Convenient syntheses of isomaltose derivatives from amygdalin

Article abstract of DOI:10.1016/j.tetlet.2004.04.057

The isomaltose trichloroacetimidate 7 was synthesized in five steps from D-amygdalin. The key step in this series of reactions was the acid catalyzed rearrangement of the inter-glycosydic bond to give the thermodynamically more stable α-anomer. The reacti

Full text of DOI:10.1016/j.tetlet.2004.04.057

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 4749–4753
Convenient syntheses of isomaltose derivatives from amygdalin
ꢀ*
Martin Chwalek and Karen Ple
CNRS FRE 2715 Isolement, Structure, Transformation et Synthꢁese de Substances Naturelles, Universitꢀe de Reims,
Champagne-Ardenne, IFR 53 Biomolꢀecules, CPCBAI, B^at. 18, Moulin de la Housse, BP 1039, 51687 Reims, France
Received 26 February 2004; revised 9 April 2004; accepted 13 April 2004
Abstract—The isomaltose trichloroacetimidate 7 was synthesized in five steps from
D-amygdalin. The key stepin this series of
reactions was the acid catalyzed rearrangement of the inter-glycosydic bond to give the thermodynamically more stable a-anomer.
The reaction was also applied to different di-, tri-, and tetrasaccharide derivatives of amygdalin giving the corresponding re-
arrangement products.
ꢀ 2004 Elsevier Ltd. All rights reserved.
The synthesis of isomaltose (1) (Fig. 1), from mono-
saccharide precursors was first reported by Wolfram in
1961.1 Since then, a multitude of papers have reported
the chemical synthesis of this simple molecule.2 Its
synthetic challenge resides in the stereoselective 1,2-cis
O-glycosylation reaction used to couple the monosac-
charide derivatives. Total selectivity in these types of
reactions has proven to be more difficult to achieve than
with the corresponding 1,2-trans reactions.3 The
majority of synthetic routes to isomaltose use a bromide
as the glycosyl donor with a nonparticipating group in
position 2, thus taking advantage of the halide ion cat-
alyzed glycosylation introduced by Lemieux et al.4 to
achieve 1,2-cis stereoselectivity in disaccharide forma-
tion. To our knowledge, only one example exists in the
literature concerning the synthesis of isomaltose from a
disaccharide precursor. In it, isomaltose was prepared
by titanium tetrachloride mediated rearrangement of
gentiobiose, followed by treatment with mercuric ace-
tate in acetic acid to achieve the final isomaltose octa-
acetate in 37% yield.5
In the course of our work related to saponin syntheses,6
we wished to prepare a trichloroacetimidate derivative
of gentiobiose (2). It was found that the treatment of
per-benzoylated gentiobiose with HBr in acetic acid,
followed by hydrolysis of the anomeric bromide and
formation of the trichloroacetimidate gave a mixture of
products in low yield. To our surprise, the major iso-
lated product proved to be benzoylated isomaltose tri-
chloroacetimidate. Since the source of our gentiobiose
was
D-amygdalin (3), we surmised that application of
the same reaction conditions to amygdalin might lead to
an isomaltose derivative. The advantage of using
amygdalin instead of gentiobiose as an access to iso-
maltose was evident when comparing the initial cost of
the two commercially available compounds. We would
now like to report an efficient synthesis of isomal-
OH
OH
OH
O
O
O
HO
HO
O
O
HO
HO
HO
HO
HO
HO
HO
HO
HO
HO
HO
O
O
O
O
CN
O
HO
HO
OH
OH
OH
HO
OH
Amygdalin 3
Isomaltose 1
Gentiobiose 2
tose from an existing disaccharide precursor,
dalin.
D-amyg-
Figure 1.
The starting point of our study was the preparation of
the per-benzoylated amygdalin derivative 4. Treatment
of D-amygdalin with benzoyl chloride in pyridine gave 4
in 93% yield. HBr (33% in acetic acid) was then added to
a solution of 4 in CH2Cl2 at room temperature. The
reaction was complete in 4–6 h, and a mixture of two
Keywords: Isomaltose; Amygdalin; Acid catalyzed rearrangement;
6-O-a- -Glucopyranosyl-a- -glucospyranose.
* Corresponding author. Tel.: +33-3-2691-8419; fax: +33-3-2691-3596;
e-mail: karen.ple@univ-reims.fr
D
D
0040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.04.057

4750
M. Chwalek, K. Plꢀe / Tetrahedron Letters 45 (2004) 4749–4753
OBz
by trichloroacetimidate formation gave the isomaltose
derivative 7 in 54% overall yield from -amygdalin
OBz
O
O
BzO
HBr (33% in AcOH),
CH2Cl2, 75%
D
O
BzO
BzO
BzO
BzO
(Scheme 2). The isomaltose derivative was fully char-
acterized at this stage because the a-trichloroacetimidate
was the unique reaction product and thus facilitated
NMR interpretation.9 The overall yield could be
improved to 60% by performing the reactions without
intermediate purification or separation of the epimers at
the amide stage. Further structural confirmation was
BzO
BzO
BzO
O
O
O
O
CN
O
BzO
BzO
O
NH2
OBz
OBz
4
5
mixture of epimers
~5:1
Scheme 1.
obtained by preparation of methyl b-D-isomaltoside (8)
1
products (5) was obtained in 75% yield (Scheme 1).
NMR analysis of the separated products showed that
the acid catalyzed rearrangement had occurred, along
with hydrolysis of the nitrile groupinto a primary amide
and epimerization at the mandeloamide moiety.7 It
should be noted that no ‘amygdalin’ amide was
detected, the reaction being completely stereoselective.
from trichloroacetimidate 7 in two steps. H and 13C
NMR data were identical in all respects to previously
published data.10
The acidic rearrangement conditions were then applied
to another amygdalin derivative 9 with an acetate on the
terminal 60-hydroxyl group. Isomerization was success-
ful, and two additional steps led to compound 10 having
a free hydroxy groupin the 6 0 position, thus creating the
possibility for further transformations on the non-
reducing sugar (Scheme 3).
The primary amide was easily converted back into the
nitrile 6 in 85% yield in the presence of trifluoroacetic
anhydride and pyridine.8 Hydrogenation of 6 followed
OBz
OBz
O
O
BzO
BzO
BzO
BzO
BzO
BzO
O
O
(CF3CO)2O, py,
O
O
O
BzO
BzO
BzO
BzO
O
O
CN
dioxane, 85%
NH2
OBz
OBz
5
6
OBz
O
OH
BzO
BzO
1. HCO2NH4, Pd/C,
acetone, reflux
O
.
HO
HO
1. CH3OH, BF3 Et2O
BzO
O
HO
O
O
2. Et3N, CH3OH, H2O
BzO
BzO
2. CCl3CN, DBU, CH2Cl2,
91% 2 steps
O
HO
HO
OCH3
BzO
O
CCl3
OH
7
8
NH
54% overall yield
Scheme 2.
OR
O
OH
1. HBr (33% in AcOH),
CH2Cl2
O
O
BzO
BzO
BzO
BzO
BzO
BzO
BzO
BzO
O
O
2. (CF3CO)2O, py,
dioxane, 64% 2 steps
O
CN
O
BzO
BzO
O
CN
OBz
9, R=Ac
11, R=H
3. AcCl, CH3OH,
80%
BzO
10
OBz
O
BzO
BzO
12
BzO
OC(NH)CCl3
OBz
TMSOTf,
O
BzO
BzO
CH2Cl2, 0˚C, 85%
BzO
OBz
O
O
O
BzO
BzO
HBr (33% in AcOH),
CH2Cl2, 50%
O
BzO
BzO
BzO
O
BzO
O
BzO
BzO
O
O
BzO
O
O
BzO
BzO
BzO
O
CN
O
BzO
NH2
OBz
OBz
14
13
Scheme 3.

M. Chwalek, K. Plꢀe / Tetrahedron Letters 45 (2004) 4749–4753
4751
To further explore the limits of this reaction, we wished
to see if it would be successful on tri- or tetrasaccharides
derived from amygdalin. The trisaccharide 13 was
synthesized from compound 11 and 2,3,4,6-tetra-O-
It has been the authors experience that glycosylation
with benzoylated glucose derivatives often gives the
orthoester before acid catalyzed rearrangement to the b-
glycosylation product. A complex mixture of products
could then be expected resulting from the formation of
all possible alcohols, as well as their eventual recombi-
nations.
benzoyl-a-D-glucopyranosyl trichloroacetimidate 12 in
85% yield. Treatment with HBr gave the desired
rearrangement at both internal anomeric centers
furnishing the isomaltotriose amide derivative 14 in 50%
yield (Scheme 3). It was interesting to observe that no
apparent epimerization took place as only one amide
was isolated. Attempts to optimize the reaction condi-
tions at this point were unsuccessful. Starting the reac-
tion at low temperature (0 ꢁC) and leaving it for long
periods of time or letting it slowly warm to room tem-
perature was detrimental. Leaving the reaction at room
temperature overnight also caused the yield to drop to
20%.
The observed outcome of the reaction is contrary to the
above arguments in that the acid catalyzed rearrange-
ment gave the thermodynamically more stable a ano-
mers in spite of possible neighboring group
participation. No benzoyl orthoesters were detected.
The reaction mixtures gave one easily isolated major
product (as a mixture of ‘amido’ epimers for the disac-
charide). We initially hypothesized that a rupture of the
internal O-(C-1) bond (endocyclic cleavage) could result
in a ‘pre-b’ linear intermediate 23 where rotation around
the C–O bond could lead to the ‘pre-a’ intermediate 24.
Ring closure would then result in the formation of the
more stable a-anomer 25 (Scheme 6).
The reaction was then tried with the tetrasaccharide
derivative 15 (synthesized from 2,3,4,6-tetra-O-benzoyl-
b-
D
-glucopyranosyl-(1 fi6)-2,3,4-tri-O-benzoyl-a-
D-gluco-
pyranosyl trichloroacetimidate11 and 11 in 68% yield).
Rearrangement occurred at three anomeric centers to
give the isomaltotetraose amide derivative 16 in 20%
yield (Scheme 4).
Based on very recent work of Deslongchamps, endo-
cyclic cleavage is possible under these conditions, but
the involvement of the free cationic species and bond
rotation appear unlikely as the b-anomer tends to
reform in nonpolar solvents.12 A modified mechanism
could be proposed13 involving the bromide ion as a
transient nucleophile/leaving group. First, endocyclic
C–O bond cleavage (23), accompanied or followed by
formation of a cationic benzoate species, ring opening
by the nucleo-philic bromide (26) followed by displace-
ment of the bromide by the free 5 hydroxyl group( 27)
(Scheme 7).
The surprising stereochemical outcome of the acid cat-
alyzed rearrangement of the di-, tri-, and tetrasacchar-
ides 4, 13, and 15, remains unexplained. The reaction
mechanism is unclear for several reasons; if bond rup-
ture occurs to give the oxonium intermediate 18, this
exists in equilibrium with the benzoxonium intermediate
19 (Scheme 5). Expected reaction products would thus
be orthoester 20 and/or the b-glycosylation product 21.
OBz
O
BzO
BzO
OBz
O
BzO
O
O
BzO
BzO
O
HBr (33% in AcOH),
CH2Cl2, 20%
BzO
BzO
BzO
BzO
BzO
O
O
BzO
O
BzO
BzO
O
O
O
BzO
BzO
BzO
BzO
BzO
O
O
BzO
O
CN
O
BzO
O
BzO
BzO
O
OBz
NH2
15
OBz
16
Scheme 4.
OBz
O
ROH
BzO
BzO
OBz
O
OBz
OBz
O
O
BzO
BzO
H+
BzO
BzO
O
20
O
BzO
BzO
OR
BzO
Ph
OR
BzO
O
O
17
18
19
and/or
OBz
O
BzO
BzO
OR
BzO
21
Scheme 5.

4752
M. Chwalek, K. Plꢀe / Tetrahedron Letters 45 (2004) 4749–4753
OBz
O
H
Br
References and notes
OBz
BzO
BzO
O
O
BzO
BzO
BzO
1. (a) Wolfram, M. L.; Pittet, A. O.; Gillam, I. C. Proc. Natl.
Acad. Sci. U.S.A. 1961, 47, 700–705; (b) Wolfram, M. L.;
Lineback, D. R. In Methods in Carbohydrate Chemistry
and Biochemistry; Whistler, R. L., Wolfram, M. L. , Eds.;
Academic: New York, London, 1963; Vol. 2, pp 341–345.
2. A few selected references are: (a) Wolfram, M. L.;
Igarashi, K.; Koizumi, K. J. Org. Chem. 1965, 30, 3841–
3844; (b) Koto, S.; Uchida, T.; Zen, S. Chem. Lett. 1972,
1049–1052; (c) Pfaffli, P. J.; Hixon, S. H.; Anderson, L.
Carbohydr. Res. 1972, 23, 195–206; (d) Takiura, K.;
Kakehi, K.; Honda, S. Chem. Pharm. Bull. 1973, 21,
523–527; (e) Koto, S.; Uchida, T.; Zen, S. Bull. Chem. Soc.
Jpn. 1973, 46, 2520–2523; (f) Berry, J. M.; Dutton, G. S.
Carbohydr. Res. 1974, 38, 339–345.
O
HBr / H+
OBz
BzO
BzO
O
O
O
R
BzO
BzO
O
R
R = CN
OBz
OBz
and/or
22
25
CONH2
OBz
OBz
Br
OR
H
H
O
Br
O
BzO
BzO
BzO
BzO
BzO
OBz
OR
23
24
Scheme 6.
3. Demchenko, A. V. Synlett 2003, 9, 1225–1240.
4. Lemieux, R. U.; Hendriks, K. B.; Stick, R. V.; James, K.
J. Am. Chem. Soc. 1975, 97, 4056–4062.
5. Lindberg, B. Acta Chem. Scand. 1949, 3, 1355–1357.
25
22
ꢀ
6. Ple, K.; Chwalek, M.; Voutquenne-Nazabadioko, L. Eur.
J. Org. Chem. 2004, 1588–1603.
7. Typical procedure for the acid catalyzed rearrangement:
Hydrogen bromide (33% in AcOH; 10 mL) was added
dropwise to a solution of amygdalin heptabenzoate 4
(2.0 g, 1.55 mmol) in CH2Cl2 (20 mL) at room tempera-
ture. The reaction was stirred for 6 h before pouring into
ice water and extracting with CH2Cl2. The combined
organic layers were washed with H2O and NaHCO3 (sat.)
before drying over Na2SO4. Column chromatography of
the crude reaction mixture (30% EtOAc, 5% CH2Cl2, 65%
cyclohexane) gave 1.25 g of the major epimer of 5 along
with 0.25 g of the minor one.
OBz
OBz
OBz
Br
OR'
H
H
O
H
Br
O
O
O
BzO
BzO
BzO
BzO
BzO
BzO
Br
OR'
OBz
BzO
27
O
OR'
23
Ph
26
Scheme 7.
Exocyclic bond cleavage with orthoester formation and
a complex mixture of products may also be a part of the
reaction mechanism and cannot be excluded. In the tri-
and tetrasaccharide reactions, small quantities of mono-
and diglucosyl bromides were detected as nonpolar
impurities. Analysis of the polar reaction impurities for
these reactions was also attempted, but the complexity
of the mixtures prevented characterization of any of
these compounds.
Compound 5 major epimer: Rf ¼ 0:4 (cyclohexane/EtOAc,
1
1:1); Selected NMR data: H NMR (500 MHz, CDCl3): d
3.82 (d, 1H, J ¼ 10:6 Hz, H-6), 3.94 (m, 1H, H-5), 4.06
(dd, 1H, J ¼ 11:6, J ¼ 5:5 Hz, H-6), 4.45 (dd, 1H,
J ¼ 12:1, J ¼ 5:9 Hz, H-60), 4.59 (m, 1H, H-50), 4.69 (d,
1H, J ¼ 7:8 Hz, H-1), 4.70 (dd, 1H, J ¼ 12:0, J ¼ 2:2 Hz,
H-60), 5.29 (m, 2H, H-2, CH), 5.42 (dd, 1H, J ¼ 10:2,
J ¼ 3:7 Hz, H-20), 5.46 (t, 1H, J ¼ 9:8 Hz, H-4), 5.57 (d,
1H, J ¼ 3:7 Hz, H-10), 5.68 (d, 1H, J ¼ 2:7 Hz, NH), 5.75
(t, 1H, J ¼ 9:6 Hz, H-3), 5.77 (t, 1H, J ¼ 9:7 Hz, H-40),
6.37 (t, 1H, J ¼ 9:9 Hz, H-30), 7.05 (d, 1H, J ¼ 2:6 Hz,
NH); 13C NMR (125 MHz, CDCl3): d 62.9 (C-60), 65.6 (C-
6), 68.2 (C-50), 68.7 (C-4), 69.3 (C-40), 70.3 (C-30), 71.9
(C-2), 72.0 (C-20), 72.3 (C-3), 73.6 (C-50), 80.2 (CH), 96.0
(C-10), 98.4 (C-1), 171.8 (CONH2); ESI-MS m=z 1227
(M+Na)þ; Anal. Calcd for C69H58NO19: C, 68.76; H, 4.85;
N, 1.16. Found: C, 68.45; H, 4.79; N, 1.28.
Compound 5 minor epimer: Rf ¼ 0:32 (cyclohexane/
EtOAc, 1:1); Selected NMR data: 1H NMR (500 MHz,
CDCl3): d 3.53 (d, 1H, J ¼ 10:3 Hz, H-6), 3.91 (m, 2H, H-
5, H-6), 4.37 (dd, 1H, J ¼ 12:0, J ¼ 5:7 Hz, H-60), 4.43
(m, 1H, H-50), 4.55 (dd, 1H, J ¼ 12:1, J ¼ 2:6 Hz, H-60),
4.93 (d, 1H, J ¼ 7:9 Hz, H-1), 5.19 (s, 1H, CH), 5.21 (d,
1H, J ¼ 3:7 Hz, H-10), 5.30 (dd, 1H, J ¼ 9:7, J ¼ 7:9 Hz,
H-2), 5.39 (dd, 1H, J ¼ 10:1, J ¼ 3:8 Hz, H-20), 5.44 (d,
1H, J ¼ 2:8 Hz, NH), 5.56 (t, 1H, J ¼ 9:7 Hz, H-4), 5.69
(t, 1H, J ¼ 9:9 Hz, H-40), 5.83 (t, 1H, J ¼ 9:7 Hz, H-3),
6.29 (t, 1H, J ¼ 9:9 Hz, H-30), 6.84 (d, 1H, J ¼ 2:8 Hz,
NH); 13C NMR (125 MHz, CDCl3): d 62.8 (C-60), 65.9 (C-
6), 68.0 (C-50), 68.6 (C-4), 69.4 (C-40), 70.5 (C-30), 71.6
(C-20), 72.3 (C-2), 72.6 (C-3), 73.5 (C-5), 82.0 (CH), 96.2
(C-10), 100.1 (C-1), 172.0 (CONH2); ESI-MS m=z 1227
(M+Na)þ.
In conclusion, isomaltose trichloroacetimidate was
conveniently synthesized from D-amygdalin in five steps
and 60% overall yield. In spite of the fact that isomaltose
is commercially available (although expensive), and can
be produced in large quantities through hydrolysis or
enzymatic processes, this synthesis is easily accessible,
and avoids the preparation of monosaccharide precur-
sors as well as the necessity of controlling the glycosyl-
ation reaction for 1,2-cis bond formation. We have
shown that this reaction is applicable to tri- and tetra-
saccharide precursors, and further investigation is
underway to better determine the mechanism as well as
the scope and limitations of this reaction.
Acknowledgements
We would like to thank the CNRS and the French
Research Ministry for financial support and the Ph.D.
scholarshipfor M.C. We would also like to thank Prof.
A. Haudrechey for useful discussions.
8. Campagna, F.; Carroti, A.; Casini, G. Tetrahedron Lett.
1977, 1813–1816.

M. Chwalek, K. Plꢀe / Tetrahedron Letters 45 (2004) 4749–4753
4753
1
9. Compound 7: Selected NMR data: H NMR (500 MHz,
CDCl3): d 3.72 (dd, 1H, J ¼ 10:9, J ¼ 1:1 Hz, H-6), 4.05
(dd, 1H, J ¼ 11:0, J ¼ 6:6 Hz, H-6), 4.39 (dd, 1H,
J ¼ 12:4, J ¼ 5:6 Hz, H-60), 4.53 (dd, 1H, J ¼ 10:4,
J ¼ 5:5 Hz, H-5), 4.63 (dd, 1H, J ¼ 12:2, J ¼ 2:2 Hz,
H-60), 4.66 (m, 1H, H-50), 5.35 (m, 2H, H-2, H-20), 5.4
(d, 1H, J ¼ 3:7 Hz, H-10), 5.54 (t, 1H, J ¼ 10:1 Hz, H-4),
5.67 (t, 1H, J ¼ 10:0 Hz, H-40), 6.22 (m, 2H, H-3, H-30),
6.69 (d, 1H, J ¼ 3:8 Hz, H-1), 9.0 (s, 1H, NH); 13C
NMR (125 MHz, CDCl3): d 62.7 (C-60), 65.4 (C-6),
67.8 (C-50), 68.6 (C-4), 69.4 (C-40), 70.3 (C-3), 70.5
(C-30), 70.6 (C-2), 71.2 (C-5), 71.7 (C-20), 92.7 (C-1), 95.4
(C-10), 160.0 (CNH); Anal. Calcd for C63H50Cl3NO18:
C, 62.26; H, 4.15; N, 1.15. Found: C, 62.21; H, 4.07; N,
1.07.
10. 1H NMR: (a) Jansson, P. E.; Kenne, L.; Kolare, I.
Carbohydr. Res. 1994, 257, 163–174; (b) 13C NMR:
Forsgren, M.; Jansson, P. E.; Kenne, L. J. Chem. Soc.,
Perkin Trans. 1 1985, 2383–2388.
11. This compound was synthesized from
three steps and 60% overall yield.
D-amygdalin in
12. Deslongchamps, P.; Li, S.; Dory, Y. L. Org. Lett. 2004, 6,
505–508.
13. Following the referee’s suggestion.