Xylogranatins F-R: Antifeedants from the Chinese mangrove, Xylocarpus granatum, a new biogenetic pathway to tetranortriterpenoids

Article abstract of DOI:10.1002/chem.200700663

Thirteen limonoids with a new carbon skeleton, the xylogranatins F-R (1-13), have been isolated from the seeds of a Chinese mangrove, Xylocarpus granatum; two recently reported compounds, xylogranatins C and D were also isolated. Their structures were elucidated on the basis of spectroscopic data and chemical methods. The absolute configurations of these compounds were determined by using the modified Mosher MTPA ester method and by quantum chemical circular dichroism (CD) calculations. Xylogranatins F-Q are the first aromatic B-ring limonoids found in nature. They belong to two substructural classes, of which one (1-3) contains a pyridine ring while the other one (4-12) contains a central furan core. Xylogranatins C and R can be considered to be key biosynthetic intermediates, while xylogranatin D, the only limonoid found so far with a carbon skeleton that conatains a C₃₀-C₉ linkage, is apparently an artifact. The structures of these compounds suggest a new biogenetic pathway to tetranortriterpenoids. Xylogranatins F, G and R were found to exhibit marked antifeedant activity against the third instar larvae of Mythimna separata (Walker) at a concentration of 1 mgmL^-1. The most potent compound tested was xylogranatin G. Its AFC₅₀ (concentration for 50% antifeedant activity) values at the exposure times of 24 and 48 h were 0.31 and 0.30 mgmL^-1, respectively.

Full text of DOI:10.1002/chem.200700663

FULL PAPER
DOI: 10.1002/chem.200700663
Xylogranatins F–R: Antifeedants from the Chinese Mangrove, Xylocarpus
granatum, A New Biogenetic Pathway to Tetranortriterpenoids
Jun Wu,*[a] Si Zhang,[a] Torsten Bruhn,[b] Qiang Xiao,[c] Haixin Ding,[c] and
Gerhard Bringmann*[b]
Dedicated to Professor Dr. Peter Welzel on the occasion of his 70th birthday
À
Abstract: Thirteen limonoids with a
new carbon skeleton, the xylograna-
tins F–R (1–13), have been isolated
from the seeds of a Chinese mangrove,
Xylocarpus granatum; two recently re-
ported compounds, xylogranatins C
and D were also isolated.Their struc-
tures were elucidated on the basis of
spectroscopic data and chemical meth-
ods.The absolute configurations of
these compounds were determined by
using the modified Mosher MTPA
ester method and by quantum chemical
circular dichroism (CD) calculations.
Xylogranatins F–Q are the first aro-
matic B-ring limonoids found in
nature.They belong to two substructur-
al classes, of which one (1–3) contains
a pyridine ring while the other one (4–
12) contains a central furan core.Xy-
logranatins C and R can be considered
to be key biosynthetic intermediates,
while xylogranatin D, the only limo-
noid found so far with a carbon skele-
ton that conatains a C30 C9 linkage, is
apparently an artifact.The structures
of these compounds suggest a new bio-
genetic pathway to tetranortriterpe-
noids.Xylogranatins F, G and R were
found to exhibit marked antifeedant
activity against the third instar larvae
of Mythimna separata (Walker) at a
concentration of 1 mgmLÀ1.The most
potent compound tested was xylogra-
natin G.Its AFC
(concentration for
50
Keywords: antifeedant activity
·
50% antifeedant activity) values at the
exposure times of 24 and 48 h were
0.31 and 0.30 mgmLÀ1, respectively.
biosynthesis configuration
·
determination · natural products ·
xylogranatins
Introduction
ring substituent located at C17 that is derived from a precur-
sor with a 4,4,8-trimethyl-17-furanylsteroid skeleton.They
are classified by the four-ring structure in the intact triter-
pene core unit, and these are usually oxidized and designat-
ed as A, B, C and D.The mangrove, Xylocarpus granatum,
is known to produce antifeedant limonoids, especially phrag-
malins and mexicanolides.Previous investigations on the
seeds of two Meliaceae plants, the mangroves, X. granatum
and X. moluccensis, uncovered one obacunol, two phragma-
lins, three andirobins, and 14 mexicanolides, including the
xyloccensins A–K.[1–5] During the course of our search for
potential lead structures from Chinese tropical mangrove
plants, we have reported the isolation and identification of a
mixture of butyrospermol fatty acid esters,[6] eleven mexica-
nolides and 13 phragmalins, named xyloccensins L–Z[7–15]
and xylogranatins A–E,[16–17] from the stem bark and fruit of
a Chinese mangrove X. granatum.Five phragmalins, four of
which were the same as reported by us earlier,[9] have been
obtained from the stem bark of the same plant.[18] On the
other hand, two mexicanolides from the fruit of X. moluc-
censis,[19] although structurally different from our xyloccen-
Limonoids, which have been found to date only in plants of
the order Rutales, are tetranortriterpenoids with a b-furyl
[a] Dr.J.Wu, Prof.S.Zhang
Guangdong Key Laboratory of Marine Materia Medica
South China Sea Institute of Oceanology
Chinese Academy of Sciences, 164 West Xingang Road
510301 Guangzhou (P.R. China)
Fax : (+86)20-8445-1672
E-mail: wwujun2003@yahoo.com
[b] Dr.T.Bruhn, Prof.Dr.G.Bringmann
Institute for Organic Chemistry, University Würzburg
Am Hubland, 97074 Würzburg (Germany)
Fax : (+49)931-888-4755
E-mail: bringmann@chemie.uni-wuerzburg.de
[c] Dr.Q.Xiao, H.Ding
Institute of Organic Chemistry
Jiangxi Science & Technology Normal University
330013 Nanchang (P.R. China)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
1129

sins X (including xyloccensins X1 and X2) and Y,[13,14] were,
nonetheless, given the same names, xyloccensins X and Y.
Furthermore, four unusual 9,10-seco-limonoids, named xy-
logranatins A–D,[20] of which two had previously been isolat-
ed by us, have been isolated from the seeds of the same
plant.In the current paper, we present the isolation and
characterization of thirteen limonoids with new carbon skel-
etons, xylogranatins F–R (1–13; Scheme 1), from the seeds
petroleum ether.The aqueous layer was further extracted
with ethyl acetate and concentrated to give a brown gum,
which was subjected to silica-gel chromatography (chloro-
form/methanol 100:0 to 2:1).The fractions that were eluted
with chloroform/methanol (25:1 to 15:1) were combined and
purified by repetitive C18 HPLC to afford xylogranatins C,
D, and F–R (1–13).
Structural elucidation ofxylogranatins F–H (1–3) : Xylogra-
natin F (1), an amorphous powder, had a molecular formula
of C26H27NO6, which was established by HR-TOFMS (m/z:
calcd: 472.1736; found: 472.1724 [M+Na]+).The 1H and
13C NMR spectroscopy data (Table 1) indicated that the
molecule contained a carbon–nitrogen double bond, five
carbon–carbon double bonds, two carbonyls, and six ring
systems.DEPT experiments revealed that 1 had four tertiary
methyls, three methylenes, eight methines (of which five are
olefinic), and eleven quaternary carbon atoms.In addition,
the NMR spectroscopic data (Table 1) showed the presence
of a hydroxyl, dH =2.24–2.28 ppm (brs), and a b-furyl ring
(dH =6.52 (brs), 7.48 (brs), 7.56 (brs), dC =110.0 (d), 119.8
(s), 141.4 (d), 143.3 ppm (d)). Three substructures 1a (from
C11 to C18 and from C20 to C23), 1b (from C3 to C7, C10,
C19, C28, and C29) and 1c (from C1 to C2, C30, and from
C8 to C9) (Figure 1a) were determined by analysis of the
2D 1H-1H COSY, HSQC, and HMBC spectral data of 1.
Substructure 1a was elucidated by starting from an a,b-un-
saturated d-lactone ring D, which was characterized by the
following NMR spectroscopic data: dH =6.57 (s), 5.22 (s);
dC =37.7 (s), 157.2 (s), 111.2 (d), 165.2 (s), 80.9 ppm (d), and
was corroborated by the HMBC correlations between H15/
C13, H15/C14, H15/C16, H17/C13, and H17/C16 (Figure 1a).
The HMBC cross-peaks from H17 to C20, C21, and C22 in-
dicated that the b-furyl ring is connected to C17 of the d-lac-
tone ring D.A methyl singlet resonance at dH =1.15 ppm
(Me18) and the methylene protons of C12, which show
HMBC correlations to the g-C (C13) of ring D along with a
proton spin system, H211–H212, which was deduced from
1
the H-1H COSY correlations, permitted us to establish the
connections of this proton spin system and Me18 with C13.
Taken together, the above data gave substructure 1a.The
second fragment, substructure 1b, could be assembled by
starting from the g-lactone ring F, which was characterized
by the NMR spectroscopic data: dH =2.97 (d, J=9.5 Hz),
2.58 (dd, J=17.2, 2.4 Hz), 2.98 (dt, J=17.2, 9.5 Hz); dC =
45.8 (d), 31.0 (t), 175.2 (s), 84.1 ppm (s), and was confirmed
Scheme 1.The xylogranatins F-R ( 1–13) isolated from X. granatum.
of X. granatum.Their constitutions and relative configura-
tions were elucidated by spectroscopic and chemical meth-
ods.The absolute configurations of these compounds were
determined by using the modified MTPA Mosher ester
method, and by circular dichroism (CD) measurements in
combination with quantum chemical CD calculations.The
structures of these limonoids hint at a new biosynthetic
pathway to tetranortriterpenoids.
1
by H-1H COSY correlations from H5 to H26 and HMBC
cross-peaks from H5 to C6, C7, and C10 (Figure 1a).The
methyl singlet resonance at d=1.83 ppm (Me19) has HMBC
correlations with b-C (C5) and g-C (C-10) of the g-lactone,
revealing that it is attached to the g-C (C10).A proton sin-
glet of a methine (H3), which bears a hydroxyl resonance at
dH =4.48 ppm and two methyl singlet resonances at d=1.16
(Me28) and 0.83 ppm (Me-29), respectively, exhibits HMBC
cross-peaks to a quaternary carbon (C4) and the b-C (C5)
of the above g-lactone; this suggests that the quaternary
carbon (C4), which bears two methyl groups is situated be-
Results and Discussion
Isolation ofxylogranatins C , D, and F–R (1–13): The dried
seeds of X. granatum were extracted with ethanol.The ex-
tract was concentrated and partitioned between water and
1130
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144

Xylogranatins F–R
FULL PAPER
Table 1. 1H (500 MHz) and 13C NMR (125 MHz) spectroscopic data for xylogranatins F–H (1–3).[a]
Position
1 (CDCl3)
2 (CDCl3)
3 (CD3OD)
1H d [ppm]
13C d
1H d [ppm]
13C d
1H d [ppm]
13C d
[ppm]
[ppm]
[ppm]
1
2
3
4
156.7 (s)
130.6 (s)
75.6 (d)
36.3 (s)
45.8 (d)
31.0 (t)
156.9 (s)
126.8 (s)
76.7 (d)
35.2 (s)
45.5 (d)
30.8 (t)
153.2 (s)
132.5 (s)
76.3 (d)
37.3 (s)
47.0 (d)
31.8 (t)
4.48 (s)
5.60 (s)
4.37 (s)
5
2.97 (d) 9.5
2.58 (dd) 17.2 2.4 2.98 (dt) 17.2
9.5
2.93 (d) 9.2
2.61 (dd) 18.0 1.5 3.05 (dd) 18.0
9.2
2.94 (d) 9.2
6a 6b
2.62 (dd) 18.0 2.9 3.13 (m)[b]
7
8
9
10
11a,
11b
175.2 (s)
124.3 (s)
158.1 (s)
84.1 (s)
174.8 (s)
124.3 (s)
158.7 (s)
83.6 (s)
174.5 (s)
134.7 (s)
158.2 (s)
86.0 (s)
3.10 (dd); 18.5 5.0, 3.20 (dd) 18.5 28.0 (t)
4.0
3.13 (dd); 18.5 5.0 3.20 (dd), 18.5 28.1 (t)
4.0
2.96 (dd); 19.0 5.0, 3.09 (dd) 19.0 28.8 (t)
4.0
12a 12b 1.87 (dd) 13.0 4.0 1.75 (dt) 13.0
30.3 (t)
1.87 (dd) 13.0 4.0 1.73 (dt) 13.0
5.0
30.2 (t)
1.96 (dd) 13.0 4.0 1.73 (dt) 13.0
5.0
31.5 (t)
5.0
13
14
15a,
15b
16
17
18
19
20
21
22
23
28
29
30
37.7 (s)
157.2 (s)
111.2 d
37.6 (s)
156.5 (s)
111.7 (d)
37.6 (s)
41.8 d
36.8 (t)
3.15[b]
m
3.02 (dd) 16.0 6.0 2.88 (dd) 16.0
11.5
6.57 (s)
6.59 (s)
165.2 (s)
80.9 (d)
15.8 (q)
28.5 (q)
119.8 (s)
141.4 (d)
110.0 (d)
143.3 d
164.8 (s)
80.9 (d)
15.8 (q)
28.6 (q)
119.9 (s)
141.4 (d)
110.1 (d)
143.3 (d)
23.9 (q)
20.3 (q)
135.7 (d)
178.5 (s)
81.0 (d)
22.5 (q)
28.6 (q)
122.4 (s)
142.6 (d)
111.0 (d)
144.7 (d)
24.1 (q)
21.4 (q)
139.6 (d)
5.22 (s)
1.15 (s)
1.83 (s)
5.21 (s)
1.14 (s)
1.80 (s)
5.44 (s)
1.03 (s)
1.79 (s)
7.56 (brs)
6.52 (brs)
7.48 (brs)
1.16 (s)
0.83 (s)
8.05 (s)
7.55 br(s)
6.51 (brs)
7.48 (brs)
1.13 (s)
0.78 (s)
8.14 (s)
7.66 br(s)
6.61 (brs)
7.59 (brs)
1.15 (s)
0.72 (s)
7.69 (s)
23.5 (q)
21.4 (q)
133.9 (d)
3-OH
3-OAc
2.24–2.28 (brs)
2.06 (s)
20.9 (q)
170.1 (s)
[a] Multiplicities are indicated in parentheses.[b] Overlapped signals without designating multiplicity.
tween C3 and C5.This connection was corroborated by the
HMBC correlation from H5 to C4.From these data, the
second substructure was unambiguously deduced as 1b.The
final fragment, 1c, was assembled as a tetrasubstituted pyri-
dine ring on the basis of the NMR spectroscopic data (dH =
8.05 (s), dC =124.3 (s), 130.6 (s), 133.9 (d), 156.7 (s),
158.1 ppm (s); Table 1) and the HMBC correlations from
H30 to C1, C2, and C9.This identification was consistent
with the result that 1 was a nitrogenous compound, which
was deduced from its molecular formula.Obviously, HMBC
correlations were crucial in assembling the gross structure of
xylogranatin F (1).Briefly, cross-peaks Me19/C1, H3/C1,
H3/C2, and H3/C30, connected substructures 1b and 1c to
that were observed in 1 (Figure 1b) H3/H5, H3/H6b, H3/
H30, H3/Me19, H3/Me28, H5/H6b, and H5/Me19 indicated
their mutual cis relationship and the cis-fused orientation of
rings A/F.The above NOE observations also helped to es-
tablish a b configuration for H3.Similarly, the interactions
H17/H22, H17/H12b, H17/H15, Me18/H21, Me18/H15, and
H30/H15, suggested that the configuration of Me18 is a and
H17 is b and the furyl ring a configured.Moreover, the
NOE interactions between H3 and H30, and between H15
and H30, established the (s)-cis-conformation that is dis-
played in the carbon series C3–C2–C30–C8–C14–C15.Based
on the above results, the relative stereostructure of 1 was
elucidated as shown in Figure 1b.
À
each other through the carbon–carbon bonds of C1 C10
Acetylation of 1 with acetic anhydride in pyridine afford-
ed 3-O-acetyl xylogranatin F (Scheme 2), which was also iso-
lated from the seeds of X. granatum as a natural product,
and was named xylogranatin G (2).This corroborated the
observation that the free hydroxyl group of 1 was located at
C3.The relative configuration of C3 in 2 was established to
be the same as in 1 on the basis of the NOE interactions
H3/H5, H3/H30, H3/Me19, and H3/Me28 (see Figure S1 in
À
and C2 C3, whilst HMBC correlations H11/C8, H11/C9,
H30/C11, H30/C14, H15/C8 linked fragments 1c and 1a to-
À
À
gether by the carbon–carbon bonds of C8 C14 and C9 C11
(Figure 1a).From all these observations, the constitution of
1 was attributed to xylogranatin F as shown in Figure 1.The
relative configuration of 1 was established on the basis of
the NOESY spectrum.The significant NOE interactions
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
1131

G.Bringmann, J.Wu et al.
C26H29NO6 (i.e., larger than that of 1 by two hydrogen
atoms); this was established by HR-TOFMS spectrometry
(m/z: calcd: 474.1893; found: 474.1884 [M+Na]+).The 1H
and 13C NMR spectroscopic data (Table 1) of 3 were similar
to those of 1.This suggests that 3 might have the same basic
molecular framework as 1.However, an olefinic proton
(dH =6.57 (s), H15 in 1) and two olefinic carbons (dC =157.2
(s), 111.2 (d), C14 and C15 in 1) that were derived from the
double bond C14=C15 in 1 were absent in 3, but an aliphatic
proton spin system, H14–H215 (dH =3.15 (m), 2.88 (dd, J=
16.0, 11.5 Hz), 3.02 ppm (dd, J=16.0, 6.0 Hz); dC =41.8 (d),
36.8 ppm (t)) appeared in 3; this was deduced from
1H-1H COSY correlations, and indicated that the C14=C15
double bond of 1 was hydrogenated in xylogranatin H (3).
HMBC correlations from H14 and H215 to C8, C13, C15,
and C16 corroborated that the C14=C15 double bond of the
compound was a hydrogenated analogue of 1.The NOE in-
teraction that was observed in 3 from Me18 to H14 indicat-
ed the a configuration of H14, and the interactions between
H3/H5, H3/H6, H3/H30, H3/Me19, and H3/Me28 (see Fig-
ure S2 in the Supporting Information) established 3b-H and
the inverse 3a-OH group.The chemical transformation
from 1 to 3 by catalytic hydrogenation (10% Pd/C in etha-
nol) (Scheme 2) further confirmed the structure that is
shown in Scheme 1.
Absolute stereostructures ofxylogranatins F–H (1–3) : The
above NOE correlation studies on compounds 1–3 and
chemical transformations from compound 1 to 2 and 3 con-
firmed that the absolute configurations of compounds 1–3
were identical (Scheme 2).As shown in Scheme 1, the cen-
tral molecular portion of compounds 1–3 is a pyridine ring.
It divides each of them into two relatively isolated, NOE-in-
dependent substructures.Within one such substructure, the
stereochemical assignment of one single stereogenic element
will unambiguously establish the absolute configuration of
the entire substructure.Each of the substructures 1b of 1
and fragment 3b of 3 (Figure 2) has a secondary alcohol at
C3.Consequently, the absolute stereostructure of this chiral
center, which was the same in compounds 1–3, could be de-
termined by using the modified Mosher MTPA [a-methoxy-
a-(trifluoromethyl)phenylacetyl] ester method.[21] On the
other hand, substructure 1a of compound 1 has two main
Figure 1.a) Partial structures, and the 1H-1H COSY or HMBC correla-
tions that were used to establish the gross structure of xylogranatin F (1)
(blue arrows indicate interactions within the fragments, green ones be-
tween them).b) Significant NOE correlations for xylogranatin F ( 1).
Scheme 2.Chemical correlations for xylogranatins F–H ( 1–3).
the Supporting Information).Based on the above results,
the relative (and thus) absolute configuration of 1 and 2 was
determined to be the same.
Xylogranatin H (3), was isolated as an amorphous
powder, and was found to have a molecular formula of
Figure 2. Dd values (Dd [ppm]=[dSÀdR]) obtained for the (3S) and (3R)-
MTPA esters of xylogranatin F (1s, 1r) and H (3s, 3r).
1132
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144

Xylogranatins F–R
FULL PAPER
chiral chromophores, one is a furan ring and the other one
is the a,b-unsaturated d-lactone that is conjugated with a
pyridine ring.Both are connected through the chiral center
of C17 in substructure 1a (Figure 2).Therefore, it should be
possible to determine the absolute configuration of C17 in
portion 1a by circular dichroism (CD) analysis.
The modified Mosher method was employed to determine
the absolute configuration of C3 in 1 and 3 (Figure 2).Un-
fortunately, in xylogranatin F (1), the Dd values of a methyl
and one proton, that is, Me28 (Dd=À0.07 ppm) in substruc-
ture 1b and H11b (Dd=+0.05 ppm) in 1a, disobeyed the
Mosher rule.Other protons, however, did obey the rule.The
Dd values of all protons in substructure 1b, except for those
of Me28, were positive, while those in 1a, except for that of
H11b, were negative.This regular arrangement indicated
that the configuration of C3 in 1 might be R.In xylograna-
tin H (3), the Dd values of all protons in substructure 3b,
except for that of Me28 (Dd=À0.13 ppm) were positive,
while those in 3a were all negative.This observation sug-
gested the 3R configuration in 3.The result is consistent
with that of the above-described chemical transformation
from 1 to 3 and the NOE correlation studies on compounds
1 and 3, which had shown that the configuration at C3 in
the two compounds was the same.The irregular Dd values
of Me28 in 1 and 3 might be due to steric interactions of the
MTPA group with Me28 and other groups of these com-
pounds.[21]
After having established the absolute configuration of 1b
by the Mosher method, and the relative configuration within
1a by NOE data, two possible isomers remained: the diaste-
reomers (3R,5S,10S,13R,17R)-1 and (3R,5S,10S,13S,17S)-1.
For the determination of the absolute configuration of 1a
(and thus of the entire molecule 1), CD investigations were
performed in combination with quantum chemical CD cal-
[22,23]
culations.These calculations at the B3LYP/6–31G(d)
Figure 3.Comparison of the calculated CD spectra (top: TDB3LYP/6–
31G(d) results; bottom: DFT/MRCI/SVP results) of the remaining possi-
ble diastereomers of 1, (3R,5S,10S,13R,17R)-1 and (3R,5S,10S,13S,17S)-1,
with the measured CD curve: Only the spectrum calculated for
(3R,5S,10S,13R,17R)-1 fits with the experimental one.
level provided six minima for (3R,5S,10S,13R,17R)-1 and
four for (3R,5S,10S,13S,17S)-1 with a relevant contribution
to the overall CD spectra (Tables S1 and Table S2).[24] In
both cases, the single CD curves of these structures were
added up by following the Boltzmann statistic, and then
were UV corrected (red shift of l=10 nm for the calculated
CD spectra).[25] The comparison of the theoretical curves
with the experimental ones (Figure 3) revealed that nearly
all the calculated Cotton effects of (3R,5S,10S,13S,17S)-1 are
opposite to the ones that were experimentally obtained,
dominate the chiroptical properties of the molecule, which
is the reason why the CD spectra of the two possible diaste-
reomers are nearly mirror-image-like.Nonetheless, with
DFT/MRCI/SVP[26,27] we also used a higher level of theory
to calculate the CD spectra of the above-found conformers.
As expected, the results of these calculations clearly proved
our first assignment of the absolute configuration from the
time-dependent DFT (TDDFT) results, because only the
curve that was calculated for (3R,5S,10S,13R,17R)-1 fits very
well the experimental one.Furthermore, no unexpected
peak was observed, which corroborated our assumption that
the non-fitting peak in the TDDFT calculations is indeed an
artifact of the method that was used.In combination with
the experimental results, the calculations thus unambiguous-
ly showed that C13 and C17 of xylogranatine F (1) are both
R-configured.
whereas
the
curve
that
was
simulated
for
(3R,5S,10S,13R,17R)-1 reproduces the experimental spec-
trum in the range from l=240–400 nm very well, expect for
the peak at l=230 nm.By model calculations in which
1
was divided into two chromophores (see Figure S3 in the
Supporting Information), it could be shown that this unex-
pected peak originates from the moiety 1b, of which the ab-
solute configuration is known by applying the Mosher
method.Therefore, this peak is most probably an artifact of
the applied theoretical method, and could be neglected.
From the model calculations, we also know that the north-
eastern chromphore with the stereocenters at C13 and C17
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
1133

G.Bringmann, J.Wu et al.
Structural elucidation ofxylogranatins I–R (4–13)
:
The
dated by starting from an a,b-unsaturated d-lactone ring D,
which is characterized by the following NMR spectroscopic
data: dH =6.39 (s), 5.52 (s); dC =41.9 (s), 153.7 (s), 113.1 (d),
167.1 (s), 79.5 ppm (d). This was corroborated by the
HMBC correlations: H15/C13, H15/C14, H15/C16, H17/C13,
and H17/C16 (Figure 4a).The HMBC cross-peaks from H17
to C20, C21, and C22 indicated that the b-furyl ring is con-
nected to C17 of the d-lactone ring D.A methyl singlet reso-
nance at dH =1.27 ppm (Me18) and the protons of C12,
which show HMBC correlations to the g-C (C13) of ring D,
established the connections between CH212 and Me18 with
C13.HMBC cross-peaks from the protons of the spin
system, H211–H212 to a carboxyl carbon atom, which was
deduced from the 1H-1H COSY correlations, linked this
proton spin system with a terminal carboxyl.These results
established the substructure 4a.
A second fragment could be assembled as 4b by starting
from the proton spin system H6–H5–H10–H19, which was
deduced from 1H-1H COSY correlations.HMBC cross-
peaks from H6 to the ester carbon atom of a methoxycar-
bonyl group (dH =3.70 ppm (s); dC =52.3 (q), 176.1 ppm (s))
connected C6 with this terminal group.A proton singlet of a
methine (C3), which bears a hydroxyl resonance at dH =
4.05 ppm and two methyl singlet resonances at dH =1.05
(Me29) and 0.76 ppm (Me28), and exhibits HMBC cross-
peaks to a quaternary carbon atom (C4) and C5 of the
above proton spin system, suggested that the quaternary
carbon atom (C4) that bears two methyl groups was situated
between C3 and C5.This connection was corroborated by
the HMBC correlation from H5 to C4; this gives substruc-
ture 4b.
TOFMS of xylogranatin I (4) showed the molecular-ion
peak at m/z: 500.Its molecular formula was established to
be C27H32O9 by HR-TOFMS (m/z: calcd: 500.2046; found:
500.2052 [M]+).This together with the 1H and 13C NMR
spectroscopic data (see the Experimental Section and
Table 2) indicated the presence of five carbon–carbon
double bonds and three carbonyls, as well as four iso- or
heterocyclic rings.DEPT experiments revealed that 4 had
five methyls, three methylenes, nine methines (five of them
olefinic), and ten quaternary carbon atoms.In addition, the
NMR spectroscopic data (see the Experimental Section and
Table 2) showed the presence of a b-furyl ring (dH =6.53
(brs), 7.53 (brs), 7.67 (brs); dC =111.1 (d), 122.0 (s), 143.6
(d), 144.7 ppm (d)).Three substructures 4a (from C9, C11 to
C18, and from C20 to C23), 4b (from C3 to C7, C10, C19,
C28, and C29), and 4c (from C1 to C2, C8, and C30) were
1
determined by analysis of the 2D H-1H COSY, HSQC, and
HMBC spectra of 4 (Figure 4a).Substructure 4a was eluci-
The last fragment, 4c, was assembled as a trisubstituted
furan ring on the basis of the NMR spectroscopic data (dH =
7.08 ppm (s); dC =159.8 (s), 123.0 (s), 116.7 (d), 149.5 ppm
(s)) (Table 2) and HMBC correlations from H30 to C1, C2,
and C8.Moreover, the HMBC cross-peaks Me19/C1, H3/
C2, H3/C30, connected substructures 4b and 4c through the
À
À
carbon–carbon bonds of C1 C10 and C2 C3, whilst HMBC
correlations H30/C14, H15/C8, linked fragments 4a and 4c
À
together through the interaction of C8 C14.From all these
observations, the planar structure of 4 was identified as
shown in Figure 4.
The relative stereostructure of 4 was established on the
basis of the NOESY spectrum.The significant NOE interac-
tions observed in 4 (Figure 4b) H3/H5, H3/H30, H3/Me29,
H5/H6b, H5/Me19, indicated they are cis to each other.The
above NOE observations also helped to establish 3b-H and
the inverse 3a-hydroxyl group.Similarly, the interactions be-
tween H17/H22, H17/H12b, H17/H15, Me18/H30, Me18/
H21, Me18/H15, H30/H15, suggested that Me18 is a, H17 is
b, and the b-furyl ring at C17 is therefore a.The observed
NOE correlations between Me18 and H17, although seem-
ingly not in agreement with the bisaxial array in the global
minimum that was found by the DFT calculations, can, how-
ever, be explained by one of the other conformers, in which
these two substituents adopt a bisequatorial orientation; this
leads to a close proximity of these two spin systems
Figure 4.(a) Partial structures, 1H-1H COSY and HMBC correlations that
are indicative of the gross structure of xylogranatin I (4) (blue arrows in-
dicate interactions within the fragments, green ones between them).
(b) Significant NOE correlations for xylogranatin I (4); to simplify the
structure, the CO2Me group at C6 is omitted).
1134
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144

Xylogranatins F–R
FULL PAPER
Table 2. 13C NMR (125MHz) data for xylogranatins I–R (4–12 in [D4]methanol and 13 in CDCl3).[a]
Position
4
5
6
7
8
9
10
11
12
13
1
2
3
4
5
6
7
8
159.8 (s)
123.0 (s)
72.9 (d)
40.2 (s)
42.7 (d)
34.5 (t)
176.1 (s)
149.5 (s)
176.4 (s)
34.5 (d)
30.1 (t)
160.4 (s)
120.9 (s)
82.3 (d)
40.2 (s)
43.2 (d)
34.5 (t)
176.1 (s)
149.3 (s)
174.9 (s)
34.8 (d)
30.1 (t)
160.3 (s)
120.8 (s)
82.3 (d)
40.2 (s)
43.2 (d)
34.5 (t)
176.1 (s)
149.2 (s)
176.5 (s)
34.8 (d)
30.2 (t)
160.2 (s)
121.5 (s)
80.7 (d)
40.2 (s)
43.4 (d)
34.6 (t)
176.2 (s)
149.3 (s)
176.8 (s)
34.8 (d)
30.5 (t)
161.1 (s)
119.9 (s)
75.0 (d)
39.1 (s)
43.6 (d)
34.3 (t)
176.0 (s)
149.7 (s)
174.8 (s)
34.5 (d)
30.1 (t)
161.0 (s)
120.1 (s)
74.8 (d)
39.2 (s)
43.7 (d)
34.2 (t)
161.1 (s)
120.1 (s)
75.0 (d)
39.5 (s)
43.7 (d)
34.1 (t)
161.0 (s)
120.0 (s)
74.8 (d)
39.3 (s)
43.7 (d)
34.2 (t)
175.9 (s)
149.8 (s)
174.9
158.9 (s)
125.2 (s)
81.6 (d)
43.8 (s)
54.7 (d)
84.3 (d)
174.3 (s)
150.6 (s)
175.9 (s)
40.8 (d)
30.5 (t)
200.5 (s)
129.3 (s)
159.2 (d)
36.9 (s)
45.6 (d)
34.7 (t)
173.6 (s)
197.8 (s)
175.5 (s)
43.1 (d)
29.8 (t)
175.9 (s)
149.7 (s)
176.0 (s)
149.7 (s)
[b]
[b]
9
–
–
10
11
12
13
14
15
16
17
18
19
20
21
22
23
28
29
30
7-OMe
9-OMe
3-R2
1’
34.4 (d)
30.8
34.4 (d)
34.5 (d)
[b]
[b]
–
–
–
–
[b]
[b]
[b]
33.1 (t)
41.9 (s)
33.1 (t)
33.1 (t)
42.0 (s)
33.3 (t)
42.0 (s)
33.2 (t)
–
33.1 (t)
42.0 (s)
29.8 (t)
42.0 (s)
153.7 (s)
113.4 (d)
167.1 (s)
79.6 (d)
21.4 (q)
16.9 (q)
122.0 (s)
143.7 (d)
111.0 (d)
144.8 (d)
20.0 (q)
25.1 (q)
117.2 (d)
52.4 (q)
52.3 (q)
41.9 (s)
153.4 (s)
114.0 (d)
166.9 (s)
79.5 (d)
21.4 (q)
16.6 (q)
122.0 (s)
143.7 (d)
111.0 (d)
144.8 (d)
19.7 (q)
24.5 (q)
116.8 (d)
52.4 (q)
52.3 (q)
42.4 (s)
42.0 (s)
42.0 (s)
41.5 (s)
158.8 (s)
124.6 (d)
163.4 (s)
78.6 (d)
20.4 (q)
11.8 (q)
119.4 (s)
141.6 (d)
109.8 (d)
143.5 (d)
20.2 (q)
28.2 (q)
41.4 (t)
153.7 (s)
113.1 (d)
167.1 (s)
79.5 (d)
21.5 (q)
16.7 (q)
122.0 (s)
143.6 (d)
111.1 (d)
144.7 (d)
20.0 (q)
25.0 (q)
116.7 (d)
52.3 (q)
153.9 (s)
113.3 (d)
167.1 (s)
79.5 (d)
21.4 (q)
16.9 (q)
122.0 (s)
143.6 (d)
111.1 (d)
144.7 (d)
20.0 (q)
25.1 (q)
117.2 (d)
52.3 (q)
154.0 (s)
113.2 (d)
167.2 (s)
79.6 (d)
21.5 (q)
16.9 (q)
122.1 (s)
143.6 (d)
111.1 (d)
144.7 (d)
20.1 (q)
25.2 (q)
117.0 (d)
52.3 (q)
153.8 (s)
113.7 (d)
167.0 (s)
79.7 (d)
21.6 (q)
16.8 (q)
122.1 (s)
143.7 (d)
111.1 (d)
144.7 (d)
19.7 (q)
24.8 (q)
116.8 (d)
52.4 (q)
153.7 (s)
113.7 (d)
167.1 (s)
79.6 (d)
21.5 (q)
16.8 (q)
122.0 (s)
143.7 (d)
111.1 (d)
144.7 (d)
19.6 (q)
24.7 (q)
116.9 (d)
52.4 (q)
153.9 (s)
113.7 (d)
167.1 (s)
79.7 (d)
21.5 (q)
16.8 (q)
122.1 (s)
143.7 (d)
111.1 (d)
144.7 (d)
19.7 (q)
24.6 (q)
116.8 (d)
52.4 (q)
153.8 (s)
113.5 (d)
167.1 (s)
79.5 (d)
21.4 (q)
17.5 (q)
121.9 (s)
143.6 (d)
111.1 (d)
144.7 (d)
22.4 (q)
26.3 (q)
115.2 (d)
52.8 (q)
52.0 (q)
57.5 (q)
57.5 (q)
66.2 (t)
15.8 (q)
172.7 (s)
21.0 (q)
177.8 (s)
42.8 (d)
27.9 (t)
11.9 (q)
17.1 (q)
169.1 (s)
129.8 (s)
139.2 (d)
12.2 (q)
14.5 (q)
178.3 (s)
35.4 (d)
19.4 (q)
19.2 (q)
2’
3’
4’
5’
[a] Multiplicites are shown in parentheses and d values are in ppm.[b] Not detected.
(Figure 5).The same NOE correlations with similar intensi-
The important NOE interaction that was observed in 4
ties had previously been observed in a closely related struc-
ture, xyloccensin O, the trans configuration of Me18 and
H17 of which was further supported by X-ray diffraction.[28]
was Me19/H15; this suggested that the configuration of
Me19 is b, which was simultaneously the key NOE juncture
between the substructures 4a and 4b.Based on the above
results, the relative structure of 4 was elucidated as shown in
Figure 4b.
The absolute configuration of 4 was established by the ap-
plication of the same two-step method that was used for xy-
logranatins F–H above.Thus, the absolute configuration of
the secondary alcohol that bears center C3 in 4b (Figure 4a)
was determined by using the modified Mosher MTPA ester
method.The chiral center at C17 in substructure 4a (Fig-
ure 4a) was again established by comparing the experimen-
tal CD spectrum of 4 with the spectra that was predicted by
quantum chemical calculations for each of the two remain-
ing possible diastereomers and their enantiomers.
Owing to the low reactivity of the secondary hydroxyl
group at C3 of 4 towards MTPACl, it was difficult to obtain
the (3S)- and (3R)-MTPA esters of 4.Thus, treatment of
4
with (R)- and (S)-MTPACl in dichloromethane solvent, with
a mixture of diethylamine, triethylamine, and (dimethylami-
no)pyridine (DMAP) as a catalyst, did not yield the desired
esterified products, but the 9-diethylamide 4’ instead
(Figure 6).To our surprise, when this derivative was treated
Figure 5.Two main conformations of ring D in xylogranatins I–Q ( 4–12)
with H17 and Me18 in bisaxial (left) and bisequatorial (right) orienta-
tion; this explains the observed NOE correlations between H17 and
Me18, although they are trans configured.For reasons of clarity, only
rings D and E are shown (cut at C30 and the furan oxygen).
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
1135

G.Bringmann, J.Wu et al.
(red shift of l=8 nm for the calculated CD spectra) the
comparison of the calculated spectra with the experimental
one showed only one match.The simulated CD spectrum of
4B has only one Cotton effect at l=350 nm that fits to the
experimental curve, one at l=230 nm that is slightly shifted
in comparison to the experiment, and two effects at l=240
and 260 nm which are opposite to the experiment
(Figure 7).Because ent-4B is the mirrored spectrum of 4B,
Figure 6. Dd values [Dd (ppm)=(dSÀdR)] that were obtained for the (3S)
and (3R)-MTPA esters (4’s, 4’r) of the 9-diethylamide derivative of xy-
logranatin I.
with (R)- and (S)-MTPACl in pyridine, with DMAP at room
temperature, its 3-(S) and 3-(R)-MTPA esters (4’s, 4’r) were
formed in high yields; this permitted us to assign the abso-
lute configuration of C3 in 4.
As shown in Figure 6, the Dd values of H5, H6, H10, H19,
H28 were positive, while those of H17, H18, H21, H22, and
H30 were negative.This regular arrangement provided evi-
dence for the 3R configuration in 4’.Because the absolute
stereostructure of 4 must be the same as that of its deriva-
tive 4’, the absolute configuration of C3 in 4 was then as-
signed as R too.Based on this 3 R configuration, the chirality
of C10 in 4 was also identified as R through NOE investiga-
tions, which led to 19bMe.This result was consistent with
the NOE interaction observed earlier in 4 from Me19 to
H15.
Me29 and H15 in 4’, by contrast, disobeyed Mosherꢁs rule.
The steric interactions with the MTPA group from Me29
and other groups inside 4’ with that in xylogranatin F and H
could explain why Me29 disobeys this rule,[21] and the differ-
À
ent rotation effects on the C8 C14 bond that is induced by
the (3S) and (3R)-MTPA esters might explain why H15 dis-
obeys Mosherꢁs rule.This might be a consequence of a
change of the dihedral angle between the central furan ring
and the a,b-unsaturated d-lactone according to the different
substituents at 3-OH.In 4’s and 4’r, this dihedral angle was
different.Thus the shielding and deshielding effects from
the benzene ring of MTPA esters to H15 was not as regular
as those of the other protons in 4’.
Because the Mosher method did not lead to an unambigu-
ous result in this case, the experimental CD spectrum of 4
had to be compared with the spectra that were computed
for all four possible diastereomers of 4, which resulted from
the combination of all enantiomeric forms of the substruc-
tures with the relative configurations that were established
by the NOE measurements above.For this purpose,
TDDFT calculations of the twelve energetically lowest con-
formers of 4A (3R,5R,10R,13R,17R-4) and of the twelve en-
ergetically lowest ones of 4B (3R,5R,10R,13S,17S-4) were
carried out (see also Tables S3 and S4 in the Supporting In-
formation).The CD spectra of the respective enantiomers,
Figure 7.Attribution of the absolute configuration of xylogranatin I ( 4)
by comparison of the CD spectra that were calculated for 4A
(3R,5R,10R,13R,17R-4) and 4B (3R,5R,10R,13S,17S-4) and their enantio-
mers ent-4A (3S,5S,10S,13S,17S-4) and ent-4B (3S,5S,10S,13R,17R-4) with
the experimental curve: The spectrum that was calculated for ent-4A is
completely opposite, while the one that was simulated for 4A fits with
the experimental one.
ent-4A
(3S,5S,10S,13S,17S-4)
and
ent-4B
(3S,5S,10S,13R,17R-4), were obtained by reflecting at the
zero line of the spectra that was simulated for 4A and 4B,
respectively.After Boltzmann weighting and UV shifting
1136
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144

Xylogranatins F–R
FULL PAPER
the peaks at l=230 and 350 nm are opposite and the peak
at 260 nm coincides with the experiment.Consequently, nei-
ther 4B nor ent-4B corresponds to the absolute stereostruc-
ture of natural xylogranatin I (4).
Xylogranatin K (6) had the molecular formula C28H34O9
as revealed by the HR-TOFMS spectrum (m/z: calcd:
537.2101; found 537.2112 [M+Na]+).The NMR spectro-
scopic data of 6 were similar to those of 4, except for the
presence of one additional O-methyl group (dH =3.35 ppm
(s); dC =57.5 ppm (q)). The HMBC correlation from H3
(d=3.63 ppm (s)) of 6 to the carbon atom of this methoxy
As Figure 7 shows, the computed CD spectrum of 4A,
however, does fit well to the experimental one, whereas the
spectrum of ent-4A is completely opposite compared with
the experiment.Therefore, xylogranatine I ( 4) has the same
absolute all-R configuration as displayed in structure 4A.
Despite the fact that some protons in 4 disobeyed Mosherꢁs
rule (as mentioned above), the assignment of the absolute
configuration of the southwestern moiety of 4 was correct
and furthermore the calculations clearly showed the R con-
figuration of C13 and C17.
The molecular formula of xylogranatin J (5) was deter-
mined to be C29H36O9 by HR-TOFMS (m/z: calcd: 551.2257;
found: 551.2237 [M+Na]+).It was larger than that of xylog-
ranatin I (4) by a C2H4 unit.The NMR spectroscopic data of
5 were similar to those of 4, except for the presence of two
more methoxy groups (dH =3.63 (s), dC =52.3 (q); dH =3.38
(s), dC =57.5 ppm (q)). HMBC correlations from the protons
(dH =3.63 ppm (s)) of one methoxy group to C9 of 5 and H3
(d=3.66 ppm (s)) of 5 to the carbon atom of another me-
thoxyl (dC =57.5 (q)) revealed that the above two methoxy
group indicated that it was attached to C3.Treatment of
6
with (trimethylsilyl)diazomethane (TMSCHN2) afforded xy-
logranatin J (5) (Scheme 2).Thus, xylogranatin K ( 6) was as-
signed as the 3-O-methyl analogue of xylogranatin I (4),
that is, the free carboxylic acid of the methyl ether of xylog-
ranatin J (5).
The HR-TOFMS spectrum (m/z: calcd: 551.2257; found:
551.2268 [M+Na]+) of xylogranatin L (7) showed that this
compound had the same molecular formula as 5.The NMR
spectroscopic data of 7 were similar to those of 4, except for
the presence of an ethoxy group (dH =1.20 (t, J=7.0 Hz),
3.52 (m), 3.67 (m); dC =15.8 (q), 66.2 ppm (t)]. The HMBC
correlation from H3 (dH =3.76 ppm (s)) of 7 to the methyl-
ene carbon atom of this ethoxy group suggested that it was
attached to C3.Treatment of 4 with bromoethane and
sodium hydride in dichloromethane afforded 7 (Scheme 3).
The configuration at C3 in 7 was determined to be the same
as in 4 by NOE interactions that were observed in 7 from
H3 to H5.Based on the above results, xylogranatin L ( 7)
was elucidated as the 3-O-ethyl analogue of xylogranatin I
(4).
groups were substituted at C9 and C3 of 5.Treatment of
with 30% hydrochloric acid in methanol, followed by
18 HPLC purification, afforded the major product 5 (yield:
4
C
>90%) (Scheme 3) and the minor one 5’ (yield: <6%) (see
Figure S4 in the Supporting Information).By careful NOE
studies, it was found that the configuration of C3 in 5 was
the same as in 4, but that in 5’ was opposite (see Figures S5
and S6 in the Supporting Information).Thus, xylogranatin J
(5) was concluded to be the 3-O-methyl-9-methylester ana-
logue of xylogranatin I (4).
The molecular formula of xylogranatin M (8) was estab-
lished as C30H36O10 by HR-TOFMS (m/z: calcd: 579.2206;
found: 579.2214 [M+Na]+).The NMR spectroscopic data of
8 were similar to those of 5, except for the absence of a 3-
methoxy group (dH =3.38 ppm (s); dC = 57.5 ppm (q) in 5)
and presence of one more acetoxy group (dH = 2.07 ppm
Scheme 3.Chemical correlations for xylogranatins I–P ( 4–11) (xyl.=xylogranatin).
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
1137

G.Bringmann, J.Wu et al.
(s); dC =21.0 (q), 172.7 ppm (s) in 8).This result was further
corroborated by the HMBC correlation from H3 (d=
5.31 ppm (s)) of xylogranatin M (8) to the carbonyl carbon
atom of this acetoxy group.Treatment of xylogranatin I ( 4)
with acetic anhydride in pyridine, followed by methylation
of the acetylated product with (trimethylsilyl)diazomethane,
gave xylogranatin M (8) (Scheme 3).The absolute configu-
ration of C3 in 8 was determined to be the same as in 4 by
NOE interactions that were observed from H3 to H5.
Therefore, the structure of xylogranatin M (8) was identified
as the 3-O-acetyl-9-methylester analogue of xylogranatin I
(4).
Similiar 1D and 2D NMR spectroscopy investigations es-
tablished the structures of xylograntines N–P (9–11) as seen
in Scheme 1 (for more details see the Experimental Section
and the Supporting Information).
The molecular formula of xylogranatin Q (12) was deter-
mined to be C27H30O9 by HR-TOFMS (m/z: calcd:
521.1788; found: 521.1796 [M+Na]+).The 1H and 13C NMR
spectroscopy data (see Experimental Section and Table 2)
indicated the presence of five carbon–carbon double bonds,
three carbonyls, and five rings.By detailed analysis of the
1
2D H-1H COSY, HSQC, and HMBC spectra of xylograna-
tin Q (12), it was found that this compound could be divided
into three substructures (Figure 8a), namely, 12a (from C9,
C11 to 18 and from C20 to C23), 12b (from C3 to C7, C10,
C19, C28, and C29), and 12c (a furan ring that comprises
C1, C2, C8, and C30).Although 12a and 12c were the same
as 4a and 4c of xylogranatin I (4), the substructure 12b was
distinctly different from that of 4b.The presence of two
oxygenated methines C3 (dH =4.51 ppm; dC =81.6 ppm) and
C6 (dH =4.21 ppm (s); dC =84.3 ppm (d)) in 12b was sug-
gested by their 1H and 13C NMR spectroscopic data.The
HMBC correlations between H6/C3 and H3/C6 indicated
that these methines were connected by an oxygen bridge.
The presence of this oxygen bridge was consistent with the
above-described unsaturation analysis of the molecular for-
mula of xylogranatin Q (12).From the view of biosynthesis,
xylogranatin Q (12) should be derived from xylogranatin I
(4), by oxidation of C6, and subsequent dehydrating ether
formation between C3 and C6 to form a tetrahydrofuran
ring (see below).
The relative configurations within 12 were established on
the basis of the NOESY spectrum.The significant NOE in-
teractions between H3/H6, H3/H30, H3/Me28, H3/Me29,
H5/H6, H5/Me19 and Me19/Me28 that were observed in 12
(Figure 8b) indicated the relative orientation of substruc-
tures 12b and 12c to each other as shown in Figure 8b.Simi-
larly, the interactions between H17/H12, H17/H15, H17/
H22, Me18/H15, Me18/H21, Me18/H30, and H30/H15 sug-
gested that Me18 is a, H17 is b, and therefore that the 17-
substituted b-furyl ring is a.Further, the NOE interactions
of Me19/H15 and H5/H15 are an additional indication of
the spatial proximity of the substructures 12a and 12b.
The absolute configuration of 12 was evidenced by CD
analysis.The CD spectrum of 12 was taken in methanol of
chromatographic grade, and it exhibited two positive and
Figure 8.(a) Substructures of xylogranatin Q ( 12) (blue arrows indicate
interactions within the fragments, green ones between them).(b) Signif-
icant NOE correlations for xylogranatin Q (12).
one negative Cotton effects at l=219 (De=+13.0), 264
(+3.2), and 326 nm (À3.2 cm2 molÀ1), which were similar to
those of xylogranatin I (4) (l=219 (De=+8.6), 264 (De=
+3.4), and 326 nm (De=À3.1 cm2 molÀ1); see Figure S8 in
the Supporting Information).The chromophores in 12 are
similiar to those in 4; ring B is the main chromophore for
the southwestern moiety, and ring D and E are the main
chromophores for the northeastern part.Despite the fact
that 12 has an additional ring (viz., ring F), the chromo-
phore of the southwestern moiety is nearly the same as in 4
and has an almost identical stereochemical orientation, as
the substituents at C3 and C5 of ring A in the main con-
formers of xylogranatin I (4) both adopt an axial array, as
was also found for ring A of xylogranatin Q (12).The only
difference is the bond between O3 and C6 in 12.Although
this bond generates an additional stereogenic center in the
southwestern moiety, this center is far away from the main
chromophore and thus should not give an additional Cotton
effect in the CD spectrum.For these reasons, the nearly
1138
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144

Xylogranatins F–R
FULL PAPER
identical CD spectra of 12 and 4 indicate that 12 should
indeed have the same absolute configuration as 4, and the
configuration of the additional stereocenter at C6 can be de-
termined from the NOE investigations.
atoms C1–C5 and C10 comprised a cyclohex-2-enone ring.
Together this gave the substructure 13b, which was the same
as that in xylogranatin C,[20], which was isolated from the
seeds of the same plant.The remaining fragment, 13c, was
assembled as a methylene connected with a ketone on the
basis of the NMR spectroscopic data: dH =3.45 (d, J=
16.8 Hz), 3.75 ppm (d, J=16.8 Hz); dC =41.4 (t), 197.8 ppm
(s) (Tables 2 and Experimental Section).The connections
between the above fragments were elucidated by HMBC
correlations between the protons and carbon atoms that
belong to different substructures.The observed HMBC
cross-peaks H30/C1, H30/C2, H30/C3, and H3/C30 connect-
ed the substructures 13b and 13c through the carbon–
Xylogranatin R (13) was isolated as an amorphous
powder.Its molecular formula was determined to be
C27H32O9 by HR-TOFMS (m/z: calcd: 523.1944; found:
523.1956 [M+Na]+).The 1H and 13C NMR spectroscopic
data (Table 2 and Experimental Section) indicate that the
molecule has four carbon–carbon double bonds, five carbon-
yls, and three rings.By a detailed analysis of the 2D
1H-1H COSY, HSQC, and HMBC spectra of xylogranatin R
(13), it was found that 13 could be divided into three sub-
structures (Figure 9), namely, 13a (from C9, C11 to 18 and
À
carbon bond of C2 C30, while the HMBC correlations
H230/C14 and H15/C8, linked fragments 13c and 13a to-
À
gether through that of C8 C14 (Figure 9).
The relative stereostructure of 13 was established by anal-
ysis of its NOESY spectrum.The significant NOE interac-
tions, H17/H12a, H17/H12b, H17/H15, H17/H22, Me18/H21,
and Me18/H15 suggested a 17b-H and a 18a-Me orientation.
The correlations H30b/H3, H30b/H5, H30b/Me19, and
Me19/Me28 indicated that H5 and Me19 of substructures
13b are b-configured (see Figure S7 in the Supporting Infor-
mation).
Absolute configuration of xylogranatins F–R (1–13): As
mentioned earlier, the absolute stereostructures of xylogra-
natins F (1) and I (4) were determined by application of the
modified Mosher MTPA ester method coupled with circular
dichroism quantum chemical calculations.The successful
chemical conversions of 1 to 2 and 3 (Scheme 2), combined
with the NOE correlation studies on the stereoarray at C3
in these compounds confirmed that they all possess the
same configurations at their chiral centers at C3, C5, C10,
C13, and C17, which were determined to be R, S, S, R, and
R, respectively.The configuration at C14 in 3 was deduced
as S by the NOE interaction from H14 to Me18.That com-
pounds 4–11 had the same R configurations at C3, 5, 10, 13,
and 17, was corroborated by the successful chemical trans-
formations of 4 and 6 to 5, of 4 to 7 and to 8, and of 9–11 to
8 (Scheme 3).Moreover, the absolute configuration at C2 in
the 2-methylbutyryl group of 9 was determined to be S by
the aD value of its acid and the E-configured double bond in
the tigloyl group (2-methylcrotonyl) of 10 was suggested by
its NOE correlation studies.Furthermore, the chiral center
at C17 of 12 was identified as the same as that of 4–11 by
analysis of its CD spectrum (see Figure S8 in the Supporting
Information), while others, including C3, C5, C6, C10, and
C13 were suggested to be R, S, S, R, and R, respectively, by
their NOE interactions.Although some of the NOE interac-
tions were quite weak and not as significant as in the case of
the xylogranatins F–P, nonetheless, this should be the most
probable stereochemical array due to the joint biogenetic
origin of 12 compared to those of 4–11.Likewise, for biosyn-
thetic reasons, we anticipate that in the case of xylograna-
tin R (13) the stereogenic centers at C5, C6, C13, and C17
should have the R configuration.
Figure 9.Partial structures, 1H-1H COSY and HMBC correlations that
were used to establish the structure of xylogranatin R (13) (blue arrows
indicate interactions within the fragments, green ones indicate interac-
tions between the fragments).
from C20 to C23), 13b (from C1 to C7, C10, C19, C28, and
C29) and 13c (from C8 and C30), among which 13a and
13b were the same as 12a and the corresponding substruc-
ture of xylogranatin C,[20] respectively.The substructure 13b
could be assembled by starting from a proton spin system,
H6–H5–H10–H319, which was deduced from 1H-1H COSY
correlations.HMBC cross-peaks from H6 to the ester
carbon of a methoxycarbonyl group (dH =3.70 ppm (s); dC =
52.0 (q), 173.6 ppm (s)) connected C6 with this terminal
group.A proton singlet of an olefinic methine (C3) reso-
nance at dH =6.54 ppm and two methyl singlet resonances at
dH =1.19 (Me29) and 1.11 ppm (Me28), which exhibit
HMBC correlations to a quaternary carbon atom (C4) and
C5 of the above-described proton spin system suggested that
the quaternary-carbon atom (C4) that bears two methyl
groups was situated between C3 and C5.On the other hand,
HMBC cross-peaks from the olefinic proton, H10 and
Me19, to a conjugated ketone (dC =200.5 ppm (s)) indicated
that this ketone was situated between C2 and C10.Based on
the above observations, it was concluded that the carbon
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
1139

G.Bringmann, J.Wu et al.
Possible joint biosynthetic origin ofxylogranatins F–R (1–
13): Xylogranatins F–R (1–13) constitute a novel type of
natural products that consist of 26 carbon atoms in their
showed that its AFC50 (concentration for 50% antifeedant
activity) values at 24 and 48 h were 0.31 and 0.30 mgmLÀ1,
respectively.
skeletons.Among them, the xylogranatins F–Q
( 1–12)
Table 3.Antifeedant bioassay results for xylogranatins C–D, F–G ( 1–2),
I–K (4–6), and P–R (12–13).
belong to two subclasses of tetranortriterpenoids, one of
which, compounds (1–3), comprises a pyridine as the B-ring
of tetranortriterpenoid, while the other subclass (4–12) con-
tains a central furan core.A plausible biogenetic pathway
for xylogranatins F–Q (1–12), with xylogranatin R (13) as a
key biosynthetic intermediate, is proposed in Scheme 4.The
biogenetic origin of them might be a mexicanolide,[4] the
Compound
Concn [mgmLÀ1
]
Antifeedant rates at different ex-
posure time (meanÆSD%)
24 h
48 h
72 h
1
2
4
5
6
12
13
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
50.0Æ5.8 55.2Æ7.1 57.7Æ10.2
74.1Æ8.7 79.9Æ6.3 73.8Æ3.5
74.5Æ4.2 59.0Æ2.5 43.3Æ2.3
38.3Æ3.1 45.4Æ3.7 34.9Æ2.7
À
retro-aldol cleavage of the C9 C10 bond of which could
6.4Æ4.8
28.5Æ8.8 32.3Æ3.6
lead to the opening of its B ring, thus generating xylograna-
45.0Æ4.0 58.5Æ8.1 44.2Æ4.4
70.6Æ2.4 58.6Æ5.3 47.5Æ7.8
48.3Æ10.3 46.9Æ2.1 51.5Æ5.2
53.3Æ7.4 57.3Æ4.7 55.7Æ9.3
88.3Æ1.2 95.6Æ2.3 100.0
À
tin C.Further cleavage of the C8 C9 bond of xylograna-
xylogranatin C
xylogranatin D
toosendanin[a]
tin C, with loss of its 30-acetoxy group, would lead to the
key intermediate, xylogranatin R (13), which, as a 1,4-dike-
tone could cyclize to produce a hypothetical heterocyclic in-
termediate, intA, and then aromatize to generate the furan
derivative xylogranatin I (4).Different substitution reactions
at C3 and at the 9-carboxylate group of xylogranatin I (4)
would produce xylogranatins J–P (5–11).While oxidation of
C6 of xylogranatin I (4) and cyclic ether formation with C3
could generate the complex structure of xylogranatin Q
(12), oxidation at C10 with g-lactonization and transforma-
tion of the C9 carboxyl group into a nitrile function would
result in the pentacyclic compound intB.By starting from
this nitrogen-containing compound, the central pyridine ring
of 1–3 might be generated by two possible subpathways:
either through a direct hetero-Diels–Alder-related ring clo-
sure to give intD and subsequent hydrogenation to give
intF, or by reduction, followed by the Diels–Alder ring clo-
sure at the level of the imine intC to give intE, and final
elimination of water via intF, thus completing the biosynthe-
sis of xylogranatin F (1), which could then be transformed
to xylogranatins G (2) and H (3) by acetylation of its 3-OH
group and hydrogenation of its C14–C15 double bond, re-
spectively.
podophyllotoxin[a] 1.0
53.2Æ2.7 57.2Æ4.7 68.4Æ3.3
[a] Antifeedant standards.
Xylogranatin G (2) is the 3-O-acetyl derivative of xylogra-
natin F (1).The introduction of this O-acetyl group at the 3-
hydroxyl of xylogranatin F (1) enhanced the antifeedant
rate significantly (16 to 25%) (Table 3).Thus, the SAR
study on the C3 analogues of xylogranatin F (1) might gen-
erate even more potent antifeedant leads.The synthesis of
these analogues for antifeedant evaluation is in progress.
Conclusion
Xylogranatins F–R, from the seeds of the Chinese man-
grove, Xylocarpus granatum, were isolated and identified as
a class of limonoids with a new carbon skeleton.Xylograna-
tins F–Q are the first aromatic B-ring limonoids to be found
in nature.They belong to two new subclasses, one comprises
a pyridine portion as the B ring of tetranortriterpenoid and
the other one contains a central furan ring.Xylogranatins C
and R are key biosynthetic intermediates, while xylograna-
tin D clearly is an artifact.The structures of these limonoids
suggest a new biogenetic pathway to tetranortriterpenoids.
Xylogranatins F, G and R exhibited pronounced antifeedant
activity against the third instar larvae of Mythimna separata
(Walker) at a concentration of 1 mgmLÀ1.The most potent
Antifeedant activity of xylogranatins: The new compounds,
xylogranatins C and D[20] (not shown), F–G (1, 2), I–K (4–
6), and P–R (12, 13), were tested for their antifeedant activi-
ties by using a conventional leaf disk method against the
third instar larvae of Mythimna separata (Walker)
(Table 3).[29] Compounds 1, 2, 4, 13, and xylogranatin D ex-
hibited marked antifeedant activity at a concentration of
1 mgmLÀ1.The antifeedant rates of these compounds at the
exposure times of 24, 48, and 72 h were over 50% in all
cases.Among the above-mentioned five compounds, 2, 4
and 13 proved to be the most potent ones; they provided an-
tifeedant rates of over 70% at the exposure time of 24 h.
The most potent compound tested was xylogranatin G (2).
Its antifeedant rate was 74–80% and was stable at different
exposure times, but those of 4 and 13 decreased with longer
exposure times.As shown in Table 3, the antifeedant activity
of xylogranatin G (2) is much stronger than that of one stan-
dard, podophyllotoxin,[30,31] but it is slightly weaker than that
compound tested was xylogranatin G.Its AFC (concentra-
50
tion for 50% antifeedant activity) values at 24 and 48 h
were 0.31 and 0.30 mgmLÀ1, respectively.This study demon-
strates that X. granatum is a rewarding new source for the
production of limonoids with novel carbon frameworks.
Experimental Section
General procedures: Optical rotations were recorded on a Polaptro-
nic HNQW5 automatic high-resolution polarimeter (Schmidt & Haensch,
Berlin, Germany) and on a J-715 spectropolarimeter (JASCO, Gross-
Umstadt, Germany) at room temperature by using a 0.1 cm standard cell
of another standard, toosendanin.[32]
A detailed study
1140
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144

Xylogranatins F–R
FULL PAPER
Scheme 4.Proposed biogenetic pathway to xylogranatins F–R (xyl =xylogranatin).
and spectrophotometric grade MeOH, and are reported in De values
(cm2 molÀ1) at the given wavelength l (nm).UV spectra were obtained
on a Beckman DU-640 UV spectrophotometer.IR spectra were recorded
on a Bruker Vertex-70 FTIR spectrophotometer.ESI-MS and TOF-MS
spectra were obtained on Bruker APEX II and BiflexIII MALDI-TOF
mass spectrometers, respectively (positive or negative mode).NMR spec-
tra were recorded in CDCl3 or [D4]methanol by using a Bruker AV-500
1
spectrometer (500 MHz for H NMR and 125 MHz for 13C NMR spectra)
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
1141

G.Bringmann, J.Wu et al.
with tetramethylsilane as the internal standard.Preparative HPLC was
carried out on ODS columns (250”10 mm i.d., YMC) with a Waters 996
photodiode array detector.For column chromatography, silica gel (200–
300 mesh) (Qingdao Mar.Chem.Ind). was used.( R)-, (S)-MTPACl and
(dimethylamino)pyridine (DMAP) were purchased from Sigma–Aldrich.
J=6.7 Hz, H19), 1.89 (dt, J=14.5, 4.0 Hz, H12a), 2.14 (dt, J=16.0, 4.0,
H11a), 2.33 (m, H6a), 2.36 (m, H5), 2.38 (m, H12b), 2.46 (dt, J=16.0,
5.0 Hz, H11b), 2.54 (m, H6b), 2.56 (m, H10), 3.35 (s, 3-OCH3), 3.63 (s,
H3), 3.68 (s, 7-OCH3), 5.52 (s, H17), 6.39 (s, H15), 6.54 (brs, H22), 7.14
(s, H30), 7.53 (brs, H23), 7.67 ppm (brs, H21); 13C NMR (see Table 2);
IR (film): n˜ =3434, 2939, 2374, 1710, 1631, 1583, 1385, 1270, 1164, 1089,
1028, 874, 804, 603 cmÀ1; UV (MeOH): lmax (loge)=212 (4.09), 335 nm
(4.19); HRTOF-MS: m/z: calcd for C28H34O9Na: 537.2101; found:
537.2112 [M+Na]+.
Xylogranatin L (7): Amorphous powder, [a]2D5 =À37 (c=0.12 in MeOH);
1H NMR (CD3OD): d=0.77 (s, H28), 1.09 (s, H29), 1.20 (t, J=7.0 Hz, 3-
OCH2CH3), 1.30 (s, H18), 1.36 (d, J=6.7 Hz, H19), 1.93 (dt, J=14.5,
4.0 Hz, H12a), 2.16 (dt, J=16.0, 4.0 Hz, H11a), 2.36 (m, H6a), 2.39 (m,
H12b), 2.40 (m, H5), 2.47 (dt, J=16.0, 5.0 Hz, H11b), 2.57 (m, H6b), 2.60
(m, H10), 3.52 (m, 3-OCH2CH3), 3.67 (m, 3-OCH2CH3), 3.71 (s, 7-
OCH3), 3.76 (s, H3), 5.55 (s, H17), 6.41 (s, H15), 6.57 (brs, H22), 7.13 (s,
H30), 7.56 (brs, H23), 7.70 ppm (brs, H21); 13C NMR (see Table 2); IR
(film): n˜ =3435, 2973, 2376, 1715, 1632, 1583, 1385, 1271, 1165, 1086,
1027, 875, 802, 603 cmÀ1; UV (MeOH): lmax (loge)=212 (4.15), 336 nm
(4.21); HRTOF-MS: m/z: calcd for C29H36O9Na: 551.2257 [M+Na]+;
found: 551.2268.
Xylogranatin M (8): Amorphous powder, [a]2D5 =À52 (c=0.15 in MeOH);
1H NMR (CD3OD): d=0.87 (s, H28), 1.04 (s, H29), 1.27 (s, H18), 1.36 (d,
J=6.7 Hz, H19), 1.94 (dt, J=14.5, 4.0 Hz, H12a), 2.07 (s, 3-OCOCH3),
2.19 (dt, J=16.0, 4.0 Hz, H11a), 2.35 (m, H12b), 2.40 (m, H5), 2.43 (m,
H6a), 2.52 (dt, J=16.0, 5.0 Hz, H11b), 2.62 (m, H6b), 2.65 (m, H10), 3.64
(s, 9-OCH3), 3.73 (s, 7-OCH3), 5.31 (s, H3), 5.53 (s, H17), 6.40 (s, H15),
6.55 (brs, H22), 7.01 (s, H30), 7.56 (brs, H23), 7.69 ppm (brs, H21);
13C NMR (see Table 2); IR (film): n˜ =3435, 2971, 2375, 1732, 1631, 1588,
1384, 1242, 1166, 1129, 1024, 875, 806, 604 cmÀ1; UV (MeOH): lmax
(loge)=212 (4.10), 336 nm (4.11); HR-TOFMS: m/z: calcd for
C30H36O10Na: 579.2206 [M+Na]+; found: 579.2214.
Plant material: Four batches of the seeds of Xylocarpus granatum were
collected in October and November 2005, and in January and March
2006 from the Hainan Island (Southern China).The identification of the
plant was performed by Prof.Yongshui Lin, Laboratory of Marine Biol-
ogy, South China Sea Institute of Oceanology, Chinese Academy of Sci-
ences.Voucher samples (nos.GKLMMM-002–4 to GKLMMM-002–7)
are maintained in the Herbarium of the South China Sea Institute of
Oceanology.
Extraction and isolation: Four batches of the dried seeds (4, 6, 6, and
8 kg, each) of X. granatum were crushed and extracted three times with
95% ethanol at room temperature.The extract was concentrated and
partitioned between water and petroleum ether.Then the aqueous layer
was further extracted with ethyl acetate and concentrated to give a
brown gum, which was subjected to silica-gel chromatography (chloro-
form/methanol 100:0 to 2:1).The fractions that were eluted with chloro-
form/methanol (25:1 to 15:1) were combined and purified by preparative
HPLC (YMC-Pack ODS-5 A, 250”20 mm i.d., acetonitrile/water 30:70
to 45:55) to yield xylogranatins F (1, 20 mg), G (2, 22 mg), H (3, 4 mg), I
(4, 40 mg), J (5, 8 mg), K (6, 20 mg), L (7, 6 mg), M (8, 3 mg), N (9,
6 mg), O (10, 7 mg), P (11, 6 mg), Q (12, 35 mg), R (13, 12 mg), and C[20]
(60 mg).Additionally, xylogranatin C was stored for two weeks dissolved
in methanol at room temperature; after this time, it was purified by
HPLC to afford xylogranatin D[20] (35 mg).
Xylogranatin F (1): Amorphous powder, [a]2D5 =+254 (c=0.21 in MeCN);
1H and 13C NMR (see Table 1); IR (film): n˜ =3420, 2970, 1772, 1712,
1595, 1440, 1365, 1239, 1161, 1129, 1061, 1027, 942, 875, 697, 602 cmÀ1
;
UV (MeOH): lmax (loge) = 210 (4.23), 268 (4.01), 310 nm (4.07); HR-
TOFMS: m/z: calcd for C26H27NO6Na: 472.1736 [M+Na]+; found:
472.1724.
Xylogranatin N (9): Amorphous powder, [a]2D5 =À41 (c=0.15 in MeOH);
1H NMR (CD3OD): d=0.88 (s, H28), 1.04 (s, H29), 1.25 (s, H18), 1.38 (d,
J=6.7 Hz, H19), 1.92 (m, H12a), 2.14 (m, H11a), 2.34 (m, H12b), 2.42
(m, H5), 2.43 (m, H11b), 2.45 (m, H6a), 2.62 (m, H6b), 2.65 (m, H10),
3.73 (s, 7-OCH3), 5.29 (s, H3), 5.53 (s, H17), 6.39 (s, H15), 6.54 (brs,
H22), 7.00 (s, H30), 7.54 (brs, H23), 7.69 ppm (brs, H21); 3-O-(2S)meth-
ylbutyryl: 0.86 (t, J=7.4 Hz, H4’), 1.13 (d, J=7.0 Hz, H5’), 1.50 (m, H3’),
1.65 (m, H3’), 2.38 ppm (m, H2’); 13C NMR (see Table 2); UV (MeOH):
lmax (loge)=212 (4.06), 335 nm (4.11); IR (film): n˜ =3434, 2970, 2375,
Xylogranatin G (2): Amorphous powder, [a]2D5 +105 (c=0.15 in MeCN);
1H and 13C NMR (see Table 1); IR (film): n˜ =3447, 2972, 1776, 1726,
1596, 1440, 1373, 1233, 1160, 1082, 1062, 1023, 937, 875, 697, 605 cmÀ1
;
UV (MeOH): lmax (loge)=208 (4.21), 270 (3.89), 310 nm (3.96); HR-
TOFMS: m/z: calcd for C28H29NO7: 491.1944 [M]+; found: 491.1947.
Xylogranatin H (3): Amorphous powder, [a]2D5 =+14 (c=0.12 in MeCN);
1H and 13C NMR (see Table 1); IR (film): n˜ =3447, 2933, 1736, 1632,
1384, 1238, 1163, 1027, 944, 875, 602 cmÀ1; UV (MeOH): lmax (loge)=217
(4.03), 280 nm (4.01); HR-TOFMS: m/z: calcd for C26H29NO6Na:
474.1893 [M+Na]+; found: 474.1884.
Xylogranatin I (4): Amorphous powder, [a]2D5 =À20 (c=0.20 in MeOH);
1H NMR (CD3OD): d=0.76 (s, H28), 1.05 (s, H29), 1.27 (s, H18), 1.36 (d,
J=6.7 Hz, H19), 1.92 (dt, J=14.5, 4.0 Hz, H12a), 2.17 (dt, J=16.0,
4.0 Hz, H11a), 2.32 (m, H5), 2.35 (m, H6a), 2.57 (m, H6b), 2.38 (m,
H12b), 2.47 (dt, J=16.0, 5.0 Hz, H11b), 2.60 (m, H10), 3.70 (s, 7-OCH3),
4.05 (s, H3), 5.52 (s, H17), 6.39 (s, H15), 6.53 (brs, H22), 7.08 (s, H30),
7.67 (brs, H21), 7.53 ppm (brs, H23); 13C NMR (see Table 2); IR (film):
n˜ =3436, 2970, 1631, 1582, 1384, 1027, 874, 603 cmÀ1; UV (MeOH): lmax
(loge)=213 (4.09), 336 nm (4.12); HR-TOFMS: m/z: calcd for C27H32O9:
500.2046 [M]+; found: 500.2052.
1721, 1631, 1587, 1384, 1267, 1186, 1027, 875, 807, 603 cmÀ1
; HR-
TOFMS: m/z: calcd for C32H40O10Na: 607.2519 [M+Na]+; found:
607.2531.
Xylogranatin O (10): Amorphous powder, [a]2D5 =À28 (c=0.11 in
MeOH); 1H NMR (CD3OD): d=0.89 (s, H28), 1.03 (s, H29), 1.27 (s,
H18), 1.39 (d, J=6.7 Hz, H19), 1.93 (m, H12a), 2.14 (m, H11a), 2.34 (m,
H12b), 2.45 (m, H11b), 2.46 (m, H6a), 2.50 (m, H5), 2.63 (m, H6b), 2.67
(m, H10), 3.73 (s, 7-OCH3), 5.33 (s, H3), 5.54 (s, H17), 6.40 (s, H15), 6.56
(brs, H22), 7.04 (s, H30), 7.55 (brs, H23), 7.70 ppm (brs, H21); 3-O-tiglo-
yl: 1.83 (d, J=7.1 Hz, H4’), 1.85 (s, H5’), 6.89 ppm (q, J=7.1 Hz, H3’);
13C NMR (see Table 2); IR (film): n˜ =3435, 2950, 2375, 1708, 1631, 1385,
1263, 1084, 1023, 875, 806, 603 cmÀ1; UV (MeOH): lmax (loge)=212
(4.18), 335 nm (4.00); HR-TOFMS: m/z: calcd for C32H38O10Na: 605.2363
[M+Na]+; found: 605.2374.
Xylogranatin J (5): Amorphous powder, [a]2D5 =À73 (c=0.10 in MeOH);
1H NMR (CD4OD): d=0.77 (s, H28), 1.08 (s, H29), 1.31 (s, H18), 1.34 (d,
J=6.7 Hz, H19), 1.95 (dt, J=14.5, 4.0 Hz, H12a), 2.20 (dt, J=16.0,
4.0 Hz, H11a), 2.37 (m, H6a), 2.38 (m, H5), 2.40 (m, H12b), 2.53 (dt, J=
16.0, 5.0 Hz, H11b), 2.58 (m, H6b), 2.59 (m, H10), 3.38 ppm (s, 3-OCH3),
3.63 (s, 9-OCH3), 3.66 (s, H3), 3.71 (s, 7-OCH3), 5.54 (s, H17), 6.41 (s,
H15), 6.56 (brs, H22), 7.16 (s, H30), 7.56 (brs, H23), 7.70 ppm (brs,
H21); 13C NMR (see Table 2); IR (film): n˜ =3435, 2932, 2375, 1631, 1581,
1406, 1385, 1027, 832, 620 cmÀ1; UV (MeOH): lmax (loge)=212 (4.12),
335 nm (4.13); HRTOF-MS m/z: calcd for C29H36O9Na: 551.2257
[M+Na]+; found: 551.2237.
Xylogranatin P (11): Amorphous powder, [a]2D5 =À36 (c=0.11 in
MeOH); 1H NMR (CD3OD): d=0.87 (s, H28), 1.04 (s, H29), 1.26 (s,
H18), 1.39 (d, J=6.7 Hz, H19), 1.95 (m, H12a), 2.13 (m, H11a), 2.33 (m,
H12b), 2.36 (m, H11b), 2.44 (m, H5), 2.44 (m, H6a), 2.63 (m, H6b), 2.65
(m, H10), 3.73 (s, 7-OCH3), 5.29 (s, H3), 5.54 (s, H17), 6.40 (s, H15), 6.56
(brs, H22), 7.01 (s, H30), 7.54 (brs, H23), 7.70 ppm (brs, H21); 3-O-iso-
butyryl: 1.13 (d, J=7.0 Hz, H4’), 1.17 (d, J=7.0 Hz, H3’), 2.58 ppm (m,
H2’); 13C NMR (see Table 2); IR (film): n˜ =3435, 2972, 2375, 1631, 1587,
1385, 1160, 1027, 875, 651 cmÀ1; UV (MeOH): lmax (loge)=212 (4.12),
335 nm (4.16); HR-TOFMS: m/z: calcd for C31H38O10Na: 593.2363
[M+Na]+; found: 593.2373.
Xylogranatin K (6): Amorphous powder, [a]2D5 =À31 (c=0.13 in MeOH);
Xylogranatin Q (12): Amorphous powder, [a]2D5 =À38 (c=0.13 in
1H NMR (CD3OD): d=0.74 (s, H28), 1.05 (s, H29), 1.28 (s, H18), 1.32 (d,
MeOH); 1H NMR (CD3OD): d=1.08 (s, H28), 1.15 (s, H29), 1.31 (s,
1142
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144

Xylogranatins F–R
FULL PAPER
H18), 1.51 (d, J=7.4 Hz, H19), 1.90 (m, H12a), 2.11 (m, H11a), 2.40 (m,
H12b), 2.49 (m, H11b), 2.52 (s, H5), 3.24 (q, J=7.4 Hz, H10), 3.79 (s, 7-
OCH3), 4.21 (s, H6), 4.51 (s, H3), 5.56 (s, H17), 6.40 (s, H15), 6.57 (brs,
H22), 7.09 (s, H30), 7.56 (brs, H23), 7.71 ppm (brs, H21); 13C NMR (see
Table 2); IR (film): n˜ =3438, 2952, 1714, 1631, 1506, 1375, 1262, 1165,
1126, 1051, 1032, 875, 805, 603 cmÀ1; UV (MeOH): lmax (loge)=213
(4.13), 336 nm (4.27); HR-TOFMS: m/z: calcd for C27H30O9Na: 521.1788
[M+Na]+; found: 521.1796.
Xylogranatin R (13): Amorphous powder, [a]2D5 =+10 (c=0.14 in
MeOH); 1H NMR (CDCl3): d=1.08 (d, J=6.5 Hz, H19), 1.11 (s, H28),
1.19 (s, H29), 1.21 (s, H18), 1.79 (m, H12a), 2.33 (m, H6a), 2.36 (m,
H12b), 2.37 (m, H5), 2.40 (m, H11a), 2.40 (m, H11b), 2.45 (m, H10), 2.50
(d, J=16.0 Hz, H6b), 3.45 (d, J=16.8 Hz, H30a), 3.70 (s, 7-OCH3), 3.75
(d, J=16.8 Hz, H30b), 5.38 (s, H17), 6.47 (brs, H22), 6.54 (s, H3), 6.68 (s,
H15), 7.46 (brs, H23), 7.55 ppm (brs, H21); 13C NMR (see Table 2); IR
(film): n˜ =3435, 2954, 2375, 1729, 1680, 1381, 1161, 1084, 1032, 875,
603 cmÀ1; UV (MeOH): lmax (loge)=216 (4.03), 232 nm (3.99); HR-
TOFMS: m/z: calcd for C27H32O9Na: 523.1944; found: 523.1956
[M+Na]+.
H329), 1.27 (s, 3H, H318), 1.34 (d, J=6.8 Hz, 3H, H319), 2.04 (m, H12b),
2.19 (m, H12a), 2.28–2.33 (4H, overlapped, H5, H6b, H11a, H11b), 2.51–
2.55 (2H, overlapped, H6a, H10), 3.15 (m, 2H, H22’), 3.32 (m, 2H, H22’),
3.71 (s, 7-OMe), 4.08 (s, H3), 5.36 (s, H17), 6.44 (s, H15), 6.49 (brs, H22),
6.87 (s, H30), 7.41 (brs, H23), 7.54 ppm (brs, H21); 13C NMR (CDCl3):
d=158.1 (C1), 121.3 (C2), 72.4 (C3), 39.1 (C4), 41.3 (C5), 33.5 (C6),
174.0 (C7), 148.3 (C8), 170.6 (C9), 33.3 (C10), 28.0 (C11), 32.5 (C12),
40.7 (C13), 151.3 (C14), 113.4 (C15), 164.6 (C16), 78.3 (C17), 21.3 (C18),
16.2 (C19), 120.5 (C20), 141.7 (C21), 110.1 (C22), 143.2 (C23), 24.3
(C28), 19.7 (C29), 41.9 (C2’), 40.4 (C2’), 14.4 (C1’), 13.1 ppm (C1’).
Ester 4’s: Amorphous powder; 1H NMR (CD3OD): d=0.89 (s, 3H,
H328), 0.96 (s, 3H, H329), 1.05 (t, J=7.1 Hz, 3H, H31’), 1.09 (t, J=7.1 Hz,
3H, H31’), 1.24 (s, 3H, H318), 1.38 (d, J=6.8 Hz, 3H, H319), 2.22 (m,
H12b), 2.29 (m, H12a), 2.37 (m, H5), 2.48 (m, H11b), 2.42 (dd, J=16.8,
4.0 Hz, H6b), 2.59 (dd, J=16.8, 10.0 Hz, H6a), 2.65 (m, H11a), 2.68 (m,
H10), 3.19 (m, 2H, H2–2’), 3.46 (m, 2H, H2–2’), 3.66 (s, 7-OMe), 5.53 (s,
H3), 5.60 (s, H17), 6.47 (s, H15), 6.55 (brs, H22), 7.09 (s, H30), 7.56 (brs,
H23), 7.70 ppm (brs, H21); ESI-MS: m/z: 538 [MÀMTPAOH]+, 560
[MÀMTPAOH+Na]+, 794 [M+Na+], 810 [M+K+].
Ester 4’r: Amorphous powder; 1H NMR (CD3OD): d=0.88 (s, 3H,
H328), 1.03 (t, J=7.1 Hz, 3H, H31’), 1.09 (s, 3H, H329), 1.09 (t, J=7.1 Hz,
3H, H31’), 1.22 (d, J=6.8 Hz, 3H, H319), 1.31 (s, 3H, H318), 2.01 (m,
H12b), 2.25 (m, H12a), 2.32 (m, H5), 2.33 (m, H11b), 2.40 (dd, J=16.8,
4.0 Hz, H6b), 2.53 (m, H11a), 2.56 (dd, J=16.8, 10.0 Hz, H6a), 2.62 (m,
H10), 3.23 (m, 2H, H2’), 3.32 (m, 2H, H2’), 3.68 (s, 7-OMe), 5.50 (s, H3),
5.62 (s, H17), 6.45 (s, H15), 6.57 (brs, H22), 7.12 (s, H30), 7.56 (brs,
H23), 7.71 ppm (brs, H21); ESI-MS: m/z: 538 [MÀMTPAOH]+, 560
[MÀMTPAOH+Na]+, 794 [M+Na]+, 810 [M+K]+.
Antifeedant bioassay: Mythimna separata Walker is a serious crop pest in
North China.First or second instar larvae of M. separata were collected
from a maize field in Shaanxi (China) where pesticides had not been ap-
plied.These larvae were reared in the laboratory under a controlled pho-
toperiod (12:12 h light/dark), temperature (T=25Æ28C), relative humid-
ity (RH=65% ꢀ80%), and fed daily with maize leaves until they
reached the earlier stage of the third instar larvae, when they were used
for antifeedant tests.Three groups of ten larvae each were used for the
antifeedant test of one compound.The tested compounds were dissolved
in acetone at a concentration of 1 mgmLÀ1.Wafer discs (1 cm diameter,
1 mm thick), which were made from maize leaves were dipped into ace-
tone solutions of each compound for 3 s and air-dried for 5 min.After
drying, one disc was placed in a Petri dish with a M. separata larva, which
had been starved for 2 h.Another disc was added after the first one had
been consumed.After 72 h, undipped discs were given to the larvae.
Discs that had been treated with acetone alone were used as a control
group.If a whole disc was consumed, food consumption was recorded as
1.If only part of a disc was consumed, the food consumption was as-
sessed by estimating the percentage of the consumed leaf surface.After
24 h, 48 h, or 72 h, the antifeedant rates (AR) were calculated by the fol-
lowing index.
Mosherꢀs MTPA esters 1s/1r and 3s/3r: Xylogranatin F (1) (2 mg) was
treated with (R)-MTPACl (10 mL) and DMAP (1 mg) in dried pyridine
(0.5 mL) at room temperature for 5 h. The reaction mixture was concen-
trated and purified by RP-HPLC (YMC-Pack ODS S-5m, 250”10 mm
i.d.) with aqueous acetonitrile to afford the (S)-MTPA ester 1s.The ( S)-
MTPA ester 3s and (R)-MTPA esters 1r and 3r were prepared in the
same way.
Ester 1s: Amorphous powder; 1H NMR (CDCl3): d=0.79 (s, 3H, H329),
1.05 (s, 3H, H328), 1.15 (s, 3H, H318), 1.62 (s, 3H, H319), 1.77 (dt, J=
13.5, 5.0 Hz, H12b), 1.89 (dd, J=13.5, 4.0 Hz, H12a), 2.55 (dd, J=18.5,
1.0 Hz, H6b), 2.75 (d, J=9.5 Hz, H5), 2.99 (dd, J=18.5, 9.5 Hz, H6a),
3.12 (dt, J=19.0, 5.0 Hz, H11a), 3.25 (dd, J=19.0, 4 Hz, H11b), 5.26 (s,
H17), 5.73 (s, H3), 6.52 (brs, H22), 6.58 (s, H15), 7.48 (brs, H23), 7.56
(brs, H21), 8.22 ppm (s, H30); ESI-MS: m/z: 666 [M+H]+, 688 [M+Na]+
.
Ester 1r: Amorphous powder; 1H NMR (CDCl3): d=0.78 (s, 3H, H329),
1.12 (s, 3H, H328), 1.15 (s, 3H, H318), 1.45 (s, 3H, H319), 1.79 (dt, J=
13.0, 5.5 Hz, H12b), 1.89 (dd, J=13.0, 5.5 Hz, H12a), 2.53 (dd, J=18.0,
1.0 Hz, H6b), 2.73 (d, J=9.5 Hz, H5), 2.97(dd, J=18.0, 9.5 Hz, H6a), 3.12
(dt, J=18.5, 5.0 Hz, H11a), 3.20 (dd, J=18.5, 5.5 Hz, H11b), 5.28 (s,
H17), 5.64 (s, H3), 6.52 (brs, H22), 6.63 (s, H15), 7.48 (brs, H23), 7.56
(brs, H21), 8.26 ppm (s, H30); ESI-MS: m/z: 666 [M+H]+.
1
Ester 3s: Amorphous powder; H NMR (CD3OD): d=0.76 (s, 3H, H29),
1.04 (s, 3H, H18), 1.08 (s, 3H, H28), 1.61 (s, 3H, H19), 1.76 (m, H12b),
2.02 (m, H12a), 2.62 (d, J=18.8 Hz, H6b), 2.76 (d, J=9.5 Hz, H5), 2.78
(dd J=16.0, 11.5 Hz, H15b), 2.89 (dd, J=18.5, 16.0 Hz, H15a), 2.99 (dd,
J=18.5, 5.0 Hz, H11a), 3.15 (H11b), 3.18 (H6a), 3.20 (H14), 5.46 (s,
H17), 5.94 (s, H3), 6.62 (brs, H22), 7.60 (brs, H23), 7.68 (brs, H21),
7.89 ppm (s, H30); ESI-MS m/z: 668 [M+H]+, 690 [M+Na]+, 706
[M+K]+.
AR=(CÀT)”100/C
in which C is the average consumption by one larva in the control and T
1
Ester 3r: Amorphous powder; H NMR (CD3OD): d=0.74 (s, 3H, H29),
1.04 (s, 3H, H18), 1.21 (s, 3H, H28), 1.28 (s, 3H, H19), 1.76 (m, H12b),
2.02 (m, H12a), 2.61 (d, J=18.8 Hz, H6b), 2.73 (d, J=9.5 Hz, H5), 2.90
(dd, J=16.0, 11.5 Hz, H15b), 2.99 (dd, J=18.5, 16.0 Hz, H15a), 3.00
(H11a), 3.15 (H11b), 3.16 (H6a), 3.20 (H14), 5.48 (s, H17), 5.82 (s, H3),
6.63 (brs, H22), 7.60 (brs, H23), 7.68 (brs, H21), 7.93 ppm (s, H30); ESI-
MS: m/z: 690 [M+Na]+, 706 [M+K]+.
in the treatment.The concentration for 50% antifeedant effect (AFC
was determined by log-probit analysis.All data were treated by log–
probit analysis and 95% fiducial limits were calculated.
)
50
Computational details: All optimizations were performed with the soft-
ware package Gaussian 03,[33] by using the DFT functional B3LYP[22] and
the basis set 6–31G(d)[23] To identify the found structures as minima fre-
quency calculations were accomplished on the same level of theory.Ro-
tational strength values for the electronic transitions from the ground
state to the singly excited states for 1 and 4 were obtained by time de-
pendent (TD) DFT calculations (B3LYP/6–31G(d)) with Gaussian 03.
Additionally Grimmeꢁs DFT/MRCI[26] approach (configuration selection
cutoff=0.8 Eh)[34] was used to calculate these values for 1.Single-point
SCF calculations of the B3LYP/6–31G(d)-optimized structures with
BHLYP/SVP,[27,35] which were needed for the DFT/MRCI calculations,
were carried out with Turbomole 5.8.[36] The rotational strength values
were summed after a Boltzmann statistical weighting and De values were
calculated by forming sums of Gaussian functions centered at the wave-
Mosherꢀs MTPA esters 4’s/4’r: Treatment of 4 (3 mg) with (R)- and (S)-
MTPACl (15 mL) in dichloromethane (0.5 mL) with a mixture of diethyl-
amine (0.2 mL), triethylamine (0.2 mL) and DMAP (2 mg) at room tem-
perature for 12 h afforded 4’ (4 mg).Then 4’ (2 mg) was treated with (R)-
MTPACl (10 mL) and DMAP (1 mg) in dried pyridine (0.5 mL) at room
temperature for 5 h.The reaction mixture was concentrated and purified
by RP-HPLC (YMC-Pack ODS S-5m, 250”10 mm i.d.) with aqueous ace-
tonitrile to afford the (S)-MTPA ester 4’s.The ( R)-MTPA ester 4’r was
prepared in the same way.
Ester 4’: Amorphous powder; 1H NMR (CDCl3): d=0.75 (s, 3H, H328),
1.03 (t, J=7.1 Hz, 3H, H31’), 1.07 (t, J=7.1 Hz, 3H, H31’), 1.08 (s, 3H,
Chem. Eur. J. 2008, 14, 1129 – 1144
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
1143

G.Bringmann, J.Wu et al.
lengths of the respective electronic transitions and multiplied by the cor-
responding rotational strengths.The CD spectra that were obtained,
were UV-corrected[25] and compared with the experimental ones.
[20] S.Yin, C.Q.Fan, X.N.Wang, L.P.Lin, J.Ding, J.M.Yue,
Org.
Lett. 2006, 8, 4935–4938.
[21] I.Ohtani, T.Kusumi, Y.Kashman, H.Kakisawa,
1991, 113, 4092–4096.
J. Am. Chem. Soc.
For further experimental details and considerations on the occurrence of
the new compounds as genuine natural products, see the Supporting In-
formation.
[22] a) C.Lee, W.Yang, RG. .Parr,
Phys. Rev. B 1988, 37, 785–789;
b) A.D.Becke, Phys. Rev. A 1988, 38, 3098–3100.
[23] a) P.C. Hariharan, J.A. Pople,
Theor. Chim. Acta 1973, 28, 213–
222; b) M.M. Francl, W.J. Pietro, W.J. Hehre, J.S. Binkley, M.S.
Gordon, D.J. DeFrees, J.A. Pople, J. Chem. Phys. 1982, 77, 3654–
3665.
[24] Structures that lie energetically higher than 3 kcalmolÀ1 above the
energetically lowest structure were found not to contribute to the
overall CD curve that was obtained by superposition of the Boltz-
mann-weighted single spectra.
Acknowledgements
Support for this work from the Important Projects of the Chinese Acade-
my of Sciences (KZCX3-SW-216 and KZCX2-YW-216), the National
High Technology Research and Development Program of China (863
Program) (2007 AA09Z407), the Natural Science Foundation of China
(20772135), the Natural Science Foundation of Guangdong Province
(04100729), the Fonds der Chemischen Industrie (Germany), and the
Deutsche Forschungsgemeinschaft (SPP 1152, “Evolution of Metabolic
Diversity”, SFB 630 “Recognition, Preparation, and Functional Analysis
of Agents against Infectious Diseases”) is gratefully acknowledged.Mass
spectra were provided by the Institute of Chemistry, Chinese Academy
of Sciences, and the NMR spectra by the Laboratory of NMR Analysis
and Measurement, South China Sea Institute of Oceanology, Chinese
Academy of Sciences. G.B and T.B. are grateful for generous allocation
of computational time from Dr.R.Koch, University of Oldenburg (Ger-
many).
[25] G.Bringmann, S.Busemann, in The Quantum Chemical Calculation
of CD Spectra: the Absolute Configuration of Chiral Compounds
from Natural or Synthetic Origin (Eds.: P. Schreier, M. Herderich,
H-.U.Humpf, W.Schwab), Vieweg, Wiesbaden,
1998, pp.195–211.
[26] S.Grimme, M.Waletzke, J. Chem. Phys. 1999, 111, 5645–5655.
[27] A.Schaefer, H.Horn, R.Ahlrichs,
2577.
J. Chem. Phys. 1992, 97, 2571–
[28] J.Wu, Q.Xiao, J.Huang, Z.Xiao, S.Qi, Q.Li, S.Zhang,
Org. Lett.
2004, 6, 1841–1844.
[29] J.Lin, Z.Q.Ma, J.T.Feng, X.Zhang,
Acta Agric. Bor.-Occid. Sin.
2005, 14, 94–97.
[30] R.Gao, X.Tian, X.Zhang,
Acta Univ. Agric. Bor.-Occid. 2000, 28,
Acta Univ. Agric. Bor.-Occid. 2000, 28,
8–13.
[31] X.Tian, R.Gao, X.Zhang,
19–24.
[32] Y.S.Xie, P.G.Fields, M.B.Isman, W.K.Chen, X.Zhang,
J. Stored
[1] A.S.Ng, A.G.Fallis, Can. J. Chem. 1979, 57, 3088–3089.
[2] I.Kubo, I.Miura, K.Nakanishi, J. Am. Chem. Soc. 1976, 98, 6704–
6705.
[3] K.A.Alvi, P.Crews, B.Aalbersberg, R.Prasad, J.Simpson, R.T.
Weavers, Tetrahedron 1991, 47, 8943–8948.
Prod. Res. 1995, 31, 259–265.
[33] Gaussian 03 (Revision D0. 1), M.J.Frisch, G.W.Trucks, H.B.Schle-
gel, G.E.Scuseria, M.A.Robb, J.R.Cheeseman, J.J.A.Montgom-
ery, T.Vreven, K.N.Kudin, J.C.Burant, J.M.Millam, S.S.Iyengar,
J.Tomasi, V.Barone, B.Mennucci, M.Cossi, G.Scalmani, N.Rega,
G.A.Petersson, H.Nakatsuji, M.Hada, M.Ehara, K.Toyota, R.
Fukuda, J.Hasegawa, M.Ishida, T.Nakajima, Y.Honda, O.Kitao,
H.Nakai, M.Klene, X.Li, J.E.Knox, H.P.Hratchian, J.B.Cross,
V.Bakken, C.Adamo, J.Jaramillo, R.Gomperts, R.E.Stratmann,
O.Yazyev, A.J.Austin, R.Cammi, C.Pomelli, J.W.Ochterski, P.Y.
Ayala, K.Morokuma, GA. .Voth, P.Salvador, JJ..Dannenberg,
VG. .Zakrzewski, S.Dapprich, AD. .Daniels, MC. .Strain, O.
Farkas, D.K.Malick, A.D.Rabuck, K.Raghavachari, J.B.Fores-
man, JV. .Ortiz, Q.Cui, A.G.Baboul, S.Clifford, J.Cioslowski,
B.B.Stefanov, G.Liu, A.Liashenko, P.Piskorz, I.Komaromi, R.L.
Martin, DJ..Fox, T.Keith, M.A.Al-Laham, C.Y.Peng, A.Na-
nayakkara, M.Challacombe, PM. W. .Gill, B.Johnson, W.Chen,
M.W.Wong, C.Gonzalez, J.A.Pople, Gaussian, Inc,. Wallingford
CT, 2004.
[4] D.A.Mulholland, B.Parel, P.H.Coombes,
Curr. Org. Chem. 2000,
4, 1011–1054.
[5] U.Kokpol, W.Chavasiri, S.Tip-pyang, G.Veerachato, F.L.Zhao, J.
Simpson, R.T.Weavers, Phytochemistry 1996, 41, 903–905.
[6] J.Wu, M.Y.Li, Z.H.Xiao, Y.Zhou,
1447–1449.
Z. Naturforsch. B 2006, 61,
[7] J.Wu, S.Zhang, Q.Xiao, Q.X.Li, J.S.Huang, Z.H.Xiao, L.J.
Long, Z. Naturforsch. B 2003, 58, 1216–1219.
[8] J.Wu, S.Zhang, Q.Xiao, Q.X.Li, J.S.Huang, L.J.Long, L.M.
Huang, Tetrahedron Lett. 2004, 45, 591–593.
[9] J.Wu, Q.Xiao, J.S.Huang, Z.H.Xiao, S.H.Qi, Q.X.Li, S.Zhang,
Org. Lett. 2004, 6, 1841–1844.
[10] J.Wu, Q.Xiao, S.Zhang, X.Li, Z.H.Xiao, H.X.Ding, Q.X.Li,
Tetrahedron 2005, 61, 8382–8389.
[11] J.Wu, S.Zhang, Y.Song, Z.H.Xiao, Q.Xiao, Q.X.Li,
forsch. B 2005, 60, 1291–1294.
Z. Natur-
[34] C.Diedrich, S.Grimme, J. Phys. Chem. A 2003, 107, 2524–2539.
[35] A.D.Becke, J. Chem. Phys. 1993, 98, 1372–1377.
[12] J.Wu, Z.H.Xiao, Y.Song, S.Zhang, Q.Xiao, C.Ma, H.X.Ding,
Q.X.Li, Magn. Reson. Chem. 2006, 44, 87–89.
[36] TURBOMOLE (Version 5.8), R. Ahlrichs, M. Bar, H.-P. Baron, R.
Bauernschmitt, S.Bçcker, N.Crawford, P.Deglmann, M.Ehrig, K.
Eichkorn, S.Elliott, F.Furche, F.Haase, M.Hꢂser, H.Horn, C.
Hꢂttig, C.Huber, U.Huniar, M.Kattanek, A.Kçhn, C.Kçlmel, M.
Kollwitz, K.May, P.Nava, C.Ochsenfeld, H.ꢃhm, H.Patzelt, D.
Rappoport, O.Rubner, A.Schꢂfer, U.Schneider, M.Sierka, O.
Treutler, B.Unterreiner, M.von Arnim, F.Weigend, P.Weis, H.
Weiss, Universitꢂt Karlsruhe, Karlsruhe, Germany, 2005.
[13] F.Cheng, Y.Zhou, J.Wu, K.Zou,
Z. Naturforsch. B 2006, 61, 626–
628.
[14] Y.Zhou, F.Cheng, J.Wu, K.Zou,
1085.
J. Nat. Prod. 2006, 69, 1083–
[15] J.Wu, Q.Xiao, Q.X.Li, Biochem. Syst. Ecol. 2006, 34, 838–841.
[16] J.Wu, S.Zhang, M.Y.Li, Y.Zhou, Q.Xiao,
2006, 54, 1582–1585.
Chem. Pharm. Bull.
[17] J.Wu, H.X.Ding, M.Y.Li, S.Zhang,
569–572.
Z. Naturforsch. B 2007, 62,
[18] J.X.Cui, Z.W.Deng, J.Li, H.Z.Fu, P.Proksch, W.H.Lin,
chemistry 2005, 66, 2334–2339.
[19] A.D.Roy, R.Kumar, P.Gupta, T.Khaliq, T.Narender, V.Aggar-
wal, R.Roy, Magn. Reson. Chem. 2006, 44, 1054–1057.
Phyto-
Received: May 2, 2007
Revised: September 9, 2007
Published online: November 28, 2007
1144
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1129 – 1144