Novel efficient synthesis of an enantiomeric pair of tricycloillicinone

Article abstract of DOI:10.1016/S0040-4039(03)01739-8

The title synthesis was achieved starting with 3-ethoxy-2-cyclohexenone by a method featuring Davis' asymmetric hydroxylation and Pauson-Khand reaction as the key steps.

Full text of DOI:10.1016/S0040-4039(03)01739-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 6875–6878
Novel efficient synthesis of an enantiomeric pair of
tricycloillicinone
Sanae Furuya and Shiro Terashima*
Sagami Chemical Research Center, 2743-1 Hayakawa, Ayase, Kanagawa 252-1193, Japan
Received 19 June 2003; revised 10 July 2003; accepted 11 July 2003
Abstract—The title synthesis was achieved starting with 3-ethoxy-2-cyclohexenone by a method featuring Davis’ asymmetric
hydroxylation and Pauson–Khand reaction as the key steps.
© 2003 Elsevier Ltd. All rights reserved.
(−)-Tricycloillicinone (1) is a novel neurotrophic sub-
stance isolated from the wood of Illicium tashiroi by
Fukuyama et al. in 1995,1 and bears a tetracyclic frame-
work characterized by three quaternary carbons, highly
substituted 2-cyclohexenone system, and 2-unsubsti-
tuted 1,3-dioxolane ring1 (Fig. 1). The structure of 1
including its absolute stereochemistry was assigned as
shown by spectral studies as well as by taking into
account its possible biosynthetic process.1
Since the current forms of therapy focus on this issue
based on the cholinergic hypothesis of memory dys-
function,3 1 is considered to have potential as a lead for
developing a novel medicine for Alzheimer’s disease.
Accordingly, 1 has attracted much interest among
medicinal and synthetic organic chemists. While Dan-
ishefsky et al. achieved the first total synthesis of a
racemic sample of 1 by employing two types of Claisen
rearrangement and intramolecular radical cyclization as
the key steps,4 the synthesis of an optically pure sample
of 1 has not been accomplished yet.
It has been reported that 1 enhances the activity of
choline acetyl transferase (ChAT) which can catalyze
the synthesis of acetylcholine (ACh) from its precur-
sors, choline and acetyl-CoA.1,2 It is well known that
senile dimentia associated with Alzheimer’s disease is
directly correlated with reduced levels of synaptic ACh
in the cortical and hippocampal areas of the brain.2,3
In order to pave the way to preparing an optically pure
sample of 1, to disclose novel aspects of the structure–
activity of 1, and, moreover, to explore novel congeners
of 1 which may show more promising activity than 1,
an efficient synthetic route to optically pure 1 was
sought which is more efficient and flexible than that
reported for the racemic compound.4 Here, we wish to
report the first total synthesis of an enantiomeric pair
of 1 in optically pure forms accomplished by featuring
Davis’ asymmetric hydroxylation5 and Pauson–Khand
reaction6 as the key steps.
Our synthetic design of optically pure 1 is outlined in
Scheme 1 in which asymmetric hydroxylation of the
6-substituted-2-cyclohexenone 6 with the Davis reagent5
and Pauson–Khand reaction6 of eneyne 3 are employed
as the key steps. The diketone 2 produced by hydro-
genation of the Pauson–Khand product will afford 1 by
sequential introduction of two methyl groups into the
C-12 position and methylenation of the C-8 carbonyl
group (tricycloillicinone numbering). The Pauson–
Khand precursor 3 can be prepared from acetal 4 by
methylenation of the C-2 position (tricycloillicinone
Figure 1. Structure of natural (−)-tricycloillicinone (1).
Keywords: natural products; asymmetric hydroxylation; Pauson–
Khand reaction; choline acetyl transferase.
* Corresponding author. Tel.: +81-467-77-4114; fax: +81-467-77-4113;
e-mail: terashima@sagami.or.jp
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01739-8

6876
S. Furuya, S. Terashima / Tetrahedron Letters 44 (2003) 6875–6878
Scheme 1. Synthetic design of natural (−)-tricycloillicinone (1).
numbering). We planned to construct 4 by acetalization
of chiral alcohol 5. It was expected that, following the
Davis protocol, the chiral center of 5 can be introduced
with high optical integrity by means of asymmetric
hydroxylation of 6 derived from commercially available
3-ethoxy-2-cyclohexenone (7).
shown by the successful synthesis of natural (−)-tricy-
cloillicinone (1) (vide infra).11,12
With completion of the asymmetric synthesis of 5,
acetalization with simultaneous cleavage of an enol
ether system was next attempted. After experimenta-
tion, it was found that treatment of
5
with
Following the designed synthetic scheme, our approach
to optically pure 1 commenced with alkylation of 7.7 As
shown in Scheme 2, alkylation of 7 with 3-bromo-1-
(trimethylsilyl)-1-propyne took place smoothly, provid-
ing 6 in 89% yield. With 6 in hand, asymmetric
hydroxylation of 6 was next attempted. At first, taking
into account the results accumulated by Davis et al. for
asymmetric hydroxylation of the a-position of ketone,5
the lithium enolate of 6 was treated with (−)-(cam-
phorylsulfonyl)oxaziridine (ent-8) in THF at −78°C8
(Fig. 2). Unexpectedly, this reaction was found to
afford ent-5,8 16% ee,9 in 48% yield. The absolute
stereochemistry of ent-5 was definitely determined by
the successful synthesis of unnatural (+)-tricycloilli-
cinone (ent-1) from ent-5 (vide infra). In order to
improve the optical yield, the asymmetric hydroxylation
was further examined using (−)-[(8,8-dicholorocam-
phoryl)sulfonyl]oxaziridine (ent-9),10 giving rise to ent-
5, 57% ee,9 in 29% yield. After experimentation, it was
finally found that the optical and chemical yield of
ent-5 can be dramatically improved by carrying out the
reaction with ent-9 in a mixture of THF and DMF (7:
3), affording ent-5, 77% ee,9 in 44% yield. At present,
the remarkable effect of DMF can not be explained.
Since the racemic alcohol dl-5 was found to be more
prone to crystallize than ent-5, removal of dl-5 by
recrystallization from pentane gave ent-5 of 88% ee9 in
77% yield based on ent-5 of 77% ee. In the same
manner, 5, 76% ee,9 was prepared from 6 in 46% yield
by employing 9 as a chiral oxidant. Recrystallization of
this sample from pentane similarly afforded 5 of 91%
ee9 in 74% yield based on 5 of 76% ee. The absolute
stereochemistry of 5 was rigorously established as
paraformaldehyde in the presence of a catalytic amount
of dl-camphorsulfonic acid in cyclohexane afforded 4 in
17% yield with 56% recovery of 5. The recovered chiral
alcohol was again treated under the same acetalization
conditions as described above, producing 4 in 34% yield
along with 18% recovery of 5. The rather lower yield of
4 might reflect the possibility that the difference in
thermodynamic stability between 5 and 4 is small.
Being different from 5, optically active 4 was found to
more readily crystallize than dl-4. Thus, recrystalliza-
tion of 4 of 91% ee from hexane–ether cleanly gave an
optically pure sample of 4, mp 112.5–113°C and [h]2D1=
−156° (c 1.05, EtOH), >99% ee,13 in 66% yield based on
4 of 91% ee. Methylenation of 4 following the protocol
reported by Danishefsky et al.14 consisting of silyl enol
formation, alkylation with the Eschenmoser reagent,
dimethyl(methylene)ammonium iodide, quarterniza-
tion, and base-promoted elimination, furnished the
Pauson–Khand precursor 3 after desilylation. The
chemical yield of 3 was 45% over five operations.
The Pauson–Khand reaction of 3 was next examined
which constitutes another key step of our novel syn-
thetic route to 1. Thus, treatment of 3 with dicobalt
octacarbonyl in toluene at room temperature readily
produced a suspension of the cobalt complex, which on
reflux smoothly produced the Pauson–Khand product
10 in 79% yield. Hydrogenation of the double bond in
8 over 5% Pd–C gave rise to the tetracyclic diketone 215
in 94% yield. Dimethylation of 2 with methyl iodide in
the presence of potassium tert-butoxide16 took place
highly regioselectively at the sterically less hindered
C-12 position, affording dimethyl ketone 11 as a sole

S. Furuya, S. Terashima / Tetrahedron Letters 44 (2003) 6875–6878
6877
Scheme 2. Synthesis of optically pure (−)-tricycloillicinone (1). Reagents and conditions: (a) LDA, TMSCꢀCCH2Br, THF, −78°C,
2 h, 98%; (b) LDA, (+)- or (−)-[(8,8-dichlorocamforyl)sulfonyl]oxazoridine, THF–DMF, −78°C, 2 h, 46%, 76% ee, then
recrystallization from pentane, 74%, 91% ee for 5; 44%, 77% ee, then recrystallization from pentane, 77%, 88% ee for ent-5; (c)
(HCHO)n, CSA, cyclohexane, reflux, 20 min, 34% (recovery of starting material 16%), then recrystallization from hexane–Et2O,
66%, >99% ee for 4; 26% (recovery of starting material 42%), then recrystallization from hexane–Et2O, 51%, >99% ee for ent-4;
(d) LHMDS, TESCl, THF, −78°C, 1 h; (e) the Eschenmoser reagent, CH2Cl2, rt, overnight; (f) MeI, CH2Cl2–Et2O, rt, 6 h; (g)
satd NaHCO3 aq, rt, 1 h; (h) TBAF, THF, rt, 1 h, 45% (five steps) for 3; 41% (five steps) for ent-3; (i) Co2(CO)8, toluene, rt, 3
t
h, then reflux, 3.5 h, 79% for 10; 74% for ent-10; (j) H2, Pd–C, EtOH, 0°C, 1.5 h, 94% for 2; 87% for ent-2; (k) KOtBu, BuOH,
CH3I, 30°C, 1 h, 35% for 11; 56% for ent-11; (l) PPh3CH3 Br, KOtBu, THF, rt, 1 h, 56% for 1; 59% for ent-1.
described for the natural product, mp 78–79°C and
[h]D20=−5.56° (c=1.15, EtOH).1,19 With the aim to
exemplify the efficiency of our novel synthetic route,
the enantiomer of 1 (ent-1) was prepared from ent-5,
>99% ee, following the same reaction scheme as
described for the preparation of 1. An optically pure
sample of ent-1 thus obtained, showed mp 95–96°C and
[h]2D0=+29.1° (c=1.07, EtOH), >99% ee18 (vide supra).
The absolute configuration of 1 was assigned as shown
by assuming that 1 is presumably biosynthesized from
(+)-illicinone A (12) which is the main component of I.
tashiroi affording 11 (Fig. 3). On the other hand,
Figure 2. Structures of the Davis reagents.
Furukawa et al. reported that (−)-illicinone A (ent-12)
was easily isolated from Illicium arborescens20 belong-
product in 35% yield. Subsequent olefination with
ing to the same species as that of I. tashiroi. Taking
methylenetriphenylphosphrane17 furnished optically
these facts into account, our results strongly suggest
pure 1,15 mp 95–96°C and [h]2D0=−29.5° (c=1.15,
that natural 1, which is ca. 20% ee, has been isolated
1
EtOH), >99% ee.18 While H (CDCl3) and 13C (C6D6)
NMR spectra of the optically pure sample for synthetic
1 were found to be identical to those reported for
natural1 and synthetic racemic 1,4b the [h]D value and
melting point observed for our synthetic sample of
optically pure 1 were completely different from those
from the natural source.21
In summary, we have succeeded in developing a novel
synthetic route to an enantiomeric pair of optically
pure tricycloillicinone (1 and ent-1) by employing asym-
metric hydroxylation and Pauson–Khand reaction as

6878
S. Furuya, S. Terashima / Tetrahedron Letters 44 (2003) 6875–6878
(i). Thus, it had been reported that oxidation of i with 8
or 9 gives ii corresponding to ent-5 (Ref. 5).
9. The optical purity of 5 and ent-5 was rigorously deter-
mined by HPLC analysis with a chiral column (Daicel
Chiralcel OD-H, i-PrOH/hexane 90/10, flow rate 0.5
ml/min, tR 22.8 min (5) and 24.2 min (ent-5), detection at
254 nm).
10. Davis et al. reported that, depending upon the structure
of the ketonic substrate, (+)-[(8,8-dichlorocam-
phoryl)sulfonyl]oxazoridine (9) sometimes gives a chiral
alcohol which is more highly optically active than that
produced by using 8 (Ref. 5).
Figure 3. Structure of (+)-illicinone A (12).
the key steps. Considering its directness and flexibility,
the explored synthetic scheme may be amenable to a
large-scale preparation of 1 as well as applicable to
synthesis of various structural types of novel congeners
of 1. Additionally, our first total synthesis of an enan-
tiomerc pair of optically pure 1 disclosed an interesting
feature concerning the optical purity of a natural sam-
ple of 1.
11. The reaction mechanism of asymmetric hydroxylation
which, for example, provides ii and iv from i and iii,
respectively, has not hitherto been presented by Davis et
al. (Ref. 5).
Acknowledgements
12. Summing up the results collected by our studies on
asymmetric hydroxylation, it appeared that 6 should be
considered structurally more close to the 2-methoxylcar-
bonyl-1-tetralone iii than to i since iii gives iv when
oxidized with 8 or 9 (Ref. 5).
The authors are grateful to Professor Y. Fukuyama,
Faculty of Pharmaceutical Sciences, Tokushima Bunri
1
University, for providing copies of H and 13C NMR
spectra of natural 1 and for many valuable suggestions.
References
1. Fukuyama, Y.; Shida, N.; Kodama, M.; Chaki, M.;
Yugami, T. Chem. Pharm. Bull. 1995, 43, 2270.
2. (a) Yoshida, M.; Hasegawa, M.; Noda, Y.; Nabeshima,
T. Dementia 1995, 9, 233; (b) Davies, P. Brain Res. 1979,
171, 319; (c) Takadoi, M.; Katoh, T.; Ishiwata, A.;
Terashima, S. Tetrahedron 2002, 58, 9903.
3. Kaneko, I.; Iwata, N. Farmashia 2001, 37, 501.
4. (a) Pettus, T. R. R.; Chen, X.-T.; Danishefsky, S. J. J.
Am. Chem. Soc. 1998, 120, 12684; (b) Pettus, T. R. R.;
Inoue, M.; Chen, X.-T.; Danishefsky, S. J. J. Am. Chem.
Soc. 2000, 122, 6160.
5. (a) Davis, F. A.; Sheppard, A. C. Tetrahedron 1989, 45,
5703; (b) Davis, F. A.; Weismiller, M. C. J. Org. Chem.
1990, 55, 3715; (c) Davis, F. A.; Sheppard, A. C.; Chen,
B.-C.; Haque, M. S. J. Am. Chem. Soc. 1990, 112, 6679;
(d) Davis, F. A.; Kumar, A.; Chen, B.-C. J. Org. Chem.
1991, 56, 1143; (e) Chen, B.-C.; Murphy, C. K.; Kumar,
A.; Reddy, R. T.; Clark, C.; Zhou, P.; Lews, B. M.; Gala,
D.; Mergelsberg, I.; Scherer, D.; Buckley, J.; Di
Benadetto, D.; Davis, F. A. Org. Synth. 1996, 73, 159.
6. (a) Ishizaki, M.; Iwahara, K.; Niimi, Y.; Satoh, H.;
Hoshino, O. Tetrahedron 2001, 57, 2729; (b) Flethcer, A.
J.; Christie, D. R. J. Chem. Soc., Perkin Trans. 1 2000,
1657; (c) Brummond, K. M.; Kent, J. L. Tetrahedron
2000, 56, 3263; (d) Sugihara, T. Kagaku 1999, 54, 58.
7. Toyota, M.; Yokota, M.; Ihara, M. J. Am. Chem. Soc.
2001, 123, 1856.
13. The optical purity of 4 and ent-4 was rigorously deter-
mined by HPLC analysis with a chiral column (Daicel
Chiralcel OD-H, i-PrOH/hexane 97.5/2.5, flow rate 1.0
ml/min, tR 25.7 min (4) and 26.8 min (ent-4), detection at
254 nm).
14. Ng, F. W.; Danishefsky, S. J. J. Am. Chem. Soc. 2002,
124, 9812.
15. The relative stereochemistry of this compound was defin-
itely determined by single-crystal X-ray analysis. See:
Furuya, S.; Ijuin, Y.; Terashima, S., to be published.
16. Metha, G.; Sreenivas, K. Chem. Commun. 2001, 1892.
17. Fitjer, L.; Quabeck, U. Synth. Commun. 1985, 15, 855.
18. The optical purity of 1 and ent-1 was definitely deter-
mined by HPLC analysis with a chiral column (Daicel
Chiralcel AD-H, i-PrOH/hexane 90/10, flow rate 0.5
ml/min, tR 13.5 min (1) and 14.5 min (ent-1), detection at
254 nm).
19. The melting point of the synthetic racemic sample for 1
was not reported in Ref. 4b.
20. Yakushiji, K.; Tohshima, T.; Kitagawa, E.; Suzuki, R.;
Sekikawa, J.; Morishita, T.; Murata, H.; Lu, S.-T.;
Furukawa, H. Chem. Pharm. Bull. 1984, 32, 11.
21. Direct comparison of natural and synthetic 1 by HPLC
analysis18 could not be performed since the natural sam-
ple had been completely consumed for biological studies
(personal communication from Professor Y. Fukuyama).
8. (−)-(Camphorylsulfonyl)oxazoridine (ent-8) was utilized
as a chiral oxidant of the first choice by considering the
structural similarity between 6 and 2-methyl-1-tetralone