Galectin-3-Binding Glycomimetics that Strongly Reduce Bleomycin-Induced Lung Fibrosis and Modulate Intracellular Glycan Recognition

Article abstract of DOI:10.1002/cbic.201600285

Discovery of glycan-competitive galectin-3-binding compounds that attenuate lung fibrosis in a murine model and that block intracellular galectin-3 accumulation at damaged vesicles, hence revealing galectin-3–glycan interactions involved in fibrosis progression and in intracellular galectin-3 activities, is reported. 3,3′-Bis-(4-aryltriazol-1-yl)thiodigalactosides were synthesized and evaluated as antagonists of galectin-1, -2, -3, and -4 N-terminal, -4 C-terminal, -7 and -8 N-terminal, -9 N-terminal, and -9 C-terminal domains. Compounds displaying low-nanomolar affinities for galectins-1 and -3 were identified in a competitive fluorescence anisotropy assay. X-ray structural analysis of selected compounds in complex with galectin-3, together with galectin-3 mutant binding experiments, revealed that both the aryltriazolyl moieties and fluoro substituents on the compounds are involved in key interactions responsible for exceptional affinities towards galectin-3. The most potent galectin-3 antagonist was demonstrated to act in an assay monitoring galectin-3 accumulation upon amitriptyline-induced vesicle damage, visualizing a biochemically/medically relevant intracellular lectin–carbohydrate binding event and that it can be blocked by a small molecule. The same antagonist administered intratracheally attenuated bleomycin-induced pulmonary fibrosis in a mouse model with a dose/response profile comparing favorably with that of oral administration of the marketed antifibrotic compound pirfenidone.

Full text of DOI:10.1002/cbic.201600285

Accepted Article
Title: Galectin-3-binding glycomimetics that strongly reduce bleomycin-induced lung
fibrosis and modulate intracellular glycan recognition
Authors: Tamara Delaine; Patric Collins; Alison MacKinnon; G Sharma; John Steg-
mayr; Vishal K Rajput; Santanu Mandal; Ian Cumpstey; Amaia Larumbe; Bader
A Salameh; Barbro Kahl-Knutsson; Hilde Van Hattum; Monique Van Scherpen-
zeel; Roland J Pieters; Tariq Sethi; Hans Schambye; Stina Oredsson; Hakon
Leffler; Helen Blanchard; Ulf Nilsson
This manuscript has been accepted after peer review and the authors have elected
to post their Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by using the Digital
Object Identifier (DOI) given below. The VoR will be published online in Early View as
soon as possible and may be different to this Accepted Article as a result of editing.
Readers should obtain the VoR from the journal website shown below when it is pu-
blished to ensure accuracy of information. The authors are responsible for the content
of this Accepted Article.
To be cited as: ChemBioChem 10.1002/cbic.201600285
Link to VoR: http://dx.doi.org/10.1002/cbic.201600285
A Journal of
www.chembiochem.org

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
Galectin-3-binding glycomimetics that strongly reduce
bleomycin-induced lung fibrosis and modulate intracellular
glycan recognition
T. Delaine,[a] P. Collins,[b] A. MacKinnon,[c] G. Sharma,[d] J. Stegmayr,[d] V. K. Rajput,[a] S. Mandal,[a] I.
Cumpstey,[a] A. Larumbe,[a] B. A. Salameh,[a]† B. Kahl-Knutsson,[d] H. van Hattum,[e] M. v.
Scherpenzeel,[e]‡ R. J. Pieters,[e] T. Sethi,[f] H. Schambye,[g] S. Oredsson,[h] H. Leffler,[d] H. Blanchard,*[b]
and U. J. Nilsson*[a]
Abstract: Discovery of glycan-competitive galectin-3-binding
compounds that attenuate lung fibrosis in a murine model and that
block intracellular galectin-3 accumulation at damaged vesicles,
hence revealing galectin-3-glycan interactions being involved in
fibrosis progression and in intracellular galectin-3 activities is
reported. Sixteen 3,3´-bis-(4-aryl-triazol-1-yl)-thiodigalactosides were
synthesized and evaluated as antagonists of galectin-1, 2, 3, 4 N-
terminal, 4 C-terminal, 7, 8 N-terminal, 9 N-terminal, and 9 C-
terminal domains. Compounds displaying low-nM affinities for
galectin-1 and 3 were identified in a competitive fluorescence
anisotropy assay. X-ray structural analysis of selected compounds in
complex with galectin-3 and galectin-3 mutant binding experiments
revealed that both aryl-triazolyl moieties and fluoro-substituents of
the compounds are involved in key interactions responsible for the
exceptional affinities for galectin-3. The most potent galectin-3
antagonist was demonstrated to act in an assay monitoring galectin-
3
accumulation upon amitriptyline-induced vesicle damage,
[a]
Dr. T. Delaine, Dr. V.K. Rajput, Dr. S. Mandal, Dr. I. Cumpstey, Dr.
A. Larumbe, Dr. B.A. Salameh, Prof. Dr. U.J. Nilsson
Centre for Analysis and Synthesis, Department of Chemistry
Lund University
POB 124, SE-221 00 Lund, Sweden.
E-mail: ulf.nilsson@chem.lu.se
visualizing biochemical/medical relevant intracellular lectin-
a
carbohydrate binding event and that it can be blocked by a small
molecule. The same antagonist administered intratracheally
attenuated bleomycin-induced pulmonary fibrosis in a mouse model
with
a dose-response profile comparing favorably to orally
[b]
[c]
[d]
[e]
Dr. P. Collins, Dr.H. Blanchard
Institute for Glycomics
Griffith University, Gold Coast Campus
Queensland 4222, Australia
E-mail: h.blanchard@griffith.edu.au
Dr. A. MacKinnon
MRC Centre for Inflammation Research, The Queen’s Medical
Research Institute
University of Edinburgh
administration of the marketed anti-fibrotic compound pirfenidone.
Introduction
The galectins are a family of proteins that have the ability to
cross-link -D-galactopyranoside-containing glycoproteins (and
other glycoconjugates) to form lattices[1] and thereby modulate
glycoprotein localization, transport, and residence times in
cellular compartments and at surfaces.[2] Cross-linking of
glycoproteins by galectins can occur due to the galectins’
capability to present multiple carbohydrate recognition sites
(CRD) depending on their type. Prototype galectins (1, 2, 7, 10,
11, 13, 14, and 15) contain one CRD but dimerize depending on
their concentration and ligand density. The tandem-repeat
galectins (4, 5, 6, 8, 9, and 12) contain and present two CRD’s,
Edinburgh, UK
Dr. G. Sharma, J. Stegmayr, Mrs. B. Kahl-Knutsson, Prof. Dr. H.
Leffler
Department of Laboratory Medicine, Section MIG
Lund University
BMC-C1228b, Klinikgatan 28, SE-221 84 Lund, Sweden
Dr. H. van Hattum, Dr. M. van Scherpenzeel, Prof. Dr. R.J. Pieters
Department of Medicinal Chemistry and Chemical Biology, Utrecht
Institute for Pharmaceutical Sciences
Utrecht University
P.O. Box 80082, 3508 TB Utrecht, The Netherlands
Prof. Dr. T. Sethi
Department of Respiratory Medicine and Allergy
Kings College, Denmark Hill Campus
London, UK
Dr. H. Schambye
Galecto Biotech ApS
COBIS
Ole Maaloes vej 3, Copenhagen N, DK-2200, Denmark
Prof. Dr. S. Oredsson
Department of Biology
Lund University
POB 118, Lund, 221 00 Sweden
[f]
and the chimera-type galectin-3 CRD is linked to
glycine/proline-rich collagen-like N-terminal domain that enables
oligomerization.
a
[g]
[h]
This organizational lattice-forming role of the galectins
influences glycoprotein activities and the duration thereof, as
well as glycoprotein intracellular trafficking and sorting. This
manifests itself in different effects on the cellular level that
depend on a match between galectin type and expression, as
well as on the glycan structures in the cell. For example,
galectin-glycoconjugate interactions control cell properties and
functions, cell adhesion, have immunomodulatory effects[3] and
effects on tumor growth and metastases.[4] The cellular
mechanisms and roles in inflammation and cancer point to the
†
‡
Present address: Chemistry Department, The Hashemite University,
PO Box 150459, Zarka 13115 Jordan
Present address: Translational Metabolic Laboratory, Radboud
University Medical Center, Nijmegen, The Netherlands
Supporting information for this article is given via a link at the end of
the document.

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
use of galectin CRD antagonists as therapeutic agents and
several ex vivo[5] and in vivo[6] studies of the most studied and
well characterized galectin-3 have corroborated such
hypotheses.
Among attempts to develop small and drug-like molecules as
galectin-3 antagonists, substitution of galactose[6a, 7] (as such, or
part of lactose or N-acetyl-lactosamine (LacNAc)) and 3,3-di-
substitution of thiodigalactoside[8] have proven to be successful.
In particular, high-affinity small-molecule galectin-3 antagonists
with sub-µM affinities have been discovered by appending
aromatic amido groups or 4-amido-1,2,3-triazolyl groups at both
C-3 carbons in thiodigalactoside.[8a, 8b, 8d] Here, we present 4-
aryl-1,2,3-triazolyl thiodigalactoside-based derivatives as
significantly improved antagonists with selectivity for galectin-1
and 3. Furthermore, an investigation based on three X-ray
structures of galectin-3 in complex with inhibitors and on
galectin-3 mutant studies revealed that the aryl-triazolyl groups
form affinity-enhancing interactions with arginine side-chains
and with ß-strand backbones. One selected compound was
demonstrated to function intracellularly in an amitriptyline-
induced vesicle damage assay and to reduce fibrosis levels in a
murine bleomycin lung fibrosis model.
Results and Discussion
Synthesis
The ditriazolyl-thiodigalactosides 3-10 and 12 were synthesized
by Cu(I)-catalyzed cycloadditions between the known diazide 1[9]
and phenylacetylenes (Scheme 1), while synthesis of the
unsubstituted phenyltriazole 2 and the phenoxyphenyltriazole 11
have been reported earlier[10] (Table 1). The 1-naphthyl-triazole
17 was synthesized essentially following a previously published
alternative procedure[8d] that involved cycloaddition of 1-
ethynylnaphthalene with the acetylated galacto azide 13[11] to
give the triazole 14. Bromination of 14 and subsequent double
substitution of the bromide 15 with sodium sulfide resulted in the
thiodigalactoside 16 in a moderate yield. De-O-acetylation of 16
gave the target 1-naphthyl-triazole 17.
Scheme 1. Reagents and conditions: (a) Alkyne, CuI, Et3N, DMF; (b) Alkyne,
CuI, DIPEA, toluene, 65-80°C; (c) HBr/AcOH; (d) Na2S, MS 4Å, MeCN; (e)
BuNH2, MeOH.
This suggests that one or both of the galectin-1 subsites that
accommodate the phenyl groups of 2-11 are tight with limited
possibilities for substitutions as suggested based on earlier
analyses of 2 and 11.[10] The preference for substitution position
of the fluorophenyl derivatives 3-5 (m<p<o) is reflected in the
corresponding trifluoromethyl series 7-9, albeit at somewhat
higher Kd values. Although the fluorophenyl-carrying 4 and 5
indeed reach low nM affinities for galectin-1, more noteworthy is
the even higher affinity of the thienyl compound 12. This
compound provided near quantitative inhibition at all
concentrations tested and an accurate dissociation constant
could not be reliably calculated. Hence, the dissociation constant
could only be estimated to be less than 10 nM, which is at least
Galectin affinities and structure-activity relationships
With a panel of bis-aryltriazolyl thiodigalactosides 2-12 and 17 at
hand, affinities towards galectin-1, 2, 3, 4 N- and C-terminal
domains, 7, 8 N-terminal domain, and 9 N- and C-terminal
domains were determined in a competitive protein-binding assay
based on fluorescence anisotropy as earlier described in
detail.[12] Except for galectin-8N, all investigated galectins bound
all, or most, of the inhibitors 2-12 and 17 with affinities
significantly greater than those of the parent unsubstituted
thiodigalactoside (Table 1). Galectin-1 bound all phenyl-triazoles,
unsubstituted or with smaller substituents, (2-9) with indeed high
affinities, while larger substituents (10 and 11) significantly
reduced affinity. Interestingly, 3- and 4-fluorinated phenyl
compounds 4 and 5 turned out to be the only ones better than
the unsubstituted phenyl 2 and the 2-fluorophenyl derivative 3,
with dissociation constants as low as 12 nM (3-fluorophenyl 4).
2400–fold
better
than
the
reference
unsubstituted
thiodigalactoside and natural disaccharide ligands.

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
fluorophenyl compound 3, reached only a moderate affinity of
about 1 µM, which is nevertheless significantly better than the
parent unsubstituted thiodigalactoside reflecting the presence of
positive interactions between 2-17 and this galectin. Galectin-3
was well inhibited by several compounds and interestingly
showed a selectivity profile similar to galectin-1; phenyl moieties
carrying small substituents (2-9), as well as the thienyl moiety
(12), conferred high affinity, while phenyls carrying larger
substituents (10-11), as well as the naphthyl (17), were less
efficiently bound by this galectin. A notable difference is,
however, that while the biphenyl 10 is virtually detrimental to
binding (as for galectin-1), the 4-phenoxy-substituted phenyl 11
is reasonably well tolerated by galectin-3, with a sub-µM affinity,
which is not the case for galectin-1. Hence, compound 11
displays, as earlier reported,[10] an important more than 50-fold
selectivity for galectin-3 over galectin-1. The reverse situations
holds for the thienyl 12, which inhibits galectin-3, albeit with an
affinity of 65 nM, but still less well than galectin-1. Hence, the
thienyl derivative 12 has a clear selectivity for galectin-1 over
galectin-3 and thus through proper choice of aryl substituents on
the triazole rings, selectivity for galectin-1 (by 12) or galectin-3
(by 11) is achieved.
Table 1. Dissociation constants (µM) for 2-12, and 17 and galectin-1, 2, 3, 4, 4
N-terminal domain, 4 C-terminal domain, 7, 8 N-terminal domain, and 9 N- and
C-terminal domain determined with a competitive fluorescence anisotropy
assay.[12]
Galectin
1
2
3
4N
4C
7
8N
9N
9C
2
0.04
9[10]
1.2±
0.27
0.044[
2.6±
0.25
0.19
±0.0
15
1.2
±0.
21
>1
00
2.3±
0.23
0.98
±0.1
6
10]
3
0.31
±0.0
45
0.98
±0.1
2
0.19±
0.033
2.9±
0.63
0.39
±0.0
83
1.6
±0.
17
83
±9.
9
1.0±
0.28
1.4±
0.22
4
0.01
2±0.
003
>5
0.014
±0.00
3
0.17
±0.0
29
0.14
±0.0
42
1.9
±0.
38
86
±8.
8
0.68
±0.3
4
0.12
±0.0
15
5
0.02
7±0.
003
>5
0.034
±0.00
58
0.65
±0.2
4
0.07
3±0.
003
4.2
±0.
79
10
4±
15
1.8±
0.20
0.58
±0.1
0
6
0.33
±0.0
30
10±
3.9
0.19±
0.044
7.2±
1.2
4.7±
0.34
3.4
±0.
56
21
0±
19
2.8±
0.76
1.8±
0.36
Both CRD of the tandem-repeat galectin-4 were evaluated and
the N-terminal domain did recognize compounds 2-12 and 17
with moderate affinities in the low-medium µM range, which for
all compounds is better than the parent thiodigalactoside.
Reflecting the difference in fine-specificity between the two
galectin-4 domains, the C-terminal domain revealed mid-nM
affinities for several compounds. As observed for galectin-1 and
3, aryltriazoles carrying no or smaller substitutents at the aryl
moiety (2-5 and 12) were identified as the best inhibitors. Again,
this suggests that one or both of the aryl-accommodating sites of
galectin-4C can harbor only smaller structures. The 4-
fluorophenyl derivative 5 stands out as a most potent galectin-
4C with a Kd of 73 nM, which suggests a specific fluorine
interaction and/or an ideal steric fit by the 4-fluoro substituent.
Most likely, efficient inhibition of one domain[13] will be sufficient
to block physiological/biological effects by galectin-4. Galectin-7
binding is enhanced by the 4-aryl-triazolyl groups of 2-9, 12, and
17, while the sterically more demanding compounds 10-11 are
virtually non-binding. Similar observations were made for
galectin-9N and 9C, which both bind several inhibitors with sub
to low µM affinities. In contrast to galectin-4, no clear selectivity
between the two domains of galectin-9 was observed.
7
1.4±
0.0.7
0
1.5±
0.44
0.38±
0.039
3.8±
0.35
4.9±
0.87
2.0
±0.
22
20
0±
11
0.69
±0.0
17
0.78
±0.2
0
8
0.45
±0.0
39
1.6±
0.24
0.23±
0.036
15±3
.0
1.5±
0.27
7.8
±1.
9
>5
00
3.7±
1.0
4.2±
1.2
9
1.2±
0.19
1.4±
0.29
0.25±
0.043
7.9±
2.5
8.6±
1.5
9.1
±3.
6
>5
00
11±0
.64
1.8±
0.14
10
11
12
17
110±
17
>50
0
770±
8.4
>10
00
650±
39
>10
00
>1
00
0
>10
00
>10
00
84[10]
32±
9.5
0.36[1
>50
0
>500
>50
0
44
0±
14
>50
0
240±
11
0]
<0.0
10
>5
nd
0.065
±0.09
3.8±
0.57
0.19
±0.0
35
2.9
±0.
49
12
0±
6.2
2.2±
0.20
0.46
±0.0
37
nd[a]
0.98±
0.023
nd
nd
nd
12
0±
1.8
0.31
±0.0
06
nd
Overall, the 4-aryl-triazolyl thiodigalactosides 2-12 and 17
delivered
inhibitors
significantly
more
potent
than
24[8a]
340
±19
49[8a]
410±
21
980±
70
160
61[
38[8a]
42±1
.1
thiodigalactoside itself against galectin-1, 2, 3, 4N, 4C, 7, 9N,
and 9C and more potent than the corresponding galactoside
monosaccharide derivative against galectin-3, 7, and 9N (c.f. e.g.
the monosaccharide corresponding to 2 show Kd of 150, 1700,
and 1300 µM, respectively, against these four galectins[7a,8d]). In
particular, galectin-1 and 3 were well inhibited with several
compounds showing low nM affinity. The inhibition potency
against galectin-3 even surpassed our earlier described
T
[8a]
8a]
D
G[
b]
[a] Not determined. [b] Thiodigalactoside.
corresponding
4-amido-triazolyl-[8d]
and
4-aryl-triazolyl-
thiodigalactoside[10] derivatives. Several compounds indeed
possess a clear selectivity for these two galectins, while the
In stark contrast, galectin-2 was inhibited with only micromolar
Kd values by any of 2-12 and 17. The best compound, the 2-

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
selectivity between them is limited except for the thienyl
derivative 12 and the phenoxy derivative 11 that displayed
moderate selectivity for galectin-1 and 3, respectively.
Structural studies
The X-ray structures were determined of three selected high-
affinity compounds, the 3- and 4-fluophenyl derivatives 4 and 5
and the thienyl derivative 12, in complex with galectin-3. Initial
refinements of X-ray diffraction data (1.5–1.6 Å resolution,
Supplementary Table S1) produced clear difference electron
density within the galectin binding sites revealing the bound
ligands 4, 5, and 12 (Figure 1). In all cases, electron density is
clearly evident for the thiodigalactoside core of all three ligand 4,
5, and 12 (except for the solvent orientated C6 hydroxyl), and for
both triazole rings. The thiodigalactoside core of each ligand 4, 5,
and 12 is in an identical binding mode to that observed in the
previously reported galectin-3-thiodigalactoside complex[14] and
forms identical protein–ligand interactions, confirming that they
do not act as divalent ligands. Electron density is also defined
for the aromatic rings extending from one of the triazole C4
atoms of 4, 5, and 12 towards the Arg144 side chain. The
second thiophene or fluorophenyl rings of the ligands 4, 5, and
12, positioned above the salt bridge between Glu165 and
Arg186, are less clearly defined by the initial difference electron
density maps compared to the rest of the ligand (upper region of
the ligands in Figure 1a–c). The position of the thiophene ring of
12 and the 3-fluorophenyl ring of 4 is evident in the region near
Glu165 and Arg186 in difference electron density maps when
scaled to 2.5 σ (calculated prior to addition of the ligand to the
model) and refinements with the ligand included in the model
show the thiophene and 3-fluorophenyl rings defined by
2mFo−DFc electron density when scaled at 0.7 σ. Additional
weak 2mFo−DFc electron density appears near Glu165 and
Arg186 of 4 after refinement, indicating a possible alternate
conformation for the ring, however, the electron density is not
clear enough to confidently model two alternate conformations
for the ligand.
The 4-fluorophenyl ring of 5 near the Glu165–Arg186 salt bridge
is poorly defined by the initial difference electron density.
Refinement of the model with the ligand 5 in place, but excluding
the 4-fluorophenyl ring near Glu165–Arg186, results in additional
difference electron density that indicates the general location of
the ring, and refinements with the ring included in the model
resulted in weak 2mFo−DFc electron density (0.7 σ) that supports
the location of the ring. However, there is clearly a higher degree
of disorder for this part of the ligand. This may initially appear
counter-intuitive as known ligands with aromatic groups near the
Glu165–Arg186 salt bridge region of galectin-3 have shown
Figure 1. Difference electron density within the galectin-3 CRD binding sites
showing bound A) 12, B) 4, and C) 5. Difference electron density calculated
from refinement with the ligand (stick representation) omitted from the model
(|Fo|-|Fc| αcalc
grey solvent-accessible surface.
The triazole ring of compounds 4, 5, and 12 located above
His158 is orientated with the nitrogens positioned towards
enhanced
affinities
(for
example
the
diamido-
8b]
thiodigalactosides[8a,
and aromatic lactose 2-O-esters[15]).
Trp181, which allows for formation of
a water-mediated
However, the interaction involves face-to-face stacking between
the aryl-triazoles onto an extended surface of the π-system of
the Glu165–Arg186 ion-pair, which could allow for the aryl-
triazole to position over different segments of the large π-system
with retained interaction free energies.
hydrogen bond between the N2 of the triazole and the nitrogen
atom of Trp181 (Figure 2 a-c). The triazole ring of compounds 4,
5, and 12 located near the in-plane Glu165-Arg186 salt bridge
shows two alternate conformations that stack face-to-face to the
in-plane Glu165-Arg186 salt bridge. In the 12 and 5 complexes,
the triazole nitrogens are directed towards Glu165, while in the

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
Figure 3. Superimposed view of the galectin-3 CRD binding site in the region
of Arg144 for the complex with 4, 5, and 12 (yellow), the 3’-(2,3,5,6-tetra-
fluoro-4-methoxybenzamido) LacNAc derivative (red, PDB ID: 1KJR), and
lactose (blue, PDB ID: 3ZSJ).
complex with 4 the ring is flipped with the triazole nitrogens
close to Arg186. The orientation of one of the triazole rings in 12
and 5 results in contacts with both Glu184 and Arg186, while in
4 the triazole is in contact with Arg186 only.
In all three galectin-3 complexes with 4, 5, and 12, the ligands
induces a conformational change in Arg144 (Figure 3) similar to
that reported for the galectin-3 CRD structure in complex with 3’-
(2,3,5,6-tetra-fluoro-4-methoxybenzamido)-LacNAc
derivative
(PDB ID: 1KJR).[7c] One of the terminal aromatic rings of the
ligands 4, 5, and 12 fits into a pocket that is exposed by the
Arg144 conformational change and move away from the protein
surface and forms a face-to-face stacking interaction with
Arg144 in a similar manner as observed for the corresponding
complex with 3’-(2,3,5,6-tetra-fluoro-4-methoxybenzamido)-
LacNAc derivative. However, in the structures of the complexes
with 4, 5, and 12, the additional length granted by the triazole
linker allows the terminal aromatic rings to extend deeper into
the pocket exposed by the Arg144 move, which allows for the
formation of an additional contact with Ala146 that is not
observed in the 3’-(2,3,5,6-tetra-fluoro-4-methoxybenzamido)-
LacNAc complex. Additionally, although the conformational
change of Arg144 in the complexes with 4, 5, and 12 is overall
similar to that earlier observed for 3’-(2,3,5,6-tetra-fluoro-4-
methoxybenzamido)-LacNAc derivative,[7c] small differences are
apparent. The Arg144 has moved in the complexes with 4, 5,
and 12 compared to the complex with the 3’-(2,3,5,6-tetra-fluoro-
4-methoxybenzamido)-LacNAc derivative (1.5–2.0 Å ζ-carbon
to ζ-carbon distance) such that the guanidino group maintains its
position directly above the aromatic ring of the ligand (Figure 3).
One thiophene ring of 12 is orientated to deeply bury the sulfur
atom in the pocket exposed by the Arg144 move, as is the
fluorine atom in the 3-fluorophenyl ring of 4. The SAD LLG map
calculated for the 12 complex confirms the orientation of the
thiophene ring showing a clear peak positioned at the location of
the sulfur atom within the pocket near Arg144. The fluorine of 4
Figure 2. Galectin-3 CRD binding site interactions with A) 12, B) 4, and C) 5.
H-bond interactions between ligand (yellow bonds) and protein/water (green
bonds) are shown as dashed lines.

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
below Arg144 is situated at a distance of 3.9Å and 3.4Å and at
angles of 155° and 147° from the backbone carbonyls of Arg144
and Ile145, respectively, which suggests the formation of two
orthogonal multipolar interactions.[16] The fluorine atom of the 4-
fluorophenyl ring in 5 is directed towards Gly238 and Ser237
and makes contact with the α-carbon of Gly238 and is also
positioned well for forming an orthogonal dipolar interaction with
the Ser237 carbonyl (distance 3.6Å and angle 140°).
Furthermore, the guanidinium ion of arginine side chains has
been proposed to be highly fluorophilic, as fluorine atoms of
The R144K mutant binds 4 only about 3-fold less well than wild-
type, which suggests that the lysine side-chain can, as an
arginine side-chain, form cation-π interactions. However, in
contrast to the arginine guanidine group, the lysine amino group
obviously lacks a π-system and π-stacking capability may be the
reason that the interaction with the 3-fluorophenyl group of 4 is
possibly slightly less productive. The R186S mutant shows a
large drop in affinity for 4, clearly revealing that a 3-fluorophenyl
stacking interaction onto Arg186 is an important contributor to
the high affinity of 4 for galectin-3. The Arg186 side-chain
guanidinim ion is, in contrast to the Arg144 side-chain, involved
in an extensive network of in-plane bi-furcated ion-pairs (Figure
2), which form an extended π-system surface onto which a 3-
fluorophenyl stacks in analogy with e.g. the acetamido group of
N-acetyl-lactosamine[7c] and aromatic rings of 2-O-benzoyl
lactose derivatives.[15] In the mutant R186S this extended π-
system of bi-furcated ion-pairs is interrupted and the 3-
fluorophenyl cannot form a beneficial stacking interaction.
Instead, a poorly solvated cavity with poor complementarity to
the 3-fluorophenyl group of 4 is present. Some binding affinity is
regained in the R186K mutant as compared to the R186S
mutant, which is presumably due the capability of the lysine
side-chain to at least partly substitute and stabilize the Arg186
side-chain’s key π-system-forming ion-pairing with the two
surrounding Glu165 and Glu184 residues, as well as providing
similar surface complementarity to the 3-fluorophenyl group of 4.
In short, the high affinity of the 4-aryl-triazolyl thiodigalactosides,
such as 4, 5, and 12, for galectin-3 can be hypothesized,
according to X-ray structural analysis of galectin-3 complexes
and galectin-3 mutant studies, originating from several factors.
First, ideal surface complementarity between the proteins and
ligands (Figure 1) are, not unexpectedly, critical as this
maximizes dispersion forces and presumably also beneficial
desolvation effects. Stacking between galectin arginine side
chain guanidinium functionalities and ligand phenyl-triazole
moieties are probably important, as are fluorine orthogonal
dipolar interactions[16c] with backbone carbonyls. Hence, while
the core thiodigalactoside disaccharide mimics natural
disaccharide ligand fragments (e.g. lactose and LacNAc) in
terms of affinity contributions and structure, the appended non-
carbohydrate aryl-triazole moieties engage in galectin-ligand
interactions not seen in natural lectin-ligand complexes (i.e.
predominantly hydrogen bonding and CH-π interactions),
resulting in drastic affinity enhancements and enhanced
selectivities.
fluorinated pharmaceuticals have been observed to be close to
17]
guanidinium moieties in proteins.[16a,
Finally, fluorination
typically results in increased lipophilicity[16a,
and fluorinated
17]
hydrocarbons are in general poorly solvated in water,[18] which
would support a conclusion that burying fluorinated lipophilic
ligand parts is important for achieving high affinity of 4 and 5 for
galectin-3. The equivalent of Arg144 is absent in some
galectins[19] and consequently targeting ligand interactions to this
region and engaging Arg144 through cation-π interactions is
proposed as a means of enhancing galectin binding selectivity.
Galectin-3 mutant studies
The X-ray structures of galectin-3 revealed that the aryl-triazoles
of 4, 5, and 12 stacked face-to-face onto two (Arg144 and
Arg186) arginine guanidinium groups. In case of galectin-3, the
two 3-fluorophenyl moieties of 4 have different stacking modes
with the two Arg144 and Arg186: One 3-fluorophenyl moiety is
stacked on top of Arg186 guanidinium group, while the other 3-
fluorophenyl moiety is inserted between the protein surface
(backbone) and Arg144 guanidinium group (Figure 1b, 2b, and
3). In order to obtain further understanding about the nature of
the aryl-triazole arginine stacking interactions, we determined
the affinity of 4 for four galectin-3 mutants, R144K, R144S,
R186K, and R186S (Table 2). The R144S and R186S mutants
were chosen because the side-chain is removed without
introducing a very non-polar surface and the R144K and R186K
mutants were chosen because the cationic nature of the side-
chain is retained while the planar π-system of the guanidino
group is removed. The effect of the R144S mutant is minimal,
which suggests that the stacking of Arg144 onto the 3-
fluorophenyl group of 4 does not contribute significantly to the
free energy of binding, while the surface complementarity and
interactions of the 3-fluorophenyl group with the rest of the
protein surface remains essentially unchanged.
Having discovered low-nM galectin antagonists, an important
question of their efficiency for antagonizing galectin-glycoprotein
interactions in biological systems was addressed with compound
4 in two models. First, an in vitro cell assay was developed with
the goal of gaining new knowledge about galectin-3 putative
intracellular glycan-binding activities and possible effects in cells
challenged with vesicle-damaging agents. Second, to achieve
further understanding of, as well as quantifying, the effects of
antagonizing galectin-3 in an in vivo mouse model of bleomycin-
induced idiopathic pulmonary fibrosis.[6b]
Table 2. Dissociation constants (µM) for
4 against galectin-3 mutants
determined with a competitive fluorescence anisotropy assay.[12]
wt
R144K
R144S
R186K
R186S
0.014±0.003
0.041±0.0045
0.017±0.0032
0.54±0.039
1.0±0.12

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
Intracellular inhibition by galectin-3 antagonist in an
amitriptyline-induced vesicle damage assay
Galectin accumulation around damaged vesicles in response to
challenge by bacteria or chemical agents has been
demonstrated in several studies and the formation of galectin-
3[20] or galectin-8[21] puncta have been proposed as a novel
marker for vesicular insult, regardless of the insult being of a
21]
bacterial[20b,
or chemical origin[20a]. Galectin-3 accumulation
around damaged vesicles has, in addition, been shown to
depend on glycan-binding, either by knock-down of certain
glycosyltransferases[20b] or via knock-in of a galectin-3 mutant
(R186S) with severely reduced affinity for endogenous
glycans[20a]
.
Antagonizing effects on such glycan binding-
dependent galectin-3 events on damaged vesicles[20b] could
provide qualitative information on intracellular availability and
activity of antagonists, such as compound 4. Cationic
amphiphilic drugs, including the tricyclic antidepressant
amitriptyline, induce phospholipidosis and are speculated to
accumulate in acidic lysosomes, and induce vesicle damage in
tumor cell lines.[20a, 22] We postulated that treatment of cells with
amitriptyline would induce formation of galectin-3 puncta in a
similar fashion as other vesicular damaging agents, such as
glycyl-l-phenylalanine 2-naphthylamide[21] and l-leucyl-l-leucine
methyl ester[23]. Amitriptyline has the advantage of being more
stable under the experimental conditions used and does not
degrade in solution as e.g. glycyl-l-phenylalanine 2-
naphthylamide. Furthermore, amitriptyline does not require the
use of DMSO as co-solvent for solubilize the more commonly
used peptidic vesicular damaging agents. Indeed, treating breast
carcinoma MCF-7 cells with amitriptyline resulted in distinct
accumulation of galectin-3 into vesicle-associated puncta,
hypothetically within galectin-3:glycoprotein lattices, in a dose-
dependent manner (Figure 4a and b). Co-treatment with 10 µM
compound 4 and 10 µM or 50 µM amitriptyline resulted in a
significant reduction in the number of galectin-3 dots compared
to amitriptyline treatment alone (Figure 4a and b), which strongly
supports that compound 4 can act as an intracellular antagonist
for galectin-3 in cell culture systems. The experimental
concentration of 4 was selected to achieve a significant effect
and possibly reflects a relatively slow cellular uptake and
intracellular concentration increase of 4 sufficient to block
intracellular galectin-3.
Pharmacological intervention in a bleomycin-induced lung
fibrosis mouse model
Galectin-3 has been shown to promote both macrophage M2
polarization[5b] and myofibroblast activation,[6b] i.e. in two key pro-
fibrotic cell types. In the case of macrophage M2 polarization,
galectin-3 association with CD98 on the macrophage cell
surface, presumably within lattices, was shown to be a plausible
molecular mechanism for regulating M2 activation via
phosphatidylinositol 3-kinase (PI3K) activation.[5b] Analogously,
transforming growth factor-ß (TGF-ß) receptor II has been
shown to bind galectin-3 on cell surfaces, which was suggested
to be a critical molecular mechanism for inducing myofibroblast
activation.[6b] Furthermore, in vivo an intratracheal single-dose of
the galectin-3 antagonist 4 (10 µg per mouse, 500 µg/kg) was
Figure 4. Inhibition galectin-3 accumulation around amitriptyline (AMI)-
damaged vesicles in MCF-7 cells. Cells were treated with combinations of 10
µM compound 4 and 10 or 50 µM amitriptyline for 24 hours, control cells were
treated with 0.1% v/v DMSO. A) Galectin-3 staining was visualized with anti-
rat Alexa Fluor® 594 (red), whereas Hoechst (blue) was used to stain the
nuclei. The immunofluorescence pictures displayed are representative for
each treatment. Scale bars are equivalent to 20 µm. Small square inserts
show which areas is magnified in each large square insert. B) The number of
galectin-3 dots were counted manually using ImageJ in four different images
for each experimental condition, and given as mean ± SEM. Each data set
represents ~250 cells. **P<0.01, ***P < 0.001, Student’s t-test.

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
Figure 6. Effects of pirfenidone and compound 4 on BAL (bronchoalveolar
lavage) fluid parameters in bleomycin-induced lung fibrosis in mice. Total
protein measured by BCA reagent, MCP-1, and galectin-3 in BAL fluid and
serum were measured by ELISA. Results represent the mean ± SEM of n=8
mice per group. (* P<0.05, statistically different from bleomycin control).
demonstrated to display an anti-fibrotic effect in a bleomycin-
induced lung fibrosis mouse model.[6b] Hence, compound 4 can
be hypothesized to possess dual anti-fibrotic effects by
disrupting lattices with CD98 on M2 macrophages and with TGF-
ß-RII on myofibroblasts and associated pro-fibrotic signaling.
However, the single-dose experiment left questions unanswered
concerning the in vivo dose-response efficacy of compound 4
and how this compared to alternative anti-fibrotic agents. Hence,
we conducted
a
dose-response study of therapeutic
administration of compound 4 in this model, in comparison with
pirfenidone, one of only two recently approved drugs for treating
idiopathic pulmonary fibrosis (IPF). Mice (n=8) received
bleomycin sulphate (1.65 mg/kg intratracheally), resulting in
inflammation and subsequent fibrosis development, followed by
either 200 mg/kg pirfenidone twice daily orally from days 18-24
or compound 4 at 500, 150, 50 or 5 µg/kg intratracheally as a
single administration every second day (days 18, 20, 22 and 24).
Lung collagen content and histopathology was determined on
Figure 5. Effects of pirfenidone and 4 on bleomycin-induced lung fibrosis in
mice. A) Total lung collagen, B) Histological inflammatory score, C)
Histological fibrosis score. Results represent the mean ± SEM of n=8 mice per
group. (* P<0.03, **P<0.05, ***P<0.01 statistically different from bleomycin
control). D) Representative Masson’s trichrome stained sections of mouse
lung from uninjured saline control, bleomycin control and bleomycin treated
with oral pirfenidone (200 mg/kg) or intratracheal 4 (500 µg/kg).
day 26.
Pirfenidone (200mg/kg) significantly reduced
bleomycin-induced collagen accumulation from 670±77 to
375±53 μg collagen/lobe (P<0.01), as did 500 and 150 µg/kg of
compound 4 (355±46 and 546±22 μg collagen/lobe P<0.01,
P<0.03, respectively) (Figure 5). In addition, compound 4 at 500
and 150 µg/kg doses and pirfenidone significantly decreased the
fibrosis score. Hence, when delivered directly into the lung,
compound 4 achieves efficacy at much lower concentrations
compared to orally delivered pirfenidone. The lower dose
needed with administration of 4 could be due to the, not
unexpectedly, improved lung targeting by intratracheal
administration in combination with the high affinity of 4 for the
target galectin-3 protein shown to be a key regulator of fibrosis

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
biochemistry. Lung availability of an orally administered
compound is likely to be lower than that of an intratracheally
administered compound, which may at least partly explain the
need for a higher dose of oral pirfenidone to achieve the same
efficacy as intratracheal 4. In addition, compound 4 did not
reduce protein in the BAL (bronchoalveolar lavage) fluid – an
indication of vascular leakage – but both pirfenidone and
strategy, although PK-ADMET properties obviously have to be
improved for intracellular/systemic availability and therapeutic
applications.
Furthermore, intratracheally delivered compound 4 attenuated
bleomycin-induced lung fibrosis in a mouse model in a dose-
dependent manner and possessed efficacy at significantly lower
doses than the approved oral anti-fibrotic pirfenidone and thus
compared favorably with pirfenidone. This may further support a
dual molecular mechanistic hypothesis in which galectin-3-
promoted macrophage and myofibroblast activation results in
sustained pro-fibrotic cell signaling and scar formation.
Finally, five-membered aromatic heterocycles are common
structural elements in many drugs and 1,2,3-triazoles in
particular are readily synthesized, which render the compounds
herein as promising leads for the development of novel galectin-
targeting therapeutics that disrupt cellular signal-sustaining
galectin-3 lattices as well as highly valuable tools for studying
galectin biology and molecular mechanisms.
compound
4 reduced MCP-1 (monocyte chemoattractant
protein-1) levels (Figure 6). Compound 4 did not produce a
clear significant decrease in galectin-3 levels in BAL fluid or
serum. The absence of a significant decrease of galectin-3 in
BAL fluid is likely due to the fact that BAL fluid samples are from
whole lung and not only the fibrotic area. As non-fibrotic tissue
has a background expression of galectin-3 this will influence the
total galectin-3 levels in BAL fluid samples, therefore BAL fluid
galectin-3 analysis may be underestimating the actual
concentration of galectin-3 in the diseased areas of the lung.
Conclusions
Experimental Section
Highly potent galectin-1 and 3 antagonists were discovered
through synthesis optimization, and structural analysis of double
C3 aryl-triazolyl-substituted thiodigalactosides. Low nM-affinities
were reached for galectin-1 and 3 and some compounds
displayed selectivity for individual galectins. Structural and
mutational studies showed evidence that the exceptional affinity
enhancement originated largely from the aryl-triazole moieties
forming stacking interactions with protein π-systems (arginine
side chains unpaired or ion-paired with glutamate or aspartate
carboxylates) and in some cases fluorine-derived orthogonal
multipolar interactions that endogenous glycoconjugate glycans
do not form. The nature of the aryl-triazole moieties has a
significant influence over galectin sub-type selectivities, which
could also be explained by small, but significant, differences
revealed in the structural studies. Overall, the results
corroborate the promising strategy for discovery of high-affinity
and selective lectin antagonists by exploring non-carbohydrate
structural elements forming interactions that glycoside fragments
of endogenous glycoconjugate ligands do not form with lectins.
Hence, drug development targeting lectins may not necessarily
involve a strategy of multimerizing ligands and antagonists to
achieve sufficient affinities and the major challenges concerning
pharmacokinetics, bioavailability, and toxicity/immunogenicity
associated with multivalent antagonists may be avoided.
One antagonist (4) was evidenced to have intracellular
availability and activity as it blocked amitriptyline-induced vesicle
damage in breast carcinoma MCF-7 cells. While it still remains
to be answered which glycoprotein binding partner is involved in
the galectin-3 accumulation on damaged vesicles, the lysosome
associated membrane proteins LAMP-1 and LAMP-2 may be
candidates for this role as they have been shown to be galectin-
3 ligands on the surface of tumor cells[24] and are thus possible
candidate glycoprotein ligands in our model. Importantly, these
observations suggest that intracellular galectin-3 glycoprotein-
binding events occur and may be biologically relevant. Targeting
such interactions with synthetic antagonists may be a viable
Expression constructs, expression, and purification of recombinant
galectins
Human galectin-1,[25] galectin-2,[26] galectin-3,[27] galectin-4N,[12a] galectin-
4C,[12a] galectin-8N,[13b] and mouse galectin-7[28] were expressed and
purified as earlier described. Human galectin-9N and galectin-9C were
produced in E. coli BL21Star (DE3) cells (Invitrogen) and purified by
affinity chromatography on lactosyl-Sepharose essentially as described
for galectin-8.[13b] DNA encoding the genes of human galectin-9N and
galectin-9C were cloned into the pET-32 Ek/LIC vector (Novagen,
Madison, WI) according to the manufacturer’s instructions. Briefly,
I.M.A.G.E. clone 2208156 (ATCC) was used as template together with
the following polymerase chain reaction (PCR) primers. The vector used
for galectin-9N encoded the N-terminal 170 amino acids of galectin-9 and
thioredoxin with the primers forward: 5`- GAC GAC GAC AAG ATG ATG
GGT TCA GCG GTT CCC AGG-3´, forward 2: 5´- GAG GAG AAG CCC
GGT TCA GGA AAC AGA CAG GCT GGG AGA ACGG C-3´, and
reverse: 5´- GAG GAG AAG CCC GGT GCC GCC TAT GTC TGC ACA
TGG G-3´. The vector used for galectin-9C encoded the C-terminal
amino acids 205-355 of galectin-9 and thioredoxin with the primers
forward: 5´-GAC GAC GAC AAG ATG GGA CAG ATG TTC TCT ACT
CCC-3´ and reverse: 5´-GAG GAG AAG CCC GGT GCG GCC TAT GTC
TGC ACA TGG G-3´. The bacteria were grown (37°C, 200 rpm) in LB
(Luria-Bertani) medium with ampicillin (1mg/l) overnight, followed by
induction with 1mM isopropyl thio-β-D-galactoside (IPTG) for 4 h (29°C,
200 rpm). The culture was centrifuged (15 min, 5000 rpm, 4°C) and the
pellet was dissolved in 50 ml MEPBS (phosphate buffered saline with 2
mM EDTA and 4 mM β-mercaptoethanol) and sonicated 10-20 x 30 s on
ice. The sonicated bacteria were centrifuged (30 min, 12000 rpm, 4°C)
and the supernatant was submitted to affinity chromatography using a
lactosyl-Sepharose column washed with MEPBS and a pre-elution of 7,5
mM lactose. The bound proteins were eluted with Lac-MEPBS (MEPBS
with 150 mM lactose) as elution buffer. Removal of lactose was done by
chromatography on a PD-10 column (Amersham Biosciences) and with
repeated ultrafiltration using Centriprep (Amicon).
Competitive fluorescence polarization experiments determining
galectin affinities

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
Fluorescence polarization experiments were performed on a POLARStar
plate reader with software FLUOstar Galaxy software or a PheraStarFS
plate reader with software PHERAstar Mars version 2.10 R3 (BMG,
Offenburg, Germany) and fluorescence anisotropy of fluorescein tagged
carbonylaminomethyl)-1H-1,2,3-triazol-1-yl]-3’-(3,5-dimethoxy-
benzamido)-1,1’-sulfanediyl-di-β-D-galactopyranoside at 0.10 µM.
Galectin-3 R144S affinities: Experiments were done at 20°C with
galectin-3 R144S at 0.30 µM and the fluorescent probe 2-(fluorescein-
5/6-yl-carbonyl)-aminoethyl 2-acetamido-2-deoxy--D-galactopyranosyl-
(1–3)-[-L-fucopyranosyl-(1–2)]-β-D-galactopyranosyl-(1–4)-β-D-
glucopyranoside at 0.02 µM.
probes measured with excitation at 485 nm and emission at 520 nm. Kd
29]
values were determined in PBS as previously described[12,
with
specific conditions for each galectin as described below. Compounds 3-
10 were dissolved in neat DMSO at 100 mM and diluted in PBS to 3-6
different concentrations to be tested in duplicates. Kd average and SEM
were calculated from 4 to 25 single point measurements showing
between 30-70% inhibition.
Galectin-3 R144K affinities: Experiments were done at 20°C with
galectin-3 R144K at 0.40 µM and the fluorescent probe 2-(fluorescein-
5/6-yl-carbonyl)-aminoethyl 2-acetamido-2-deoxy--D-galactopyranosyl-
(1–3)-[-L-fucopyranosyl-(1–2)]-β-D-galactopyranosyl-(1–4)-β-D-
glucopyranoside at 0.02 µM.
Galectin-1 affinities: Experiments were done at 20°C with galectin-1 at
0.50 µM and the fluorescent probe 3,3’-dideoxy-3-[4-(fluorescein-5-yl-
carbonylaminomethyl)-1H-1,2,3-triazol-1-yl]-3’-(3,5-dimethoxy-
benzamido)-1,1’-sulfanediyl-di-β-D-galactopyranoside[25] at 0.10 µM.
Galectin-3 R186S affinities: Experiments were done at 20°C with
galectin-3 R186S at 3.50 µM and the fluorescent probe 2-(fluorescein-
5/6-yl-carbonyl)-aminoethyl 2-acetamido-2-deoxy--D-galactopyranosyl-
(1–3)-[-L-fucopyranosyl-(1–2)]-β-D-galactopyranosyl-(1–4)-β-D-
glucopyranoside at 0.1 µM.
Galectin-2 affinities: Experiments were done at 20°C with galectin-2 at
10 µM and the fluorescent probe 3,3’-dideoxy-3-[4-(fluorescein-5-yl-
carbonylaminomethyl)-1H-1,2,3-triazol-1-yl]-3’-(3,5-dimethoxy-
benzamido)-1,1’-sulfanediyl-di-β-D-galactopyranoside at 0.10 µM.
Galectin-3 R186K affinities: Experiments were done at 20°C with
galectin-3 R186K at 0.90 µM and the fluorescent probe -D-
galactopyranosyl(1—4)-2- acetamido-2-deoxy--D-glucopyranosyl(1—3)-
-D-galactopyranosyl(1—4)-(N1-fluorescein-5-yl-
Galectin-3 affinities: Experiments were done at 20°C with galectin-3 at
0.20 µM and the fluorescent probe 3,3’-dideoxy-3-[4-(fluorescein-5-yl-
carbonylaminomethyl)-1H-1,2,3-triazol-1-yl]-3’-(3,5-
dimethoxybenzamido)-1,1’-sulfanediyl-di-β-D-galactopyranoside at 0.02
µM or with galectin-3 at 1.0 µM and 2-(fluorescein-5/6-yl-carbonyl)-
carbonylaminomethylcarbonyl)--D-glucopyranosylamine at 0.1 µM.
aminoethyl
2-acetamido-2-deoxy--D-galactopyranosyl-(1–3)-[-L-
Crystallization
fucopyranosyl-(1-2)]-β-D-galactopyranosyl-(1–4)-β-D-glucopyranoside[12]
at 0.10 µM.
Compounds 4, 5, and 12 were prepared in the galectin-3 crystallization
conditions by initially solubilizing in 55% w/v polyethylene glycol (PEG
6000), before addition of other crystallisation reagents to give a final
concentration of 20 mM of 4, 5, and 12 in the galectin-3 crystallisation
condition (31% w/v PEG 6000, 100 mM Tris-HC pH 7.5, 100 mM MgCl2
for galectin-3). Galectin-3-CRD lactose or galactose co-crystals
(prepared as previously described[31]) were soaked for 2–8 days in drop
containing a 1:1 ratio of the ligand-containing crystallisation condition and
20 mg/mL human galectin-3-CRD in 10 mM Tris-HCl pH 7.5 (pre-
equilibrated co-crystallisation drops that had not produced crystals).
Galectin-4N affinities: Experiments were done at 20°C with galectin-4N
at 3.0 µM and the fluorescent probe 3,3’-dideoxy-3-[4-(fluorescein-5-yl-
carbonylaminomethyl)-1H-1,2,3-triazol-1-yl]-3’-(3,5-dimethoxy-
benzamido)-1,1’-sulfanediyl-di-β-D-galactopyranoside at 0.10 µM.
Galectin-4C affinities: Experiments were done at 20°C with galectin-4C
at 0.50 µM and the fluorescent probe 2-(fluorescein-5/6-yl-carbonyl)-
aminoethyl
2-acetamido-2-deoxy--D-galactopyranosyl-(1–3)-[-L-
fucopyranosyl-(1-2)]-β-D-galactopyranosyl-(1–4)-β-D-glucopyranoside at
0.1 µM.
X-ray diffraction analysis and structure determination
Galectin-7 affinities: Experiments were done at 4°C with galectin-7 at
2.00 µM and the fluorescent probe -D-galactopyranosyl(1—4)-2-
acetamido-2-deoxy--D-glucopyranosyl(1—3)--D-galactopyranosyl(1—
4)-(N1-fluorescein-5-yl-carbonylaminomethylcarbonyl)--D-
glucopyranosylamine[30] at 0.1 µM.
X-ray diffraction data sets were collected at room temperature from
human galectin-3-CRD crystals mounted in 0.7 mm quartz capillaries on
a ProteumR (Bruker AXS, Madison, WI, USA) diffractometer with a
MacScience M06XCE rotating-anode generator (wavelength 1.5418 Å)
equipped with a SMART6000 CCD detector. X-ray diffraction data were
integrated using SAINT (Bruker AXS, Madison, WI, USA) and scaled and
merged using SCALA[32] within the CCP4 suite of crystallographic
software.[33] Structures were solved by initial rigid body refinement using
a previously published galectin-3-CRD structure (1A3K),[34] with ligand
and waters removed, as the initial model. TLS and restrained refinement
was performed using REFMAC5.[35] Anomalously scattering elements
were identified using single-wavelength anomalous dispersion log-
Galectin-8N affinities: Experiments were done at 20°C with galectin-8N
at 0.40 µM and the fluorescent probe 2-(fluorescein-5-yl-
carbonylamino)ethyl -D-galactopyranosyl(1—4)-2-acetamido-2-deoxy--
D-glucopyranosyl(1—3)--D-galactopyranosyl(1—4)--D-
glucopyranoside[13b] at 0.1 µM.
[36]
Galectin-9N affinities: Experiments were done at 20°C with galectin-9N
at 1.0 µM and the fluorescent probe 2-(fluorescein-5-yl-
carbonylamino)ethyl -D-galactopyranosyl(1—4)-2-acetamido-2-deoxy--
D-glucopyranosyl(1—3)--D-galactopyranosyl(1—4)--D-
glucopyranoside at 0.1 µM.
likelihood gradient maps (SAD LLG maps); calculated using Phaser
(in experimental phasing mode within CCP4) in the ‘SAD with molecular
replacement partial structure’ mode with purely anomalous scatterers
and zero LLG-map completion cycles using the current model and F+
and F− structure factor amplitudes as input. Visualization of electron
density and model building was performed using Coot.[37] Ligand
geometry topologies for refinement were initially created by REFMAC5
within CCP4 (LIBCHECK) or using the Dundee PRODRG2 Server.[38] In
most cases minor to moderate manual editing of the automatically
Galectin-9C affinities: Experiments were done at 20°C with galectin-9C
at 2.0 µM and the fluorescent probe 3,3’-dideoxy-3-[4-(fluorescein-5-yl-

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
generated topologies was performed to ensure correct atom and bond
types. Model validation and analysis was performed using MolProbity.[39]
Figures were created using the CCP4 molecular-graphics project
(CCP4MG).[40]
Bleomycin-induced fibrosis
Bleomycin was purchased from Apollo Scientific and reconstituted in
sterile saline at a concentration of 0.66 mg/mL and aliquots were stored
at -20oC. Pirfenidone was purchased from Tocris Biochemicals and was
Accession codes
dissolved in 0.5% carboxymethyl cellulose (Sigma Aldrich) to a
concentration of 20 mg/mL. Compound 4 was dissolved in 100% DMSO
at a concentration of 10 mg/mL and aliquots stored at -20oC. For each
day, compound 4 for instillation was diluted in sterile saline to give a final
concentration of DMSO in the instillate of 2%. Female C57/Bl6 mice 10
weeks of age were purchased from Charles River and were maintained in
12-hour light/12-hour dark cycles with free access to food and water. All
procedures were performed in accordance with Home Office guidelines
(Animals (Scientific Procedures) Act 1986).
PDB: The atomic coordinates and structure factors of galectin-3 in
complex with 4, 5, and 12 have been deposited with accession codes
5E89, 5E8A, and 5E88, respectively.
Site-directed Mutagenesis
Mutants of human galectin-3 were made using the QuickChange® II site-
directed mutagenesis kit (Stratagene, Amsterdam, The Netherlands),
produced in E.coli BL21Star (DE3) cells (Invitrogen, Lidingö, Sweden)
and purified by affinity chromatography on lactosyl-Sepharose as
previously described.[41] Mutagenic primers for PCR were as follows: Gal-
3R186K (AGA→AAA) sense (5´-CTG GGG AAG GGA AGA AAA ACA
GTC GGT TTT CCC-3´) and antisense (5´-GGG AAA ACC GAC TGT
TTT TCT TCC CTT CCC CAG-3´) and Gal-3R144K (AGA→AAA) sense
(5´-GAA GCC CAA TGC AAA CAA AAT TGC TTT AGA TTT CCA AAG
AG-3´) and antisense (5´-CTC TTT GGA AAT CTA AAG CAA TTT TGT
TTG CAT TGG GCT TC-3´). Successful mutagenesis was confirmed by
sequencing by GATC Biotech (Konstanz, Germany) in the forward
direction from the T7 promotor primer and in the reverse direction from
the pET-RP primer. Galectin-3 R144S and R186S were prepared as
earlier reported.[41]
Mice were randomised into
8 treatment groups (n=8) and were
saline. Mice were monitored closely over the next 26 days. Pirfenidone
treated mice received pirfenidone 200 mg/kg by oral gavage twice daily
from days 18-24. Mice treated with compound 4
intratracheally commencing day 18 every 48 hours for a total of 4
administrations. Control mice received vehicle (2% DMSO). Mice were
culled on day 26. The lungs were perfused (via the right ventricle) with 5
ml saline and the lungs lavaged with 3 x 0.8 mL PBS containing 1 mM
EDTA. BAL cells were combined and pelleted and lavage fluid from the
first lavage was snap frozen. The lungs were removed and the entire left
lobe removed and stored at -80oC for analysis of total collagen. Two
upper right lobes were removed and snap frozen and stored at -80oC for
subsequent RNA analysis. The remaining lung was inflated with 10%
formalin and fixed for 24 hours prior to removal into 70% ethanol before
embedding in paraffin wax for histological examination.
Cell culture and immunocytochemistry
MCF-7 cells were maintained in RPMI-1640 (Biochrom, Berlin, Germany)
supplemented with 10% fetal bovine serum (Biochrom, Berlin, Germany),
10 µg/mL insulin (Sigma-Aldrich, Stockholm, Sweden), 100 µg/mL
streptomycin and 100 units/mL penicillin (Hyclone). The cells were kept
in a 37 C humidified incubator supplied with 5% CO2 in air. For the
experiments stock solutions of 4 mM compound 4 in 40% dimethyl
sulfoxide (DMSO) was used, while for amitriptyline (Sigma-Aldrich) stock
solutions of 20 mM was made in sterile water. Both compound 4 and
amitriptyline were serially diluted in RPMI-1640 before treatment of cells,
such that the DMSO concentration did not exceed 0.1% v/v. MCF-7 (105
cells) were seeded onto sterile coverslips (placed in multiwell plates) and
cultured for 24 hours. Cells were then treated with either 10 or 50 µM
amitriptyline either alone or in combination with 10 µM of compound 4 for
24 hours. After fixation with 2% paraformaldehyde in phosphate-buffered
saline (PBS) for 10 minutes, cells were permeabilized using 0.4% v/v
Triton X-100 in PBS for 5 minutes. Non-specific binding was inhibited by
blocking the cells with blocking buffer (1% w/v BSA, 0.1% v/v Tween 20
in PBS) for 10 minutes. Cells were then incubated with rat anti-mouse
galectin-3 antibody (anti-Mac-2[42]) in a humidified chamber for 1 hour at
room temperature. After three washes with PBS, goat anti-rat Alexa
Fluor® 594 (Invitrogen, Carlsbad, USA) was added. Hoechst (10 ng/mL)
was used to stain the nuclei. Cells were visualized by obtaining z-stacks
of high magnification single optical planes using a LSM510 confocal laser
scanning microscope (Carl Zeiss Microscopy GmbH, Oberkochen,
Germany), conjugated with Hamamatsu R6357 (Hamamatsu Photonics
K.K., Hamamatsu, Japan) photomultiplier. Galectin-3 dots were counted
manually using ImageJ 1.47v and the plug-in Cell Counter (Wayne
Rasband, National Institutes of Health, USA). Bar graphs representing
galectin-3 dots/nuclei are expressed as mean values of different image
areas ± SEM. For measuring statistical significance between a pair of
data sets, Student’s t-test (two tailed, unpaired) was employed. P<0.05
was considered to be significant.
Total lung collagen
Frozen left lobes were thawed, weighed and minced finely with scissors
and placed in 5ml of 3mg/ml pepsin in 0.5M acetic acid. Samples were
incubated for 24 hours at 4oC and 0.2 mL of cleared extract was
incubated with 0.8 mL Sircol reagent (Biocolor) for 60 minutes at room
temperature. Collagen was sedimented by centrifugation at 13000 rpm
for 5 minutes and the pellets resuspended in 0.5 mL of 0.5M NaOH.
Samples were examined for absorbance at 560 nm with reference to a
collagen standard curve.
Estimation of vascular leakage
Vascular leakage was determined by measuring total protein in the
lavage fluid by BCA assay (Pierce) using bovine serum albumin as
standard.
Histological lung inflammation and fibrosis score
Fibrosis and histological score was carried out in Masson’s trichrome
stained sections.
Inflammation (peribronchiolar, perivascular, and
alveolar wall thickness) scored in > 5 random fields at magnification X630
using the following system (peribronchiolar and perivascular, 1 = no cells,
2 = <20 cells, 3 = 20 – 100 cells, 4 = > 100 cells; alveolar wall thickness,
1 = no cells, 2 = 2 – 3 cells thick, 3 = 4 – 5 cells thick, 4 = > 5 cells thick).
The combined inflammatory score is the sum of these scores. Fibrosis
score was evaluated as the area of the section positively stained for
collagen (1 = none, 2 = <10%, 3 = < 50%, 4 = > 50%). Only fields where
the majority of the field is composed of alveoli were scored.

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
Determination of galectin-3 levels in BAL and serum by ELISA
Chem. 2011, 19, 3280-3287; h) D. Giguere, M. Bonin, P. Cloutier, R.
Patnam, C. Stpierre, S. Sato, R. Roy, Bioorg. Med. Chem. 2008, 16,
7811-7823; i) D. Giguere, R. Patnam, M.-A. Bellefleur, C. St.-Pierre, S.
Sato, R. Roy, Chem. Commun. 2006, 2379-2381; j) D. Giguere, S. Sato,
C. Stpierre, S. Sirois, R. Roy, Bioorg. Med. Chem. Lett. 2006, 16, 1668-
1672; k) M. F. Marchiori, D. E. P. Souto, L. O. Bortot, J. F. Pereira, L. T.
Kubota, R. D. Cummings, M. Dias-Baruffi, I. Carvalho, V. L. Campo,
Bioorg. Med. Chem. 2015, 23, 3414-3425; l) G. A. Rabinovich, A.
Cumashi, G. A. Bianco, D. Ciavardelli, I. Iurisci, M. D'Egidio, E. Piccolo,
N. Tinari, N. Nifantiev, S. Iacobelli, Glycobiology 2005, 16, 210-220; m)
S. R. Rauthu, T. C. Shiao, S. André, M. C. Miller, É. Madej, K. H. Mayo,
H.-J. Gabius, R. Roy, ChemBioChem 2015, 16, 126-139.
Samples of BAL fluid and serum were assayed for galectin-3 and MCP-1
levels by ELISA (R&D systems)..
Acknowledgements
This work was supported by grants to U.J.N. and H.L. from the
Swedish Research Council (Grants No. 621-2003-4265, 621-
2006-3985,
and
621-2009-5326),
the
programs
[8]
[9]
a) I. Cumpstey, E. Salomonsson, A. Sundin, H. Leffler, U. J. Nilsson,
Chem. Eur. J. 2008, 14, 4233-4245; b) I. Cumpstey, A. Sundin, H.
Leffler, U. J. Nilsson, Angew. Chem. Int. Ed. 2005, 44, 5110-5112; c) T.
Delaine, I. Cumpstey, L. Ingrassia, M. Le Mercier, P. Okechukwu, H.
Leffler, R. Kiss, U. J. Nilsson, J. Med. Chem. 2008, 51, 8109-8114; d) B.
A. Salameh, I. Cumpstey, A. Sundin, H. Leffler, U. J. Nilsson, Bioorg.
Med. Chem. 2010, 18, 5367-5378.
“Glycoconjugates in Biological Systems” and “Chemistry for Life
Sciences” sponsored by the Swedish Strategic Research
Foundation, the foundation “Olle Engkvist Byggmästare”, the
Royal Physiographic Society, Lund, by the European
Community's Seventh Framework Program (FP7-2007-2013)
under grant agreement n° HEALTH-F2-2011-256986–project
acronym PANACREAS, and a project grant awarded by the Knut
and Alice Wallenberg Foundation (KAW 2013.0022). The in vivo
experiments were supported by Galecto Biotech AB, Sweden.
H.B gratefully acknowledges the financial support from the
Cancer Council Queensland, Australia (grants : ID1043716 and
ID1080845). R.J.P. gratefully acknowledges the financial
support from the Dutch Technology Foundation STW, applied
science division of NOW. Barbro Kahl-Knutsson is
M. Van Scherpenzeel, E. E. Moret, L. Ballell, R. M. J. Liskamp, U. J.
Nilsson, H. Leffler, R. J. Pieters, ChemBioChem 2009, 10, 1724-1733.
[10] H. van Hattum, H. M. Branderhorst, E. E. Moret, U. J. Nilsson, H. Leffler,
R. J. Pieters, J. Med. Chem. 2013, 56, 1350-1354.
[11] T. L. Lowary, O. Hindsgaul, Carbohydr. Res. 1994, 251, 33-67.
[12] a) P. Sörme, B. Kahl-Knutsson, M. Huflejt, U. J. Nilsson, H. Leffler, Anal.
Biochem. 2004, 334, 36-47; b) P. Sörme, B. Kahl-Knutsson, U. Wellmar,
U. J. Nilsson, H. Leffler, Meth. Enzymol. 2003, 362, 504-512.
[13] a) S. Carlsson, M. C. Carlsson, H. Leffler, Glycobiology 2007, 17, 906-
912; b) S. Carlsson, C. T. Öberg, M. C. Carlsson, A. Sundin, U. J.
Nilsson, D. Smith, R. D. Cummings, J. Almkvist, A. Karlsson, H. Leffler,
Glycobiology 2007, 17, 663-676.
acknowledged
experiments.
for
performing
fluorescence
anisotropy
[14] K. Bum-Erdene, I. A. Gagarinov, P. M. Collins, M. Winger, A. G.
Pearson, J. C. Wilson, H. Leffler, U. J. Nilsson, I. D. Grice, H.
Blanchard, Chembiochem 2013, 14, 1331-1342.
Keywords: galectin • antagonist • thiodigalactoside • vesicle •
fibrosis
[15] I. Cumpstey, E. Salomonsson, A. Sundin, H. Leffler, U. J. Nilsson,
ChemBioChem 2007, 8, 1389-1398.
[16] a) K. Müller, C. Faeh, F. Diederich, Science 2007, 317, 1881-1886; b)
R. Paulini, K. Müller, F. Diederich, Angew. Chem. Int. Ed. 2005, 44,
1788-1805; c) M. Zürcher, F. Diederich, J. Org. Chem. 2008, 73, 4345-
4361.
[1]
[2]
P. Lajoie, J. G. Goetz, J. W. Dennis, I. R. Nabi, J. Cell Biol. 2009, 185,
381-385.
C. Boscher, J. W. Dennis, I. R. Nabi, Curr. Opin. Cell Biol. 2011, 23,
383-392.
[17] P. Zhou, J. Zou, F. Tian, Z. Shang, J. Chem. Inf. Mod. 2009, 49, 2344-
2355.
[3]
[4]
[5]
F. T. Liu, G. A. Rabinovich, Ann. N. Y. Acad. Sci. 2010, 1183, 158-182.
F.-T. Liu, G. A. Rabinovich, Nat. Rev. Cancer 2005, 5, 29-41.
a) C.-I. Lin, E. E. Whang, D. B. Donner, X. Jiang, B. D. Price, A. M.
Carothers, T. Delaine, H. Leffler, U. J. Nilsson, V. Nose, F. D. Moore, D.
T. Ruan, Mol. Cancer Res. 2009, 7, 1655-1662; b) A. C. MacKinnon, S.
L. Farnworth, P. S. Hodkinson, N. C. Henderson, K. M. Atkinson, H.
Leffler, U. J. Nilsson, C. Haslett, S. J. Forbes, T. Sethi, J. Immunol.
2008, 180, 2650-2658.
[18] J. C. Biffinger, H. W. Kim, S. G. DiMagno, ChemBioChem 2004, 5, 622-
627.
[19] D. N. Cooper, Biochim Biophys Acta 2002, 1572, 209-231.
[20] a) S. Aits, J. Kricker, B. Liu, A.-M. Ellegaard, S. Hämälistö, S.
Tvingsholm, E. Corcelle-Termeau, S. Høgh, T. Farkas, A. H. Jonassen,
I. Gromova, M. Mortensen, M. Jäättelä, Autophagy 2015, 11, 1408-
1424; b) I. Paz, M. Sachse, N. Dupont, J. Mounier, C. Cederfur, J.
Enninga, H. Leffler, F. Poirier, M. C. Prevost, F. Lafont, P. Sansonetti,
Cell Microbiol. 2010, 12, 530-544.
[6]
[7]
a) V. V. Glinsky, G. Kiriakova, O. V. Glinskii, V. V. Mossine, T. P.
Mawhinney, J. R. Turk, A. B. Glinskii, V. H. Huxley, J. E. Price, G. V.
Glinsky, Neoplasia 2009, 11, 901-909; b) A. C. Mackinnon, M. A.
Gibbons, S. L. Farnworth, H. Leffler, U. J. Nilsson, T. Delaine, A. J.
Simpson, S. J. Forbes, N. Hirani, J. Gauldie, T. Sethi, Am. J. Resp. Crit.
Care Med. 2012, 185, 537-546.
[21] T. L. Thurston, M. P. Wandel, N. von Muhlinen, A. Foeglein, F. Randow,
Nature 2012, 482, 414-418.
[22] N. H. Petersen, O. D. Olsen, L. Groth-Pedersen, A. M. Ellegaard, M.
Bilgin, S. Redmer, M. S. Ostenfeld, D. Ulanet, T. H. Dovmark, A.
Lonborg, S. D. Vindelov, D. Hanahan, C. Arenz, C. S. Ejsing, T.
Kirkegaard, M. Rohde, J. Nylandsted, M. Jaattela, Cancer Cell 2013, 24,
379-393.
a) B. A. Salameh, H. Leffler, U. J. Nilsson, Bioorg. Med. Chem. Lett.
2005, 15, 3344-3346; b) B. A. Salameh, A. Sundin, H. Leffler, U. J.
Nilsson, Bioorg. Med. Chem. 2006, 14, 1215-1220; c) P. Sörme, P.
Arnoux, B. Kahl-Knutsson, H. Leffler, J. M. Rini, U. J. Nilsson, J. Am.
Chem. Soc. 2005, 127, 1737-1743; d) P. Sörme, Y. Qian, P.-G. Nyholm,
H. Leffler, U. J. Nilsson, ChemBioChem 2002, 3, 183-189; e) J. Tejler,
B. Salameh, H. Leffler, U. J. Nilsson, Org. Biomol. Chem. 2009, 7,
3982-3990; f) S. Fort, H. S. Kim, O. Hindsgaul, J. Org. Chem. 2006, 71,
7146-7154; g) D. Giguère, S. André, M.-A. Bonin, M.-A. Bellefleur, A.
Provencal, P. Cloutier, B. Pucci, R. Roy, H.-J. Gabius, Bioorg. Med.
[23] S. Aits, J. Kricker, B. Liu, A. M. Ellegaard, S. Hamalisto, S. Tvingsholm,
E. Corcelle-Termeau, S. Hogh, T. Farkas, A. H. Jonassen, I. Gromova,
M. Mortensen, M. Jaattela, Autophagy 2015, 0.
[24] V. Sarafian, M. Jadot, J. M. Foidart, J. J. Letesson, F. van den Brûle, V.
Castronovo, R. Wattiaux, S. Wattiaux-De Coninck, Int. J. Cancer 1998,
75, 105-111.

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
[25] E. Salomonsson, A. Larumbe, J. Tejler, E. Tullberg, H. Rydberg, A.
Sundin, T. Khabut, T. Frejd, Y. D. Lobsanov, J. M. Rini, U. J. Nilsson, H.
Leffler, Biochemistry 2010, 49, 9518-9532.
[34] J. Seetharaman, A. Kanigsberg, R. Slaaby, H. Leffler, S. H. Barondes,
J. M. Rini, J. Biol. Chem. 1998, 273, 13047-13052.
[35] G. N. Murshudov, P. Skubak, A. A. Lebedev, N. S. Pannu, R. A. Steiner,
R. A. Nicholls, M. D. Winn, F. Long, A. A. Vagin, Acta Crystallogr. D
Biol. Crystallogr. 2011, 67, 355-367.
[26] M. A. Gitt, S. M. Massa, H. Leffler, S. H. Barondes, J. Biol. Chem. 1992,
267, 10601-10606.
[27] S. M. Massa, D. N. Cooper, H. Leffler, S. H. Barondes, Biochemistry
1993, 32, 260-267.
[36] R. J. Read, A. J. McCoy, Acta Crystallogr. D Biol. Crystallogr. 2011, 67,
338-344.
[28] T. Magnaldo, D. Fowlis, M. Darmon, Differentiation 1998, 63, 159-168.
[29] I. Cumpstey, S. Carlsson, H. Leffler, U. J. Nilsson, Org. Biomol. Chem.
2005, 3, 1922-1932.
[37] P. Emsley, B. Lohkamp, W. G. Scott, K. Cowtan, Acta Crystallogr. D
Biol. Crystallogr. 2010, 66, 486-501.
[38] A. W. Schuttelkopf, D. M. van Aalten, Acta Crystallogr.
Crystallogr. 2004, 60, 1355-1363.
D Biol.
[30] C. T. Öberg, S. Carlsson, E. Fillion, H. Leffler, U. J. Nilsson, Bioconj.
Chem. 2003, 14, 1289-1297.
[39] V. B. Chen, W. B. Arendall, 3rd, J. J. Headd, D. A. Keedy, R. M.
Immormino, G. J. Kapral, L. W. Murray, J. S. Richardson, D. C.
Richardson, Acta Crystallogr. D Biol. Crystallogr. 2010, 66, 12-21.
[40] S. McNicholas, E. Potterton, K. S. Wilson, M. E. Noble, Acta Crystallogr.
D Biol. Crystallogr. 2011, 67, 386-394.
[31] P. M. Collins, K. I. P. J. Hidari, H. Blanchard, Acta Crystallogr. D Biol.
Crystallogr. 2007, 63, 415-419.
[32] P. Evans, Acta Crystallogr. D Biol. Crystallogr. 2006, 62, 72-82.
[33] M. D. Winn, C. C. Ballard, K. D. Cowtan, E. J. Dodson, P. Emsley, P. R.
Evans, R. M. Keegan, E. B. Krissinel, A. G. Leslie, A. McCoy, S. J.
McNicholas, G. N. Murshudov, N. S. Pannu, E. A. Potterton, H. R.
Powell, R. J. Read, A. Vagin, K. S. Wilson, Acta Crystallogr. D Biol.
Crystallogr. 2011, 67, 235-242.
[41] E. Salomonsson, M. Carlsson, V. Osla, R. Hendus-Altenburger, B. Kahl
Knutson, C. Oberg, A. Sundin, R. Nilsson, E. Nordberg-Karlsson, U.
Nilsson, A. Karlsson, J. M. Rini, H. Leffler, J. Biol. Chem. 2010, 285,
35079-35091.
[42] M. K. Ho, T. A. Springer, J. Immunol. 1982, 128, 1221-1228.

ChemBioChem
10.1002/cbic.201600285
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Dot removal: Synthetic ligands
capitalizing on unnatural lectin
interactions efficiently inhibit
intracellular galectin-3 accumulation in
dots on damaged vesicles and
attenuates experimental lung fibrosis
T. Delaine, P. Collins, A. MacKinnon, G.
Sharma, J. Stegmayr, V. K. Rajput, S.
Mandal, I. Cumpstey, A. Larumbe, B. A.
Salameh, B. Kahl-Knutsson, H. van
Hattum, M. v. Scherpenzeel, R. J.
Pieters, T. Sethi, H. Schambye, S.
Oredsson, H. Leffler, H. Blanchard,* and
U. J. Nilsson*
Page No. – Page No.
Galectin-3-binding glycomimetics
that strongly reduce bleomycin-
induced lung fibrosis and modulate
intracellular glycan recognition
((Insert TOC Graphic here: max.
width: 5.5 cm; max. height: 5.0 cm))