Biomimetic synthesis of acid-sensitive (-)-caparrapi oxide and (+)-8-epicaparrapi oxide induced by artificial cyclases

Article abstract of DOI:10.1021/ol050295r

(Chemical Equation Presented) Asymmetric total syntheses of acid-sensitive (-)-caparrapi oxide (1) and (+)-8-epicaparrapi oxide (2) from farnesol (9) were achieved using Sharptess-Katsuki epoxidation and Lewis acid-assisted chiral Bronsted acid (chiral LBA)-induced polyene cyclization as key steps. Furthermore, (-)-1 could be directly synthesized from (S)-nerolidol (3) and (R)-LBA with 88% ds by reagent control which overcame substrate control, while (-)-2 was obtained from (R)-3 and (R)-LBA with >99% ds by the double-asymmetric induction.

Full text of DOI:10.1021/ol050295r

ORGANIC
LETTERS
2005
Vol. 7, No. 8
1601-1604
Biomimetic Synthesis of Acid-Sensitive
(−)-Caparrapi Oxide and
(
+
)-8-Epicaparrapi Oxide Induced by
Artificial Cyclases
Muhammet Uyanik,† Hideaki Ishibashi,† Kazuaki Ishihara,*,† and
Hisashi Yamamoto*,‡
Graduate School of Engineering, Nagoya UniVersity, Furo-cho, Chikusa,
Nagoya 464-8603, Japan, and Department of Chemistry, The UniVersity of Chicago,
5735 South Ellis AVenue, Chicago, Illinois 60637
ishihara@cc.nagoya-u.ac.jp; yamamoto@uchicago.edu
Received February 12, 2005
ABSTRACT
Asymmetric total syntheses of acid-sensitive (
−)-caparrapi oxide (1) and (+)-8-epicaparrapi oxide (2) from farnesol (9) were achieved using
Sharpless Katsuki epoxidation and Lewis acid-assisted chiral Brønsted acid (chiral LBA)-induced polyene cyclization as key steps. Furthermore,
−
(
(
−
−
)-1 could be directly synthesized from (S)-nerolidol (3) and (R)-LBA with 88% ds by reagent control which overcame substrate control, while
)-2 was obtained from (R)-3 and (R)-LBA with >99% ds by the double-asymmetric induction.
Natural bicyclic sesquiterpene ethers such as (5S,8S,10S)-
(-)- and (5R,8R,10R)-(+)-caparrapi oxides (1)1,2 and 8-epi-
caparrapi oxide (2)3 can be formally derived by biomimetic
proton-induced cyclization of (S)-(+)- or (R)-(-)-nerolidol
(3) (Scheme 1). (-)-1 has been isolated from the neutral
fraction of the essential oil of Ocotea caparrapi Nates
(Dugand).1 On the other hand, (+)-1 has been isolated from
the sponge Dysidea fragilis Montagu (family Dysideidae).2
8-Epicaparrapi oxide 2 has been isolated as a minor constitu-
ent of the defense secretion of the termite Amitermes
eVuncifer.3 Unfortunately, it has not yet been confirmed
Scheme 1. Formal Biosynthetic Routes for Bicyclic
Sesquiterpene Ethers 1 and 2
† Nagoya University.
‡ The University of Chicago.
(1) For the isolation of (-)-1, see: (a) Appel, H. H.; Brooks, C. J. W.;
Campbell, M. M. Perf. Essent. Oil Record 1967, 776-781. (b) Brooks, C.
J. W.; Campbell, M. M. Phytochemistry 1969, 8, 215-218.
(2) For the isolation of (+)-1, see: Shen, Y.-C.; Hsieh, P.-W. Chin.
Pharm. J. 1999, 51, 213-218.
(3) For the isolation of natural product 2, see: (a) Wadhams, L. J.; Baker,
R.; Howse, P. E. Tetrahedron Lett. 1974, 1697-1700. (b) Baker, R.; Evans,
D. A.; McDowell, P. G. Tetrahedron Lett. 1978, 4073-1076.
whether the absolute configuration of natural product 2 by
analogy to (3R,5R,8S,10R)-(+)-3â-bromo-8-epicaparrapi ox-
ide4 is (5R,8S,10R)-(+). According to Zefirov and co-
workers, the cyclization of (()-3 induced by 5 equiv of
10.1021/ol050295r CCC: $30.25
© 2005 American Chemical Society
Published on Web 03/12/2005

HSO3F gives (()-2 diastereoselectively (via substrate con-
trol).5 However, there have been no successful examples of
the diastereoselective cyclization of (()-3 to (()-1. Kametani
and co-workers obtained a 1:1 diastereomeric mixture of
(()-1 and (()-2 through the cyclization of â-hydroxy
phenylselenide derived from 10,11-epoxynerolidol induced
by 5.7 equiv of CF3CO2H.6 To concisely synthesize (+)-1
and (-)-1 through the polyene cyclization of (R)-3 and (S)-
3, respectively, asymmetric control with artificial cyclases
should be able to overcome substrate control, and both
enantiomers of artificial cyclases should be readily available.
Recently, we demonstrated that Lewis acid assisted chiral
Brønsted acids (chiral LBAs) prepared in situ from chiral
alcohols and tin(IV) chloride were highly effective as
artificial cyclases for the enantioselective biomimetic cy-
clization of polyprenoids.7 For example, tri-, tetra-, and
pentacyclic terpenpoids bearing a chroman skeleton give
products with up to 91% ee by enantioselective cyclization
of the corresponding 2-(polyprenyl)phenol derivatives in-
duced by chiral catechol derivative 4‚SnCl4 (Figure 1).7f We
methane at -78 °C (Table 1). Cyclization of (()-3 bearing
an acid-sensitive allylic hydroxy group gave a complex
Table 1. Double-Asymmetric Induction in the Cyclization of
(()-3 with (R)-4‚SnCl4
yielda
(%)
ArOH solvent 1+2
ratiob
(+)-1/(-)-1/
(+)-2/(-)-2
from
from
(S)-3b
(R)-3b
(-)-1/(+)-2 (+)-1/(-)-2
5c
CH2Cl2 <10 18.5:18.5:31.5:31.5
37:63
37:63
5c
toluene
0
32
13
(R)-4 CH2Cl2
(R)-4 toluene
0.4:8.2:9.9:81.5
0.4:27.5:3.7:68.4
45:55
88:12
<1:>99
<1:>99
a Isolated yield. b The ratio was determined by GC analysis (PEG and
â-DM columns). c 2-Methoxyphenol (5).
reaction mixture, and the desired trans-fused 2-oxabicyclo-
[4.4.0]decanes were obtained in less than 10% yield as a
37:63 mixture of (()-1 and (()-2, which were stable under
the reaction conditions. This diastereomeric ratio is due to
substrate control. When (R)-4 was used as a Brønsted acid
instead of 5, a 9:91 mixture of (-)-1 (91% ee) and (-)-2
(78% ee) was obtained in 32% yield. This result indicates
that (+)-2 and (-)-2 were obtained from (S)-3 and (R)-3
with 55% and >99% diastereoselectivity, respectively. In
the former case, low diastereoselectivity was observed due
to the mismatch in asymmetric induction between substrate
control and reagent control. In the latter case, high diaste-
reoselectivity was observed due to the double asymmetric
induction of substrate control and reagent control. The use
of toluene in place of CH2Cl2 lowered the chemical yield of
1 and 2 but raised their enantioselectivities to 97% ee and
90% ee. Notably, (-)-1 was obtained from (S)-3 with 88%
diastereoselectivity due to reagent control, which overcame
substrate control. The activated proton in (R)-4‚SnCl4
preferentially attacked the si face of the terminal isoprenyl
group because the OH/π interaction between (R)-4‚SnCl4 and
3 in the initial protonation step should be stronger in less
polar solvents such as toluene.7f
To improve the chemical yield of 1 or 2, (()-(E)-3,7,11-
trimethyl-6,10-dodecadiene-1,3-diol derivatives 6a-f, which
were less acid-sensitive than (()-3, were examined as
substrates for cyclization with (R)-4‚SnCl4 (Table 2). Al-
though the cyclizations of 1,3-diol 6a and 1-tert-butyldiphe-
nylsilyl ether 6b were carried out in the presence of 2 equiv
of (R)-4‚SnCl4 in toluene at -78 °C for 1 day, no desired
bicyclic ethers were obtained, probably due to the tight
bidentate chelation between the substrates and SnCl4 (entries
1 and 2). This undesirable chelation disturbs not only the
generation of (R)-4‚SnCl4 but also the internal nucleophilic
attack of the 3-hydroxy group in the final step of the
cyclization of 6.8 In the course of screening various protecting
groups for the 1-hydroxy group of 6a, we found that
1-acylates such as 1-benzoate 6e and 1-phenylacetate 6f were
effective for the cyclization of 6 and gave trans-fused
Figure 1. Artificial cyclases that are available in both enantiomeric
forms.
describe here a concise total synthesis of acid-sensitive
bicyclic sesquiterpenes (-)-1 and (+)-2 based on a bio-
mimetic pathway induced by the chiral LBAs (R)-4‚SnCl4
and (S)-4‚SnCl4.7
First, the diastereoselective cyclization of (()-3, which
was obtained commercially, was examined with 1 equiv of
the achiral LBA, 2-methoxyphenol (5)‚SnCl4, in dichloro-
(4) For the isolation of (+)-2, see: (a) Faulkner, D. J. Phytochemistry
1976, 15, 1993-1994. (b) Kato, T.; Ishii, K.; Ichinose, I.; Nakai, Y.;
Kumagai, T. J. Chem. Soc., Chem. Commun. 1980, 1106-1109.
(5) For the diastereoselective cyclization of (()-3 to (()-2, see:
Polovinka, M. P.; Korchagina, D. V.; Gatilov, Y. V.; Bagrianskaya, I. Y.;
Barkhash, V. A.; Shcherbukhin, V. V.; Zefirov, N. S.; Perutskii, V. B.;
Ungur, N. D.; Vlad, P. F. J. Org. Chem. 1994, 59, 1509-1517.
(6) (a) Kametani, T. Tetrahedron Lett. 1981, 22, 3655-3656. (b)
Kametani, T.; Kurobe, H.; Nemoto, H.; Fukumoto, K. J. Chem. Soc., Perkin
Trans. 1 1982, 1085-1087.
(7) (a) Ishihara, K.; Nakamura, S.; Yamamoto, H. J. Am. Chem. Soc.
1999, 121, 4906-4907. (b) Nakamura, S.; Ishihara, K.; Yamamoto, H. J.
Am. Chem. Soc. 2000, 122, 8131-8140. (c) Ishihara, K.; Ishibashi, H.;
Yamamoto, H. J. Am. Chem. Soc. 2001, 123, 1505-1506. (d) Ishihara, K.;
Ishibashi, H.; Yamamoto, H. J. Am. Chem. Soc. 2002, 124, 3647-3655.
(e) Kumazawa, K.; Ishihara, K.; Yamamoto, H. Org. Lett. 2004, 6, 2551-
2554. (f) Ishibashi, H.; Ishihara, K.; Yamamoto, H. J. Am. Chem. Soc. 2004,
126, 11122-11123.
1602
Org. Lett., Vol. 7, No. 8, 2005

was obtained in 65% yield (entry 9). These experimental
results indicate that the substrate control of 6 is relatively
lower than that of 3 because of little difference in the
thermodynamic stabilities of 7 and 8 (Table 1 versus Table
2). Fortunately, 7f and 8f were easily separable by column
chromatography on silica gel. In contrast, it was difficult to
separate 1 and 2 without any chemical modification.6
Table 2. Double-Asymmetric Induction in the Cyclization of
(()-6 with (R)-4‚SnCl4
(S)-6f had to be prepared to synthesize (-)-7f, which is a
synthetic precursor of (-)-caparrapi oxide 1.9 (S)-6f was
prepared with 90% ee in 91% overall yield from farnesol
(9) in three steps (Scheme 2): (a) Sharpless-Katsuki
Scheme 2
yield
(%)a
7 + 8
ratiob
(+)-7/(-)-7/
(+)-8/(-)-8
entry
R, 6
solvent
toluene
toluene
toluene
1
2
3
4
5
6
7
8
9
H, 6a
0
0
0
Si-t-BuPh2, 6b
CO-i-Bu, 6c
CO(CH2)2Ph, 6d toluene
COPh, 6e
COBn, 6f
COBn, 6f
COBn, 6f
COBn, 6f
16
21
29
78
41
65
3.7:49.3:4.5:42.5
5.8:49.2:4.0:41.0
4.0:58.0:3.4:34.6
8.5:38.5:8.7:44.3
3.1:48.9:2.6:45.4
4.0:40.0:5.0:51.0
toluene
toluene
CH2Cl2
PrCl
CH2Cl2-PrClc
a Isolated yield. b The ratio was determined by GC (PEG column) and
HPLC analyses (AD-H columns). c A 1:1 (v/v) mixed solvent.
epoxidation of 9 to (2S,3S)-(-)-epoxyfarnesol (10) with 90%
ee,10 (b) regioselective reduction of (-)-10 to (S)-6a (>99%
regioselectivity) with Red-Al (65% sodium bis(2-methoxy-
ethoxy)aluminum hydride in toluene),11 and (c) regioselective
acylation of (S)-6a with phenylacetyl chloride to (S)-6f
(>99% regioselectivity).12
The asymmetric cyclization of (S)-6f induced by 2 equiv
of (R)-4‚SnCl4 gave an 81:19 mixture of (-)-7f (>99% ee)
and (+)-8f (21% ee) in 74% yield. On the other hand, the
asymmetric cyclization of (S)-6f induced by 2 equiv of (S)-
4‚SnCl4 gave a 14:86 mixture of (-)-7f (27% ee) and (+)-
8f (98% ee) in 73% yield. These experimental results indicate
that the substrate control of 6f was much lower than the
reagent control by 4‚SnCl4. Optically pure (-)-7f and (+)-
8f were easily separated by column chromatography on silica
gel (Scheme 3).
2-oxabicyclo[4.4.0]decanes 7 and 8 (entries 5-9). Interest-
ingly, aliphatic esters such as isovalerate 6c were inert under
the same reaction conditions (entry 3), and 3-phenylpropi-
onate 6d was less reactive than 6e and 6f (entry 4). These
experimental data suggest the existence of some attractive
interaction between Sn(IV) and a phenyl group of 6e and
6f.8 The cyclization of (()-6f with (R)-4‚SnCl4 gave a
62:38 mixture of (-)-7f (87% ee) and (-)-8f (82% ee) in
29% yield (entry 6). Judging from the enantioselectivity and
chemical yield of 7 and 8, (()-6f gave slightly better results
than (()-6e (entry 5 versus entry 6). Next, the solvent effect
was investigated in the cyclization of (()-6f with (R)-4‚SnCl4
(entries 6-8): the enantioselectivity was higher in the order
CH2Cl2 , toluene < chloropropane, while the chemical yield
of 7 and 8 increased in the order toluene < chloropropane
, CH2Cl2. Thus, chloropropane was superior to toluene with
respect to both enantioselectivity and reactivity. Finally, when
a 1:1 mixed solvent of chloropropane and CH2Cl2 was used,
a 44:56 mixture of (-)-7f (82% ee) and (-)-8f (82% ee)
Scheme 3. Diastereoselective Preparation of (-)-7f and
(+)-8f from (S)-6f
(8) Predicatable chelation structures of 6 with SnCl4 are shown below.
Further studies to elucidate the existence of some attractive interaction
between Sn(IV) and a phenyl group of 6e and 6f are currently in progress
in our laboratory, and our results will be reported in due course.
Org. Lett., Vol. 7, No. 8, 2005
1603

subsequent Grieco elimination to (-)-1 through alkyl o-
nitrophenyl selenide 11.13 In the same manner, (+)-8-
epicaparrapi oxide 2 (98% ee) was obtained in 91% overall
yield from (+)-8d: (a) hydrolysis of (+)-8f to (+)-8a
(>99%), (b) o-nitrophenylselenylation of 8a (96%), and (c)
oxidative elimination of 12 to (+)-3 (95%).
Scheme 4
In summary, we have demonstrated that the chiral LBA
4‚SnCl4 is an artificial cyclase that is useful for both achiral
and chiral substrates: (-)-caparrapi oxide 1 and (+)-8-
epicaparrapi oxide 2 could be diastereoselectively synthesized
from (S)-6f by the reagent control of (R)-4‚SnCl4 and (S)-
4‚SnCl4, respectively, regardless of the chirality of (S)-6f.
Furthermore, in the cyclization of (()-3 induced by
(R)-4‚SnCl4, (-)-1 was diastereoselectively obtained from
(S)-3 by reagent control which overcame substrate control,
while (-)-2 was highly diastereoselectively obtained from
(R)-3 by the double-asymmetric induction of substrate control
and reagent control.
Acknowledgment. Financial support for this project was
provided by SORST, Japan Science and Technology Agency
(JST), JSPS.KAKENHI (15205021), and the 21st Century
COE program “Nature-Guided Materials Processing” of
MEXT. H.I. also acknowledges a JSPS Fellowship for
Japanese Junior Scientists.
Optically pure (-)-caparrapi oxide 1 was obtained in 92%
overall yield from (-)-7f in three steps (Scheme 4):
hydrolysis of (-)-7f to (-)-7a under basic conditions and
(9) For the asymmetric synthesis of (-)-1 from (-)-sclareol, see:
Barrero, A. F.; Alvarez-Manzaneda, E. J.; Chahboun, R.; Pa´iz, M. C.
Tetrahedron Lett. 1998, 39, 9543-9544.
(10) (a) Katsuki, T.; Sharpless, K. B. J. Am. Chem. Soc. 1980, 102,
5974-5976. (b) Kigoshi, H.; Ojika, M.; Shizuri, Y.; Niwa, H.; Yamada,
K. Tetrahedron 1986, 42, 3789-3792. (c) Gao, Y.; Hanson, R. M.; Klunder,
J. M.; Ko, S. Y.; Masamune, H.; Sharpless, K. B. J. Am. Chem. Soc. 1987,
109, 5765-5780. (d) Dittmer, D. C.; Discordia, R. P.; Zhang, Y.; Murphy,
C. K.; Kumar, A.; Pepito, A. S.; Wang, Y. J. Org. Chem. 1993, 58, 718-
731.
Supporting Information Available: Experimental details
and spectroscopic data. This material is available free of
charge via the Internet at http://pubs.acs.org.
OL050295R
(12) Ishihara, K.; Kurihara, H.; Yamamoto, H. J. Org. Chem. 1993, 58,
3791-3793.
(11) (a) Viti, S. M. Tetrahedron Lett. 1982, 23, 4541-4544. (b) Hyatt,
J. A.; Kottas, G. S.; Effler, J. Org. Process Res. DeV. 2002, 6, 782-787.
(13) Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org. Chem. 1976, 41,
1485-1486.
1604
Org. Lett., Vol. 7, No. 8, 2005