A novel synthesis of α-D-Galp-(1→3)-β-D-Galp-1-Ο-(CH2) 3-NH2, its linkage to activated matrices and absorption of anti-αGal xenoantibodies by affinity columns

Article abstract of DOI:10.1016/S0008-6215(00)00010-0

Pig organs transplanted into primates are rapidly rejected because of the interaction between Galα(1→3)Gal epitopes carried by the graft and natural antibodies (anti-αGal antibodies) present in the blood of the recipient. This report describes a simplified synthesis of the xenogeneic disaccharide and its linkage to activated gel matrices. The digalactosides α-D-Galp-(1→3)-α,β-D-Galp-Galp were synthesized by the condensation of the trichloroacetimidoyl 2,3,4,6-tetra-Ο-benzyl-β-D-galactopyranoside donor with the 3,4-unprotected allyl 2,6-di-Ο-benzyl-α- or β-D-galactopyranoside acceptor precursor. Deallylation and hydrogenolysis led to the free digalactoside, whereas hydrogenolysis alone resulted in the 1-Ο-propyl digalactoside. Both products were tested by inhibition ELISA of natural anti-Gala(1→3)Gal antibodies. The α-D-Galp-(1→3)-β-D-Galp-OPr was found to be the best inhibitor. Thus, the allyl group of the partially benzylated α-D-Galp-(1→3)-β-D-Galp-OAll was engineered, via the hydroxy-, the tosyloxy- and the azidopropyl intermediates, into an aminopropyl group amenable to binding to N-hydroxysuccinimide-activated agarose gel matrices in order to obtain specific immunoabsorption columns. Columns made of gel substituted with 5 μmol of disaccharide per milliliter were found efficient for the immunoabsorption of anti-αGal antibodies from human plasma.

Full text of DOI:10.1016/S0008-6215(00)00010-0

www.elsevier.nl/locate/carres
Carbohydrate Research 325 (2000) 265–277
A novel synthesis of
a-
D
-Galp-(1“3)-b- -Galp-1-O-(CH2)3ꢀNH2, its linkage
D
to activated matrices and absorption of anti-aGal
xenoantibodies by affinity columns
Je´roˆme Liaigre a,b, Didier Dubreuil b, Jean-Paul Prade`re b, Jean-Franc¸ois Bouhours a,*
a Unite´ de Recherche sur l’Immunointer6ention dans les allo- et les xe´notransplantations, INSERM Unite´ 437,
Centre Hospitalier Uni6ersitaire de Nantes, 30 boule6ard Jean Monnet, F-44093 Nantes, France
b Laboratoire de Synthe`se Organique, CNRS UMR 6513, Faculte´ des Sciences et des Techniques de Nantes,
BP 92208, 2, rue de la Houssinie`re, F-44322 Nantes, France
Received 1 July 1999; accepted 18 December 1999
Abstract
Pig organs transplanted into primates are rapidly rejected because of the interaction between Gala(1“3)Gal
epitopes carried by the graft and natural antibodies (anti-aGal antibodies) present in the blood of the recipient. This
report describes a simplified synthesis of the xenogeneic disaccharide and its linkage to activated gel matrices. The
digalactosides a-
D
-Galp-(1“3)-a,b-
D
-Galp-OAll were synthesized by the condensation of the trichloroacetimidoyl
-galac-
2,3,4,6-tetra-O-benzyl-b-
D
-galactopyranoside donor with the 3,4-unprotected allyl 2,6-di-O-benzyl-a- or b-
D
topyranoside acceptor precursor. Deallylation and hydrogenolysis led to the free digalactoside, whereas hydrogenoly-
sis alone resulted in the 1-O-propyl digalactoside. Both products were tested by inhibition ELISA of natural
anti-Gala(1“3)Gal antibodies. The a-
D
-Galp-(1“3)-b-
D
-Galp-OPr was found to be the best inhibitor. Thus, the
-Galp-OAll was engineered, via the hydroxy-, the
allyl group of the partially benzylated a-
D
-Galp-(1“3)-b-
D
tosyloxy- and the azidopropyl intermediates, into an aminopropyl group amenable to binding to N-hydroxysuccin-
imide-activated agarose gel matrices in order to obtain specific immunoabsorption columns. Columns made of gel
substituted with 5 mmol of disaccharide per milliliter were found efficient for the immunoabsorption of anti-aGal
antibodies from human plasma. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Affinity column; 3,4-Unprotected b-
plantation
D
-galactopyranoside; (3-Amino)propyl glycoside; Immunoadsorption; Xenotrans-
1. Introduction
obstacle to be overcome is the hyperacute
rejection, which occurs after recognition of
Xenotransplantation in humans, especially
of porcine organs, is presently being investi-
gated as a possible means to alleviate the
shortage of human organ donors. The first
antigens of the vascular endothelium of the
grafted organ by natural antibodies of the
host. Such a situation is identical with what is
observed in human allotransplantation be-
tween non-concordant ABO blood group indi-
viduals [1], suggesting the involvement of
anti-carbohydrate antibodies for the inter-spe-
cies hyperacute rejection. It has been found
* Corresponding author.
E-mail address: jfb@nantes.inserm.fr (J.-F. Bouhours)
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(00)00010-0

266
J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
that the major culprit of hyperacute rejection of
pig organs by primates and man is the
Gala(1“3)Gal epitope [2–4]. It is present in all
mammals except man, apes, and Old World
monkeys, which, in turn, express natural anti-
bodies (anti-aGal antibodies) against it [5,6].
Therefore, it was suggested that ex vivo absorp-
tion of the plasma of a xenograft recipient on
affinity columns carrying the proper synthetic
oligosaccharide would prevent hyperacute re-
jection of the graft [1,2,7]. For this purpose,
affinity columns bearing the xenogenic determi-
nant Gala(1“3)Gal are required.
pared by modifications of previously described
methods in order to simplify the procedures,
improve their reproducibility, diminish the
cost, and keep high yields in the scope of scaling
up the synthesis. They were both synthesized
starting from the allyl
The allyl a- -galactopyranoside 1a was iso-
lated in 35% yield by crystallization from the
mixture resulting from the treatment of
galactose with allylic alcohol in acidic condi-
tions [17]. The allyl b- -galactopyranoside 1b
[18] was synthesized in 80% overall yield from
the readily available tetra-O-acetyl-a- -galac-
D
-galactopyranosides.
D
D
-
D
D
The digalactoside Gala(1“3)Gal was first
topyranoside bromide [19] by treatment with
allyl alcohol in the presence of silver tri-
fluoromethanesulfonate [20], and subsequent
potassium carbonate-catalyzed deacetylation
[21].
synthesized as part of the human blood group
B
determinant, Gala(1“3)[Fuca(1“2)]-
Galb(1“4)GlcNAc, using bromide [8], imidate
[9] or trichloroacetimidate [10] galacto-
side donor precursors. Recently, the stereo-
selective a(1“3) glycosylation was obtained
using new donor galactoside derivatives [11–
13]. Occasionally, the regioselective a-(1“3)
linkage was obtained from 2,3-unprotected ac-
The trichloroacetimidoyl 2,3,4,6-tetra-O-
benzyl-b- -galactopyranoside (2) was synthe-
D
sized in three steps: per-O-benzylation of 1a
[22], palladium(II) chloride-catalyzed ano-
meric deallylation [23,24], and reaction of the
ceptors [10,14,15]. In this work, the a-
Galp-(1“3)-b- -Galp-OPrNH2, which could
easily bind to matrix support, was prepared
-Galp-OAllyl
D
-
resulting protected
D-galactopyranose with
D
trichloroacetonitrile in the presence of potas-
sium carbonate [25]. It was found that the
per-O-benzylation of 1a with benzyl bromide
by a phase-transfer method, using powdered
potassium hydroxide and tetraoctylammonium
as a catalyst [22], was more efficient and easier
to perform than the previously described
method using sodium hydride [18]. The access
from the a-
D
-Galp-(1“3)-b-
D
precursor, which was synthesized from a
trichloroacetimidate-activated galactoside do-
nor and a 3,4-unprotected allyl b- -galactoside
D
acceptor. The predominant reactivity of the
equatorial over the axial hydroxyl was ex-
ploited to favor the (1“3) intergalactoside
linkage.
to the tetra-O-benzyl-
originally obtained from the allyl tetra-O-ben-
zyl- -galactopyranosides via the prop-1-enyl
D
-galactopyranose was
The allyl group was either removed to give
reducing saccharides, hydrogenated to yield
propyl glycosides, or easily functionalized into
a small spacer arm for binding to matrices.
Soluble propyl glycoside anomers were tested
by inhibition ELISA on human plasma. The
D
intermediates [18,26]. The catalytic deallylation
method of Corey and Suggs [23] was completed
at room temperature in a few hours, and gave
a more straightforward access to the tetra-O-
a-
D
-Galp-(1“3)-b- -Galp-OPrꢀNH2 was re-
D
benzyl- -galactopyranose in 85% yield [24]
D
acted with N-hydroxysuccinimide ester-acti-
vated agarose gels. The resulting affinity
columns were evaluated for their capacity to
retain anti-aGal antibodies from human
plasma.
(Scheme 1).
The glycoside acceptors, allyl 2,6-O-benzyl-
a,b- -galactopyranosides 5a [16] and 5b [28],
D
were prepared in three steps from 1a and 1b,
respectively, following a procedure similar to
that described by Nashed et al. [16]. However,
cyclohexylidene derivatives 3 and 4 were pre-
ferred because their hydrophobicity made them
easier to purify than homologous isopropyli-
dene derivatives (Scheme 2).
2. Results and discussion
Synthesis of h-
D
-Galp-(1“3)-h,i- -Galp-
D
OAll.—Precursors 2 [10] and 5 [16] were pre-

J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
267
Scheme 1. (i) 1,1-Dimethylcyclohexane, H2SO4, DMF, a 53%, b 71%; (ii) BnBr, KOH, Oct4NBr, CH2Cl2, 94%; (iii) HCl aq,
MeOH, quant.
Scheme 2. BF3·Et2O, CH2Cl2; 6a, 33%, 7a, 15%; 6b, 25%, 25%, 7b, 20%; (ii) Me3SiOTf, CH2Cl2, 6b, 60%, 7b, B5%.
The glycosylation reaction between 2 and
either 5a or 5b was performed according to
Schmidt and Michel [27]. Boron trifluoride
etherate promoted the glycosylation of 5a
with the trichloroacetimidate 2 to give a mix-
ture of disaccharides 6a and 7a in 33 and 15%
yield, respectively. Starting from 5b, the corre-
sponding glycosides 6b and 7b were produced
in 25 and 20% yields, respectively, under
similar conditions. Trimethylsilyl trifluoro-
methanesulfonate appeared more effective for
the stereocontrol of the glycosylation of com-
pound 5b, providing the a-(1“3) digalac-
toside 6b in 60% yield. In terms of the
regioselectivity of the reaction, the (1“3)
linkage was largely favored, as compounds
with the a- or b-(1“4) linkage were not de-
tected in any of the above reactions, indicating
that they were formed in less than 5% yield.
This approach using 3,4-unprotected galac-
toside acceptor intermediates was also effec-
tive for the synthesis of Gala(1“3)-
Galb(1“3)Gala(1“1)All1.
Synthesis of h-
OPrꢀNH2.—Among
D
-Galp-(1“3)-i-
D
-Galp-
various derivatives
[29,30], allyl glycosides can readily lead to two
types of end product: (i) a reducing saccha-
ride, and (ii) a propyl glycoside. A third possi-
bility was investigated, i.e., engineering of the
allyl into an aminopropyl group that can be
bound to a solid matrix (Scheme 3).
From the allyl glycosides 6, the free
Gala(1“3)Gal 8 was produced in two steps:
palladium-catalyzed deallylation, prior to the
hydrogenolysis of the benzyloxy groups (65%
overall yield). Direct catalytic hydrogenation
gave rise to both propyloxy isomers 9a and
9b, from the corresponding parents 6a and 6b,
in quantitative yields.
As the propyl glycoside 9b was found to be
a better inhibitor of anti-aGal antibodies than
the anomer 9a, the allyl glycoside 6b was
transformed into the aminopropyl glycoside
15 in order to make affinity columns. The allyl
group of the allyl 2,6-di-O-benzyl-3-O-
(2,3,4,6-tetra-O-benzyl-a- -galactopyranosyl)-
D
b- -galactopyranoside (6b) was first converted
D
into a hydroxypropyl group by a hydrobora-
tion–oxidation reaction [31–33], leading to
1 Liaigre et al. unpublished.

268
J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
the alcohols 10 and 11 with 40 and 36% yields,
respectively. The use of a more voluminous
reagent, such as 9-borabicyclo(3.3.1)nonane
(9-BBN) in THF, favored the formation of the
anti-Markovnikov products 10 and 11 in 60
and 5% yields, respectively. The regioselectiv-
ity of the reaction was improved, but a notice-
able part of the products was degraded,
probably because of the long time required for
the reaction to proceed due to the weak reac-
tivity of 9-BBN in solution in THF. A higher
yield might be obtained as the hydrobora-
tion–oxidation of 1b with solid 9-BBN gave
the primary alcohol in a 81% yield.
The isomer 11 was deprotected by hy-
drogenolysis, and the resulting digalactoside
12 was tested as an inhibitor of the binding of
anti-aGal antibodies to their substrate. The
functionalized compound 10 could be
modified in a number of ways in order to
synthesize neoglycoconjugates. It was chosen
here to obtain a primary amine aimed at
forming amide bonds with spacer arms of
available N-hydroxysuccinimide ester-acti-
vated agarose gels. For this purpose, the pri-
mary hydroxyl of 10 was converted into the
tosylate 13, which was displaced by sodium
azide to give the azidopropyl digalactoside 14
(85% overall yield). Alternatively, the conver-
sion of the primary hydroxyl into an azide
function could be carried out using the Mit-
sunobu procedure [34]. Finally, the target
Gala(1“3)Galb(1“1)PrꢀNH2 (15) was gen-
erated by hydrogenolysis of the intermediate
14.
Inhibition of anti-hGal antibodies by synthe-
sized carbohydrates.—The digalactosides were
tested for their capacity to inhibit anti-aGal
antibodies of human plasma (Fig. 1). The best
inhibitor was Gala(1“3)Galb(1“1)(2-OH)Pr
Scheme 3. (i) PdCl2, MeOH; (ii) H2/PdꢀC, EtOH; (iii) 9-BBN, THF, then H2O2, NaOH; (iv) TsCl, Et3N, CH2Cl2; (v) NaN3,
DMF.

J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
269
Fig. 1. Inhibition ELISA of anti-Gal(a1“3)Gal human IgG by various alpha galactoside disaccharides. A human plasma was
diluted 1/16 and incubated for 30 min with the saccharides at the indicated concentrations before ELISA on rabbit erythrocytes
glycolipids. The secondary antibody was a peroxidase-conjugated anti-human g chain.
The aminopropyl digalactoside 15 dissolved in
one vol of dimethyl sulfoxide was added to the
gel. The column was capped and stirred end
over end at room temperature overnight. The
column was washed with water, blocked with
a 1 M ethanolamine solution (pH 8), and
equilibrated with PBS pH 7.4. It was found
that the coupling efficiency was 80% by mea-
suring the amount of sugar exposed to the gel
and the residual concentration after the cova-
lent binding. The gel final vol was 1.3 mL.
The column was stored with NaN3 0.05% in
PBS.
Disaccharide load of affinity columns.—
Four columns were prepared. Column A was
a control column without bound saccharide.
Columns B, C, and D were loaded with 2, 4,
and 15 mmol of disaccharide, respectively.
Then, 20 mL of fresh plasma were applied on
each column and the amount of anti-aGal
antibodies was measured by ELISA. It was
found that the unsubstituted gel did not retain
any antibody, and that the more important
the sugar load, the more efficient was the
column at removing IgG and IgM (Fig. 2). It
was also found that none of the columns was
overloaded; the titer of anti-aGal antibodies
was the same in the first and the last milliliters
(12), followed by Gala(1“3)Galb(1“1)Pr
(9b). The disaccharide Gala(1“3)-Gala-
(1“1)Pr (9a) was much less inhibitory than
its anomer 9b (390 mM versus 50 mM for
OD=0.5), and the free disaccharide
Gala(1“3)Gal 8 had an intermediary in-
hibitory activity. Melibiose, Gala(1“6)Glc,
which was formerly used as inhibitor of anti-
aGal antibodies [5], was by far the worst
inhibitor (700 mM for OD=0.5). The largest
discriminating effect on inhibition of IgG was
seen with saccharide concentrations ranging
from 0.05 to 0.5 mM. These concentrations
were insufficient to inhibit IgM. Using con-
centrations of inhibitors ten times higher than
for IgG, anti-aGal IgM was inhibited by each
saccharide in the same order as for IgG. These
results were taken into account to select the
protected Gala(1“3)Galb(1“1)All (6b) as
precursor of Gala(1“3)Galb(1“1)PrNH2
(15) for binding to NHS-activated agarose
gels.
Affinity column with the (3-amino)propyl 3-
O-(h-
D
-galactopyranosyl)-i- -galactopyran-
D
oside.—An aliquot (2 mL) of Affi-Gel 10
(Bio-Rad) was poured in a small column,
washed with 5 vol of anhyd 2-propanol, and
equilibrated with anhyd dimethyl sulfoxide.

270
J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
coming out of the column. From these experi-
ments, it was concluded that the optimal load
of digalactoside was 5 mmol/mL of gel. A
similar digalactoside load was also found to
be optimal for the capacity of a glass beads-
based matrix at retaining both anti-aGal of
human plasma in vitro [35] and baboon
plasma ex vivo [36].
These absorptions reflected the ability of the
columns to bind the antibodies under a flow
Fig. 4. Capacity of affinity columns. Undiluted plasma (150
mL) was applied to each of the columns (1 mL) of either
Affigel (A) or Sepharose Fast Flow (S) loaded with 5 mmol of
Gala(1“3)Galb(1“1)PrꢀNH2. The plasma was collected by
10 ml fractions that were assayed for anti-Gala(1“3)Gal
antibodies by ELISA. The titer was defined as the reciprocal
of the dilution giving the same absorbance in ELISA than the
original plasma at the dilution 1/100.
of 1 mL/min. During ex vivo immunoabsorp-
tion, the plasma is recirculated several times
through the column. Therefore, plasma (20
mL) was applied five times onto column B, the
least efficient column (Fig. 3). It is obvious
that the column capacity was better used un-
der these conditions, i.e., on one passage the
column absorbed 50% of anti-aGal IgG and
20% of IgM, whereas after five passages the
plasma was depleted from 90% of anti-aGal
IgG and 70% of IgM.
Fig. 2. Absorption of anti-Gala(1“3)Gal antibodies on
columns with various loads of Gala(1“3)Galb(1“1)-
PrꢀNH2. Undiluted plasma (20 mL) was applied to columns
of Affigel (1.3 mL) loaded with 0 (A), 2 (B), 4 (C), and 15
mmol (D) of Gala(1“3)Galb(1“1)PrꢀNH2. The non-re-
tained fractions were assayed for anti-Gala(1“3)Gal by
ELISA using anti IgG or anti-IgM as secondary antibodies,
and compared with the unabsorbed plasma (0% absorption,
100% activity).
Large-size columns.—It was decided to
make a 25 mL affinity column with Affigel-10
(Bio-Rad) loaded with 50 mg (5 mmol/mL) of
15. A 1 mL-aliquot of gel was poured into a
small (5 mm diameter) column, and tested
with 150 mL of plasma. The plasma was
collected after the column by 10 mL fractions,
which were assayed for anti-aGal activity
(Fig. 4). The column appeared saturated with
anti-aGal IgG after 150 mL of plasma had
run through it, i.e., 50–60% of anti-aGal IgM
was retained, without great variation during
the process.
Therefore, another 25 mL column was
made of Sepharose Fast Flow (Pharmacia
Biotech), also loaded with 50 mg of 15. The
performance of the Sepharose column was
tested on a 1 mL minicolumn as for the
Affigel column. It retained antibodies better
Fig. 3. Improvement of the absorption of anti-Gala(1“3)Gal
antibodies by recirculation of the plasma on the affinity
column. Plasma (20 mL) was applied one to five times on
column B (Fig. 2) without intermediary elutions. The anti-
Gala(1“3)Gal antibodies of the non-retained fraction were
assayed by ELISA.

J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
271
than the Affigel column. In the 70–80 mL
fraction, 80% of the anti-aGal IgG and 68% of
the IgM were retained. In the 140–150 mL
fraction, 68% of the anti-aGal IgG and 59% of
the IgM were still retained. It was concluded
that a 25 mL column would be able to deplete
more than 68% of anti-aGal antibodies of 2 L
of plasma, and more than 59% of 4 L of
plasma, after only one passage through the
column.
In the most recent report on anti-aGal im-
munoabsorption [37], the authors found that a
50 mL column of silica particles linked to the
type 6 trisaccharide maximally absorbed anti-
aGal antibodies from 1.25 L of pooled human
plasma. This type of column was able to delay
the rejection of a pig kidney transplanted into
a cynomolgus monkey by more than 48 h
[37]. By comparison, the 25 mL Sepharose
Gal-a(1“3)Galb(1“1)PrꢀNHꢀCOꢀR column
that was built in the present work had a larger
capacity and a smaller volume, both important
properties for an efficient immunoabsorption
operation. Furthermore, its capacity might be
even greater ex vivo than in vitro, as increasing
the number of passages through the column
improves the antibody depletion. Both large
columns, Sepharose- and Affigel-based, are
presently being tested for baboon-specific anti-
aGal ex vivo immunoabsorption.
hydrate, 35 mL of a 5% aq phenol solution,
and 180 mL of concd H2SO4. Each well content
was thoroughly mixed. The reaction was al-
lowed to proceed for 30 min. Plates were read
at 490 nm with a Dynatech MRX microplate
reader. Concentrations were calculated from a
standard curve established with a titrated
galactose solution.
Inhibition ELISA.—For inhibition experi-
ments the plasma, diluted in phosphate-
buffered saline (PBS) pH 7.4, was incubated
for 1 h in the presence of various concentra-
tions of sugars. Residual activities were mea-
sured by ELISA using microtiter plates
Immulon I (Dynatech, Poly Labo, F-67 Stras-
bourg, France) coated with purified neutral
glycosphingolipids from rabbit erythrocyte
membranes [39]. Anti-aGal antibodies bound
to glycolipids were detected with peroxidase-
conjugated anti-IgG or anti-IgM (Dako S.A.,
F-78190 Trappes, France). Enzyme activity
was detected with ABTS (Boehringer Mann-
heim, F-38243 Meylan, France) as substrate.
Absorption experiments.—For testing the
absorption of anti-aGal antibodies on
columns, undiluted plasma was applied to
columns and recovered as indicated. Anti-
aGal antibodies were measured by ELISA on
serial dilution of plasma before and after ab-
sorption. The titer was defined as the dilution
for which half maximal OD of the undiluted
plasma was obtained.
General methods.—1H and 13C NMR spec-
tra were recorded with a Bruker ARX-400
spectrometer. Coupling constants (J) are given
in Hertz (Hz). Multiplicities are recorded as s
(singlet), d (doublet), t (triplet), q (quartet),
and m (multiplet or complex). Mass spectra
were recorded on a low-resolution MS Engine
(Hewlett–Packard, CH-1217 Meyrin, Switzer-
land) with NH3-chemical ionization in a ther-
mospray source. Alternatively, mass spectra
were obtained in the electrospray ionisation
(ESMS) mode with an ESQUIRE-LC ion-trap
mass spectrometer (Bruker Daltonik, Bremen,
Germany) at 0.6 mass unit resolution. Thin-
layer chromatography (TLC) was performed
on Silica Gel 60 plates (E. Merck, D-64293
Darmstadt, Germany). Column chromatogra-
phy was carried out as indicated using SiO2
unless otherwise specified.
3. Experimental
Materials.—Chemical reagents were pur-
chased from Sigma–Aldrich–Fluka (F-38297
St. Quentin Fallavier, France) and Acros Or-
ganics (F-93166 Noisy-le-Grand, France). An-
hydrous solvents were purchased and kept
,
under Ar on 4 A molecular sieves. Solvents
for chromatography were distilled in the
laboratory. N-Hydroxysuccinimide-activated
agarose gels were purchased from BioRad (F-
94200 Ivry-sur-Seine, France) and from Amer-
sham Pharmacia Biotech (F-91940 Les Ulis,
France).
Carbohydrate quantitation.—Digalactoside
concentrations were assayed in microtiter
plates by a modification of the method of
Dubois et al. [38]. To each well were added 35
mL of solution containing 0.5–15 mg of carbo-

272
J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
2
2 H, OCH2Ph), 4.47 and 4.42 (2 d, 2 H, J
Allyl 2,3,4,6-tetra-O-benzyl-h-
anoside.—Allyl a- -galactopyranoside 1a (2.5
D
-galactopyr-
3
11.8 Hz, OCH2Ph), 4.09 (dd, 1 H, J 8.0, 9.6
D
3
Hz, H-2), 3.98 (de, 1 H, J 3.3 Hz, H-4), 3.73
g, 11.3 mmol) was dissolved in benzyl bromide
(8.5 mL, 72.2 mmol) and CH2Cl2 (3 mL).
Powdered KOH (3.8 g, 67.8 mmol) and cata-
lytic amounts of tetraoctylammonium bro-
mide were added. The mixture was vigorously
stirred for 6 days at room temperature (rt).
Stirring was maintained for 3 h after addition
of MeOH (10 mL). The mixture was filtered
through Celite with CH2Cl2 and the filtrate
was evaporated to dryness. Column chro-
matography (1:19 EtOAc–petroleum ether) of
the residue gave syrupy allyl 2,3,4,6-tetra-O-
3
(te, 1 H, J 6.1 Hz, H-5), 3.63 (m, 3 H, H-3,
H-6, H-6%). 13C NMR data (CDCl3): l 161.5
(CNH), 138.5, 138.2, 137.8, 128.4, 128.2,
127.9, 127.8, 127.6, 127.5 (aromatics), 98.7
(C-1), 82.2 (C-3), 78.1 (C-2), 75.2, 74.8 (2
OCH2Ph), 74.4 (C-5), 73.4 (C-4), 73.4, 73.0 (2
OCH2Ph), 68.1 (C-6).
Allyl 3,4-O-cyclohexylidene-h- -galactopyr-
D
anoside (3a).—Compound 1a (2.0 g, 9.1
mmol) was dissolved in DMF (15 mL).
Dimethoxycyclohexane (2.5 mL, 18.2 mmol)
and concd H2SO4 (0.2 mL) were added. The
mixture was stirred for 17 h at rt, then diluted
in CH2Cl2 (100 mL), washed with satd aq
NaHCO3, dried over MgSO4 and concentrated
under reduced pressure. Column chromatog-
raphy (1:1 EtOAc–petroleum ether) of the
residue gave syrupy 3a (1.4 g, 53%). Rf 0.60
benzyl-a- -galactopyranoside (4.4 g, 70%). Rf
D
0.55 (1:4 EtOAc–petroleum ether). CIMS
(NH3): m/z 598 [M+NH4]+, 540, 458.
2,3,4,6-Tetra-O-benzyl-
—Allyl 2,3,4,6-tetra-O-benzyl-a-
D
-galactopyranose.
-galactopyr-
D
anoside (3.0 g, 5.2 mmol) was dissolved in
MeOH (50 mL) and PdCl2 (305 mg, 1.7
mmol) was added. Competitive formation of
methyl glycosides was prevented by the addi-
tion of a few drops of water. The mixture was
stirred for 3.5 h at rt, filtered through Celite
with CH2Cl2 and the filtrate was evaporated
to dryness. Column chromatography (3:17
EtOAc–petroleum ether) of the residue gave
1
(EtOAc). H NMR data (CDCl3): l 5.95 (m, 1
2
3
H, CH2CHCH2), 5.32 (dd, 1 H, J 1.4, J 17.2
2
3
Hz, CH2CHCH2), 5.24 (dd, 1 H, J 1.4 Hz, J
10.4 Hz, CH2CHCH2), 4.96 (d, 1 H, 3J 3.8 Hz,
H-1), 4.30–4.25 (m,
3
H, H-3, H-4,
CH2CHCH2), 4.13–4.07 (m, 2 H, H-5,
3
2
CH2CHCH2), 3.99 (dd, 1 H, J 6.1, J 11.8
Hz, H-6), 3.89–3.80 (m, 2 H, H-2, H-6%),
1.67–1.40 (m, 10 H, cyclohexyl). 13C NMR
data (CDCl3): l 133.6 (CH2CHCH2), 117.8
(CH2CHCH2), 96.9 (C-1), 75.9 (C-3), 73.4 (C-
4), 69.9 (C-2), 68.7 (CH2CHCH2), 68.3 (C-5),
62.6 (C-6), 37.5, 35.1, 24.9, 24.0, 23.6 (cyclo-
hexyl). Calcd for C15H24O6: Mw 300.3; CIMS
(NH3): m/z 318 [M+NH4]+, 301 [M+H]+,
243.
syrupy
2,3,4,6-tetra-O-benzyl- -galactopyr-
D
anose (2.4 g, 86%). Rf 0.10 (1:4 EtOAc–
petroleum ether). CIMS (NH3): m/z 558
[M+NH4]+, 450, 342.
Trichloroacetimidoyl 2,3,4,6-tetra-O-benzyl-
i-D
-galactopyranoside (2).—2,3,4,6-Tetra-O-
benzyl-D-galactopyranose (2.3 g, 4.2 mmol)
was dissolved in CH2Cl2 (20 mL) under Ar.
Potassium carbonate (2.3 g, 17.0 mmol) and
trichloroacetonitrile (2 mL, 20.0 mmol) were
added. The mixture was vigorously stirred for
5 h at rt and filtered through Celite with
CH2Cl2 (30 mL). The filtrate was washed with
satd aq NaHCO3, dried over MgSO4 and con-
centrated under reduced pressure. Column
chromatography (1:4 EtOAc–petroleum ether
containing 1% Et3N) of the residue gave
syrupy 2 (2.3 g, 80%). Rf 0.65 (3:7 EtOAc–
Allyl 3,4-O-cyclohexylidene-i- -galactopyr-
D
anoside (3b).—The same experimental proce-
dure as above but starting from 1b gave
1
syrupy 3b (2.1 g, 71%). Rf 0.65 (EtOAc). H
NMR data (CDCl3): l 5.94 (m, 1 H,
2
3
CH2CHCH2), 5.32 (dd, 1 H, J 1.4, J 17.2
2
3
Hz, CH2CHCH2), 5.22 (dd, 1 H, J 1.4, J
3
10.4 Hz, CH2CHCH2), 4.39 (dd, 1 H, J 5.4,
2J 12.6 Hz, CH2CHCH2), 4.25 (d, 1 H, J 8.2
3
1
petroleum ether). H NMR data (CDCl3): l
Hz, H-1), 4.16–4.08 (m, 3 H, H-3, H-4,
3
2
8.62 (s, 1 H, NH), 7.34–7.25 (m, 20 H, aro-
CH2CHCH2), 4.02 (dd, 1 H, J 8.1, J 12.7
Hz, H-6), 3.86 (m, 2 H, H-5, H-6%), 3.58 (dd, 1
matics), 5.75 (d, 1 H, 3J 8.0 Hz, H-1), 4.94 and
2
3
4.63 (2 d, 2 H, J 11.5 Hz, OCH2Ph), 4.90 and
H, J 7.2, 8.2 Hz, H-2), 1.74–1.35 (m, 10 H,
2
4.81 (2 d, 2 H, J 10.7 Hz, OCH2Ph), 4.73 (s,
cyclohexyl). 13C NMR data (CDCl3): l 133.7

J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
273
H, 2J 12.0 Hz, OCH2Ph), 4.47 (dd, 1 H, 3J 5.1,
(CH2CHCH2), 118.0 (CH2CHCH2), 111.0
(CIV cyclohexyl), 101.3 (C-1), 78.5 (C-3), 73.9
(C-2), 73.5 (C-5), 73.5 (C-4), 70.2
(CH2CHCH2), 62.4 (C-6), 37.8, 35.4, 24.8,
23.9, 23.6 (cyclohexyl). CIMS (NH3): m/z 318
[M+NH4]+, 301 [M+H]+, 243.
3
2J 13.0 Hz, CH2CHCH2), 4.41 (d, 1 H, J 8.1
Hz, H-1), 4.19 (m,
3
H, H-3, H-4,
CH2CHCH2), 3.96 (te, 1 H, H-5), 3.86 (m, 2
3
H, H-6, H-6%), 3.48 (dd, 1 H, J 6.1, 8.1 Hz,
H-2), 1.66–1.39 (m, 10 H, cyclohexyl). 13C
NMR data (CDCl3): l 138.2 (aromatics),
134.0 (CH2CHCH2), 128.3, 128.1, 127.6,
127.5, 127.4 (aromatics), 117.1 (CH2CHCH2),
110.4 (CIV cyclohexyl), 101.8 (C-1), 80.0
(C-2), 78.7 (C-3), 73.6 (OCH2Ph), 73.6
(OCH2Ph), 73.3 (C-4), 72.3 (C-5), 69.7
(CH2CHCH2), 69.6 (C-6), 37.6, 35.5, 25.0,
24.0, 23.7 (cyclohexyl). CIMS (NH3): m/z 498
[M+NH4]+, 440.
Allyl 3,4-O-cyclohexylidene-2,6-di-O-ben-
zyl-h- -galactopyranoside (4a).—Compound
D
3a (850 mg, 2.8 mmol) was dissolved in
CH2Cl2 (5 mL) and benzyl bromide (1.3 mL,
11.3 mmol). Powdered KOH (640 mg, 11.3
mmol) and catalytic amounts of tetraoctylam-
monium bromide were added. The mixture
was vigorously stirred for 3 days at rt. Stirring
was maintained for 2 h after the addition of
MeOH (5 mL). The mixture was filtered
through Celite and evaporated to dryness.
Column chromatography (1:9 EtOAc–
petroleum ether) of the residue gave 4a (1.1 g,
81%). Rf 0.60 (1:4 EtOAc–petroleum ether).
1H NMR data (CDCl3): l 7.36–7.27 (m, 10 H,
aromatics), 5.93 (m, 1 H, CH2CHCH2), 5.32
Allyl 2,6-di-O-benzyl-h- -galactopyranoside
D
(5a).—Compound 4a (1.2 g, 2.5 mmol) was
dissolved in MeOH (32 mL) and 1 M aq HCl
(8 mL). The mixture was stirred for 1 h at
60 °C. After cooling, satd aq NaHCO3 (10
mL) was added and MeOH was evaporated.
The residue was diluted in CH2Cl2 (50 mL),
washed with satd aq NaHCO3, dried over
MgSO4 and the filtrate was evaporated to
dryness to give syrupy 5a (920 mg, 92%). Rf
0.15 (3:7 EtOAc–petroleum ether). CIMS
(NH3): m/z 418 [M+NH4]+, 360.
2
3
(dd, 1 H, J 1.4, J 17.2 Hz, CH2CHCH2),
5.20 (dd,
1
H, 2J 1.4, 3J 10.4 Hz,
3
CH2CHCH2), 4.83 (d, 1 H, J 3.5 Hz, H-1),
2
4.80 and 4.71 (2 d, 2 H, J 12.6 Hz, OCH2Ph),
2
4.64 and 4.54 (2 d, 2 H, J 12.0 Hz, OCH2Ph),
3
4.36 (dd, 1 H, J 5.4, 7.9 Hz, H-3), 4.22–4.15
Allyl 2,6-di-O-benzyl-i- -galactopyranoside
D
(m, 3 H, H-4, H-5, CH2CHCH2), 4.01 (dd, 1
(5b).—The same experimental procedure as
above but starting from 4b gave syrupy 5b
(2.5 g, 98%). Rf 0.20 (3:7 EtOAc–petroleum
ether). CIMS (NH3): m/z 418 [M+NH4]+,
360.
3
2
H, J 6.2, J 13.0 Hz, CH2CHCH2), 3.76 (dd,
3
2
3
1 H, J 5.3, J 10.1 Hz, H-6), 3.71 (dd, 1 H, J
6.9, 2J 10.1 Hz, H-6%), 3.52 (dd, 1 H, 3J 3.5, 7.9
Hz, H-2), 1.65–1.37 (m, 10 H, cyclohexyl). 13C
NMR data (CDCl3): l 133.7 (CH2CHCH2),
128.3, 128.3, 127.7, 127.6, 127.5, 127.4 (aro-
matics), 117.7 (CH2CHCH2), 96.0 (C-1), 76.7
(C-2), 75.5 (C-3), 73.4 (OCH2Ph), 73.4 (C-4),
72.2 (OCH2Ph), 69.6 (C-6), 68.3 (CH2-
CHCH2), 66.8 (C-5), 38.0, 35.5, 25.0, 24.1,
23.7 (cyclohexyl). CIMS (NH3): m/z 498
[M+NH4]+, 243.
Allyl
O-benzyl-h-
2,6-di-O-benzyl-3-O-(2,3,4,6-tetra-
-galactopyranosyl)-h- -galacto-
D
D
pyranoside (6a).—Compounds 2 (910 mg, 1.3
mmol) and 5a (800 mg, 2.0 mmol) were dis-
solved in dry CH2Cl2 (20 mL) and activated 4
,
A molecular sieves (2 g) were added under Ar.
The mixture was stirred for 30 min at
−30 °C. Then boron trifluoride diethyl ether-
ate (50 mL, 0.4 mmol) was added, and stirring
was maintained for 24 h at −30 °C. The
mixture was filtered through Celite, washed
with satd aq NaHCO3, dried over MgSO4 and
the filtrate was concentrated under reduced
pressure. Column chromatography (1:9
EtOAc–petroleum ether) of the residue gave
syrupy 6a (400 mg, 33%). Rf 0.60 (3:7
EtOAc–petroleum ether). 1H NMR data
(CDCl3): l 7.41–7.22 (m, 30 H, aromatics),
Allyl 3,4-O-cyclohexylidene-2,6-di-O-ben-
zyl-i- -galactopyranoside (4b).—The same
D
experimental procedure as above but starting
from 3b gave syrupy 4b (3.1 g, 94%). Rf 0.50
1
(1:4 EtOAc–petroleum ether). H NMR data
(CDCl3): l 7.45 (m, 10 H, aromatics), 5.95 (m,
2
3
1 H, CH2CHCH2), 5.40 (dd, 1 H, J 1.0, J
2
17.2 Hz, CH2CHCH2), 5.25 (dd, 1 H, J 1.0,
3J 10.4 Hz, CH2CHCH2), 4.91 and 4.87 (2 d, 2
2
H, J 11.9 Hz, OCH2Ph), 4.70 and 4.62 (2 d, 2

274
J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
Hz, H-1II), 4.74 (s, 2 H, OCH2Ph), 4.65 and
3
5.94 (m, 1 H, CH2CHCH2), 5.30 (dd, 1 H, J
2
3
4.64 (2 d, 2 H, J 11.8 Hz, OCH2Ph), 4.62 and
17.2 Hz, CH2CHCH2), 5.18 (dd, 1 H, J 10.9
2
3
4.58 (2 d, 2 H, J 12.0 Hz, OCH2Ph), 4.42
Hz, CH2CHCH2), 4.85 (d, 1 H, J 3.9 Hz,
(ddt, 1 H, 4J 1.4, 3J 5.0, 2J 12.9 Hz,
3
H-1I), 4.84 (d, 1 H, J 3.8 Hz, H-1II), 4.88 and
3
CH2CHCH2), 4.37 (d, 1 H, J 6.0 Hz, H-1I),
2
4.54 (2 d, 2 H, J 11.7 Hz, OCH2Ph), 4.78 and
2
2
4.31 and 4.24 (2 d, 2 H, J 11.8 Hz, OCH2Ph),
4.74 (2 d, 2 H, J 11.8 Hz, OCH2Ph), 4.63 and
4
3
4.13 (t, 1 H, H-5II), 4.12 (ddt, 1 H, J 1.4, J
2
4.57 (2 d, 2 H, J 12.0 Hz, OCH2Ph), 4.57 (d,
2
2
6.0, J 12.9 Hz, CH2CHCH2), 4.01 (dd, 1 H,
2 H, J 11.8 Hz, OCH2Ph), 4.37 and 4.33 (2 d,
2
2 H, J 11.7 Hz, OCH2Ph), 4.20 (te, 1 H,
3J 3.7, 9.7 Hz, H-2II), 3.87 (m, 3 H, H-3II,
3
2
H-5II), 4.14 (dd, 1 H, J 5.2, J 12.8 Hz,
CH2CHCH2), 4.12–3.93 (m, 6 H, H-2II, H-3II,
H-4I, H-4II, H-5I, CH2CHCH2), 3.82 (dd, 1 H,
H-4I, H-4II), 3.80 (dd, 1 H, J 5.9, J 10.0 Hz,
3
2
H-6I), 3.74 (dd, 1 H, J 6.2, J 10.0 Hz, H-6%I),
3 2
3
3
3.64 (d, 1 H, J 5.3 Hz, H-3I), 3.55 (t, 1 H, J
6.0 Hz, H-5I), 3.41 (m, 3 H, H-2I, H-6II,
H-6%II). 13C NMR data (CDCl3): l 138.6,
138.5, 138.4, 138.1, 138.0, 137.8 (aromatics),
134.0 (CH2CHCH2), 128.4, 128.3, 128.2,
128.1, 128.0, 127.9, 127.8, 127.7, 127.6, 127.5,
127.4, 127.3 (aromatics), 116.9 (CH2CHCH2),
102.6 (C-1I), 94.8 (C-1II), 78.9 (C-3II), 77.6
(C-3I), 77.4 (C-2I), 75.5 (C-2II), 74.9, 74.7 (2
OCH2Ph), 74.6 (C-4I), 74.2, 73.5 (2 OCH2Ph),
73.0 (C-5I), 72.9, 72.4 (2 OCH2Ph), 69.8
(CH2CHCH2), 69.3 (C-5II), 69.2 (C-6I), 68.3
(C-6II), 64.9 (C-4II). ESMS: m/z 945.4 [M+
Na]+. Anal. Calcd for C57H62O11: C, 74.17; H,
6.77. Found: C, 74.00; H, 6.76.
3
2
3J 3.6, 9.8 Hz, H-3I), 3.74 (dd, 1 H, J 5.7, J
3
2
10.0 Hz, H-6I), 3.68 (dd, 1 H, J 6.9, J 10.0
Hz, H-6%I), 3.53 (m, 3 H, H-2I, H-6II, H-6%II).
13C NMR data (CDCl3): l 138.3, 138.1, 137.9
(aromatics) 134.4 (CH2CHCH2), 128.9, 128.4,
128.3, 128.2, 128.1, 128.0, 127.9, 127.8, 127.5,
127.3 (aromatics), 117.8 (CH2CHCH2), 95.9
(C-1I), 94.9 (C-1II), 79.9 (C-3II), 77.0 (C-3I),
75.7 (C-2I), 75.0 (C-2II), 74.9, 74.8 (2
OCH2Ph), 74.6 (C-4I), 73.5, 73.4 (2 OCH2Ph),
73.1 (C-5I), 72.9, 72.3 (2 OCH2Ph), 69.7
(CH2CHCH2), 69.8 (C-5II), 68.2 (C-6I), 68.1
(C-6II), 65.9 (C-4II). ESMS: m/z 945.4 [M+
Na]+.
Allyl
O-benzyl-h-
2,6-di-O-benzyl-3-O-(2,3,4,6-tetra-
-galactopyranosyl)-i- -galacto-
Propyl 3-O-(h-
D
-galactopyranosyl)-h- -gal-
D
D
D
actopyranoside (9a).—Palladium 10% on acti-
vated charcoal (200 mg) was suspended in
EtOH (30 mL). Compound 6a (140 mg, 0.15
mmol) dissolved in CH2Cl2 (2 mL) was added
and the mixture was stirred at rt for 16 h
under H2 (2 atm). The mixture was filtered on
Celite and the filtrate was concentrated under
reduced pressure. Lyophilization of the
pyranoside (6b).—Compounds 2 (4.7 g, 6.9
mmol) and 5b (2.3 g, 5.7 mmol) were dis-
solved in dry CH2Cl2 (60 mL) and activated 4
,
A molecular sieves (5 g) were added under Ar.
Tetramethylurea (0.8 mL, 6.9 mmol) was
added and the mixture was stirred for 30
min at −10 °C. Then trimethylsilyl trifluo-
romethanesulfonate (0.1 mL, 0.55 mmol) was
added and stirring was maintained for 48 h at
−10 °C and for 24 h at rt. The mixture was
filtered through Celite, washed with satd aq
NaHCO3, dried over MgSO4 and the filtrate
was concentrated under reduced pressure.
Column chromatography (1:4 EtOAc–
petroleum ether) of the residue gave syrupy 6b
(3.2 g, 60%). Rf 0.60 (3:7 EtOAc–petroleum
1
residue gave syrupy 9a (58 mg, 99%). H
3
NMR data (Me2SO-d6): l 4.98 (d, 1 H, J 3.7
3
Hz, H-1I), 4.81 (d, 1 H, J 2.9 Hz, H-1II), 4.06
3
(s, 1 H, H-3I), 4.00 (t, 1 H, J 6.1 Hz, H-3II),
3.85–3.76 (m, 5 H, H-2II, H-4I, H-4II, H-5I,
3
H-5II), 3.69 (dd, 1 H, J 3.7, 10.2 Hz, H-2I),
3.58 (m, 4 H, H-6I, H-6%I, H-6II, H-6%II), 3.51
3
2
(dt, 1 H, J 7.0, J 9.6 Hz, CH2CH2CH3), 3.33
3
2
(dt, 1 H, J 6.2, J 9.6 Hz, CH2CH2CH3), 1.46
1
3
ether). H NMR data (CDCl3): l 7.35–7.22
(m, 2 H, CH2CH2CH3), 0.75 (t, 3 H, J 7.4
(m, 30 H, aromatics), 5.94 (m, 1 H,
Hz, CH2CH2CH3). ESMS: m/z 407.2 [M+
Na]+.
2
3
CH2CHCH2), 5.31 (dd, 1 H, J 1.6, J 17.2
2
3
Propyl
3-O-(h-
D
-galactopyranosyl)-i- -
D
Hz, CH2CHCH2), 5.16 (dd, 1 H, J 1.4, J
10.4 Hz, CH2CHCH2), 4.89 and 4.53 (2 d, 2
galactopyranoside (9b).—The same experi-
2
mental procedure as above but starting from
H, J 11.3 Hz, OCH2Ph), 4.87 and 4.85 (2 d, 2
1
2
3
6b gave 9b (98%). H NMR data (Me2SO-d6):
H, J 10.9 Hz, OCH2Ph), 4.80 (d, 1 H, J 3.7

J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
275
(2 OCH2Ph), 74.6 (C-4I), 74.4, 73.6, 73.1 (3
OCH2Ph), 73.0 (C-5I), 72.5 (OCH2Ph), 69.4
(C-5II), 69.3 (C-6I), 68.3 (C-6II), 67.3
(CH2CH2CH2OH), 65.0 (C-4II), 59.9 (CH2-
3
l 4.81 (d, 1 H, J 3.1 Hz, H-1II), 4.15 (d, 1 H,
3J 7.3 Hz, H-1I), 4.00 (t, 1 H, 3J 5.8 Hz, H-3II),
3.83–3.33 (m, 14 H, H-2I, H-2II, H-3I, H-4I,
H-4II, H-5I, H-5II, H-6I, H-6%I, H-6II, H-6%II,
CH2CH2CH3), 1.44 (m, 2 H, CH2CH2CH3),
CH2CH2OH),
32.4
(CH2CH2CH2OH).
3
ESIMS: m/z 963.4 [M+Na]+.
0.74 (t, 3 H, J 7.4 Hz, CH2CH2CH3). ESIMS:
m/z 407.2 [M+Na]+.
(3-Tosyloxy)propyl
(2,3,4,6-tetra-O-benzyl-h-
2,6-di-O-benzyl-3-O-
-galactopyranosyl)-
(3-Hydroxy)propyl
(2,3,4,6-tetra-O-benzyl-h-
i- -galactopyranoside (10).—Compound 6b
2,6-di-O-benzyl-3-O-
-galactopyranosyl)-
D
D
i- -galactopyranoside (13).—Compound 10
D
D
(2.4 g, 2.5 mmol) was dissolved in CH2Cl2 (60
mL) under Ar. Triethylamine (2 mL, 14.4
mmol) and tosyl chloride (2.1 g, 11.0 mmol)
were added and the mixture was stirred for 4
days at rt. The mixture was washed with satd
aq NH4Cl, dried over MgSO4 and the organic
layer was concentrated under reduced pres-
sure. Column chromatography (1:9 EtOAc–
petroleum ether) of the residue gave syrupy 13
(2.5 g, 94%). Rf 0.40 (3:7 EtOAc–petroleum
(4.1 g, 4.4 mmol) was dissolved in THF (60
mL) under Ar at 0 °C. A 0.5 M solution of
9-BBN in THF (20 mL, 10.0 mmol) was
added and the mixture was stirred for 24 h
during which time the temperature was al-
lowed to reach rt. As TLC control showed
residual presence of precursor, an additional
20 mL of the 0.05 M solution of 9-BBN in
THF was added and stirring was maintained
for 24 h at rt. The mixture was cooled to 0 °C.
Then 60 mL of a cooled aq 2 M NaOH soln
and 15 mL of cooled H2O2 35% were added
dropwise simultaneously. After 30 min of stir-
ring, the mixture was diluted with EtOAc (100
mL), washed with satd aq NH4Cl, dried over
MgSO4, and the organic layer was concen-
trated under reduced pressure. Column chro-
matography (1:4 EtOAc–petroleum ether) of
the residue gave syrupy 10 (2.5 g, 60%). Rf
1
ether). H NMR data (CDCl3): l 7.75 (d, 2 H,
aromatics), 7.34–7.18 (m, 32 H, aromatics),
2
4.89 and 4.53 (2 d, 2 H, J 11.3 Hz, OCH2Ph),
2
4.85 and 4.65 (2 d, 2 H, J 11.7 Hz, OCH2Ph),
3
4.78 (d, 1 H, J 3.7 Hz, H-1II), 4.73 (s, 2 H,
2
OCH2Ph), 4.67 and 4.52 (2 d, 2 H, J 11.1 Hz,
2
OCH2Ph), 4.60 and 4.56 (2 d, 2 H, J 12.1 Hz,
2
OCH2Ph), 4.32 and 4.26 (2 d, 2 H, J 11.8 Hz,
3
OCH2Ph), 4.21 (d, 1 H, J 7.7 Hz, H-1I),
4.17–4.06 (m, 3 H, H-5II, CH2CH2CH2OTs),
4.01 (dd, 1 H, 3J 3.7, 9.9 Hz, H-2II), 3.92 (ddd,
1 H, 3J 5.6, 6.2, 2J 10.0 Hz, CH2CH2CH2OTs),
3.86–3.81 (m, 3 H, H-3II, H-4I, H-4II), 3.76
1
0.15 (3:7 EtOAc–petroleum ether). H NMR
data (CDCl3): l 7.36–7.19 (m, 30 H, aromat-
2
ics), 4.89 and 4.53 (2 d, 2 H, J 11.3 Hz,
2
OCH2Ph), 4.85 and 4.82 (2 d, 2 H, J 10.9 Hz,
3
2
(dd, 1 H, J 5.9, J 9.9 Hz, H-6I), 3.70 (dd, 1
OCH2Ph), 4.79 (d, 1 H, J 3.7 Hz, H-1II), 4.73
3
3
2
3
H, J 6.2, J 9.9 Hz, H-6%I), 3.59 (dd, 1 H, J
3.3, 9.4 Hz, H-3I), 3.56–3.48 (m, 3 H, H-2I,
H-5I, CH2CH2CH2OTs), 3.41 (m, 2 H, H-6II,
H-6%II), 2.36 (s, 3 H, CH3PhS), 1.94 (m, 2 H,
2
(s, 2 H, OCH2Ph), 4.65 and 4.62 (2 d, 2 H, J
2
12.0 Hz, OCH2Ph), 4.64 and 4.56 (2 d, 2 H, J
3
12.0 Hz, OCH2Ph), 4.32 (d, 1 H, J 7.6 Hz,
H-1I), 4.31 and 4.24 (2 d, 2 H, J 11.8 Hz,
2
13
OCH2Ph), 4.11 (te, 1 H, 3J 6.8 Hz, H-5II), 4.02
(m, 2 H, H-2II, CH2CH2CH2OH), 3.88–3.70
(m, 7 H, H-3II, H-4I, H-4II, H-6I, H-6%I,
CH2CH2CH2OH, CH2CH2CH2OH), 3.65–
3.55 (m, 4 H, H-2I, H-3I, H-5I, CH2-
CH2CH2CH2OTs). C NMR data (CDCl3): l
144.5, 138.6, 138.4, 138.3, 138.0, 137.7, 132.9,
129.7, 128.4, 128.4, 128.3, 128.3, 128.2, 128.1,
128.0, 127.9, 127.8, 127.7, 127.6, 127.5, 127.5,
127.3 (aromatics), 103.6 (C-1I), 94.9 (C-1II),
79.1 (C-3II), 77.7 (C-3I), 77.4 (C-2I), 75.6 (C-
2II), 75.0, 74.8 (2 OCH2Ph), 74.6 (C-4I), 74.4,
73.6, 73.1 (3 OCH2Ph), 73.0 (C-5I), 72.5
(OCH2Ph), 69.4 (C-5II), 69.2 (C-6I), 68.4 (C-
6II), 67.7 (CH2CH2CH2OTs), 65.2 (CH2CH2-
CH2OTs), 64.9 (C-4II), 29.5 (CH2CH2CH2-
OTs), 21.5 (CH3PhS). ESMS: m/z 1117.3
[M+Na]+.
3
2
CH2CH2OH), 3.42 (dd, 1 H, J 7.4, J 9.2 Hz,
H-6II), 3.40 (dd, 1 H, J 6.2, J 9.2 Hz, H-6%II),
1.83 (m, 2 H, CH2CH2CH2OH). 13C NMR
data (CDCl3): l 138.6, 138.4, 138.1, 138.0,
137.7, 128.4, 128.4, 128.3, 128.3, 128.2, 128.1,
128.0, 127.8, 127.7, 127.6, 127.5, 127.4 (aro-
matics), 103.7 (C-1I), 94.9 (C-1II), 79.0 (C-3II),
77.8 (C-3I), 77.5 (C-2I), 75.6 (C-2II), 75.0, 74.8
3
2

276
J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
3
NMR data (Me2SO-d6): l 4.82 (d, 1 H, J 3.1
(3-Azido)propyl
(2,3,4,6-tetra-O-benzyl-h-
i- -galactopyranoside (14).—Compound 13
2,6-di- O -benzyl-3- O-
Hz, H-1II), 4.16 (d, 1 H, J 7.6 Hz, H-1I), 3.99
3
D-galactopyranosyl)-
(te, 1 H, 3J 6.4 Hz, H-3II), 3.85–3.30 (m, 13 H,
H-2I, H-2II, H-3I, H-4I, H-4II, H-5I, H-5II,
H-6I, H-6%I, H-6II, H-6%II, CH2CH2CH2NH2),
D
(2.4 g, 2.1 mmol) was dissolved in DMF (60
mL) under Ar. Sodium azide (280 mg, 4.3
mmol) was added and the mixture was stirred
for 1 day at rt. The mixture was diluted with
CH2Cl2 (50 mL), washed with satd aq
NaHCO3, dried over MgSO4 and the organic
layer was concentrated under reduced pres-
sure. Column chromatography (1:4 EtOAc–
petroleum ether) of the residue gave syrupy 14
(1.9 g, 90%). Rf 0.65 (3:7 EtOAc–petroleum
3
2.85 (t, 1 H, J 7.1 Hz, CH2CH2CH2NH2),
2.79 (t, 1 H, CH2CH2CH2NH2), 2.71 (q, 1 H,
CH2CH2CH2NH2), 1.77 (m, 2 H, CH2CH2-
3
CH2NH2, 1.09 (t, 1 H, J 7.1 Hz, CH2CH2-
CH2NH2)).
1
Acknowledgements
ether). H NMR data (CDCl3): l 7.34–7.18
(m, 30 H, aromatics), 4.89 and 4.53 (2 d, 2 H,
2J 11.3 Hz, OCH2Ph), 4.86 and 4.80 (2 d, 2 H,
J.L. was supported by the Fondation
Transvie, the Re´gion de Pays de Loire, and
the Centre National de la Recherche
Scientifique.
3
2J 11.8 Hz, OCH2Ph), 4.79 (d, 1 H, J 3.7 Hz,
H-1II), 4.74 (s, 2 H, OCH2Ph), 4.65 and 4.63
2
(2 d, 2 H, J 11.8 Hz, OCH2Ph), 4.62 and 4.58
2
(2 d, 2 H, J 12.0 Hz, OCH2Ph), 4.31 and 4.24
2
(2 d, 2 H, J 11.8 Hz, OCH2Ph), 4.30 (d, 1 H,
3
3J 7.6 Hz, H-1I), 4.11 (te, 1 H, J 6.8 Hz,
H-5II), 4.00 (m, 2 H, H-2II, CH2CH2CH2N3),
References
3.88 (m, 3 H, H-3II, H-4I, H-4II), 3.78 (dd, 1
[1] D.K.C. Cooper, E. Koren, R. Oriol, Immunol. Re6., 141
(1994) 31–58.
[2] A.H. Good, D.C.K. Cooper, A.J. Malcolm, R.M. Ip-
polito, E. Koren, F.A. Neethling, Y. Ye, N. Zuhdi, L.R.
Lamontagne, Transplant Proc., 24 (1992) 559–562.
[3] M.S. Sandrin, I.F.C. McKenzie, Immunol. Re6., 141
(1994) 169–190.
[4] W.L. Fodor, B.L. Williams, L.A. Matis, J.A. Madri,
S.A. Rollins, J.W. Knight, W. Velander, S.P. Squinto,
Proc. Natl. Acad. Sci. USA, 91 (1994) 11153–11160.
[5] U. Galili, M.R. Clark, S.B. Shohet, J. Buehler, B.A.
Macher, Proc. Natl. Acad. Sci. USA, 84 (1987) 1369–
1373.
[6] U. Galili, S.B. Shohet, E. Kobrin, C.L.M. Stults, B.A.
Macher, J. Biol. Chem., 263 (1988) 17755–17762.
[7] D.K.C. Cooper, Xenotransplantation, 5 (1998) 6–17.
[8] R.U. Lemieux, H. Driguez, J. Am. Chem. Soc., 97 (1985)
4069–4075.
[9] J.C. Jacquinet, P. Sinay, Tetrahedron, 35 (1979) 365–371.
[10] P. Zimmermann, U. Greilich, R.R. Schmidt, Tetrahedron
Lett., 31 (1990) 1849–1852.
[11] A.K. Sarkar, K.L. Matta, Carbohydr. Res., 233 (1992)
245–250.
[12] T. Zhu, G.J. Boons, J. Chem. Soc., Perkin Trans. 1,
(1998) 857–861.
[13] G.V. Reddy, R.K. Jain, B.S. Bhatti, K.L. Matta, Carbo-
hydr. Res., 263 (1994) 67–77.
3
2
3
H, J 5.8, J 9.9 Hz, H-6I), 3.72 (dd, 1 H, J
6.2, 2J 9.9 Hz, H-6%I), 3.67–3.56 (m, 4 H, H-2I,
H-3I, H-5I, CH2CH2CH2N3), 3.41 (m, 2 H,
H-6II, H-6%II), 3.36 (t, 2 H, 3J 6.8 Hz,
CH2CH2CH2N3), 1.86 (m, 2 H, CH2CH2-
CH2N3). 13C NMR data (CDCl3): l 138.6,
138.5, 138.1, 137.8, 128.5, 128.4, 128.4, 128.2,
128.2, 128.0, 128.0, 127.9, 127.7, 127.7, 127.6,
127.6, 127.5, 127.4 (aromatics), 103.6 (C-1I),
94.9 (C-1II), 79.0 (C-3II), 77.8 (C-3I), 77.5 (C-
2I), 75.6 (C-2II), 75.1, 74.8 (2 OCH2Ph), 74.6
(C-4I), 74.4, 73.6, 73.1 (3 OCH2Ph), 73.1 (C-
5I), 72.5 (OCH2Ph), 69.4 (C-5II), 69.3 (C-6I),
68.3 (C-6II), 66.7 (CH2CH2CH2N3), 65.0 (C-
4II), 48.3 (CH2CH2CH2N3), 29.2 (CH2CH2-
CH2N3). ESIMS: m/z 988.4 [M+Na]+.
(3-Amino)propyl
osyl)-i- -galactopyranoside (15).—Palladium
3-O-(h- -galactopyran-
D
D
10% on activated charcoal (300 mg) was sus-
pended in EtOH (30 mL). Compound 14 (300
mg, 0.31 mmol) dissolved in CH2Cl2 (2 mL)
was added and the mixture was stirred at rt
for 16 h under H2 (2 atm). The mixture was
filtered on Celite and the filtrate was concen-
trated. Lyophilization of the residue gave
syrupy 15 (120 mg, 97%) for which MS data
[14] E.Y. Korchagina, N.V. Bovin, So6. J. Bioorg. Chem., 18
(1992) 153–165.
[15] R.T. Lee, Y.C. Lee, Carbohydr. Res., 251 (1994) 69–79.
[16] M.A. Nashed, M.S. Chowdhary, L. Anderson, Carbo-
hydr. Res., 102 (1982) 99–110.
[17] E.A. Talley, M.D. Vale, E. Yanovsky, J. Am. Chem.
Soc., 67 (1945) 2037–2039.
[18] K. James, R.V. Stick, Aust. J. Chem., 29 (1976) 1152–
1159.
1
were not available (note of the editor). H

J. Liaigre et al. / Carbohydrate Research 325 (2000) 265–277
[19] U. Reichman, K.A. Watanabe, J.J. Fox, Carbohydr. [31] C.F. Lane, J. Org. Chem., 39 (1974) 1437–1438.
277
Res., 42 (1975) 233–240.
[20] U. Nilsson, A. Wendler, G. Magnusson, Acta Chem.
Scand., 48 (1994) 356–361.
[21] J.J. Plattner, R.D. Gless, H. Rapoport, J. Am. Chem.
Soc., 94 (1972) 8613–8615.
[22] D. Dubreuil, J. Cleophax, A. Loupy, Carbohydr. Res.,
252 (1994) 149–157.
[23] E.J. Corey, J.W. Suggs, J. Org. Chem., 38 (1973) 3224.
[24] T. Ogawa, H. Yamamoto, Agric. Biol. Chem., 49 (1985)
475–482.
[25] R.R. Schmidt, J. Michel, M. Roos, Justus Liebigs Ann.
Chem., (1984) 1343–1357.
[26] J. Gigg, R. Gigg, J. Chem. Soc. C, (1966) 82–86.
[27] R.R. Schmidt, J. Michel, Angew. Chem., Int. Ed. Engl.,
19 (1980) 731–732.
[28] A.V. Kornilov, L.O. Kononov, G.V. Zatonskii, A.S.
Shashkov, N.E. Nitant’ev, Russ. J. Bioorg. Chem., 8
(1997) 597–607.
[32] H.C. Brown, P.J. Geoghegan, J. Org. Chem., 35 (1970)
1844–1850.
[33] F.G. Bordwel, M.L. Douglas, J. Am. Chem. Soc., 88
(1966) 993–999.
[34] O. Mitsunobu, Synthesis, (1981) 1–28.
[35] R. Rieben, E. von Allmen, E.Y. Korchagina, U.E.
Nydegger, F.A. Neethling, M. Kujundzic, E. Koren,
N.V. Bovin, D.C.K. Cooper, Xenotransplantation, 2
(1995) 98–106.
[36] T. Taniguchi, F.A. Neethling, E.Y. Korchagina, N.
Bovin, Y. Ye, T. Kobayashi, M. Niekrasz, S. Li, E.
Koren, R. Oriol, D.C.K. Cooper, Transplantation, 62
(1996) 1379–1384.
[37] Y. Xu, T. Lorf, T. Sablinski, P. Gianello, M. Bailin, R.
Monroy, T. Kozlowski, M. Awwad, D.K.C. Cooper,
D.H. Sachs, Transplantation, 65 (1998) 172–179.
[38] M. Dubois, K.A. Gilles, J.K. Hamilton, P.A. Rebers, F.
Smith, Anal. Chem., 28 (1956) 350–356.
[39] J.F. Bouhours, C. Richard, N. Ruvoen, N. Barreau, J.
Naulet, D. Bouhours, Glycoconjugate J., 15 (1998) 93–
99.
[29] R.T. Lee, Y.C. Lee, Carbohydr. Res., 37 (1974) 193–201.
[30] P.B. van Seeventer, J.A.L.M. van Dorst, J.F. Siemerink,
J.P. Kamerling, J.F.G. Vliegenthart, Carbohydr. Res.,
300 (1997) 369–373.
.