Cobalt-mediated solid phase synthesis of 3-O-alkynylbenzyl galactosides and their evaluation as galectin inhibitors

Article abstract of DOI:10.1016/j.tet.2006.06.057

Methyl β-d-galactoside was converted to the corresponding 3,4-O-stannylene acetal, which was selectively benzylated with 3-iodobenzyl bromide and coupled to a polymer-bound propargylic ether via a Sonogashira reaction. The polymer-bound carbohydrate subst

Full text of DOI:10.1016/j.tet.2006.06.057

Tetrahedron 62 (2006) 8309–8317
Cobalt-mediated solid phase synthesis of 3-O-alkynylbenzyl
galactosides and their evaluation as galectin inhibitors
a,
Anders Bergh,a Hakon Leffler,c Anders Sundin,b Ulf J. Nilssonb, and Nina Kann
*
*
aOrganic Chemistry, Department of Chemical and Biological Engineering, Chalmers University of Technology,
SE-41296 Go€teborg, Sweden
bOrganic Chemistry, Lund University, PO Box 124, SE-22100 Lund, Sweden
cSection MIG, Department of Laboratory Medicine, Lund University, SE-22362 Lund, Sweden
Received 16 March 2006; revised 2 June 2006; accepted 15 June 2006
Available online 7 July 2006
Abstract—Methyl b-D-galactoside was converted to the corresponding 3,4-O-stannylene acetal, which was selectively benzylated with
3-iodobenzyl bromide and coupled to a polymer-bound propargylic ether via a Sonogashira reaction. The polymer-bound carbohydrate substrate
was cleaved from the resin with different carbon nucleophiles in a cobalt-mediated Nicholas reaction. The product 3-O-alkynylbenzyl galac-
tosides were screened towards galectin-1, -3, -7, -8N and -9N in a competitive fluorescence polarisation assay. Particularly potent inhibitors
were identified against galectin-7 with affinity enhancements up to one order of magnitude due to the 3-O-alkynylbenzyl moiety.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
Cells can communicate with the outer world by exchanging
information via their surfaces. Cell surface glycoconjugates,
e.g., glycoproteins and glycolipids, code for information
such as cell identity and are involved in cell signalling path-
ways, which are the reasons why carbohydrate-recognising
Figure 1. LacNAc (N-acetyllactosamine).
proteins, lectins, are potential targets for pharmaceutical
galectins5 show highly conserved CRDs as well as extended
research. The galectins are a sub-class of lectins defined by
binding sites that could accommodate the additional sugar
having an affinity for b-galactosides, a carbohydrate recog-
residues at the galactose C-3. We recently showed that an
nition domain (CRD) of approximately 130 amino acids
affinity-enhancing effect can be achieved by derivatising
and with a conserved amino acid sequence motif of about
LacNAc with 3- or 4-substitued benzamides or a 4-methoxy-
seven residues.1 There are today 14 known galectins that
benzyl ether.6 Hence, synthesis of galactosides carrying 3- or
can be found in mammals,2 and they have a wide variety
4-substituted benzyl ethers at O-3 of galactose emerged as
of biological functions, such as inducing apoptosis of T-
a route towards galectin inhibitors (Fig. 2). Within this con-
cells, antiapoptotic and pro-inflammatory functions as well
text, the use of alkyne-substituted benzyl ethers appeared
as modulation of cell adhesion and migration. The galectins
attractive, as alkyne derivatives can be exploited in various
can be found in almost all types of tissues and organs.3 The
diversifying reactions, among others the Nicholas reaction.
biological importance of galectins discovered over the past
few years makes the discovery of novel and potent galectin
inhibitors more interesting than ever.4 Among natural
saccharide ligands, galectins bind lactose, LacNAc (Fig. 1)
and related disaccharides. Most galectins also bind to longer
saccharides more efficiently. These longer saccharides are
typically characterised by having an additional sugar residue
added to C-3 of Gal in lactose or LacNAc. X-ray studies of
* Corresponding authors. Tel.: +46 46 2228218; fax: +46 46 2228209
(U.J.N.); tel.: +46 31 7723070; fax: +46 31 7723657 (N.K.); e-mail
addresses: ulf.nilsson@organic.lu.se; kann@chalmers.se
Figure 2. Targeted potential galectin inhibitors.
0040–4020/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.06.057

8310
A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
An important tool for investigating biologically active com-
pounds is parallel synthesis on solid phase.7 By using solid
phase techniques, many tedious purification steps can be
avoided thus, accelerating access to new compounds. One
disadvantage, however, is that extra steps are needed for ini-
tial attachment of the substrate to the resin as well cleavage
from the solid support at the end of the sequence. If addi-
tional diversity could be introduced during cleavage, this
reaction step could be seen as an advantage rather than an
impediment, especially if new carbon–carbon bonds could
be formed. Our group has recently developed a method to
this effect using a solid phase variant of the Nicholas reac-
tion8 and we herein describe its use in the preparation of
potential galectin inhibitors, using different carbon and
oxygen nucleophiles to introduce diversity.
3-iodobenzyl bromide as the electrophile to afford 2 in
56% yield (Scheme 1). Propargylic alcohol was attached
to Merrifield resin in a Williamson reaction,11 affording 4,
verified by IR analysis. The iodobenzyl derivative 2 was
then attached to resin 4 under Sonogashira conditions. IR
analysis revealed the disappearance of the terminal alkyne
C–H vibration, while a new broad peak in the region of
3500 cmꢀ1 showed that the carbohydrate had been success-
fully attached, affording 5. Resin 5 could be stored under
dry conditions for several months without any sign of
deterioration.
Complex formation was effected by treating the substrate
with an excess of cobalt octacarbonyl, forming the desired
alkyne–cobalt complex 6, indicated by the deep red colour-
ing of the resin.
2. Results and discussion
In almost all examples of the Nicholas reaction in solution as
well as on solid phase, the solvent of choice has been
dichloromethane.12 However, our initial experiments in di-
chloromethane always resulted in unidentified by-product
formation. Hence, we made a comparison of different
solvents and found that the Nicholas reaction failed in tetra-
hydrofuran and acetonitrile, but gave a clean reaction in
toluene. Toluene also allowed us to run the reaction at
room temperature.
The enticing prospect of performing the synthesis without
protecting the carbohydrate hydroxyl groups could be
achieved by regioselective benzylation at O-3 of methyl
b-D-galactoside 1.9 Functionalisation with meta- or para-
iodobenzyl ether allows the use of the Sonogashira reaction
to introduce an alkyne functionality on the aromatic ring.
This approach would also allow us to connect the carbo-
hydrate onto a solid phase under mild conditions, by using
a polymer-bound propargylic ether in the Sonogashira reac-
tion.10 Furthermore, attaching the galactoside to a solid
phase should minimise the risk of galactose hydroxyl groups
acting as nucleophiles in the Nicholas reaction.
Polymer-bound scaffold 6 was treated with 10 different
nucleophiles (Table 1, entries 1–10) under the modified
Nicholas conditions. The solutions typically turned red
within 2 min indicating initiation of the reaction. After
17 h, the reactions were quenched via the addition of tri-
ethylamine, and the cobalt–alkyne complex was then cleaved
with iodine.13 After workup, the crude products 7–16 were
Methyl b-D-galactoside 1 was converted to the 3,4-O-stan-
nylene acetal, which was selectively benzylated with
Scheme 1. Preparation of meta-substituted 3-O-alkynylbenzyl galactosides 7–16. Reaction conditions: (i) Bu2SnO, MeOH, reflux, 2 h; (ii) 3-iodobenzyl bro-
mide, n-Bu4NBr, 1,4-dioxane, reflux, 2 h; (iii) NaH, propargyl alcohol, 15-crown-5, THF, 16 h; (iv) Pd(PPh3)4, CuI, 4, THF/Et3N (1:1), rt, 16 h; (v) Co2(CO)8,
dichloromethane, rt, 4 h; (vi) BF3$OEt2, nucleophile (see Table 1, entries 1–10), toluene, rt, 16 h or two cycles of 10 min/30 min; (vii) I2, THF, 0 ꢁC, 1.5 h
(yields, see Table 1).

A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
8311
Table 1. Nicholas reaction with polymer-bound substrate 6
Entry Nucleophile
methoxy-substituted benzenes in mind, we opted to run the
whole set with the methodology used for 11 and 12.
Product Overall
yield (%)b
Unfortunately, this methodology failed to afford any product
for 1,3-di- and 1,3,5-trimethoxybenzene (only unidentified
decomposition products were observed). For the other
nucleophiles products were formed, but the yields were sub-
stantially lower. The poor yield may be caused by the para-
alkenyl in combination with the unprotected hydroxyls,
rendering the benzyl ether less stable.
1
N-Methylindole
Benzo[b]thiophene
7
8
9
10
11
12
13
14
25
<5
18
14
48
2
3a
4
3-Trimethylsilyl-1-cyclohexene
1-Trimethylsilyloxy-cyclohexene
1,3,5-Trimethoxybenzene
1,3-Dimethoxybenzene
4-Chloroanisole
5a
6a
7
40
n.r.
n.r.
<5
50
8
Fluorene
9a
10
4-(Trimethylsilyoxy)-3-penten-2-one 15
Allyltrimethylsilane 16
To avoid possible difficulties caused by the unprotected
hydroxyls we decided to acetylate 17, affording 23 in 99%
yield, followed by attachment to the solid phase 3 as before,
yielding 24 (Scheme 2). After complexation of 24 with
dicobalt hexacarbonyl, the nucleophiles that had given the
highest yields with substrate 6 earlier were applied in the
Nicholas reaction. Oxidative removal of the cobalt complex
and deacetylation under Zemplen conditions14 (catalytic
amount of sodium methoxide in methanol) afforded prod-
ucts 19–22 without any by-product formation (Table 2,
entries 1–4).
a
Two short consecutive reactions.
Isolated yield after chromatography for the four-step sequence calculated
from 2.
b
purified either by flash chromatography or by HPLC. 1,3-Di-
methoxybenzene and 1,3,5-trimethoxybenzene initially
gave products, but these decomposed rapidly under the con-
ditions used (entries 5, 6). Products 11 and 12 were formed
within 1 min according to TLC analysis and decomposition
products could be detected after about 15 min. Fortunately,
a shorter reaction time, i.e., 15 min, in combination with
a second reaction cycle of 30 min resulted in good yields
for 1,3,5-tri- (11) and 1,3-di-methoxybenzene (12). Also
included in the second run was 4-(trimethylsilyloxy)-3-
penten-2-one (15) but only traces of nucleophile could be
isolated. The reaction with 1,3-dimethoxybenzene not
surprisingly yielded 12 as a mixture of two regioisomers,
inseparable by HPLC (entry 6). Having completed the syn-
thesis of the meta-substituted 3-O-benzylated galactosides
7–16, our attention turned towards synthesis of the para-
substituted analogues. Starting again from 1, the key
alkyne-complex 18 was synthesised following the same pro-
tocol as for 6 (Scheme 2). With the earlier problems for the
Our initial fears that the hydroxyl groups would compete
with the C-nucleophiles during the Nicholas reaction turned
Table 2. Nicholas reaction with polymer-bound substrate 24
Entry
Nucleophile
Product
Overall yield
(%)a
1
N-Methylindole
19
20
21
22
24
45
51
11
2
Allyltrimethylsilane
1,3-Dimethoxybenzene
1,3,5-Trimethoxybenzene
3a
4
a
Isolated yield after chromatography for the five-step sequence calculated
from 23.
Scheme 2. Preparation of para-substituted 3-O-alkynylbenzyl galactosides 19–22. Reaction conditions: (i) Bu2SnO, MeOH, reflux, 2 h; (ii) 4-iodobenzyl bro-
mide, n-Bu4NBr, 1,4-dioxane, reflux, 2 h; (iii) Pd(PPh3)4, CuI, 4, THF/Et3N (1:1), rt, 16 h; (iv) Co2(CO)8, dichloromethane, rt, 3 h; (v) BF3$OEt2, nucleophile
(see Table 2, entries 1–4), toluene, rt, 2 h; (vi) Ac2O, pyridine, rt, 16 h; (vii) I2, THF, 0 ꢁC, 2 h; (viii) NaOMe, MeOH, rt, 16 h (yields, see Table 2).

8312
A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
conformational freedom. Hence, conclusive analysis of the
conformations of the alkynyl chains proved to be challeng-
ing. Nevertheless, some general conclusions could be drawn.
The two compounds carrying benzyl ethers substituted in the
meta-position with straight-chain substituents, 16 and 25,
preferred similar conformations and similar interactions
with galectin-7. While the alkynylbenzyl moiety of 16 dis-
played a better surface complementarity with galectin-7
(Fig. 4a), the hydroxyl group of the alkyne moiety of 25
could form a hydrogen bond with Asn-2 (Fig. 4b).
Figure 3.
out to be unwarranted. We could never detect any other
carbohydrate-containing products when using 6 as the sub-
strate. The lower yields for the unprotected carbohydrates
are probably caused by boron trifluoride etherate coordinat-
ing to the free hydroxyl groups, as well as material loss in the
isolation steps.
The low-energy complexes with 21, which carries a dimeth-
oxybenzyl-substituted alkyne moiety at the para-position,
revealed preferred interaction modes significantly different
from those of the meta-substituted 16 and 25. The dimeth-
oxybenzyl group of 21 fills a narrow cleft with good surface
complementarity (Fig. 4c). The favoured complex calculated
between 21 and galectin-7 could also explain why the struc-
turally similar 1,3,5-trimethoxy-substituted 22 has a Kd twice
as high as 21. The third methoxy group of 22 would inevita-
bly sterically interfere with the alkyne moiety if a complex
similar to that between 21 and galectin-7 was formed.
The final compounds 2, 7, 9–12, 16 and 19–22 as well as
compound 25 (Fig. 3) were tested against galectin-1, -3,
-7, -8N (N-terminal domain) and -9N (N-terminal domain)
using fluorescence polarisation techniques (Table 3).15,16
While the results for galectin-1, -3 and -8N were less impres-
sive, inhibitors of galectin-7 and -9N with more than an
order magnitude improved affinity were discovered (entries
16, 25, 21 and 2, respectively).
The best inhibitor of galectin-9N was 2, which carries
a structurally simple meta-iodobenzyl ether at the 3-O posi-
tion. Compound 2 is almost 13 times higher in affinity than
methyl b-D-galactoside itself. Disappointingly, additional
substitution on the benzyl group lowers the affinity.
Although 2 might not be a good structure for further devel-
opment into galectin-9N inhibitors, it shows significant
selectivity over the other galectins investigated. This con-
firms that the galectin binding sites close to galactose C-3
differ enough to allow for the much desired development
of selective inhibitors.
Three sub-millimolar inhibitors, compounds 16, 25 and 21,
were found for galectin-7. The Kd value of 0.39 mM for 16
is more than one order of magnitude improvement over the
underivatised reference galactoside 1 and remarkably potent
for a monosaccharide derivative. The best inhibitors are the
simple straight-chain allyl- and hydroxymethyl-substituted
alkynes 16 and 25, which suggests that the binding pocket
of galectin-7 close to galactose O-3 is relatively small and
does not allow larger cyclic structures to bind. Several struc-
tures of galectin-7 as complexes with ligands have been
solved,17 this allowed computational analysis of galectin-7
in complexation with compounds 16, 25 and 21. Although
conformational searches of the complexes gave several
energy minima for each inhibitor/galectin-7 complex, a
coherent picture emerged. The conformations and positions
of the galactose ring and the 3-O-benzyl group were in all
cases similar. The N-terminal of galectin-7, which was close
to the alkynyl groups, possessed a high degree of
3. Conclusions
We have demonstrated that the Nicholas reaction of alkynyl
benzyl ethers can be used to provide a straightforward and
flexible route to novel galectin inhibitors. The solid phase
approach simplified the purification steps and enabled the
use of unprotected carbohydrate in the formation of the
meta-substituted products. The inhibitors were tested for
their affinity towards galectin-1, -3, -7, -8N and -9N using
fluorescence polarisation techniques and the majority ex-
hibited affinity for one or more of the tested galectins. In
particular, potent monosaccharide inhibitors of galectin-7
were discovered.
Table 3. Kd (mM) values against galectin-1, -3, -7, -8N and -9N measured in
a competitive fluorescence polarisation assay
Galectin
1
3
7
8N 9N
1
Methyl b-D-galactoside
10 4.4 4.8 5.3
3.4
4. Experimental
meta-Substituted benzyl ethers, R¼
2
7
9
I
3-(N-Methylindol-3-yl)prop-1-ynyl
(Cyclohexen-3-yl-methyl)prop-1-ynyl
11 2.1 1.3 2.7
0.26
n.i.a n.i. n.i. n.i. n.i.
4.1. General methods
n.i. n.i. 8.4 1.8
5.7
10 (Cyclohexanon-2-yl-methyl)prop-1-ynyl 11 2.4 4.6 1.5
3.0
4.2
All reactions involving moisture or air sensitive compounds
were performed under a nitrogen or argon atmosphere using
oven-dried glass equipment. Solvents were either used as
purchased from commercial sources or purified using stan-
dard procedures18 as appropriate. All reagents were used as
purchased. Column chromatography was carried out using
11 (1,3,5-Trimethoxyphenyl)prop-1-ynyl
12 (1,3-Dimethoxyphenyl)prop-1-ynyl
16 Allylprop-1-ynyl
n.i. 4.5 1.5 1.9
n.i. 4.8 1.2 n.i. 1.5
27 2.4 0.39 1.0
6.9 2.9 0.65 3.8
1.0
1.9
25b Hydroxymethylprop-1-ynyl
para-Substituted benzyl ethers, R¼
19 3-(N-Methylindol-3-yl)prop-1-ynyl
20 Allylprop-1-ynyl
n.i. n.i. n.i. n.i. n.i.
n.i. 4.3 4.9 n.i. 13
˚
Matrix Si 60 A, 35–70 mm. Thin layer chromatography
21 (1,3-Dimethoxyphenyl)prop-1-ynyl
22 (1,3,5-Trimethoxyphenyl)prop-1-ynyl
n.i. 5.4 0.74 2.4 2.0
2.2 1.3 1.5 0.93 1.8
(TLC) was performed using precoated alumina-backed
plates (Merck 25 DC Alufolien Kieselgel 60 F254).
Visualisation was effected either by UV fluorescence
(y¼254 nm) or by heating the plates after treatment with
a
Non-inhibitory.
Prepared via solution phase methods.
b

A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
8313
a stepwise gradient of 0.1 M NH4OAc (5% acetonitrile) to
pure acetonitrile as the mobile phase. Analytical HPLC
was conducted on a Waters 600E system using a Chromasil
100 C8 3.5 mm column with the same mobile phase as the
preparative system. NMR spectra were recorded using either
a Varian Unity 500 or a JEOL eclipse+ (1H 400 MHz and 13
C
100 MHz). All spectra were recorded either in chloroform-
d1 or in methanol-d4. All chemical shifts (dH and dC) are
quoted in parts per million relative to tetramethylsilane (dH
0.00). Optical rotations were measured using a Perkin–
Elmer polarimeter 341 C, which has a thermally jacketed
10ꢀ1 cm cell (path length of 1 dm) and were given in units
of 10ꢀ1 deg cm2 gꢀ1 at 589 nm (sodium D-line). All infra-
red spectra were obtained using a Perkin–Elmer 1600
FTIR as KBr tablet samples. Only selected absorbancies
are reported. Galectin production and fluorescence polarisa-
tion experiments with rat galectin-1 and human galectin-3,
-7, -8N and -9N were performed as described earlier.15,16
4.1.1. Methyl 3-O-(3-iodobenzyl)-b-D-galactopyranoside
(2). Methyl b-D-galactoside (3.00 g, 15.5 mmol) was
dissolved in MeOH (50 mL) and n-Bu2SnO (4.23 g,
17.0 mmol) was added. The mixture was refluxed for 2 h
whereafter the solvent was evaporated under reduced pres-
sure. To the crude residue were added 1,4-dioxane
(50 mL), 3-iodobenzyl bromide (9.17 g, 31.0 mmol) and
n-Bu4NBr (4.98 g, 15.5 mmol). The mixture was sonicated
for 10 min and then refluxed overnight. The solvent was
evaporated under reduced pressure and the sticky residue
was dissolved in MeOH (70 mL) and cooled to 0 ꢁC for
2 h. A white solid was filtered off and the clear solution
was concentrated under reduced pressure. The crude residue
was diluted with CH2Cl2 (100 mL) and cooled to 0 ꢁC, af-
fording methyl [3-O-[3-(3-propargyl)-benzyl]]-b-D-galacto-
pyranoside 2 in the form of white crystals (3.56 g, 56%).
[a]2D5 +22.2 (c 0.021, CD3OD). 1H NMR (400 MHz;
CD3OD) d 7.83 (s, 1H), 7.58 (s, J¼9.0 Hz, 1H), 7.40 (d,
J¼7.1 Hz, 1H), 7.06 (t, J¼7.8 Hz, 1H), 4.69 (d, J¼
12.1 Hz, 1H), 4.56 (d, J¼12.1 Hz, 1H), 4.11 (d, J¼7.8 Hz,
1H), 4.02 (d, J¼3.2 Hz, 1H), 3.71 (m, 2H), 3.62 (dd,
J¼9.6, 7.8 Hz, 1H), 3.49 (s, 3H), 3.43 (t, J¼6.1 Hz, 1H),
3.32 (dd, J¼9.7, 3.3 Hz, 1H); 13C NMR (100 MHz;
CD3OD) d 142.61, 137.93, 137.71, 131.15, 128.18,
105.10, 94.80, 82.74, 76.55, 71.74, 71.48, 67.03, 62.50,
57.28. Anal. Calcd for C14H19IO6: C, 40.99; H, 4.67; I,
30.94. Found: C, 40.86; H, 4.82; I, 30.78.
4.1.2. Polymer-bound propargyl ether (4). Merrifield resin
HL 200–400 mesh (2 g, 2.20 mmol) was mixed with THF
(50 mL), 15-crown-5 (874 mL, 4.40 mmol), KI (73 mg,
0.44 mmol) and NaH (60% in mineral oil, 440 mg,
11 mmol) under N2 for 10 min. Propargyl alcohol (640 mL,
11 mmol) was then added dropwise. The mixture was
shaken overnight whereupon the resin was washed with
water, DMF, MeOH and CH2Cl2 (2ꢂ200 mL each) and
then dried under N2. The resin was analysed using IR and ex-
Figure 4. Energy minimised structure of galectin-7 in complex with (a) 16,
(b) 25 and (c) 21. Molecular modelling was performed with MMFFs force
field in water implemented in Macromodel 9.0. Starting conformations were
built from the galactose/galectin-7 crystal structure.12
hibited the characteristic alkyne C–H stretch at 3294 cmꢀ1
.
4.1.3. Polymer-bound methyl 3-O-[3-(3-propargyl)-
benzyl]-b-D-galactopyranoside (5). To a solid phase reac-
tion vessel were added polymer-bound propargyl ether 4
(2.20 mmol) and THF/Et3N (50 mL, 1:1), whereupon the
mixture was degassed with N2 for 10 min. Pd(PPh3)4
10% H2SO4. Preparative HPLC was conducted on a Waters
600E system using an XTerra Ms C18 5 mm column and
equipped with a Waters 490 Multiwavelength detector, using

8314
A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
(250 mg, 10 mol %) was then added and the mixture was
again degassed for 5 min and then shaken overnight. The
resin was washed with water, DMF, MeOH, CH2Cl2 and
Et2O (2ꢂ100 mL) and dried under N2. The resin was ana-
lysed using IR and the alkyne C–H vibration had been
replaced by a broad peak around 3500 cmꢀ1, indicating
that reaction had taken place.
4.1.5. Methyl [3-O-[3-(3-(1-methyl-indol-3-yl)-prop-1-
ynyl)]-benzyl]-b-D-galactopyranoside (7). White solid.
Yield: 31 mg (25% over four steps, calculated from 2).
[a]2D5 +11.9 (c 0.007, CD3OD). 1H NMR (400 MHz;
CD3OD) d 7.58 (d, J¼7.9 Hz, 1H), 7.50 (s, 1H), 7.36 (d,
J¼7.4 Hz, 1H), 7.31 (m, 2H), 7.26 (t, J¼7.6 Hz, 1H), 7.17
(dd, J¼7.0, 1.1 Hz, 1H), 7.09 (s, 1H), 7.03 (dd, J¼7.0,
0.9 Hz, 1H), 4.70 (d, J¼12.0 Hz, 1H), 4.58 (d, J¼12.0 Hz,
1H), 4.12 (d, J¼7.4 Hz, 1H), 4.01 (dd, J¼2.5, 0.7 Hz, 1H),
3.85 (s, 2H), 3.75–3.72 (m, 4H), 3.61 (dd, J¼7.8, 1.9 Hz,
1H), 4.51–3.49 (m, 4H), 3.42 (dt, J¼5.6, 0.8 Hz, 1H), 3.32
(dd, J¼9.6, 3.3 Hz, 1H); 13C NMR (100 MHz; CD3OD)
d 140.19, 138.82, 132.04, 131.72, 129.32, 128.60, 128.40,
127.92, 125.35, 122.66, 119.81, 119.65, 110.92, 110.28,
105.94, 89.30, 82.52, 81.87, 76.49, 72.01, 71.73, 67.05,
62.50, 57.26, 32.73, 16.55. HRMS (ES) m/z (M+Na) calcd
for C26H29O6NNa m/e: 474.1893. Found: 474.1907.
4.1.4. Representative procedures for meta-substituted
products 7–16. Method A: To a solid phase vessel contain-
ing polymer-bound methyl [3-O-[3-(3-propargyl)]-benzyl]-
b-D-galactopyranoside (5) (2.20 mmol) was added CH2Cl2
(50 mL) and the mixture was degassed with N2 for 10 min.
Co2(CO)8 (2.57 g, 6.6 mmol) was added and the vessel
was shaken at ambient temperature for 4 h whereupon the
resin was washed with CH2Cl2 (6ꢂ50 mL) followed by tolu-
ene (6ꢂ50 mL), affording 6, which was used directly in the
Nicholas reaction. The resin was split into eight parts. To
each part were added toluene (5 mL) and one nucleophile
(3 equiv). The mixture was shaken for 5 min and then
BF3$OEt2 (171 mL, 1.35 mmol) was added in one portion.
The reactions were shaken overnight whereupon Et3N was
added. The resins were washed with THF and MeOH in al-
teration until no colourisation of the wash fluid was seen.
The solvent was then evaporated under reduced pressure
and the crude residues treated with I2 (10 equiv) in a THF/
MeOH mixture (10 mL, 8:2) at ambient temperature until
TLC showed complete conversion (normally around
1.5 h). To the reaction mixtures was then added 20 mL of
a 1:1 solution of satd NaHCO3 and satd Na2SO3. The
mixture was extracted with 2ꢂ30 mL EtOAc, dried over
Na2SO4 and concentrated under reduced pressure. Reversed
phase HPLC afforded products 7, 10 and 16. Remaining
nucleophiles gave only traces of product.
4.1.6. Methyl [3-O-[3-(3-(cyclohex-1-enyl)-prop-1-ynyl)]-
benzyl]-b-D-galactopyranoside (9). White solid. Yield:
15 mg (18% over four steps, calculated from 2). [a]D25
1
+32.8 (c 0.005, CD3OD). H NMR (400 MHz; CD3OD)
d 7.43 (s, 1H), 7.33 (dt, J¼4.0, 1.3 Hz, 1H), 7.23–7.22 (m,
2H), 5.73–5.74 (m, 2H), 4.69 (d, J¼11.9 Hz, 1H), 4.57 (d,
J¼12.0 Hz, 1H), 4.11 (d, J¼7.7 Hz, 1H), 4.00 (d,
J¼2.8 Hz, 1H), 3.75 (m, 2H), 3.62 (dd, J¼7.8, 1.9 Hz,
1H), 3.48 (s, 3H), 3.42 (t, J¼6.55 Hz, 1H), 3.32 (dd,
J¼9.7, 3.3 Hz, 1H), 2.34–2.29 (m, 3H), 1.98–1.94 (m,
2H), 1.92–1.85 (m, 1H), 1.78–1.69 (m, 1H), 1.59–1.49 (m,
1H), 1.42–1.34 (m, 1H); 13C NMR (100 MHz; CD3OD)
d 140.15, 132.01, 131.70, 131.48, 129.30, 129.05, 128.35,
125.40, 105.95, 89.63, 82.53, 82.46, 76.522, 72.05, 71.74,
67.05, 62.49, 57.28, 36.57, 30.00, 26.91, 26.23, 22.38.
HRMS (FAB) m/z (M+Na) calcd for C23H30O6 m/e:
425.1940. Found: 425.1937. Anal. Calcd for C23H30O6: C,
68.64; H, 7.51. Found: C, 68.84; H, 7.18.
Method B: To a solid phase vessel containing polymer-bound
methyl [3-O-[3-(3-propargyl)]-benzyl]-b-D-galactopyrano-
side (5) (1.65 mmol) was added CH2Cl2 (50 mL) and the
mixture was degassed with N2 for 10 min. Co2(CO)8
(1.96 g, 4.95 mmol) was added and the vessel was shaken
at ambient temperature for 4 h whereupon the resin was
washed with CH2Cl2 (6ꢂ50 mL) followed by toluene
(6ꢂ50 mL). The resin was then split into four parts. To
each part were added toluene (4 mL) and one nucleophile
(1.5 equiv). The mixture was shaken for 5 min and then
BF3$OEt2 (131 mL, 1.03 mmol) was added. The mixtures
were shaken for 10 min and then the resins were washed
with toluene (3ꢂ2 mL). The combined wash fluids from
each mixture were quenched with an excess of triethyl-
amine. The resins were then again treated with nucleophile
(1.5 equiv) and BF3$OEt2 (131 mL, 1.03 mmol) in toluene
(2 mL) for 30 min and then washed with toluene and THF.
The solvent was evaporated and the crude residues treated
with I2 (10 equiv) in THF/MeOH (10 mL, 8:2) at ambient
temperature until TLC showed complete conversion. To
the reaction mixtures was then added 20 mL of a 1:1
solution of satd NaHCO3 and satd Na2SO3, and the aqueous
phase was extracted with 2ꢂ30 mL EtOAc, dried over
Na2SO4 and concentrated under reduced pressure. The
residue was subjected to chromatography on silica with
EtOAc as eluent to afford 9, 11 and 12. Only traces of
15 were formed.
4.1.7. Methyl [3-O-[3-(3-(2-cyclohexanonyl)-prop-1-
ynyl)-benzyl]]-b-D-galactopyranoside (10). White solid.
Yield: 16 mg (14% over four steps, calculated from 2).
[a]2D5 +14.5 (c 0.010, CD3OD). 1H NMR (400 MHz;
CD3OD) d 7.45 (s, 1H), 7.37 (dt, J¼4.3, 1.7 Hz, 1H),
7.26–7.24 (m, 2H), 4.72 (d, J¼12.0 Hz, 1H), 4.60 (d,
J¼12.0 Hz, 1H), 4.14 (d, J¼7.7 Hz, 1H), 4.04 (dd, J¼2.4,
0.9 Hz, 1H), 3.78–3.70 (m, 2H), 3.65 (dd, J¼7.8, 1.9 Hz,
1H), 3.52 (s, 3H), 3.46 (dd, J¼5.6, 1.1 Hz, 1H), 3.35 (dd,
J¼5.5, 3.3 Hz, 1H), 2.74 (dd, J¼11.9, 4.9 Hz, 1H), 2.69–
2.61 (m, 1H), 2.48–2.40 (m, 1H), 2.40 (d, J¼6.2 Hz, 1H),
2.36 (d, J¼7.8 Hz, 1H), 2.15–2.08 (m, 1H), 1.96–1.90 (m,
1H), 1.84–1.73 (m, 1H), 1.71–1.60 (m, 1H), 1.54–1.40 (m,
1H); 13C NMR (100 MHz; CD3OD) d 212.05, 138.83,
130.64, 130.31, 127.92, 127.05, 123.87, 104.62, 87.60,
81.29, 81.20, 75.18, 70.67, 70.40, 65.70, 61.13, 55.91,
49.50, 41.38, 33.31, 27.70, 24.72, 19.12. Anal. Calcd for
C23H30O7: C, 66.01; H, 7.23. Found: C, 66.15; H, 7.18.
4.1.8. Methyl [3-O-[3-(3-(2,4,6-trimethoxy-phenyl)-prop-
1-ynyl)]-benzyl]-b-D-galactopyranoside (11). White solid.
Yield: 48 mg (48% over four steps, calculated from 2).
[a]2D5 +22.8 (c 0.015, CD3OD). 1H NMR (400 MHz;
CD3OD) d 7.40 (s, 1H), 7.31 (t, J¼3.6 Hz, 1H), 7.23–7.16
(m, 2H), 6.20 (s, 2H), 4.67 (d, J¼11.9 Hz, 1H), 4.56 (d, J¼
11.9 Hz, 1H), 4.12 (d, J¼7.8 Hz, 1H), 4.00 (d, J¼2.7 Hz,

A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
8315
1
1H), 3.82 (s, 6H), 3.77 (s, 3H), 3.73 (t, J¼6.6 Hz, 2H), 3.59 (t,
J¼9.5 Hz, 1H), 3.56 (s, 2H), 3.48 (s, 3H), 3.35 (s, 2H); 13C
NMR (100 MHz; CD3OD) d 161.84, 159.92, 139.97,
132.05, 131.70, 129.20, 128.02, 125.75, 107.03, 105.90,
91.84, 90.57, 82.44, 78.91, 76.47, 72.06, 71.70, 67.04,
62.48, 57.26, 56.37, 55.81, 13.63. HRMS (FAB) m/z
(M+Na) calcd for C26H32O9 m/e: 511.1944. Found:
511.1942.
(c 0.012, CD3OD). H NMR (400 MHz; CD3OD) d 7.63
(d, J¼8.3 Hz, 2H), 7.20 (d, J¼8.3 Hz, 2H), 4.68 (d, J¼
12.2 Hz, 1H), 4.56 (d, J¼12.2 Hz, 1H), 4.10 (d, J¼7.8 Hz,
1H), 4.00 (d, J¼3.1 Hz, 1H), 3.75–3.66 (m, 2H), 3.62 (dd,
J¼7.8, 1.8 Hz, 1H), 3.48 (s, 3H), 3.42 (dt, J¼6.6, 1.0 Hz,
1H), 3.31 (dd, J¼9.6, 3.3 Hz, 1H); 13C NMR (100 MHz;
CD3OD) d 139.88, 138.50, 131.01, 105.98, 93.53, 82.63,
76.53, 71.75, 67.05, 62.47, 57.28. Anal. Calcd for
C14H19IO6: C, 40.99; H, 4.67. Found: C, 41.12; H, 4.65.
4.1.9. Methyl [3-O-[3-(3-(2,4-dimethoxy-phenyl)-prop-1-
ynyl)]-benzyl]-b-D-galactopyranoside and methyl [3-O-
[3-(3-(2,6-dimethoxy-phenyl)-prop-1-ynyl)]-benzyl]-b-D-
galactopyranoside (12). White solid. Isomeric mixture.
Yield: 39 mg (40% over four steps, calculated from 2).
[a]2D5 +17.2 (c 0.023, CD3OD). 1H NMR (400 MHz;
CD3OD) d 7.51 (br s, 2H), 7.40–7.36 (m, 3H), 7.32–7.16
(m, 3H), 6.62 (d, J¼8.37 Hz) and 6.48 (d, J¼2.4 Hz, 1H),
6.51 (br s, 2H, 0.5H), 4.73 (d, major isomer, J¼11.9 Hz,
2H), 4.61 (d, major isomer, J¼11.9 Hz, 2H), 4.14 (d, major
isomer, J¼7.8 Hz, 1H), 4.03 (d, major isomer, J¼2.8 Hz,
1H), 3.85 (s, 2H), 3.82 (s, 3H), 3.77 (s, 3H), 3.75–3.71 (m,
4H), 3.70–3.66 (m, 2H), 3.65–3.58 (m, 4H), 3.51 (s, major
isomer, 3H), 3.48–3.41 (m, 2H), 3.37–3.32 (m, 1H); 13C
NMR (100 MHz; CD3OD) d 161.50, 159.10, 140.22,
132.09, 131.74, 130.23, 129.34, 129.24, 129.20, 128.45,
128.142, 125.33, 118.44, 105.97, 105.39, 105.10, 99.21,
88.85, 86.00, 82.55, 76.54, 72.04, 71.76, 67.07, 62.50,
61.58, 57.28, 56.45, 55.96, 55.83, 162.68, 20.04. HRMS
(FAB) m/z (M+Na) calcd for C25H30O8 m/e: 481.1838.
Found: 481.1837.
4.1.12. Methyl 2,4,5-O-triacetyl-3-O-(4-iodobenzyl)-b-D-
galactopyranoside (23). Methyl 3-O-(4-iodobenzyl)-b-D-
galactopyranoside (0.85 g, 2.1 mmol) was dissolved in
pyridine (25 mL) and Ac2O (25 mL) was added. The mixture
was refluxed for 2 h whereupon the solvent was evaporated
under reduced pressure and the crude residue was subjected
to chromatography on silica gel using toluene/EtOAc (5:2)
as eluent to afford methyl 2,4,5-O-triacetyl-3-O-(4-iodo-
benzyl)-b-D-galactopyranoside 23 as a white solid (1.10 g,
99%). [a]2D5 +49.1 (c 0.0085, CDCl3). 1H NMR (400 MHz;
CDCl3) d 7.64 (d, J¼7.4 Hz, 2H), 7.00 (d, J¼7.4 Hz, 2H),
5.49 (s, 1H), 5.10 (t, J¼9.0 Hz, 1H), 4.64 (d, J¼12.4 Hz,
1H), 4.33 (d, J¼12.4 Hz, 1H), 4.30 (d, J¼8.0 Hz, 1H),
4.20–4.13 (m, 2H), 3.80 (t, J¼6.4 Hz, 1H), 3.53–3.45 (m,
4H), 2.13 (s, 3H), 2.07 (s, 3H), 2.04 (s, 3H); 13C NMR
(100 MHz; CDCl3) d 170.70, 170.58, 169.58, 137.61,
137.30, 129.65, 102.15, 93.50, 77.23, 70.93, 70.83, 70.45,
65.92, 62.03, 56.98, 21.15, 21.04, 20.96. Anal. Calcd for
C20H25IO9: C, 44.79; H, 4.70. Found: C, 44.89; H, 4.88.
4.1.13. Polymer-bound methyl [2,4,5-O-triacetyl-3-O-(4-
iodobenzyl)]-b-D-galactopyranoside (24). To a solid phase
reaction vessel were added polymer-bound propargylic
ether 4 (1.10 mmol), THF/Et3N (20 mL, 1:1 mixture), CuI
(21 mg, 10 mol %) and methyl O-[2,4,5-triacetyl-3-(4-iodo-
benzyl)]-b-D-galactopyranoside 23 (649 mg, 1.21 mmol).
The mixture was thoroughly degassed by bubbling N2 for
5 min. Pd(PPh3)4 (127 mg, 10 mol %) was added and the
mixture was again degassed for 5 min and then shaken over-
night. The resin was washed with THF, CH2Cl2 and MeOH
repeatedly and thereafter dried under N2. The resin 24 was
analysed using IR and did not exhibit the alkyne C–H stretch
at 3294 cmꢀ1 seen in 23, indicating that Sonogashira
coupling had taken place.
4.1.10. Methyl [3-O-(3-hex-5-en-1-ynyl)-benzyl]-b-D-gal-
actopyranoside (16). White solid. Yield: 50 mg (50% over
four steps, calculated from 2). [a]2D5 +19.5 (c 0.030,
1
CD3OD). H NMR (400 MHz; CD3OD) d 7.46 (s, 1H),
7.36 (t, J¼4.4 Hz, 1H), 7.26 (d, J¼4.0 Hz, 2H), 5.95 (m,
1H), 5.12 (d, J¼17.6 Hz, 1H), 5.05 (d, J¼10.4 Hz, 1H),
4.73 (d, J¼12.0 Hz, 1H), 4.61 (d, J¼12.0 Hz, 1H), 4.14 (d,
J¼8.0 Hz, 1H), 3.74 (d, J¼3.2 Hz, 1H), 3.67 (m, 2H), 3.66
(dd, J¼8.0, 1.8 Hz, 1H), 3.52 (s, 3H), 3.45 (t, J¼4.8 Hz,
1H), 3.45 (dd, J¼8.0, 3.2 Hz, 1H), 2.48 (t, J¼6.4 Hz, 2H),
2.34 (m, 2H); 13C NMR (100 MHz; CD3OD) d 138.94,
137.076, 130.82, 130.42, 128.08, 127.173, 124.12, 114.93,
104.75, 88.99, 81.30, 80.74, 75.33, 70.80, 70.53, 65.80,
61.27, 56.10, 33.00, 18.79. HRMS (ES) m/z (M+Na) calcd
for C20H26O6Na m/e: 385.1627. Found: 385.1611.
4.1.14. Representative procedure for para-substituted
products 19–22. To a solid phase vessel containing polymer-
bound methyl O-[2,4,5-triacetyl-3-(4-iodobenzyl)]-b-D-
galactopyranoside 24 (0.22 mmol) was added CH2Cl2
(10 mL) and the mixture was degassed with N2 for 5 min.
Co2(CO)8 (291 mg, 0.66 mmol) was added and the vessel
was shaken at ambient temperature for 3 h whereafter the
resin was washed with CH2Cl2 (4ꢂ10 mL) followed by tol-
uene (4ꢂ10 mL), forming polymer-bound complex 25. To
the resin were then added toluene (10 mL) and the nucleo-
phile (0.66 mmol). The mixture was degassed for 5 min
and then BF3$OEt2 (42 mL, 0.66 mmol) was added. The
mixture was shaken for 2 h and then quenched with Et3N.
The resin was washed with CH2Cl2 and THF. The solvent
was then evaporated under reduced pressure and the crude
residue treated with I2 (558 mg) in THF (20 mL) at ambient
temperature until TLC showed complete conversion (nor-
mally around 2 h). To the reaction mixtures was then added
4.1.11. Methyl 3-O-(4-iodobenzyl)-b-D-galactopyrano-
side (17). Methyl O-b-D-galactoside (3.00 g, 15.5 mmol)
was dissolved in MeOH (50 mL) and n-Bu2SnO (4.23 g,
17.0 mmol) was added. The mixture was refluxed for 2 h
whereafter the solvent was evaporated. To the crude residue
were added 1,4-dioxane (50 mL), 4-iodobenzyl bromide
(9.17 g, 31.9 mmol) and n-Bu4NBr (4.98 g, 15.5 mmol).
The mixture was sonicated for 10 min and then refluxed
overnight. The solvent was evaporated under reduced pres-
sure and the sticky residue was dissolved in MeOH
(50 mL) and cooled to 0 ꢁC for 2 h. Awhite solid was formed
and filtered off, and the clear solution was then concentrated
under reduced pressure. The crude residue was diluted with
CH2Cl2 (100 mL) and cooled to 4 ꢁC overnight, affording,
after filtration, methyl 3-O-(4-iodo)-benzyl-b-D-galacto-
pyranoside 17 as white crystals (4.06 g, 64%). [a]2D5 +18.2

8316
A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
20 mL of a 1:1 solution of satd NaHCO3 and satd Na2SO3.
Extraction with 2ꢂ30 mL EtOAc, drying over Na2SO4 and
evaporation of the solvent under reduced pressure was
followed by flash chromatography (SiO2, CH2Cl2/MeOH
20:1). Removal of the acetyl moieties was effected by dis-
solving the product in MeOH (1 mL) and adding 1 mL of
a solution of NaOMe in MeOH (20 mg NaOMe/10 mL
MeOH). The mixture was stirred overnight at room temper-
ature and neutralised with Amberlyst 120 H+ when TLC
indicated complete conversion. The solvent was removed
in vacuo and the residue was subjected to flash chromato-
graphy to afford products 19–22.
118.48, 106.00, 105.99, 105.98, 105.39, 105.12, 99.22,
90.01, 88.80, 83.04, 82.56, 82.51, 76.54, 76.53, 72.16,
72.13, 71.79, 71.77, 67.10, 62.49, 62.48, 57.28, 57.26,
56.44, 55.95, 55.82, 20.04, 13.91. HRMS (FAB) m/z
(M+Na) calcd for C26H34O8 m/e: 481.1838. Found: 481.1841.
4.1.18. Methyl 3-O-[4-(2,4,6-trimethoxyphenyl-prop-1-
ynyl)]-benzyl-b-D-galactopyranoside (22). White solid.
Yield: 12.0 mg (11% over five steps, calculated from 23).
[a]2D5 +19.5 (c 0.033, CD3OD). 1H NMR (400 MHz;
CD3OD) d 7.30 (d, J¼8.2 Hz, 2H), 7.23 (J¼8.2 Hz, 2H),
6.18 (s, 2H), 4.68 (d, J¼12.1 Hz, 1H), 4.58 (d, J¼12.1 Hz,
1H), 4.09 (d, J¼7.8 Hz, 1H), 3.98 (d, J¼3.2 Hz, 1H), 3.80
(s, 6H), 3.74–3.65 (m, 2H), 3.60 (dd, J¼7.8, 1.8 Hz, 1H),
3.56 (s, 2H), 3.42 (dt, J¼5.6, 0.9 Hz, 1H), 3.29 (dd, J¼6.4,
3.3 Hz, 1H); 13C NMR (100 MHz; CD3OD) 159.98,
132.40, 128.88, 125.10, 107.17, 105.99, 91.90, 90.50,
82.50, 76.54, 72.18, 71.77, 67.10, 62.48, 57.27, 56.38,
55.82, 13.62. HRMS (FAB) m/z (M+Na) calcd for
C26H34O8 m/e: 511.1944. Found: 511.1942.
4.1.15. Methyl 3-O-[4[(1-methyl-indol-3-yl)-prop-
1-ynyl]-benzyl]-b-D-galactopyranoside (19). White solid.
Yield: 23.4 mg (24% over five steps, calculated from 23).
[a]2D5 +20.0 (c 0.01, CD3OD). 1H NMR (400 MHz; CD3OD)
d 7.58 (d, J¼8.0 Hz, 1H), 7.36–7.31 (m, 4H), 7.28 (d,
J¼8.3 Hz, 1H), 7.13 (dt, J¼7.1, 1.0 Hz, 1H), 7.06 (s, 1H),
7.01 (dt, J¼7.9, 0.9 Hz, 1H), 4.69 (d, J¼12.1 Hz, 1H), 4.58
(d, J¼12.1 Hz, 1H), 4.09 (d, J¼7.8 Hz, 1H), 3.98 (d,
J¼3.1 Hz, 1H), 3.82 (s, 2H), 3.74–3.66 (m, 5H), 3.61 (dd,
J¼7.8, 1.8 Hz, 1H), 3.48 (s, 3H), 3.40 (t, J¼5.7 Hz, 1H),
3.31 (dd, J¼9.7, 3.3 Hz, 1H); 13C NMR (100 MHz;
CD3OD) d 139.59, 138.85, 132.45, 128.95, 128.62, 127.88,
124.61, 122.66, 119.79, 119.66, 110.96, 110.28, 105.97,
89.27, 82.52, 81.78, 76.52, 72.11, 71.76, 67.08, 62.49, 57.28,
32.73, 16.56. HRMS (ES) m/z (M+Na) calcd for
C26H29O6NNa m/e: 474.1893. Found: 474.1891.
4.1.19. Methyl 3-O-[3-(3-hydroxypropargyl)-benzyl]-b-
D-galactopyranoside (25). Methyl 3-O-(3-iodobenzyl)-b-
D-galactopyranoside 2 (200 mg, 0.49 mmol) was dissolved
in THF/Et3N (1:1, 10 mL) together with CuI (9 mg,
10 mol %) and propargylic alcohol (34 mL, 0.59 mmol).
The mixture was thoroughly degassed with N2 for 10 min
whereafter Pd(PPh3)4 (28 mg, 5 mol %) was added. The so-
lution was degassed for an additional 5 min and then stirred
at ambient temperature until TLC indicated total conversion.
The solvent was evaporated under reduced pressure and
the residue subjected to multiple chromatographies (SiO2,
EtOAc/MeOH 10:1) to afford methyl 3-O-[3-(3-hydroxy-
propargyl)-benzyl]-b-D-galactopyranoside 32 (137 mg,
4.1.16. Methyl [3-O-(4-hex-5-en-1-ynyl)-benzyl]-b-D-gal-
actopyranoside (20). White solid. Yield: 21.3 mg (45%
over five steps, calculated from 23). [a]2D5 +21.9 (c 0.014,
CD3OD). 1H NMR (400 MHz; CD3OD) d 7.34 (d,
J¼8.2 Hz, 2H), 7.27 (d, J¼8.2 Hz, 2H), 5.95–5.84 (m,
1H), 5.09 (d ab-quart, J¼17.1, 1.8 Hz, 1H), 5.00 (d ab-quart,
J¼10.2, 1.8 Hz, 1H), 4.71 (d, J¼12.1 Hz, 1H), 4.60 (d, J¼
12.1 Hz, 1H), 4.10 (d, J¼7.8 Hz, 1H), 3.99 (d, J¼3.2 Hz,
1H), 3.75–3.66 (m, 2H), 3.61 (dd, J¼7.8, 1.9 Hz, 1H),
3.49 (s, 3H), 3.41 (t, J¼6.5 Hz, 1H), 3.31 (dd, J¼9.6,
3.3 Hz, 1H), 2.44 (t, J¼7.2 Hz, 2H), 2.29 (quart, J¼
6.45 Hz, 2H); 13C NMR (100 MHz; CD3OD) d 139.54,
138.29, 132.39, 128.92, 124.61, 116.10, 105.98, 90.19,
82.54, 81.86, 76.53, 72.11, 71.78, 67.08, 62.49, 57.27,
34.21, 19.97. HRMS (FAB) m/z (M+Na) calcd for
C26H34O8 m/e: 385.1627. Found: 385.1636.
1
87%) as a white solid. [a]2D5 +19.8 (c 0.016, CD3OD). H
NMR (400 MHz; CD3OD) d 7.51 (br s, 1H), 7.40 (d, J¼
6.8 Hz, 1H), 7.32–7.28 (m, 2H), 4.73 (d, J¼12.0 Hz, 1H),
4.60 (d, J¼12.0 Hz, 1H), 4.36 (s, 2H), 4.12 (d, J¼8.0 Hz,
1H), 4.03 (d, J¼3.2 Hz, 1H), 3.74–3.70 (m, 2H), 3.65–3.61
(m, 1H), 3.50 (s, 3H), 3.44 (tr, J¼6.0 Hz, 1H), 3.37 (dd,
J¼9.6, 3.2 Hz, 1H); 13C NMR (100 MHz; CD3OD)
d 140.08, 131.68, 131.40, 128.09, 128.69, 123.97, 105.64,
88.50, 85.14, 82.29, 76.18, 71.58, 71.42, 66.70, 62.17,
57.08, 56.94, 50.90, 49.37. HRMS (FAB) m/z (M+Na) calcd
for C17H22NaO7 m/e: 361.1263. Found: 361.1259.
Acknowledgements
4.1.17. Methyl 3-O-[4-(2,4-dimethoxyphenyl-prop-1-
ynyl)]-benzyl-b-D-galactopyranoside and methyl 3-O-
[4-(2,6-dimethoxyphenyl-prop-1-ynyl)]-benzyl-b-D-gal-
actopyranoside (21). White solid. Yield: 15.1 mg (51% over
five steps, calculated from 23). Isomeric mixture. [a]2D5 +17.3
(c 0.01, CD3OD). 1H NMR (400 MHz; CD3OD) d 7.41–7.32
(m) and 7.19 (t, J¼8.3 Hz, 5H), 7.26 (d, J¼8.3 Hz, 1H), 6.63
(d, J¼8.4 Hz, 1H), 6.51 (br s, 1H), 6.49 (d, J¼2.4 Hz, 1H),
4.76–4.70 (m, 1H), 4.62–4.59 (m, 1H), 4.13 (dd, J¼7.8,
5.7 Hz, 1H), 4.03 (dd, J¼3.1, 0.9 Hz) and 4.01 (dd, J¼3.1,
0.9 Hz, 1H), 3.85 (s, 3H), 3.83 (s, 2H), 3.78 (s, 2H), 3.75–
3.71 (s, 3H), 3.70–3.67 (m, 1H), 3.65–3.60 (m, 3H), 3.52–
3.51 (m, 4H), 3.46–3.42 (m, 2H), 3.36 (dd, J¼39.6,
3.3 Hz, 1H); 13C NMR (100 MHz; CD3OD) d 161.51,
159.43, 159.11, 139.66, 139.21, 132.46, 132.46, 132.41,
130.19, 129.23, 128.98, 128.97, 128.88, 125.00, 124.57,
A 2-azidoethyl glycoside precursor for the probe used for
galectins-8N and -9N was provided by the Consortium for
Functional Glycomics (grant number NIH GM62116). We
thank Barbro Kahl-Knutsson and Susanne Carlsson for help
with galectin production and fluorescence polarisation mea-
˚
surements, and Daniel Aberg for performing some test reac-
tions. We gratefully acknowledge financial support from
€
AstraZeneca R&D Molndal, the Swedish Foundation for
Strategic Research and the Swedish Research Council (VR).
References and notes
1. Barondes, S. H.; Castronovo, V.; Cooper, D. N. W.; Cummings,
R. D.; Drickamer, K.; Feizi, T.; Gitt, M. A.; Hirabayashi, J.;

A. Bergh et al. / Tetrahedron 62 (2006) 8309–8317
8317
Hughes, C.; Kasai, K.; Leffler, H.; Liu, F. T.; Lotan, R.;
Mercurio, A. M.; Monsigny, M.; Pillai, S.; Poirer, F.; Raz,
A.; Rigby, P. W. J.; Rini, J. M.; Wang, J. L. Cell 1994, 76,
597–598.
Chemistry; Nicolaou, K. C., Hanko, R., Hartwig, W., Eds.;
€
Wiley-VCH: Weinheim, 2002; (c) Zaragoza Dorwald, F.
Organic Chemistry on Solid Phase; Wiley-VCH: Weinheim,
2000.
2. (a) Leffler, H.; Carlsson, S.; Hedlund, M.; Qian, Y.; Poirier, F.
Glycoconjugate J. 2004, 19, 433–440; (b) Houzelstein, D.;
Goncalves, I. R.; Fadden, A. J.; Sidhu, S. S.; Cooper, D. N.;
Drickamer, K.; Leffler, H.; Poirier, F. Mol. Biol. Evol. 2004,
21, 1177–1187.
3. (a) Molinero, L. L.; Fuertes, M. B.; Rabinovich, G. A.;
Fainboim, L.; Zwirner, N. W. J. Leukocyte Biol. 2002, 71,
791–797; (b) Baum, L. G.; Rabinovich, G. A.; Tinari, R.;
Paganelli, C.; Natoli, F. T.; Liu, B. S.; Iacobelli, S. Trends
Immunol. 2002, 23, 313–320; (c) Liu, F. T.; Rabinovich,
G. A. Nat. Rev. Cancer 2005, 5, 29–41.
8. (a) Cassel, J. A.; Leue, S.; Gachkova, N. I.; Kann, N. C. J. Org.
Chem. 2002, 67, 9460–9463; (b) Gachkova, N.; Cassel, J.;
Leue, S.; Kann, N. J. Comb. Chem. 2005, 7, 449–457.
9. (a) Auge, C.; David, S.; Veyrieres, A. J Chem. Soc., Chem.
Commun. 1976, 375–376; Tsuda, Y.; Haque, M. E.;
Yoshimoto, K. Chem. Pharm. Bull. 1983, 31, 1612–1624; (b)
Alais, J.; Maranduba, A.; Veyrieres, A. Tetrahedron Lett.
1983, 24, 2383–2386.
10. Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett.
1975, 4467–4470.
11. Williamson, A. W. J. Chem. Soc. 1852, 4, 229.
12. Teobald, B. J. Tetrahedron 2002, 58, 4133–4170.
13. Tanaka, S.; Tsukiyama, T.; Isobe, M. Tetrahedron Lett. 1993,
34, 5757–5760.
4. Pieters, R. J. Chembiochem 2006, 7, 721–728.
5. Rini, J. M.; Lobsanov, Y. D. Curr. Opin. Struct. Biol. 1999, 9,
578–584.
€
6. (a) Sorme, P.; Kahl-Knutsson, B.; Wellmar, U.; Magnusson,
B.-G.; Leffler, H.; Nilsson, U. J. Methods Enzymol. 2003,
14. Zemplen, G. Ber. 1927, 60B, 1555–1564.
€
15. Sorme, P.; Kahl-Knutsson, B.; Huflejt, M.; Nilsson, U. J.;
Leffler, H. Anal. Biochem. 2004, 334, 36–47.
€
363, 157–169; (b) Sorme, P.; Qian, Y.; Nyholm, P.-G.;
Leffler, H.; Nilsson, U. J. Chembiochem 2002, 3, 183–189;
16. Cumpstey, I.; Carlsson, S.; Leffler, H.; Nilsson, U. J. Org.
Biomol. Chem. 2005, 3, 1922–1932.
€
(c) Sorme, P.; Arnoux, P.; Kahl-Knutsson, B.; Leffler, H.;
Rini, J. M.; Nilsson, U. J. J. Am. Chem. Soc. 2005, 127,
1737–1743.
7. (a) Combinatorial Chemistry; Bannwarth, W., Hinzen, B., Eds.;
Wiley-VCH: Weinheim, 2006; (b) Handbook of Combinatorial
17. Leonidas, D. D.; Vatzaki, E. H.; Vorum, H.; Celis, J. E.;
Madsen, P.; Acharya, K. R. Biochemistry 1998, 37, 13930–13940.
18. Perrin, D. D.; Armarego, L. F. Purification of Laboratory
Chemicals; Pergamon: Oxford, 1988.