NMR and x-ray conformational study of artemisiifolin and three other related germacranolides

Article abstract of DOI:10.1002/mrc.1352

From an acetone extract of Staehelina dubia, large quantities of the previously known germacranolide artemisiifolin were isolated. The conformations of the 10-membered germacra-1(10)E,4Z-diene ring system of this compound and those of its derivatives 11,13-dihydroartemisiifolin, isabelin and 6α-hydroxy-15-oxogermacra-1(10)E,4Z,11(13)-trien-12,8α-olide were studied by NMR spectroscopic methods. Low-energy conformations were obtained by quantum mechanical calculations. An x-ray diffraction analysis of artemisiifolin established that, in the crystalline state, it possesses a unique conformation that corresponds to the majority one existing in acetone-d₆ solution. Copyright

Full text of DOI:10.1002/mrc.1352

MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2004; 42: 474–483
Published online 18 February 2004 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1352
NMR and x-ray conformational study of artemisiifolin
and three other related germacranolides
1∗
2
2
´
´
´
Marıa L. Jimeno, Marıa del Carmen Apreda-Rojas, Felix H. Cano and
3
´
´
Benjamın Rodrıguez
1
´
´
Centro de Quımica Organica ‘Manuel Lora-Tamayo’, CSIC, Juan de la Cierva 3, E-28006 Madrid, Spain
2
3
´
Departamento de Cristalografıa, Instituto ‘Rocasolano’, CSIC, Serrano 119, E-28006 Madrid, Spain
Instituto de Quımica Organica, CSIC, Juan de la Cierva 3, E-28006 Madrid, Spain
´
´
Received 29 July 2003; Revised 5 November 2003; Accepted 12 November 2003
From an acetone extract of Staehelina dubia, large quantities of the previously known germacranolide
artemisiifolin were isolated. The conformations of the 10-membered germacra-1(10)E,4Z-diene ring system
of this compound and those of its derivatives 11,13-dihydroartemisiifolin, isabelin and 6a-hydroxy-15-
oxogermacra-1(10)E,4Z,11(13)-trien-12,8a-olide were studied by NMR spectroscopic methods. Low-energy
conformations were obtained by quantum mechanical calculations. An x-ray diffraction analysis of
artemisiifolin established that, in the crystalline state, it possesses a unique conformation that corresponds
to the majority one existing in acetone-d6 solution. Copyright  2004 John Wiley & Sons, Ltd.
KEYWORDS: NMR; 1H NMR; 13C NMR; 2D NMR; variable temperature; conformational analysis; molecular modeling; x-ray
crystallography; Staehelina dubia; sesquiterpenoids; germacranolides; artemisiifolin
1
9
INTRODUCTION
10
5
8
7
O
O
2
3
14
O
12
O
11
Several naturally occurring sesquiterpene compounds pos-
sessing a 10-membered germacra-1(10),4-diene ring system
have been isolated from the aerial parts of Staehelina dubia L.
(Compositae). These germacranolides show broadened NMR
spectra or even multiple NMR signals indicative of their con-
formational mobility that causes the existence of two or
more conformers in solution at room temperature.1–6 In this
paper we describe the isolation, structure corroboration and
6
4
13 HB
15R
OH
OH
HO
O
HA
1 R = CH2OH
4 R = CHO
2
O
complete conformational analysis by means of H and 13C
1
HB
O
HA
O
NMR spectroscopy and quantum mechanical calculations
of the sesquiterpene germacranolides artemisiifolin (1), its
11,13-dihydro derivative (2), isabelin (3) and 6˛-hydroxy-15-
oxogermacra-1(10)E,4Z,11(13)-trien-12,8˛-olide (4), and also
the x-ray diffraction analysis of 1.
Although studies on the conformations of isabelin in
CDCl3 solution have already been reported,2,3 we have now
obtained additional data on this compound that complete
previous results.
3
of large quantities (1.8% on dry plant material) of the
germacrane-type sesquiterpenoid artemisiifolin (1), found
for the first time in Ambrosia artemisiifolia L.1 Stereoselec-
tive reduction of 1 with sodium borohydride quantitatively
yielded 11,13-dihydroartemisiifolin (2), previously obtained
from 1 under milder hydrogenation conditions.1 Oxida-
tion of 1 with pyridinium dichromate gave a mixture of
two compounds, which were easily separated by column
chromatography and characterized as isabelin (3) (a ger-
macratriene dilactone previously isolated from Ambrosia
psilostachya DC, whose two conformational isomers in CDCl3
solution have already been studied2,3) and 6˛-hydroxy-15-
oxogarmacra-1(10)E,4Z,11(13)-trien-12,8˛-olide (4).
RESULTS AND DISCUSSION
Column chromatography of an acetone extract of the aerial
parts of Staehelina dubia L. (Compositae) allowed the isolation
Ł
´
´
Correspondence to: Marıa L. Jimeno, Centro de Quımica Orga´nica
‘Manuel Lora-Tamayo’, CSIC, Juan de la Cierva 3, E-28006 Madrid,
Spain. E-mail: iqolj13@iqog.csic.es
Structural corroboration
Contract/grant sponsor: Direccio´n General de Investigacio´n
´
The 1H and C NMR spectra of 1 at 25 C were obtained
for CDCl3 and Me2CO-d6 solutions. The spectra showed
broad signals, indicating the presence, in each case, of four
Cientıfica y Te´cnica (DGICYT); Contract/grant number: BQU
13
°
2000-0868.
Contract/grant sponsor: Comisio´n Interministerial de Ciencia y
´
Tecnologıa (CICYT); Contract/grant number: AGL 2001-1652.
Copyright  2004 John Wiley & Sons, Ltd.

NMR and x-ray conformational study of germacranolides
475
14
5
10
3
2
9
1
8
4
6
12
7
15
11
13
1A (UD)
1
9
14
10
2
10
4
8
7
12
15
3
5
6
11
2
4
9
14
15
12
8
7
11
3
1
6
5
13
13
1B (UU)
1D (DD)
8
7
12
15
1
9
11
2
10
6
4
13
3
5
14
1C (DU)
Figure 1. AM1 conformers of artemisiifolin (1). Relevant experimental NOEs are indicated by arrows.
from the 13C signals at υ 171.35, 127.44 and 135.13 and from
the two characteristic doublets at υ 6.31 and 6.06 in the H
slowly interconverting conformers (1A, 1B, 1C and 1D; Fig. 1)
with different conformer ratios depending on the solvent
(60 : 15 : 14 : 11 for CDCl3 and 29 : 30 : 27 : 14 for Me2CO-d6).
Confirmation of the structure of 1 was carried out with a
1
spectrum. The gHMBC cross peak between H-8ˇ and the
lactone carbon at υ 171.35 confirmed the position of the
lactone ring. The existence of two olefinic double bonds was
confirmed by the two vinyl proton multiplets at υ 5.00 and
5.28. The gHMBC spectrum showed cross peaks between
H-1 and C-10 (υ 129.60), C-2 (υ 24.91), C-3 (υ 36.09), C-9
(υ 40.98) and C-14 (υ 21.09), and between H-5 and C-3 (υ
36.09), C-4 (υ 140.39), C-6 (υ 69.21) and C-15 (υ 61.46). The
absence of NOE effects between the Me-14 group and H-
1, and between the hydroxymethylene protons H-15a and
H-15b and H-5, confirmed a trans relationship for these
protons and consequently a 1(10)E,4Z-stereochemistry in the
10-membered ring. These results confirmed that compound
CDCl3 solution, owing to the unequal conformer ratio, which
1
facilitated the peak assignments. At ꢀ30 C, the H and 13C
°
spectra showed sharp resonances and the assignments for the
major conformer were carried out at this temperature using
a combination of DQCOSY, TOCSY, gHSQC and gHMBC 2D
NMR sequences. The 1H and 13C chemical shifts are collected
in Tables 1 and 3.
The 13C NMR spectrum of 1 showed 15 signals assignable
to a sesquiterpene moiety. The edited gHSQC spectrum
showed the presence of four quaternary carbons (υ 171.35,
140.39, 135.13 and 129.60), one methyl (υ 21.09), five
methylene (υ 127.44, 61.46, 40.98, 36.09 and 24.91) and five
methine carbons (υ 128.89, 128.89, 81.06, 69.21 and 50.65). The
presence of an ˛-methylene-ꢀ-lactone moiety was evident
1 was artemisiifolin.1
13
The 1H and C NMR spectra of 2 in DMSO at 25 C
°
showed broad signals, corresponding to a mixture of two
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

476
M. L. Jimeno et al.
1
°
Table 1. H NMR data (υ, ppm; J, Hz) for conformers 1A–1D (500 MHz, at ꢀ30 C)
a
CDCl3
Me2CO-d6
Parameter
1A
1A
1B
1C
1D
υ H-1
5.00
4.7
11.0
2.16
2.25
5.16
6.6
9.8
5.00
3.7
12.7
5.13
10.7
4.3
2.38
2.06
5.22
8.1
8.1
J(H-1,H-2˛)
J(H-1,H-2ˇ)
υ H-2˛
υ H-2ˇ
υ H-3
υ H-5
J(H-5,H-6ˇ)
υ H-6ˇ
J(H-6ˇ,H-7˛)
J(H-6ˇ,OH-6)
υ OH-6
υ H-7˛
J(H-7˛,H-8ˇ)
υ H-8ˇ
J(H-8ˇ,H-9˛)
J(H-8ˇ,H-9ˇ)
υ H-9˛
υ H-9ˇ
υ H-13
υ Me-14
υ H-15a
2.18
2.41
2.49b, 2.14b
5.39
5.6
2.07
2.33
2.58b, 1.86b
4.88
10.0
4.52
9.3
2.41b
2.15b
2.35b, 1.97b
5.19
5.1
2.24b, 1.93b
5.28
2.37b, 1.90b
4.86
10.0
4.40
10.0
4.6
4.59
2.74
5.2
3.99
11.5
4.2
1.93
2.74
6.14, 6.11
1.68
4.39
4.08
5.1
5.8
3.97
4.03
10.2
6.6
5.75
2.88
3.4
4.59
11.6
5.8
1.77
2.86
6.02, 6.15
1.66
3.88
4.12
5.9
4.12
7.5
5.8
5.64
2.98
4.6
10.2
c
—
—
3.9
c
4.46
2.94
7.0
4.02
9.4
2.91
3.4
4.47
11.5
5.8
1.74
2.96
6.06, 6.31
1.63
4.02
c
—
—
c
1.4
2.58
2.50
6.27, 6.14
1.53
3.75
4.20
6.5
2.48b
2.54b
6.16, 6.14
1.65
c
4.15, 4.05
—
c
υ H-15b
—
—
4.37
5.4
5.4
c
J(H-15a,OH-15)
J(H-15b,OH-15)
υ OH-15
4.8
4.21
4.3
4.08
4.4
5.13
c
—
4.77
a Major conformer.
b Assignments interchangeable.
c Not measured.
1
13
°
conformers, 2A and 2B (the major conformer is shown
in Fig. 2), in an 85 : 15 conformer ratio. The assignment of
the major conformer by gHSQC and gHMBC experiments
confirmed the structure (Tables 2 and 3). The attachment
of a methyl group to the 11-position was established by
The H and C NMR spectra of 4 at 25 C showed the
presence of only one conformer in CDCl3 solution. The NMR
assignment by gHSQC and gHMBC experiments confirmed
its structure (Tables 2 and 3). The attachment of an aldehyde
group to the 4-position was in agreement with the presence
of a proton at υ 9.47 directly bonded to a methine carbon at
υ 194.83, the disappearance in the 1H and 13C spectra of the
signals corresponding to the 15-hydroxymethylene group of
1 and the presence of cross peaks between H-15 and C-3, C-4
and C-5 in the gHMBC spectrum of 4.
1
the disappearance in the H and 13C spectra of the signals
corresponding to the exo-methylene group, the strong upfield
shift of the carbon signal corresponding to C-11, and the
appearance of a new carbon signal at υ 17.33 and a 3H
doublet signal at υ 1.24, which was coupled with H-11ˇ.
3
The value of the coupling constant, JꢁH-7, H-11ꢂ D 9.3 Hz,
and the observed cross peak between H-7˛ and the Me-13
protons in a NOESY experiment established an ˛ orientation
for the Me-13 group (11R-configuration).
Conformational analysis
Conformational analysis of 1 was performed in two steps.
The first step consisted of the full assignment of the four 1H
and 13C subspectra co-existing in solution. The second step
consisted of the assignment of the different spin systems to
each particular conformer by means of NOE experiments
and coupling constants values.
The 1H and 13C NMR spectra of 3 in CDCl3 at 25 C
°
showed the presence of two set of sharp resonances
corresponding to conformers 3A and 3B (Fig. 2) in a 55 : 45
ratio. The 1H NMR spectrum of 3, characterized as isabelin,
has previously been partially assigned.2,3 The presence in 3
of a second lactone moiety was evident from the appearance
of a 13C signal at υ 171.71, which showed gHMBC cross
peaks with H-3ˇ and H-5. Fully assigned 1H and 13C
NMR data for 3 are reported in Tables 2 and 3 for the first
time.
For 1, the study was performed at two temperatures (25
°
and ꢀ30 C), using acetone-d6 as the most suitable solvent.
The NOESY spectrum at room temperature proved to be
particularly useful in the assignments, showing the exchange
peaks (EXSY) between equivalent protons in the slowly
interconverting conformers.7 These peaks have the same
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

NMR and x-ray conformational study of germacranolides
477
Table 2. 1H NMR data (υ, ppm; J, Hz) for 2–4
Parameter
2A
2B
3A
3B
4
υ H-1
4.88
3.7
10.9
2.05
2.15
1.72
2.48
4.69
4.7
4.27
9.3
10.0
4.67
1.90
9.3
5.03
8.1
8.1
4.92
12.3
4.1
4.82
8.0
8.0
2.72b
2.50b
2.32b
2.50b
6.67
1.0
5.04
7.0
—
—
2.65
7.2
5.29
8.5
8.5
1.99b
2.46b
2.67b
2.16b
6.62
J(H-1,H-2˛)
J(H-1,H-2ˇ)
υ H-2˛
υ H-2ˇ
υ H-3˛
2.28
2.15
1.98
2.53
2.20
2.72b
2.30b
6.83
1.7
5.20
1.7
—
a
υ H-3ˇ
υ H-5
—
4.62
4.3
3.97
J(H-5,H-6ˇ)
υ H-6ˇ
J(H-6ˇ,H-7˛)
J(H-6ˇ,OH-6)
υ OH-6
10.2
4.49
1.3
a
—
—
—
a
a
—
—
a
a
—
3.23
8.5
υ H-7˛
1.76
3.09
8.1
a
J(H-7˛,H-8ˇ)
J(H-7˛,H-11ˇ)
υ H-8ˇ
J(H-8ˇ,H-9˛)
J(H-8ˇ,H-9ˇ)
υ H-9˛
υ H-9ˇ
υ H-11ˇ
υ H-13
υ Me-14
υ H-15a
υ H-15b
J(H-15a,OH-15)
J(H-15b,OH-15)
υ OH-15
—
—
a
9.3
4.11
9.0
3.89
11.0
5.0
1.78
2.73
2.49
1.29
1.60
3.81
4.36
4.41
11.1
5.0
1.91
3.08
—
6.41, 6.14
1.60
—
4.39
11.2
2.4
2.09
2.78
—
6.45, 5.85
1.69
—
4.39
11.3
4.0
2.07
2.84
<1
2.35
2.50
2.70
1.24
1.39
3.64
4.01
6.2
6.47, 5.85
1.78
9.47
—
—
—
—
—
a
—
—
—
—
a
4.6
4.49
a
—
—
—
a Not measured.
b These assignments could be reversed.
Table 3. 13C NMR chemical shifts (υ, ppm) of 1–4
1
2
3
4
°
°
°
°
°
CDCl3 ꢀ30 C
Me2CO-d6 ꢀ30 C
DMSO 25 C
CDCl3 25 C
CDCl3 25 C
Carbon
1A
1A
1B
1C
1D
2A
3A
3B
4A
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
128.89
24.91
36.09
140.39
128.89
69.21
50.65
81.06
40.98
129.60
135.13
171.35
127.44
21.09
131.29
26.06
32.14
141.63
129.72
71.35
52.26
81.28
42.73
128.64
138.13
170.72
126.26
21.50
131.41
26.33
34.76
139.34
136.13
69.18
54.50
80.33
48.10
131.34
138.36
170.87
125.94
17.41
131.44
24.77
36.70
138.13
133.51
70.46
51.53
84.00
43.19
130.27
138.44
170.91
126.24
21.82
—
129.54
25.08
33.36
135.88
134.74
68.11
58.31
78.91
46.89
130.75
40.35
178.35
17.33
16.32
57.97
130.46
23.49
25.37
131.75
147.16
81.25
50.59
81.46
41.52
131.10
136.74
168.12
123.17
20.76
130.13
23.87
22.41
134.01
151.60
81.88
55.45
75.49
47.05
131.96
136.89
168.39
124.77
17.99
128.64
25.65
23.43
142b
152.60
67.35
50.92
78.59
45.35
131.20
132b
25.33
35.53
a
—
—
a
72.26
53.42
78.96
47.69
a
—
—
—
a
a
169.94
125.32
17.31
194.83
126.52
17.65
61.74
61.46
61.07
59.56
61.96
171.71
172.54
a Not measured.
b Broad signal.
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

478
M. L. Jimeno et al.
15
14
10
2
4
1
9
8
12
6
3
5
7
11
13
2A (UU)
14
12
11
8
1
10
9
10
9
7
8
7
2
12
5
6
2
1
13
3
4
14
5
11
6
3
4
15
13
15
3A (DD)
3B (UD)
12
15
13
11
3
8
1
2
9
4
10
7
5
6
14
4 (DU)
Figure 2. AM1 conformers of 2–4. Relevant experimental NOEs are indicated by arrows.
phase as the diagonal peaks and were easily recognized in the
phase-sensitive NOESY spectrum (Fig. 3). This experiment,
together with the COSY, TOCSY, gHSQC and gHMBC
the usual exchange models established for trans,trans-
germacradienolides.9 Examination of the models suggests
that the 1A–1B, 1A–1D, 1B–1C and 1C–1D interconversions
(Fig. 1) are relatively easy through a low-energy exchange
barrier. On the other hand, the interconversion between
1B and 1D or 1A and 1C requires much more significant
°
experiments obtained at ꢀ30 C, allowed us to obtain the
full assignment of the four 1H and 13C conformer subspectra
of 1, as can be seen in Tables 1 and 3.
°
In order to identify these conformers, we performed a
molecular modeling study. First, we used the conformational
search module, as implemented in the Hyperchem program,8
to determine low-energy conformers. We found two families
of conformers depending on the relative orientation (crossed
or parallel) of the double bonds. Four conformations, UU,
UD, DU and DD, where U (up) and D (down) refer to
the orientation of the C-10 methyl and C-4 substituient
moieties on the ring, were considered in accord with
rearrangement. 1D NOESY spectra at ꢀ30 C provided the
assignment of the different subspectra to each conformer
(see Fig. 1). Additional information about the geometry of
the conformers could be obtained from the experimental
vicinal coupling constants. Thus, the coupling constants
3J(H-5,H-6ˇ), 3J(H-6ˇ,H-7˛), 3J(H-7˛,H-8ˇ), 3J(H-8ˇ,H-9˛)
and J(H-8ˇ,H-9ˇ) are sensitive to changes in the geometry
of the conformers. Using the dihedral angles calculated
by the Altona equation10 and the NOE information, we
3
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

NMR and x-ray conformational study of germacranolides
479
(a)
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5
6A
15bA
15bC
8C
8B
15aA
1A 1C
(b)
5B 5C
1B
6
C 15bB
15aB
5A
8A 6
B 15aC
3.6
3.8
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
5.6
6B /6A
6B/6C
1A /1B
1B /1C
5A/5B
5.6 5.4 5.2 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6
Figure 3. (a) 500 MHz phase-sensitive EXSY spectrum of 1 at 298 K. (b) Expansion of the EXSY spectrum showing the υ 5.7–3.6
ppm region. Slight differences in the chemical shifts of the H-15, OH-15 and OH-6 protons can be observed between the room
temperature and 243 K spectra. The EXSY experiment was obtained with a long mixing time in order to map the whole
exchange mechanism.
generated starting structures for each conformer. The
geometry optimization was performed for each conformer
using the AM1 method, obtaining good agreement between
experimental and calculated data (Table 4).
For conformer 1A, irradiation of the methyl-14 protons
at υ 1.66 showed enhancements in the H-9ˇ and H-2ˇ
resonances at υ 2.86 and 2.41, respectively. Furthermore,
irradiation of the H-15a proton (υ 3.88) showed enhancements
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

480
M. L. Jimeno et al.
Table 4. Experimental and calculated coupling constants (J, Hz) for 1
1A
Jcalc.
1B
Jcalc.
1C
Jcalc.
1D
Jcalc.
Fragment
 AM1
Jexp.
 AM1
Jexp.
 AM1
Jexp.
 AM1
Jexp.
H-5, H-6ˇ
38.8
177.8
132.9
ꢀ169.0
74.3
4.2
9.8
6.2
11.5
2.2
5.6
10.2
3.4
11.4
5.8
ꢀ178.0
ꢀ168.9
136.8
ꢀ151.9
90.7
9.8
9.9
6.8
10.1
1.2
10.0
9.3
7.0
9.4
1.4
176.8
ꢀ179.9
133.3
ꢀ174.5
68.9
9.9
9.9
6.2
11.6
2.9
10.0
10.0
5.2
11.5
4.2
30.3
171.5
115.3
168.9
52.9
5.3
9.4
3.5
10.8
5.2
5.1
7.5
4.6
H-6ˇ, H-7˛
H-7˛, H-8ˇ
H-8ˇ, H-9˛
H-8ˇ, H-9ˇ
a
—
a
—
a Not measured.
in the H-15b, H-1 and OH-15 resonances (υ 4.12, 5.16 and
4.21, respectively). These results suggest that conformer 1A
has an up–down (UD) disposition between the methyl and
the hydroxymethylene groups. In addition, the 3J(H-5,H-6ˇ)
(5.6 Hz), 3J(H-7˛,H-8ˇ) (3.4 Hz) and 3J(H-6ˇ,H-7˛) (10.2 Hz)
point to a gauche arrangement between H-5/H-6ˇ and H-
7˛/H-8ˇ, and an anti arrangement between H-6ˇ/H-7˛.
For 1B, irradiation of the methyl-14 protons (υ 1.53) caused
enhancements in the H-8ˇ, H-15b, H-9ˇ and H-2ˇ resonances
(υ 4.02, 3.75, 2.50 and 2.33, respectively) and irradiation of
H-13 exo (υ 6.27) produced enhancements in H-13 endo, OH-6
and H-7˛ (υ 6.14, 4.46, and 2.94, respectively), suggesting
that the 1B conformer has an up–up (UU) disposition. The
3J(H-5,H-6ˇ) (10.0 Hz) and 3J(H-6ˇ,H-7˛) (9.3 Hz) indicate an
anti arrangement for these fragments. For 1C, irradiation of
the methyl-14 protons at υ 1.68 caused enhancements in the
H-5, H-7˛, H-9˛ and H-2˛ resonances (υ 4.86, 2.74, 1.93 and
2.38, respectively), suggesting that the 1C conformer has a
down–up (DU) disposition. The 3J(H-5,H-6ˇ) (10.0 Hz) and
3J(H-6ˇ,H-7˛) (10.0 Hz) values show the same disposition as
1B for these fragments. Finally, irradiation of the methyl-14
protons of 1D at υ 1.65 caused an enhancement in the H-
15a signal (υ 4.37), suggesting that the 1D conformer has a
down–down (DD) disposition.
b-axis through (ꢀy C 1, x ꢀ y, z ꢀ 2/3), (ꢀy C 1, x ꢀ y, z C
1/3), (ꢀx C y, ꢀx, z ꢀ 1/3) and (ꢀy, x ꢀ y, z C 1/3), and
along the a C b diagonal distribution through (ꢀy, x ꢀ y, z C
1/3) and (ꢀy, x ꢀ y, z ꢀ 2/3).16
The conformational study of 2 was carried out in
DMSO-d6 at room temperature on the basis of the value
of the coupling constants and the NOE effects. The EXSY
H13A
C13
O4
H13B
C12
C11
O1
H7
O3
H1
C7
C5
C8
H6
H9B
H5
H5
C5
H9A
C9
The structure of artemisiifolin (1) was later confirmed by
an x-ray crystallographic analysis. Figure 4 is an ORTEP11
view of the molecule of 1 in the crystalline state, show-
ing only a conformation almost identical with that of the
major conformer detected in acetone-d6 solution (1B, see
Fig. 1). In the crystal of 1 there are three formula units in the
unit cell. The bond distances and angles are as usual.9,12–14
The conformation of the 10-membered ring is such that
the torsion angle around the virtual bond C-5—C-10 is
H1A
H15A
C15
C4
C10
O2
H14C
H14B
C1
C3
H2
H3A
C14
C2
H15B
H14A
H2A
H2B
Figure 4. ORTEP11 molecular structure of artemisiifolin (1)
with the numbering scheme. Displacement ellipsoids are
drawn at the 30% probability level.
°
ꢀ70.5(3) , a value similar to that found in a structurally
related compound14 (ꢀ70.1 ), and also that conformation
°
is characterized by amplitudes and phases expressed by a
Fourier synthesis.15 In the crystalline state there are two
intermolecular hydrogen interactions (O—HÐ Ð ÐO) which
build columns along the c-axis that determine the secondary
motif in the packing (Table 5). This distribution is rein-
forced by other contacts mainly hydrophobic through the
symmetries (ꢀy C 1, x ꢀ y C 1, z ꢀ 2/3), (ꢀx C y, ꢀx C 1, z ꢀ
1/3), (ꢀx C y, ꢀx C 1, z C 2/3), (ꢀy C 1, x ꢀ y C 1, z C 1/3)
and (x, y, z C 1) (Fig. 5). The columns pack to form layers
along the a-axis determined by van der Waals interac-
tions through the symmetries (ꢀx C y C 1, ꢀx C 1, z ꢀ 1/3)
and (ꢀx C y C 1, ꢀx C 1, z C 2/3); the layers pile along the
Table 5. Hydrogen bonds for artemisiifolin (1) in the crystalline
statea
D—H HÐ Ð ÐA
DÐ Ð ÐA
D—HÐ Ð ÐA
˚
˚
˚
°
( )
D—HÐ Ð ÐA
(A)
(A)
(A)
O(1)—H(1)Ð Ð ÐO(2)i
0.82
0.82
1.94
1.99
2.748(3)
2.790(3)
169.2
164.0
O(2)—H(2)Ð Ð ÐO(1)ii
a Symmetry transformations used to generate equivalent atoms:
ꢀx C y, ꢀx C 1, z ꢀ 1/3; ꢁiiꢂ ꢀx C y, ꢀx C 1, z C 2/3.
ꢁiꢂ
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

NMR and x-ray conformational study of germacranolides
481
between H-1/H-8ˇ and H-5/H-8ˇ for 3A suggests that
this conformer has a down–down (DD, Fig. 2) disposition,
3
and the coupling constant values J(H-5,H-6ˇ) (1.7 Hz) and
3J(H-6ˇ,H-7˛) (1.7 Hz) are in accordance with an orthogonal
arrangement of the dihedral angles H5—C5—C6—H6ˇ and
H6ˇ—C6—C7—H7˛. For 3B, the cross peaks H-5/Me-14
and H-8ˇ/Me-14 suggest that this conformer has a up–down
(UD) disposition (see Fig. 2) and the coupling constant value
3J(H-5,H-6ˇ) (1.0 Hz) is in accordance with an orthogonal
arrangement of the dihedral angle H5—C5—C6—H6ˇ
(Table 6). These results are in agreement with the published
data.2,3
Finally, the only conformation present in CDCl3 solution
for 4 was studied. For 4, irradiation of the H-1 proton (υ
5.29) showed an enhancement in the H-8ˇ signal (υ 4.39),
and the irradiation of H-9˛ showed enhancements in Me-14
and H-7˛. This result suggests that the 4 conformer has a
down–up (DU) disposition (see Fig. 2). However, differences
from the similar conformer 1C in the conformation around
the C5—C6 bond were observed (Table 6).
EXPERIMENTAL
General experimental procedures
Melting-points were determined on a Kofler block and are
uncorrected. Optical rotations were measured on a Perkin-
Elmer 241 MC polarimeter. IR spectra were obtained on a
Perkin-Elmer Spectrum One spectrophotometer. UV spec-
tra were recorded on a Perkin-Elmer Lambda 2 UV–visible
spectrophotometer. Mass spectra were registered in the pos-
itive electron ionization (EI) mode on a Hewlett-Packard
model 5973 instrument (70 eV). Elemental analyses were
made with a Carlo Erba EA1108 apparatus. Merck silica gel
No. 7734 (70–230 mesh) was used for column chromatog-
raphy. Merck 5554 Kieselgel 60 F254 sheets were used for
thin-layer chromatographic (TLC) analysis.
Figure 5. The crystal of artemisiifolin (1) displaying the
packing mode.
experiment showed the exchange peaks between equivalent
protons in the two slowly interconverting conformers.
Reliable coupling constants and NOESY correlations were
only obtained for the major conformer, precluding the
possibility of determining the geometry of the minor
conformer. For 2A, irradiation of the Me-14 protons (υ 1.39)
showed enhancements in the H-8ˇ, H-15a, H-9ˇ and H-2ˇ
resonances (υ 4.11, 3.64, 2.50 and 2.15, respectively). These
results suggest that the 2A conformer has an up–up (UU)
disposition (see Fig. 2). From the coupling constant values
3J(H-5,H-6ˇ) (4.7 Hz), 3J(H-6ˇ,H-7˛) (9.3 Hz) and 3J(H-7˛,H-
8ˇ) (9.3 Hz), a gauche arrangement between H-5/H-6ˇ and
an anti arrangement between H-6ˇ/H-7˛ and H-7˛/H-8ˇ
could be determined for this conformer (Table 6).
Partial studies of the conformations of 3 in CDCl3
solution have already been reported.2,3 On basis of the
full assignment of the 1H NMR spectrum of 3 and using
NOESY experiments and coupling constant values, we
determined the global conformation for the two conformers.
Thus, in the NOESY spectrum, the existence of cross peaks
NMR spectra
NMR measurements were performed on a Varian UNITY-500
spectrometer equipped with a 5 mm gradient inverse detec-
tion probe, operating at 499.88 MHz (1H) and 125.71 MHz
(13C). Data were obtained from CDCl3, DMSO-d6 and
CD3OD solutions. Standard Varian pulse programs were
used throughout. One-dimensional 1H and 13C spectra were
recorded under standard conditions. 1D-NOESY experi-
ments were acquired using mixing times of 500 ms. Homonu-
clear 2D spectra, gCOSY, TOCSY and NOESY, were acquired
in the phase-sensitive mode. Data were collected in a
Table 6. Experimental and calculated coupling constants (J, Hz) for 2–4
2A
Jcalc.
3A
Jcalc.
3B
Jcalc.
4A
Jcalc.
Fragment
 AM1
Jexp.
 AM1
Jexp.
 AM1
Jexp.
 AM1
Jexp.
H-5, H-6ˇ
139.0
ꢀ155.1
147.9
ꢀ160.8
82.0
5.4
9.9
8.8
9.9
1.4
4.7
9.3
9.3
9.0
78.3
110.4
143.9
177.9
61.8
0.6
1.6
8.3
10.9
3.0
1.7
1.7
8.5
11.1
5.0
87.4
138.5
143.7
ꢀ167.1
76.3
0.6
5.9
8.2
10.5
1.8
1.0
7.0
7.2
11.2
2.4
ꢀ174.7
87.1
ꢀ174.8
ꢀ176.4
62.3
11.1
0.8
11.1
10.9
3.0
10.2
1.3
8.1
11.3
4.0
H-6ˇ, H-7˛
H-7˛, H-8ˇ
H-8ˇ, H-9˛
H-8ˇ, H-9ˇ
<1
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

482
M. L. Jimeno et al.
columnchromatography [silicagel deactivatedwith10% H2O(w/v),
light petroleum–EtOAc (7 : 3) as eluent], yielding 60 fractions (15 ml
each). Fractions 28–37 contained isabelin (3) (320 mg, 32% yield) and
1024 ð 256 matrix with a spectral width of 5000 Hz and a
2 s recycle delay. The data were linear predicted forward to
512 points in the t1 dimension and zero-filled to 1024 points.
Inverse proton detected heteronuclear 2D spectra, gHSQC
and gHMBC, were acquired in the phase-sensitive mode.
Data were collected in a 4096 ð 128 matrix with a spec-
tral width of 8000 Hz in the proton domain and 25000 Hz
in the carbon domain. The data were linear predicted for-
ward to 256 points in the t1 dimension and zero-filled to
512 points. The gHSQC experiment was optimized for a
one-bond heteronuclear coupling constant of 145 Hz. The
gHMBC experiment was optimized for long-range coupling
constants of 8 Hz.
fractions 52–59 provided 4 (270 mg, 27% yield).
20
°
°
Isabelin (3): m.p. 168–170 C (Me2CO–pentane), [˛]D ꢀ 56.7
25
2
°
°
(CHCl3; c 0.213); literature: m.p. 169–170 C, [˛]D ꢀ57.2 (CHCl3; c
0.87). IR and UV spectra were identical to those reported previously.2
EIMS: m/z (relative intensity, %) 260 [M]C (3), 231 (4), 215 (4), 164
(17), 149 (14), 135 (24), 124 (59), 96 (100), 77 (41), 68 (77), 55 (44),
43 (78), 39 (74). For 1H and 13C NMR spectra, see Tables 2 and 3,
respectively.
6˛-Hydroxy-15-oxogermacra-1(10)E,4Z,11(13)-trien-12,8˛-olide
20
°
°
(4): m.p. 153–156 C (EtOAc–n-hexane), [˛]D ꢀ63.9 (CHCl3; c
0.668); IR, ꢃmax: 3435, 2935, 2862, 1762, 1687, 1635, 1312, 1252, 1144,
1103, 1072, 1040, 1007, 970, 934, 893, 817, 763 cmꢀ1; UV (MeOH),
ꢄmax (log ε): 223 (3.55); EIMS: m/z (relative intensity, %) 262 [M]C
(12), 244 (4), 233 (31), 215 (13), 187 (12), 159 (13), 148 (39), 137 (25),
123 (69), 105 (30), 97 (65), 91 (84), 69 (94), 68 (100), 67 (82), 53 (44), 41
(88), 39 (73). Found: C, 68.49; H, 6.98. C15H18O4 requires C, 68.68; H,
6.92%. For 1H and 13C NMR spectra, see Tables 2 and 3, respectively.
Molecular modeling
Semi-empirical calculations were performed using the
HyperChem package.8 The computations were carried out
on a Windows 2000 workstation with a Pentium III PC. The
starting structures were generated according to previous
molecular mechanics calculations and to NMR restrictions.
Final geometries were optimized using the AM1 method.
X-ray crystallography of artemisiifolin (1)
Crystals of artemisiifolin (1) suitable for x-ray diffraction
analysis were obtained by recrystallization from EtOAc. A
colourless prism of 1 (0.47 ð 0.36 ð 0.08 mm) was selected
for the data collection. Crystal data: C15 H20 O4; Mr
D
264.32 g molꢀ1; trigonal a D 11.0254(3) A, b D 11.0254(3) A,
˚
˚
Plant material
˚
°
°
°
c D 9.7614(5) A, ˛ D 90 , ˇ D 90 , ꢀ D 120 , V D
Aerial parts of Staehelina dubia L. were collected in August
2000, near Ciruelos del Pinar, Guadalajara province, Spain.
A voucher specimen (No. 479/2000) is deposited in the
Herbarium of the Department of Pharmacognosy, Faculty of
Pharmacy, Complutense University, Madrid, Spain.
1027.6(1) A , space group P31, Z D 3, Dcal D 1.2813 Mg mꢀ3
.
3
˚
Data collection: Seifert XRD 3000S diffractometer, 273.(2)
K; 1289 independent reflection intensities were collected
°
between 4.63 and 66.80 in Â, in the ω/2Â scan mode, with
˚
Cu K˛ monochromated radiation (ꢄ D 1.54180 A). No decay
was observed in two reference reflections measured every
150 min, and 1284 reflections were considered as observed
at the 2ꢅ(I) level.
Extraction and isolation of artemisiifolin (1)
Dried and powdered aerial parts of the plant (1.5 kg) were extracted
with Me2CO (3 ð 10 l) at room temperature for 1 week. Filtration
and evaporation of the solvent yielded a residue (86 g) which was
subjected to column chromatography [silica gel 1080 g, deactivated
The structure was solved by direct methods using
SHELX-9717 and difference Fourier techniques; no absorption
correction was applied (ꢆ D 0.251 mmꢀ1). The structure
was refined using full-matrix least-squares on F2 through
SHELXL-97.17 All non-H atoms were refined with anisotropic
thermal parameters. Since 1 crystallizes in a polar space
group, polar axis restraints were applied.18 The H atoms
were assigned geometrically and treated using appropriate
riding models. The refinement converged to R final indices
[I > 2ꢅ(I)] R1 D 0.036, ωR2 D 0.100 and R indices (all data)
R1 D 0.036, ωR2 D 0.100. All calculations were done with the
program SHELX-97.17 All the geometric calculations were
performed with the program PARST,19 and scattering factors,
anomalous dispersion and absorption coefficients were
taken from International Tables for X-ray Crystallography.20
Crystallographic data for the structure reported in this paper
have been deposited with the Cambridge Crystallographic
Data Centre (CCDC 193230).
°
with 10% (w/v) H2O], eluting with light petroleum (b.p. 60–70 C)
and then with a light petroleum–EtOAc gradient from 9 : 1 to 1 : 1.
The chromatographic fractions eluted with light petroleum–EtOAc
(3 : 2) furnished a crystalline residue, which was recrystallized
from EtOAc, giving pure artemisiifolin (1) (27 g, 1.8% on dry
22
°
°
plant material): m.p. 135–137 C, [˛]D C 58.9 (CHCl3; c 1.557);
25
literature:1 m.p. 131 C, [˛]D C 54.6 (MeOH). IR, UV and mass
spectra were identical with those reported previously.1 For 1H and
13C NMR spectra, see Tables 1 and 3, respectively.
°
°
Sodium borohydride reduction of 1 to give
11,13-dihydroartemisiifolin (2)
A stirred solution of 1 (330 mg) in MeOH (20 ml) was treated with
an excess of NaBH4 (200 mg) for 30 min at room temperature.
Then water (50 ml) was added and the reaction mixture was
extracted with CH2Cl2 (4 ð 25 ml). The combined extracts were
dried (Na2SO4), filtered and evaporated to dryness, giving a residue
(328 mg, 98.6% yield), which was crystallized from EtOAc: m.p.
20
1
°
°
175–177 C, [˛]D C 89.6 [CHCl3 –MeOH (1 : 1); c 0.556]; literature:
25
°
°
m.p. 183–185 C, [˛]D C 94 (MeOH). IR and UV spectra were
identical with those reported previously.1 EIMS: m/z (relative
intensity, %) 266 [M]C (1), 248 (5), 235 (14), 219 (15), 217 (17),
121 (45), 107 (67), 93 (65), 79 (100), 69 (65), 55 (36), 53 (36), 41 (64).
Found: C, 67.72; H, 8.60. C15H22O4 requires C, 67.64; H, 8.33%. For
1H and 13C NMR spectra, see Tables 2 and 3, respectively.
Acknowledgements
The authors thank the Spanish Direccio´n General de Investigacio´n
´
Cientıfica y Te´cnica (DGICYT) (grant No. BQU 2000-0868) and
´
Comisio´n Interministerial de Ciencia y Tecnologıa (CICYT) (grant
No. AGL 2001-1652) for financial support.
Pyridinium dichromate oxidation of 1: isabelin (3)
and compound 4
Pyridinium dichromate (PDC) (2.25 g) was added to a solution of
1 (1.01 g) in anhydrous THF (55 ml) under argon. The mixture was
stirred for 48 h at room temperature. Then Et2O (25 ml) was added
and the mixture was filtered through a pad of Celite. Evaporation
of the solvents gave a residue (630 mg), which was subjected to
REFERENCES
1. Porter TH, Mabry TJ, Yoshioka H, Fischer NH. Phytochemistry
1970; 9: 199.
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483

NMR and x-ray conformational study of germacranolides
483
2. Yoshioka H, Mabry TJ. Tetrahedron 1969; 25: 4767.
3. Tori K, Horibe I, Yoshioka H, Mabry TJ. J. Chem. Soc. B 1971;
1084.
4. Tashkhodzhaev B, Abduazimov BK. Khim. Prir. Soedin. 1997; 33:
497.
13. Quijano L, Caldero´n JS, Go´mez GF, Lo´pez PJ, Rios T,
Froncsek FR. Phytochemistry 1984; 23: 1971.
14. Toscano RA, Maldonado E, Ortega A. J. Crystallogr. Spectrosc.
Res. 1992; 2: 461.
´
15. Cano FH, Foces-Foces C, Garcıa-Blanco S. Acta Crystallogr., Sect.
5. Ru¨ngeler P, Brecht V, Tamayo-Castillo G, Merfort I. Phytochem-
istry 2001; 56: 475.
A 1978; 34: S91.
´
´
16. Martınez-Alca´zar MP, Cano FH, Martınez-Ruiz P. J. Mol. Struct.
2001; 562: 79.
6. Wong H-F, Brown GD. Phytochemistry 2002; 59: 529.
7. Perrin CL, Dwyer TJ. Chem. Rev. 1990; 90: 935.
8. HyperChem TM Release 5.1. Windows Molecular Modelling
System. HyperCube: Waterloo, Ontario, Canada, 1997.
9. Watson WH, Kashyap RP. J. Org. Chem. 1986; 51: 2521.
10. Haasnoot CAG, de Leeuw FAAM, Altona C. Tetrahedron 1980;
36: 2783.
17. Sheldrick GM. SHELX-97 (SHELXS and SHELXL). Manual of
Programs for Refinement of Crystal Structures. University of
Go¨ttingen: Go¨ttingen, 1997.
18. Flack HD, Schwarzenbach D. Acta Crystallogr., Sect. A 1988; 44:
499.
19. Nardelli M. Comput. Chem. 1983; 7: 95.
20. International Tables for X-ray Crystallography. Hamilton WC (ed),
vol. C. Kluwer: Dordrecht, 1992.
11. Hall SR, Flack HD, Stewart JM. Xtal System. University of
Western Australia, 1999.
12. Rychlewska U. J. Chem. Soc., Perkin Trans. 2 1982; 1641.
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 474–483