Palladium-Catalyzed Asymmetric Allylic Substitution of 2-Arylcyclohexenol Derivatives: Asymmetric Total Syntheses of (+)-Crinamine, (-)-Haemanthidine, and (+)-Pretazettine

Article abstract of DOI:10.1021/jo030309b

Much interest has been shown in Amaryllidaceae alkaloids as synthetic targets due to their wide range of biological activities. Over 100 alkaloids have been isolated from members of the Amaryllidaceae family; most of them can be classified into eight skel

Full text of DOI:10.1021/jo030309b

P a lla d iu m -Ca ta lyzed Asym m etr ic Allylic Su bstitu tion of
2-Ar ylcycloh exen ol Der iva tives: Asym m etr ic Tota l Syn th eses of
(+)-Cr in a m in e, (-)-Ha em a n th id in e, a n d (+)-P r eta zettin e
Toyoki Nishimata, Yoshihiro Sato, and Miwako Mori*
Graduate School of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812, J apan
mori@pharm.hokudai.ac.jp
Received September 30, 2003
Much interest has been shown in Amaryllidaceae alkaloids as synthetic targets due to their wide
range of biological activities. Over 100 alkaloids have been isolated from members of the
Amaryllidaceae family; most of them can be classified into eight skeletally homogeneous groups.
We have succeeded in the first asymmetric total syntheses of the crinane-type alkaloids (+)-
crinamine (1), (-)-haemanthidine (2), and (+)-pretazettine (3). The starting cyclohexenylamine 14
was obtained from allyl phosphonate 11c by palladium-catalyzed asymmetric amination in 82%
yield and with 74% ee. The product was recrystallized from MeOH. Interestingly, (-)-14 with 99%
ee was obtained from the mother liquor (74% recovery). Intramolecular carbonyl-ene reaction of
(-)-10 proceeds in a highly stereoselective manner to give hexahydroindole derivative 9 as the
sole product. In the Lewis-acid-catalyzed carbonyl-ene reaction, an interesting rearrangement
product, 20, was isolated in high yield. From 9, (+)-crinamine was synthesized. Thus, the asymmetric
total synthesis of (+)-crinamine was achieved in 10 steps from 11c, and the overall yield is 19%.
The total synthesis of (-)-haemanthidine was also achieved from 9 by a short sequence of steps.
SCHEME 1
Amaryllidaceae alkaloids constitute an important group
of naturally occurring bases, and much interest has been
shown in them due to the wide range of biological
activities they exhibit.1 Over 100 alkaloids have been
isolated from members of the Amaryllidaceae family, and
most of these compounds can be classified into eight
skeletally homogeneous groups. Crinine-type alkaloids
such as crinamine (1),2 haemanthidine (2),3 and pretazet-
tine (3)4 (Scheme 1) have a perhydroindole skeleton
connected to an aromatic ring at the ring junction. The
antineoplastic activity exhibited, in particular, by pre-
(1) (a) Martin, S. F. In The Alkaloids; Brossi, A., Ed.; Academic
Press: New York, 1987; Vol. 30, p 251. (b) Hoshino, O. In The
Alkaloids; Brossi, A., Ed.; Academic Press: New York, 1998; Vol. 51,
p 323. (c) J in, Z.; Li, Z.; Huang, R. Nat. Prod. Rep. 2002, 19, 454.
(2) Isolation of (+)-crinamine: (a) Mason, L. H.; Puschett, E. R.;
Wildman, W. C. J . Am. Chem. Soc. 1955, 77, 1253. (b) Kobayashi, S.;
Tokumoto, T.; Kihara, M.; Imakura, Y.; Shingu, T.; Taira, Z. Chem.
Pharm. Bull. 1984, 32, 3015. Total synthesis of (()-crinamine: (c)
Isobe, K.; Taga, J .; Tsuda, Y. Tetrahedron Lett. 1976, 2331.
(3) Isolation of (-)-haemanthidine: (a) Boit, H. G. Chem. Ber. 1954,
87, 1339. (b) Takagi, S.; Yamaki, M. Yakugaku Zasshi 1974, 94, 617;
Chem. Abstr. 1974, 81, 74924y. Isolation of (+)-pretazettine: (c)
Wildman, W. C.; Bailey, D. T. J . Org. Chem. 1968, 33, 3749. Total
syntheses of (()-haemanthidine and (()-pretazettine: (d) Hendrickson,
J . B.; Bogard, T. L.; Fisch, M. E. J . Am. Chem. Soc. 1970, 92, 5538. (e)
Hendrickson, J . B.; Bogard, T. L.; Fisch, M. E.; Grossert, S.; Yoshimura,
N. J . Am. Chem. Soc. 1974, 96, 7781. (f) Tsuda, Y.; Ukai, A.; Isobe, K.
Tetrahedron Lett. 1972, 3153. (g) Martin, S. F.; Davidsen, S. K. J . Am.
Chem. Soc. 1984, 106, 6431. (h) Martin, S. F.; Davidsen, S. K.;
Puckette, T. A. J . Org. Chem. 1987, 52, 1962. Formal total syntheses
of (()-haemanthidine and (()-pretazettine: (i) Ishibashi, H.; Nakatani,
H.; Iwatani, S.; Sato, T.; Nakamura, N.; Ikeda, M. J . Chem. Soc., Chem.
Commun 1989, 1767. (j) Ishibashi, H.; Uemura, N.; Nakatani, H.;
Okazaki, M.; Sato, T.; Nakamura, N.; Ikeda, M. J . Org. Chem. 1993,
58, 2360. Total syntheses of (-)-haemanthidine, (+)-prettazetine, and
(+)-tazettine: (k) Baldwin, S. W.; Debenham, J . S. Org. Lett. 2000, 2,
99.
tazettine, has stimulated interest in the synthesis of
these compounds. The total synthesis of haemanthidine
was achieved by Hendrickson.3d Several groups later
succeeded in the total syntheses of these alkaloids.2-4
Very recently, Baldwin reported the asymmetric synthe-
ses of these alkaloids.3k
We have already reported5 the asymmetric synthesis
of a 2-arylcyclohexenylamine derivative, 6, via π-allylpal-
ladium complex 5 generated from 4a , Pd(0), and (S)-
BINAPO. In the palladium-catalyzed allylic substitution
of 4 using Pd(0), the starting material 4 is racemic and
10.1021/jo030309b CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/14/2004
J . Org. Chem. 2004, 69, 1837-1843
1837

Nishimata et al.
SCHEME 3. Retr osyn th etic An a lysis of
SCHEME 2. Asym m etr ic Allylic Su bstitu tion
Cr in in e-Typ e Alk a loid s
the previous natural product syntheses,5a we used zirco-
nium-mediated cyclization for the synthesis of a hexahy-
droindole derivative. However, for the synthesis of crinine-
type alkaloids, a hydroxymethyl group of 7 must be
converted into a hydroxyl group. Thus, an alternative
route was considered. The retrosynthetic analysis of these
alkaloids is shown in Scheme 3.
If aldehyde (S)-10 was obtained from 11 and 12 using
this palladium-catalyzed asymmetric allylic substitution,
the intramolecular carbonyl-ene reaction7 of 10 would
construct a quaternary carbon center of 9 in a stereose-
lective manner via I. From this compound, the target
alkaloids 1-3 would be synthesized as chiral forms in a
short numbers of steps.
the intermediary π-allylpalladium complex 5 is meso type
(Scheme 2). Thus, a nucleophile attacks from both sides
of π-allylpalladium complex 5 to give racemic 6. However,
if a chiral ligand is used for this reaction, the intermedi-
ary π-allylpalladium complex should be a chiral form.
Thus, the nucleophile attacks preferentially from one side
to give 6 enantioselectively. Using this strategy, we
synthesized 6a in an enantioriched form from 4a . After
conversion of 6a into 6b, zirconium-mediated cyclization
of 6b was carried out, and hexahydroindole derivative 7
was obtained. From 7, (-)-mesembrane and (-)-mesem-
brine were synthesized.5a
Here, we report the first asymmetric total syntheses6
of the crinine-type alkaloids (+)-crinamine ((+)-1), (-)-
haemanthidine ((-)-2), and (+)-pretazettine ((+)-3). In
Resu lts a n d Discu ssion
Asym m etr ic Allylic Su bstitu tion of a Cycloh ex-
en ol Der iva t ive Ha vin g a n Ar om a t ic R in g a t t h e
2-P osition . Initially, 4a was used as a model compound
to synthesize a chiral cyclohexenylamine derivative, 14.
When a THF solution of 4a (1 equiv), acetal 128 (1.1
equiv), Pd2dba3‚CHCl3 (2.5 mol %), and dppb (5.0 mol %)
was stirred at 60 °C for 2 h, a cyclohexenylamine
derivative, 13, was obtained in 40% yield (Table 1, run
1). When the ligand was changed to (S)-BINAPO9and the
reaction was carried out under similar conditions, the
yield was increased to 83% and the ee10 was 76% (run
2). The ee was increased to 84% when the reaction was
carried out at room temperature (run 3), but the reaction
rate decreased at the lower temperature (run 4). In the
(4) Total syntheses of (()-tazettine and/or (()-6a-epipretazettine:
(a) Danishefsky, S.; Morris, J .; Mullen, G.; Gammill, R. J . Am. Chem.
Soc. 1980, 102, 2838. (b) Danishefsky, S.; Morris, J .; Mullen, G.;
Gammill, R. J . Am. Chem. Soc. 1982, 104, 7591. (c) White, J . D.; Chong,
W. K. M.; Thirring, K. J . Org. Chem. 1983, 48, 2300. (d) Abelman, M.
M.; Overman, L. E.; Tran, V. D. J . Am. Chem. Soc. 1990, 112, 6959.
(e) Rigby, J . H.; Cavezza, A.; Ahmed, G. J . Am. Chem. Soc. 1996, 118,
12848. (f) Rigby, J . H.; Cavezza, A.; Heeg, M. J . J . Am. Chem. Soc
1998, 120, 3664. For the formal total synthesis of (()-6a-epipretazettine
(synthetic compound by Wildman) see: Wildman, W. C.; Bailey, D. T.
J . Am. Chem. Soc. 1969, 91, 150). (g) Overman, L. E.; Wild, H.
Tetrahedron Lett. 1989, 30, 647. Efforts for syntheses: (h) Pearson,
W. H.; Postich, M. J . J . Org. Chem. 1994, 59, 5662. (i) Baldwin, S. W.;
Aube, J .; McPhail, A. T. J . Org. Chem. 1991, 56, 6546. (j) Watson, D.
J .; Meyers, A. I. Tetrahedron Lett. 2000, 41, 1519.
(5) (a) Mori, M.; Kuroda, S.; Zhang, C.-S.; Sato, Y. J . Org. Chem.
1997, 62, 3263. (b) Preliminary report: Nishimata, T.; Mori, M. J . Org.
Chem. 1998, 63, 7586. (c) Nishimata, T.; Yamaguchi, K.; Mori, M.
Tetrahedron Lett. 1999, 40, 5713. (d) Mori, M.; Nishimata, T.; Na-
gasawa, Y.; Sato, Y. Adv. Synth. Catal. 2001, 343, 34.
(6) Asymmetric total syntheses of crinine-type alkaloids: (a) [(+)-
Maritidine] Yamada, S.-I.; Tomioka, K.; Koga, K. Tetrahedron Lett.
1976, 57; Chem. Pharm. Bull. 1977, 25, 2681. (b) [(-)-Crinine]
Overman, L. E.; Sugai, S. Helv. Chem. Acta 1985, 68, 745. (c) [(-)-
Amabiline and (-)-augustamine] Pearson, W. H.; Lovering, F. E. J .
Am. Chem. Soc. 1995, 117, 12336; J . Org. Chem. 1998, 63, 3607. (d)
Formal total synthesis of (+)-maritidine: Kita, Y.; Takeda, T.; Gyoten,
M.; Tohma, H.; Zenk, M. H.; Eichhorn, J . J . Org. Chem. 1996, 61, 5857.
(7) Recent review: Mikami, K.; Shimizu, M. Chem. Rev. 1992, 92,
1021.
(8) Marchant, J .; Oliveira-Campos, A. M. F.; Queiroz, M. J . R. P.;
Raposo, M. M.; Shannon, P. V. R. J . Chem. Soc., Perkin Trans. 1 1993,
1879.
(9) Grubbs, R. H.; DeVries, R. A. Tetrahedron Lett. 1977, 26, 1879.
(10) The ee values of 13 and 14 were determined by HPLC analysis
(DAICEL Chiralpak AD, hexane/2-propanol (9:1) and DAICEL CHIRAL-
PAK AS, hexane/2-propanol (9:1), respectively).
1838 J . Org. Chem., Vol. 69, No. 6, 2004

Allylic Substitution of 2-Arylcyclohexenols
TABLE 1. P a lla d iu m -Ca ta lyzed Asym m etr ic Am in a tion
SCHEME 4. P la n for th e Use of Vin yl Ca r bon a te
a s a Lea vin g Gr ou p
temp time yielda
solvent (°C) (h) (%)
ee
(%)
run substrate
ligand
dppb
1
2
4a
4a
THF
60
60
rt
0
0
60
rt
0
2
2
19
40
83 76
87 84
(S)-BINAPO THF
(S)-BINAPO THF
(S)-BINAPO THF
(S)-BINAPO THF
3
4a
4
4a
330 30 (31) 88
2
3
5
4b
84 88
51
68 60
6
7
8
9
10
11
12
13
11a
11a
11a
11b
11b
11b
11b
11b
dppb
THF
(S)-BINAPO THF
(S)-BINAPO THF
(S)-BINAPO THF
(S)-BINAPO THF
(S)-BINAPO DMF
(S)-BINAPO EtCN
18
complex IIIa is formed from II and Pd(0), and it is
converted into complex IIIb by abstraction of a proton
from the nucleophile (Scheme 4). Thus, the nucleophile
can react with the π-allylpalladium complex without a
base to form IV. If vinyl carbonate IIa is used for this
reaction, oxo-π-allylpalladium complex IIIc is formed.13
Since IIIc should be a stable palladium enolate complex,
it was expected that the reaction rate might be acceler-
ated.
106 31 (38) 68
0
4
48
91
73 69
80 74
65 67
-20
-20
-20
rt
310 48 (21) 68
160 43 (28) 75
(S)-BINAP
THF
a
The numbers in parentheses show the yields of recovered
starting material.
case of phosphonate 4b, the desired compound 13 with
high ee was obtained in high yield (run 5). When a
cyclohexenol derivative, 11a , having a methylenediox-
yphenyl group at the 2-position was treated in a similar
manner, both the yield and the ee were decreased (runs
7 and 8).
The reaction was carried out under various conditions.
Surprisingly, when phosphonate 11b was used instead
of methyl carbonate 11a , the reaction rate increased, and
after 4 h, the desired compound 14 was obtained in 73%
yield (run 9). Even at the lower temperature, the reaction
proceeded, and desired compound 14 with 74% ee was
obtained in 80% yield after 48 h (run 10). THF was found
to be the most suitable solvent (runs 10-12), and the use
of (S)-BINAP as a ligand gave 14 with 75% ee, but the
yield was only moderate (run 13).
Develop m en t of a Novel Lea vin g Gr ou p for P a l-
la d iu m -Ca ta lyzed Allylic Su bstitu tion . The reaction
of an allylic compound with a nucleophile using a
stoichiometric amount of palladium catalyst was devel-
oped by Tsuji.11 Subsequently, this reaction was carried
out as a catalytic reaction using alkyl allyl carbonate.12
Although phosphonate is a good leaving group in
palladium-catalyzed allylic substitution, we searched for
another effective leaving group. Among the many leaving
groups for palladium-catalyzed allylic substitution, alkyl
allyl carbonate is unique because an alkoxide anion is
generated and thus a base is not required. The reaction
of palladium-catalyzed allylic substitution proceeds via
the formation of a π-allylpalladium complex. If alkyl allyl
carbonate II is used for this reaction, π-allylpalladium
When a THF solution of allyl vinyl carbonate 4c and
tosyl amide 12 was stirred in the presence of Pd2dba3‚
CHCl3 and (S)-BINAPO at 0 °C, surprisingly, the desired
product 13 with 88% ee was obtained in 90% yield after
only 2 h, while the reaction of methyl carbonate 4a gave
13 in 30% yield after 330 h under the same reaction
conditions (Table 2, runs 1 and 2). Even at -20 °C, the
reaction proceeded and the desired product 13 with 92%
ee was obtained in 78% yield (run 3). The reactivity of
isopropenyl carbonate 4d was slightly lower than that
of vinyl carbonate 4c because of the steric hindrance
(runs 2 and 4).
Similar results were obtained when vinyl carbonate
11c was used for this reaction. The desired product 14
was obtained in 69% yield after 2.5 h (run 7), while the
reaction of methyl carbonate 11a with 12 afforded 4 in
31% yield after 106 h (run 6). The reactivity of isopro-
penyl carbonate 11d was lower than that of vinyl
carbonate 11c (runs 8 and 9). In each case, the leaving
group did not affect the enantioselectivity of 13 or 14,
but it was affected by the reaction temperature. Thus,
compound 14 was obtained in high yield by use of vinyl
carbonate compared with use of methyl carbonate, and
the reactivity would be almost the same as that of
phosphonate.
Syn th esis of Hexa h yd r oin d ole Der iva tives Usin g
Ca r bon yl-En e Rea ction . Since the desired compound
14 was obtained from 11b or 11c in an enantioriched
form, the usefulness of a carbonyl-ene reaction for
(13) A palladium-catalyzed reaction using allyl carbonate has been
reported: (a) Tsuji, J .; Shimizu, I.; Minami, I.; Ohashi, Y. Tetrahedron
Lett. 1982, 23, 4809. (b) Tsuji, J .; Shimizu, I.; Minami, I.; Ohashi, Y.;
Sugiura, T.; Takahashi, K. J . Org. Chem. 1985, 50, 1523. Palladium-
catalyzed allylation using allyl vinyl carbonate has been reported: (c)
Tsuji, J .; Minami, I.; Shimizu, I. Tetrahedron Lett. 1983, 24, 1793. (d)
Shimizu, I.; Minami, I.; Tsuji, J . Tetrahedron Lett. 1983, 24, 1797.
(11) Tsuji, J . Palladium Reagents and Catalysts; J ohn Wiley &
Sons: West Sussex, England, 1995; p 290.
(12) (a) Hata, G.; Takahashi, K.; Miyake, A. Chem. Commun. 1970,
1392. (b) Atkin, K. E. W.; Walker, E.; Manyik, R. M. Tetrahedron Lett.
1970, 3821.
J . Org. Chem, Vol. 69, No. 6, 2004 1839

Nishimata et al.
TABLE 3. Lew is-Acid -P r om oted Rea ction of 15
TABLE 2. Rea ction of Vin yl Ca r bon a te w ith 12 in th e
P r esen ce of P d (0)
yield (%)
Lewis
acid
run
equiv
16a
16b
17
1
2
3
4
AlCl3
AlCl3
SnCl4
SnCl4
1
1
1
0.1
3
67
13
84
2
5
6
5
50
a
29
a
i
The reaction mixture was treated with Pr2NEt.
SCHEME 6. P ossible Mech a n ism of
Lew is-Acid -P r om oted Rea r r a n gem en t
temp
(°C)
time
(h)
yield
(%)
ee
(%)
run
substrate
product
1
2
3
4
5
6
7
8
9
4a
4c
4c
4d
0
0
330
2
116
5
145
106
2.5
53
13
13
13
13
13
14
14
14
14
30
90
78
67
50
31
69
82
39
88
88
92
88
91
68
68
74
74
-20
0
4d
-20
0
0
-20
-20
11a
11c
11c
11d
On the other hand, when the reaction mixture was
treated with Pr2NEt, two compounds, 16a and 16b, were
245
i
obtained as an inseparable mixture (run 2). Treatment
of 15 with SnCl4 afforded the same products, but the yield
was low (run 3). However, the use of a catalytic amount
of SnCl4 (10 mol %) afforded 16 in 85% yield (run 4).
The structures of 16 and 17 were determined by NMR
spectroscopic data, including NOESY and HMBC data.
The possible reaction course for the formation of 16 and
17 is shown in Scheme 6. Coordination of a Lewis acid
to the carbonyl group of 15 gives V and V′. Carbon-
carbon bond formation between the carbonyl carbon
coordinated by the Lewis acid and the double bond occurs
to form the six-membered cation VI, and the generated
cation VI should be stabilized by an adjacent aryl group.15
Carbon-carbon bond fission of VI is accelerated by the
formation of imminium cation, and then oxygen attacks
the imminium cation of VII or VII′, resulting in the
formation of 16a or 16b. Compound 17 would be formed
by hydrolysis of VII.
SCHEME 5. Tr ea tm en t of 15 w ith Lew is Acid
Since Lewis-acid-catalyzed carbonyl-ene reactions
gave undesired rearrangement products, carbonyl-ene
reactions were carried out under thermal conditions.
When a toluene solution of 15 was heated in a sealed
tube at 180 °C for 15 h, only a small amount of desired
hexahydroindole derivative 18 was obtained (Table 4, run
1). A higher reaction temperature increased the yield of
the desired compound 18 (run 2), but reproducibility was
not achieved. Thus, molecular sieves were added to
the reaction mixture. When the solution was heated at
230 °C for 45 min, 18 was obtained in 40% yield in
the presence of 4 Å molecular sieves (run 3). Further-
more, when dried molecular sieves were used, 18 was
obtained in 60% yield and reproducibility was achieved
(run 4).
constructing a hexahydroindole skeleton was next ex-
amined. A model compound, 13, was converted into 15
in high yield by treatment with FeCl3‚SiO214 (Scheme 5).
Initially, Lewis-acid-promoted carbonyl-ene reactions of
15 were carried out. When a CH2Cl2 solution of 15 was
stirred in the presence of a Lewis acid such as TiCl4,
FeCl3, Me2AlCl, Et2AlCl, and MAD at room temperature,
the desired product 18 was not obtained and a complex
mixture was formed. However, when AlCl3 was used as
a Lewis acid, an unexpected product, 17, was obtained
after hydrolysis with aqueous NaHCO3 solution (Table
3, run 1).
Since the thermal carbonyl-ene reaction of 15 afforded
desired hexahydroindole derivative 18, an attempt was
made to convert 14 into hexahydroindole derivative 9 for
(15) A similar rearrangement has been reported. See: Csuzdi, E.;
Pallagi, I.; J erkovich, G.; Solyom, S. Synlett 1994, 429.
(14) Fadel, A.; Yefsah, R.; Salaun, J . Synthesis 1987, 37.
1840 J . Org. Chem., Vol. 69, No. 6, 2004

Allylic Substitution of 2-Arylcyclohexenols
TABLE 4. In tr a m olecu la r Ca r bon yl-En e Rea ction s
u n d er Th er m a l Con d ition s
F IGURE 1. Formation of 22b and allyl cation.
SCHEME 8. Sn Cl4-Ca ta lyzed Rea r r a n gem en t
temp
(°C)
time
(h)
SM yield
entry
additive
(%)
(%)
1
2
3
4
180
200-230
230
15
3
0.5
1.1
5
26
40
60
4 Å molecular sieves
4 Å molecular sievesa
51
230
a
4 Å molecular sieves were dried at 180 °C under reduced
pressure.
SCHEME 9. Allylic Oxid a tion of 19
SCHEME 7. Con str u ction of a
cis-3a -Ar ylocta h yd r oin d ole Moiety
and the acetoxy group of 19 is consistent with that of
the target alkaloids 1-3. In contrast to thermal condi-
tions, the SnCl4-catalyzed (20 mol %) carbonyl-ene
reaction of 10 in CH2Cl2 at 0 °C for 3 h gave a rearrange-
ment product, 20, in 86% yield (Scheme 8).
Total Syn th eses of (+)-Cr in am in e, (-)-Haem an th i-
d in e, a n d (+)-P r eta zetin e. Finally, we turned our
attention to the syntheses of crinine-type alkaloids such
as (+)-1, (-)-2, and (+)-3. Since these alkaloids have a
hydroxyl group or methoxy group at the 3-position, allylic
oxidation was carried out. Treatment of 19 with CrO3 in
the presence of 3,5-dimethylpyrazole16 gave desired
compound 21 in only 10% yield (Scheme 9).
Subsequently, 19 was treated with SeO2 in dioxane to
afford allyl alcohol 22b in high yield, but the stereochem-
istry of the hydroxyl group could not be determined
(Scheme 10). Methylation of this compound with KH
afforded an unexpected product, 23, whose acetyl group
migrates onto the 3-position of 23. Since the â-hydroxyl
group at the 3-position in the concave face is near the
â-acetoxy group at the 11-position, migration of the acetyl
group to the hydroxyl group at the 3-position occurred,
and then methylation proceeded as shown in Figure 1.
This means that the hydroxyl group at the 3-position
should be placed at the â-position. Mesylation followed
by treatment with MeOH at 0 °C afforded â-methoxylated
compound 24b (ratio of R to â 1:6) as a major product.
Presumably, allylic cation VIII is formed (Figure 1), and
then methanol attacks from the concave face to produce
â-hydroxyl compound. This is an unexpected result
the total syntheses of crinine-type alkaloids as enantio-
merically pure forms. Surprisingly, when compound 14
with 74% ee was recrystallized from MeOH, racemic 14
was obtained as a colorless crystal, and concentration of
the mother liquor gave oily (-)-14 with 99% ee in 74%
recovery.
Next, the intramolecular carbonyl-ene reaction of
enantiopure 10 was carried out. Deacetalization of 14
14
(99% ee) with FeCl3‚SiO2 gave the aldehyde 10 in high
yield (Scheme 7).
As expected, when a toluene solution of 10 was heated
at 230-240 °C in the presence of dried 4 Å molecular
sieves for several hours, the desired hexahydroindole 9
was obtained in 59% yield as an identifiable product.
Results of the NOE experiments using acetylated com-
pound 19 indicated that the ring junction of the five,six-
membered ring is cis and that an acetoxy group is trans
to an aryl group. The stereochemistry of the ring junction
(16) (a) Corey, E. J .; Fleet, W. J . Tetrahedron Lett. 1973, 4499. (b)
Salmond, W. G.; Barta, M. A.; Havens, J . L. J . Org. Chem. 1978, 21,
1357.
J . Org. Chem, Vol. 69, No. 6, 2004 1841

Nishimata et al.
SCHEME 10. Meth oxyla tion of th e C-3 P osition of
22b
SCHEME 11. Tota l Syn th eses of (+)-Cr in a m in e,
(-)-Ha em a n th id in e, a n d (+)-P r eta zetin e
and 24b was obtained as the main product (ratio of 6:1)
via SN1. However, the reaction of 22a , obtained by
Mitsunobu reaction with 22b, with HC(OMe)3 gave 24a
and 24b in a ratio of 2.2:1. Moreover, the reaction of 22b
with HC(OMe)3 gave 24a as the main product in a ratio
of 4:1.
It is difficult to explain these results. Presumably, the
reaction of 22a or 22b with HC(OMe)3 in the presence of
Montmorillonite K-10 proceeds via both SN1 and SN2.
Finally, we tried to synthesize (+)-1, whose methoxy
group is placed at the â-position. Detosylation of 24b
followed by methylenation with an Eschenmoser reagent
afforded 26 (Scheme 11). Deacetylation of 26 afforded 1,
whose [R]D value and spectral data agreed with those of
(+)-crinamine reported in the literature.2b Thus, we
succeeded in the first asymmetric total synthesis of (+)-
crinamine from cyclohexenol derivative 11c via 10 steps
in 19% overall yield.
Subsequently, we tried to synthesize crinine-type
alkaloids having R-methoxy groups at the C-3 positions.
Detosylation of 24a followed by treatment with methyl
formate19 afforded 27, which was treated with POCl3
followed by deacetylation to give alcohol 2, whose [R]D
values and spectral data agreed with those reported in
the literature.3a-c (+)-3 was synthesized from (-)-2 by
the known method.3c-f
because it was thought that methanol would attack from
the convex face to give an R-methoxylated compound.17
For the synthesis of crinine-type alkaloids, it is important
to introduce R- and â-methoxy groups at the C-3 position
in a stereoselective manner. Thus, an R-methoxylated
compound was synthesized. Mitsunobu reaction of 22b
in the presence of formic acid smoothly proceeded to give
formylated compound 25, which was treated with NH3
in MeOH to afford R-hydroxylated compound 22a in high
yield. Many attempts for methylation of the R-hydroxyl
group were then made, but all attempts were fruitless.
Fortunately, a methyl orthformate solution of 22b was
refluxed in the presence of Montmorillonite K-1018 for 1
min to give R-methoxylated compound 24a as a major
product (ratio of R to â 2.2:1) in quantitative yield. In a
similar treatment of â-hydroxylated compound 22b, 24a
was also obtained as a major product (ratio of R to â 4:1)
in 95% yield. In the reaction of a mesylated compound
of 22b with MeOH, an allylic cation, VIII, was formed,
Thus, we succeeded in short asymmetric total syn-
theses of crinine-type alkaloids (+)-1, (-)-2, and (+)-3,
using palladium-catalyzed asymmetric allylic substitu-
tion developed by us and carbonyl-ene reaction as key
steps.
(17) In previous studies,3g,h,4a-c a compound having the methyl group
on nitrogen (in our case, the tosyl group) was treated under the same
reaction conditions to give only an R-methoxylated product.
(18) Kumar, H. M. S.; Reddy, B. V. S.; Mohanty, J . S.; Yadav, J . S.
Tetrahedron Lett. 1997, 38, 3619.
(19) Schmidhammer, H.; Brossi, A. Can. J . Chem. 1982, 60, 3055.
1842 J . Org. Chem., Vol. 69, No. 6, 2004

Allylic Substitution of 2-Arylcyclohexenols
d , 9, 10, 11a -d , 13-21, 23a ,b, 24, 25a ,b, 26, and 28 and (+)-
crinamine, (-)-haemanthidine, and (+)-pretazettine. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Ack n ow led gm en t. This work was supported by a
Grant-in-Aid for Scientific Research (No. 09470478)
from the Ministry of Education, Science, Sports and
Culture, J apan, which is gratefully acknowledged.
Su p p or t in g In for m a t ion Ava ila b le: Information on
experimental procedures and spectral data for substrates 4c-
J O030309B
J . Org. Chem, Vol. 69, No. 6, 2004 1843