Stereoselective synthesis of polyhydroxylated quinolizidines from C-glycosides by one-pot double-conjugate addition

Article abstract of DOI:10.1021/jo062057v

(Chemical Equation Presented) An effective one-pot synthesis of polyhydroxylated quinolizidines from 1-C-(2′-oxo-4′-pentenyl)-5- azido-C-glycofuranosides was developed. Reduction of the 5-azido group using triphenylphosphine followed by base treatment pro

Full text of DOI:10.1021/jo062057v

Stereoselective Synthesis of Polyhydroxylated Quinolizidines from
C-Glycosides by One-Pot Double-Conjugate Addition
Wei Zou,* Mahendra Sandbhor, and Milan Bhasin
Institute for Biological Sciences, National Research Council of Canada, Ottawa,
Ontario, Canada K1A 0R6
wei.zou@nrc-cnrc.gc.ca
ReceiVed October 4, 2006
An effective one-pot synthesis of polyhydroxylated quinolizidines from 1-C-(2′-oxo-4′-pentenyl)-5-azido-
C-glycofuranosides was developed. Reduction of the 5-azido group using triphenylphosphine followed
by base treatment produced quinolizidines in good yield. The base-mediated ring-opening â-elimination
produced an acyclic R,â-conjugated ketone as a Michael acceptor, which was followed by an intramolecular
nitrogen conjugate addition to form an aza-C-glycopyranoside intermediate. Meanwhile, the â,γ-double
bond of the aglycon migrated under the basic conditions to form another R,â-conjugated ketone. The
subsequent intramolecular conjugate addition by the azasugar nitrogen led to the formation of the
quinolizidines in a highly stereoselective manner. The stereoselectivity of the first conjugate addition
giving azasugar is affected by the stereochemistry of the monosaccharide substrate, whereas the
stereoselectivity in the second conjugate addition was likely directed entirely by steric repulsion from
the azasugar.
Introduction
synthesis of polyhydroxylated indolizidines has been often
accomplished in a stepwise approach via pyrrolidine and
piperidine intermediates.9 Few examples exist where the nitrogen
bicyclic ring system was constructed by a one-pot reaction. They
include an intramolecular 1,3-dipolar addition of azide to an
Quinozilidine and indolizidine alkaloids have provided a rich
source of synthetic targets because of their potential pharma-
ceutical applications.1 The construction of nitrogen bicycles has
been achieved by ring-closing metathesis,2 Diels-Alder reac-
tions,3 other inter- and intramolecular cycloadditions,4 and
reactions based on N-nitrones5 and N-sulfinylimines.6 Polyhy-
droxylated indolizidines have also been desirable synthetic
targets because they are able to selectively inhibit glyco-
processing enzymes7 and thus may be used as therapeutic agents
for the treatment of diabetes, viral infections, and cancers.8 The
(2) For examples, see: (a) Katoh, M.; Mizutani, H.; Honda, T.
Tetrahedron Lett. 2005, 46, 5161-5163. (b) Ma, S. M.; Ni, B. K.; Liang,
Z. Q. J. Org. Chem. 2004, 69, 6305-6309. (c) Skaanderup, P. R.; Madsen,
R. J. Org. Chem. 2003, 68, 2115-2122. (d) Lesma, G.; Crippa, S.; Danieli,
B.; Passarella, D.; Sacchetti, A.; Silvani, A.; Virdis, A. Tetrahedron 2004,
60, 6437-6442. (e) Randl, S.; Blechert, S. J. Org. Chem. 2003, 68, 8879-
8882. (f) Wijdeven, M. A.; Botman, P. N. M.; Wijtmans, R.; Schoemaker,
H. E.; Rutjes, F. P. J. T.; Blaauw, R. H. Org. Lett. 2005, 7, 4005-4007.
(g) Kinderman, S. S.; de Gelder, R.; van Maarseveen, J. H.; Schoemaker,
H. E.; Hiemstra, H.; Rutjes, F. P. J. T. J. Am. Chem. Soc. 2004, 126, 4100-
4101. (h) Barluenga, J.; Mateos, C.; Aznar, F.; Valdes, C. Org. Lett. 2002,
4, 1971-1974.
* To whom correspondence should be addressed. Tel: 1-(613)-991-0855.
Fax: 1-(613)-952-9092.
(1) Michael, J. P. Nat. Prod. Rep. 2005, 22, 603-626 and previous reports
from this author.
10.1021/jo062057v CCC: $33.50 Published 2007 American Chemical Society
Published on Web 01/23/2007
1226
J. Org. Chem. 2007, 72, 1226-1234

Synthesis of Polyhydroxylated Quinolizidines
olefin followed by a nucleophilic substitution,10 a double
intramolecular N-alkylation involving an epoxide,11 and a
reductive double cyclization to bicyclic lactam.12 The majority
of these syntheses started from a carbohydrate, which takes
advantage of pre-existing stereocenters.
SCHEME 1. Nitrogen Bicycles by Double-Michael Addition
Surprisingly, polyhydroxylated quinozilidines have so far not
been discovered in Nature. Consequently, their synthesis has
attracted much less interest and is only scarcely reported. The
synthetic methods are based on those for quinozilidines/
indolizidines and polyhydroxylated indozilidines, which include
double-reductive amination,13 hetero-Diels-Alder reaction,14
intramolecular SN2 reactions involving a mesyl leaving group15
and an epoxide,11 cycloaddition via nitrones16 and sulfones,17
ring-closure metathesis,18 and ring expansion via an aziridinium
intermediate.19 Both carbohydrates and non-carbohydrates were
used as starting material.
Ideally, one would like to construct the nitrogen bicyclic
system in a single step in a highly stereoselective manner.
Although double-reductive amination of an azido-R,ω-dialde-
hyde13 and double-intramolecular N-alkylation11 are two excel-
lent examples, the synthesis of these intermediates itself was a
challenge that required careful control of the stereochemistry
during multistep manipulations. Our goal is to develop a more
general method that does not require a sophisticated technique,
uses common chemicals, and is applicable to different monosac-
charides.
During the course of our synthetic work toward aza-C-
glycosides20 via a base-mediated intramolecular nitrogen Michael
addition,21 we envisioned that C-glycosides equipped with an
amino group (e.g., at C5 of furanosides) may produce polyhy-
droxylated nitrogen bicycles (such as quinolizidines) as il-
lustrated in Scheme 1. This double-conjugate addition synthetic
pathway results in the creation of two stereocenters during the
carbon-nitrogen bond formation. Thus, we are also interested
in investigating the stereoselectivity of the reaction and how
substrate chirality of the monosaccharide and conformation of
the azasugar intermediate affect the stereochemical outcome.
Stereoselective conjugate addition of nitrogen nucleophiles
to R,â-unsaturated carbonyl compounds is widely used in
organic synthesis,22 e.g., as one of the most effective methods
for the synthesis of â-amino acids and â-lactams,23 and is often
(3) For examples, see: (a) Alcaide, B.; Almendros, P.; Alonso, J. M.;
Aly, M. F. Chem. Eur. J. 2003, 9, 3415-3426. (b) Palacios, F.; Alonso,
C.; Amezua, P.; Rubiales, G. J. Org. Chem. 2002, 67, 1941-1946. (c)
Barluenga, J.; Mateos, C.; Aznar, F.; Valde´s, C. Org. Lett. 2002, 4, 1971-
1974.
(4) For examples, see: (a) Maloney, K. M.; Danheiser, R. L. Org. Lett.
2005, 7, 3115-3118. (b) Yu, S. H.; Zhu, W.; Ma, D. W. J. Org. Chem.
2005, 70, 7364-7370. (c) Back, T. G.; Hamilton, M. D.; Lim, V. J. J.;
Parvez, M. J. Org. Chem. 2005, 70, 967-972. (d) Zhu, W.; Dong, D. P.;
Pu, X. T.; Ma, D. W. Org. Lett. 2005, 7, 705-708. (e) Huang, H. L.; Sung,
W. H.; Liu, R. S. J. Org. Chem. 2001, 66, 6193-6196. (f) Wrobleski, A.;
Sahasrabudhe, K.; Aube, J. J. Am. Chem. Soc. 2004, 126, 5475-5481. (g)
Amat, M.; Llor, N.; Hidalgo, J.; Escolano, C.; Bosch, J. J. Org. Chem.
2003, 68, 1919-1928. (h) Gracias, V.; Zeng, Y. B.; Desai, P.; Aube, J.
Org. Lett. 2003, 5, 4999-5001. (i) Gravier, P. C.; Maton, W.; Le Merrer,
Y. Synlett 2003, 333-336. (j) Huang, H. B.; Spande, T. F.; Panek, J. S.
J. Am. Chem. Soc. 2003, 125, 626-627. (k) Koskinen, A. M. P.; Kallatsa,
O. A. Tetrahedron 2003, 59, 6947-6954. (l) Reddy, P. G.; Varghese, B.;
Baskaran, S. Org. Lett. 2003, 5, 583-585. (m) Rejzek, M.; Stockman,
R. A. Tetrahedron Lett. 2002, 43, 6505-6506.
(5) (a) Alcaide, B.; Pardo, C.; Saez, E. Synlett 2002, 85-88. (b) Broggini,
G.; La, Rosa, C.; Pilati, T.; Terraneo, A.; Zecchi, G. Tetrahedron 2001,
57, 8323-8332.
(6) (a) Davis, F. A.; Yang, B. J. Am. Chem. Soc. 2005, 127, 8398-
8407. (b) Davis, F. A.; Mang, Y.; Anilkumar, G. J. Org. Chem. 2003, 68,
8061-8064.
(7) (a) Watson, A. A.; Fleet, G. W. J.; Asano, N.; Molyneux, R. J.; Nash,
R. J. Phytochemistry 2001, 56, 265-295. (b) Asano, N. Curr. Top. Med.
Chem. 2003, 3, 471-484. (c) Asano, N.; Nash, R. J.; Molyneux, R. J.;
Fleet, G. W. J. Tetrahedron: Asymmetry 2000, 11, 1645-1680.
(8) (a) Jacob, G. S. Curr. Opin. Struct. Biol. 1995, 5, 605-611. (b) Goss,
P. E.; Reid, C. L.; Bailey, D.; Dennis, J. W. Clin. Cancer Res. 1997, 3,
1077-1086. (c) Goss, P. E.; Baker, M. A.; Carver, J. P.; Dennis, J. W.
Clin. Cancer Res. 1995, 1, 935-944.
(9) (a) Nemr, A. L. Tetrahedron 2000, 56, 8579-8629. (b) Stu¨tz, A.
Iminosugars as glycosidase inhibitors; Wiley-VCH: Weinheim, 1999. (c)
Look, G. C.; Fotsch, C. H.; Wong, C.-H. Acc. Chem. Res. 1993, 26, 182-
190. (d) Wong, C.-H.; Halcomb, R. L.; Ichikawa, Y.; Kajimoto, T. Angew.
Chem., Int. Ed. 1995, 34, 521-546.
(17) Carretero, J. C.; Arrayas, R. G.; deGracia, I. S. Tetrahedron Lett.
1997, 38, 8537-8540.
(18) (a) Dhavale, D. D.; Jachak, S. M.; Karche, N. P.; Trombini, C.
Synlett 2004, 1549-1552. (b) Overkleeft, H. S.; Bruggeman, P.; Pandit,
U. K. Tetrahedron Lett. 1998, 39, 3869-3872.
(19) Verhelst, S. H. L.; Martinez, B. P.; Timmer, M. S. M.; Lodder, G.;
van der Marel, G. A.; Overkleeft, H. S.; van Boom, J. H. J. Org. Chem.
2003, 68, 9598-9603.
(20) Yi, T.; Wu, A.; Wu, S.-H.; Zou, W. Tetrahedron 2005, 61, 11716-
11722.
(10) Bennett, R. B.; Choi, J. R.; Montgomery, W. D.; Cha, J. K. J. Am.
Chem. Soc. 1989, 111, 2580. (b) Pearson, W. H.; Lin, K. C. Tetrahedron
Lett. 1990, 31, 7571-7574.
(11) (a) Pearson, W. H.; Hembre, E. J. J. Org. Chem. 1996, 61, 5537-
5545. (b) Pearson, W. H.; Hembre, E. J. J. Org. Chem. 1996, 61, 5546-
5556.
(21) (a) Zou, W.; Lacroix, E.; Wang, Z. R.; Wu, S. H. Tetrahedron Lett.
2003, 44, 4431-4433. (b) Wang, Z.; Shao, H.; Lacroix, E.; Wu, S. H.;
Jennings, H. J.; Zou, W. J. Org. Chem. 2003, 68, 8097-8105. (c) Shao,
H.; Wang, Z.; Lacroix, E.; Wu, S. H.; Jennings, H. J.; Zou, W. J. Am.
Chem. Soc. 2002, 124, 2130-2131.
(22) (a) Perlmutter, P. Conjugate Addition Reactions in Organic Syn-
thesis; Pergamon Press: Oxford, 1992. (b) Rossiter, B. E.; Swingle, N. M.
Chem. ReV. 1992, 92, 771-806. (c) Tomioka, K.; Nagaoka, Y. In
ComprehensiVe Asymmetric Catalysis; Jacobsen, E. N., Pfaltz, A., Yama-
moto, H., Eds.; Springer: Berlin, 1999; Vol. III, Chapter 31.1. (d) Tomioka,
K. In Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley-VCH: Weinheim,
2000; Chapter 12. (e) Sibi, M. P.; Manyem, S. Tetrahedron 2000, 56, 8033-
8061. (f) Krause, N.; Hoffmann-Ro¨der, A. Synthesis 2001, 171-196.
(23) (a) Cole, D. C. Tetrahedron 1994, 50, 9517-9582. (b) Juaristi, E.;
Quintana, D.; Escalante, J. Aldrichim. Acta 1994, 27, 3-11. (c) Cardillo,
G.; Tomasini, C. Chem. Soc. ReV. 1996, 117-128. (d) Liu, M.; Sibi, M. P.
Tetrahedron 2002, 58, 7991-8035. (e) Recent Progress in the Chemical
Synthesis of Antibiotics; Lukacs, G., Ohno, M., Eds.; Springer-Verlag:
Berlin, 1990.
(12) (a) Setoi, H.; Takeno, H.; Hashimoto, M. J. Org. Chem. 1985, 50,
3948-3950. (b) Hembre, E. J.; Pearson, W. H. Tetrahedron 1997, 53,
11021-11032.
(13) Gradnig, G.; Berger, A.; Grassberger, V.; Stutz, A. E. Tetrahedron
Lett. 1991, 32, 4889-4892.
(14) Herczegh, P.; Kova´cs, I.; Szila´gyi, L.; Sztaricskaia, F.; Amaya,
Berecibar, A.; Claude, Riche, C.; Chiaroni, A.; Olesker, A.; Lukacs, G.
Tetrahedron 1995, 51, 2969-2978.
(15) Liu, P. S.; Rogers, R. S.; Kang, M. S.; Sunkara, P. S. Tetrahedron
Lett. 1991, 32, 5853-5856.
(16) (a) Gebarowski, P.; Sas, W. Chem. Commun. 2001, 915-916. (b)
Dhavale, D. D.; Jachak, S. M.; Karche, N. P.; Trombini, C. Tetrahedron
2004, 60, 3009-3016.
J. Org. Chem, Vol. 72, No. 4, 2007 1227

Zou et al.
achieved using chiral acceptors,24 chiral amines,25 and chiral
catalysts.26 The R,â-unsaturated carbonyl generated by â-elim-
ination from amino-C-glycoside (see Scheme 1) can be con-
sidered as both chiral amine and chiral Michael acceptor;
therefore, one would expect the conjugate addition to be
stereoselective. Using ribose- and arabinose-derived amino-C-
glycosides as substrates, we describe here the feasibility of this
double-Michael addition for the synthesis of nitrogen bicycles
and the factors affecting the stereochemistry.
SCHEME 2. Synthesis of Polyhydroxylated Quinolizidines
Synthesis of Polyhydroxylated Quinolizidines. Instead of
an R,â-conjugated ketone as an aglycon, we decided to
synthesize â,γ-conjugated ketones, namely 1-C-(2′-oxo-4′-
pentenyl)-glycosides,27 as substrates to further exploit our
previous observation that the â,γ-double bond of a 2′-ketone
C-glycoside migrated under basic conditions to form an R,â-
conjugated ketone.21b,c Therefore, the first test substrate was the
ribose derivative 7 (see Scheme 2) since we have previously
obtained 5-azido-C-riboside 2 via allyl-1-C-riboside 1.21
The allyl double bond of 2 was oxidized to aldehyde 3 by
ozonolysis in good yield. The resultant aldehyde was reacted
with allylmagnesium bromide (AllylMgBr) to afford a mixture
of two diastereomers (4) in a ratio of ca. 5:1. No investigation
on stereochemistry of the products was attempted. Instead, the
mixture of alcohols was oxidized using PCC to ketone 7 in 77%
yield. However, the Grignard reaction, which converts 3 to 4,
was somewhat unpredictable and often gave low yields (20-
54%) likely due to a side reaction involving an azido group
since we observed decomposition of azido-C-glycoside 2 when
treated with AllylMgBr. One of the difficulties associated with
the Grignard reaction was the adjustment of the amount of
AllylMgBr to be used to compensate for deterioration of the
commercial reagent. We often found either starting material
remained or the product significantly decomposed.28 In order
to circumvent this problem, we modified the reaction sequence
by introducing the azido group after Grignard reaction. Thus,
C-glycoside 1 was converted quantitatively to aldehyde 5 by
ozonolysis without chromatographic purification. The aldehyde
was then reacted with AllylMgBr to give a diastereomeric
mixture of 2′,5-diols (6) in 59% yield. After the 5-OH was
converted to the 5-azido group by O-tosylation followed by
treatment with sodium azide to give 4, the 2′-OH was
subsequently oxidized to obtain azidoketone 7. Although this
procedure gave a moderate yield, few chromatography purifica-
tions were required. Actually, we were able to just purify 6 in
this synthetic pathway to 7.
Nevertheless, with substrate 7 in hand, we performed a
Staudinger reduction of the azido group to amine using Ph3P.29
After removal of solvent, the resulting amine, without further
purification, was treated with base (4% NaOMe or 1% K2CO3
in methanol). The first Michael addition was complete within
4 h as indicated by the appearance of a faster moving spot on
TLC, and this intermediate was then further converted overnight
to an even less polar product 8, a quinolizidine derivative
isolated by chromatography (76% from 7) as a single diaste-
reomer.
This one-pot transformation from C-glycoside 7 to nitrogen
bicyclic 8 likely starts with base-mediated â-elimination (see
Scheme 3). The resultant R,â-conjugated ketone 11 in turn
underwent an intramolecular nitrogen conjugate addition to form
azasugar 12. Meanwhile, as expected, the â,γ-double bond
spontaneously migrated under the same conditions to produce
13 with an R,â-conjugated ketone functionality, which under-
went a second intramolecular conjugate addition by azasugar
nitrogen to afford 8. In fact, the formation of only one
diastereomer indicates both intramolecular conjugate additions
(24) (a) Shimano, M.; Meyers, A. I. J. Am. Chem. Soc. 1994, 116, 6437-
6438. (b) Dumas, F.; Mezrhab, B.; d’Angelo, J. J. Org. Chem. 1996, 61,
2293-2304. (c) Asao, N.; Shimada, T.; Sudo, T.; Tsukada, N.; Yazawa,
K.; Gyoung, Y. S.; Uyehara, T.; Yamamoto, Y. J. Org. Chem. 1997, 62,
6274-6282.
(25) (a) Furukawa, M.; Okawara, T.; Terawaki, Y. Chem. Pharm. Bull.
1977, 25, 1319-1325. (b) Hawkins, J. M.; Fu, G. C. J. Org. Chem. 1986,
51, 2820-2822. (c) Davies, S. G.; Ichihara, O. Tetrahedron: Asymmetry
1991, 2, 183-186. (d) Rico, J. G.; Lindmark, R. J.; Rogers, T. E.; Bovy,
P. R. J. Org. Chem. 1993, 58, 7948-7951. (e) Enders, D.; Wahl, H.; Bettray,
W. Angew. Chem., Int. Ed. Engl. 1995, 34, 455-457. (f) Sewald, N.; Hiller,
K. D.; Helmreich, B. Liebigs Ann. 1995, 925-928. (g) Leroux, M. L.; Gall,
T.; Mioskowski, C. Tetrahedron: Asymmetry 2001, 12, 1817-1823. (h)
Bull, S. D.; Davies, S. G.; Robert, P. M.; Savory, E. D.; Smith, A. D.
Tetrahedron 2002, 58, 4629-4642.
(26) (a) Sibi, M. P.; Shay, J. J.; Liu, M.; Jasperse, C. P. J. Am. Chem.
Soc. 1998, 120, 6615-6616. (b) Myers, J. K.; Jacobsen, E. N. J. Am. Chem.
Soc. 1999, 121, 8959-8960. (c) Nelson, S. G.; Spencer, K. L. Angew.
Chem., Int. Ed. 2000, 39, 1323-1325. (d) Zhuang, W.; Hazell, R. G.;
Jorgensen, K. A. Chem. Commun. 2001, 1240-1241. (e) Sundararajan, G.;
Prabagaran, N. Org. Lett. 2001, 3, 389-392. (f) Nakama, K.; Seki, S.;
Kanemasa, S. Tetrahedron Lett. 2001, 42, 6719-6722. (g) Cardillo, G.;
Gentilucci, L.; Gianotti, M.; Kim, H.; Perciaccante, R.; Tolomelli, A.
Tetrahedron: Asymmetry 2001, 12, 2395-2398. (h) Sugihara, H.; Daikai,
K.; Jin, X. L.; Furuno, H.; Inanaga, J. Tetrahedron Lett. 2002, 43, 2735-
2739. (i) Guerin, D. J.; Miller, S. J. J. Am. Chem. Soc. 2002, 124, 2134-
2136.
(27) To avoid confusion in the assignment of NMR data the carbon
numbering used in this paper corresponds to the numbering of the original
carbohydrates.
(28) Better yield might be obtainable by stoicheiometric reaction of
freshly prepared AllylMgBr with azidoaldehyde.
(29) Patai, S. The Chemistry of the Azido Group; John Wiley & Sons:
London, 1971.
1228 J. Org. Chem., Vol. 72, No. 4, 2007

Synthesis of Polyhydroxylated Quinolizidines
SCHEME 3. Proposed Pathway of the Base-Mediated
One-Pot Reaction
FIGURE 1. Major NOEs observed in compounds 8, 9, 20, and 21.
as well as an anomerically pure 17R from 14R to investigate
the influence of the anomeric configuration, if any, on the
stereoselectivity of the conjugate additions. After subjecting
17R/â and 17R to Staudinger reduction and base treatment, we
isolated, from both reactions, a mixture of two diastereomers
(18 and 19) in a ratio of 3:1, which were inseparable by silica
gel chromatography, in 50-60% yield.
SCHEME 4. Synthesis of Polyhydroxylated Quinolizidines
The Stereochemistry of Quinolizidines. It is apparent that
the stereoselectivity of the conjugate additions can be affected
by the substrate chirality but is independent of the anomeric
configuration of azidoketone substrates. Double conjugate
addition could produce four diastereoisomers; however, we
obtained a single diastereomer 8 from 7 and a pair of
stereoisomers (18 and 19) from 17, indicating both intramo-
lecular additions as being stereoselective. The stereochemistry
of 8 and 18/19 was determined on the basis of various 1D- and
2D-NMR spectroscopic analyses and confirmed by the NMR
analysis on their de-O-benzylated products, namely, 9 and 20/
21, respectively.
The proton spectrum of 8 revealed H1 resonance at 3.10 ppm
with a coupling constant J1,2 ) 9.2 Hz and small couplings
between H2 and H3, and H3 and H4, which indicates the
azasugar ring in 8 was in a 4C1 conformation with an equatorial
substitution at C1. The stereochemistry at C4′ was assigned on
the basis of the observation of NOE between H1 and Me at
C4′ (see Figure 1). After removal of the O-benzyl groups, the
azasugar ring of quinolizidine 9 also adopts a 4C1 conformation
with 1,2-diaxial configuration clearly indicated by the coupling
constants, J1,2 ) 9.6 Hz, and the NOE between H2 and H4 (see
Figure 1).
The structures of 18 and 19 in a mixture were also determined
by various NMR analyses. The major compound 18 was
assigned on the basis of proton-proton couplings (COSY and
TOCSY) and NOEs. We observed small coupling constants
between adjacent azasugar ring protons (H1, H2, and H3) and
NOEs between H1 and H2, H2 and H3, and H3 and H4. These
results indicate that the azasugar ring must adopt a 4C1
conformation with an equatorial substitution at C1 (1,2-cis
configuration).31 In contrast, structural determination of the
minor compound 19 was less straightforward because of many
overlapping signals with 18. However, we observed a doublet
of doublet at 3.58 (J ) 8.8 and 9.2 Hz), a characteristic signal
of an axial proton that couples with two adjacent axial protons.
being highly stereoselective. Removal of O-benzyl groups by
catalytic hydrogenation provided trihydroxylated quinolizidine
derivative 9 in good yield.
In order to investigate the utility of this one-pot procedure,
we selected C-arabinoside azidoketone 17 as the next test
substrate. Using aldehyde 1430 as starting material, 17 was
synthesized by the same procedure as described above for
substrate 7 (see Scheme 4), which included (a) Grignard reaction
to 2′,5-diol 15 (76%); (b) azido-substitution at C5 to azide 16;
and (c) PCC oxidation of 2′-OH to 17 (35% for two steps). We
prepared an anomeric mixture of 17R/â (R/â 1:1) from 14R/â
(31) In the previous report (ref 20), we presumed a similar L-arabinoa-
zasugar adopting 4C1 conformation (equatorial orientation at C2 and C3
and axial at C4) and thus made an error in the assignment of the anomeric
stereochemistry.
(30) Yi, T.; Wu, S. H.; Zou, W. Carbohydr. Res. 2005, 340, 235-244.
J. Org. Chem, Vol. 72, No. 4, 2007 1229

Zou et al.
FIGURE 2. 400 MHz 1H NMR spectra of a mixture of 20/21 recorded in D2O at room temperature (A) and in D2O (pH 8.5) with the same sample
after 2 days at room temperature (B).
1D-TOCSY experiments provided evidence suggesting this
proton as H2, and the structure of 19 was proposed in which
the azasugar ring has a 1C4 conformation with a 1,2-trans
configuration as illustrated in Scheme 4. Both structures (18
and 19) were unambiguously confirmed by determination of
the structures of their de-O-benzylated products 20 and 21.
Although they were still a mixture, removal of the O-benzyl
25% of H3′ protons remained. Thus, regioselective modification
at C1′ may be possible.
In contrast to the chair conformation of azasugar ring, the
conformation of the ketone ring of 9, 20, and 2132 was less
certain. Slow conformational interchange around nitrogen,
especially in 20, results in broadened 13C NMR signals of C1′,
C3′, and Me at C4′ (see the Supporting Information). Conse-
quently, we were unable to detect these 13C-1H correlations
by HSQC spectroscopic analysis.33
Stereochemistry of Conjugate Additions. The difference
in diastereoselectivity resulting from the two substrates (ribose
and arabinose) suggests the stereochemistry of the sugar played
an important role, particularly in the first conjugate addition,
where ribo-substrate 7 gave single product 8, whereas arabino-
substrate 17 produced two isomers (18 and 19). The effect of
substrate chirality during the first conjugate addition is illustrated
in Figure 3. Both thermodynamic and steric effects contribute
to the stereoselectivity of azasugar formation. The thermody-
1
groups provided a H NMR spectrum which clearly indicated
that the azasugar ring of the major compound (20) adopted 4C1
conformation with a 1,2-cis configuration as evidenced by the
small coupling constant between H1 and H2, H2 and H3, and
H3 and H4 (actually broad singlets were obtained) due to their
axial-equatorial and equatorial-equatorial relationships (Figure
2A). A simpler 1H NMR spectrum was obtained by deuterium
replacement of the R-protons adjacent to the ketone group at
C1′ and C3′ (see Figure 2B); this transformation was completed
by dissolving the compounds (20/21) in D2O at pH 8.5 (adjusted
by the addition of Na2HPO4). Thus, we were able to unambigu-
ously assign all the protons of 20 and 21 based on various 2D-
NMR spectra (COSY, TOCSY, NOESY). The axial-axial
coupling constants of H1-H2 and H2-H3 (J1,2 ) 8.8 Hz and
J2,3 ) 9.6 Hz) and NOEs observed in 21 (see Figure 1) not
only revealed the structure of 21 but also confirmed the previous
structural assignment of 19. It is also noteworthy to mention
that the R-protons at C1′ were more readily exchanged with
deuterium than those at C3′ since we observed that the H1′
proton resonances had completely disappeared within 24 h while
(32) It is noteworthy that the hydroxylated quinolizidines 9, 20, and 21
described here possess the same or similar skeleton as naturally occurring
myrtine from plants and xestosin found from marine sponges. See: (a)
Slosse, P.; Hootele´, C. Tetrahedron 1981, 37, 4287-4294. (b) Iwagawa,
T.; Kaneko, M.; Okamura, H.; Nakatani, M.; van Soest, R. W. M.; Shiro,
M. J. Nat. Prod. 2000, 63, 1310-1311.
(33) We attempted to record NMR spectra at 60 °C; unfortunately,
epimerization at C4′ likely occurred in both 20 and 21, as indicated by the
proportional decrease of 4′-Me signals of 20 and 21 and by the appearance
of a methyl resonance at 1.24 ppm (J ) 5.9 Hz). See ref 32a.
1230 J. Org. Chem., Vol. 72, No. 4, 2007

Synthesis of Polyhydroxylated Quinolizidines
FIGURE 3. Stereoselectivity of the first Michael addition.
TABLE 1. Stereocontrol by Chelation
namic effect likely decides the initial conformation of the
transition state, where ribose intermediate 11 adopts a more
stable 4C1 conformation and arabinose intermediate favors 1C4
conformation, where both accommodate two equatorial substitu-
tions and one axial substitution.34 The key to the stereoselectivity
is the orientation of the Michael acceptor. For example, 11 may
lead to two transition states, 11-eq and 11-ax (see Figure 3),
but 11-eq which gave the equatorially substituted C-glycoside
was favored thermodynamically and sterically, while 11-ax with
an axial Michael acceptor was disfavored due to 1,3-diaxial
interactions. Consequently, aza-â-C-glycoside (12) was formed
as a single distereomeric intermediate. Similarly, two transition
base
NaOMe
80:20
LiOH
70:30
DBU
60:40
2,6-lutidine
55:45
TEA
26/27
45:55
intermediate 23 with 1,4-trans configuration, while 22-eq
provided minor intermediate 24. The metal chelation and/or
hydrogen bonding were proposed because we observed, in a
similar intramolecular conjugate addition using L-arabinose 25
as a substrate, the influence of base to the stereoselectivity (see
Table 1). Apparently, because of the depleted chelation due to
use of organic base, we were able to increase the ratio of R- to
â-aza-C-glycosides (27/26). The significant amount of 26
produced with organic base was likely due to the role played
by the intramolecular hydrogen-bonding effect (C-O-H‚‚‚
-NH) that could not be suppressed using organic base. Ad-
ditionally, it has also been reported in the literature that the
axial 4-O-acyl substitution of a glycosyl donor favors 1,4-trans
glycosylation as a result of remote participation that stabilizes
1
states with C4 conformation, 22-ax and 22-eq, were proposed
(see Figure 3). Because we obtained 18 as the major product
and 19 as the minor product from 17, azasugars 23 and 24 were
likely the intermediates. Therefore, the major transition state
must be 22-ax rather than 22-eq. Although 22-eq with an
equatorial Michael acceptor is favored thermodynamically, the
steric repulsion, likely resulting from metal ion chelation and/
or hydrogen bonding between the vicinal axial 4-OH and amino
groups in the transition state, acted in the opposite direction.
This steric repulsion appears more critical than the thermody-
namic effect to the transition state. In addition, the axial Michael
acceptor was not interfered with by 1,3-diaxial interaction. Thus,
the transition state 22-ax became favored to give the major
36
the anomeric oxacarbenium ion.35 However, there was no
evidence to suggest such participation in our case other than
(34) The conformation should unlikely be affected by the stereoelectronic
substituent effects which were observed upon protonation of azasugars.
See: (a) Jensen, H. H.; Bols, M. Acc. Chem. Res. 2006, 39, 259-265. (b)
Jensen, H. H.; Lyngbye, L.; Jensen, A.; Bols, M. Chem. Eur. J. 2002, 8,
1218-1226.
(35) (a) De Meo, C.; Kamat, M. N.; Demchenko, A. V. Eur. J. Org.
Chem. 2005, 706-711. (b) Cheng, Y. P.; Chen, H. T.; Lin, C. C.
Tetrahedron Lett. 2002, 43, 7721-7723. (c) Demchenko, A. V.; Rousson,
E.; Boons, G. J. Tetrahedron Lett. 1999, 40, 6523 -6526.
J. Org. Chem, Vol. 72, No. 4, 2007 1231

Zou et al.
The solvent was removed, and the crude was purified by chroma-
tography (hexane-EtOAc 3:1) to give 3 as a syrup (0.7 g, 70%).
1H NMR (CDCl3) δ: 2.81 (bdd, 1H, 1′-CH2, J ) 6.0, 15.6 Hz),
2.89 (dd, 1H, 1′-CH2, J ) 6.4, 15.6 Hz), 3.17 (dd, 1H, H-5a, J )
4.0, 13.2 Hz), 3.55 (dd, 1H, H-5b, J ) 3.6, 13.2 Hz), 4.02 (dd, 1H,
H-3, J ) 4.4, 7.6 Hz), 4.13 (dd, 1H, H-2, J ) 4.4, 4.4 Hz), 4.18
(m, 1H, H-5), 4.50 (d 1H, CH2Ph, J ) 11.6 Hz), 4.53 (m, 1H,
H-1), 4.53 (d 1H, CH2Ph, J ) 11.6 Hz), 4.68 (d 1H, CH2Ph, J )
11.6 Hz), 4.78 (d 1H, CH2Ph, J ) 11.6 Hz), 9.74 (s, 1H, CHO).
5-C-(5-Azido-2,3-di-O-benzyl-5-deoxy-R-D-ribofuranosyl)-1-
penten-4-ol (4). Method A. To a solution of aldehyde 3 (0.3 g,
0.79 mmol) in diethyl ether (10 mL) at -78 °C was dropwise added
a solution of 1 M AllMgBr in diethyl ether (2 mL, 2 mmol). The
mixture was stirred overnight at room temperature and diluted by
the addition of saturated 0.5 M aqueous HCl/EtOAc (1:1, 100 mL).
The organic phase was washed with water, dried, and concentrated.
Purification by chromatography (EtOAc-hexanes 1:3) afforded a
mixture of diastereomeric alcohols 4 (0.18 g, 54%) as a syrup.
Method B. To a solution of 1 M AllMgBr in diethyl ether (15
mL, 15 mmol) at -40 °C was dropwise added a solution of
aldehyde 5 (0.9 g, 2.53 mmol) in anhydrous THF (5 mL). After 5
min at -40 °C, the mixture was stirred at room temperature for 2
h. The reaction was quenched by the addition of aqueous NH4Cl,
and the solution was extracted with dichloromethane. The organic
phase was dried and concentrated. Purification by chromatography
(EtOAc-hexanes 1:1) gave alcohol 6 (0.6 g, 59%) as a syrup.
To a solution of 6 (0.14 g) in dichloromethane (10 mL) was
added a solution of TsCl (0.14 g) in pyridine (1.5 mL) at 0 °C.
The mixture was kept at 4 °C for 2 days. The excess amount of
TsCl was destroyed by the addition of MeOH, and the solution
was washed with 0.5 M HCl and brine. The organic phase was
dried and concentrated to a residue. To a solution of the residue in
DMF (7 mL) was added NaN3 (0.1 g), and the mixture was stirred
at 80 °C for 4 h. Upon cooling, the mixture was diluted by the
addition of water, and the aqueous solution was extracted with ethyl
acetate. The combined organic solution was dried and concentrated.
Purification by chromatography (EtOAc-hexanes 1:3) gave 4 as a
syrup (60 mg, 41%). 1H NMR (CDCl3) δ: 1.72 (m, 1H), 1.92 (m,
1H), 2.23 (m, 2H), 3.14 (m, 1H), 3.54 (m, 1H), 3.77-3.85 (m,
1H), 3.99-4.03 (m, 2H), 4.19-4.32 (m, 2H), 4.47-4.82 (m, 4H,
2 × CH2Ph), 5.08-5.14 (m, 2H), 5.75 (m, 1H), 7.29-7.37 (m,
10H).
FIGURE 4. Stereoselectivity of the second Michael addition.
the steric repulsion. The mechanism proposed in Figure 3 is
different from those in a typical Michael addition where the
metal ion chelation occurs between carbonyl group and the
nucleophile.36
The second conjugate addition was totally stereoselective
because the existing azasugar ring likely directed the stereo-
chemistry. The â,γ-double bond migration prior to the conjugate
addition must have produced the thermodynamic stable E-
isomers, e.g., 13-E (see Figure 4). Subsequently, the steric
repulsion between the methyl group and the azasugar ring
dictated the transition-state topicity. Although the lone electron
pair of nitrogen may approach the Michael acceptor from both
axial and equatorial direction, regardless, both transition states
afforded the same conjugation addition product. As a result,
we were able to obtain complete diastereoselectivity in the
conjugation addition.
5-C-(5-Azido-2,3-di-O-benzyl-5-deoxy-R-D-ribofuranosyl)-1-
penten-4-one (7). To a mixture of alcohol 4 (0.13 g, 0.31 mmol),
NaOAc (0.1 g), and 4 Å molecular sieves (0.2 g) in DCM (10 mL)
was added PCC (0.15 g, 0.7 mmol) at room temperature. The
mixture was stirred for 5 h, and the filtrate was washed and dried.
Purification by chromatography (EtOAc-hexanes 1:3) afforded
In summary, we described here an effective method for the
synthesis of polyhydroxylated quinolizidines by a double-
conjugate addition. The evidence from our work suggests that
the transition state of the first conjugate addition of nitrogen,
which produces azasugar, is controlled by both the steric and
thermodynamic effects, whereas the second conjugate addition
is highly specific and entirely controlled by steric effect. This
synthesis takes advantage of substrate chirality rather than use
of chiral catalysts and is simple and easy to reproduce. In
addition, the intermediates with hydroxyl and ketone function-
alities should allow us to further modify these molecules to
generate molecular diversities, which may provide critical value
to lead discovery.
1
ketone 7 (0.1 g, 77%) as a syrup. [R]D: +36 (c 0.4, MeOH); H
NMR (CDCl3) δ: 2.82 (dd, 1H, CH2, J ) 6.0, 17.6 Hz), 2.94 (dd,
1H, CH2, J ) 8.0, 17.6 Hz), 3.09 (bd, 1H, CH2CHdCH2, J ) 6.8
Hz), 3.11 (bd, 1H, CH2CHdCH2, J ) 6.8 Hz), 3.16 (dd, 1H, H-5a,
J ) 4.4, 13.2 Hz), 3.52 (dd, 1H, H-5b, J ) 3.2, 13.2 Hz), 3.99 (dd,
1H, H-3, J ) 4.0, 7.6 Hz), 4.14 (dd, 1H, H-4, J ) 4.0, 7.6 Hz),
4.17 (dd, 1H, H-2, J ) 3.6, 4.4 Hz), 4.49 (m, 1H, H-1), 4.44 and
4.79 (d and d, 1H each, CH2Ph, J ) 11.2 Hz), 4.51 and 4.68 (d
and d, 1H each, CH2Ph, J ) 12.0 Hz), 5.07 (d, 1H, CHdCH2, J )
16.4 Hz), 5.14 (d, 1H, CHdCH2, J ) 10.4 Hz), 5.84 (m, 1H,
CH)CH2), 7.28-7.36 (m, 10 H, 2 × Ph). 13C NMR (CDCl3) δ:
42.8 (C-1′), 48.4 (C-3′), 52.3 (C-5), 73.2 (CH2Ph), 74.1 (CH2Ph),
74.4 (C-1), 76.5 (C-2), 78.7 (C-4), 80.6 (C-3), 119.3 (C-5′), 128.0,
128.3, 128.6, 127.8, 130.4 (C-4′), 137.3 (Ph), 138.3 (Ph), 207.3
(C-2′). HRMS: calcd for C24H28 N3O4 [M + H] 422.2084, found
422.2079.
Experimental Section
2-C-(5-Azido-2,3-di-O-benzyl-5-deoxy-R-D-ribofuranosyl)ac-
etaldehyde (3). A solution of 2 (1.0 g, 2.64 mmol) in CH2Cl2 (40
mL) was bubbled with ozone at -78 °C until a blue color appeared
(10 min). The solution was concentrated to a residue, which was
treated with dimethyl sulfide (5 mL) at room temperature overnight.
Di-O-benzylquinolizidine 8. To a solution of azide 7 (55 mg,
0.131 mmol) in THF-H2O (20:1, 6 mL) was added Ph3P (60 mg),
and the mixture was stirred at room temperature overnight.
Complete reduction of the azide to amine was detected by TLC
(36) Gawley, R. E.; Aube, J. Principles of asymmetric synthesis, 1st ed.;
Pergamon: Oxford, U.K., 1996.
1232 J. Org. Chem., Vol. 72, No. 4, 2007

Synthesis of Polyhydroxylated Quinolizidines
(EtOAc-MeOH 5:1, Rf 0.15). The solvent was removed by
evaporation under diminished pressure to a residue, to which was
added 4% NaOMe (3.6 mL). After 2 h at room temperature, the
starting material had disappeared and an intermediate with a slightly
higher Rf was formed. The mixture was continuously stirred
overnight to give a much less polar product. The mixture was diluted
with water and extracted with ethyl acetate. The combined organic
solution was dried and concentrated. Purification by chromatography
(EtOAc to EtOAc/MeOH 5:1) gave compound 8 as a syrup (39
mg, 76%). [R]D: -19.2 (c 2.0, CHCl3). 1H NMR (CDCl3) δ: 0.97
(d, 3H, CH3, J ) 6.8 Hz), 2.00 (dd, 1H, H-1′a, J ) 10.8, 14.8 Hz),
2.18 (ddd, 1H, H-3′a, J ) 2.0, 2.4, 14.0 Hz), 2.26 (bs, 1H, 4-OH),
2.60-2.75 (m, 4H, H-5a, 5b, 1′b, 3′b), 3.10 (ddd, 1H, H-1, J )
4.0, 9.2, 10.8 Hz), 3.23 (dd, 1H, H-2, J ) 2.0, 9.2 Hz), 3.36 (m,
1H, H-4′), 3.75 (m, 1H, H-4), 4.05 (bs, 1H, H-3), 4.50 and 4.74 (d
and d, 1H each, CH2Ph, J ) 11.6 Hz), 4.56 and 5.03 (d and d, 1H
each, CH2Ph, J ) 11.6 Hz), 7.28-7.36 (m, 10 H, 2 × Ph). 13C
NMR (CDCl3) δ: 11.7 (CH3), 44.4 (C-1′), 48.6 (C-3′), 51.8 (C-1),
52.8 (C-5), 56.5 (C-4′), 67.8 (C-4), 72.8 (CH2Ph), 74.6 (CH2Ph),
74.9 (C-3), 83.5 (C-2), 128.0, 128.1, 128.3, 128.8, 137.5, 138.7,
209.2 (C-2′). HRMS: calcd for C24H30 NO4 [M + H] 396.2174,
found 396.2174.
chromatography (EtOAc-hexanes 1:3) afforded ketone 17R as a
1
wax (0.13 g, 35% over three steps). [R]D: +6 (c 4.0, MeOH). H
NMR (CDCl3) δ: 2.75 (dd, 1H, H-1′a, J ) 7.6, 16.8 Hz), 2.82
(dd, 1H, H-1′b, J ) 6.4, 16.8 Hz), 3.17 (bd, 2H, 3′-CH2, J ) 7.2
Hz), 3.31 (dd, 1H, H-5a, J ) 5.6, 12.8 Hz), 3.41 (dd, 1H, H-5b,
J ) 6.8, 12.8 Hz), 3.89 (bs, 1H, H-2), 3.91 (bs, 1H, H-3), 4.15 (m,
1H, H-4), 4.56 (m, 1H, H-1), 4.47 and 4.51 (d and d, 1H each,
CH2Ph, J ) 13.2 Hz), 4.55 and 4.63 (d and d, 1H each, CH2Ph,
J ) 11.8 Hz), 5.12 (d, 1H, H-5′a, J ) 17.2 Hz), 5.18 (d, 1H, H-5′b,
J ) 10.4 Hz), 5.87 (m, 1H, H-4′), 7.28-7.36 (m, 10 H, 2 × Ph).
13C NMR (CDCl3) δ: 41.9 (C-1′), 48.4 (C-3′), 52.7 (C-5), 72.0
(CH2Ph), 72.2 (CH2Ph), 79.3 (C-1), 82.4 (C-4), 85.5 (C-2), 86.4
(C-3), 119.4 (C-4′), 127.8, 127.9, 128.2, 128.6, 127.7, 130.4 (C-
5′), 137.5, 137.8, 206.6 (C-2′). HRMS: calcd for C24H28 N3O4
[M + H] 422.2084, found 422.2088.
The anomeric mixture, 17 with 1:1 R/â ratio, was prepared by
same procedure and used for the one-pot reaction as well. For the
NMR spectra, see the Supporting Information.
Protected Quinolizidines 18 and 19. To a solution of ketone
17R (and 17) (100 mg, 0.238 mmol) in THF-H2O (20:1, 3 mL)
was added Ph3P (100 mg), and the solution was stirred at room
temperature overnight. The solvent was evaporated to a residue,
which was dissolved in 4% NaOMe (3 mL). The mixture was stirred
at room temperature overnight, diluted with water, and extracted
with ethyl acetate. The combined organic solution was dried and
concentrated. Purification by chromatography (hexanes-EtOAc 1:4)
afforded a mixture of 18 and 19 as a syrup in a ratio of 3:1 (55
mg, 59%). HRMS: calcd for C24H30 NO4 [M + H] 396.2174, found
396.2152.
Trihydroxyquinolizidine 9. A mixture of 8 (19 mg, 0.048 mmol)
and 10% Pd-C (50% wet, 20 mg) in MeOH-HOAc (10:1, 5 mL)
was subjected to hydrogenation (40 psi) overnight. The catalyst
was removed by centrifugation, and the solvent was evaporated to
a residue, which was dissolved in water. The product obtained by
lyophilization was dissolved in water, and the solution was adjusted
to pH 8 with 0.1 N NaOH before lyophilization. Purification by
chromatography (EtOAc-MeOH 5:1) gave the final trihydroxylated
quinolizidine 9 as a white solid (8 mg, 77%). [R]D: -6.5 (c 0.1,
1
18 (Major). H NMR (CDCl3) δ: 0.97 (d, 3 H, Me, J ) 6.8
Hz), 1.94 (bd, 1H, H-1′a, J ) 14.0 Hz), 2.12-2.17 (m, 1H, H-3′a),
2.62-2.70 (m, 2H, H-1′b, 5a), 2.76-2.80 (m, 2H, H-3′b, 5b), 3.09
(bd, 1H, H-1, J ) 10.8 Hz), 3.34 (bs, 1H, H-2), 3.44 (m, 1H, H-4′),
3.69 (dd, 1H, H-3, J ) 4.0, 3.2 Hz), 4.12 (m, 1H, H-4), 4.45 and
4.51 (d and d, 1H each, CH2Ph, J ) 12.0 Hz), 4.48 and 4.57 (d
and d, 1H each, CH2Ph, J ) 12.0 Hz), 7.24-7.60 (m, 10H, 2 ×
Ph). 13C NMR (CDCl3) δ:11.9 (CH3), 41.7 (C-1′), 47.9 (C-3′), 52.2
(C-1), 52.5 (C-5), 56.7 (C-4′), 66.0 (C-4), 73.2 (CH2Ph), 73.8 (CH2-
Ph), 75.3 (C-3), 75.8 (C-2), 128.1, 128.2, 128.4, 128.6, 128.7, 128.8,
128.9, 129.0, 137.3, 138.0, 210.1 (C-2′).
1
MeOH). H NMR (D2O) δ: 1.00 (d, 3H, CH3, J ) 6.8 Hz), 2.27
(m, 1H, H-1′a), 2.29 (m, 1H, H-3′a), 2.60 (m, 1H, H-5eq), 2.64
(m, 1H, H-3′b), 2.75 (dd, 1H, H-5ax, J ) 10.8, 10.8 Hz), 2.82 (m,
1H, H-1′b), 2.92 (m, 1H, H-1), 3.44 (dd, 1H, H-2, J ) 2.8, 9.6
Hz), 3.48 (m, 1H, H-4′), 3.87 (m, 1H, H-4), 4.02 (dd, 1H, H-3,
J ) 2.4, 2.8 Hz). 13C NMR (D2O) δ: 10.8 (CH3), 43.2 (C-3′), 47.5
(C-1′), 49.3 (C-5), 52.0 (C-1), 55.9 (C-4′), 66.6 (C-4), 70.3 (C-3),
73.4 (C-2), 214.4 (C-2′). HRMS: calcd for C10H18 NO4 [M + H]
216.1226, found 216.1235.
1
19 (Minor). H NMR (CDCl3, assigned based on 1D-TOCSY)
5-C-(5-Azido-2,3-di-O-benzyl-5-deoxy-R-D-arabinofuranosyl)-
1-penten-4-one (17R). To a solution of aldehyde 14R (0.45 g, 1.26
mmol) in anhydrous THF (5 mL) at -40 °C was dropwise added
a solution of AllylMgBr in diethyl ether (7 mL, 7 mmol). After 5
min at -40 °C, the mixture was stirred at room temperature for 3
h. The reaction was quenched by the addition of 0.5 M HCl, and
the solution was extracted with dichloromethane. The organic phase
was dried and concentrated. Purification by chromatography
(EtOAc-hexanes 1:3) afforded a diastereomeric mixture of 15R
(0.38 g, 76%) as a syrup.
δ: 0.93 (d, 3H, Me, J ) 6.8 Hz), 2.20 (dd, 1H, H-1′a, J ) 10.4,
14.0 Hz), 2.62-2.70 (m, 2H, H-1, 1′b, 5a), 2.76-2.80 (m, 2H,
H-3′b, 5b), 3.09 (bd, 1H, H-1, J ) 10.8 Hz), 3.42 (dd, 1H, H-3,
J ) 3.2, 9.2 Hz), 3.44 (m, 1H, H-4′), 3.58 (dd, 1H, H-2, J ) 8.8,
9.2 Hz), 4.12 (m, 1H, H-4), 4.60 and 4.94 (d and d, 1H each, CH2-
Ph, J ) 12.0 Hz), 4.69 and 4.74 (d and d, 1H each, CH2Ph, J )
12.0 Hz), 7.24-7.60 (m, 10H, 2 × Ph). 13C NMR (CDCl3) δ: 11.2
(CH3), 44.4 (C-1′), 48.5 (C-3′), 54.4 (C-1), 56.5 (C-5), 57.4 (C-
4′), 65.8 (C-4), 72.0 (CH2Ph), 75.3 (CH2Ph), 81.4 (C-2), 82.7 (C-
3), 128.1, 128.2, 128.4, 128.6, 128.7, 128.8, 128.9, 129.0, 138.0,
138.2, 208.7 (C-2′).
To a solution of 15R (0.35 g, 0.88 mmol) in dichloromethane
(15 mL) was added a solution of TsCl (0.35 g, 1.8 mmol) in pyridine
(1.5 mL) at 0 °C. The mixture was kept at 4 °C overnight and then
at room temperature for additional 6 h. The excess amount of TsCl
was destroyed by the addition of MeOH, and the solution was
washed with 0.5 M HCl and brine. The organic phase was dried
and concentrated to a residue. To a solution of above residue in
DMF (15 mL) was added NaN3 (0.3 g). The mixture was stirred at
80 °C for 6 h. Upon cooling, the mixture was diluted by the addition
of water, and the aqueous solution was extracted with ethyl acetate.
The combined organic phase was dried and concentrated to azide
16R as a syrup.
Trihydroxyquinolizidines 20 and 21. A mixture of 18/19 (24
mg, 0.06 mmol) and 10% Pd-C (50% wet, 20 mg) in MeOH-
HOAc (10:1, 5 mL) was subjected to hydrogenation (40 psi)
overnight. The catalyst was removed by centrifugation, and the
solvent was evaporated. The residue was dissolved in water and
lyophilized. The solid obtained was redissolved in water, and the
solution was adjusted to pH 8 with 0.1 N NaOH and lyophilized
again to give crude product. Purification by chromatography
(EtOAc-MeOH 5:1) gave the final product 20/21 as a white solid
(9 mg, 69%). HRMS: calcd for C10H18 NO4 [M + H] 216.1235,
found 216.1221.
1
To a solution of above syrupy 16R in dichloromethane (20 mL)
were added NaOAc (0.2 g), 4 Å molecular sieves (0.4 g), and PCC
(0.4 g, 1.86 mmol) at room temperature. The mixture was stirred
for 5 h, and the filtrate was washed and dried. Purification by
20 (Major). H NMR (CDCl3) δ: 1.04 (d, 3H, Me, J ) 6.8
Hz), 2.22-2.28 (m, 2H, H-1′a, 3′a), 2.64-2.90 (m, 4H, H-1′b, 3′b,
5a, 5b), 3.24 (bs, 1H, H-1), 3.48 (m, 1H, H-4′), 3.75 (bs, 1H, H-2),
3.96 (bs, 1H, H-3), 4.14 (m, 1H, H-4). 13C NMR (CDCl3) δ: 11.0
J. Org. Chem, Vol. 72, No. 4, 2007 1233

Zou et al.
(Me), 49.8 (C-5), 52.1 (C-1), 56.3 (C-4′), 65.3 (C-4), 69.6 (C-3),
70.7 (C-2), 216.1 (C-2′).
21 (Minor). H NMR (CDCl3) δ: 0.97 (d, 3H, Me, J ) 6.8
Acknowledgment. This is NRCC publication no. 42513. We
thank Ken Chan for mass spectroscopic analysis and Dean
Williams for helpful discussions.
1
Hz), 2.45 (m, 1H, H-3′a), 2.46 (dd, 1H, H-1′a, J ) 9.8, 14.4 Hz),
2.68 (m, 1H, H-1), 2.78-2.81 (m, 2H, H-3, 5a), 2.91 (dd, 1H, H-5b,
J ) 3.2, 13.2 Hz), 3.43-3.50 (m, 2H, H-3, 4′), 3.58 (dd, 1H, H-2,
J ) 8.8, 10.0 Hz), 4.08 (m, 1H, H-4). 13C NMR (CDCl3) δ: 10.2
(Me), 54.1 (C-5), 56.0 (C-4′), 57.6 (C-1), 67.9 (C-4), 73.29 (C-3),
73.32 (C-2), 215.2 (C-2′).
1
Supporting Information Available: General methods and H
and 13C NMR spectra of products 7-9 and 17-21. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO062057V
1234 J. Org. Chem., Vol. 72, No. 4, 2007