Concise synthesis of Cannabisin G

Article abstract of DOI:10.1016/j.bmcl.2010.07.028

Cannabisin G (1), a naturally occurring lignanamide, was synthesized in 45% overall yield starting from 3-tert-butyl ethyl ferulate (6). An oxidative coupling by potassium ferricyanide in an alkaline media serves as the key step to construct the biphenylb

Full text of DOI:10.1016/j.bmcl.2010.07.028

Bioorganic & Medicinal Chemistry Letters 20 (2010) 5095–5098
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Concise synthesis of Cannabisin G
a,
Dawei Li a,c, Wenling Li a,b, Qian Wang a, Zhaoqi Yang c, , Zijie Hou
*
*
a Department of Chemistry, State Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, PR China
b School of Chemical and Biological Engineering, Lanzhou Jiaotong University, Lanzhou 730070, PR China
c Division of Chemistry and Biological Chemistry, School of Physical and Mathematical Sciences, Nanyang Technological University, Singapore 637371, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history:
Cannabisin G (1), a naturally occurring lignanamide, was synthesized in 45% overall yield starting from 3-
tert-butyl ethyl ferulate (6). An oxidative coupling by potassium ferricyanide in an alkaline media serves
as the key step to construct the biphenylbutadiene skeleton of 1 with high regioselectivity.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 21 June 2010
Revised 4 July 2010
Accepted 7 July 2010
Available online 11 July 2010
Keywords:
Cannabisin G
Lignanamide
Regioselectivity
Oxidative coupling
Cannabisin G (1), a naturally occurring lignanamide, was iso-
lated from the fruits of Cannabis sativa in 1995.1 It has been used
as a purgative in China and Japan. As shown in Figure 1, structur-
ally, Cannabisin G and its other analogs such as Cannabisin D
(2),2 grossamide (3)3 and Cannabisin E (4),1 are the dimeric forms
of N-trans-feruloyltyramine (5), formed via a single-electron oxida-
tive coupling in vivo. The most concise biosynthetic approach is the
direct oxidative dimerization of 5 (Fig. 2). However, literature re-
ports1,4 indicated that this approach was complicated due to the
lack of regioselectively as several coupling modes5 such as 8–5,
8–8 and 8-O-4 couplings could take place; furthermore, cyclization
reactions could also occur making this approach impractical. As an
example, the 8–8 coupling with further cyclization produces 2
while the similar 8–5 coupling produces 3; both 2 and 3 were
apply the same strategy to synthesize Cannabisin G with outstand-
ing regioselectivity and high yield.
As discussed above, there are two problems associated with the
oxidative coupling of N-trans-feruloyltyramine: one is the regiose-
lectivity of the oxidative coupling and another is the subsequent
cyclization. We hope that the application of our tert-butyl protect-
ing strategy would improve the yields of the desired 8–8 coupled
products. When examining the conditions reported in the litera-
ture,1,4 we discovered that all the reactions were carried out in
acidic (FeCl3 and Cu(NO3)2ꢀ3H2O) or neutral (horseradish peroxi-
dase) conditions to produce the acyclic lignanamides including
the desired 1, but the subsequent cyclizations consumed the ini-
tially formed 1 making the reaction mixture complex.1 We specu-
lated that the 8–8 coupled product of quinone can give the 1 and 2
under alkaline and acidic conditions respectively, on the other
hand, 1 is probably unstable under acidic conditions thus rearrange
to 2 as shown in Figure 3. Therefore, we planned to carry out the
coupling in alkaline conditions using potassium ferricyanide as
the oxidant.
1
the dominant products formed when FeCl3 was used as the re-
agent. The use of other oxidants such as horseradish peroxidase
(HRP)-H2O2,4 and Cu(NO3)2ꢀ3H2O4 gave similar results. In all cases,
only trace amounts of 1 were detected. Up till now, no other syn-
thetic pathway toward 1 has been reported.
Recent work in our laboratory has been directed toward the
synthesis of important natural products such as lignans6,7 and stil-
bene dimers8 using regioselective phenolic couplings. We intro-
duced a tert-butyl as the protecting group into our coupling
precursors which has greatly improved the 8–8 couplings in the
lignan and stilbene dimer syntheses and significantly enhanced
the yields of the desired products. In this study, we continued to
As depicted in Figure 4, our initial strategy involved the direct
oxidative coupling of amide 7 that bears a tert-butyl group at C-3
and a subsequently debutylation of coupled dimer 8. Compound
7 was successfully prepared from the 3-tert-butyl ferulic acid (6)7
with tyramine hydrochloride using DCC as the reagent. Unfortu-
nately, oxidative coupling reaction of 7 using alkaline potassium
ferricyanide gave a complicated mixture, possibly due to the extra-
neous active phenolic substitute on the tyramine moiety. We
therefore abandoned this route.
A revised synthetic strategy is formulated and illustrated in
Figure 5. In the new route, we decided to use 3-tert-butyl ethyl
* Corresponding authors. Tel.: +65 8349 6232.
E-mail addresses: Yang0213@ntu.edu.sg (Z. Yang), houzj@lzu.edu.cn (Z. Hou).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.07.028

5096
D. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5095–5098
OH
OH
OH
O
O
O
MeO
N
H
MeO
HO
N
H
H
HO
HO
N
O
H
N
MeO
OMe
OH
OH
OH
2
Cannabisin D ( )
Cannabisin G (1)
O
OH
N
H
OH
O
MeO
MeO
MeO
HO
N
H
H
O
N
O
H
N
O
OH
MeO
O
OH
OH
Cannabisin E (4)
Grossamide (3)
Figure 1. Chemical structures of lignanamides 1, 2, 3, and 4.
8-5 coupled product-3 (13%)
8-8 coupled product-2 (20%)
8-8 coupled product-1 (0.34%)
8-O-4 coupled product-4 (0.53%)
FeCl3
OH
2''
O
2
7
α
MeO
HO
N
8
6''
HRP-H2O2
H
β
8-5 coupled product-3 (18%)
8-8 coupled product-2 (10%)
6
N-trans-feruloyltyramine (5)
Cu(NO3)2 • 3H2O
8-5 coupled product-3 (14%)
8-8 coupled product-2 (7%)
Figure 2. Product distribution profile from oxidation of 5.1,4
R
R
R
R
8
H
7
H
H
H
H
H
H
H
-H+, -e
2
6
OMe
OMe
OMe
OMe
OH
O
O
O
MeO
HO
OH
R
MeO
O
R
alkaline conditions
R
OMe
1
R
8-8 coupling
H+
(taking no account of
other coupling modes)
MeO
HO
R
acidic conditions
OH
OMe
O
R
O
2
R=
OMe
N
H
OH
Figure 3. Purposed mechanism for 8–8 coupling in alkaline and acidic conditions.5
ferulate (9) as the coupling precursor and subsequently introduce
the tyramine moieties to the coupled intermediate. Compound 9
was reported in our previous work.6 As expected, the oxidation
of 9 in a benzene–water two-phase system using alkaline potas-
sium ferricyanide as the oxidation reagent afforded regioselec-
tively the desired 8–8 coupling product 10 in excellent yield. The
tert-butyl group was then successfully removed from 10 via a Fri-
edel–Crafts reaction using AlCl3 as the regent in moderate yield.9

D. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5095–5098
5097
OH
OH
O
+
O
MeO
HO
MeO
HO
OH
N
H
6
O
NH
t-Bu
t-Bu
t-Bu
debutylation
coupling
OH
condensation
1
HO
NH3Cl
H
N
HO
MeO
MeO
t-Bu
O
(+) tyramine hydrochloride
OH
8
7
Figure 4. A failed trial path to prepare 1.
O
O
MeO
OEt
MeO
OEt
O
OEt
HO
HO
HO
°
AlCl3, C6H6, 50 C
t-Bu
t-Bu
K3[Fe(CN)6], KOH
C6H6/ H2O, r.t.
60%
HO
OEt
O
MeO
92%
t-Bu
MeO
OEt
MeO
OH
9
11
O
10
3''
OH
2''
O
a
2
7
O
MeO
8
N
H
MeO
6''
β
OH
DCC, TEA/THF, reflux
6
HO
HO
3
3'
KOH, EtOH/H2O, reflux HO
2'
HO
( )tyramine hydrochloride
86%
H
N
β'
2'''
95%
3'''
OH
MeO
8'
OH
6' 7'
a'
MeO
O
6'''
O
12
Cannabisin G ( )
1
Figure 5. The optimized path to synthesize 1.
The ethyl esters of 11 were saponified using KOH in refluxing aque-
ous EtOH to produce the intermediate diacid 12 in 95% yield. Con-
densation of the diacid with tyramine hydrochloride in refluxing
THF using DCC as the coupling reagent afforded the targeted prod-
uct 1 in high yields. All spectral data of 1 obtained were in good
agreement with those reported in literature.1
In conclusion, a facile method for the synthesis of Cannabisin G
(1) was reported by using a highly regioselective coupling reaction,
the benefit of using tert-butyl as a transient protecting group is
cleared demonstrated by suppressing the undesired coupling
modes. Furthermore, we have effectively eliminated the subse-
quent cyclizations by conducting the coupling in an alkaline media
using potassium ferricyanide as the reagent. The overall yield was
45%. We expect that the new synthetic strategy could be applied to
the synthesis of other lignanamides with similar structures.
General experimental procedures
10. Structural determinations of the isolated compounds were based on 1H, 13C
NMR, MS, IR, and UV analysis. All NMR spectra were recorded on a Varian
Mercury 300 or 400 MHz instrument. EIMS (70 eV) spectra were obtained on a
VG ZAB-HS spectrometer. FAB HRMS spectra were obtained on an Autostec-
3090 mass spectrometer. Melting points were measured on a Kofler hot stage
without calibration. IR spectra were obtained on a Nicolet NEXUS 670 FT-IR
spectrometer. UV spectra were measured using TU-1901 UV–vis
spectrophotometer. All solvents were ACS grade or higher and freshly
purified and dried by standard techniques prior to use as needed. Silica gel
(200–300 mesh) was purchased from QingDao Marine Chemical Co. (QingDao,
China).
Synthesis of E-3-(3-tert-butyl-4-hydroxy-5-methoxyphenyl)-N-2-[4-hydroxyl-
phenylethyl]-2-propenamide (7): TEA (5 mL) was added dropwise to
suspension of ( ) tyramine hydrochloride (345 mg, 0.20 mmol) in THF
(20 mL), then compound (500 mg, 0.20 mmol) and DCC (412 mg,
a
6
0.20 mmol) were added to the solution. The reaction mixture was stirred at
room temperature for 2 h, and then the solvent was evaporated in vacuo. The
residue was purified by column chromatography using petroleum ether and
ethyl acetate (3:1, v/v) as eluent to afford 7 (605 mg, 82%) as an amorphous
solid; 1H NMR (CD3COCD3, 300 MHz) d 1.39 (9H, s, C(CH3)3), 2.76 (2H, t,
References and notes
J = 7.2 Hz, H-b), 3.51 (2H, dd, J = 7.2, 13.2 Hz, H-a), 3.86 (3H, s, OCH3), 6.54 (1H,
d, J = 15.6 Hz, H-7), 6.77 (2H, d, J = 8.1 Hz, H-300, H-500), 7.06 (2H, d, J = 8.1 Hz, H-
200, H-600), 7.07 (2H, s, H-2, H-6), 7.37 (1H, br t, J = 13.2 Hz, NH), 7.49 (2H, d,
J = 15.6 Hz, H-7); 13C NMR (CD3COCD3, 75 MHz) d 30.1 (CH3, C(CH3)3), 35.2 (C,
1. Sakakibara, I.; Ikeya, Y.; Hayashi, K.; Okada, M.; Maruno, M. Phytochemistry
1995, 38, 1003.
2. Sakakibara, I.; Ikeya, Y.; Hayashi, K.; Mitsuhashi, H. Phytochemistry 1992, 31,
3219.
3. Yoshihara, T.; Yamaguchi, K.; Takamatsu, S. Agric. Biol. Chem. 1981, 45, 2593.
4. Lajide, L.; Escoubas, P.; Mizutani, J. Phytochemistry 1995, 40, 1105.
5. Sarkanen, K. V.; Wallis, A. F. J. Chem. Soc., Perkin Trans. 1 1973, 1869.
6. Wang, Q.; Yang, Y.; Li, Y.; Yu, W.; Hou, Z. J. Tetrahedron 2006, 62, 6107.
7. Zhu, F. Q.; Li, W. L.; Wang, Q.; Hou, Z. J. Synlett 2006, 1780.
8. Li, W. L.; Li, H.; Li, Y.; Hou, Z. J. Angew. Chem., Int. Ed. 2006, 45, 7609.
9. Tashiro, M.; Yamato, T. J. Org. Chem. 1979, 44, 3037.
C(CH3)3), 35.7 (CH2, C-b), 41.9 (CH2, C-a), 56.3 (CH3, OCH3), 108.4 (CH, C-2),
116.0 (CH, C-300, C-500), 119.4 (CH, C-6), 120.4 (CH, C-8), 126.6 (C, C-1), 130.4
(CH, C-200, C-600), 130.9 (C, C-100), 136.1 (C, C-5), 141.1 (CH, C-7), 147.3 (C, C-2),
148.4 (C, C-3), 156.7 (C, C-400), 166.6 (C, C@O); EIMS m/z: 369 (M+), 354, 260,
151,137.
Synthesis of E-3-(3-tert-butyl-4-hydroxy-5-methoxyphenyl)-propenoate (9):
Compound
9 was prepared in three steps in 65% overall yield using
commercially available creosol as the starting material.6

5098
D. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5095–5098
Synthesis of diethyl (E,E)-bis(3-tert-butyl-4-hydroxy-5-methoxybenzylidene)
was removed under reduced pressure and the syrup was diluted with H2O
(5 mL). The mixtures were acidified to pH 1 using HCl (2.0 N) in an ice-bath.
White solid precipitated from the solution. The solid was extracted with ethyl
ether (5 ꢂ 15 mL). Evaporation of the ether solution afforded the crude product
which was recrystallized from EtOH–H2O to yield the diacid 12 (0.88 g, 96%) as
white needles. Mp: 171–172 °C; 1H NMR (DMSO, 300 MHz) d 3.64 (6H, s,
OCH3), 6.71 (2H, dd, J = 4.8, 5.1 Hz, H-5, H-50), 7.01 (2H, dd, J = 1.8, 4.8 Hz, H-6,
H-60), 7.19 (2H, d, J = 1.8 Hz, H-2, H-20), 7.65 (2H, s, H-7, H-70), 9.56 (2H, s,
COOH); 13C NMR (DMSO, 75 MHz) d 55.9 (CH3, OCH3), 113.8 (CH, C-2, C-20),
116.2 (CH, C-5, C-50), 124.8 (CH, C-6, C-60), 125.6 (C, C-1, C-10), 126.8 (C, C-8, C-
80), 138.3 (CH, C-7, C-70), 147.9 (C, C-4, C-40), 149.0 (C, C-3, C-30), 169.0 (C,
C@O); HRMS: m/z 387.1072 (calcd for C20H18O8+H, 387.1074).
Synthesis of Cannabisin G (1): Triethylamine (5 mL) was added dropwise to a
suspension of ( ) tyramine hydrochloride (224 mg, 0.13 mmol) in THF (20 mL),
then compound 12 (500 mg, 0.13 mmol) and DCC (268 mg, 0.13 mmol) were
added. The reaction mixture was heated to reflux for 2 h. The solvent was
removed under reduced pressure, and the residue was purified by column
chromatography using petroleum ether and ethyl acetate (2:1, v/v) as eluent to
succinate (10): Compound 9 (200 mg, 0.72 mmol) in benzene (7.2 mL) was
vigorously stirred with an aqueous solution (1.45 mL) containing potassium
ferricyanide (600 mg) and potassium hydroxide (220 mg) for 0.5 h under
nitrogen. The organic layer was washed with water, brine, and dried over
MgSO4. The solvent was evaporated under vacuum and the residue was
purified by column chromatography using petroleum ether and ethyl acetate
(5:1, v/v) as eluent to afford 10 (183 mg, 92%) as a yellow oil; IR (KBr) mmax
:
3412, 2957, 1700, 1366, 978 cmꢁ1 1H NMR (CDCl3, 300 MHz) d 1.10 (6H, t,
;
J = 6.9 Hz, CH3CH2), 1.34 (18H, s, C(CH3)3), 3.77 (6H, s, OCH3), 4.14 (4H, q,
J = 6.9 Hz, CH3CH2), 6.17 (2H, s), 7.03 (2H, s), 7.11 (2H, s), 7.85 (2H, s, H-7, H-70).
EIMS m/z 554 (M+), 278, 57. HRMS: m/z 577.2790 (calcd for C32H42O8 + Na,
577.2772).
Synthesis of diethyl (E,E)-bis(4-hydroxy-3-methoxybenzylidene) succinate (11): To
a solution of compound 10 (1.8 g, 0.32 mmol) in dry benzene (40 mL) was
added AlCl3 (8 equiv, 3.5 g) at 50 °C. The reaction mixture was stirred for 1 h,
and then quenched with ice-water. After extraction with benzene, the
combined organic layers were washed with brine and dried over Na2SO4.
After evaporation under vacuum, the residue was purified by column
chromatography using petroleum ether and ethyl acetate (3:1, v/v) as eluent
to afford 11 (858 mg, 60%) as an amorphous solid; 1H NMR (CDCl3, 300 MHz) d
1.11 (6H, t, J = 6.9 Hz, CH3CH2), 3.74 (6H, s, OCH3), 4.15 (4H, q, J = 6.9 Hz,
CH3CH2), 6.84 (2H, d, J = 8.1 Hz, H-5, H-50), 7.06 (2H, d, J = 8.1 Hz, H-6, H-60),
7.86 (2H, s, H-2, H-20); 13C NMR (CDCl3, 75 MHz) d 13.9 (CH3, CH3CH2O), 55.5
(CH2, CH3CH2O), 60.9(CH3, OCH3), 111.3 (CH, C-2, C-2’), 114.5 (CH, C-5, C-50),
124.6 (CH, C-6, C-60), 124.5 (C, C-1, C-10), 127.1 (C, C-8, C-80), 142.2 (CH, C-7, C-
70), 146.4 (C, C-4, C-40), 147.3 (C, C-3, C-30), 167.3 (C, C@O); EIMS m/z 422 (M+),
296, 260, 151, 137.
afford 1 (694 mg, 86%) as a pale yellow solid; IR (KBr)
1560 cmꢁ1 1H NMR (CD3COCD3, 400 MHz) d 2.42 (2H, ddd, J = 13.6, 8.4, 2.0 Hz,
H-b), 2.52 (2H, ddd, J = 13.6, 8.4, 2.0 Hz, H-b0), 3.25 (2H, dddd, J = 13.6, 6.4, 8.4,
mmax: 3360, 2948, 1658,
;
2.0 Hz, H-a
), 3.48 (2H, dddd, J = 13.6, 8.4, 2.0 Hz, H-a0), 3.74 (6H, s, OCH3), 6.68
(4H, d, J = 8.4 Hz, H-300, H-500, H-3000, H-5000), 6.83 (2H, d, J = 8.0 Hz, H-5, H-50),
6.84 (4H, d, J = 8.4 Hz, H-200, H-2000, H-600, H-6000), 7.08 (2H, dd, J = 8.4, 2.0 Hz, H-6,
H-60), 7.12 (2H, br t, J = 5.6 Hz, NH), 7.28 (2H, d, J = 2.0 Hz, H-2, H-20), 7.88 (2H,
13
s, H-7, H-70); C NMR (CD3COCD3,100 MHz) d 35.3 (CH2, C-b, C-b0), 42.6 (CH2,
C-a
, C-a0), 56.0 (CH3, OCH3), 113.2 (CH, C-2,C-20), 116.0 (CH, C-5, C-50), 125.8
(CH, C-6, C-60), 127.9 (C, C-8, C-80), 130.3 (CH, C-200, C-2000, C-600, C-6000), 130.8 (C,
C-100, C-1000), 140.3 (CH, C-7, C-70), 148.2 (C, C-4, C-40), 149.1 (C, C-3, C-300), 156.5
(C, C-400, C-4000), 166.0 (C, C@O); HRMS: m/z 623.2390 (calcd for C36H36N2O8ꢁH,
623.2399).
Synthesis of 4,40-dihydroxy-3,30-dimethoxy-b,b0-bicinnamic acid (12): Compound
11 (1.0 g, 0.23 mmol) was dissolved in EtOH–H2O (4:1, v/v, 50 mL) and KOH
(0.39 g, 3.5 equiv) was added. The reaction mixture was refluxed for 8 h. EtOH