 provided only
two conformers for each diastereoisomer. For (4R,10S)-1 a syn
axial−equatorial conformer (96.9%) and a syn diaxial conformer
(3.1%) were found (Figure 5), while for (4R,10R)-1
Figure 3. Computed (DFT/B3LYP/TZVP) most stable conformers
of (R)-3. In parentheses ΔG (kcal mol−1)/population in gas phase.
Figure 5. Computed (DFT/B3LYP/TZVP) most stable conformers
of (4R,10S)-1. In parentheses ΔG (kcal mol−1)/population in gas
phase; in square brackets values obtained in the Et2O IEFPCM
solvation model.
equatorial conformer was found to be predominant (92.3%).
The optical rotation at 589 nm for (R)-3 was then calculated at
the TDDFT/B3LYP/aug-cc-pVDZ level19 using the IEFPCM20
implicit solvation model for CH2Cl2, affording a value of −15.9,
in good agreement with the experimental value of [α]D −19.9
(c 0.2, CH2Cl2), thus supporting the assignment of the (4R)-
configuration. Further evidence for this assignment was
provided by ECD analysis. The ECD spectrum of (R)-3 was
simulated at the TDDFT/CAM-B3LYP/aug-cc-pVDZ level in
the gas phase and compared with the experimental data for
(−)-3.21 The theoretical ECD spectrum for (R)-3 was in good
agreement with the experimental data for (−)-3 (Figure 4).
calculations provided an anti axial−equatorial (91.5%) and an
anti diaxial (8.5%) conformer (Figure 6). The ECD spectrum
Figure 6. Computed (DFT/B3LYP/TZVP) most stable conformers
of (4R,10R)-1. In parentheses ΔG (kcal mol−1)/population in gas
phase; in square brackets values obtained in the Et2O IEFPCM
solvation model.
of (−)-1 recorded in Et2O under argon, to prevent its
decomposition, showed three main features (Figure 7 and
Table 2): a negative CE at 225 nm, a positive CE at 288 nm,
and a broad negative band in the 309−400 nm region. The
ECD spectra of (4R,10S)-1 and (4R,10R)-1 were simulated at
the TDDFT/CAM-B3LYP/aug-cc-pVDZ level using the
IEFPCM20 implicit solvation model for Et2O. The computed
spectrum of (4R,10S)-1 displayed CEs with the same sequence
(negative/positive/negative), shape, and relative intensity as
the experimental spectrum of (−)-1, while the ECD spectrum
calculated for (4R,10R)-1 was significantly different from the
experimental data. These data supported the assignment of the
(4R,10S) absolute configuration to (−)-1.
Such an assignment is not definitive because experimental
and calculated spectra did not fit perfectly. Calculations carried
out with a different long-range-corrected functional such as lc-
ωpbe22 (see Figure S21 in the Supporting Information) did not
lead to significant improvement of the spectra agreement (see
Supporting Information). Therefore, to further support the
above assignment, we attempted to computationally reproduce
the ORD curve of (−)-1. In fact, it has been shown that in the
case of complex molecules the use of more than a single
chiroptical technique can provide more reliable results.23 The
Figure 4. Experimental ECD and UV spectra of (−)-3 in CH3CN
(solid red line) and calculated (TDDFT/CAM-B3LYP/aug-cc-pVDZ/
gas phase) ECD and UV spectra of (R)-3 (dashed blue line).
Calculated spectra are red-shifted by 15 nm.
The sequence of experimental Cotton effects (CEs) observed,
strongly negative at 190 nm and negative at 335 nm, was
satisfactorily reproduced by the calculations in sign, position,
and relative intensity. Such analysis unambiguously allows the
assignment of the (4R) absolute configuration for (−)-3. As the
transformation of (−)-1 to (−)-(R)-3 does not involve a
change of C-4 configuration, the same (R) absolute
configuration at this carbon was assumed for (−)-1. Therefore,
assignment of absolute configuration to (−)-1 required
comparison with chiroptical data computed only for (4R,10S)
1256
dx.doi.org/10.1021/np400114b | J. Nat. Prod. 2013, 76, 1254−1259

Journal of Natural Products
Article
times higher than experimental ones (Table 3). This result
further confirmed the assignment of a (4R,10S) absolute
configuration for (−)-1.
Table 3. Comparison of Experimental and Calculated
Optical Rotations for 1
calculated optical
a
b
experimental optical rotations
rotations
wavelength (nm)
(−)-1
(4R,10S)-1 (4R,10R)-1
589
546
435
405
−40.4
−52.0
−227.6
−453.2
−36.3
−50.4
−198.6
−394.0
−260.7
−341.6
−980.9
−1616.6
b
a
Optical rotations recorded in Et2O (c 0.25 g/100 mL). Optical
rotations obtained as Boltzmann averages calculated at the TDDFT/
B3LYP/aug-cc-pVDZ level employing the Et2O implicit IEFPCM
solvation model.
Figure 7. Experimental ECD spectrum of (−)-1 in Et2O (solid red
line) and calculated (TDDFT/CAM-B3LYP/aug-cc-pVDZ/IEFPCM
Et2O) ECD spectra of (4R,10S)-1 (dashed blue line) and (4R,10R)-1
(dotted green line). Calculated spectra are red-shifted by 25 nm.
Agarsenolides A and B (2a and 2b, respectively) were
obtained as an inseparable mixture of isomers (ratio 2.4:1) by
autoxidation of agarsenone (1).24 All attempts to separate the
mixture failed, leading to the decomposition of the compounds;
therefore the mixture was subjected to structure elucidation.
The molecular formula of both isomers was established to be
C15H16O3 by HRMS, indicating eight indices of hydrogen
deficiency. The 13C NMR spectrum of the mixture (Table 1),
supported by a JMODXH experiment, showed the presence of
three methyls, two aliphatic methylenes, and two aliphatic
methines. In the sp2 region, we observed one methine, five
quaternary carbons, and two carbonyls. The 13C NMR
spectroscopic data of the dihydronaphthalenone core were
similar to those of 3.10
Table 2. Main UV and ECD Features of (−)-3 and (−)-1
compd
UV λ (nm) [ε × 10−3
]
ECD λ (nm) [Δε]
a
1
232[70.4], 238[68.5],
284[49.6]
225[−2.2], 276sh[+2.5], 288[+3.1],
301sh[+1.5], 318[−0.8]
b
3
196[29.0], 240[60.0],
260sh[25.0], 310[7.5]
190[−13.3], 222[+3.4], 242[−1.5],
269[+1.9] 335[−2.1]
a
b
Spectra recorded in Et2O (c ∼1 × 10−3 M). Spectra recorded in
MeCN (c ∼1 × 10−3 M).
optical rotation of (−)-1 was recorded at several wavelengths in
Et2O, obtaining the ORD curve in Figure 8, which was
The significant differences were the replacement of the
olefinic carbon at C-11 with an aliphatic methine resonating at
δ 38.1/38.0 (δH 3.74/3.71, q, J = 7.5/7.6 Hz) and the presence
of a second carbonyl at δ 177.7. The presence of the β-methyl
lactone was evidenced by the resonances at δ 177.7 (C-12),
124.0/124.1 (C-7), 155.6/155.9 (C-8), 38.1/38.0 (C-11), and
16.1/16.0 (C-13)25 and by the absorbance at 1807 cm−1 in the
IR spectrum. The peaks at δ 113.0/112.9 (C-9), 127.5/127.7
(C-10), and 24.3/24.2 (C-15) were indicative of the presence
of the methyl-substituted aromatic ring. The connections from
C-3 to C-5 and between C-11 and C-13 were confirmed by
COSY data. All these features were characteristic of the
presence of a 2,3-dihydrobenzo[b]furan-2-one system. HMBC
correlations permitted establishment of the structure of 2 as
depicted in Figure 1. Particularly diagnostic were correlations
between H-3/C2, H-14/C4 and C5, H-5/C4 and C6, H-9/C8,
C10, and C15, H-11/C7, C12, and C13, and H-15/C1, C9, and
C10. The relative configurations of 2a and 2b were determined
by analysis of NOESY data and comparison with 1’s NMR data
(Figure 9). The absolute configuration at C-4 was assigned as
R, the same as agarsenone 1, from which these compounds
arise by decomposition. From all these observations we
assigned a β-orientation of the C-11 methyl group in 2a and
α in 2b.
Figure 8. Experimental ORD curve of (−)-1 in Et2O (solid red line,
◆
) and calculated (TDDFT/B3LYP/aug-cc-pVDZ/IEFPCM Et2O)
▲
ORD curves for (4R,10S)-1 (dotted green line, ) and (4R,10R)-1
●
(dashed blue line, ).
compared with ORD curves calculated for (4R,10S)-1 and
(4R,10R)-1 diastereoisomers. ORD calculations were per-
formed at the TDDFT/B3LYP/aug-cc-pVDZ level on geo-
metries obtained at the DFT/B3LYP/TZVP level, performing
both calculations in an Et2O implicit IEFPCM solvation
model.20 As shown in Figure 8 the calculated ORD curve for
(4R,10S)-1 fits very well with experimental data, while the
ORD calculation for (4R,10R)-1 provided values about five
CONCLUSION
■
We reported the isolation and structure identification of
agarsenone (1), a new furanosesquiterpenoid that can easily
decompose during extraction and purification to give myrrhone
(3) and, by autoxidation, agarsenolides (2a and 2b). The
1257
dx.doi.org/10.1021/np400114b | J. Nat. Prod. 2013, 76, 1254−1259

Journal of Natural Products
Article
Figure 9. Key NOESY correlations for agarsenolides (2a and 2b).
Composition and amounts of the fractions are reported in Table 1S
(Supporting Informations).
structures of 1−3 were elucidated by NMR methods, and their
relative and absolute configurations were determined by
TDDFT computational simulation of ORD and ECD spectra.
This study provides the first assignment of absolute
configuration for bioactive compounds from myrrh. Moreover,
our observations may suggest that, in certain cases, myrrhone, a
well-known compound found in Commiphora spp. myrrh
extracts,26 could be an experimental artifact and not a genuine
plant metabolite.
Fr22 (65 mg) was further purified by preparative TLC (elution with
n-hexane−Et2O 25%), affording agarsenone (1) (34 mg) and rel-
1S*,2S*-epoxy-4R-furanogermacr-10(15)-en-6-one (10) (15 mg).
Agarsenone (1): colorless oil; [α]22 −78.3 (c 0.64 in TBME);
D
[α]26 −30.6 (c 1.10, benzene); FT-IR νmax 2956, 1762, 1652, 1549
D
cm−1; 1H NMR (CDCl3, 400 MHz) δ 0.91 (3H, d, J = 7.0 Hz, H-14),
1.11 (3H, d, J = 6.3 Hz, H-15), 2.15 (3H, brs, H-13), 2.20−2.30 (3H,
m, H4, H-5β), 2.38 (1H, dd, J = 10.0, 17.5 Hz, H-3β), 2.52 (1H, dd, J
= 2.1, 15.8 Hz, H-5α), 2.54 (1H, d, J = 17.0 Hz, H-9α), 2.75 (1H, dd, J
= 4.5, 17.5 Hz, H-3α), 2.86 (1H, dd, J = 7.7, 17.0 Hz, H-9β), 3.30 (1H,
m, J = 7.7 Hz, H-10), 7.11 (1H, brs, H-12); 1H NMR and 13C NMR in
benzene-d6 see Table 1; C15H18O2 HREIMS m/z 231.1367 [M + H]+,
calcd 213.1385.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured with a JASCO DIP-1000 digital polarimeter. Absorption
and ECD spectra were recorded at room temperature on a JASCO
J815 spectropolarimeter, using 0.5 mm cells and concentrations of
about 1 × 10−3 M in MeCN for (−)-3 and in Et2O for (−)-1. IR
spectra were recorded on a JASCO 410 spectrophotometer equipped
with a diffuse reflectance accessory. High-resolution MS spectra were
acquired on a Q-TOF Premier instrument (Waters, Milford, MA,
USA), equipped with a nanospray ion source and provided with a lock-
mass apparatus to perform a real-time calibration correction. GC
analyses (Table S2) were performed on a Hewlett-Packard HP 6890
gas chromatograph equipped with a Hewlett-Packard MS 5973 mass
selective detector and a fused silica capillary column (HP-5MS; 30 m
× 0.25 mm i.d., 0.25 μm film thickness). NMR spectra were recorded
using a Bruker Avance DPX-400 spectrometer operating at frequencies
of 400 MHz (1H) and 100 MHz (13C) and a Bruker DRX-600
spectrometer operating at 600 MHz (1H) and 150.85 MHz (13C). The
Agarsenolides (2a and 2b): yellowish oil; FT-IR νmax 2953, 1807,
1676, 1587, 1451 cm−1; 1H and 13C NMR see Table 1; HREIMS m/z
245.1158 [M + H]+, calcd 245.1178.
Myrrhone (3): colorless oil; [α]24 −19.9 (c 0.2, CH2Cl2); FT-IR,
D
1H and 13C NMR were in agreement with reported data.9,10
Computational Details. Preliminary conformational analyses
were performed by the Spartan02 package27 employing MMFF94s
molecular mechanics force field with Monte Carlo searching and fixing
arbitrarily assigned absolute configurations (R) for 3 and (4R,10S) and
(4R,10R) for 1. All possible conformers were searched, considering the
degrees of freedom of the system, retaining only the structures within
an energy range of 30 kcal/mol with respect to the most stable one.
The minimum energy conformers found by molecular mechanics were
further fully optimized by using the DFT at the DFT/B3LYP/TZVP
level in either the gas phase or solvent with the IEFPCM implicit
solvation model,20 as implemented in the Gaussian09 package.28 For
each compound two conformers were found (Figures 3, 5, and 6). All
conformers are real minima, no imaginary vibrational frequencies have
been found, and the free energy values have been calculated and used
to get the Boltzmann population of conformers at 298.15 K. Optical
rotations were computed by using the TDDFT/B3LYP/aug-cc-pVDZ
level on the DFT-optimized geometries. Solvent effects were modeled
with the IEFPCM implicit solvation model20 with parameters of Et2O
for 1 and CH2Cl2 for 3. The DFT/B3LYP/TZVP geometries were
also employed as input geometries for calculation of UV and ECD
spectra at the TDDFT/CAM-B3LYP/aug-cc-pVDZ level. TDDFT
calculations employing the long-range corrected CAM-B3LYP func-
tional29 provided a good reproduction of Cotton effects observed in
the theoretical ECD spectra. The theoretical ORD, UV, and ECD
spectra were obtained as averages over the conformers’ Boltzmann
populations. The ECD spectra were obtained from calculated
excitation energies and rotational strengths, as a sum of Gaussian
functions centered at the wavelength of each transition, with a
parameter σ (width of the band at half-height) of 0.5 eV for 1 and 0.3
eV for 3. To guarantee origin independence and to evaluate the quality
of the molecular wave functions employed,30 theoretical ECD spectra
were obtained in both the length and velocity representation, using the
lowest 30 states. The velocity/length calculated spectra were almost
coincident, indicating a good level of calculation. Therefore, in all
figures only the velocity-form predicted spectra are reported. UV and
1
spectra were measured in CDCl3 and C6D6. The H and 13C NMR
chemical shifts (δ) were expressed in ppm with reference to the
solvent signals [CDCl3 δH 7.26 and δC 77.1, C6D6 δH 7.16 and δC
128.0]. Column chromatography was performed using Davisil LC60A
60−200 μm silica gel. Fractions were monitored by TLC (silica gel 60
F254, Merck), and spots on TLC were visualized under UV light and
after staining with p-anisaldehyde−H2SO4−EtOH (1:1:98) followed
by heating at 110 °C. For preparative TLC silica gel 60 F254 plates
(Merck) were used, and the compounds visualized by UV.
Plant Material. The resin of Commiphora erythraea (Agarsu grade
I) commercialized by Agarsu Liben Cooperative and imported into
Italy by IPO (Increasing People Opportunities) Association (www.
ipoassociazione.org) was studied. A voucher specimen (#MCM-1) of
the resin (Agarsu grade I) is deposited at the Dipartimento di Chimica
e Tecnologia del Farmaco-Sez. Chimica Organica, University of
Perugia.
Extraction and Isolation. The ground resin (15 g) was extracted
by maceration in MeOH (3 × 250 mL) at room temperature for 24 h.
After filtration, the organic solutions were concentrated at 40 °C to
give the MeOH extract (7.52 g). The extract was divided into two
fractions by column chromatography (SiO2) using n-hexane for the
elution of fraction M1 (0.42 g) and Et2O for the elution of fraction M2
(6.19 g). An aliquot of the M2 fraction (3.5 g) was purified on a SiO2
gel chromatography column using a gradient of n-hexane−Et2O (0−
100%), collecting 7 mL fractions that were evaluated by TLC and
combined upon similar appearance, yielding 31 fractions (Fr1−Fr31).
1258
dx.doi.org/10.1021/np400114b | J. Nat. Prod. 2013, 76, 1254−1259

Journal of Natural Products
Article
ECD calculations provided electronic transitions at slightly higher
energy than experimental; therefore, for a clearer comparison with
experimental data in Figures 4 and 7, calculated UV and ECD spectra
of 3 and 1 have been red-shifted by 15 and 25 nm, respectively.
(13) Cenci, E.; Messina, F.; Rossi, E.; Epifano, F.; Marcotullio, M. C.
Nat. Prod. Commun. 2012, 7, 143−144.
(14) Brieskorn, C. H.; Noble, P. Phytochemistry 1983, 22, 187−189.
(15) Zhu, N.; Kikuzaki, H.; Sheng, S.; Sang, S.; Rafi, M. M.; Wang,
M.; Nakatani, N.; DiPaola, R. S.; Rosen, R. T.; Ho, C.-T. J. Nat. Prod.
2001, 64, 1460−1462.
(16) For general discussion of the ab initio calculation of chiroptical
properties: (a) Autschbach, J.; Nitsch-Velasquez, L.; Rudolph, M. Top.
Curr. Chem. 2011, 298, 1−98. (b) Stephens, P. J. Computational
Medicinal Chemistry for Drug Discovery; Bultnick, P., de Winter, H.,
Langenaecker, W., Tollenaere, J., Eds.; Dekker: New York, NY, 2003;
pp 699−725. (c) Autschbach, J. Chirality 2009, 21, E116−E152.
(17) Autoxidation of 1 could occur through a hetero-Diels-Alder
reaction between oxygen and a furan, followed by hydrolysis and acid
(from chlorinated solvent decomposition)-catalyzed dehydration into
the stable benzofuran. See for example: Chen, K.-S.; Wu, Y.-C.
Tetrahedron 1999, 55, 1353−1366.
(18) Bigler, P. NMR Spectroscopy; Wiley: Weinheim, 2000; p 57.
(19) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (b) Lee,
C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789. (c) Stephens,
P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem.
1994, 98, 11623−11627.
(20) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999−3094.
(21) (a) For synthetic compounds: Xu, D.-Q.; Wang, Y.-F.; Luo, S.-
P.; Zhang, S.; Zhong, A.-G.; Chen, H.; Xu, Z.-Y. Adv. Synth. Catal.
2008, 350, 2610−2616. (b) For natural products: Evidente, A.;
Superchi, S.; Cimmino, A.; Mazzeo, G.; Mugnai, L.; Rubiales, D.;
ASSOCIATED CONTENT
■
S
* Supporting Information
Purification details, NMR spectra of the described compounds,
conformational analysis, and ECD data of 1 and 3 are available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*(M.C.M.) Tel: +39-075-585-5104. Fax: +39-075-585-5116. E-
mail: marcotu@unipg.it or mariacarla.marcotullio@unipg.it.
(S.S.) Tel: +39-0971-20-6098. Fax: +39-0971-20-5678. E-
mail: stefano.superchi@unibas.it.
Present Address
§Department of Organic Chemistry, Arrhenius Laboratory,
Stockholm University, SE-106 91 Stockholm, Sweden.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Andolfi, A.; Villegas-Fernan
5570.
́
dez, A. M. Eur. J. Org. Chem. 2011, 5564−
The authors wish to thank Fondazione Cassa di Risparmio
(Project Code: 2010.011.0410), Universita
(R.I.L. 2010 grant), and MIUR (Rome) (PRIN project
2008LYSESR) for financial support, Ipoassociazione, Agarsu
Liben Cooperative for providing the resin, and Infarmazone.
̀
della Basilicata
(22) Vydrov, O. A.; Scuseria, G. E. J. Chem. Phys. 2006, 125,
234109−234117.
(23) See for example: Mazzeo, G.; Santoro, E.; Andolfi, A.; Cimmino,
A.; Troselj, P.; Petrovic, A. G.; Superchi, S.; Evidente, A.; Berova, N. J.
Nat. Prod. 2013, 76, 588−599.
̀
We thank Prof. L. Rastrelli (Universita di Salerno) for the use
of HRMS and Bruker 600 MHz facilities. A. Palmier OLJ
BPharm (Hons) Melit. (University of Malta) and M. Kerrigan
MA (Cantab) are gratefully acknowledged for valuable
linguistic suggestions.
(24) Ulubelen, A.; Oksuz, S. J. Nat. Prod. 1984, 47, 177−178.
(25) Harrowven, D. C.; Lucasa, M. C.; Howes, P. D. Tetrahedron
2001, 57, 791−804.
(26) (a) Shen, T.; Wan, W.; Yuan, H.; Kong, F.; Guo, H.; Fan, P.;
Lou, H. Phytochemistry 2007, 68, 1331−1337. (b) Shen, T.; Wan, W.-
Z.; Wang, X.-N.; Yuan, H.-Q.; Ji, M.; Lou, H.-X. Helv. Chim. Acta
2009, 92, 645−652.
REFERENCES
■
(1) Zorloni, A. MD Dissertation, University of Pretoria, Pretoria,
South Africa, 2007.
(27) SPARTAN ’02; Wavefunction Inc.: Irvine, CA; http://www.
wavefunction.com.
(28) Frisch, M. J.; et al. Gaussian 09, Revision A.02; Gaussian, Inc.:
Wallingford, CT, 2009.
(2) Maradufu, A. Phytochemistry 1982, 21, 677−680.
(3) Carroll, J. F.; Maradafu, A.; Warthen, J. D. Entomol. Exp. Appl.
1989, 53, 111−116.
(4) Marcotullio, M. C.; Santi, C.; Oball-Mond Mwankie, G. N.;
Curini, M. Nat. Prod. Commun. 2009, 4, 1751−1754.
(5) Fraternale, D.; Sosa, S.; Ricci, D.; Genovese, S.; Messina, F.;
Tomasini, S.; Montanari, F.; Marcotullio, M. C. Fitoterapia 2011, 82,
654−661.
(29) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393,
51−57.
(30) Moscowitz, A. In Modern Quantum Chemistry; Sinanoglu, O.,
Ed.; Academic Press: London, 1965; Part III, p 31.
(6) GC-MS analysis of these oils and extracts revealed the presence
of a compound with retention index (RI) of 1894 on a HP-5MS
column (ref 4). Every attempt to isolate it failed. After chromato-
graphic purification using methylene chloride it could not be detected
anymore. Conversely, myrrhone (3) was either not present or present
only in traces in the starting oil/extract (refs 4 and 5).
(7) Dekebo, A.; Dagne, E.; Sterner, O. Fitoterapia 2002, 73, 48−55.
(8) Hikino, H.; Konno, C.; Takemoto, T.; Tori, K.; Ohtsuru, M.;
Horibe, I. Chem. Commun. 1969, 662−663.
(9) Barreira, E. S.; Quieroz Monte, F. J.; Braz-Filho, R. Nat. Prod.
Lett. 1996, 8, 285−289.
(10) Zhu, N.; Sheng, S.; Rosen, R. T.; Ho, C.-T. Flav. Frag. J. 2003,
18, 282−285.
(11) Konig, W. A.; Bulow, N.; Fricke, C.; Melching, S.; Rieck, A.;
̈
̈
Muhle, H. Phytochemistry 1996, 4, 629−633.
(12) Marcotullio, M. C.; Messina, F.; Curini, M.; Macchiarulo, A.;
Cellanetti, M.; Ricci, D.; Giamperi, L.; Bucchini, A.; Minelli, A.; Mierla,
A. L.; Bellezza, I. Molecules 2011, 16, 10357−10369.
1259
dx.doi.org/10.1021/np400114b | J. Nat. Prod. 2013, 76, 1254−1259