Targeting matrix metalloproteinases: Design of a bifunctional inhibitor for presentation by tumour-associated galectins

Article abstract of DOI:10.1002/chem.201203794

A new strategy to exploit galectin presence to target matrix metalloproteinases (MMPs) is presented. A bifunctional conjugate with lactose and an inhibitor for MMPs is able to bind MMP and Gal-3 simultaneously. This compound might allow the lectin to attract the MMP inhibitor to the tumour site and to block protumoural activities of the lectin at the same time. Copyright

Full text of DOI:10.1002/chem.201203794

DOI: 10.1002/chem.201203794
Targeting Matrix Metalloproteinases: Design of a Bifunctional Inhibitor for
Presentation by Tumour-Associated Galectins
Marco Bartoloni,[a] Blanca E. Domꢀnguez,[b, e] Elisa Dragoni,[a] Barbara Richichi,[a]
Marco Fragai,[a, c] Sabine Andrꢁ,[d] Hans-Joachim Gabius,[d] Ana Ardꢂ,[b]
Claudio Luchinat,[a, c] Jesffls Jimꢁnez-Barbero,*[b] and Cristina Nativi*[a, c]
In memory of Prof. Ivano Bertini
Matrix metalloproteinases (MMPs) have long been con-
sidered promising therapeutic targets and catch the atten-
tion of many researchers but with disappointing results.
Nonetheless, increasingly strong evidence for the involve-
ment of MMPs in the development of diverse pathologic
states, such as tumour invasion, gives reason to consider
these proteases worthy of a renewed interest, though in a
new perspective.[1] In this context, we recently reported the
rational design and the synthesis of a family of water-soluble
Scheme 1. Structure of NNGH and inhibitors 1a,b and 2.
MMP inhibitors (MMPIs) 1a,b and 2 that are structurally
related to NNGH but featuring a hydrophilic rather than
hydrophobic substituent R.[2] Low-nanomolar affinities were
reached with these new inhibitors towards therapeutically
relevant MMPs, including MMP-9 (Scheme 1).
MMP activity, namely to interfere with MMPs selectively at
sites of disease progression.[4] In principle, installing a hydro-
philic group at the sulfonamidic nitrogen, which can account
for a preferential presentation of the inhibitor conjugate, for
example in or around tumours, would increase therapeutic
efficiency and reduce the extent of side effects. Along this
line, it was necessary to identify a receptor–ligand pair that
would fulfill these two prerequisites; that is, to probe
tumour cell or stroma MMPs overexpression and featuring a
hydrophilic substituent to direct the conjugate to galectins
with sufficient affinity.
The main improvement gained by inhibitors 1 and 2 with
respect to the large number of reported MMPIs[3] concerns
their solubility in water. This is achieved by replacing the
isobutyl residue of NNGH with hydrophilic groups. The
choice of the hydrophilic residue enables the solubility of
the inhibitors to be modulated without affecting their affini-
ty to MMPs. Importantly, this possibility for synthetic flexi-
bility opens the route to address a major issue in targeting
[a] Dr. M. Bartoloni, Dr. E. Dragoni, Dr. B. Richichi, Dr. M. Fragai,
Prof. C. Luchinat, Prof. C. Nativi
The far-reaching functionality of endogenous lectins com-
bined with their intriguing selectivity for glycans[5] directed
our interest to establish a proof-of-principle for a physiologi-
cally relevant test case. Clinical association of expression of
galectins to tumour progression[6] attracted our attention.
Furthermore, the presence of galectins in the zone of
tumour invasion and the role in tumour progression by up-
regulating MMP production reveal a functional connection
between these two effector proteins and their potential for
the suitable positioning of a ligand in situ, even with the
added benefit to block clinically unfavourable galectin activ-
ities.[7] As the term implies, these adhesion/growth-regulato-
ry lectins show a considerable affinity to b-galactosides so
that the introduction of carbohydrates as residue in MMP li-
gands favourably meets the requirement for the presence of
a hydrophilic substituent in sulfonamidic-based inhibitors. In
this pilot study, we focus on galectin-3.
Department of Chemistry, University of Florence
Via della Lastruccia, 13 50019 Sesto F.no Firenze (Italy)
E-mail: cristina.nativi@unifi.it
[b] Dr. B. E. Domꢀnguez, Dr. A. Ardꢁ, Prof. J. Jimꢂnez-Barbero
Centro de Investigationes Biolꢃgicas, CSIC
Ramiro de Maetzu 9, 28040 Madrid (Spain)
E-mail: jjbarbero@cib.csic.es
[c] Dr. M. Fragai, Prof. C. Luchinat, Prof. C. Nativi
CERM, University of Florence
Via Sacconi, 6 50019 Sesto F.no Firenze (Italy)
[d] Dr. S. Andrꢂ, Prof. H.-J. Gabius
Institute of Physiological Chemistry, Faculty of Veterinary Medicine
Ludwig-Maximilians-University, Veterinꢄrstr. 13
80539 Mꢅnchen (Germany)
[e] Dr. B. E. Domꢀnguez
Centro de Investigaciones Quꢀmicas
Universidad Autꢆnoma del Estado de Morelos
Av. Universidad 1001, 62210, Cuernavaca, Morelos (Mꢂxico)
This tumour-associated lectin is itself a substrate of MMP-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201203794.
2, MMP-7, MMP-9 and MMP-14.[8] The removal of the col-
1896
ꢇ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1896 – 1902

COMMUNICATION
lagenous tail from the lectin
domain impairs the capacity for
oligomerization in the presence
of multivalent ligands;[9] lectin
oligoconjugation is known to
promote angiogenesis and os-
teoclastogenesis
in
breast
cancer and bone metastases.[10]
By this process, the affinity for
lactose is not altered so that
both full-length and truncated
forms of galectin-3, as well as
other tumour-associated galec-
tins, preserve the binding prop-
erties versus the ligand.[11] Of
note, work with substituted lac-
tosides has revealed that the
aglycone extension does not
impair ligand activity; converse-
ly, it may even increase the re-
activity and introduce inter-ga-
lectin selectivity.[12] Having thus
Scheme 3. Synthesis of sulfonamide 9.
explained the selection for lectin type and substituent, we
report the synthesis of the bifunctional MMP inhibitor 3,
which is related to 2 and endowed with an assumedly strate-
gic bioactive lactose moiety (Scheme 2). Its ligand proper-
ties were analysed using the carbohydrate recognition
domain (CRD) of chicken galectin-3 (CG-3),[13] both experi-
mentally by NMR spectroscopy studies and computationally
by molecular modelling.
Cbz protecting group (H2/Pd(OH)2/C) afforded amino deriv-
ative 7, which was treated with adipic ester 8[17] and N-
methyl morpholine (NMM) in DMF as solvent to give 9
(66% over two steps; Scheme 3). Glycosyl donor 10, ob-
tained from peracetylated lactose and trichloroacetonitrile
under Schmidt reaction conditions,[18] was first processed
with N-Cbz-protected ethanolamine 11 in the presence of
trimethylsilyl triflate as promoter to form the b-O-glycosyl
derivative 12 (40%). After removal of the protecting group
(H2-Pd(OH)2/C) the resulting lactose derivative 13 was re-
acted with 9 and NMM in DMF as solvent to afford the sul-
fonamide 14 (90% over two steps).
Deacetylation on the saccharidic portion (K2CO3, MeOH)
transformed 14 into 15 (>90%). Methyl ester 15 was treat-
ed with hydroxylamine hydrochloride and potassium hy-
droxide in methanol as solvent to generate the desired bi-
functional product 3 (31%; Scheme 4). This compound en-
Scheme 2. Structure of ligand 3.
AHCTUNGTREGaNNUN bled the testing of the assumed dual activity, that is, to in-
This new ditopic molecule is considered as lead com-
pound, combining the MMP inhibition property with the ga-
lectin-targeting and -blocking capacity owing to a galacto-
side substituent. Upregulated presence in tumours and also
in inflammation (galectin-3 is also known as MAC-2 antigen,
a macrophage marker),[14] together with MMP activity on
galectin-3, make this lectin family member ideal for testing.
Compound 3 is a sulfonamide presenting a hydroxamic
acid as zinc-binding group linked to lactose through an
amidic spacer. The length of the spacer was chosen to allow
interactions with both the lectin and the catalytic domain of
MMP and to prevent the two proteins to clash.[15] The syn-
thesis of compound 3 was performed by a strategy shown in
Scheme 3 and 4. The sulfonamide of the glycine methylester
4 was reacted with a benzyloxycarbonyl (Cbz)-protected
aminohexanol 5 under Mitsunobu reaction conditions[16] to
form methylester 6 in quantitative yield. Removal of the
hibit MMPs and to bind galectin.
The inhibitory potency of compound 3 was tested on a
panel of six MMPs: MMP-1, 7, 8, 9, 12 and 13 (see the Sup-
porting Information for details). Assays used the catalytic
domain of the proteins. As previously observed,[2b] except
for MMP-1 and MMP-7, compound 3 showed low-nanomo-
lar Ki values for the MMPs tested. This confirms that the
functionalization with lactose did not impair the inhibitory
activity of the ligand.[2a]
To investigate the ability of 3 to bind MMP and Gal-3
simultaneously, a series of NMR experiments were carried
out, focusing our attention on the ligand (STD and
trNOESY) as well as on the proteins (HSQC titration on
15N-labelled proteins). The catalytic domain of MMP-12
(catMMP-12) was chosen as a model, since a full assignment
of the respective signals in solution is available and the
binding affinity of the sulfonamide scaffold for MMP-12 is
Chem. Eur. J. 2013, 19, 1896 – 1902
ꢇ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1897

J. Jimꢂnez-Barbero, C. Nativi et al.
CG-3/ligand 3 mixture (molar
ratio 1:20; see the Supporting
Information for details), and
compared with the NOESY
spectrum of the ligand. As
shown in Figure 2, and in con-
trast with the observations for
the free ligand, the cross-peaks
of the galactose moiety of 3
were negative in the presence
of CG-3. However, the NOEs
corresponding to protons of the
spacer, the phenylsulfonamide
and the hydroxamic acid moiet-
ies remained positive in the
presence of the lectin. The sign
change of the cross-peaks for
the lactose part of 3 in the pres-
ence of the lectin indicated that
a molecular recognition takes
place and exclusively involves
Scheme 4. Synthesis of ligand 3.
similar to those for gelatinases. As a model for these inter-
action studies, we used the CRD of CG-3, first studying the
binary complexes of both proteins with compound 3.
For the CG-3$ligand 3 binary complex, the STD spec-
trum (Figure 1), recorded under standard conditions (see
also the Supporting Information), clearly showed that the
Figure 2. Left: region of the 500 MHz NOESY (500 ms mixing time)
spectrum of compound 3 in solution at 298 K. Right: the same region of
the 500 MHz trNOESY (250 ms mixing time) spectrum of a sample of
CG-3 (100 mm) and compound 3 (2 mm) at 298 K. The regions show the
NOE correlations for the anomeric protons: positive (blue) when the
ligand is free, and negative (black) in the presence of CG-3.
Figure 1. Saturation-transfer difference (STD) spectrum for a sample of
CG-3 (100 mm) and compound 3 (5 mm). Top: the off-resonance reference
spectrum. Below: the STD spectrum. An STD of CG-3 acquired under
the same conditions was subtracted to the STD of the CG-3+ligand
sample to remove the protein background signals. The protons of the
ligand showing STD are annotated.
lactose. Consistently, the simultaneous presence of positive
and negative cross-peaks for the different portions of 3 in
the presence of the lectin accounts for distinct effective cor-
relation times for each region within 3. The negative sign
for the cross-peaks of lactose reveals that the correlation
time for this portion is longer than for those of the spacer or
the sulfonamidic substructures. This prolonged correlation
time is most likely due to a close contact of the sugar with
the protein, as the STD data document, while the rest of the
ligand likely retains high flexibility responsible for positive
NOEs, even in the presence of CG-3. Apparently, some
transient interactions between the lectin and the methoxy-
phenyl portion might be established, as suggested by the
presence of signals in the aromatic region of the STD spec-
trum (protons A, B and C in Figure 1).
galactose residue is the portion of the ligand receiving high-
saturation transfer from the lectin, in particular protons H-5
(100%), H-6 (60%) and H-4 (53%). In fact, these protons
are prominent when testing free lactose, and also human ga-
lectins.[19] Noteworthy, the protons of the methoxyphenyl
group also received saturation transfer from the protein, in-
dicating additional carbohydrate-independent binding of
this part of the molecule with CG-3.
To the further define structural aspects of CG-3$ligand 3
recognition, a trNOESY experiment was run on a sample of
1898
www.chemeurj.org
ꢇ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1896 – 1902

Targeting Matrix Metalloproteinases
COMMUNICATION
1
Figure 3. Average chemical shift perturbation DdNH,H of H and 15N chemical shifts of CG-3 versus the amino acid sequence upon addition of 10 equiv
of lactose (top) and 3 (bottom). The 3D model for the CRD for CG-3 (right) shows the amino acids for which the backbone H/N HSQC cross-peaks
shifted the most; 10 mm Tris at pH 7.5, 298 K, 600 MHz for all spectra.
The recognition process was next monitored from the per-
spective of galectin by carrying out HSQC titration experi-
ments of 15N-CG-3 with increasing amounts of ligand 3 (see
the Supporting Information for details). The results were
compared with those obtained from parallel titration per-
formed with unsubstituted lactose as ligand. Figure 3 shows
ing of the indolyl ring of W79 with the hydrophobic patch
established by galactopyranose H-3, H-4 and H-5 became
evident along the simulation and exhibited an average dis-
tance between the two rings of 4.9 ꢈ. Glycosidic torsional
angles did not vary, indicating stability of the initial exoano-
meric-F/syn-Y conformer. Having herewith revealed and
characterized binding of 3 to the galectin, we next ascer-
tained the interaction of 3 with MMP-12.
We thus analysed the 3$catMMP-12 binary complex.
Taking into account the similar affinity of 3 for MMP-9 (Ki
80 nm) and MMP-12 (Ki 17 nm) and the availability of the
full assignment of the NMR signals of the MMP-12 catalytic
domain, either free of substrates or complexed with
NNGH,[20] the set of experiments to investigate the forma-
tion of the compound-3$MMP-12 complex were carried
out using the unlabelled and 15N, 13C-labelled catalytic
domain of macrophage metalloelastase (MMP-12; see also
the Supporting Information).
1
the average chemical-shift perturbation on H and 15N back-
bone resonances of CG-3 upon addition of lactose (top) and
compound 3 (below). It is clear that the sets of perturbed
backbone amino acids are very similar in both cases and,
furthermore, of comparable extent. This means that ligand 3
binds to CG-3 at the same site as lactose does, leading us to
assume that the binding mode of 3 might be very similar to
that of lactose, with the galactose ring stacked on top of
W79 residue and establishing a network of hydrogen bonds
(Supporting Information, Figure S1).
A putative three-dimensional model of the complex
formed was deduced by running MD simulations. The start-
ing structure was built by manually docking lactose at the
canonical lactose-binding site, as experimentally assessed by
the combined STD/trNOESY/HSQC procedure. The stack-
STD and trNOESY were not helpful to shed light onto
the interaction mode between the two partners. Very proba-
bly, the strong interaction provided by the aryl sulfonamidic
Chem. Eur. J. 2013, 19, 1896 – 1902
ꢇ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1899

J. Jimꢂnez-Barbero, C. Nativi et al.
hydroxamic acid with MMP-12 is responsible for the lack of
signals observed in both experiments. Indeed, tight binding
causes koff to be very slow. With 15N-labelled MMP-12
(0.1 mm) in Tris buffer (20 mm; see the Supporting Informa-
tion for details), titrations with a stock solution of 3 in D2O
were performed using different MMP-12/ligand 3 molar
ratios: 1:0.2, 1:0.5 and 1:1. As expected, for 1:0.2 and 1:0.5
ratios, two different states of the protein, in slow equilibri-
um on the chemical-shift timescale, were detected (shown
for G178 in Figure 4). These two states correspond to ligand
tion of catMMP-12 (0.1 mm; Supporting Information, Fig-
ure S3), allowing any direct interaction between the two
proteins to be excluded.
To delineate the bifunctionality of 3, the HSQC spectra of
the 15N-labelled catalytic domain of MMP-12 (0.1 mm) in
the presence of an equimolar concentration of ligand 3 was
recorded. As already described (see above), MMP-12 is
fully loaded under these conditions. Upon addition of CG-3
(0.2 mm), a new HSQC spectrum was recorded. Of particu-
lar importance, it presents a dramatic decrease of the inten-
sity of the cross-peaks corresponding to the 15N-labelled
MMP-12 amino acid residues located at the a-helix-B and in
the adjacent b-strands V, IV, III, II and I. Conversely, the in-
tensity of the peaks corresponding to the amino acids in the
a-helix-A maintained their intensity (Figure 5).
Figure 4. Superimposition of the 1H–15N-labelled HSQC spectra
(600 MHz) of 0.1 mm 15N catMMP-12 (black) and upon addition of
0.1 mm of 3 (blue). The arrows indicate some of the largest cross-peak
shifts observed and the corresponding amino acid residue. The cross-
peak corresponding to residue G178 (in dashed box) is shown at ratios
1:0, 1:0.2, 1:0.5 and 1:1.
Figure 5. The 600 MHz 1H–15N HSQC spectra. Left: 0.1 mm of 15N-label-
led catMMP-12+0.1 mm ligand 3 (ns=64). Right: the same sample after
the addition of 0.2 mm of CG-3 (ns=96).
The decrease of the intensity of the signals in the region
of the hydroxamate binding site and in the adjacent b-strand
strongly suggested that a large molecular weight entity, that
is a complex with the lectin, was formed in solution. The ex-
change process involving the large-size ternary complex
should cause a decrease of the effective T2 relaxation time
of many residues with the concomitant broadening of the
corresponding NMR signals.
Of note, all of the residues belonging to the a-helix-A
sited behind the binding site appeared to keep their relaxa-
tion properties. In aggregate, these data indicated that the
ternary complex was formed in solution and that 3 is able to
bind CG-3 and MMP-12 side-by-side.
In summary, we have established compound 3 as a lead
substance for the design of effective bifunctional probes, tar-
geting galectins for directing MMPI to specific sites. Clinical
documentation of the status of negative prognostic predic-
tors for members of both families, for example, MMP-9/ga-
lectin-7 in hypopharyngeal cancer,[7c,21] gives direction to
suited in vitro test models. Furthermore, the presence of a
galectin network in tumours with functionally antagonistic
activities, for example known between galectins-1 and
-3,[6,9a,22] can make it necessary to implement selectivity in
the sugar part. Appropriate substitutions, along with dendri-
meric display, offer ways toward selectivity and affinity in-
crease, as established for the noted case.[23] The formation of
3-free and ligand 3-loaded MMP-12 and indicate that the ex-
change process is slow on the chemical-shift timescale, as a
consequence of the strong binding between both partners.
The superimposition of the HSQC spectra of the 15N-label-
led catalytic domain of MMP-12 (0.1 mm) and a 1:1 15N
MMP-12 (0.1 mm)/ligand 3 (0.1 mm) mixture allowed the
cross-peaks of the protein to be identified that showed the
larger chemical-shift perturbation upon ligand 3 binding. As
depicted in Figure 4, these cross-peaks correspond to resi-
dues G178, I180, L181 located in the b-sheet IV, T215 and
H 218 located in the a-helix 2 and T239, Y240, R241 and
Y242 situated in the L8 loop, all around the catalytic bind-
ing site (Supporting Information, Figure S2). These spec-
ACHTUNGTRENNUNGtroscopic data clearly indicate that the binding mode of
ligand 3 is fairly similar to that previously described for
other MMPIs structurally related to NNGH.[2b] Also, they
add structural information to the proven inhibitor activity.
Together, these data set the stage to test whether a ternary
complex is formed.
Before investigating the possibility for formation of the
CG-3$ligand 3$catMMP-12 ternary complex, we checked
whether a direct interaction between CG-3 and catMMP-12
was possible. In the absence of compound 3, no chemical-
shift perturbations of the HSQC cross-peaks of 15N-labelled
CG-3 (0.1 mm in Tris buffer) were observed upon the addi-
1900
www.chemeurj.org
ꢇ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1896 – 1902

Targeting Matrix Metalloproteinases
COMMUNICATION
the ternary complex in situ will then not only exploit the
lectin for attracting MMPI to the tumour site but also, as an
added benefit, block protumoural activities of the lectin at
the same time. The same strategy can be applied to other
pathologic states with concomitant upregulation of MMPs/
galectins, such as rheumatoid arthritis, thus broadening the
applicability of this concept. The synthesis of C-lactosides as
well as of thioglycosides analogues of compound 3 in den-
drimeric display is actually in progress in order to obtain
heterobifunctional ligands stable against glycosidases.[12,24]
[1] a) B. Fingleton, Sem. Cell Dev. Biol. 2008, 19, 61–68; b) A. Noꢊl, M.
Jost, E. Maquoi, Sem. Cell Dev. Biol. 2008, 19, 52–60.
[2] a) E. Attolino, V. Calderone, E. Dragoni, M. Fragai, B. Richichi, C.
Luchinat, C. Nativi, Eur. J. Med. Chem. 2010, 45, 5919–5925; b) I.
Bertini, V. Calderone, M. Fragai, A. Giachetti, M. Loconte, C. Lu-
chinat, M. Maletta, C. Nativi, K.-J. Yeo, J. Am. Chem. Soc. 2007,
129, 2466–2475.
[3] a) M. Egeblad, Z. Werb, Nat. Rev. Cancer 2002, 2, 161–174;
b) W. C. Parks, C. L. Wilson, Y. S. Lopez-Boado, Nat. Rev. Immunol.
2004, 4, 617–629.
[4] S. H. Park, H. S. Min, B. Min, J. Myung, S. H. Paek, Neuropathology
2008, 28, 497–506.
[5] a) The Sugar Code: Fundamentals of Glycosciences (Ed.: H.-J.
Gabius), Wiley-VCH, Weinheim, 2009; b) H.-J. Gabius, S. Andrꢂ, J.
Jimꢂnez-Barbero, A. Romero, D. Solꢀs, Trends Biochem. Sci. 2011,
36, 298–313.
Experimental Section
[6] a) H. Barrow, J. M. Rhodes, L.-G. Yu, Int. J. Cancer 2011, 129, 1–8;
b) H. Kaltner, H.-J. Gabius, Histol. Histopathol. 2012, 27, 397–416;
c) G. Radosavljevic, V. Volarevic, I. Jovanovic, Immunol. Res. 2012,
52, 100–110.
Synthesis of compound 14: Pd(OH)2/C (20% Pd content, 70.0 mg) was
added to a stirred solution of 12 (0.150 g, 0.184 mmol) in a mixture of
EtOAc/MeOH 3:1 (5.0 mL). The mixture was stirred at RT for 6 h under
a H2 atmosphere then filtered through a pad of Celite. The filtrate was
concentrated to dryness to give 0.130 g of crude product 13, which was
used for the next step without further purification. Compound 9 (0.334 g,
0.191 mmol) and NMM (32 mL, 0.287 mmol) were added to a stirred solu-
tion of 13 (0.130 mg) in dry DMF (2.0 mL). The mixture was stirred at
RT for 6 h then diluted with CH2Cl2 (100 mL) and washed with brine (2ꢉ
10 mL) and water (2ꢉ10 mL). The organic phase was dried over Na2SO4
and concentrated to dryness to give 0.270 mg of crude product. The
crude product was purified by flash chromatography on silica gel (EtOAc
then CHCl3/MeOH 15:1) to give 14 (0.190 g, 90% over two steps) as a
yellow oil (see the Supporting Information for details of spectroscopy).
[7] a) S. Rorive, N. Belot, C. Decaestecker, F. Lefranc, L. Gordower, S.
Micik, C. A. Maurage, H. Kaltner, M. M. Ruchoux, A. Danguy, H.-J.
Gabius, I. Salmon, R. Kiss, I. Camby, Glia 2001, 33, 241–255; b) M.
Demers, T. Magnaldo, Y. St-Pierre, Cancer Res. 2005, 65, 5205–
5210; c) O. Roda, E. Ortiz-Zapater, N. Martinez-Bosch, R. Gutiꢂr-
rez-Gallego, M. Vila-Perellꢆ, C. Ampurdanꢂs, H.-J. Gabius, S.
Andrꢂ, D. Andreu, F. X. Real, P. Navarro, Gastroenterology 2009,
136, 1379–1390; d) S. Saussez, S. Cludts, A. Capouillez, G. Mortu-
aire, K. Smetana, H. Kaltner, S. Andrꢂ, X. Leroy, H.-J. Gabius, C.
Decaestecker, Int. J. Oncol. 2009, 34, 433–439.
[8] a) J. Herrmann, C. W. Turck, R. E. Atchison, M. E. Huflejt, L. Poult-
er, M. A. Gitt, A. L. Burlingame, S. H. Barondes, H. Leffler, J. Biol.
Chem. 1993, 268, 26704–26711; b) J. Ochieng, R. Fridman, P.
Nangia-Makker, D. E. Kleiner, L. A. Liotta, W. G. Stetler-Stevenson,
A. Raz, Biochemistry 1994, 33, 14109–14114; c) G. S. Butler, R. A.
Dean, E. M. Tam, C. M. Overall, Mol. Cell. Biol. 2008, 28, 4896–
4914; d) M. Puthenedam, F. Wu, A. Shetye, A. Michaels, K. J. Rhee,
J. H. Kwon, Inflamm. Bowel Dis. 2011, 17, 260–267.
[9] a) J. Kopitz, C. von Reitzenstein, S. Andrꢂ, H. Kaltner, J. Uhl, V.
Ehemann, M. Cantz, H.-J. Gabius, J. Biol. Chem. 2001, 276, 35917–
35923; b) N. Ahmad, H.-J. Gabius, S. Andrꢂ, H. Kaltern, S. Sabesan,
R. Roy, B. Liu, F. Macaluso, C. F. Brewer, J. Biol. Chem. 2004, 279,
10841–10847.
Synthesis of compound 15: K2CO3 (0.005 g, 0.036 mmol) was added to a
stirred solution of 14 (0.190 g, 0.165 mmol) in MeOH (1.5 mL). The mix-
ture was stirred at RT for 4 h then concentrated to dryness to give
0.145 g of crude product. The crude product was purified by flash column
chromatography on silica gel (CH2Cl2/MeOH 4:1) to give 15 (0.130 g,
92%) as a white solid (see the Supporting Information for details of
spectroscopy).
Synthesis of ligand 3: A suspension of KOH (0.200 g, 3.564 mmol) in
MeOH (700 mL) and a suspension of NH2OH·HCl (0.100 g, 1.439 mmol)
in MeOH (860 mL) were separately stirred for 10 min at 608C. Then
280 mL of the obtained 5.0m solution of KOH were added to the ob-
tained 1.6m solution of NH2OH·HCl. The reaction mixture was stirred
for 10 min at 508C, then 15 (0.130 g, 0.152 mmol) was dissolved in 490 mL
of the freshly obtained suspension of NH2OH. The mixture was stirred at
RT for 3 h then diluted with MeOH (10 mL) and the solid was filtered
off. The filtrate was concentrated to dryness to give 0.123 g of crude
product. The crude product was purified by HPLC (column Zorbax
[10] a) P. Nangia-Makker, Y. Wang, T. Raz, L. Tait, V. Balan, V. Hogan,
A. Raz, Int. J. Cancer 2010, 127, 2530–2541; b) E. Pivetta, M. Scapo-
lan, M. Pecolo, B. Wassermann, I. Abu-Rumeileh, L. Balestreri, E.
Borsatti, C. Tripodo, A. Colombatti, P. Spessotto, Breast Cancer Res.
2011, 13, R105.
300SB-C18, 9.4ꢉ250,
5 um, H2O/MeOH 60:40 to 10:90) to give 3
[11] T. K. Dam, H.-J. Gabius, S. Andrꢂ, H. Kaltner, M. Lensch, C. F.
Brewer, Biochemistry 2005, 44, 12564–12571.
(0.040 g, 31%) as a glassy solid (see the Supporting Information for de-
tails of spectroscopy).
[12] a) S. Andrꢂ, D. Giguꢂre, T. K. Dam, F. Brewer, H.-J. Gabius, R.
Roy, New J. Chem. 2010, 34, 2229–2240; b) F. J. MuÇoz, J. I. Santos,
A. Ardꢁ, S. Andrꢂ, H.-J. Gabius, J. V. Siniterra, J. Jimꢂnez-Barbero,
M. J. Hernꢋiz, Org. Biomol. Chem. 2010, 8, 2986–2992; c) D. Gi-
guꢌre, S. Andrꢂ, M. A. Bonin, M. A. Bellefleur, A. Provencal, P.
Cloutier, B. Pucci, R. Roy, H.-J. Gabius, Bioorg. Med. Chem. 2011,
19, 3280–3287.
[13] H. Kaltner, L. Lꢆpez-Merino, M. Lohr, J. C. Manning, M. Lensch, J.
Seidler, W. D. Lehmann, S. Andrꢂ, D. Solꢀs, H.-J. Gabius, Anat. Rec.
2011, 294, 427–444.
[14] a) M.-K. Ho, T. A. Springer, J. Immunol. 1982, 128, 112–124;
b) R. C. Hughes, Glycobiology 1994, 4, 5–12; c) M. Krzeminski, T.
Singh, S. Andrꢂ, M. Lensch, A. M. Wu, A. M. Bonvin, H.-J. Gabius,
Biochim. Biophys. Acta 2011, 1810, 150–161.
Acknowledgements
The group at Madrid thanks MINECO of Spain (grant CTQ2012–32025),
Comunidad de Madrid (MHit project). We also thank the EU for the
GlycoHit and Glycopharm projects, as well as COST actions CM1102
and BM1003.
[15] E. Dragoni, V. Calderone, M. Fragai, R. Juiswal, C. Luchinat, C.
Nativi, Bioconjugate Chem. 2009, 20, 719–727.
[16] V. Calderone, M. Fragai, C. Luchinat, C. Nativi, B. Richichi, S. Roe-
lens, ChemMedChem 2006, 1, 598–601.
Keywords: glycoconjugates · inhibitors · ligand design ·
metalloproteins · NMR spectroscopy
Chem. Eur. J. 2013, 19, 1896 – 1902
ꢇ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1901

J. Jimꢂnez-Barbero, C. Nativi et al.
ˇ
[17] X. Wu, C.-C. Ling, D. R. Bundle, Org. Lett. 2004, 6, 4407–4410.
[18] R. R. Schmidt, W. Kinzy, Adv. Carbohydr. Chem. Biochem. 1994, 50,
21.
topathology 2007, 51, 681–690; b) Z. Cada, K. Smetana, L. Lacina,
ˇ
Z. Plzakova, J. Stork, H. Kaltner, R. Russwurm, M. Lensch, S.
Andrꢂ, H.-J. Gabius, Folia Biol. 2009, 55, 145–152; c) H. Sanchez-
Ruderisch, C. Fischer, K. M. Detjen, M. Welzel, A. Wimmel, J. C.
Manning, S. Andrꢂ, H.-J. Gabius, FEBS J. 2010, 277, 3552–3563;
d) S. Saussez, L. de Leval, C. Decaestecker, N. Sirtaine, S. Cludts, A.
Duray, D. Chevalier, S. Andrꢂ, H.-J. Gabius, M. Remmerlink, X.
Leroy, Histol. Histopathol. 2010, 25, 541–550; e) M. Remmelink, L.
De Leval, C. Decaestecker, A. Duray, E. Crompot, N. Sirtaine, S.
Andrꢂ, H. Kaltner, X. Leroy, H.-J. Gabius, S. Saussez, Histopatholo-
gy 2011, 58, 543–556.
[19] a) D. Solꢀs, M. J. Matꢂ, M. Lohr, J. P. Ribeiro, L. Lꢆpez-Merino, S.
Andrꢂ, E. Buzamet, F. J. CaÇada, H. Kaltner, M. Lensch, F. M.
Ruiz, G. Haroske, U. Wollina, M. Kloor, J. Kopitz, J. L. Sꢋiz, M. Me-
nꢂndez, J. Jimꢂnez-Barbero, A. Romero, H.-J. Gabius, Int. J. Bio-
chem. Cell Biol. 2010, 42, 1019–1029; b) S. Martꢀn-Santamarꢀa, S.
Andrꢂ, E. Buzamet, R. Caraballo, G. Fernꢋndez-Cureses, M. Moran-
do, J. P. Ribeiro, K. Ramirez-Gualito, B. de Pascual-Teresa, F. J.
CaÇada, M. Menꢂndez, O. Ramstrçm, J. Jimꢂnez-Barbero, D. Solꢀs,
H.-J. Gabius, Org. Biomol. Chem. 2011, 9, 5445–5454.
[20] a) I. Bertini, V. Calderone, M. Cosenza, M. Fragai, M. Lee, C. Lu-
chinat, S. Mangani, B. Terni, P. Turano, Proc. Natl. Acad. Sci. USA
2005, 102, 5334–5339; b) I. Bertini, V. Calderone, M. Fragai, C. Lu-
chinat, S. Mangani, B. Terni, J. Mol. Biol. 2004, 336, 707–716.
[21] a) S. Saussez, D. R. Cucu, C. Decaestecjeer, D. Chevalier, H. Kalt-
ner, S. Andrꢂ, A. Wacreniez, G. Toubeau, I. Camby, H.-J. Gabius, R.
Kiss, Ann. Surg. Oncol. 2006, 13, 999–1006; b) S. Saussez, C. Deca-
estecker, F. Lorfevre, D. Chevalier, G. Mortuaire, H. Kaltner, S.
Andrꢂ, G. Toubeau, H.-J. Gabius, X. Leroy, Histopathology 2008,
52, 483–493.
[23] a) S. Andrꢂ, F. Sansone, H. Kaltner, A. Casnati, J. Kopitz, H.-J.
Gabius, R. Ungaro, ChemBioChem 2008, 9, 1649–1661; b) S. Andrꢂ,
C. Grandjean, F. M. Gautier, S. Bernardi, F. Sansone, H.-J. Gabius,
R. Ungaro, Chem. Commun. 2011, 47, 6126–6128.
[24] a) J. L. Asensio, J. F. Espinosa, H. Dietrich, F. J. CaÇada, R. R.
Schmidt, M. Martin-Lomas, S. Andrꢂ, H.-J. Gabius, J. Jimenꢂz-Bar-
bero, J. Am. Chem. Soc. 1999, 121, 8995–9000; b) V. Garcꢀa-Apari-
cio, M. Sollogoub, Y. Blꢂriot, V. Colliou, S. Andrꢂ, J. L. Asensio,
F. J. CaÇada, H.-J. Gabius, P. Sinaꢍ, J. Jimenꢂz-Barbero, Carbohydr.
Res. 2007, 342, 1918–1928.
[22] a) S. Langbein, J. Brade, J. K. Badawi, M. Hatzinger, H. Kaltner, M.
Lensch, K. Specht, S. Andrꢂ, U. Brinck, P. Alken, H.-J. Gabius, His-
Received: October 24, 2012
Published online: December 27, 2012
1902
www.chemeurj.org
ꢇ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1896 – 1902