Aryl O- and S-galactosides and lactosides as specific inhibitors of human galectins-1 and -3: Role of electrostatic potential at O-3

Article abstract of DOI:10.1016/j.bmcl.2005.12.010

Phase transfer catalyzed reaction was used for the high yielding synthesis of aryl 1-thio-β-d-galacto- and lacto-pyranosides carrying a panel of substituents on the phenyl groups. Best galectin-1 inhibitors were simple p-nitrophenyl thiogalactoside 5a for

Full text of DOI:10.1016/j.bmcl.2005.12.010

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1668–1672
Aryl O- and S-galactosides and lactosides as specific inhibitors
of human galectins-1 and -3: Role of electrostatic potential at O-3
a
b
b
a
a,
*
`
´
Denis Giguere, Sachiko Sato, Christian St-Pierre, Suzanne Sirois and Rene Roy
a
`
Departement of Chemistry, Universite´ du Que´bec A Montre´al, PO Box 8888, Succ. Centre-Ville Montreal, Que., Canada H3C 3P8
bResearch Centre for Infectious Diseases, Faculty of Medicine, Universite´ Laval 2705 boul. Laurier,
RC-9700 Sainte-Foy, Que., Canada G1V 4G2
Received 31 October 2005; revised 2 December 2005; accepted 5 December 2005
Available online 5 January 2006
Abstract—Phase transfer catalyzed reaction was used for the high yielding synthesis of aryl 1-thio-b-D-galacto- and lacto-pyrano-
sides carrying a panel of substituents on the phenyl groups. Best galectin-1 inhibitors were simple p-nitrophenyl thiogalactoside 5a
for the monosaccharide and o-nitrophenyl thiolactoside 6f or napthylsulfonyl lactoside 8c, both being 20 times better relative to
natural ligands. Relative inhibitory properties as low as 2500 and 40 lM were observed, respectively. The electronic effects of the
lactoside aglycons directly influenced the electrostatic potential at O-3, which was associated with the inhibitory potencies against
galectin-1.
ꢀ 2006 Elsevier Ltd. All rights reserved.
The galectins are a family of cytosolic b-D-galactoside
binding proteins for which fourteen members have been
identified in mammals.1–4 Galectins-1 and -2 possess one
carbohydrate recognition domain (CRD) per subunit
and exist as dimers, in comparison to galectins-4, -8,
and -9 which are connected to a short linker region.
Galectin-3 (Gal-3) is unique and exists as chimera type
composed of one CRD connected to non-CRD domain
consisting of collagen-like repeats of a peptide sequence
rich in proline and glycine, and is capable of self-associ-
ation. The C-terminal CRD domain of Gal-3 is homol-
ogous to that of Gal-1.5 The role of the galectins is not
yet clear, but a striking common feature of all galectin
members is the strong modulation of their expression
during development, differentiation stages, and under
various physiological or pathological conditions.2 Gal-
3 possesses diverse biological activities and is involved
in colon cancer metastasis,6 brain tumor progression,7
inhibits metastasis-associated cancer cell adhesion,8
and may play a key role in innate immunity.9,10 Recent
reports suggest that Gal-311 and Gal-112–14 can regulate
cell apoptosis.15 Also, Gal-1 acts as a soluble host factor
that promotes HIV-1 infectivity through stabilization of
virus attachment to host cells.16 Thus, selective inhibi-
tion of galectins may lead to anti-cancer, anti-inflamma-
tory, or even anti-HIV properties.2,16–18
Naturally occurring carbohydrate ligands for galectins
have low affinities, are too polar to be used as drugs,
and possess poor physiological stabilities due to their
acid-sensitive glycosidic bonds.19 For instance, galactose
and lactose have inhibitory properties of 50 and 0.8 mM
for both Gal-1 and Gal-3, respectively. Combining
experimental information from the binding data and
high resolution X-ray crystal structures of galectin–
carbohydrate ligand complexes enables rational design
approach for the development of new classes of gly-
comimetic inhibitors with high affinity, stability, and
specificity.20–25 Nilsson et al. have explored the
30-position of a panel of lactoside derivatives for the syn-
thesis of high affinity inhibitors of Gal-3 using benzami-
do20–22 or triazole23 functionalities. We report herein the
efficient synthesis of small libraries of aryl thiogalacto-
sides and lactosides using phase transfer catalysis reac-
tion (PTC) with the aim to further explore the effect of
the aglycons on the relative capacity toward galec-
tins.24,25 Encouraging inhibitory properties were ob-
tained for inhibitors having anomeric sulfides that
conferred better stability under physiological condi-
tions. The attachment of aryl groups also increased the
lipophilicities of the molecules, a desirable property
for better cell permeability in in vivo assays. All the
Keywords: Galectins; Aryl lactosides; Hemagglutination; Electrostatic
potential; Inhibitors.
*
Corresponding author. Tel.: +1 514 987 3000x2546; fax: +1 514 987
4054; e-mail: roy.rene@uqam.ca
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.010

`
D. Giguere et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1668–1672
1669
reactions were efficiently accomplished at room temper-
ature under mildly basic conditions, proceeded with
good to excellent yields, and were essentially completed
within three hours under non-anhydrous conditions.26
Exploration of anomeric sulfones from anomeric sulfide
oxidation has also been accomplished.
and 6a–f were obtained in good to excellent yields
(Table 1) and were ready to be evaluated on the
galectins using an inhibition of hemagglutination assay
described below, with only two synthetic steps.
With aryl thioglycosides 3b, 4c, and 4d in hand, synthe-
sis of anomeric sulfones by oxidation of the sulfide
groups using mCPBA (2.1 equiv, CH2Cl2) afforded sulf-
ones 7b, 8c, and 8d, respectively, after de-O-acetylation
(Scheme 2) (Table 2).
Reactions toward Gal-1 and -3 inhibitors started from
commercially available acetobromogalactose 1 and per-
acetylated lactosyl bromide 2, which was obtained in
quantitative yield by treatment of lactose peracetate
with HBr in AcOH. As expected, the phase transfer
catalyzed nucleophilic displacements of both glycosyl
bromides by the aryl alcohols or aryl thiols occurred
with complete anomeric inversion to afford only the
corresponding b-glycoside derivatives 3a–e and 4a–f,
respectively (Scheme 1). After de-O-acetylations
with methanolic sodium methoxide, compounds 5a–e
All new compounds and control 9 (galactose) and 10
(lactose) were tested by inhibition of hemagglutination
assay at a concentration of 1 lM for both galectins.
Hemagglutination assays were performed using red
blood cells, type O, fixed with 3% glutaraldehyde–
0.0025% NaN3 in PBS,16,27 to confer both lectins equal
relative affinities. Table 3 shows inhibitory properties
and relative activities of our derivatives toward Gal-1
and -3. The first overall observation was that none of
our compounds bound to human Gal-4, indicating that
phenol and thiophenol derivatives were better inhibitors
and improved specifity toward Gal-1 and -3.28 S-Galac-
tosides improved the inhibitory properties against Gal-1
(5a vs 5d) and the anomeric sulfones did not have bene-
ficial effect (7b). Amongst all galactosides tested, p-nitro-
thiophenyl galactoside 5a demonstrated the best affinity
(2500 lM) with a relative inhibitory potency 20 times
better than the control galactose 9. Lactosides 6f and
8c were more specific toward Gal-1 than Gal-3.
1
1
R
R
OAc
O
OAc
O
2
2
R
R
a
X
3
AcO
AcO
R
AcO
AcO
Br
X = O or S
1
1
1 R = OAc, R = H
2
3a-e R = OAc, R = H
2
1
1
2 R = H
4a-f R = H
2
2
R
= β-D-Gal(OAc)
4
R
= β-D-Gal(OAc)
4
b
1
5a-e R = OH, R = H
2
1
6a-f R = H
2
R
= β-D-Gal(OH)
4
Most O-aryl lactosides were estimated as being 3 times
better than the natural analogue 10, thus indicating a
preference for aromatic aglycons.24,25 S-Lactosides were
Scheme 1. Reagents and conditions: (a) HXR3 (X = O or S), TBAHS,
1 M Na2CO3, DCM; (b) NaOMe, MeOH.
Table 1. Synthesis of different O- and S-galacto- and lacto-pyranosides 5–6 using PTC reactions followed by treatment with methanolic sodium
methoxide
Compound
R1
R2
H
R3
X
S
Yield (%)
9226
5a
5b
OH
OH
NO2
H
S
86
OMe
5c
5d
OH
OH
H
H
S
78
a
O
NO2
O2N
5e
OH
H
O
73
6a
6b
H
H
b-D-Gal(OH)4
b-D-Gal(OH)4
S
S
9124
98
NO2
OMe
6c
H
b-D-Gal(OH)4
S
93
6d
6e
H
H
b-D-Gal(OH)4
b-D-Gal(OH)4
S
88
85
Br
O
NO2
O2N
6f
H
b-D-Gal(OH)4
O
75
a Commercially available.

`
D. Giguere et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1668–1672
1670
1
1
R
R
OAc
O
OH
O
a, b
O
2
2
R
R
S
3
S
3
AcO
HO
R
R
AcO
HO
O
1
1
1
3b
R
= OAc, R = H
7b R = OH, R = H
2
2
1
4c,d R = H
8c,d R = H
2
2
R
= β-D-Gal(OH)
R
= β-D-Gal(OAc)
4
4
Scheme 2. Reagents and conditions: (a) mCPBA (2.1 equiv), DCM,
0 ꢁC; (b) NaOMe, MeOH.
Table 2. Synthesis of sulfones 7–8 using mCPBA as an oxidizing agent
Sulfides
Sulfones
R3
Yield (%)
Figure 1. Connolly surface using a grid spacing of 0.75 colored by
Active Lone Pair and docking of 6f in Gal-1 showing the o-nitrophenyl
aglycon pointing out of the CRD pocket and O-3 (yellow) into an H-
bonding network. The blue regions are hydrophobic, the red are
hydrophilic, while the white indicate regions through which hydrogen
bonds are likely to form.
3b
4c
4d
7b
8c
8d
p-MeOPh
b-Napht
p-BrPh
89
92
98
Table 3. Inhibitory properties of compounds 5–10 against Gal-1
and -3
This suggested that the inhibitory specificity was provid-
ed from indirect contributions. Hydrogen bondings be-
tween Gal-1 and natural lactose only occurred through
its O-3 hydroxyl group (Fig. 2) with two coming from
Arg48, one from Glu71, and one from the endocyclic
oxygen of the galactoside moiety.31 These interactions
create a network of three hydrogen bonds which are
common features in carbohydrate–lectin recogni-
tions.32–34 Semi-empirical calculations correlated with
the chemical modifications done at the anomeric posi-
tion with the effect on charge density at the O-3 oxygen
of the glucoside residue of lactose.
Compound
Inhibitory properties
(mM)
Relative activitya,b
Galectin-1 Galectin-3 Galectin-1 Galectin-3
5a
5b
2.5
>5
>5
2.5
20
>10
10
>10
20
5c
5d
5
10
5
10
10
5
>10
5
>10
5e
6a
>5
>5
0.313
0.313
2.6
2.6
6b
6c
0.313
0.313
0.313
0.313
0.08
0.313
0.313
0.313
0.625
0.625
2.6
2.6
2.6
2.6
2.6
2.6
2.6
1.3
1.3
6d
6e
The O-3 electron density calculations were made on all
lactoside derivatives evaluated on Gal-1. Electron with-
drawing aglycons decreased the O-3 electronic densities
and favored better interactions with Glu71 which
accounted for 50% of the hydrogen bonding.
6f
7b
10
>5
>5
>10
20
>10
8c
8d
0.04
0.313
0.313
0.313
2.6
2.6
1
2.6
9 (Gal)
10 (Lac)
50
0.8
50
0.8
1
1
1
Hence, a direct correlation exists between the inhibitory
potencies and the charge density at O-3 calculated on
electrostatic potential. Figure 3 reveals the correlation
between the inhibitory properties and the electron densi-
a Compounds 5 and 7 were compared to galactose 9, compounds 6 and
8 were compared to lactose 10.
b Lactose is ꢀ50· better than galactose.
less potent than O-lactosides for Gal-3 (6a compared to
6e). For Gal-1 (8c,29 40 lM), the presence of an anomer-
ic sulfone provided a compound 20 times better than
lactose 10. The position of the nitro group on the phenyl
moiety seemed to play a key role for the specificity be-
tween Gal-1 and -3. There was no increase of inhibitory
potency for Gal-3 when the nitro group was para- or
ortho-substituted (6e and 6f). Gal-1 was sensitive to this
change and became 10 times more potent than the refer-
ence 10 when the nitro group was ortho-substituted (6f,
80 lM).
The nitro group position might be important not only
from a steric point of view but also from an electrostatic
one. Figure 1 shows the O-linked o-nitrophenyl lactoside
6f extending out of the CRD pocket, thus lacking direct
interactions with the lectin itself.30
Figure 2. O-3 (yellow) hydrogen bonding interactions from lactose 10,
and Arg48 and Glu71 residues of Gal-1 based on the X-ray crystal
structure.30

`
D. Giguere et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1668–1672
1671
2. (a) Horrie, H. Curr. Drug Targets 2005, 6, 373; (b)
Chiariotti, L.; Salvatore, P.; Frunzio, R.; Bruni, C. B.
Glycoconjugate J. 2004, 19, 441.
3. Cooper, D. N. W.; Barondes, S. H. Glycobiology 1999, 9,
979.
4. Hirabayashi, J.; Kasai, K. Glycobiology 1993, 3, 297.
5. Elola, M. T.; Chiesa, M. E.; Alberti, A. F.; Mordoh, J.;
Flink, N. E. Biomed. Sci. 2005, 12, 13.
1,0
0,8
0,6
0,4
0,2
0,0
10
6e
6f
8c
6. Bresalier, R. S.; Mazurek, N.; Stenberg, L. R.; Byrd, J. C.;
Yunker, C. K.; Nangia-Makker, P.; Raz, A. Gastroente-
roloy 1998, 115, 287.
-0,355
-0,345
-0,335
-0,325
-0,315
O-3 Electron Charge Density
7. Stillman, B. N.; Mischel, P. S.; Baum, L. G. Brain Pathol.
2005, 15, 124.
Figure 3. Correlation between O-3 electron charge density calculated
on electrostatic potential of lactosides 6e, 6f, 8c, and reference 10,
and the inhibitory properties (mM) of those compounds evaluated on
Gal-1.
8. Zou, J.; Glinsky, V. V.; Landon, L. A.; Matthews, L.;
Deutscher, S. L. Carcinogenesis 2005, 26, 309.
9. (a) Sato, S.; Nieminem, J. Glycoconjugate J. 2004, 19, 441;
(b) Sato, S. Trends Glycosci. Glycotechnol. 2002, 14, 285.
10. Liu, F. T. Int. Arch. Allergy Immunol. 2005, 136, 385.
11. (a) Califice, S.; Castronovo, V.; Van Den Brule, F. Int. J.
Oncol. 2004, 25, 983; (b) Nakahara, S.; Oka, N.; Raz, A.
Apoptosis 2005, 10, 267.
12. Perillo, N. L.; Pace, K. E.; Seilhamer, J. J.; Baum, L. G.
Nature 1995, 378, 736.
13. Szoke, T.; Kayser, K.; Baumhakel, J. D.; Trojan, I.;
Furak, J.; Tiszlaviez, L.; Horvath, A.; Szluha, K.; Gabius,
ty of compounds 6e, 6f, 8c, and reference 10 (6a–d not
shown).35 O-Lactoside 6e (p-nitrophenyl) possesses a
theoretical electron density of ꢁ0.346 compared to that
of o-nitrophenyl 6f of ꢁ0.322. This difference of electron
density could explain the relative inhibitory potency var-
iation on Gal-1. Compound 8c, having the lowest O-3
electron density, has also the highest inhibitory potency.
´
H.-J.; Andre, S. Oncology 2005, 69, 167.
In conclusion, preliminary data showed a novel direc-
tion toward more specific Gal-1 and -3 inhibitors. Most
compounds were made in high yields with only 2 or 3
synthetic steps from commercially available starting
materials. The PTC reaction has proven its practicality
for the synthesis of more stable and less polar inhibitors
compared to natural ligands. The best inhibitors 5a
(2500 lM) and 8c (40 lM) have shown 20 times better
affinity toward Gal-1 as compared to galactose 9 and
lactose 10, respectively. The strategy used for the synthe-
sis of inhibitors provides a large potential for further
improvements by changing the nucleophiles in the
PTC reaction. Additionally, the actual modifications,
combined to those already published,20–25,36–38 can lead
to improved pharmacological properties. Finally, inhib-
itory data correlated with the electron density at the O-3
group of the glucose unit within the lactoside ligand that
led to a greater inhibitory capacity for Gal-1. Current
progresses are made by taking advantage of this effect
and are guiding the design of more selective inhibitors
against human galectins.37 Although the above com-
pounds are notably less efficient than those described
by Nilsson et al.,20–23 we used inhibition of hemaggluti-
nation assays, known to require higher concentrations
of ligand.
14. Battig, P.; Saudan, P.; Gunde, T.; Bachmann, M. F. Mol.
Immunol. 2004, 41, 9.
15. (a) Hsu, D. K.; Liu, F. T. Glycoconj. J. 2004, 19, 507; (b)
Hernandes, J. D.; Baum, L. G. Glycobiology 2002, 12, 127R.
16. Ouellet, M.; Mercier, S.; Pelletier, I.; Bounou, S.; Roy, J.;
Hirabayashi, J.; Sato, S.; Tremblay, M. J. J. Immunol.
2005, 174, 4120.
17. Rabinovich, G. A. Br. J. Cancer 2005, 92, 1188.
18. Kato, T.; Ren, C. H.; Wada, M.; Kawanami, T. Curr.
Drug Targets 2005, 6, 407.
19. Leffler, H.; Barondes, S. H. J. Biol. Chem. 1986, 22, 10119.
20. So¨rme, P.; Qian, Y.; Nyholm, P.-G.; Leffler, H.; Nilsson,
U. J. ChemBioChem. 2002, 3, 183.
21. So¨rme, P.; Arnoux, P.; Kahl-Knutsson, B.; Leffler, H.; Rini,
J. M.; Nilsson, U. J. J. Am. Chem. Soc. 2005, 127, 1747.
22. Cumpstey, I.; Sundin, A.; Leffler, H.; Nilsson, U. J.
Angew. Chem. Int. Ed. 2005, 44, 5110.
23. (a) Salameh, B. A.; Leffler, H.; Nilsson, U. J. Bioorg. Med.
Chem. Lett. 2005, 15, 3344; (b) Cumpstey, I.; Carlsson, S.;
Leffler, H.; Nilsson, U. J Org. Biomol. Chem. 2005, 3,
1922; (c) Tejler, J.; Leffler, H.; Nilsson, U. J Bioorg. Med.
Chem. Lett. 2005, 15, 2343.
24. Lee, R. T.; Ichikawa, Y.; Allen, H. J.; Lee, Y. C. J. Biol.
Chem. 1990, 265, 7864.
25. Ahmed, H.; Allen, H. J.; Sharma, A.; Matta, K. L.
Biochemistry 1990, 29, 5315.
26. Roy, R.; Tropper, F. D.; Cao, S.; Kim, J. M. ACS Symp.
Ser. 1997, 659, 163.
27. Butler, W. T. J. Immunol. 1963, 90, 663.
28. Concentration at 0.5 lM for Gal-4 which is known to
have different binding requirements see: Ideo, H.; Seko,
A.; Ohkura, T.; Matta, K. L.; Yamashita, K. Glycobiology
2002, 12, 199.
29. This compound was difficult to dissolve, 5% of DMSO
was used at which hemagglutination by galectins was not
affected.
Acknowledgments
This work was supported from Natural Sciences and
Engineering Research Council of Canada (NSERC)
and a Canadian Research Chair in Therapeutic Chemis-
try to R.R. and from Canadian Institutes for Health
Research to S.S.
30. Docking was made with the MOE program. The PDB
entry for Gal-1 was 1GZW.
´
31. Lopez-Lucendo, M. F.; Solis, D.; Andre, S.; Hirabayashi,
J.; Kasai, K.-i.; Kaltner, H.; Gabius, H.-J.; Romero, A.
J. Mol. Biol. 2004, 343, 957.
References and notes
32. Manoj, N.; Sirnivas, V. R.; Surolia, A.; Vijayan, M.;
Suguna, K. J. Mol. Biol. 2000, 302, 1129.
1. Barondes, S. H.; Cooper, D. N. W.; Gitt, M. A.; Leffler,
H. J. Biol. Chem. 1994, 269, 20807.

`
D. Giguere et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1668–1672
1672
´
33. Imberty, A.; Gauthier, C.; Lescar, J.; Perez, S.; Wynsi, L.;
Loris, R. J. Biol. Chem. 2000, 275, 17541.
38. Compound 8c: mp 181.5–182.5 ꢁC; [a]D + 6.5 (c = 1,
DMSO); 1H NMR (CDCl3) d (ppm): 8.51 (s, 1H,
HAR), 8.03–7.87 (m, 4H, HAR), 7.66–7.58 (m, 2H, HAR),
4.42 (d, J = 9.34 Hz, H-1), 4.21 (d, J = 7.42 Hz, 1H, H-
10), 3.65–3.26 (m, 12H, H-2-6, H-20-60); 13C NMR
(CDCl 3) d (ppm) : 135.2, 133.5, 132.8, 130.7, 130.6,
129.9, 129.04, 128.7, 128.7, 125.5 (7 C@O), 104.9 (C-10),
92.9 (C-1), 81.1 (C-4), 78.9 (C-3), 77.4 (C-2), 77.1 (C-5),
74.8 (C-30), 72.5 (C-50), 71.3 (C-20), 70.3 (C-40), 62.5
(C-6), 61.5 (C-60).
34. Ng, K. K.-S.; Weis, W. I. Biochemistry 1997, 36, 979.
35. All electrostatic potential calculations were made with the
CaChe software.
´
36. Ahmad, N.; Gabius, H.-J.; Andre, S.; Kaltner, H.;
Sabesan, S.; Roy, R.; Liu, B.; Macaluso, F.; Brewer, C.
F. J. Biol. Chem. 2004, 279, 10841.
´
37. Andre, S.; Liu, B.; Gabius, H.-J.; Roy, R. Org. Biomol.
Chem. 2003, 1, 3909.