A concise synthesis of (+)-cassiol

Article abstract of DOI:10.1016/S0957-4166(01)00222-1

A synthesis of the anti-ulcerogenic compound (+)-cassiol 1b with 43% overall yield has been achieved. This short and efficient synthesis features the one-pot Julia olefination reaction of lactol (S)-2 with sulfone 3b through the key intermediate (-)-4b.

Full text of DOI:10.1016/S0957-4166(01)00222-1

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 12 (2001) 1251–1253
A concise synthesis of (+)-cassiol
Mar´ıa I. Colombo, Juan Zinczuk, Mirta P. Mischne and Edmundo A. Ru´veda*
Instituto de Qu´ımica Orga´nica de S´ıntesis (CONICET-UNR), Facultad de Ciencias Bioqu´ımicas y Farmace´uticas,
Casilla de Correo 991, 2000 Rosario, Argentina
Received 27 March 2001; accepted 30 April 2001
Abstract—A synthesis of the anti-ulcerogenic compound (+)-cassiol 1b with 43% overall yield has been achieved. This short and
efficient synthesis features the one-pot Julia olefination reaction of lactol (S)-2 with sulfone 3b through the key intermediate
(−)-4b. © 2001 Elsevier Science Ltd. All rights reserved.
As depicted retrosynthetically in Scheme 1, three main
strategies were used for the synthesis of (+)-cassiol. The
one involving disconnection a, based on the assembly
During the course of a pharmacological analysis of the
aqueous extract of the dried stem bark of Cinnamomum
cassia Blume, Fukaya et al.1 isolated the serotonin-
induced anti-ulcerogenic glucoside, cassioside 1a, whose
enzymatic hydrolysis afforded the aglycone (+)-cassiol
1b, exhibiting a more potent anti-ulcer activity than
cassioside itself.
of
a chiral and adequately functionalized cyclo-
hexenone/cyclohexanone intermediate, and its coupling
with a side-chain precursor, was used by several
research groups.2–8 Disconnection bb% involves a chiral
Diels–Alder cycloaddition reaction9,10 and disconnec-
tion c, a cycloisomerization in an ene type fashion.11
A
The structural features and pharmacological activity of
cassiol have prompted extensive synthetic efforts, which
have culminated in several syntheses of 1b over the past
decade.
comparative analysis of these approaches indicates that
the sequence in which the key step is a catalyzed
enantioselective Diels–Alder reaction9 is, by far, the
most efficient with respect to the number of synthetic
steps and the overall yield (40%).
We have recently developed a rather simple sequence
for the preparation of both enantiomers of lactol 212
which, as a racemic mixture, had been used by White
et al. for the synthesis of ( ) trisporic acids.13 The
Scheme 1.
* Corresponding author. E-mail: eruveda@fbioyf.unr.edu.ar
0957-4166/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(01)00222-1

1252
M. I. Colombo et al. / Tetrahedron: Asymmetry 12 (2001) 1251–1253
Scheme 2. Reagents and conditions: (a) TBDMSOTf, Et3N, Cl2CH2, 0°C, then rt, 1 h; (b) DIBAL-H, THF, −78°C, 1 h; (c) TBAF,
THF, rt, 1 h; (d) 6N HCl, MeOH, rt, 1.5 h.
availability of this key intermediate prompted us to
study a convergent approach for the synthesis of (+)-
cassiol 1b, via an olefination sequence which has not
previously been explored.
ing the sequence used for racemic cassiol, (+)-1b was
obtained in 43% overall yield. The sequence with opti-
cally active compounds was carried out without isola-
tion and purification of the intermediates 5a, 5b and 6,
requiring only chromatographic purification of (−)-4b22
and of (+)-1b.23
Our first choice for the synthesis of 1b was to use a
simple Wittig reaction however, due to the difficulties
found in the preparation of the appropriate phospho-
nium bromide, we decided to apply the one-pot olefina-
tion reaction recently reported by Julia et al.14 and
successfully used by Kociensky et al. in the synthesis of
several natural products.15,16
The synthesis of (+)-1b described in this report, featur-
ing an excellent approach for the olefination of lactols,
is short and efficient and uses simple reactions that
allow good reproducibility and material throughput.
Although the anti-ulcerogenic activity of 1b is probably
due to the presence of a cyclohexenone moiety in the
molecule,24 the availability of (R)-2 and the easy prepa-
ration of different benzothiazolylsulfones will allow the
synthesis of (−)-1b and its analogs in order to elucidate
the effect of the configuration at C(4) and the side chain
on the pharmacological activity.
We found that by the addition of lactol ( )-2 to the
anion of the benzothiazolylsulfone 3a generated with
LDA in THF at −80°C, the desired product 4a was
obtained in only 18% yield after treatment of the crude
product with excess diazomethane and purification by
column chromatography.
Careful analysis of the reaction mixture allowed us to
identify the methyl ester of unchanged starting mate-
rial, together with products that suggest that 2 under-
goes an unwanted Cannizzaro-type reaction under these
conditions.17 However, if a solution of ( )-2 in THF
was first treated with equimolar sodium hydride and
then added to the anion of the benzothiazolylsulfones
3a or 3b,18 the coupling products 4a and 4b are, respec-
tively obtained in ca. 75% yields, without formation of
the side products.19,20
Acknowledgements
We thank Drs. Manuel Gonza´lez-Sierra and Gerardo
Burton for H NMR determinations and Dr. Jose´ A.
1
Bacigaluppo for his contributions to some aspects of
this project. This work was supported by Consejo
Nacional de Investigaciones Cient´ıficas y Te´cnicas
(CONICET), Universidad Nacional de Rosario (UNR)
and Agencia Nacional de Promocio´n Cient´ıfica y
Tecnolo´gica
With the coupling products 4a and 4b in hand, we
analyzed several approaches to ( )-1b, including its
reduction to a mixture of diastereomeric diols, selective
protection of the primary alcohol followed by oxidation
of the allylic alcohol or selective oxidation of the allylic
alcohol and adjustment of the diol protecting group,
finally we found that formation of the silyl enol ether
5a,21 from 4b, followed by reduction with DIBAL-H to
5b21 and deprotection through 6,21 gave ( )-1b in good
overall yield (Scheme 2). Starting with (S)-2 and follow-
References
1. Shiraga, Y.; Okano, K.; Akira, T.; Fukaya, C.;
Yokoyama, K.; Tanaka, S.; Fukui, H.; Tabata, M. Tet-
rahedron 1988, 44, 4703–4711.
2. Takemoto, T.; Fukaya, C.; Yokoyama, K. Tetrahedron
Lett. 1989, 30, 723–724.

M. I. Colombo et al. / Tetrahedron: Asymmetry 12 (2001) 1251–1253
1253
3. Uno, T.; Watanabe, H.; Mori, K. Tetrahedron 1990, 46,
5563–5566.
4. Taber, D. F.; Meagly, R. P.; Doren, D. J. J. Org. Chem.
1996, 61, 5723–5728.
more, since the attack of any nucleophilic base present in
the reaction medium to the neutral lactol cannot occur
under these conditions, the formation of Cannizaro-type
side products is avoided.
5. Irie, O.; Fujiwara, Y.; Nemoto, H.; Shishido, K. Tetra-
21. Compounds ( )-5a, ( )-5b and ( )-6 were fully character-
ized by 1H and 13C NMR spectroscopy. The data of ( )-6
are identical with those reported in the literature for
(−)-6.9
hedron Lett. 1996, 37, 9229–9232.
6. Maiti, S.; Achari, B.; Banerjee, A. K. Synlett 1998,
129–130.
7. Miyaoka, H.; Kajiwara, Y.; Hara, M.; Suma, A.;
Yamada, Y. Tetrahedron: Asymmetry 1999, 10, 3189–
3196.
22. Compound (−)-4b: colorless oil; [h]2D7=−25.3 (c 3,
CHCl3), e.e.=97%, the enantiomeric purity was estab-
1
lished by H NMR analysis employing the shift reagent
8. Momose, T.; Toyooka, N.; Nishio, M.; Shinoda, H.;
Fujii, H.; Yanagino, H. Heterocycles 1999, 51, 1321–
1343.
9. Corey, E. J.; Guzman-Perez, A.; Luh, T.-P. J. Am. Chem.
Soc. 1994, 116, 3611–3612.
10. Boeckman, Jr., R. J.; Liu, Y. J. Org. Chem. 1996, 61,
7984–7985.
11. Trost, B. M.; Li, Y. J. Am. Chem. Soc. 1996, 118,
6625–6633.
12. Bacigaluppo, J. A.; Colombo, M. I.; Preite, M. D.;
Zinczuk, J.; Ru´veda, E. A. Tetrahedron: Asymmetry 1996,
7, 1041–1057.
13. White, J. D.; Takabe, K.; Prisbylla, M. P. J. Org. Chem.
1985, 50, 5233–5244.
tris(3-[heptafluoropropyl-hydroxymethylene]-d-camphor-
ato)europium(III) derivative [Eu(hfc)3]; IR (neat) 1728,
1
1664, 1634, 1570 cm−1; H NMR (CDCl3, 200 MHz): l
6.27 (1H, d, J=16.4), 5.67 (1H, dd, J=16.4 and 8.2 Hz),
3.95–3.85 (2H, m), 3.80–3.60 (2H, m), 3.68 (3H, s),
2.60–2.30 (4H, m), 1.95 (1H, m), 1.87 (3H, s), 1.47 (3H,
s), 1.43 (3H, s), 1.42 (3H, s). 13C NMR (CDCl3, 50
MHz): l 197.78 (s), 175.85 (s), 151.69 (s), 133.92 (d),
132.41 (s), 128.86 (d), 97.62 (s), 63.66 (t, two carbons),
52.22 (q), 46.78 (s), 39.33 (d), 34.05 (t), 33.39 (t), 26.59
(q), 22.66 (q), 20.92 (q), 12.39 (q). HRMS calcd for
C18H26O5 (M+): 322.1780, found: 322.1784.
23. Compound (+)-1b: colorless oil, [h]3D0=+8.3 (c 0.35,
MeOH); IR (neat) 3368, 2930, 2876, 1644, 1594, 1456,
14. Baudin, J. B.; Hareau, G.; Julia, S. A.; Ruel, O. Tetra-
−1
1044 cm
;
1H NMR (D2O, 200 MHz): l 6.28 (1H, d,
hedron Lett. 1991, 32, 1175–1178.
15. Billington, R.; Jarowicki, K.; Kocienski, P.; Martin, V.
Synthesis 1996, 285–296.
J=16.3 Hz), 5.67 (1H, dd, J=16.3 and 8.4 Hz), 3.76 (1H,
d, J=11.4 Hz), 3.75 (2H, dd, A part of ABX, J=11.5
and 5.9 Hz), 3.66 (2H, dd, B part of ABX, J=11.1 and
7.0 Hz), 3.43 (1H, d, J=11.5 Hz), 2.71–2.55 (3H, m), 2.17
(1H, ddd, J=13.4, 9.8 and 6.0 Hz), 1.81 (3H, d, J=0.9
Hz), 1.74 (1H, ddd, J=13.6, 6.1 and 6.1 Hz), 1.12 (3H,
s). The signals at l 3.75 and 3.66 simplify into an AB
16. Smith, N. D.; Kocienski, P. J.; Street, S. D. A. Synthesis
1996, 652–666.
17. Colombo, M. I.; Zinczuk, J.; Mischne, M. P.; Ru´veda, E.
A. Pure Appl. Chem. 2001, 73, 623–626 and references
cited therein.
1
quartet upon irradiation at l 2.67. This H NMR spec-
1
18. Compound 3b: mp 112.8–113.3°C, H NMR (CDCl3, 200
trum is coincident with that reported by Fukaya et al. for
(+)-1b at 250 MHz.1 1H NMR (D2O, 500 MHz): l 6.28
(1H, dt, J=16.3 and 1.0 Hz), 5.67 (1H, dd, J=16.2 and
8.4 Hz), 3.76 (1H, d, J=11.4 Hz), 3.745 (1H, dd, J=11.2
and 5.9 Hz), 3.740 (1H, dd, J=11.2 and 5.9 Hz), 3.670
(1H, dd, J=11.2 and 7.1 Hz), 3.666 (1H, dd, J=11.2 and
7.1 Hz), 3.43 (1H, d, J=11.4 Hz), 2.68–2.52 (3H, m), 2.17
(1H, ddd, J=13.5, 10.4 and 5.3 Hz), 1.81 (3H, d, J=0.9
Hz), 1.75 (1H, ddd, J=13.5, 6.6 and 5.8 Hz), 1.12 (3H,
s). This 1H NMR spectrum is essentially identical with
that reported by Corey et al. for synthetic (+)-1b at 500
MHz.9 13C NMR (D2O, 50 MHz): l 207.12 (s), 164.87
(s), 139.17 (d), 134.35 (s), 131.37 (d), 70.44 (t), 64.55 (t,
two carbons), 50.30 (d), 43.18 (s), 35.89 (t), 33.28 (t),
23.02 (q), 15.63 (q). These data are consistent with those
reported by Corey et al. for (+)-1b at 100 MHz.9
MHz): l 8.24–8.20 (1H, m), 8.10–8.00 (1H, m), 7.66–7.60
(2H, m), 4.12 (2H, dd, A part of ABX, J=12.0 and 3.5
Hz), 3.81 (2H, dd, B part of ABX, J=12.0 and 4.6 Hz),
3.76 (2H, d, J=6.10 Hz), 2.41 (1H, m), 1.43 (3H, s), 1.39
(3H, s). 13C NMR (CDCl3, 50 MHz): l 165.68 (s), 152.40
(s), 136.49 (s), 127.98 (d), 127.57 (d), 125.27 (d), 122.20
(d), 98.25 (s), 63.09 (t, two carbons), 54.30 (t), 29.45 (d),
25.57 (q), 21.56 (q). Anal. calcd for C14H17NO4S2: C,
51.36; H, 5.23; N, 4.28; S, 19.58. Found: C, 51.16; H,
5.23; N, 4.35; S, 19.56%.
19. We thank Professor S. V. Ley (Cambridge) for this
helpful suggestion.
20. By treatment of 2 with sodium hydride in THF solution
at room temperature the corresponding anion is readily
formed. We believe that under these conditions an equi-
librium between the alkoxide and its open form is estab-
lished, allowing in this way fast attack of the lithiated
benzothiazolylsulfone to the carbonyl of the free alde-
hyde group, leading to the coupling product. Further-
24. Maria, A. O. M.; Donald, O.; Wendel, G. H.; Guzman,
J. A.; Guerreiro, E.; Giordano, O. S. Biol. Pharm. Bull.
2000, 23, 555–557.
.