Total synthesis of (±)-cocculolidine

Article abstract of DOI:10.1016/S0040-4039(01)01650-1

A non-aromatic erythrina alkaloid, cocculolidine (1), which has insecticidal activity, was synthesized efficiently employing coupling reaction of imine with tetronic acid and intramolecular 1,6-addition as the key steps.

Full text of DOI:10.1016/S0040-4039(01)01650-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 8003–8006
Total synthesis of ( )-cocculolidine
Tsuneomi Kawasaki, Naoko Onoda, Hidenori Watanabe and Takeshi Kitahara*
Department of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, The University of Tokyo,
1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
Received 25 July 2001; revised 29 August 2001; accepted 31 August 2001
Abstract—A non-aromatic erythrina alkaloid, cocculolidine (1), which has insecticidal activity, was synthesized efficiently
employing coupling reaction of imine with tetronic acid and intramolecular 1,6-addition as the key steps. © 2001 Elsevier Science
Ltd. All rights reserved.
In 1966, Wada et al. isolated cocculolidine (1) from
Cocculus trilobus DC as an insecticide.1–4 It belongs to
erythrina alkaloids which occur in the seeds and plant
part of the erythrina species and causes paralysis of
smooth muscles similar to the effect of curare.5–7 Ery-
thrina alkaloids are classified into two groups according
to their structural features; those whose D-rings are
aromatic, and the others whose D-rings are an unsatu-
rated lactone. The synthesis of aromatic erythrina alka-
loids has been achieved by many groups, but the
synthesis of non-aromatic erythrina alkaloids had not
been reported yet.8 In this paper we describe the first
total synthesis of cocculolidine (1), a non-aromatic
erythrina alkaloid.
could be transformed into cocculolidine (1) via intro-
duction of olefin and methoxy groups using the car-
bonyl group in the A-ring. Compound A would be
constructed by intramolecular 1,6-addition of B which
would be given from C by vinylation. Compound C
would be synthesized by a coupling reaction of tetronic
acid with imine D which might be derived from 4-
hydroxycyclohexanone.
The synthesis of imine (6) is outlined in Scheme 2. We
started
from
known
4-t-butyldimethylsilyloxy-
cyclohexanone9 (2) and it was alkylated with bromoace-
tonitrile to give nitrile (3) (ca. 1:1 cis/trans) in good
yield. Both isomers were separable. We employed both
isomers for the successive steps without separation,
because in the later stage this hydroxyl group will be
oxidized to ketone. The carbonyl group was protected
Our retrosynthetic analysis is illustrated in Scheme 1.
We decided to utilize the key intermediate A, which
B
N
N
NH
RO
A
C
O
D
O
MeO
O
intramolecular
O
O
1,6-addition
O
O
Cocculolidine 1
A
B
OH
O
O
NH
O
O
RO
N
OH
RO
coupling reaction
of imine with
tetronic acid
O
OH
C
D
Scheme 1. Retrosynthetic analysis of cocculolidine 1.
Keywords: erythrina alkaloids; cocculolidine; coupling reaction of imine with tetronic acid; intramolecular 1,6-addition.
* Corresponding author. Fax: 81-3-5841-8019; e-mail: atkita@mail.ecc.u-tokyo.ac.jp
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01650-1

8004
T. Kawasaki et al. / Tetrahedron Letters 42 (2001) 8003–8006
O
O
O
O
NH2
a)
b)
c)
CN
OTBS
O
OTBS
2
3
4
OTBS
N
H
N
CF3
d)
O
TBSO
5
6
OTBS
Scheme 2. (a) (i) LDA, bromoacetonitrile, THF −78 to 0°C (70% based on recovered 2); (b) (i) TsOH, ethylenglycol, benzene
reflux; (ii) LAH, AlCl3, Et2O 0°C; (c) (i) TFAA, pyr., CH2Cl2 0°C; (ii) TsOH, acetone reflux; (d) 5N NaOH, MeOH (five steps
in 52%).
as an ethylenacetal, and reduction of nitrile with alu-
minum hydride afforded the corresponding amine (4).
At first, we tried removal of ethylenacetal from amine
(4), but the undesired removal of the TBS group
occurred. On the other hand, after protection of the
amino group with a trifluoroacetyl group, deprotection
of ethylenacetal with catalytic TsOH in acetone was
successful. The next intramolecular condensation took
place to give imine (6) when the trifluoroacetyl group
was removed.
Next, we examined functionalization of the A-ring, as
shown in Scheme 4. At first, we attempted the enoliza-
tion of 11 by several methods, but the desired com-
pounds were not obtained. So, after ketone (11) was
transformed to dimethylacetal, elimination of methanol
was executed by refluxing in o-dichlorobenzene to give
methyl enyl ether (12, 74%) and its regioisomer (14%).
The regioselectivity was changed to give methyl enyl
ether (12, 4%) and its regioisomer (76%) by using
TMSOTf and DIPEA.13 Both isomers were separable
by silica gel column chromatography. We attempted
the conversion of 12 to a-selenoketone by treatment
with phenylselenenyl chloride14 followed by enolization
The coupling reaction10 of imine (6) with tetronic acid
(7) was a key reaction, as shown in Scheme 3. To an
ethereal solution of imine (6), a solution of tetronic acid
in acetonitrile was added dropwise. The coupling reac-
tion proceeded quickly, and the product (8) was precip-
itated as a white solid. So equilibration tends toward
product (8), which was obtained by filtration in 91%
yield. Secondary amine (8) was protected as t-butylcar-
bamate, and the enolic hydroxyl group was converted
into trifluoromethanesulfonate which was transformed
to a vinyl group under Stille’s condition11 to give 9. The
next intramolecular 1,6-addition of amine was also a
key reaction. In the presence of TFA, the removal of
t-butylcarbamate and subsequent intramolecular 1,6-
addition proceeded successively to construct C-ring.
Then water was added to the reaction mixture to give
alcohol (10) in nearly quantitative yield. This alcohol
was oxidized to key intermediate (11) with TPAP and
NMO.12 X-Ray analysis of (11) is shown in Fig. 1.
Figure 1. X-Ray analysis of 11.
OH
NH
O
7
a)
O
TBSO
N
OH
TBSO
TBSO
MeCN-Et2O
O
O
6
8
NBoc
O
N
N
TFA (neat)
b)
HO
O
then H2O
(5 steps, 52%)
O
O
O
O
O
9
10
11
Scheme 3. (a) (i) Boc2O, cat. DMAP, CH2Cl2 rt; (ii) Tf2O, pyr., CH2Cl2 −78 to rt 8 h; (iii) n-Bu3SnCHꢀCH2, (Ph3P)2PdCl2, DMF
,
35°C 4 h; (b) TPAP, NMO, 4 A MS, CH2Cl2–MeCN (10:1) 0°C to rt (84%).

T. Kawasaki et al. / Tetrahedron Letters 42 (2001) 8003–8006
8005
PhSe
MeO
1
N
N
N
a)
b)
2
O
MeO
3
O
O
O
O
O
O
O
11
12
13
N
N
N
c)
d)
O
MeO
+
HO
HO
MeO
O
O
O
O
O
14
15
(±)-1
Scheme 4. (a) (i) TsOH, CH(OMe)3, MeOH reflux (quant); (ii) o-dichlorobenzene reflux (74%); (b) PhSeCl, DIPEA, CH2Cl2
(70%); (c) OsO4, H2O2, THF 0°C then NMO 40°C (52%); (d) (i) TsOH, ethanedithiol, MeOH reflux (90%); (ii) Ac2O, pyr (95%);
(iii) Raney-Ni, THF (47%); (iv) K2CO3, MeOH (72%); (v) CH2N2, SiO2, CH2Cl2–Et2O (34%).
to give compound (13). Although a-selenoketone was
obtained, it could not be transformed into 13. So
phenylselenylation was carried out in the presence of
DIPEA, and in this case concomitant deprotonation of
the 3-position occurred to give 1315 in good yield.
Selenide (13) was oxidized to selenoxide with a catalytic
amount of OsO4 and 1 equivalent of hydrogen perox-
ide. After the formation of selenoxide, 2.2 equivalents
of NMO was added to give enone (14)15 in 52% yield
via dihydroxylation along with a side product (15)15 in
42% yield via sigmatropic rearrangement of selenoxide.
We tried to suppress this side reaction, but the yield of
the desired 14 did not exceed 52%. All attempts for the
methylation of the alcohol (14) to afford methyl ether
were unsuccessful. Therefore, 14 was converted to the
corresponding dithioacetal for removal of the ketone
carbonyl group. The hydroxyl group was protected
with an acetyl group to avoid the loss by absorption on
Raney-Ni in the next reduction. After the desulfuriza-
tion, the acetyl group was removed to give demethyl-
cocculolidine, and we applied various conditions of
methylation to this hydroxyl group. Under basic condi-
tions, the desired cocculolidine (1) was not obtained
and presumably tetra alkyl ammonium salt was formed
because of high nucleophilicity of the tertiary amine.
Fortunately, methylation was successful only under an
References
1. Wada, K.; Marumo, S.; Munakata, K. Tetrahedron Lett.
1966, 42, 5179.
2. Wada, K.; Munakata, K. Agric. Biol. Chem. 1967, 31,
336.
3. Wada, K.; Marumo, S.; Munakata, K. Agric. Biol. Chem.
1967, 31, 452.
4. Wada, K.; Marumo, S.; Munakata, K. Agric. Biol. Chem.
1968, 32, 1187.
5. Unna, K.; Kniazuk, M.; Greslin, J. G. J. Pharmacol.
Exp. Therap. 1944, 80, 39.
6. Megirian, D.; Leary, D. E.; Slater, I. H. J. Pharmacol.
Exp. Therap. 1955, 113, 212.
7. Lullmann, H.; Mondon, A.; Seidel, P. R. Arch. Exp.
Pathol. Pharmakol. 1967, 258, 91.
8. Tsuda, Y.; Sano, T. The Alkaloids 1996, 48, 249 and
references cited therein.
9. Pitt, C. G.; Gu, Z.; Ingram, P.; Hendren, R. W. J. Polym.
Sci., Part A: Polym. Chem. 1987, 25, 955.
10. Maischein, J.; Vilsmaier, E. Liebigs Ann. Chem. 1988,
355.
11. Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508.
12. Ley, S. V.; Norman, J.; Grifith, W. P.; Marsden, S. P.
Synthesis 1994, 639.
13. Gassman, P. G.; Burns, S. J. J. Org. Chem. 1988, 53,
5574.
16
acidic condition using CH2N2 and SiO2 to give coccu-
lolidine (1) in 34% yield. The synthetic ( )-1 was com-
14. Nicolaou, K. C.; Magolda, R. L.; Sipi, W. J. Synthesis
pletely identical with natural authentic sample.17
1979, 982.
15. Although compounds (13), (14) and (15) were produced
as single isomers, the related stereochemistry of them
were not decided exactly. But we predicted that they had
the stereochemistry shown in Scheme 4, because PhSeCl
and OsO4 approached from the less hindered face. Details
will be discussed in a full account.
In conclusion, we have completed the first total synthe-
sis of ( )-cocculolidine in 0.42% overall yield from
4-t-butyldimethylsilyloxycyclohexanone (2) through 21
steps. Work is now under way to improve the whole
scheme of this synthesis, and the results will be reported
in a full account.
16. Ohno, K.; Nishiyama, H.; Nagase, H. Tetrahedron Lett.
1979, 4405.
1
17. IR and H NMR (60 MHz) charts were kindly provided
Acknowledgements
by Professor Yoji Sakagami. IR data of natural and
1
synthetic samples were completely identical. Although H
NMR (300 MHz) data was consistent with its structure, it
was impossible to compare with that (60 MHz) of a
natural sample. We did not have full assurance. So, we
isolated natural cocculolidine (250 mg) from fresh leaves
of Cocculus trilobus DC (500 g) under a reported
We sincerely thank Professor Yoji Sakagami, Nagoya
University, for a kind gift of the spectral charts of
natural cocculolidine. We are much indebted to Mr.
Masahiko Bando, Otsuka Pharmaceutical Co., Ltd., for
X-ray analysis of compound 11.

8006
T. Kawasaki et al. / Tetrahedron Letters 42 (2001) 8003–8006
1
1
procedure2 and measured H NMR (300 MHz). The H
NMR spectrum of our synthetic sample was identical
with that of cocculolidine. IR data and 300 MHz 1H
NMR data of the synthetic sample are shown below.
Hz), 1.93–2.04 (1H, m), 2.16 (1H, dd, J=5.4, 19.5 Hz),
2.27 (1H, dd, J=3.6, 10.8 Hz), 2.31–2.52 (2H, m), 2.57–
2.73 (2H, m), 2.75–2.86 (1H, m), 2.96 (1H, dt, J=3.6, 8.7
Hz), 3.19–3.40 (2H, m), 3.35 (3H, s), 4.02–4.12 (1H, m),
4.64 (1H, dd, J=0.9, 17.4 Hz), 4.73 (1H, d, J=17.4 Hz),
5.69 (1H, m).
IR (CCl4, solution): w 2930, 2836, 1766, 1653 cm−1 1H
,
NMR (300 MHz, CDCl3): l=1.53 (1H, dd, J=11.4, 12.0