Revised structure for a neolignan from Brucea javanica

Article abstract of DOI:10.1016/S0031-9422(98)00449-X

On the basis of ¹H and ¹³C NMR spectral comparison with natural and synthetic compounds, the neolignan isomers isolated from Brucea javanica for which unusual 8-O-6'-structures were proposed have been reassigned to the conventional 8-O-4'-structures (erythro and threo-guaiacylglycerol-β- coniferyl ether), respectively. Thus far, there is no evidence for the existence of 8-O-6'-neolignans in nature.

Full text of DOI:10.1016/S0031-9422(98)00449-X

Phytochemistry Vol. 49, No. 7, pp. 2125±2128, 1998
# 1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
0031-9422/98/$ - see front matter
PII: S0031-9422(98)00449-X
REVISED STRUCTURE FOR A NEOLIGNAN FROM
BRUCEA JAVANICA
SHIMING LI, KNUT LUNDQUIST and ADRIAN F. A. WALLIS *
Department of Organic Chemistry, Chalmers University of Technology, 412 96 Goteborg, Sweden,
CSIRO Forestry and Forest Products, Private Bag 10, South Clayton MDC, Vic. 3169, Australia
(Received 22 January 1998; in revised form 15 May 1998)
Key Word IndexÐBrucea javanica; Simaroubaceae; 8-O-4'-neolignans; dilignols; guaiacylgly-
1
cerol-b-coniferyl ether; H and 13C NMR spectra.
AbstractÐOn the basis of 1H and 13C NMR spectral comparison with natural and synthetic compounds,
the neolignan isomers isolated from Brucea javanica for which unusual 8-O-6'-structures were proposed
have been reassigned to the conventional 8-O-4'-structures (erythro and threo-guaiacylglycerol-b-coniferyl
ether), respectively. Thus far, there is no evidence for the existence of 8-O-6'-neolignans in nature. # 1998
Elsevier Science Ltd. All rights reserved
INTRODUCTION
RESULTS AND DISCUSSION
Both erythro and threo neolignans 4 and 5 were
prepared through a reaction sequence starting from
the intermediate 6a. Oxidation of 6a with 2,3-
Lignans and neolignans belong to an important
group of optically active natural products consisting
of two phenylpropane monomers linked through
carbon±carbon or carbon±oxygen bonds [1]. In one
class of neolignans, the phenylpropane monomers
are linked through the 8-O-4' atoms, as in formula
1. Luyengi et al. [2] have isolated from Brucea java-
nica a mixture of neolignan isomers which were
found to induce cell di€erentiation with human pro-
myelocytic leukemia (HL-60) cells. They proposed
the unusual structures 2 and 3, ether-linked from C-
8 to C-6' rather than to C-4', for the compounds on
dichloro-5,6-dicyanobenzoquinone
using
a
modi®cation [5] of a procedure used for the prep-
aration of coniferaldehyde [6] gave the formylvinyl
compound 6b. Reduction of 6b with sodium boro-
hydride and subsequent alkaline hydrolysis gave
rise to a mixture of erythro and threo neolignans 4
and 5. Separation of 4 and 5 was achieved by ion
exchange chromatography of their borate com-
plexes, the threo isomer 5 eluting from the column
in advance of the erythro isomer 4 [7]. NMR spec-
tral comparison with related arylglycerol-b-aryl
ethers examined by X-ray crystallography [8] pro-
vides unambiguous proof of the steric assignments
of 4 and 5. Compounds 4 and 5 are important sub-
structures of lignin and they have been previously
prepared as an isomeric mixture by two research
groups [9, 10].
1
the basis of H and 13C NMR spectroscopy [2]. We
have recently shown [3] that compounds isolated
from Arum italicum for which similar unusual struc-
tures were proposed on the basis of NMR
spectroscopy [4] were actually compounds 4 and 5,
isomers of the known 8-O-4' neolignan guaiacylgly-
cerol-b-coniferyl ether. Because the structures pro-
posed for 2 and 3 represent linkages not previously
found in neolignans, we undertook to synthesize
compounds 4 and 5 and to compare their NMR
spectra with those of the neolignan mixture. Herein
we present evidence to show that the compounds
isolated from Brucea javanica are the known iso-
meric neolignans 4 and 5, rather than compounds 2
and 3, respectively.
1
The H and 13C NMR spectral data for deuterio-
dimethylsulfoxide solutions of compounds 4 and 5
and those reported by Luyengi et al. [2] for the neo-
lignan mixture of 2 and 3 are given in Tables 1 and
2, respectively. It is clear from Table 1 that, with
the exception of the signals assigned to the hydroxyl
1
protons, the H NMR signals reported for the mix-
ture of 2 and 3 are almost identical to those of the
erythro 8-O-4' neolignan 4 and are similar to those
of the threo isomer 5. Similarly, the 13C NMR sig-
*Author to whom correspondence should be addressed.
2125

2126
S. LI et al.
nals for 2 and 3 correspond closely to those of the of each isomer to the mixture (Table 1). The multi-
synthetic neolignans 4 and 5.
plet assigned to the 8' proton at 6.23 ppm was in
fact a mixture of two double triplets centred at 6.23
The correspondence of the 1H NMR data of the
neolignan mixture to that of the erythro isomer 4 is
and 6.25 ppm for 4 and 5, respectively, the double
explicable because the neolignan mixture of
doublet at 6.67 ppm was a composite of two doub-
Luyengi et al. was estimated to be a 3:1 mixture of
lets for each isomer and the broad doublet at
erythro:threo isomers [2], so that the signals
assigned to the latter would be relatively minor.
6.76 ppm and triplet at 6.85 ppm were each com-
The 1H NMR signals of the neolignan mixture
prised of two double doublets. It is well recognised
could be reinterpreted in terms of the contribution that the chemical shifts of the hydroxyl protons are

Revised structures for neolignan from B. javanica
2127
Table 1. 1H NMR chemical shifts of compounds in DMSO-d6
H (2 and 3)*
2 and 3*
3.58 (bs)
4
5
H (4 and 5)
9 (2H)
3.59 (m)
3.59 (m)
4.08 (m)
4.30 (m)
4.70 (m)
3.23 (m)
3.57 (m)
4.09 (m)
4.26 (m)
4.71 (m)
9a (1H)
9b (1H)
9' (2H)
8 (1H)
9' (2H)
8 (1H)
7 (1H)
8' (1H)
7' (1H)
5 (1H)
6 (1H)
4' (1H)
5' (1H)
2,3' (2H)
4.08 (bs)
4.29 (m)
4.68 (m)
7 (1H)
6.23 (m)
6.23 (dt) (15.9, 5.2)
6.43 (bd) (15.9)
6.67 (d) (8.0)
6.77 (dd) (8.0, 1.6)
6.84 (dd) (8.4, 1.8)
6.92 (d) (8.4)
6.99 (m)
6.99 (m)
3.72 (s)
3.73 (s)
4.62 (t) (5.4)
4.81 (t) (5.4)
5.33 (d) (4.4)
8.78 (s)
6.25 (dt) (15.9, 5.2)
6.45 (bd) (15.9)
6.68 (d) (8.2)
6.76 (dd) (8.2, 2.0)
6.86 (dd) (8.4, 2.0)
6.94±6.98 (m)
6.94±6.98 (m)
7.04 (d) (2.0)
3.72 (s)
8' (1H)
7' (1H)
5 (1H)
6.44 (d) (16.0)
6.67 (dd) (8.0, 2.5)
6.76 (bd) (8.0)
6.85 (t) (8.0)
6.92 (d) (8.0)
6.98 (m)
6 (1H)
6' (1H)
5' (1H)
2 or 2' (1H)
2 or 2' (1H)
OCH3
OCH3
9-OH
9'-OH
7-OH
Ar±OH
OCH3 (6H)
3.72 (s)
3.79 (s)
9-OH
3.00 (bm)
4.85 (bs)
5.30 (m)
4.65 (t) (5.2)
4.81 (t) (5.4)
5.28 (d) (4.4)
8.80 (s)
9'-OH
7-OH
Ar±OH
*Data from Ref. [2].
dependent on the conditions under which the NMR al. [2] to the isomeric neolignan mixture are incor-
rect and that these substances are the known neo-
lignans 4 and 5. Compounds 4 and 5 have been
reported recently as components of hydrolysates
from treatment of spruce (Picea abies) wood and
high yield pulps derived from spruce wood with an
aqueous solution at pH 4 [11]. A mixture of the
dilignol isomers 4 and 5 was previously obtained by
mild acid treatment of spruce wood [12]. One of the
isomers (4 or 5) has been isolated as a neolignan
extractive from the wood of Larix leptolepis [13]
and both isomers have been found as components
of the bark of Ehretia ovalifolia [14].
experiments are conducted.
The 13C NMR signals of the neolignan mixture 2
and 3 generally match those of isomers 4 and 5
(Table 2), although Luyengi et al. [2] reported an
additional signal for C-5' at 115.01 ppm in the spec-
trum of the mixture which did not correspond clo-
sely with
a signal in either of the synthetic
isomers [2]. No attempt was made to assign the 13C
NMR signals to the aromatic carbon atoms in 4
and 5 in the present work.
Because of the close correspondence of the 1H
and 13C NMR data of the isomeric mixture and
those of the synthetic neolignans 4 and 5, it is clear
that the structures 2 and 3 given by Luyengi et
The use of NMR spectroscopy for structural de-
termination of neolignans has led to several
instances
of
incorrect
assignments
being
reported [3, 15, 16]. It is thus recommended that
complementary methods should be investigated in
addition to the NMR measurements for structural
determinations. Thus far, there is no evidence to
support the existence of 8-O-6'-linked neolignans in
nature.
Table 2. 13C NMR chemical shifts of compounds in
DMSO-d6
C (2 and 3)*
2 and 3*
4
5
C (4 and 5)
OCH3
OCH3
9
9'
7
7
8
8
3'
2
5
55.41
55.57
60.07
61.62
70.88
71.56
83.64
84.23
109.76
111.35
114.53
115.01
115.35
119.03
119.46
128.52
55.3
55.4
60.0
61.5
55.3
55.5
60.0
61.5
70.8
OCH3
OCH3
9
9'
7 (threo)
7 (erythro)
8 (erythro)
8 (threo)
71.5
83.5
EXPERIMENTAL
84.2
109.6
110.8
114.5
Spectra
109.6
111.2
114.4
1H NMR spectra were recorded at 400 MHz and
13C NMR spectra at 100.6 MHz with a Varian XL-
400 (VXR-5000) instrument. Chemical shifts are
given in ppm from tetramethylsilane (d 0.00).
5'
5'
4'
6
115.3
118.9
119.3
128.3
128.4
129.8
133.1
145.3
146.8
147.4
149.5
115.3
118.8
118.9
128.4
128.4
130.0
132.8
145.3
146.8
147.7
149.5
7',8'
7' or 8'
7' or 8'
1-(4-Benzoyl-3-methoxyphenyl)-3-hydroxy-2-{2-
methoxy-4-[(E)-1-propenyl]phenoxy}-1-propanone
(6a)
1'
1
4
3
6'
2'
130.05
132.89
145.60
146.91
147.78
149.65
This was prepared starting from acetoguaiacone
and (E)-isoeugenol according to methods applied
to the preparation of 1-(4-benzoyl-3-methoxy-
phenyl)-3-hydroxy-2-(2-methoxyphenoxy)-1-pro-
*Data from Ref. [2].

2128
S. LI et al.
panone [17, 18] 1H NMR (CDCl3) d 1.86 (3H, dd, H-9), 4.30 (1H, dd, J = 4.6, 12.0 Hz, H-9), 4.63
J = 1.6, 6.6 Hz, H-9'), 3.07 (1H, br t, OH), 3.86 (1H, m, H-8), 4.71 (2H, d, J = 6.7 Hz, H-9'), 6.11
(6H, s, 2Â OCH3), 4.09 (2H, m, H-9), 5.41 (1H, dd, (1H, d, J = 6.3 Hz, H-7), 6.18 (1H, dt, J = 6.7,
J = 4.5, 5.6 Hz, H-8), 6.12 (1H, qd, J = 6.6, 15.9 Hz, H-8'), 6.59 (1H, br d, J = 15.9 Hz, H-7')
15.7 Hz, H-8'), 6.32 (1H, qd, J = 1.6, 15.7 Hz, H-7') and 6.8±7.1 (6H, m, H±Ar). Lit. [3, 11] identical
and 6.7±8.3 (11H, m, Ar±H).
NMR spectra for tetra-acetates of 4 and 5.
1-(4-benzoyl-3-methoxyphenyl)-2-{4-[(E)-2-formyl-
vinyl]-2-methoxyphenoxy}-3-hydroxy-1-propanone
(6b)
AcknowledgementsÐFinancial support from the
Jacob Wallenbergs Forskningsstiftelse is gratefully
acknowledged.
Compound 6a (370 mg) was oxidised with 2,3-
dichloro-5,6-dicyano-p-benzoquinone
using
a
REFERENCES
modi®cation [5] of a procedure used for the prep-
aration of coniferaldehyde [6]. Adsorption of the
reaction mixture on a column of silica gel and
elution with dichloromethane±ethyl acetate 2:1 gave
compound 6b (270 mg) 1H NMR (CDCl3) d 3.87
(3H, s, OCH3), 3.90 (3H, s, OCH3), 4.17 (2H, m,
H-9'), 5.60 (1H, dd, J = 3.8, 6.0 Hz, H-8), 6.61 (1H,
dd, J = 7.7, 15.7 Hz, H-8'), 7.39 (1H, d,
J = 15.7 Hz, H-7'), 6.80±8.30 (11H, m, Ar±H) and
9.66 (1H, d, J = 7.7 Hz, CHO).
1. Whiting, D. A., Natural Products Report, 1985,
2, 191.
2. Luyengi, L., Suh, N., Fong, H. H. S., Pezzuto,
J. M. and Kinghorn, A. D., Phytochemistry,
1996, 43, 409.
3. Gellerstedt, G., Lundquist, K., Wallis, A. F. A.
and Zhang, L., Phytochemistry, 1995, 40, 263.
4. Greca, M. D., Molinaro, A., Monaco, P. and
Previtera, L., Phytochemistry, 1994, 35, 777.
5. Iliefski, T., Li, S. and Lundquist, K.,
Tetrahedron Letters, 1998, 39, 2413.
Erythro- and threo-1-(4-hydroxy-3-methoxyphenyl)-
2-{4-[(E)-3-hydroxy-1-propenyl]-2-methoxyphe-
noxy}-1,3-propanediol (4 and 5)
6. Nakamura, Y. and Higuchi, T., Wood
Research±60, 1976, 59, 101.
Sodium borohydride (150 mg) was added to a
solution of the aldehyde 6b (270 mg) in ethanol
(12 ml) and the mixture was kept at 208C for 16 h.
Sodium hydroxide (0.4 M, 7 ml) was added to the
reaction mixture and, after a further 2 h at 208C,
the reaction mixture was acidi®ed with 1 M hydro-
7. Li, S., Lundquist, K. and Soubbotin, N.,
Holzforschung, 1994, 48, 509.
8. Lundquist, K., Li, S. and Stomberg, R., Nordic
Pulp and Paper Research Journal, 1996, 11, 43.
9. Freudenberg, K. and Eisenhut, W., Chemische
Berichte, 1955, 88, 626.
chloric acid and extracted with chloroform. 10. Nakatsubo, F. and Higuchi, T., Wood
Evaporation of the extract gave an oil [230 mg, ery- Research, 1980, 66, 23.
thro:threo 4:5 was 2:1 (1H NMR)], which after 11. Gellerstedt, G. and Zhang, L., Journal of Wood
anion exchange chromatography and elution with Chemistry and Technology, 1992, 12, 387.
0.06 M potassium tetraborate in ethanol±water 12. Nimz, H., Chemische Berichte, 1965, 98, 533.
1:4 [7, 19] gave, successively, the threo and erythro 13. Miki, K., Takehara, T., Sasaya, T. and
compounds 5 and 4 (53 and 96 mg), respectively. 4
Tetra-acetate, 1H NMR (CDCl3) d 2.03 (3H, s, 14. Yoshikawa, K., Kageyama, H. and Arihara, S.,
CH3CO), 2.095 (3H, s, CH3CO), 2.101 (3H, s, Phytochemistry, 1995, 39, 659.
CH3CO), 2.31 (3H, s, CH3CO), 3.80 (3H, s, OCH3), 15. Reilly, M. J., Wallis, A. F. A., Lundquist, K.
Sakakibara, A., Phytochemistry, 1980, 19, 449.
3.82 (3H, s, OCH3), 4.25 (1H, dd, J = 4.1, 12.0 Hz,
H-9), 4.45 (1H, dd, J = 5.8, 12.0 Hz, H-9), 4.66
and Stomberg, R., Journal of Wood Chemistry
and Technology, 1992, 12, 471.
(1H, m, H-8), 4.71 (2H, dd, J = 1.0, 6.5 Hz, H-9'), 16. Li, S., Iliefski, T., Lundquist, K. and Wallis, A.
6.07 (1H, d, J = 5.3 Hz, H-7), 6.17 (1H, dt, F. A., Phytochemistry, 1997, 46, 929.
J = 6.5, 15.9 Hz, H-8'), 6.57 (1H, br d, 17. Adler, E. and Eriksoo, E., Acta Chemica
J = 15.9 Hz, H-7') and 6.7±7.1 (6H, m, H±Ar). 5 Scandinavica, 1956, 9, 341.
Tetra-acetate 1H NMR (CDCl3) d 2.00 (3H, s, 18. Landucci, L. L., Geddes, S. A. and Kirk, T.
CH3CO), 2.05 (3H, s, CH3CO), 2.10 (3H, s, K., Holzforschung, 1981, 35, 66.
CH3CO), 2.31 (3H, s, CH3CO), 3.82 (3H, s, OCH3), 19. Li, S., Lundquist, K. and Paulsson, M., Acta
3.83 (3H, s, OCH3), 4.06 (1H, dd, J = 5.8, 12.0 Hz, Chemica Scandinavica, 1995, 49, 623.