Synthesis of octopus glycosides: Core molecules for the construction of glycoclusters and carbohydrate-centered dendrimers

Article abstract of DOI:10.1016/S0008-6215(98)00155-4

Allyl α-D-glucopyranoside was perallylated to allyl 2,3,4,6-tetra-O-allyl-α-D-glucopyranoside and this was converted into an array of uniformly functionalized spacer glycosides of an 'octopus' type, taking advantage of the rich chemistry of the allyl ethe

Full text of DOI:10.1016/S0008-6215(98)00155-4

Carbohydrate Research 310 (1998) 35±41
Synthesis of octopus glycosides: core molecules
for the construction of glycoclusters and
carbohydrate-centered dendrimers
Michael Dubber, Thisbe K. Lindhorst *
Department of Organic Chemistry, University of Hamburg, Martin-Luther-King-Platz 6, D-20146
Hamburg, Germany
Received 21 February 1998; accepted 6 June 1998
Abstract
Allyl ꢀ-d-glucopyranoside was perallylated to allyl 2,3,4,6-tetra-O-allyl-ꢀ-d-glucopyranoside
and this was converted into an array of uniformly functionalized spacer glycosides of an ``octo-
pus'' type, taking advantage of the rich chemistry of the allyl ether function. Thus, carbohydrate-
derived pentaaldehydes, pentaalcohols and pentaamines with di€erent spacer lengths were
obtained by ozonolysis, reductive amination, hydroboration or photoaddition of cysteamine
hydrochloride, respectively. The new octopus glycosides are useful core molecules for the synthesis
of glycoclusters and for the construction of carbohydrate-centered dendrimers. # 1998 Elsevier
Science Ltd. All rights reserved
Keywords: Multivalent glycomimetics; Glycoclusters; Multivalency; Spacer glycosides; Allyl ether modi®cations
1. Introduction
which have recently been shown to serve as high
anity ligands in many carbohydrate-protein
interactions due to the ``multivalency e€ect'' [3]
operating in such interactions. Furthermore, mul-
tivalent carbohydrate derivatives are of interest as
initiator cores for the synthesis of carbohydrate-
centered dendrimers in order to obtain chiral den-
drimers (Fig. 1), an area of research only recently
explored [4].
To turn carbohydrates into suitable cores for
glycocluster and dendrimer synthesis, the uniform
derivatization of all hydroxyl groups of the sugar
ring is desirable. At the same time, the hydroxyl
functions should be equipped with spacers of var-
ious chain lengths and di€erent functionalities
attached at the !-position. Allyl ethers o€er a great
Carbohydrates have been extensively used as
chiral pool starting materials for enantioselective
synthesis in numerous reports [1]. Usually the
number of functionalities is reduced during these
modi®cations to remove an ``overload'' of chiral
centers as well as hydroxyl groups. Conversely,
carbohydrates can be of great use as polyfunctional
molecules for the design of multivalent core mole-
cules. These are of interest for the synthesis of
carbohydrate-centered multivalent glycomimetics,
such as glycoclusters and glycodendrimers [2],
* Corresponding author. Fax: 49 40 41234325; e-mail:
tklind@chemie.uni-hamburg.de
0008-6215/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII S0008-6215(98)00155-4

36
M. Dubber, T.K. Lindhorst/Carbohydrate Research 310 (1998) 35±41
Fig. 1. Glycosides, uniformly modi®ed at all hydroxyl groups of the sugar ring have been designed to serve as core molecules for
the construction of glycoclusters or carbohydrate-centered dendrimers.
potential for the generation of diverse spacers with
di€erent functionalities [5] under conditions that
do not harm glycosidic linkages as exempli®ed in
Fig. 2. As such, per-allylated glycosides appear to
be suitable precursors for the synthesis of the
desired polyfunctional carbohydrate cores. The
functionalities resulting from the allyl ether mod-
i®cations can be used as tethers for the attachment
of other glycosides leading to glycoclusters [6] or
for the multiplication of functional groups in a
dendritic manner in order to form carbohydrate-
centered hyperbranched molecules for the synthesis
of chiral hybrid dendrimers.
described. However, the complete conversion of all
allyl functions of a per-allylated glycoside to uni-
formly functionalized octopus-like derivatives has
not yet been investigated. To outline this chem-
istry, the synthesis and characterization of various
octopus glycosides from the per-O-allylated allyl
glucoside 2 is described in this paper.
2. Results and discussion
For the synthesis of a per-O-allylated glucoside,
the anomerically pure allyl ꢀ-d-glucopyranoside 1
was used as inexpensive starting material (Scheme
1). Allylation of 1 was carried out under phase
transfer conditions with allyl chloride as the
organic phase. Under these conditions the reaction
rate increases with the extent of allylation and thus
side products which are only partially allylated are
not obtained [8]. Puri®cation on silica gel is not
necessary for further derivatizations of 2. Thus, the
perallylated glucoside 2 served directly as starting
material for exhaustive ozonolysis, photoaddition
of cysteamine hydrochloride as well as for hydro-
boration.
The chemistry of carbohydrate O-allyl ethers has
been employed in many cases [7] and is well
Initially, ozonolysis of 2 was carried out to form
the pentaaldehyde derivative 3. The reaction yiel-
ded di€erent products, depending on the applied
solvent mixture in accordance with the Criegee
mechanism for the ozonolysis reaction (Fig. 3).
When 2 was treated with ozone in an aprotic sol-
vent such as dichloromethane, a pentacarbonyl
Fig. 2. Versatile transformations of the allyl ether function.

M. Dubber, T.K. Lindhorst/Carbohydrate Research 310 (1998) 35±41
37
product of type A (Fig. 3) was expected. However,
only an insoluble polymeric mixture was obtained.
The same reaction in methanol was expected to
give the corresponding acetal derivative (type B),
however, the product could not be characterized by
NMR. Finally, ozonolysis in bu€ered dichlor-
omethane±methanol [9] a€orded the desired pro-
duct mainly as the hemiacetal (type C). Upon
reaction with ozone at 78 ꢀC and using one
equivalent of sodium hydrogencarbonate per allyl
group to bu€er the reaction mixture, complete
conversion of all allyl groups of 2 was achieved
with no side product formation or decomposition
observed. The yield of this reaction can only be
estimated to be at least 75% as the hemiacetal-type
product also contained traces of the corresponding
1
carbonyl and hydrate derivatives. The H NMR
spectrum of the ozonolysis product 3 in deuterium
oxide displays the expected peaks for the penta-
hydrate and for released methanol.
The ozonolysis product 3 could be reduced to
the pentaalcohol 4. However, it was more feasible to
treat the hydroperoxide, formed during the ozono-
lysis reaction, with sodium borohydride on alumi-
nium oxide to give the penta-(2-hydroxyethyl)
derivative 4 directly from 2 in 85% overall yield.
For further derivatization the ozonolysis pro-
duct 3 was treated with benzylamine under reduc-
tive conditions to e€ect reductive amination to 5.
After extensive optimization studies, treatment of 3
with sodium triacetoxyborohydride [10] as the
reducing agent in tetrahydrofuran and acetic acid,
at low temperature, was selected as an e€ective
procedure. The reductive amination product was
puri®ed by size-exclusion chromatography on
Sephadex LH-20 with 1:1 methanol±dichloromethane
as eluent. Dichloromethane was an important com-
ponent, because it increased the solubility of the
product and eliminated adsorption at the LH-20
gel, which otherwise led to low yields. However,
careful MS analysis of the puri®ed material clearly
revealed ill-de®ned compounds in addition to the
desired product 5, due to intramolecular ring closure
reactions to cyclic amines. This problem was over-
come by application of dibenzylamine instead of
benzylamine under the afore-mentioned optimized
reaction conditions, which led to pure 6 in 74%
yield. NMR and MS analyses proved the uniform
functionalisation to the dibenzylamine cluster 6.
Debenzylation of 6 to the free amine 7 was achieved
by transfer hydrogenation [11] with ammonium
formate and Pd-C in methanol in 89% yield.
Fig. 3. Schematic representation of the Criegee mechanism of
Scheme 1. Reagents and conditions: (a) allyl chloride (5
ꢀ
equiv), aq 40% NaOH, TBABr (1 equiv), 35 C, 16 h; (b) O3,
NaHCO3 (>5 equiv), 6:1 CH2Cl2±MeOH, PPh3; (c), O3,
NaHCO3 (>5 equiv), 5:1 CH2Cl2±MeOH, NaBH4; (d)
BnNH2, NaHB(OAc)3, HOAc, THF, 10 ꢀC; (e) Bn2NH,
NaHB(OAc)3, HOAc, THF, 10 ꢀC; (f) Pd-C, NH4HCO2,
MeOH, Á; (g) 9-BBN, THF, Á; (h) cysteamine hydrochloride
(25 equiv) in MeOH (1 g/mL), 254 nm.

38
M. Dubber, T.K. Lindhorst/Carbohydrate Research 310 (1998) 35±41
Regioselective hydroboration of allyl groups is
being evaluated for the synthesis of glycoclusters
and carbohydrate-centered dendrimers. Further-
more, the presented octopus glycosides and
hyperbranched derivatives thereof imply the pos-
sibility of combinatorial approaches for future
developments.
described in the literature using 9-bora-
bicyclo[3.3.1]nonane or disiamylborane [12]. For
the conversion of the allyl groups of glucoside 2
into 3-hydroxy-propyl spacers, 9-borabicyclo-
[3.3.1]nonane was used in re¯uxing tetrahydro-
furan [13]. This a€orded the pentaalcohol 8 nearly
quantitatively, which could be puri®ed on silica gel.
Even though 8 carries ®ve hydroxyl groups, it is
readily solubilized in chloroform due to the lipo-
philic spacers.
4. Experimental
General methods.ÐOptical rotations were deter-
mined with a Perkin±Elmer 241 polarimeter (10 cm
cells, Na-D-line: 589 nm). NMR spectra were recor-
ded at 400 or 500 MHz on Bruker AMX-400 and
Bruker DRX-500 instruments with Me4Si (ꢁ 0) as
the internal standard. All reactions were monitored
by TLC on silica gel FG254 (Merck) with detection
by UV light, by charring with 10% ethanolic sul-
phuric acid or by a hydroperoxide speci®c reagent
[15]. Flash column chromatography was performed
on silica gel 60 (200±400 mesh, Macherey Nagel &
Co). Gel permeation chromatography was carried
out on Sephadex G-15 (Pharmacia). Photochemical
irradiations were carried out under N2. The degas-
sed solutions were placed in quartz glass tubes and
irradiated at 254 nm in a RPR-100 reactor from
Rayonet. Lyophilisation was performed with a Ley-
bold±Heraeus Lyovac GT 2 apparatus. Elemental
analyses were determined by the Microanalytical
Laboratory of the Department of Organic Chem-
istry at the University of Hamburg.
The functionalization of allyl groups by photo-
addition reaction of cysteamine hydrochloride has
been described in water [14] and is typically per-
formed with a large excess of cysteamine hydro-
chloride. As the pentaallyl derivative 2 is not
water-soluble, another solvent is required for the
photoaddition reaction, which can also dissolve
high concentrations of cysteamine hydrochloride.
Thus, methanol was used to convert the allyl gly-
coside 2 into the amino-terminated core molecule 9
by photoaddition. The large excess of cysteamine
hydrochloride could be removed by gel permeation
chromatography on Sephadex LH-20. Unfortu-
nately, our attempts to apply this procedure to
derivatives, insoluble in methanol or water, were
not successful.
NMR spectra of the described octopus glyco-
sides are characterized by a number of overlapping
peaks due to several magnetically similar OCH2
groups on each molecule. The assignments are
1
facilitated by H-13C HMBC NMR, which fre-
Allyl 2,3,4,6-tetra-O-allyl-a-d-glucopyranoside
(2).ÐTo a solution of 1 (2.08 g, 9.4 mmol) [16] in
aq 33% NaOH (140 mL) the phase transfer cata-
lyst tetrabutylammonium bromide (TBABr, 2.3 g)
was added. Then, allyl chloride (4.6 mL,
quently allows the assignment of all peaks for each
individual spacer. Interestingly, in the 13C NMR
spectra the anomeric OCH2 group is the most high
®eld shifted of all OCH2 groups bound to the sugar
ring. Additionally it is at higher ®eld also when
compared to the C-6 methylene group, in many
cases.
ꢀ
56.5 mmol) was added dropwise at 30 C over 1 h.
The mixture was stirred for 12 h, additional allyl
chloride (1.2 mL) was added and the mixture was
stirred for another 6 h. When the reaction was
complete, toluene (100 mL) was added, the organic
layer was separated and washed ®ve times with
water. The neutral organic phase was dried
(MgSO4), ®ltered, and concentrated to yield 2
3. Conclusions
The methodology presented in this paper allows
the transformation of all hydroxyl groups of a
sugar ring into the corresponding hydroxy ethyl,
hydroxy propyl, amino ethyl and amino-3-thia-
hexyl derivatives with great ease. It is expected that
the approach shown in this paper can be exten-
ded to more complex systems such as disaccharides
and oligosaccharides. The resulting homologous
functionalized octopus glycosides are currently
(2.57 g, 72%) as a colourless oil; [ꢀ]19 +105.3ꢀ (c
d
1.04, CHCl3); 1H NMR (400 MHz, CDCl3): ꢁ 5.85±
6.02 (dddd&m, 5 H, 5 OCH2CHCH2), 5.23±5.34
(ddd, 5 H, 5 OCH2CHCHH), 5.12±5.22 (ddd, 5 H,
5 OCH2CHCHH), 4.93 (d, 1 H, J1,2 3.6 Hz, H-1),
4.37, 4.32, and 4.25 (3 dddd, each 1 H, 3
OCHHCHCH2), 4.20±4.02 (m, 6 H, 6 OCHH
CHCH2), 3.99 (dddd, 1 H, OCHHCHCH2), 3.74

M. Dubber, T.K. Lindhorst/Carbohydrate Research 310 (1998) 35±41
39
(dd&t, 1 H, J3,4 9.7 Hz, H-3), 3.71 (ddd, 1 H, H-5),
3.65 (dd, 1 H, J5,6a 3.6, J6a,6b 10.2 Hz, H-6a), 3.61
(dd, 1 H, J5,6b 2.6 Hz, H-6b), 3.43 (dd&t, 1 H, J4,5
9.7 Hz, H-4), 3.40 (dd, 1 H, J2,3 9.7 Hz, H-2); 13C
NMR (100.67 MHz, CDCl3): ꢁ 135.0, 134.6, 134.5,
134.2, and 133.4 (allyl-C-2), 117.5, 116.9, 116.7,
116.3, and 115.9 (allyl-C-3), 95.3 (C-1), 81.1 (C-3),
79.0 (C-2), 77.0 (C-4), 73.8, 73.4, 72.0, and 71.8
(allyl-C-1), 69.7 (C-5), 68.1 (C-6), 67.6 (allyl-C-1).
Anal. Calcd for C21H32O6: C, 66.28; H, 8.48.
Found: C, 66.09; H, 8.48.
84.0 (C-3), 83.2 (C-2), 80.6 (C-4), 76.9, 76.5, 75.2,
and 74.9 (4 OCH2CH2OH), 73.0 (C-5), 72.0 and
71.8 (C-6, OCH2CH2OH), 63.9, 63.9, 63.6, 63.4,
and 63.3 (5 OCH2CH2OH); 1H-13C HMQC: ꢁ
1
3.735 (H-5), 3.685 (H-3); H-13C HMBC: ꢁ 3.76
(H-6a), 3.68 (H-6b).
(2-N,N-Dibenzyl-amino-ethyl) 2,3,4,6-tetra-O-
(2-N,N-dibenzyl-amino-ethyl)-a-d-glucopyranoside
(6).ÐThe perallylated allyl glucoside 2 (340 mg,
0.894 mmol) was transformed to 3 as described
above and dissolved in THF (20 mL). Dibenzyla-
mine (0.39 mL, 6.8 mmol), HOAc (0.39 mL,
6.8 mmol) and sodium triacetoxyborohydride
(2.3 g, 10.8 mmol) wereꢀadded and the mixture was
stirred for 3 h at 10 C. When the reaction was
complete, CH2Cl2 (100 mL) was added, the organic
layer was washed three times with 1 M aq NaOH
(100 mL) and ®ve times with water. The neutral
organic phase was dried (MgSO4), ®ltered, and
concentrated. After puri®cation on Sephadex LH-
20 (1:1 CH2Cl2±MeOH) 6 was obtained as a col-
(2-Oxo-ethyl) 2,3,4,6-tetra-O-(2-oxo-ethyl)-a-d-
glucopyranoside (3).ÐA solution of 2 (200 mg,
0.53 mmol) and NaHCO3 (340 mg, 4.0 mmol) in 6:1
CH2Cl2±MeOH (70 mL) was treated with ozone at
78 ꢀC until the colour of the solution turned blue.
N2 was bubbled through the solution until the blue
colour had disappeared and then triphenyl-
phoshine (790 mg, 3.0 mmol) was added. When the
reduction of the hydroperoxide was complete, the
solution was puri®ed by ¯ash chromatography
(6:1!5:1 CH2Cl2±MeOH) to yield 3 (210 mg); H
ourless oil (854 mg, 74%); [ꢀ]27 +35.2ꢀ (c 1.175,
1
d
NMR (500 MHz, D2O, MeOH 3.35 ppm): ꢁ 9.64,
9.61, 9.61, 9.60, and 9.56 (5 s, traces of CHO),
5.23±5.10 (m, 6 H, H-1, 5 CH(OH)2), 3.85±3.74,
3.69±3.63, and 3.59±3.49 (each m, 7 H, 4 H, 5 H,
H-2,3,4,5,6a,6b, 10 OCHHCH(OH)2); 13C NMR
(100.62 MHz, D2O, MeOH 50.2 ppm ): ꢁ 97.9 (C-1),
89.8, 89.7, 89.7, 89.7, and 89.6 (5 CH(OH)2), 82.8
(C-3), 81.4 (C-2), 79.0 (C-4), 77.0, 76.6, 75.0, 74.9,
and 72.0 (5 CH2CH(OH)2), 71.0 (C-5), 70.6 (C-6);
1H-13C HMQC: ꢁ 5.13 (H-1), 3.84 (H-5), 3.76 (H-
CDCl3); Mr (FAB): 1297.2 (M+1)+-ion; MALDI-
1
TOF: 1299.7 (M+1)+-ion; H NMR (500 MHz,
CDCl3): ꢁ 7.34±7.13 (m, 50 H, arom H), 4.72 (d, 1
H, J1,2 3.6 Hz, H-1), 3.85 (m, 1 H, OCHHCH2N),
3.75 (m, 1 H, OCHHCH2N), 3.71±3.36 (m, 32 H,
H-3,5,6a,6b, 8 OCHHCH2N, 10 NCH2C6H5), 3.13
(dd, 1 H, H-4), 3.10 (dd, 1 H, J2,3 9.2 Hz, H-2),
2.71±2.46 (m, 10 H, 5 OCH2CH2N); 13C NMR
(125.83 MHz, CDCl3): ꢁ 139.0, 128.1, 128.0, 127.6,
127.5, 126.2, and 126.1 (60 arom C), 96.2 (C-1),
81.3 (C-3), 80.1 (C-2), 77.5 (C-4), 69.7 (C-5), 71.0,
70.9, 69.7, 69.2, and 65.6 (5 OCH2CH2N), 69.0 (C-
6), 58.3, 58.3, 58.1, 58.1, and 57.9 (5 NCH2C6H5),
1
3), 3.55 (H-2), 3.48 (H-4); H-13C HMBC: ꢁ 3.83
(H-6a), 3.80 (H-6b).
(2-Hydroxy-ethyl) 2,3,4,6-tetra-O-(2-hydroxy-
ethyl)-a-d-glucopyranoside (4).ÐA solution of 2
(210 mg, 0.55 mmol) and NaHCO3 (570 mg,
6.8 mmol) in 5:1 CH2Cl2±MeOH (60 mL) was
1
52.6 (3Â), 52.1, and 51.7 (5 OCH2CH2N); H-1H
1
COSY: ꢁ 3.43 (H-3), 3.50 (H-5); H-13C HMBC: ꢁ
3.46 (H-6a), 3.40 (H-6b).
ꢀ
treated with ozone at 78 C until the colour of
(2-Amino-ethyl) 2,3,4,6-tetra-O-(2-amino-ethyl)-
a-d-glucopyranoside (7).ÐTo a solution of 6
(181 mg, 0.14 mmol) in dry MeOH (50 mL), Pd
(10% on charcoal, 240 mg) and ammonium for-
mate (2.5 g, 39.6 mmol) were added under N2 and
the suspension was stirred under re¯ux for 2 h.
Then additional ammonium formate (2.5 g,
39.6 mmol) was added and stirring at re¯ux tem-
perature was continued for 1 h. When the reduc-
tion was complete, the solution was ®ltered
through Celite, and concentrated to yield 7 (49 mg,
the reaction mixture turned blue. After decolour-
isation by a N2 stream, NaBH4 on Al2O3 (1.18 g,
2.95 mmol) was added. After the reduction was
complete, the mixture was puri®ed by ¯ash chro-
matography (5:1!4:1 CH2Cl2±MeOH) to yield 4
(190 mg, 85%) as a colourless oil; [ꢀ]26 +90.2ꢀ (c
d
0.95, MeOH); 1H NMR (400 MHz, MeOH-d4
3.35 ppm): ꢁ 5.03 (d, 1 H, J1,2 3.6 Hz, H-1), 3.98±
3.88 (m, 3 H, 3 OCHHCH2OH), 3.83±3.56 (m, 21
H, 7 OCHHCH2OH, 10 OCH2CHHOH, H-6a,
5,3,6b), 3.42 (dd&t, 1 H, J2,3 9.7 Hz, H-2), 3.41
(ddꢁt, 1 H, J3,4=J4,5 = 9.7 Hz, H-4); 13C NMR
(100.62 MHz, MeOH-d4 50.2 ppm): ꢁ 99.4 (C-1),
89%) as a colourless oil; [ꢀ]28 +76.8ꢀ (c 0.6,
d
MeOH); 1H NMR (500 MHz, MeOH-d4 3.35 ppm):
ꢁ 5.02 (d, 1 H, J1,2 3.6 Hz, H-1), 3.93±3.51 (m, 14

40
M. Dubber, T.K. Lindhorst/Carbohydrate Research 310 (1998) 35±41
H, 5 OCH2CH2NH2, H-3,5,6a,6b), 3.41 (dd, 1 H,
J2,3 9.7 Hz, H-2), 3.40 (dd&t, 1 H, H-4), 2.96±2.79
+45.1ꢀ (c 1.04, H2O); Mr (FAB): 766.5 (M+1)+-
ion without hydrochlorine; H NMR (500 MHz,
1
(m, 10 H,
5
OCH2CH2NH2); 13C NMR
D2O, acetone 2.22 ppm): ꢁ 5.07 (d, 1 H, J1,2 3.6 Hz,
H-1), 3.61±3.91 (m, 13 H, 5 OCH2CH2CH2S, H-
6a,6b,5), 3.57 (dd&t, 1 H, H-3), 3.32±3.42 (m, 2 H,
H-2,4), 3.23 (m&t, 5 H, SCH2CH2NH2HCl), 2.84±
2.90 (m, 5 H, SCH2CH2NH2HCl), 2.65±2.72 (m, 5
H, OCH2CH2CH2S), 1.84±1.98 (m, 5 H, OCH2
CH2CH2S); 13C NMR (100.61 MHz, D2O): ꢁ 96.4
(C-1), 81.5 (C-3), 79.9 (C-2), 78.0 (C-4), 72.5, 70.1,
69.8, 69.1, and 66.8 (5 OCH2CH2CH2S), 72.1 (C-
6), 70.0 (C-5), 38.8 (5 SCH2CH2NH2 HCl), 29.7,
29.5, 29.3, 28.9, and 28.7 (5 OCH2CH2CH2S), 28.6
(3Â) and 28.5 (2Â) (5 SCH2CH2NH2 HCl), 27.8
(2Â), 27.8, and 27.7 (2Â) (5 OCH2CH2CH2S);
1H-1H COSY: ꢁ 3.72 (H-5), 3.40 (H-2), 3.37 (H-4);
1H-13C HMQC: ꢁ 3.91 (H-6a), 3.76 (H-6b).
(125.83 MHz, MeOH-d4 50.2 ppm): ꢁ 99.2 (C-1),
83.9 (C-3), 83.0 (C-2), 80.7 (C-4), 76.1, 76.0, 74.0,
74.7, 71.9, and 71.8 (C-6, 5 OCH2CH2NH2), 73.1
(C-5), 43.9, 43.9, 43.5, 43.2, and 43.2 (5
OCH2CH2NH2); 1H-13C HMQC: ꢁ 3.74 (H-5),
1
3.70 (H-3), 71.9 or 71.8 (C-6); H-13C HMBC: ꢁ
3.77 (H-6a), 3.70 (H-6b).
(3-Hydroxy-propyl) 2,3,4,6-tetra-O-(3-hydroxy-
propyl)-a-d-glucopyranoside (8).ÐTo a solution of
2 (300 mg, 0.79 mmol) in dry THF (8 mL), 9-BBN
(0.5 M solution in THF; 23.5 mL, 11.75 mmol) was
added under N2 atmosphere, and the solution was
stirred at re¯ux temperature for 6 h. Then the
excess of 9-BBN was destroyed by dropwise addi-
ꢀ
tion of water at 0 C. The hydroboration mixture
was oxidized by adding 3 M aq NꢀaOH (11.75 mL)
and 30% H2O2 (11.75 mL) at 0 C, followed by
stirring at room temperature overnight. The aq
phase was saturated with K2CO3 and the THF phase
was separated. The aq phase was extracted twice
with THF (40 mL). The combined organic phases
were dried (MgSO4), ®ltered, and concentrated to
Acknowledgements
We are grateful to the Fonds der Chemischen
Industrie (FCI), the Bundesministerium fur Bil-
dung und Forschung (BMBF), and to Professor
Dr. J. Thiem for their support. We are indebted to
Dr. V. Sinnwell for performing the advanced NMR
experiments.
a€ord 8 (360 mg, 97%) as a colourless oil; [ꢀ]19
d
+87.7ꢀ (c 1.08, CHCl3); Mr (FAB): 493.5
(M+Na)+; 1H NMR (500 MHz, MeOH-d4): ꢁ 4.96
(d, 1 H, J1,2 3.5 Hz, H-1), 3.94±3.82 (m, 4 H, 2
OCH2CH2CH2OH), 3.78±3.55 (m, 19 H, 3 OCH2-
CH2CH2OH, 5 OCH2CH2±CH2OH, H-5,6a,6b),
3.54 (dd&t, 1 H, J3,4 9.5 Hz, H-3), 3.29 (dd, 1 H,
J2,3 9.5 Hz, H-2), 3.26 (dd&t, 1 H, J4,5 9.2 Hz, H-
4), 1.91±1.79 (m, 10 H, 5 OCH2CH2CH2OH); 13C
NMR (125.77 MHz, MeOH-d4): ꢁ 99.1 (C-1), 84.1
(C-3), 83.0 (C-2), 80.6 (C-4), 72.9 (C-5), 72.6, 72.1,
71.9, 70.7, 70.2, and 67.4 (5 OCH2CH2CH2OH, C-
6), 61.6, 61.4, 61.4, 61.4, and 61.2 (5 OCH2CH2-
CH2OH), 35.7, 35.5, 35.2, 34.8, and 34.5 (5
References
[1] S. Hanessian, Total Synthesis of Natural Products:
The Chiron Approach, Pergamon Press, New York,
1983; M. Bols, Carbohydrate Building Blocks, J.
Wiley & Sons, New York, 1996; R.J. Ferrier and S.
Middleton, Chem. Rev., 93 (1993) 2779±2831; B.
Fraser-Reid, Acc. Chem. Res., 29 (1996) 57±66 and
references cited therein.
[2] R. Roy, Curr. Opin. Struct. Biol., 6 (1996) 692±702;
R. Roy, Topics Curr. Chem., 187 (1997) 241±274.
[3] Y.C. Lee and R.T. Lee, Acc. Chem. Res., 28 (1995)
321±327.
[4] H.W. Peerlings and E.W. Meijer, Chem. Eur. J., 3
(1997) 1563±1570; G.R. Newkome, C.N. Moore-
®eld, and F. Vogtle, Dentritic Molecules, Concepts
Syntheses Perspectives, VCH Verlagsgesellschaft,
Weinheim, 1996.
[5] R. Roy and H.J. Jennings, Carbohydr. Res., 112
(1983) 63±72 and references cited therein.
[6] Ch. Kieburg, M. Dubber, and Th.K. Lindhorst,
Synlett, (1997) 1447±1449.
[7] M.A. Bernstein and L.D. Hall, Carbohydr. Res., 78
(1980) C1±C3; R.C. Beier, B.P. Mundy, and G.A.
Strobel, Carbohydr. Res., 93 (1981) 141±143; S.-C.
1
OCH2CH2CH2OH); H-1H COSY: ꢁ 3.66 (H-5).
(Amino-3-thia-hexyl) 2,3,4,6-tetra-O-(amino-3-
thia-hexyl)-a-d-glucopyranoside pentahydrochloride
(9).ÐTo a solution of 2 (230 mg, 0.60 mmol) in
degassed MeOH (1 mL), cysteamine hydrochloride
(1.1 g, 9.7 mmol) was added and the resulting sus-
pension was irradiated at 254 nm for 2 h. Addi-
tional cysteamine hydrochloride (1.1 g, 9.7 mmol)
was added and the suspension was irridated at
254 nm for another 2 h. It was puri®ed by gel ®l-
tration chromatography on Sephadex G-15 using
water as eluent. After lyophilisation 9 was isolated
as an almost colourless glass (550 mg, 97%); [ꢀ]25
d

M. Dubber, T.K. Lindhorst/Carbohydrate Research 310 (1998) 35±41
41
Ats, J. Lehmann, and S. Petry, Carbohydr. Res.,
233 (1992) 141±150.
[12] H.C. Brown, Organic Syntheses via Boranes, J.
Wiley and Sons, New York, 1975.
[8] R. Nouguier and M. Mchich, J. Org. Chem., 50
(1985) 3296±3298.
[13] E.F. Knights, C.G. Scoutern, and H.C. Brown, J.
Am. Chem. Soc., 96 (1974) 7765±7770.
[9] S.L. Schreiber, R.E. Claus, and J. Reagan, Tetra-
hedron Lett., 23 (1982) 3867±3870.
[10] A.F. Abdel-Magid, K.G. Carson, B.D. Harris,
C.A. Maryano€, and R.D. Shah, J. Org. Chem., 61
(1996) 3849±3862.
[14] R.T. Lee and Y.C. Lee, Carbohydr. Res., 37 (1974)
193±201; J. Thiem, P. Stangier, P. Matschulat, and
H. Paulsen, Carbohydr. Res., 225 (1992) 147±150.
[15] W. Huber and E. Frohlke, Chromatographia, 5
(1972) 256±257.
[11] S. Ram and L.D. Spicer, Tetrahedron Lett., 28
(1987) 515±516.
[16] H.P. Wessel, J. Carbohydr. Chem., 7 (1988) 263±
269.