Design of a nonreductive method for chemoselective cleavage of hydrazines in the presence of unsaturations: Application to a stereoconvergent three-component synthesis of (-)-methyl palustramate

Article abstract of DOI:10.1021/jo048581o

A chemoselective hydrazine (N-N) cleavage methodology that preserves the integrity of alkenes was developed based on a mild acid-promoted fragmentation of tetrasubstituted 1-(trimethylsilylmethyl)-1-benzylhydrazines. This strategy was applied to a concise

Full text of DOI:10.1021/jo048581o

Design of a Nonreductive Method for Chemoselective Cleavage of
Hydrazines in the Presence of Unsaturations: Application to a
Stereoconvergent Three-Component Synthesis of (-)-Methyl
Palustramate
Barry B. Toure´ and Dennis G. Hall*
Department of Chemistry, Gunning-Lemieux Chemistry Centre, University of Alberta,
Edmonton, Alberta, T6G 2G2, Canada
dennis.hall@ualberta.ca
Received August 13, 2004
A chemoselective hydrazine (N-N) cleavage methodology that preserves the integrity of alkenes
was developed based on a mild acid-promoted fragmentation of tetrasubstituted 1-(trimethylsilyl-
methyl)-1-benzylhydrazines. This strategy was applied to a concise asymmetric synthesis of
(-)-methyl palustramate (4), which featured a convergent stereo- and regioselective sequential
three-component aza[4+2]cycloaddition/allylboration/retro-sulfinyl-ene rearrangement between
diene 1f, dienophile 2b, and propionaldehyde to afford cis-2-carboxy-6-hydroxyalkylpiperidine 25.
The acid-promoted hydrazinolysis of 25 cleanly afforded key intermediate 31, and the latter led to
target 4 in four steps after a series of functional group transformations.
Introduction
In recent years, synthetic chemists have become in-
creasingly aware of the power of multicomponent reac-
tions1 for the stereoselective construction of complex
natural products.2 By rapidly generating molecular com-
plexity from readily available starting materials, these
reactions represent attractive step-economical strate-
gies in target-oriented synthesis. Inspired by the origi-
nal carbocyclic variant of Vaultier and co-workers,3 we
have recently reported the tandem hetero[4+2]cycloaddi-
tion/allylboration three-component reaction for the prepa-
ration of R-hydroxyalkylated piperidines4 and pyrans5a
(eq 1).
Ghosez and co-workers were first to demonstrate the
use of hydrazines as electron releasing groups to reverse
the polarity of 1-azadienes and access piperidine deriva-
tives via [4+2] cycloaddition reactions.6 Our design of a
tandem aza[4+2]/allylboration strategy was based on this
seminal work, and relied on the hybrid 1-dialkyl-
amino-1-aza-4-boronobutadienes (1).
* Address correspondence to this author. Fax: (+1) 780-492-8231.
(1) For recent reviews on multicomponent reactions, see: (a) Weber,
L.; Illgen, K.; Almstetter, M. Synlett 1999, 366-374. (b) Do¨mling, A.
Comb. Chem., High Throughput Screening 1999, 1, 1. (c) Do¨mling, A.;
Ugi, I. Angew. Chem., Int. Ed. 2000, 39, 3169-3210. (d) Bienayme´,
H.; Hulme, C.; Oddon, G.; Schmitt, P. Chem. Eur. J. 2000, 6, 3321-
3329. (e) Zhu, J. P. Eur. J. Org. Chem. 2003, 1133-1144.
(2) For recent examples of applications of multicomponent reactions
in target-oriented synthesis, see: (a) Trost, B. M.; Higuchi, R. I. J.
Am. Chem. Soc. 1996, 118, 10094-10105. (b) Tietze, L. F.; Zhou, Y. F.
Angew. Chem., Int. Ed. 1999, 38, 2045-2047. (c) Dierkes, T.; Furstner,
A. Org. Lett. 2000, 2, 2463-2466. (d) Stragies, R.; Blechert, S. J. Am.
Chem. Soc. 2000, 122, 9584-9591. (e) Saito, S.; Yamazaki, S.; Yama-
moto, H. Angew. Chem., Int. Ed. 2001, 40, 3613-3617. (f) Arnold, L.
A.; Naasz, R.; Minnaard, A. J.; Feringa, B. L. J. Am. Chem. Soc. 2001,
123, 5841-5842. (g) Smith, A. B., III; Doughty, V. A.; Sfouggatakis,
C.; Bennett, C. S.; Koyanagi, J.; Takeuchi, M. Org. Lett. 2002, 4, 783-
786. (h) Powel, D. A.; Batey, R. A. Org. Lett. 2002, 4, 2913-2916. (i)
Mi, Y.; Schreiber, J. V.; Corey, E. J. J. Am. Chem. Soc. 2002, 124,
11290-11291.
Thus, the one-pot three-component process shown in
eq 1 is initiated by a hetero-Diels-Alder reaction between
azadienes 1 and electron-poor dienophiles such as ma-
leimides. The formation of the resulting cycloadduct
unmasks an allylboronate that adds in situ onto alde-
hydes to provide polysubstituted R-hydroxyalkyl piperi-
(3) (a) Vaultier, M.; Truchet, F.; Carboni, B.; Hoffmann, R. W.;
Denne, I. Tetrahedron Lett. 1987, 28, 4169. (b) Six, Y.; Lallemand, J.-
Y. Tetrahedron Lett. 1999, 40, 1295.
(5) (a) Gao, X.; Hall, D. G. J. Am. Chem. Soc. 2003, 125, 9308-
9309. (b) Deligny, M.; Carreaux, F.; Carboni, B.; Toupet, L.; Dujardin,
G. Chem. Commun. 2003, 276-277. (c) Deligny, M.; Carreaux, F.;
Toupet, L.; Carboni, B. Adv. Synth. Catal. 2004, 345, 1215-1219.
(6) Serckx-Poncin, B.; Hesbain-Frisque, A.-M.; Ghosez, L. Tetrahe-
dron Lett. 1982, 32, 3261-3264.
(4) (a) Tailor, J.; Hall, D. G. Org. Lett. 2000, 2, 3715-3718. (b) Toure´,
B. B.; Hoveyda, H. R.; Tailor, J.; Ulaczyk-Lesanko, A.; Hall, D.G. Chem.
Eur. J. 2003, 9, 466-474.
10.1021/jo048581o CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/05/2004
J. Org. Chem. 2004, 69, 8429-8436
8429

Toure´ and Hall
pot three-component tandem aza[4+2]/allylboration de-
livers this key structural unit in a highly convergent and
stereocontrolled fashion.4a,b Efforts to apply this reaction
strategy in target-oriented synthesis have recently cul-
minated in a concise asymmetric synthesis of methyl
dihydropalustramate (5).11 Waldner’s sulfinimide dieno-
phile,12 a chiral derivative of 2, played a key role in our
successful synthesis of 5. Unfortunately, a synthesis of
4 could not be envisaged at this time due to the lack of
reliable methodology for the chemoselective cleavage of
the tetraalkyl hydrazine remnant of diene 1. Such a
transformation is required to liberate the free piperidine
while preserving the double bond, the only feature
distinguishing 4 from 5.
Current methodologies that preserve the integrity of
the alkene functionality during the cleavage of an N-N
bond first consist of activating the hydrazine in a regio-
selective manner via acylation, followed by reductive
cleavage of the resulting hydrazide with sodium or
lithium13a-d in liquid ammonia, samarium iodide,14 or
through an oxidative cleavage with magnesium mono-
peroxyphthalate (MMPP).15 It should be noted that most
of these preactivation strategies are not applicable to
unsymmetrical alkylhydrazines such as those employed
in our studies. Herein, we report the synthesis of a new
class of hydrazines, namely the 1-(trimethylsilylmethyl)-
1-benzylhydrazines, which can be cleaved nonreductively
under mild acidic conditions in the presence of unsat-
urations. This unique methodology was successfully
applied to a concise total synthesis of (-)-methyl palus-
tramate (4). A mechanistic insight pertaining to this new
N-N cleavage strategy is also discussed.
FIGURE 1. Selected R-hydroxyalkyl piperidine containing
natural products.
dine products in a highly stereoselective fashion. This
motif is encountered in the structure of a number of
natural products such as the palustrine family exempli-
fied by palustrine itself (3),7 methyl palustramate (4) and
its saturated degradation product methyl dihydropalus-
tramate (5), and other alkaloids such as (+)-can-
nabisativine (6)8 and quinine (7)9 (Figure 1). Although a
number of total syntheses of these natural products have
appeared in the literature,10 the stereocontrolled intro-
duction of the 6-hydroxyalkyl side chain has often neces-
sitated the recourse to a linear approach. Our one-
Results and Discussion
(7) For early isolation and structural elucidation through synthetic
and degradation studies (note that the originally postulated C4-C5
dehydro structure of palustrine was wrong, and was later corrected to
C3-C4 dehydro on the basis of refs 10a-c), see: (a) Karrer, P.; Eugster,
C. H. Helv. Chim. Acta 1948, 31, 1062-1066. (b) Mayer, C.; Trueb, J.;
Wilson, J.; Eugster, C. H. Helv. Chim. Acta 1968, 51, 661. (c) Eugster,
C. H. Heterocycles 1976, 4, 51-105. (d) Ru¨edi, P.; Eugster, C. H. Helv.
Chim. Acta 1978, 61, 899-904. (e) Mayer, C.; Green, C. L.; Trueb, W.;
Wa¨lchli, P. C.; Eugster, C.H. Helv. Chim. Acta 1978, 61, 905-921. (f)
Wa¨lchli, P. C.; Mukherjee-Mu¨ller, G.; Eugster, C. H. Helv. Chim. Acta
1978, 61, 921-928.
(8) For isolation of (+)-cannabisativine: (a) Lotter, H. L.; Abraham,
D. J.; Turner, C. E.; Knapp, J. E.; Schiff, P. L.; Slatkin, D. J.
Tetrahedron Lett. 1975, 7, 2815-2818. (b) Turner, C. E.; Hsu, M.-F.;
Knapp, J. E.; Schiff, P. L.; Slatkin, D. J. J. Pharm. Sci. 1976, 65, 1084-
1085.
Scope of the Aza-Diels-Alder Reaction. Our three-
component aza[4+2]cycloaddition/allylboration has shown
a very broad substrate scope in terms of hydrazine and
aldehyde components, which makes it particularly suited
for applications in diversity-oriented synthesis (DOS).4b
Unfortunately, in the normal electron-demand [4+2]
cycloaddition manifold, the bulky electron-withdrawing
pinacol boronate substituent exerts a strong deactivating
effect on the diene. Thus, the thermal cycloaddition works
well only with very electron-poor diactivated dienophiles
such as N-substituted maleimides; monoactivated dieno-
philes such as acrylates and vinyl sulfones are unreac-
tive.4b However, as targets 3-7 do not bear any substitu-
ent at the 3-position (Figure 1), their syntheses would
require further fine-tuning of the dienes’ reactivity or the
(9) Turner, R. B.; Woodward, R. B. The Chemistry of the Cinchona
Alkaloids. In The Alkaloids; Manske, R. H. F., Ed.; Academic Press:
New York, 1953; Vol. 3, Chapter 16.
(10) Syntheses of members of the palustrine family: Racemic
syntheses of the wrong structure of palustrine: (a) Natsume, M.;
Ogawa, M.; Yoda, I.; Shiro, M. Chem. Pharm. Bull. 1984, 32, 812-
814. (b) Wasserman, H. H.; Leadbetter, M. R.; Kopka, I. E. Tetrahe-
dron Lett. 1984, 25, 2391-2394. Synthesis of racemic palustrine
and structure revision: (c) Natsume, M.; Ogawa, M. Chem. Pharm.
Bull. 1984, 32, 3789-3791. Total synthesis of (-)-dihydropalustra-
mic acid: (d) Muraoka, O.; Zheng, B.-Z.; Okumura, K.; Tanabe, G.;
Momose, T.; Eugster, C. H. J. Chem. Soc., Perkin Trans. 1 1996,
1567-1575. A prospective intermediate for the synthesis of (+)-
palustrine: (e) Hirai, Y.; Watanabe, J.; Nozaki, T.; Yokoyama, H.;
Yamaguchi, S. J. Org. Chem. 1997, 63, 7776-777. Total synthesis of
(-)-methyl palustramate: (f) Angle, S. R.; Henry, R. M. J. Org. Chem.
1998, 63, 7490-7497. For a recent asymmetric synthesis of cannabi-
sativine see: (g) Kuethe, J. T.; Comins, D. L. Org. Lett. 2000, 2, 855-
857. For recent asymmetric syntheses of quinine see: (h) Stork, G.;
Niu, D.; Fujimoto, A.; Koft, E. R.; Balkovec, J. M.; Tata, J. R.; Dake,
G. J. Am. Chem. Soc. 2001, 123, 3239-3242. (i) Raheem, I. T.;
Goodman, S. N.; Jacobsen, E. N. J. Am. Chem. Soc. 2004, 126, 706-
707.
(11) Toure´, B. B.; Hall, D. G. Angew. Chem., Int. Ed. 2004, 43, 2001-
2004.
(12) Waldner, A. Tetrahedron Lett. 1989, 30, 3061-3064.
(13) For use of Li/NH3, see: (a) Mellors, J. M.; Smith, N. M. J. Chem.
Soc., Perkin Trans. 1 1984, 96, 2927. (b) Denmark, S. E.; Nicaise, O.;
Edwards, J. P. J. Org. Chem. 1990, 55, 6219. (c) Enders, D.; Tiebes,
J.; De Kimpe, N.; Keppens, M.; Stevens, C.; Smagghe, G.; Bertz, O. J.
Org. Chem. 1993, 58, 4881-4184. (d) Job, A.; Janeck, C. F.; Bettray,
W.; Peters, R.; Enders, D. Tetrahedron 2002, 58, 2253.
(14) (a) Burk, M. J.; Feaster, J. E. J. Am. Chem. Soc. 1992, 114,
6266-6267. (b) Sturino, C. F.; Fallis, A. G. J. Am. Chem. Soc. 1994,
116, 7447-7448. (c) Ding, H.; Friestad, G. K. Org. Lett. 2004, 6, 637-
640.
(15) (a) Ferna´ndez, R.; Ferrete, A.; Lassaletta, J. M.; Llera, J. M.;
Monge, A. Angew. Chem., Int. Ed. 2000, 39, 2893-2897. (b) Ferna´ndez,
R.; Ferrete, A.; Lassaletta, J. M.; Llera, J. M.; Martin-Zamora, E.
Angew. Chem., Int. Ed. 2002, 41, 831-833.
8430 J. Org. Chem., Vol. 69, No. 24, 2004

Synthesis of (-)-Methyl Palustramate
SCHEME 1. Retrosynthetic Analysis for
(-)-Methyl Palustramate (4)
the sulfoxide group. Moreover, the use of Waldner’s chiral
dienophile 2b, obtained from the inexpensive and com-
mercially available optically active (R)-(+)-R-methylben-
zylamine,12 afforded products that were virtually optically
pure. The stereochemical outcome is explained by the
endo transition structure whereby the diene approaches
the dienophile from the face opposite to the S-O bond.
Optimization of the Retro-Sulfinyl-Ene Fragmen-
tation. To the best of our knowledge, the retro-sulfinyl-
ene rearrangement process (eq 2) has never been em-
ployed in target-oriented synthesis, and only one study
examined cyclic substrates.17c Thus, our design strategy
to 2,6-disubstituted piperidines and the palustrine alka-
loids relied on the successful implementation of such a
retro-sulfinyl-ene fragmentation17 involving the cyclic
adducts 10 from the aza[4+2]/allylboration reaction
between dienes 1, dienophiles 2, and aldehydes (Table
1). In this scenario, SO2 extrusion from 10 would be
concomitant with a migration of the C4-C5 unsaturation
to the C3-C4 position, which is necessary for accessing
methyl palustramate (4). Toward this end, 10e (entry 5,
Table 1) was arbitrarily selected for early exploratory
work. The latter was submitted to standard retro-ene
reaction conditions consisting of a treatment with 5% aq
NaOH, followed by acidification to generate in situ the
sulfinic acid intermediate 11. The desired retro-ene
product 12e was obtained, albeit in low yield. To our
surprise, when the tert-butyl group of the dienophile was
replaced with the n-propyl-substituted analogue 10c
(entry 3, Table 1), no retro-ene product was observed.
Likewise, when the aldehyde side chain was replaced
with a less sterically demanding group (e.g 10f, entry 6,
Table 1), no product formation was observed either.
Longer stirring time in the acidic media or attempt to
warm the solution to room temperature mainly resulted
in decomposition. Clearly, an improved retro-ene proce-
dure was required. This was achieved by carefully
quenching the acidic solution with a weak base (aqueous
NaHCO3), followed by removal of the solvent. The sulfinic
acid intermediate 11 was then dissolved in chloroform
and heated. Only then was the desired product isolated
in all cases, although the reaction time and tempera-
ture were still a function of the steric bulk of R3 and R4.
In general, as the size of these groups increases, the
reaction time and temperature can be decreased signifi-
cantly.
use of dienophiles bearing easily removable activating
groups. Unfortunately, all our attempts to increase the
reactivity of dienes 1 only met with failure. Accordingly,
we next considered the use of suitable diactivated dieno-
philes to overcome the limitations of our methodology
with respect to the mitigated reactivity of dienes 1.
However, any such dienophiles would have to meet the
following requirements: (1) possess the requisite elec-
tronic characteristics to react with heterodienes 1, (2)
provide high enantiofacial selectivity, and (3) provide
high regioselectivity and lead to a cycloadduct convertible
to the C3-C4 dehydro unit of 4 via an allylic transposi-
tion (i.e., 8 to 9) as depicted in the retrosynthetic analysis
shown in Scheme 1.
With these criteria in mind, we turned our attention
to the 4-isothiazolin-3-one 1-oxide dienophiles (2) first
reported by Weiler and Brennan.16 We anticipated that
the hydrolysis of the MCR adducts of these compounds
would provide allylic sulfinic acids capable of undergoing
a retro-ene fragmentation17 concomitant with the re-
quired alkene migration (eq 2).
Thus, sulfinimide dienophiles 2a-d bearing a variety
of substituents at the nitrogen were obtained in excellent
yield, and in only three steps following some small
modifications to the original procedure.16 Gratifyingly,
these dienophiles reacted smoothly with heterodienes 1
in the presence of aldehydes under our previously opti-
mized reaction conditions (Table 1). This process afforded
the three-component adducts 10 in moderate to good yield
following aqueous sodium bicarbonate workup and flash
chromatography purification. It is remarkable that only
a single regioisomer and stereoisomer of adducts 10 was
observed in the crude reaction mixture. Obviously, the
endo pathway operates in the aza-Diels-Alder reaction,
and the regiochemistry is controlled by the electron-
donating 1-amino substituent of the diene along with the
carbonyl of the dienophile, which overrides the effect of
Although the reasons for this reactivity trend remain
speculative, conformational effects may be at play to
explain the different behavior of 10a-10i. To reach the
six-membered transition state for a concerted retro-
sulfinyl-ene fragmentation,17d the sulfinic acid substitu-
ent must occupy a pseudoaxial orientation (conformer B)
(Figure 2). This reactive conformer also features two
disfavored gauche interactions between the bulky hy-
drazine substituents (R1 and R2) and both the R-hydroxy-
alkyl chain and the carboxamide. To minimize this type
(16) Weiler, E. D.; Brennan, J. J. J. Heterocycl. Chem. 1978, 15,
1299-1302.
(17) Selected references: (a) Wucherpfennig, W. Tetrahedron Lett.
1967, 3235. Wucherpfennig, W. Ann. Chem. 1971, 761, 16-27. (b)
Mock, W. L.; Nugent, R. M. J. Org. Chem. 1978, 43, 3433-3434. (c)
Rogic, M. M.; Masilamani, D. J. Am. Chem. Soc. 1977, 99, 5219-5220.
(d) Weinreb, S. M.; Staib, R. R. Tetrahedron 1982, 38, 3087-3128. (e)
Garigipati, R. S.; Morton, J.A.; Weinreb, S. M. Tetrahedron Lett. 1983,
24, 987-990.
J. Org. Chem, Vol. 69, No. 24, 2004 8431

Toure´ and Hall
TABLE 1. Scope of the Three-component Aza[4+2]/Allyboration and the Retro-Sulfinyl-Ene Fragmentation Reaction
Sequencea
diene
R1
dienophile
R3
aldehyde
R4
entry
no.
R2
no.
product
yieldb (%)
product
yieldc (%)
1
2
1a
1a
1a
1a
1a
1a
1a
1b
1c
1d
1e
1g
Me
Me
Me
Me
Me
Me
Me
Ph
Bn
H
Me
Me
Me
Me
Me
Me
Me
Me
Bn
Boc
Ac
2a
2b
2c
2d
2a
2a
2b
2b
2b
2b
2b
2b
t-Bu
Cy
Ph
Ph
n-Bu
Ph
Et
10a
10b
10c
10d
10e
10f
10g
10h
10i
65
50
51
52
49
41
44
41
62
0
12a
12b
12c
12d
12e
12fd
12g
12h
12i
55
65
52
65
69
PhCH(Me)
3
n-Pr
4
PhCH(Me)
t-Bu
5
6
t-Bu
7
PhCH(Me)
PhCH(Me)
PhCH(Me)
PhCH(Me)
PhCH(Me)
PhCH(Me)
Et
80
80
78
8
Et
9
Et
10
11
12
Et
10j
12j
H
Ph
Ph
10k
10l
0
12k
12l
e
0
a All aza[4+2]/allylboration reactions were carried out by heating a 1:1:2 mixture of diene/dienophile/aldehyde, except for propionaldehyde
where 5 equiv were used, in anhydrous toluene [0.2-0.3 M] at 80 °C for 72 h. Racemic 2b was used throughout except for making 10b,
10g, 12g, and 12i, which were accessed from optically pure 2b (â-sulfoxide diastereomer) derived from (R)-(+)-R-methylbenzylamine (see
experimental details in the Supporting Information). b Yields of isolated products after flash chromatography purification. c The products
were obtained after reflux for 12-48 h in CHCl3 and flash chromatography purification (12e was simply stirred at room temperature).
d Reaction not attempted. e Diene made from O-methylhydroxylamine, R1NR2 ) OMe. For 1f, see Scheme 3.
all the requisite stereocenters of the palustrine natural
products (3-5) in only two synthetic operations (com-
pounds 12, Table 1). From there, the completion of the
synthesis of 3 and 4 requires the chemoselective cleavage
of the N-N bond followed by a one-carbon homologation
of the amide side chain (Scheme 1). Strategically, we first
decided to tackle what we perceived as the major chal-
lenge: the chemoselective cleavage of the N-N bond in
the presence of the C3-C4 olefin.
FIGURE 2. Suggested conformational equilibrium to explain
the influence of the amide substituent (R3) of intermediates
11 in the retro-sulfinyl-ene rearrangement.
Model Studies toward Methyl Palustramate (4):
Examination of Traditional Methods for Hydrazine
Cleavage. Our synthetic plan, from the outset, first
stressed the need to cleave the N-N bond in a chemose-
lective manner that would preserve the C3-C4 double
bond. We envisioned three main strategies to accomplish
this task: (1) the use of hydrazides or N-alkoxyamines,
which usually can be cleaved under mild reaction condi-
tions to afford the desired amines, (2) activation of
trialkylhydrazines after the retro-ene reaction step, or
(3) exploring the feasibility of a direct cleavage of
tetraalkylhydrazines by using either the oxidative or
reduction cleavage methods mentioned above. To test the
first hypothesis, we prepared the known dienes 1d-e
(entries 11 and 12, Table 1)4b and 1g, and tested their
ability to undergo the aza-Diels-Alder reaction with
dienophile 2. Despite significant efforts, no desired cy-
cloaddition product was obtained probably due to the
lower reactivity of these dienes, which are less electron
of strain, closely related cis-2,6-disubstituted piperidines
have been shown to adopt a “diaxial” conformation of type
A.18 In this nonreactive conformer A, the carboxamide
group occupies a pseudoaxial orientation. Thus, we
hypothesize that bulkier N-alkyl substituents on the
amide may affect the conformational equilibrium and
facilitate the retro-ene fragmentation by destabilizing
conformer A to the benefit of reactive conformer B.
Another possible explanation is epimerization of the
sulfinic acid substituent in 11, forcing a highly unfavor-
able reactive conformer with all-axial substituents.
Overall, this tandem aza[4+2]/allylboration/retro-ene
sequence allowed the construction of the entire core with
(18) For example, N-acylated cis-2,6-disubstituted piperidines exist
in the “diaxial” conformation to escape A1,3 strain between the planar
exocyclic amide group and the neighboring substituents in the “diequa-
torial” conformation: Natsume, M.; Ogawa, M. Chem. Pharm. Bull.
1982, 30, 3442-3445 and references therein. In the case of compounds
11, the planar hydrazine can be considered isosteric to an acyl group.
8432 J. Org. Chem., Vol. 69, No. 24, 2004

Synthesis of (-)-Methyl Palustramate
rich than 1a-c. We next examined the activation of the
hydrazine moiety following the MCR/retro-ene sequence.
Toward this goal, we hoped to take advantage of the well-
established N-debenzylation of tertiary amines using
chloroformates, although this strategy had never been
applied to hydrazines before.19a,b Accordingly, 1-diben-
zylhydrazine was synthesized in two steps then con-
densed with 3-boronoacrolein,4b and the resulting diene
1c was carried through our sequential tandem [4+2]/
allylboration/retro-sulfinyl-ene reaction to afford 12i
(entry 9, Table 1). However, when the latter was treated
with ethylchloroformate or R-chloro ethylchloroformate
under known reaction conditions,19c none of the expected
debenzylated product 14 was obtained (eq 3).
FIGURE 3. Design of a nonreductive hydrazine cleavage
strategy.
Design of a Nonreductive Hydrazine Cleavage
Strategy. The failures of traditional N-N cleavage
methods with tetraalkylhydrazines 12 exposed an im-
portant methodological gap in organic synthesis. A po-
tential solution to this problem arose from an under-
standing of the reactivity of compounds 12. In the course
of our studies, we had realized that these compounds are
unstable to basic conditions, giving pyridines as products.
Indeed, treatment of 12 with bases such as Et3N or DBU
results in the effective cleavage of the N-N bond to give
17, presumably via deprotonation of the acidic C2 proton,
followed by air oxidation to afford 18 (eq 6, Figure 3).
Along this line, we reasoned that compound 20 could
be obtained if a suitable anion precursor could be
introduced next to the exocylic nitrogen. After careful
consideration, we chose to include a Me3SiCH2- sub-
stituent to the hydrazine 19, and test the feasibility of a
fragmentation process reminiscent of the removal of a
[2-(trimethylsilyl)ethoxy]methyl (SEM) protective group21a,b
(eq 7, Figure 3). This type of silylated hydrazine, however,
was apparently unknown. The discovery of a convenient
synthetic route to this class of reagent became the next
focus of our efforts.
A New Acid-Cleavable Hydrazine. The commer-
cially available ((trimethylsilyl)methyl)benzylamine was
first nitrosated under standard reaction conditions in an
aqueous acetic acid solution (Scheme 2). The unstable
nitroso intermediate 21 undergoes slow decomposition at
room temperature, which can be accelerated by heating.
Consequently, it was rapidly reduced to the desired
hydrazine 22 at low temperature. The latter was obtained
in sufficiently high purity that it could be carried through
the next step without the need for any purification.
Alternatively, the hydrazine could be purified by simple
acid-base extraction or bulb-to-bulb distillation.
Instead, the intramolecular lactonization product 13 was
observed. The latter most likely results from indirect
amide activation through intramolecular acyl transfer
between the first formed, quaternized acylated hydrazine
and the secondary amide.
These failures prompted us to explore the use of more
direct strategies. Our first attempt relied on the oxidative
cleavage methodology.15 On the basis of the proposed
mechanism for this reaction, we reasoned that the
oxidant would preferentially react with the least hindered
exocyclic nitrogen, hence leading to a selective N-N
fragmentation. In this approach, the high reactivity of
the nitrogen toward electrophilic peroxide reagents should
also preclude any competitive epoxidation of the C3-C4
olefin. However, when we put this hypothesis to the test
with substrate 12g under various reaction conditions (e.g.
MMPP, mCPBA), only decomposition product was ob-
served.
Likewise, treatment of 12g with zinc, a reductive N-N
cleavage method,6 did not yield any of the desired N-N
cleaved product. However, powerful reducing agents,
typified by Red-Al, did yield the desired N-N cleavage
product 15 albeit the carbonyl functionality could not be
preserved during this operation, thereby rendering the
one carbon homologation difficult (eq 4). Finally, an
attempt to achieve the one-pot amide debenzylation/N-N
cleavage under Bouveault-Blanc-type conditions, and
variants thereof,20 only resulted in the reduction of the
aromatic moiety to afford 16. The latter result is par-
ticularly telling of the difficult challenge that represents
the cleavage of tetraalkylhydrazines (eq 5).
(19) (a) Olofson, R. A.; Schnur, R. C.; Bunes, L.; Pepe, J. P.
Tetrahedron Lett. 1977, 1567-1570. (b) Olofson, R. A.; Martz, J. T.;
Senet, J. P.; Piteau, M.; Malfoot, T. J. Org. Chem. 1984, 49, 2081-
2082. (c) Yang, B. V.; O’Rourke, D.; Li, J. Synlett 1993, 195-196.
(20) See the Supporting Information for the preparation of compound
1 in: Buckmelter, A. J.; Kim, A. I.; Rychnovsky, S. D. J. Am. Chem.
Soc. 2000, 122, 9386-9390.
(21) (a) Lipshutz, B. H.; Pegram, J. J. Tetrahedron Lett. 1980, 21,
3343-3346. (b) Whitten, J. P.; Matthews, D. P.; McCarthy J. R. J. Org.
Chem. 1986, 51, 1891-1894.
J. Org. Chem, Vol. 69, No. 24, 2004 8433

Toure´ and Hall
SCHEME 2. Synthesis of
1-(Trimethylsilylmethyl)-1-benzylhydrazine
SCHEME 4. Undesired TBAF-Promoted C-C
Cleavage
SCHEME 5. Attempted TBAF-Promoted N-N
Cleavage with MOM-Protected Substrate 29
SCHEME 3. Sequential Aza[4+2]/Allylboration
Route to the Key Hydrazine Cleavage Precursor
25
tion via the proposed mechanism depicted in Scheme 4.
To prevent the reaction from proceeding via this undes-
ired pathway, the alcohol was protected as its meth-
oxymethyl ether 29. When the latter product was sub-
mitted to the above reaction conditions, however, none
of the desired product 30 was obtained (Scheme 5).
We then turned our attention to the acid-promoted
desilylation conditions.21b Gratifyingly, when key inter-
mediate 25 was stirred in ethanolic aqueous HCl solution,
a very clean conversion of the starting hydrazine to the
desired product 31 along with benzylamine coproduct
were observed in the 1H NMR of the crude reaction
mixture (eq 8).
Synthesis of (-)-Methyl Palustramate (4): Con-
struction of the [4+2]/Allylboration/Retro-Sulfinyl-
Ene Adduct 25. Hydrazine 22 was first condensed with
boronoacrolein 23 in refluxing dichloromethane under
dehydrating reaction conditions to afford diene 1f as a
mixture of different geometrical isomers. The latter was
carried through the three-component [4+2]/allylboration
reaction involving dienophile 2b and freshly distilled
propionaldehyde (Scheme 3). The MCR-adduct 24 was
obtained as a single stereoisomer, albeit, in moderate
yield due to partial decomposition of the product under
the conditions involving long reaction times. Compound
24 was then submitted to our optimal conditions for the
retro-sulfinyl-ene reaction to afford key intermediate 25.
The latter was found to be slightly unstable on silica gel,
and thus was best carried through the next step without
any purification from 24. The stage was now set for
testing the hydrazine cleavage strategy.
Chemoselective N-N Fragmentation on Inter-
mediate 25. We first explored the tetrabutylammonium
fluoride (TBAF)-mediated desilylation reaction condi-
tions.21a,b Compound 25 was heated in THF in the
presence of 3 equiv of TBAF overnight. To our dismay,
after workup, the 1H NMR spectrum of the crude product
revealed a clean conversion of the starting material to
pyridine 28, which is consistent with a C-C fragmenta-
Completion of the Synthesis of Methyl Palustra-
mate (4). With this key N-N cleaved product 31 in hand,
we next addressed the one-carbon homologation issue
that we hoped to accomplish via the previously described
Arndt-Eistert reaction sequence.11 For this purpose, the
amino alcohol 31 was first converted in excellent yield
to its carbamate derivative with use of carbonyl diimi-
dazole (Table 2). Selective hydrolysis of the amide group
could only be performed through formation of the N-
nitroso derivative 32.22a Unfortunately, epimerization
occurred at C2 in this operation, and the desired 2,6-cis-
configured acid product was only obtained as the minor
(22) (a) White, E. M. J. Am. Chem. Soc. 1955, 77, 6011-6014. (b)
Evans, D. A.; Carter, P. H.; Dinsmore, C. J.; Barrow, J. C.; Katz, J. L.;
Kung, D. W. Tetrahedron Lett. 1997, 38, 4535-4538.
8434 J. Org. Chem., Vol. 69, No. 24, 2004

Synthesis of (-)-Methyl Palustramate
TABLE 2. Optimization of the Hydrolysis of Amide 31
SCHEME 6. Synthesis of Methyl Palustramate (4)
ratiob
33:epi-33
entry
base
temp (°C)
time (h)
1
2
3
4
5
6
7
8
NaOH
23
23
16
16
16
16
8
16
8
15
1:7
1:4
KOH
LiOH
23
1:5
Ba(OH)2
Ba(OH)2
Ba(OH)2
LiOOH
LiOOH
23
0
-10
0
-10
1:2.3
1:1.6
1:1
1:1
3.3:1
SCHEME 7. Construction of Model Compound 38
for Investigation of the Scope and Mechanism of
the Hydrazine Cleavage Methodology
a The reaction was performed on a 10-15 mg sample of 32.
b The ratio of 33:epi-33 was determined by proton NMR of the
crude reaction mixture after workup.
isomer in a 2:1 ratio. This result contrasted with our
earlier observations with the saturated analogue, whereby
the 2,6-cis-piperidine was always observed as the major
product for a nitrosation time of 3 h or less.11 It is
interesting to note that the nitroso compound 32 is prone
to epimerization even in the absence of any external base.
Indeed, complete inversion of the C2 stereochemistry
could be observed by simply maintaining a chloroform
solution of 32 at room temperature overnight. Moreover,
when the N-nitrosation was allowed to proceed at room
temperature for longer reaction times, the undesired 2,6-
trans-isomer epi-33 was obtained almost exclusively.
To reverse this trend, we thus set out to investigate
the effect of base and temperature on the stereochemical
outcome of this hydrolysis reaction. The results are
summarized in Table 2.
The best result was obtained at low temperature, using
the highly nucleophilic lithium hydrogen peroxide previ-
ously reported by Evans and co-workers (entry 8, Table
2).22b However, in this case, the product was always
contaminated by an inseparable pyridine byproduct.
Thus, for the completion of the synthesis, we opted for
the barium hydroxide conditions, which provided a clean
reaction with a 1:1 ratio of C2 epimers (entry 6, Table
2). The required homologation was performed on the
1:1 epimeric mixture of carboxylic acids 33, using an
Arndt-Eistert sequence (Scheme 6). The two epimers
34 were directly subjected to the final step of amino
alcohol deprotection with the method of Weinreb and
co-workers, using barium hydroxide.23 Re-esterifica-
tion of the resulting amino acid and effective chromato-
graphic separation afforded (-)-methyl palustramate (4),
which possessed spectral characteristics and an optical
rotation value in agreement with reported literature
data.10f
mentation during the TBAF-promoted N-N cleavage (i.e.
25 to 28) prompted us to investigate any possible
implications of the proximal side chain alcohol under the
hydrazinolysis conditions. Toward this end, bicyclic
compound 37 was prepared via an aza-Diels-Ader reac-
tion between diene 36 and N-methyl maleimide (Scheme
7). The sodium cyanoborohydride reduction of the enam-
ine 37 resulted in a 2:1 diasteromeric mixture at the
methyl stereocenter, which was inconsequential for our
study. When the latter compound (38) was subjected to
our N-N cleavage conditions, the desired reaction prod-
uct 39 was obtained in a nonoptimized yield of 63%. The
reaction time, however, was significantly longer. In light
of this result, it seems plausible that the alcohol func-
tionality in the R-hydroxyalkyl derivative 25 participates
and accelerates the N-N cleavage via nucleophilic attack
of the silicon atom. Considering the failure of the TBAF-
promoted method, protonation of the endocyclic nitrogen
seems essential for a successful N-N cleavage in this
context. Despite the longer reaction time, the successful
N-N cleavage of compound 38 suggests a wide scope of
application for this novel class of acid-cleavable hydra-
zine.
Preliminary Study of the Scope and Mechanism
of the Hydrazine Cleavage Method. The C-C frag-
(23) Bailey, T. R.; Garigipati, R. S.; Morton, J. A.; Weinreb, S. M. J.
Am. Chem. Soc. 1984, 106, 3240-3245.
J. Org. Chem, Vol. 69, No. 24, 2004 8435

Toure´ and Hall
Conclusion
hydrazines including the development of a chiral variant
are currently under investigation and will be reported
in due course.
In summary, we have described a three-component
aza[4+2] cycloaddition/allylboration/retro-sulfinyl-ene se-
quential reaction approach to access cis-2,6-disubstituted
piperidines in a highly regio-, diastereo-, and enantiose-
lective fashion. Few multicomponent reaction strategies
demonstrate such a high level of stereocontrol in the
formation of complex, functionalized compounds. The
utility of this powerful step-economical process was
successfully demonstrated with a concise enantioselective
synthesis of the palustrine relative (-)-methyl palustra-
mate (4). The entire sequence to reach target 4 was
accomplished with only two purification steps, and in only
10 linear synthetic operations from commercial 3,3′-
diethoxypropyne.4b Along the way, we have developed the
first class of hydrazines that can be cleaved nonreduc-
tively under mild acidic conditions in the presence of
unsaturations. This new trimethylsilylmethyl-substituted
hydrazine played a pivotal role in the completion of the
synthesis of 4. Further applications of this class of
Acknowledgment. Financial support for this re-
search by the Natural Sciences and Engineering Re-
search Council (NSERC) of Canada (Discovery Grant
to D.G.H.), AstraZeneca (Chemistry Award to D.G.H.),
and the University of Alberta is gratefully acknowl-
edged. B.B.T. thanks the Canadian Institutes for Health
Research (CIHR) for a graduate scholarship. We are
grateful to Prof. Steven R. Angle (University of Cali-
fornia-Riverside) for a copy of the 13C NMR spectrum
of 4. We thank Glen Bigam and Lai Kong for NMR
analyses.
Supporting Information Available: Full experimental
details with synthetic procedures and analytical data for all
new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
JO048581O
8436 J. Org. Chem., Vol. 69, No. 24, 2004