The inhibition of liposaccharide heptosyltransferase WaaC with multivalent glycosylated fullerenes: A new mode of glycosyltransferase inhibition

Article abstract of DOI:10.1002/chem.201102052

l,d-Heptosides (L-glycero-D-manno-heptopyranoses) are found in important bacterial glycolipids such as lipopolysaccharide (LPS), the biosynthesis of which is targeted for the development of novel antibacterial agents. This work describes the synthesis of

Full text of DOI:10.1002/chem.201102052

FULL PAPER
DOI: 10.1002/chem.201102052
The Inhibition of Liposaccharide Heptosyltransferase WaaC with Multivalent
Glycosylated Fullerenes: A New Mode of Glycosyltransferase Inhibition
Maxime Durka,[a] Kevin Buffet,[a] Julien Iehl,[b] Michel Holler,[b]
Jean-FranÅois Nierengarten,*[b] and Stꢀphane P. Vincent*[a]
Abstract: l,d-Heptosides (l-glycero-d-
manno-heptopyranoses) are found in
important bacterial glycolipids such as
lipopolysaccharide (LPS), the biosyn-
thesis of which is targeted for the de-
velopment of novel antibacterial
agents. This work describes the synthe-
sis of a series of fullerene hexa-adducts
bearing 12 copies of peripheral sugars
displaying the mannopyranose core
structure of bacterial l,d-heptoside.
The multimers were assembled through
an efficient copper-catalyzed alkyne–
azide cycloaddition reaction as the
final step. The final fullerene sugar
balls were assayed as inhibitors of hep-
tosyltransferase WaaC, the glycosyl-
transferase catalyzing the incorporation
of the first l-heptose into LPS. Inter-
estingly, the inhibition of the final mol-
ecules was found in the low micromo-
lar range (IC50 =7–45 mm), whereas the
corresponding monomeric glycosides
displayed high micromolar to low milli-
molar inhibition levels (IC50 always
above 400 mm). When evaluated on a
“per-sugar” basis, these inhibition data
showed that, in each case, the average
affinity of a single glycoside of the ful-
lerenes towards WaaC was significantly
enhanced when displayed as a multi-
mer, thus demonstrating an unexpected
multivalent effect. To date, such a mul-
tivalent mode of inhibition had never
Keywords: cell wall · fullerenes ·
heptose · inhibitors · multivalence ·
virulence
been evidenced with glycosyltrans
AHCTUNGTERGaNNUN ses.
ACHTUNGTRENNUNGfer-
Introduction
ogy.[5] Importantly, the high local concentration of carbohy-
drates around the C60 core in such derivatives has been
found perfectly suited to the binding of lectins[7–8] and dra-
matic multivalent effects have been observed in some
cases.[8] On the other hand, surprising effects have been evi-
denced for the inhibition of carbohydrate-processing en-
zymes with a fullerene hexakis-adduct decorated with 12
iminosugar residues.[9] Specifically, multivalent effects have
been observed with isomaltase of bakerꢀs yeast and a-man-
nosidase of Jack bean with Ki values improved by two and
three orders of magnitude, respectively, over the corre-
sponding monomeric iminosugar. This first evidence of dra-
matic multivalent effects in glycosidase inhibition not only
opens a new field of research, but also reveals that the mul-
tivalent approach may be an appealing tool for innovative
drug discoveries. As part of this research, we now report the
synthesis of dodecaglycosylated fullerenes and their evalua-
tion as a new class of multivalent inhibitors of a biologically
relevant bacterial glycosyltransferase (GT). Indeed, the bio-
synthesis of complex glycans involves the participation of a
huge number of GTs that catalyzes the regio- and stereose-
lective transfer of a saccharide from a donor (usually a nu-
cleotide sugar) to an acceptor (mainly a “growing” oligosac-
charide, a lipid or a protein; Figure 1).
Fullerene derivatives substituted with sugar residues, that is,
glycofullerenes, exhibit a combination of interesting proper-
ties related to water solubility and biological relevance.[1–2]
Most of the glycofullerenes reported to date[3–4] are however
amphiphilic molecules, with a hydrophobic fullerene subunit
occupying a part of the outer architecture of the molecule.
As a result, low solubility in aqueous media and/or aggrega-
tion phenomena remain major limitations for some biologi-
cal applications with such compounds. Recently, we have
shown that the amphiphilic character of glycofullerenes can
be avoided[5] by taking advantage of the unique globular
structure of fullerene hexakis-adducts with a Th-symmetrical
octahedral addition pattern.[6] These glycoclusters display
twelve sugar residues on their periphery in a globular topol-
[a] Dr. M. Durka, K. Buffet, Prof. S. P. Vincent
Chemistry Department, University of Namur (FUNDP)
rue de Bruxelles 61, 5000 Namur (Belgium)
Fax : (+32)81-72-45-17
E-mail: stephane.vincent@fundp.ac.be
[b] Dr. J. Iehl, Dr. M. Holler, Dr. J.-F. Nierengarten
Laboratoire de Chimie des Matꢁriaux Molꢁculaires
Universitꢁ de Strasbourg et CNRS (UMR 7509)
Ecole Europꢁenne de Chimie, Polymꢂres et Matꢁriaux (ECPM)
25 rue Becquerel, 67087 Strasbourg Cedex 2 (France)
Fax : (+33)368-85-27-74
Many of these enzymes are involved in key biological pro-
cesses such as cell adhesion and recognition, signaling, and
some of them represent attractive targets for the develop-
ment new treatments of various pathologies such as cancer,
inflammation and infectious diseases.[10–12] Given their bio-
logical significance, the discovery of GT inhibitors[13] as well
E-mail: nierengarten@unistra.fr
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201102052.
Chem. Eur. J. 2012, 18, 641 – 651
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
641

Figure 1. Top: Heptosylation of Lipid A-Kdo2 catalyzed by heptosyltransferase WaaC. Bottom: Partial view of the crystallographic structure of heptosyl-
transferase WaaC complexed with ADP-2F-heptose (pdb 2H1H) highlighting the proximity of the heptose unit of the substrate with the amine of lysine
residues (Lys 192) and the presence of cationic residues in the binding pocket (Lys98, Arg120, Arg 60, Arg189).
as the understanding of their intimate mechanism[14] are of
major importance but still represent challenging tasks.[15]
Indeed, very potent inhibitors have been only obtained in a
few cases which is in deep contrast with other glycosyl-
processing enzymes or lectins for which many low nanomo-
lar or even picomolar inhibitors have been developed.[16]
Taking into account the medicinal relevance of GTs,
novel strategies to discover and develop potent inhibitors
are clearly needed.
Results and Discussion
The targeted glycosyltransferase and design of the multiva-
lent inhibitors: Lipopolysaccharide (LPS) is a key compo-
nent of the outer membrane of Gram-negative bacteria.[17]
From
a pharmaceutical viewpoint, its biosynthesis has
become a very important research field if one considers that
the mortality of many infectious diseases is closely related
to the amount of circulating LPS found in patient sera.[18]
LPS is an amphipathic molecule that can be decomposed
into three main substructures: lipid A, the oligosaccharide
core, and the O-antigen. The oligosaccharide core unit can
itself be divided into two parts: the inner core, which is
642
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 641 – 651

A New Mode of Glycosyltransferase Inhibition
FULL PAPER
formed at least by one molecule of 3-deoxy-a-d-manno-oct-
2-ulopyranosonic acid (Kdo) and two molecules of l-
glycero-a-d-manno-heptopyranose (heptose), and the outer
core which is composed of hexoses. Lipid A and one Kdo is
the minimal structure for maintaining cell viability. Gram-
negative bacteria lacking the heptose units display the deep-
rough phenotype[19] and show a reduction in outer mem-
brane protein content, an increased sensitivity towards de-
tergents or hydrophobic antibiotics and are much more sus-
ceptible to phagocytosis by macrophages.[20]
Therefore, the heptosyltransferases implied in the LPS
biosynthesis represent attractive targets, the inhibition of
which could attenuate the virulence of diverse bacterial
strains. As depicted in Figure 1, heptosyltransferase WaaC
catalyzes the incorporation of the first heptosyl-subunit onto
the growing core moiety of LPS.[21]
To provide a proof-of-principle that fullerene sugar balls
would be inhibitors of WaaC, we have decided to decorate
the fullerene core with l-heptose moieties (Figure 2, mole-
cules C60(A)12 and C60(B)12) sharing the d-mannopyranose
substructure found in ADP-heptose, the donor substrate of
WaaC or in the product LipidA-Kdo2-heptose. As outlined
in Figure 1, WaaC is an inverting glycosyltransferase: the
donor substrate being b-configured, thus the product is an
a-heptoside. Having in hand the three a-mannosides 1–3
and their corresponding fullerenes,[7] we decided to synthe-
size a-configured heptosides and octosides to compare a ho-
mogeneous series and evaluate whether WaaC could be in-
hibited in a multivalent fashion.
Figure 2. Targeted multimeric glycofullerenes and their corresponding
control monomers.
Based on the knowledge of the contacts between the hep-
tose residue and the enzyme deduced from the crystallo-
graphic structure of heptosyltransferase WaaC in a complex
with ADP-2F-heptose (Figure 1),[22] we have also decided to
explore the binding properties of an octose bearing an addi-
tional ionic functionality (Figure 2, molecule C60(C)12).
Indeed, the amine of a lysine residue is located at the vicini-
ty of the 6-position of the heptose (Figure 1)[23] and a prop-
erly positioned carboxylate group in the inhibitor structure
might provide additional strong coulombic interactions.
To complete the study, we also assayed compounds
C60(D)12, C60(E)12, and C60(F)12 bearing peripheral mannosyl
subunits. Finally, to define whether these multimers display
a multivalent effect or not, the corresponding monomeric
structures A and C–F were also investigated. Furthermore,
fullerene derivative C60(Ph)12 lacking the carbohydrate resi-
dues was used as a negative control compound for the bio-
logical assays.
Synthesis: The synthesis of glycofullerenes C60ACTHNUTRGEN(UGN A–F)12 relies
on the direct grafting of unprotected sugar derivatives onto
preconstructed fullerene hexa-adduct derivatives under the
copper-catalyzed azide–alkyne cycloaddition (CuAAC) con-
ditions.[5] The mannoside precursors bearing the appropriate
alkyne or azide function allowed the preparation of C
F)12 (compounds 1–3). The preparation of the alkyne build-
ing blocks requested to prepare C60A(A–C)12 is shown in
60ACHTUNGTRENNUNG(D–
CTHUNGTRENNUNG
Scheme 1 and Scheme 2. Among the various synthetic pro-
cedures that have been developed for the construction of l-
heptoses,[24] the most appealing strategy for us was the se-
quence developed by Van Boom and collaborators.[25] Com-
pound 4 (Scheme 1) was obtained by the addition of a sily-
lated Grignard reagent to the corresponding aldehyde pre-
cursor. In the original procedure of Van Boom, compound 5
was prepared from 4 under Fleming–Tamao oxidation condi-
Chem. Eur. J. 2012, 18, 641 – 651
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
643

S. P. Vincent, J.-F. Nierengarten et al.
AHCTUNGTREUNaGNN cACHTUNEGRTNNUNGeACTHNUGTRENtNNGU ate 12 was coupled to propargyl alcohol under standard
conditions to give intermediate 13 in 72% yield. Deacetyla-
tion under Zemplꢁn conditions followed by saponification
of the methyl ester group afforded fully unprotected octo-
side 14 that was further coupled to azide 11 under CuAAC
conditions to yield model compound C.
The preparation of fullerene sugar balls C60ACTHNUTRGEN(UGN A–F)12 is de-
picted in Scheme 3 and Scheme 4. The hexa-substituted full-
erene building blocks 15[27] and 17[28] were prepared accord-
Scheme 1. Synthesis of l,d-heptoside building blocks 9 and 10, and prepa-
ration of model compound A. Reagents and conditions: a) AcOK,
HgACHTUNGTRENNUNG(TFA)2, AcOOH, AcOH, 108C to RT; b) Ac2O, Py, DMAP, RT;
c) Ac2O, AcOH, H2SO4, 08C; d) Pd/C 10%, H2, MeOH/H2O 5:1, RT
then Ac2O, Py, DMAP, RT; e) Propargyl alcohol (n=1) or hexynyl alco-
hol (n=4), BF3·Et2O, CH2Cl2, 08C to RT; f) MeONa, MeOH, RT;
g) CuI, DIEA, DMF, microwave irradiation 808C.
Scheme 3. Preparation of glycocluster C60(F)12 and of model compound F.
Reagents and conditions: a) 3, CuSO4·5H2O, AscNa, DMSO, RT; b) 3,
CuI, DIEA, DMF, microwave irradiation 808C.
Scheme 2. Synthesis of propargyl octoside 14 and preparation of model
compound C. Reagents and conditions: a) Propargyl alcohol, BF3·Et2O,
CH2Cl2, 08C to RT; b) MeONa, MeOH, RT; c) LiOH 1M, RT; d) 11,
CuI, DIEA, DMF, microwave irradiation.
ing to reported procedures. The grafting of azide 3 onto the
hexa-substituted fullerene core was achieved under the
CuAAC conditions we have optimized for the preparation
of fullerene sugar balls.[5] Compound 15 was treated with a
slight excess of azide 3 (13 equiv) in the presence of tetrabu-
tylammonium fluoride (TBAF), CuSO4·5H2O and sodium
ascorbate (AscNa) in DMSO. At the end of the reaction,
the product was precipitated by addition of MeOH, filtered
and extensively washed with MeOH and CH2Cl2. To remove
eventual copper-based impurities remaining from the
CuAAC reaction, glycoconjugate C60(F)12 was further puri-
fied by filtration on a SephadexTM column (H2O), precipitat-
ed with MeOH and dried under high vacuum. Compound
C60(F)12 was thus isolated in 84% yield. Finally, the corre-
sponding model compound (F) was prepared in 67% yield
from 16 and 3.
tions. We found that the scaling-up of the latter step was
highly problematic. Actually, we have recently reported a
modification of the Fleming–Tamao procedure allowing the
synthesis of l,d-heptoside 5 on a multigram scale,[26] this
compound was thus selected as the starting material for the
preparation of alkynylated heptoses 9 and 10 (Scheme 1).
Precursor 5 was engaged in an acetylation of its primary
alcohol function followed by acetolysis, hydrogenolysis, and
peracetylation to afford 6. Subsequent glycosylations with
propargyl alcohol and 5-hexynyl alcohol gave alkynes 7 and
8, respectively. Subsequent deacetylation yielded the two
unprotected heptosides 9 and 10, ready for ligation to the
fullerene core. The copper-catalyzed cycloaddition of azide
11 with alkyne 9 gave model compound A in 65% yield.
The synthesis of the targeted octoside building block 14 was
performed from the known molecule 12[26] (Scheme 2). Per-
The chemical structure of compound C60(F)12 was con-
1
firmed by H and 13C NMR analyses. As shown in Figure 2,
the 13C NMR spectrum of fullerene hexakis-adduct C60(F)12
is in full agreement with its T-symmetrical structure and
shows all the expected signals for the six equivalent malo-
644
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 641 – 651

A New Mode of Glycosyltransferase Inhibition
FULL PAPER
nate addends (Figure 3). Importantly, the signals corre-
sponding to the sp C atoms detected at d=85.6 and
105.0 ppm for precursor 15 are not present anymore for
Scheme 4. Preparation of glycoclusters C
pounds D and E. Reagents and conditions: a) CuSO4·5H2O, AscNa,
DMSO/CH2Cl2/H2O (1:1:1), RT 60A(D–E)12 b) CuSO4·5H2O, AscNa,
60ACTHNUGTREN(UNGN A–C)12 and of model com-
C
C
;
DMSO/CH2Cl2/H2O (1:1:1), microwave irradiation 1008C, C
c) CuI, DIEA, DMF, microwave irradiation 808C.
60ACHTUNGTRENNUNG(A–C)12;
photolysis of azide groups and the nitrene residues thus gen-
erated are then responsible of the intermolecular reactions
leading to polymers. Cycloaddition of the azide group onto
the hexasubstituted fullerene core can also be at the origin
of this slow polymerization process. Upon purification, the
best is to use polyazide 17 for the CuAAC reactions within
the next 24 h to obtain good yields. Reaction of freshly pre-
pared 17 with alkynes 1, 2, 9, 10, and 14 was first attempted
under the classical CuAAC conditions previously optimized
for the preparation of fullerene sugar balls (CuSO4·5H2O/
AscNa in a CH2Cl2/H2O/DMSO (1:1:1) mixture at room
temperature).[5] These conditions were efficient for the prep-
aration of C60(D)12 and C60(E)12 from alkynylated manno-
Figure 3. 13C NMR spectrum of C60(F)12 recorded in [D6]DMSO
(100 MHz, 258C); unambiguous assignment was achieved with the help
of the corresponding DEPT spectrum (N=CH2Cl2).
C60(F)12, whereas two additional resonances are detected at
d=121.7 and 145.6 ppm. The latter signals are attributed to
the sp2 C atoms of the 12 equivalent 1,2,3-triazole subunits.
Finally, only three signals out of the five expected ones are
observed for the fullerene C atoms (d=68.8 for the sp3 C
atom; d=140.7 and 145.0 ppm for the sp2 C atoms). Indeed,
these three signals are reminiscent of those of the three
non-equivalent fullerene C atoms of the hexakis-adduct car-
rying achiral addends (overall Th symmetry). No influence
of the overall symmetry of C60(F)12 (T symmetry) could be
deduced and the two pairs of diastereotopic sp2 C atoms are
pseudo-equivalent. Similar observations have been already
reported for related C60 derivatives.[5]
sides 1 and 2, respectively. In contrast, compounds C60ACHTUNGTRENNUNG(A–
C)12 thus prepared were always polluted with defected by-
products even after one week of reaction. Indeed, their IR
spectra revealed minor traces of azide groups (typical signal
observed at ca. 2092 cmÀ1). Furthermore, as in the case of
building block 17, these compounds became completely in-
soluble after a few days of storage. In other words, the poly-
merization reactions occurred due to the presence of un-
reacted azide subunits.
The preparation of compounds C
(A–E)12 required the
This prompted us to attempt the CuAAC reactions of 17
with 9, 10, and 14 under microwave irradiation. Interesting-
ly, complete functionalization was thus achieved in 2 h. The
use of polyazide 17[28] (Scheme 4). As stated in our previous
reports,[8b,28,29] compound 17 must be handled with special
care owing to its high number of azide residues. Upon evap-
oration, compound 17 has never been extensively dried
under high vacuum and the use of metallic spatula avoided.
Furthermore, this compound has been always prepared on a
small scale (less than 500 mg). It can also be noted that com-
pound 17 cannot be stored in the solid state as slow decom-
position into insoluble polymers occurs. The latter observa-
tion is most probably associated to the slow thermolysis or
IR spectrum of C60ACHTNUGTRNEG(UN A–C)12 prepared under microwave irra-
diation confirmed that no azide residues remain in the final
products (Figure 4). This was further confirmed by the sta-
bility of the prepared compounds. Finally, it is also worth
noting that glycoconjugates C60ACHTNUTRGNEUNG(A–E)12 were all purified by
precipitation followed by filtration on a SephadexTM column
to remove residual copper-based impurities from the
CuAAC reactions. Actually, in most of the cases, two well
Chem. Eur. J. 2012, 18, 641 – 651
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
645

S. P. Vincent, J.-F. Nierengarten et al.
Figure 4. IR spectra of compounds C60(A)12 (A), C60(C)12 (B) and 17 (C).
Figure 5. MALDI TOF mass spectrum of compound C60(A)12. A series of
typical fragments resulting from successive retro-Bingel reactions are ob-
served. Further fragments resulting from the cleavage of a malonic ester
unit followed by a decarboxylation are also observed.
separated fractions were collected upon SephadexTM filtra-
tion. The high molecular weight one was orange and corre-
sponded to the glycofullerene whereas the low molecular
weight one was pale blue and contained the residual copper
salts.
Additionally, model compounds D and E were prepared
from 11 and the corresponding alkylynated mannose deriva-
tives 1 and 2.
corresponds to the hydrolysis of a malonic ester unit fol-
lowed by a decarboxylation, leading to a loss of C14H22N3O9
with respect to the parent peak.[32] Despite the high level of
fragmentation observed in the MALDI-TOF spectrum of
C60(C)12, no significant signals corresponding to defected
products could be observed thus providing definitive conclu-
sion about the monodispersity of the isolated product as al-
ready suggested from its IR spectrum in which no azide resi-
dues could be detected.
The structures of the final compounds C60ACTHNUTRGEN(NUG A–E)12 were
1
confirmed by their H and 13C NMR spectra. Inspection of
1
the H NMR spectra shows the typical signal of the triazole
unit at d=7.81–8.06 ppm. As discussed for C60(F)12, the
13C NMR spectra of fullerene hexakis-adducts C
60ACHTNUGTRENUNG(A-E)12
were particularly helpful to evidence their T-symmetrical
structure. In all the cases, the expected signals for the hexa-
substituted fullerene core and the six equivalent malonate
addends were clearly observed.
Inhibition of WaaC: The inhibition of WaaC by the glycoful-
lerenes and their control monomers was measured by the
assay we previously developed and applied to measure the
IC50 of ADP-2F-heptose (Figure 1).[22,23] In brief, the specific
activity of WaaC was followed by a coupled enzymatic assay
involving pyruvate kinase to transform the ADP reaction
product into ATP. The concentration of the latter was then
monitored either by luminescence using luciferase or by
fluorescence using lactate dehydrogenase. The results are
summarized in Table 1. Interestingly, the glycosylated ful-
The structure of glycoclusters C
firmed by their MALDI-TOF mass spectra. The analysis of
glycoclusters of high molecular weight such as C60A(A–E)12 is
60ACHTUNGTRENN(NUG A–E)12 was also con-
CTHUNGTRENNUNG
generally rather difficult due to high degree of fragmenta-
tion and/or to the formation of matrix adducts.[30] Despite
these problems, the expected molecular ion peak could be
evidenced in the case of compounds C60ACHTUNRGTNEUNG(A–E)12. As a typical
example, the MALDI-TOF mass spectrum of compound
C60(A)12 is shown in Figure 5. In addition to the expected
molecular ion peak at m/z=5331.6 [M+Na]+, a series of
typical fragments resulting from successive retro-Bingel re-
actions[31] are observed ([M+NaÀ(C29H44N6O4)n]+, with n=
1 to 3). This fragmentation pathway is indeed classical for
hexasubstituted fullerene derivatives.[27–29] It is also worth
noting that an additional peak is always associated with the
ones corresponding to [M+NaÀ(C29H44N6O4)n]+. The latter
lerenes C60ACHTUGNTERNNU(G A–F)12 displayed all low micromolar inhibition
levels with IC50ꢀs ranging from 7 to 47 mm (Table 1, entries 1,
2, 5, 8, 10, and 12). By itself, this result is remarkable when
analyzed in the global context of GTs inhibition. Effectively,
the best GT inhibitors bind generally these enzymes with
low mM Kiꢀs, typically in the range of the Km of the donor
substrate.[13]
In the particular case of heptosyltransferase WaaC, the
IC50 of the donor analogue ADP-2F-heptose is 30 mm
646
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 641 – 651

A New Mode of Glycosyltransferase Inhibition
Table 1. Inhibition data.
FULL PAPER
values obtained with the two dodecamannosides C60(D)12
and C60(E)12 with different spacers (Table 1 entries 8 and
10). Moreover, C60(F)12 is the most potent mannosylated in-
hibitor (Table 1, entry 12). Thus, it appears that the incorpo-
ration of an aromatic moiety in the spacer improves the
binding properties.
Entry
IC50
Inhibition
Enhancement
per sugar[d]
[mM]
enhancement
1
2
3
4
C60(A)12
C60(B)12
A
11Æ0.3
47Æ6.6
19%[a]
0%[b]
>45
>11
>3.7
>0.9
10
To assess whether some multivalent effect operates for
the inhibition of heptosyltransferase WaaC with glycofuller-
5
6
C60(C)12
14
6.7Æ2.0
61
5.1
0%[b]
7
8
9
10
11
12
13
14
C
409Æ28
18Æ1.3
19%[a]
36Æ2.1
18%[a]
9.5Æ0.3
8%[a]
enes C60ACHTNUGTRENUNG(A–F)12, the binding properties of the corresponding
C60(D)12
D
C60(E)12
E
C60(F)12
F
C60(Ph)12
>28
>14
>53
–
>2.3
>2.2
>4.4
–
monomeric triazoles A to F (Figure 2) were evaluated. For
comparison purposes, heptoside 10 and octoside 14 bearing
an alkyne functionality were also assessed. The model com-
pounds A–F bear an additional arm functionalized with a
pivaloate ester to mimic as closely as possible the attach-
ment to the fullerene core. The monomeric structures A–F
only weakly inhibited WaaC. Interestingly, the two alkyny-
lated derivatives 10 and 14 did not display any inhibition
properties at concentrations up to 500 mm (Table 1, entries 4
and 6). The latter observation suggests an active role of the
aromatic triazole subunit in the recognition process, this is
also in line with the fact that C60(F)12 possessing an aromatic
subunit in the linker is the most potent mannosylated inhibi-
tor (see below). The IC50 of the best monomeric inhibitor,
octoside C, is 409 mm, a value 61-fold higher than for the cor-
responding multivalent fullerene derivative C60(C)12, which
corresponds to a 5.1 enhancement factor of inhibition per
octose residue. For the other monomeric structures, the IC50
values were too high to be accurately measured under our
experimental conditions. Therefore, we only report here the
inhibition percentages measured at an inhibitor concentra-
tion of 500 mm and only a minimal value of the enhancement
factors for the multimers on a “per-sugar” basis could be es-
timated. Globally, a clear multivalent effect is observed for
all the fullerene derivatives with binding affinities enhanced
by at least one order of magnitude when compared with the
corresponding monomeric derivatives. Such an enhancement
resulting from the multivalent presentation of carbohydrates
around a central core structure had never been observed or
even evidenced for glycosyltransferase inhibition.
In contrast to lectins, the inhibition of enzymes by multi-
valent inhibitors has been poorly explored to date. The only
reported examples are multimeric iminosugars that have
been assayed against glycosidases. In the two first pioneer-
ing studies, a nojirimycine[33] and a azafagomine tetramer[34]
were found to be micromolar glycosidase inhibitors but a
multivalent inhibition effect could not be evidenced. A simi-
lar result was later obtained on the enzymes involved in the
glucosylceramide biosynthesis with a dimeric nojirimyci-
ne,[35a] and against E. coli b-galactosidase by multimeric thio-
galactosides.[35b] Importantly, Gouin and Kovensky described
the synthesis of a trimeric iminosugar that displayed a low
micromolar inhibition against Jack bean a-mannosidase and,
for the first time, an inhibition enhancement factor of 2.5
per iminosugar unit.[36] More recently, Compain and Nieren-
garten described a fullerene iminosugar ball that was as-
sayed against a range of glycosidases and displayed a rela-
tive inhibition potency of 179 compared with the monomeric
0%[c]
[a] Inhibition percentage at 500 mm. [b] No inhibition at 500 mm. [c] No in-
hibition at 250 mm. For solubility reasons, this fullerene could not be as-
sayed at 500 mm. [d] Calculated on
(IC50fullerene/12) with IC50monomer=500 when this value could not be
measured above 500 mm.
a per-sugar basis=IC50monomer/
(Figure 1). Moreover, the most potent GT inhibitors are
always charged species that provide strong coulombic inter-
actions with the protein. In our case, the grafting of neutral
monosaccharides (except for fullerene C60(C)12) is sufficient
to allow a tight binding with the transferase. As a negative
control, we measured the inhibition of WaaC by fullerene
C60(Ph)12 bearing phenyl rings in place of carbohydrates
(Table 1, entry 14). No inhibition could be observed within
the solubility limits of this molecule, thus showing that the
origin of the inhibition of WaaC by the glycofullerenes does
not originate from the fullerene core structure.
The peripheral residues of glycofullerenes C60(A)12,
C60(C)12, and C60(D)12 are all mannopyranosides attached to
the central core through the same spacer. The only structur-
al variation between these molecules is the nature of the
substituent at the 5-position: CH2OH (C60(D)12), CH(OH)-
CH2OH (C60(A)12) and CH(OH)CH(OH)CO2H (C60(C)12).
The observed decrease in IC50 along the series when going
from the hexose derivative to the octose one thus demon-
strates the beneficial effect of additional interactions of the
substituent at the 5-position of the carbohydrate unit on
WaaC inhibition. Octoside C60(C)12 displayed the best inhib-
ition profile of the whole series (IC50 =6.7 mm, Table 1,
entry 5). This result was anticipated based on the X-ray crys-
tal structure of the enzyme. As illustrated in Figure 1, sever-
al cationic residues (lysines 98 and 192 and arginines 60,
120, and 189) are located at the vicinity of the binding
pocket thus allowing positive electrostatic interactions with
the carboxylate groups of C60(C)12. Overall, if one considers
that these molecules are a-configured, the observed inhibi-
tion might be related to product inhibition (see Figure 1).
A comparison of the inhibition properties of the two mul-
tivalent heptosides C60(A)12 and C60(B)12 bearing the same
peripheral heptose units (Table 1, entries 1 and 2) indicates
that the fullerene with the shortest spacer is a better inhibi-
tor. A similar trend is observed by comparing the IC50
Chem. Eur. J. 2012, 18, 641 – 651
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
647

S. P. Vincent, J.-F. Nierengarten et al.
structure, also against Jack bean a-mannosidase.[9] These lit-
erature data clearly indicate that all enzymes are not prone
to multivalent inhibition and that the nature of both the
central core structure and the spacer unit tethering the in-
hibitory monomers play critical roles in the inhibition pro-
cess. The inhibition data presented in this study demonstrate
that heptosyltransferase WaaC is the second enzyme ever
reported for which significant levels of multivalent inhibi-
tion have been measured. In contrast to many lectins that
are themselves multimeric and for which multimeric inhibi-
tors were found extremely potent thanks to a “cluster
effect”, multivalent enzyme inhibition is much more difficult
to rationalize.[16h,37] WaaC is a monomeric enzyme that has
the specific property to have a glycolipid (Lipid A-Kdo2) as
the acceptor substrate (Figure 1). Due to the lipidic nature
of the substrate, the reaction takes place at a surface: either
of a vesicle for in vitro experiments or at the cytoplasmic
inner membrane (in vivo). Therefore, clustering or local
concentration effects might explain why multimeric struc-
tures such as glycofullerenes do inhibit WaaC in a multiva-
lent manner. At the molecular level, inspection of the struc-
ture of WaaC in complex with an inhibitor shows that the
binding pockets (donor and acceptor) are buried in a large
open cleft that could allow quick rebinding processes with
heptosides presented in the multimeric fashion. However,
given the complexity of both the glycosyltransferase and the
fullerenes, it is not reasonable, at this stage, to speculate on
a specific binding mode. A novel series of multimeric glyco-
sides, especially b-glycosides, will certainly be helpful to de-
lineate the binding strength and selectivity of the sugar balls
with WaaC.
spectroscopic apparatus used for the characterization of the compounds
is given in the Supporting Information.
Synthesis of compounds: Compounds C60(Ph)12,[28] 1,[38] 2,[39] 3,[7] 4,[26] 5,[26]
11,[7] 12,[26] and 15[27] were prepared according to previously reported pro-
cedures, whereas the preparation of compounds 6, 7, 8, 9, 10, 13, and 14
is given in the Supporting Information.
Compound A: A mixture of 9 (15 mg, 60.7 mmol), 11 (13 mg, 72.8 mmol),
diisopropyl ethylamine (DIEA, 16 mg, 124 mmol) and CuI (5 mg,
0.26 mmol) in dry DMF (120 mL) was heated for one hour under micro-
wave irradiation (808C). The solvent was evaporated under reduced pres-
sure. Column chromatography (SiO2, EtOAc to EtOAc/MeOH 9:1) gave
1
A (17 mg, 65%) as a yellow oil. [a]2D0 = +25.9 (MeOH, c=0.5). H NMR
(400 MHz, D2O): d=7.94 (s, 1H), 4.83 (s, 1H), 4.65 (AB, J=12 Hz, 1H),
4.53 (AB, J=12 Hz, 1H), 4.43 (t, J=6 Hz, 2H), 3.98 (t, J=6 Hz, 2H),
3.90 (app t, J=6.5 Hz, 1H), 3.77 (app brs, 1H), 3.73 (app t, J=10 Hz
,1H), 3.64 (dd, J=10 and 3 Hz, 1H), 3.55 (dd, J=11 and 8 Hz, 1H), 3.49
(m, 2H), 2.18 (quint., J=6 Hz, 2H), 0.98 ppm (s, 9H). 13C NMR
(100 MHz, D2O): d=181.8, 143.6, 125.3, 99.6, 71.4, 70.7, 70.0, 68.7, 66.0,
62.8, 62.7, 59.8, 48.2, 38.5, 28.3, 26.2 ppm. MS (ESI): m/z: 456.2
AHCTUNGTRENNUNG
[M+Na]+; HRMS (ESI): calcd for C18H31N3O9Na [M+Na]+: 456.1953;
found: 456.1970.
Compound C: As described for A, with 11 (12 mg, 0.07 mmol) and 14
(15 mg, 0.06 mmol). Column chromatography (SiO2, EtOAc to EtOAc/
EtOH 7:3, 0.05% NEt3) afforded the triethylammonium salt of C (19 mg,
79%) as a yellow oil. [a]2D0 = +9.3 (MeOH, c=0.8); 1H NMR (400 MHz,
D2O): d=7.98 (s, 1H), 4.83 (brs, 1H), 4.74–4.53 (m, 2H), 4.43 (t, J=
6 Hz, 2H), 3.97 (t, J=5.5 Hz, 2H), 3.77–3.66 (m, 4H), 3.05 (q, J=7 Hz,
2H), 2.18 (tt, J=5.5 and 6 Hz, 2H), 1.13 (t, J=7 Hz, 3H), 0.95 ppm (s,
9H); 13C NMR (101 MHz, D2O): d=181.8, 99.0, 71.6, 70.5, 69.7, 69.1,
62.7, 59.6, 58.8, 48.18, 46.7, 38.5, 28.2, 26.2, 8.24 ppm; MS (ESI+): m/z
(%): 478.2 [M+H]+ (100), 500.2 [M+Na]+ (95), HRMS (ESI+): calcd
for C19H32O11N3 [M+H]+: 478.2031; found: 478.2050.
Compound D: As described for A, with 1 (22 mg, 0.1 mmol, 1 equiv) and
11 (18 mg, 0.12 mmol, 1.2 equiv). Column chromatography (SiO2, EtOAc
to EtOAc/EtOH 8:2) gave D (14 mg, 35%) as a yellow oil. [a]2D0 (MeOH,
c=0.5)= +26.2. 1H NMR (400 MHz, D2O): d=7.95 (s,1H), 4.82 (d, J=
1.5 Hz, 1H), 4.62 (AB, J=12 Hz, 2H), 4.43 (t, J=6 Hz, 2H), 3.97 (t, J=
5.5 Hz, 2H), 3.78 (dd, J=1.5 and 3 Hz, 1H), 3.71 (m, 1H), 3.64–3.58 (m,
1H), 3.62 (m, 1H), 3.52 (t, J=10 Hz, 1H), 3.54–3.48 (m, 1H), 2.17
(quint., J=6 Hz, 2H), 0.95 ppm (s, 9H); 13C NMR (100 MHz, D2O): d=
181.8, 125.4, 99.5, 72.9, 70.5, 69.9, 66.7, 62.6, 60.8, 59.7, 48.1, 38.5, 28.2,
26.18 ppm. MS (TOF-MS-ES+): m/z (%) 242.1 [MÀC6H11O6 +NH4]+
(100), 404.2 [M+H]+ (88). HRMS: calcd for C17H30N3O8 [M+H]+:
404.2027, found: 404.2041.
Conclusion
Although the mechanism by which WaaC is inhibited by
these multimeric structures remains obscure, we are con-
vinced that these results are significant because glycosyl-
transferases, in general, constitute a fantastic class of targets
for drug development, and, in particular, because the glyco-
syltransferases of the LPS biosynthesis have been over-
looked for the discovery of novel antibacterial agents. This
work paves the way to novel generations of glycosyltransfer-
ase inhibitors for systematic studies on the role of the cen-
tral core and on the peripheral ligands. Work in this direc-
tion is under progress in our laboratories.
Compound E: As described for A, with 2 (15 mg, 0.058 mmol, 1 equiv)
and 11 (12 mg, 0.069 mol, 1.2 equiv). Column chromatography (SiO2,
EtOAc to EtOAc/EtOH 9:1) gave E (15 mg, 58%) as a yellow oil. [a]D20
(MeOH, c=0.2)= +4.7. 1H NMR (400 MHz, D2O): d=7.73 (s,1H), 4.74
(s, 1H), 4.29 (t, J=6 Hz, 2H), 3.99 (t, J=5.5 Hz, 2H), 3.82 (m, 1H), 3.75
(d, J=12 Hz, 1H), 3.69–3.62 (m, 3H), 3.57–3.43 (m, 3H), 2.65 (t, J=
7 Hz, 2H), 2.20 (m, 2H), 1.67–1.51 (m, 4H), 1.00 ppm (s, 9H). 13C NMR
(100 MHz, D2O): d=181.7, 99.7, 72.7, 70.6, 70.1, 67.4, 66.7, 62.6, 60.9,
48.0, 38.5, 28.2, 27.9, 26.2, 25.2, 24.3 ppm. MS (TOF-MS-ES+): m/z (%):
468.23 [M+Na]+ (100), 913.47 [2M+Na]+ (21); HRMS: calcd for
C20H35N3O8Na): 468.2316 [M+Na]+; found: 468.2292.
Compound F: As described for A, with 3 (67 mg, 0.15 mmol, 1 equiv)
and 16 (63 mg, 0.37 mmol). Column chromatography (SiO2, EtOAc to
EtOAc/EtOH 85/15) gave F (41 mg, 67%) as a yellow oil. [a]2D0 (MeOH,
c=1.2)= +20.2. 1H NMR (400 MHz, CD3OD): d=7.74 (s, 1H), 7.36 (d,
J=8.5 Hz, 2H), 7.31 (d, J=8.5 Hz, 2H), 4.81 (d, J=1.5 Hz, 1H), 4.61
(AB, J=12 Hz, 2H), 4.37 (t, J=7 Hz, 2H), 4.15 (s, 2H), 4.07 (t, J=6 Hz,
2H), 3.82 (m, 1H), 3.81 (dd, J=2 and 3 Hz, 1H), 3.70 (dd, J=4 and
9 Hz, 1H), 3.67 (m, 1H), 3.60 (brt, J=10 Hz, 1H), 3.54 (ddd, J=2, 6 and
10 Hz, 1H), 2.74 (t, J=8 Hz, 2H), 2.26 (t, J=7 Hz, 2H), 2.00–1.88 (m,
4H), 1.61 ppm (quint., J=7 Hz, 2H), 1.17 (s, 9H). 13C NMR (100 MHz,
CD3OD): d=178.7, 173.8, 138.2, 131.3, 127.7, 122.2, 99.5, 84.9, 81.8, 73.7,
71.3, 70.8, 68.0, 67.3, 63.4, 61.6, 49.6, 34.6, 29.4, 28.9, 28.2, 26.2, 22.3,
21.6 ppm. MS (TOF-MS-ES+): m/z (%): 617.32 [M+H]+ (100), 639.30
Experimental Section
General techniques: Reagents and solvents were purchased as reagent
grade and used without further purification. All reactions were per-
formed using purified and dried solvents: tetrahydrofuran (THF) was re-
fluxed over sodium-benzophenone, dichloromethane (CH2Cl2), triethyla-
mine (NEt3), and pyridine were refluxed over calcium hydride (CaH2).
All reactions were monitored by thin-layer chromatography (TLC) car-
ried out on Merck aluminum roll silica gel 60-F254 using UV light and a
molybdate-sufuric acid solution as revelator. Detailed description of the
648
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 641 – 651

A New Mode of Glycosyltransferase Inhibition
FULL PAPER
[M+Na]+ (100); HRMS: calcd for C31H45N4O9 [M+H]+: 617.3181,
found: 617.3189.
Na]+, 3921.3 [MÀC39H60N9O24 +Na]+, 3562.0 [MÀC54H80N12O32 +Na]+,
3215.9 [MÀC67H100N15O40 +Na]+, 2856.7 [MÀC81H120N18O48 +Na]+.
Compound
C60(E)12: As described for C60(D)12, with 17 (295 mg,
Compound C60(A)12: A mixture of 17 (166 mg, 0.071 mmol), 9 (230 mg,
0.926 mmol), CuSO4·5H2O (1 mg, 0.007 mmol) and sodium ascorbate
(4 mg, 0.021 mmol) in CH2Cl2/H2O/DMSO (1:1:1, 3 mL) was heated 2 h
under microwave irradiation (1008C). The product was precipitated by
addition of acetone (20 mL). The precipitate was washed with MeOH,
then CH2Cl2 and dried under reduced pressure. Compound C60(A)12 was
further purified by filtration on a SephadexTM column (H2O), precipitated
with MeOH and dried under high vacuum in order to totally remove
eventual traces of residual copper-based impurities from the CuAAC re-
actions. Compound C60(A)12 (169 mg, 45%) was thus obtained as a red
0.127 mmol) and 2 (360 mg, 1.65 mmol). Compound C60(E)12 (509 mg,
81%) was thus obtained as a red–orange powder. 1H NMR ([D6]DMSO,
300 MHz):d=7.81 (s, 12H), 4.70 (m, 24H), 4.55 (m, 12H), 4.43 (m,
12H), 4.33 (m, 60H), 3.62 (m, 24H), 3.40 (m, 24H partially masked by
the H2O signal), 2.59 (m, 24H), 2.19 (m, 24H), 1.55 ppm (m, 48H).
13C NMR ([D6]DMSO, 75 MHz): d=162.6, 146.9, 145.0, 140.6, 121.7,
99.7, 73.8, 71.0, 70.3, 68.6, 67.0, 65.9, 64.3, 61.2, 48.5, 45.9, 45.2, 28.5, 25.7,
À1
À
24.7 ppm. IR (neat): n˜ =3332 (O H), 1740 cm (C=O). UV/Vis (H2O):
247 (79700), 269 (64300), 283 (58700), 320 (sh, 37200), 337 nm (sh,
29200 molÀ1 dm3 cmÀ1). MALDI-TOF MS: 5476.8 [M+Na]+, 5088.0
1
solid. H NMR (300 MHz, [D6]DMSO): d=8.02 (s, 12H), 4.61 (m, 24H),
[MÀC16H26N3O8 +Na]+,
4688.3
[MÀC33H52N6O16 +Na]+,
4299.9
4.49 (24H), 4.07 (m, 12H), 4.61 (m, 48H), 4.49 (m, 48H), 4.07 (m, 12H),
3.35 (m, 36H), 3.13 (m, 24H), 2.04 ppm (m, 24H). 13C NMR (100 MHz,
[D6]DMSO): d=162.7, 145.2, 143.6, 140.6, 124.2, 98.9, 71.2, 70.1, 68.6,
[MÀC49H78N9O24 +Na]+, 3898.5 [MÀC66H104N12O32 +Na]+, 3511.2
[MÀC82H130N15O40 +Na]+), 3110.0 [MÀC99H156N18O48 +Na]+, 3721.8
[MÀC115H182N21O56 +Na]+.
À
65.9, 64.3, 62.5, 58.8, 46.1, 28.6 ppm. IR (neat): n˜ =3340 (O H),
1739 cmÀ1 (C=O). UV/Vis (H2O): 232 (66000), 326 (sh, 24000), 379 nm
(sh, 9500 molÀ1 dm3 cmÀ1). MALDI-TOF MS: 5331.6 [M+Na]+, 4956.3
Compound
C60(F)12: A 1m solution of TBAF in THF (0.37 mL,
0.37 mmol) was added to a mixture of 15 (80 mg, 0.027 mmol), 3 (156 mg,
0.35 mmol), CuSO4·5H2O (0.4 mg, 0.003 mmol) and sodium ascorbate
(1.6 mg, 0.008 mmol) in CH2Cl2/MeOH/DMSO (1:1:1, 3 mL). The result-
ing mixture was stirred at room temperature. After 24 h, methanol
(10 mL) was added to the mixture and the resulting orange precipitate
filtered, extensively washed with methanol then CH2Cl2 and dried under
high vacuum. Compound C60(F)12 was further purified by filtration on a
SephadexTM column (H2O), precipitated with MeOH and dried under
high vacuum to totally remove eventual traces of residual copper-based
impurities from the CuAAC reactions. Compound C60(F)12 (168 mg,
[MÀC14H22N3O9 +Na]+,
4566.2
[MÀC29H44N6O18 +Na]+,
4190.9
([MÀC43H66N9O24 +Na]+), 3801.7 [MÀC58H88N12O36 +Na]+, 3426.5
[MÀC72H110N15O45 +Na]+, 3036.3 [MÀC87H132N18O54 +Na]+, 2662.1
[MÀC101H154N21O63 +Na]+.
Compound C60(B)12
: As described for C60(A)12, with 17 (100 mg,
0.043 mmol) and 10 (162 mg, 0.56 mmol). Compound C60(B)12 (122 mg,
49%) was thus obtained as a red–orange solid. 1H NMR (300 MHz,
[D6]DMSO): d=7.81 (s, 12H), 4.61 (m, 12H), 4.30 (m, 72H), 3.76 (m,
24H), 3.62 (m, 48H), 3.35 (m, 96H), 2.18 (m, 24H), 1.55 ppm (m, 48H).
13C NMR (100 MHz, [D6]DMSO): d=162.7, 146.9, 145.0, 140.6, 121.7,
99.7, 71.3, 71.2, 70.7, 70.4, 68.7, 65.9, 65.8, 64.3, 62.6, 48.5, 46.4, 45.9, 28.5,
25.7, 24.8 ppm. IR (neat): n˜ =3362 (O-H), 1740 cmÀ1 (C=O). UV/Vis
1
84%) was thus obtained as a red solid. H NMR ([D6]DMSO, 300 MHz):
d=8.33 (br s, 12H), 7.82 (br s, 12H), 7.26–7.46 (m, 48H), 4.40–4.80 (m,
84H), 4.20–4.38 (m, 36H), 4.04–4.16 (m, 24H), 3.60–3.75 (m, 24H), 3.47–
3.57 (m, 24H), 3.33–3.45 (m, 36H), 2.55–2.75 (m, 24H), 2.10–2.20 (m,
24H), 1.88–2.00 (m, 24H), 1.70–1.85 (m, 24H), 1.40–1.55 ppm (m, 24H).
13C NMR ([D6]DMSO, 100 MHz): d=171.5, 162.7, 145.6, 145.0, 140.7,
138.3, 131.2, 127.7, 121.7, 121.3, 99.1, 87.0, 81.4, 74.2, 71.0, 70.2, 68.8, 67.1,
(H2O):
232
(64000),
326
(sh,
22000),
379 nm
(sh,
8500 molÀ1 dm3 cmÀ1).MALDI-TOF MS: 5851.5 [M+K]+, 5434.0
[MÀC17H28N3O9 +K]+,
[MÀC35H56N6O18 +K]+,
[M-C70H112N12O36 +K]+,
[M-C105H168N18O54 +K]+.
5419.9
4987.8
3736.9
[MÀC17H28N3O9 +Na]+,
[MÀC35H56N6O18 +Na]+,
[M-C87H139N15O45 +K]+,
5002.7
4154.1
3305.6
67.0, 66.5, 61.2, 48.9, 45.6, 40.3, 34.3, 29.2, 28.5, 27.6, 22.0, 21.3 ppm. IR
À1
À
ꢀ
(neat): n˜ =3385 (O H), 2129 (C C), 1740 (C=O), 1648 (C=O) cm
.
Enzymatic assays: For determination of specific activity, WaaC activity
was monitored by a coupled assay involving pyruvate kinase (PK) and lu-
ciferase as coupling enzymes.[22,23] Basically, the reaction product ADP
was converted by PK to ATP, which was detected by luciferase activity.
The reaction took place in a white 96-well plate (Costar) in a final
volume of 60 mL. The assay buffer contained Hepes (50 mm, pH 7.5),
MgCl2 (10 mm), KCl (50 mm), dithiothreitol (1 mm), myelin basic protein
(0.1 mm; Sigma) and Triton X-100 (0.1% (v/v)), which generated support-
ing micelles for Re-LPS. The pre-incubation mixture contained 3 mL of
inhibitors (at eight different concentrations) in DMSO and of WaaC
(27 mL; 0.1 nm final concentration). After 30 min of preincubation at
room temperature, ADP-heptose (1 mm) and Re-LPS (1 mm; Sigma, puri-
fied from Salmonella minnesota) were added and the resulting reaction
mixture (60mL) was incubated at room temperature until a 30% conver-
sion rate was reached. The following readout mixture (100 mL) was final-
ly added in each well: pyruvate kinase (5 unitsmLÀ1, Sigma), phospho-
Compound
0.038 mmol) and 14 (147 mg, 0.502 mmol). Compound C60(C)12 (156 mg,
69%) was thus obtained as
dark red solid. 1H NMR (300 MHz,
[D6]DMSO): d=8.01 (s, 12H), 4.66 (m, 12H), 4.45 (m, 12H), 4.37 (m,
36H), 4.29 (m, 12H), 3.95 (m, 12H), 3.80 (m, 72H, partially masked by
the H2O signal), 2.22 ppm (m, 24H). 13C NMR (100 MHz, [D6]DMSO):
d=175.1, 163.2, 145.6, 144.8, 141.1, 124.7, 99.5, 75.7, 72.2, 71.5, 71.0, 70.4,
C60(C)12: As described for C60(A)12, with 17 (90 mg,
a
À
69.9, 64.4, 59.4, 46.6, 45.7, 29.2 ppm. IR (neat): n˜ =3355 (O H),
1736 cmÀ1 (C=O). UV/Vis (H2O): 232 (65000), 326 (sh, 24000), 379 nm
(sh, 8900 molÀ1 dm3 cmÀ1). MALDI-TOF MS: 5860.0 [M+Na]+, 5439.8
[MÀC15H22N3O11 +Na]+,
[MÀC46H66N9O33 +Na]+,
[MÀC77H110N15O55 +Na]+.
5007.6
4154.5
[MÀC31H44N6O22 +Na]+,
4586.7
3734.4
[MÀC62H88N12O44 +Na]+,
Compound C60(D)12: A mixture of 17 (176 mg, 0.072 mmol), 1 (203 mg,
0.93 mmol), CuSO4·5H2O (1 mg, 0.007 mmol) and sodium ascorbate
(4 mg, 0.021 mmol) in CH2Cl2/H2O/DMSO (1:1:1, 3 mL) was stirred at
RT for 96 h. Methanol (15 mL) was added to the mixture and the result-
ing orange precipitate filtered, extensively washed with methanol then
CH2Cl2 and dried under high vacuum. Compound C60(D)12 was further
purified by filtration on a SephadexTM column (H2O), precipitated with
MeOH and dried under high vacuum to totally remove eventual traces of
residual copper-based impurities from the CuAAC reactions. Compound
C60(D)12 (261 mg, 73%) was thus obtained as a red–orange powder.
1H NMR ([D6]DMSO, 300 MHz): d=8.06 (s, 12H), 4.81 (m, 12H), 4.71
(m, 12H), 4.16 (m, 24H), 4.38 (m, 48H), 3.20 (m, 48H partially masked
by the H2O signal), 2.07 ppm (m, 24H). 13C NMR ([D6]DMSO, 75 MHz):
ACHUTNGRENUeNG nolAHCUTNGTRENpNNGU yruvate (50 mm; Sigma), luciferase (Sigma), 20 mm luciferin
(10000 unitsmLÀ1, Sigma), N-acetylcysteamine (0.1 mm, Aldrich). The lu-
minescence intensity was read on a Luminoskan (Thermo). IC50 curves
were fit to a classical Langmuir equilibrium model using XLFIT (IDBS).
Acknowledgements
d=163.0, 145.3, 143.9, 140.8, 124.4, 99.3, 74.1, 71.0, 70.3, 68.9, 67.1, 64.6,
The authors thank the EU (“Marie Curie Training Network” grant, “Dy-
namic Interactive Chemical Biology & Biomedicine”, MRTN-CT-2005–
019561), the FNRS (FRFC grant 2.4.625.08.F and “Crꢁdit au Chercheur ”
1.5.167.10), the CNRS (UMR 7509) and the University of Strasbourg for
financial support. We also thank V. Oerthel for the initial preparation of
À1
À
61.4, 59.3, 46.5, 46.0, 28.8 ppm. IR (neat): n˜ =3332 (O H), 1740 cm
(C=O). UV/Vis (H2O): 245 (77700), 268 (62200), 282 (57500), 321 (sh,
35800), 341 nm (sh, 26400 molÀ1 dm3 cmÀ1). MALDI-TOF MS: 4971.4
[M+Na]+, 4625.9 [MÀC13H20N3O8 +Na]+, 4266.5 [MÀC27H40N6O16
+
Chem. Eur. J. 2012, 18, 641 – 651
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
649

S. P. Vincent, J.-F. Nierengarten et al.
compound C60(D)12. We further thank M. Schmitt for NMR measure-
ments and Julien Dewinter (UMH) for having recorded the mass spectra.
1989, 264, 17174; f) T. Kajimoto, M. Node, Synthesis 2009, 3179;
g) M. D. Burkart, S. P. Vincent, C.-H. Wong, Chem. Commun. 1999,
1525; h) H. Y. Sun, S. W. Lin, T. P. Ko, J. F. Pan, C. L. Liu, C. N. Lin,
A. H. J. Wang, C. H. Lin, J. Biol. Chem. 2007, 282, 9973; i) K.
Takaya, N. Nagahori, M. Kurogochi, T. Furuike, N. Miura, K.
Monde, Y. C. Lee, S. I. Nishimura, J. Med. Chem. 2005, 48, 6054;
j) L. V. Lee, M. L. Mitchell, S. J. Huang, V. V. Fokin, K. B. Sharpless,
C. H. Wong, J. Am. Chem. Soc. 2003, 125, 9588; k) K. Hosoguchi, T.
Maeda, J. Furukawa, Y. Shinohara, H. Hinou, M. Sekiguchi, H.
Togame, H. Takemoto, H. Kondo, S. I. Nishimura, J. Med. Chem.
2010, 53, 5607; l) T. Pesnot, R. Jorgensen, M. M. Palcic, G. K.
Wagner, Nat. Chem. Biol. 2010, 6, 321; m) N. L. Rose, G. C. Comple-
to, S.-H. Lin, M. McNeil, M. Palcic, T. L. Lowary, J. Am. Chem. Soc.
2006, 128, 6721; n) M. B. Lazarus, Y. Nam, J. Jiang, P. Sliz, S.
Walker, Nature 2011, 469, 564.
[1] Y. M. Chabre, R. Roy, Adv. Carbohydrate Chem. and Biochem.
2010, 63, 165.
[2] For reviews on biological applications of fullerenes, see: a) T.
Da Ros, M. Prato, Chem. Commun. 1999, 663; b) E. Nakamura, H.
Isobe, Acc. Chem. Res. 2003, 36, 807; c) S. Bosi, T. Da Ros, G. Spal-
luto, M. Prato, Eur. J. Med. Chem. 2003, 38, 913; d) A. Bianco, T.
Da Ros in Fullerenes: Principles and Applications (Eds.: F. Langa,
J.-F. Nierengarten), RSC Nanoscience and Nanotechnology Series,
Cambridge, 2007, pp. 301–328.
[3] a) A. Vasella, P. Uhlmann, C. A. A. Waldraff, F. Diederich, C. Thiel-
gen, Angew. Chem. 1992, 104, 1383; Angew. Chem. Int. Ed. Engl.
1992, 31, 1388; b) Y. Wang, J. Cao, D. I. Schuster, S. R. Wilson, Tet-
rahedron Lett. 1995, 36, 6843; c) M. R. Banks, J. I. G. Cadogan, I.
Gosney, P. K. G. Hodgson, P. R. R. Langridge-Smith, J. R. A. Millar,
A. T. Taylor, J. Chem. Soc. Chem. Commun. 1995, 885; d) A. Ya-
shiro, Y. Nishida, M. Ohno, S. Eguchi, K. Kobayashi, Tetrahedron
Lett. 1998, 39, 9031; e) F. Cardullo, F. Diederich, L. Echegoyen, T.
Habicher, N. Jayaraman, R. M. Leblanc, J. F. Stoddart, S. Wang,
Langmuir 1998, 14, 1955; f) H. Kato, A. Yashiro, A. Mizuno, Y.
Nishida, K. Kobayashi, H. Shinohara, Bioorg. Med. Chem. Lett.
2001, 11, 2935; g) A. Dondoni, A. Marra, Tetrahedron Lett. 2002, 43,
1649; h) Y. Mikata, S. Takagi, M. Tanahashi, S. Ishii, M. Obata, Y.
Miyamoto, K. Wakita, T. Nishisaka, T. Hirano, T. Ito, M. Hoshino,
C. Ohtsuki, M. Tanihara, S. Yano, Bioorg. Med. Chem. Lett. 2003,
13, 3289; i) H. Ito, T. Tada, M. Sudo, Y. Ishida, T. Hino, K. Saigo,
Org. Lett. 2003, 5, 2643; j) H. Isobe, H. Mashima, H. Yorimitsu, E.
Nakamura, Org. Lett. 2003, 5, 4461; k) Y. Nishida, A. Mizuno, H.
Kato, A. Yashiro, T. Ohtake, K. Kobayashi, Chem. Biodiversity
2004, 1, 1452; l) H. Kato, N. Kaneta, S. Nii, K. Kobayashi, N. Fukui,
H. Shinohara, Y. Nishida, Chem. Biodiversity 2005, 2, 1232; m) H.
Kato, C. Bçttcher, A. Hirsch, Eur. J. Org. Chem. 2007, 2659; n) S.
Tanimoto, S. Sakai, S. Matsumura, D. Takahashi, K. Toshima, Chem.
Commun. 2008, 5767.
[14] a) J. C. Errey, S. S. Lee, R. P. Gibson, C. M. Fleites, C. S. Barry,
P. M. J. Jung, A. C. OꢀSullivan, B. G. Davis, G. J. Davies, Angew.
Chem. 2010, 122, 1256; Angew. Chem. Int. Ed. 2010, 49, 1234;
b) J. F. May, R. A. Splain, C. Brotschi, L. L. Kiessling, Proc. Nat.
Acad. Sci. USA 2009, 106, 11851; c) L. L. Lairson, B. Henrissat, G. J.
Davies, S. G. Withers, Annu. Rev. Biochem. 2008, 77, 521; d) C.
Breton, L. Snajdrova, C. Jeanneau, J. Koca, A. Imberty, Glycobiolo-
gy 2006, 16, 29R; e) T.-S. A. Wang, T. J. Lupoli, Y. Sumida, H. Tsu-
kamoto, Y. Wu, Y. Rebets, D. E. Kahne, S. Walker, J. Am. Chem.
Soc. 2011, 133, 8528.
[15] a) S. G. Withers, Carbohydr. Polym. 2001, 21, 325; b) L. L. Lairson,
C. P. C. Chiu, H. D. Ly, S. He, W. W. Wakarchuk, N. C. J. Strynadka,
S. G. Withers, J. Biol. Chem. 2004, 279, 28339; c) M. D. Burkart, S. P.
Vincent, A. Dꢆffels, B. W. Murray, S. V. Ley, C.-H. Wong, Bioorg.
Med. Chem. 2000, 8, 1937; d) P. A. Frantom, J. K. Coward, J. S. Blan-
chard, J. Am. Chem. Soc. 2010, 132, 6626.
[16] a) M. Lahmann, Top Curr. Chem. 2009, 288, 17; b) J. J. Lundquist,
E. J. Toone, Chem. Rev. 2002, 102, 555; c) R. J. Pieters, Org. Biomol.
Chem. 2009, 7, 2013; d) M. Touaibia, R. Roy in Application of Mul-
tivalent Mannosylated Dendrimers in Glycobiology, Vol. 3 (Ed.: J. P.
Kamerling), Elsevier, Oxford, 2007, pp. 781–829; e) V. H. Lillelund,
H. H. Jensen, J. Liang, M. Bols, Chem. Rev. 2002, 102, 515; f) S. Ce-
cioni, O. A. Argintaru, T. Docsa, P. Gergely, J. P. Praly, S. Vidal,
New J. Chem. 2009, 33, 148; g) A. Imberty, Y. M. Chabre, R. Roy,
Chem. Eur. J. 2008, 14, 7490; h) L. L. Kiessling, J. E. Gestwicki,
L. E. Strong, Angew. Chem. 2006, 118, 2408; Angew. Chem. Int. Ed.
2006, 45, 2348.
[17] a) S. Gronow, H. Brade, J. Endotoxin Res. 2001, 7, 3; b) A. Silipo, A.
Molinaro, T. Ierano, A. De Soyza, L. Sturiale, D. Garozzo, C. Al-
dridge, P. A. Corris, C. M. A. Khan, R. Lanzetta, M. Parrilli, Chem.
Eur. J. 2007, 13, 3501.
[18] C. M. Kahler, A. Datta, Y.-L. Tzeng, R. W. Carlson, D. S. Stephens,
Glycobiology 2005, 15, 409.
[19] E. T. Rietschel, T. Kirikae, F. U. Schade, U. Mamat, G. Schmidt, H.
Loppnow, A. J. Ulmer, U. Zꢇhringer, U. Seydel, F. Di Padova, M.
Schreier, H. Brade, FASEB J. 1994, 8, 217.
[20] a) W. G. Coleman, Jr., L. Leive, J. Bacteriol. 1979, 139, 899; b) S. Es-
caich, Curr. Opin. Chem. Biol. 2008, 12, 400; c) F. Moreau, N.
Desroy, J. M. Genevard, V. Vongsouthi, V. Gerusz, G. Le Fralliec, C.
Oliveira, S. Floquet, A. Denis, S. Escaich, K. Wolf, M. Busemann,
A. Aschenbrenner, Bioorg. Med. Chem. Lett. 2008, 18, 4022.
[21] A. Zamyatina, S. Gronow, C. Oertelt, M. Puchberger, H. Brade, P.
Kosma, Angew. Chem. 2000, 112, 4322; Angew. Chem. Int. Ed. 2000,
39, 4150.
[4] For glycofullerenes prepared under CuAAC conditions, see: a) H.
Isobe, K. Cho, N. Solin, D. B. Werz, P. H. Seeberger, E. Nakamura,
Org. Lett. 2007, 9, 4611; b) G. Rocha Pereira, L. J. Santos, I. Luduvi-
co, R. Brondi Alves, R. Pereira de Freitas, Tetrahedron Lett. 2010,
51, 1022.
[5] J.-F. Nierengarten, J. Iehl, V. Oerthel, M. Holler, B. M. Illescas, A.
Munoz, N. Martin, J. Rojo, M. Sanchez-Navarro, S. Cecioni, S.
Vidal, K. Buffet, M. Durka, S. P. Vincent, Chem. Commun. 2010, 46,
3860.
[6] For a review on fullerene hexa-adducts, see: A. Hirsch, O. Vostrow-
sky, Eur. J. Org. Chem. 2001, 829.
[7] M. Durka, K. Buffet, J. Iehl, M. Holler, J.-F. Nierengarten, J. Tagan-
na, J. Bouckaert, S. P. Vincent, Chem. Commun. 2011, 47, 1321.
[8] a) M. Sꢄnchez-Navarro, A. MuÇoz, B. M. Illescas, J. Rojo, N. Martꢅn,
Chem. Eur. J. 2011, 17, 766; b) S. Cecioni, V. Oerthel, J. Iehl, M.
Holler, D. Goyard, J.-P. Praly, A. Imberty, J.-F. Nierengarten, S.
Vidal, Chem. Eur. J. 2011, 17, 3252.
[9] P. Compain, C. Decroocq, J. Iehl, M. Holler, D. Hazelard, T. Mena
Barragꢄn, C. Ortiz Mellet, J.-F. Nierengarten, Angew. Chem. 2010,
122, 5889; Angew. Chem. Int. Ed. 2010, 49, 5753.
[10] D. H. Dube, C. R. Bertozzi, Nat. Rev. Drug Discov. 2005, 4, 477.
[11] E. E. Simanek, G. J. McGarvey, J. A. Jablonowski, C. H. Wong,
Chem. Rev. 1998, 98, 833.
[22] H. Dohi, R. Pꢁrion, M. Durka, M. Bosco, Y. Rouꢁ, F. Moreau, S.
Grizot, A. Ducruix, S. Escaich, S. P. Vincent, Chem. Eur. J. 2008, 14,
9530.
[23] S. Grizot, M. Salem, V. Vongsouthi, L. Durand, F. Moreau, H. Dohi,
S. Vincent, S. Escaich, A. Ducruix, J. Mol. Biol. 2006, 363, 383.
[24] a) S. Oscarson in Glycoscience Synthesis of Oligosaccharides and
Glycoconjugates, Vol. 186 (Eds.: H. Driguez, J. Thiem), Springer,
Berlin, 1997, p.171; b) S. Oscarson, Carbohydr. Polym. 2001, 44,
305; c) T. Ohara, A. Adibekian, D. Esposito, P. Stallforth, P. H. See-
[12] S. Berg, D. Kaur, M. Jackson, P. J. Brennan, Glycobiology 2007, 17,
35r.
[13] a) P. Compain, O. R. Martin, Bioorg. Med. Chem. 2001, 9, 3077;
b) M. L. Mitchell, F. Tian, L. V. Lee, C. H. Wong, Angew. Chem.
2002, 114, 3167; Angew. Chem. Int. Ed. 2002, 41, 3041; c) S. H.
Khan, S. C. Crawley, O. Kanie, O. Hindsgaul, J. Biol. Chem. 1993,
268, 2468; d) B. Mꢆller, C. Schaub, R. R. Schmidt, Angew. Chem.
1998, 110, 3021; Angew. Chem. Int. Ed. 1998, 37, 2893; e) M. M.
Palcic, L. D. Heerze, O. P. Srivastava, O. Hindsgaul, J. Biol. Chem.
650
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 641 – 651

A New Mode of Glycosyltransferase Inhibition
FULL PAPER
berger, Chem. Commun. 2010, 4106; d) D. Crich, A. Banerjee, J.
Am. Chem. Soc. 2006, 128, 8078.
New J. Chem. 2000, 24, 687; b) Y. Rio, G. Enderlin, C. Bourgogne,
J.-F. Nierengarten, J.-P. Gisselbrecht, M. Gross, G. Accorsi, N. Ar-
maroli, Inorg. Chem. 2003, 42, 8783; c) M. Gutierrez-Nava, G. Ac-
corsi, P. Masson, N. Armaroli, J.-F. Nierengarten, Chem. Eur. J.
2004, 10, 5076; d) H. Herschbach, K. Hosomizu, U. Hahn, E. Leize,
A. Van Dorsselaer, H. Imahori, J.-F. Nierengarten, Anal. Bioanal.
Chem. 2006, 386, 46; e) J. Iehl, M. Vartanian, M. Holler, J.-F. Nier-
engarten, B. Delavaux-Nicot, J.-M. Strub, A. Van Dorsselaer, Y. Wu,
J. Mohanraj, K. Yoosaf, N. Armaroli, J. Mater. Chem. 2011, 21, 1562.
[33] B. A. Johns, C. R. Johnson, Tetrahedron Lett. 1998, 39, 749.
[34] A. Lohse, K. B. Jensen, K. Lundgren, M. Bols, Bioorg. Med. Chem.
1999, 7, 1965.
[35] a) T. Wennekes, R. J. B. H. N. van den Berg, K. M. Bonger, W. E.
Donker-Koopman, A. Ghisaidoobe, G. A. van der Marel, A. Strij-
land, J. M. F. G. Aerts, H. S. Overkleeft, Tetrahedron: Asymmetry
2009, 20, 836; b) A. J. Cagnoni, O. Varela, S. G. Gouin, J. Kovensky,
M. L. Uhrig, J. Org. Chem. 2011, 76, 3064.
[25] a) G. J. P. H. Boons, M. Overhand, G. A. van der Marel, J. H. van
Boom, Carbohydr. Res. 1989, 192, c1; b) G. J. P. H. Boons, M. Over-
hand, G. A. van der Marel, J. H. van Boom, Angew. Chem. 1989,
101, 1538; Angew. Chem. Int. Ed. Engl. 1989, 28, 1504; c) G. J. P. H.
Boons, G. A. van der Marel, J. H. van Boom, J. T. Poolman, Recl.
Trav. Chim. Pays-Bas 1989, 108, 339.
[26] M. Durka, A. Tikad, R. Pꢁrion, M. Bosco, M. Andaloussi, S. Flo-
quet, E. Malacain, F. Moreau, M. Oxoby, V. Gerusz, S. P. Vincent,
Chem. Eur. J. 2011, 17, 11305.
[27] J. Iehl, J. F. Nierengarten, Chem. Eur. J. 2009, 15, 7306.
[28] J. Iehl, R. Pereira de Freitas, B. Delavaux-Nicot, J.-F. Nierengarten,
Chem. Commun. 2008, 2450.
[29] a) J. Iehl, J.-F. Nierengarten, Chem. Commun. 2010, 46, 4160; b) J.
Iehl, M. Holler, J.-F. Nierengarten, K. Yoosaf, J. M. Malicka, N. Ar-
maroli, J.-M. Strub, A. Van Dorsselaer, B. Delavaux-Nicot, Aust. J.
Chem. 2011, 64, 153; c) P. Fortgang, E. Maisonhaute, C. Amatore, B.
Delavaux-Nicot, J. Iehl, J.-F. Nierengarten, Angew. Chem. 2011, 123,
2412; Angew. Chem. Int. Ed. 2011, 50, 2364.
[36] J. Diot, M. I. Garcia-Moreno, S. G. Gouin, C. Ortiz Mellet, K.
Haupt, J. Kovensky, Org. Biomol. Chem. 2009, 7, 357.
[37] J. E. Gestwicki, C. W. Cairo, L. E. Strong, K. A. Oetjen, L. L. Kies-
sling, J. Am. Chem. Soc. 2002, 124, 14922;
[30] D. J. Harvey, Mass Spectrom. Rev. 1999, 18, 349.
[31] a) R. Kessinger, J. Crassous, A. Herrmann, M. Rꢆttiman, L. Eche-
goyen, F. Diederich, Angew. Chem. 1998, 110, 2022; Angew. Chem.
Int. Ed. 1998, 37, 1919; b) N. N. P. Moonen, C. Thilgen, L. Echegoy-
en, F. Diederich, Chem. Commun. 2000, 335.
[38] C. Ferrandiz-Huertas, J. Isac-Garcꢅa, F. Pꢁrez-Balderas, F. Santoyo-
Gonzꢄlez, Synthesis 2005, 6, 939.
[39] Y. Yamada, K. Matsuura, K. Kobayashi, Bioorg. Med. Chem. 2005,
13, 1913.
[32] a) D. Felder, H. Nierengarten, J.-P. Gisselbrecht, C. Boudon, E.
Leize, J.-F. Nicoud, M. Gross, A. Van Dorsselaer, J.-F. Nierengarten,
Received: July 4, 2011
Published online: December 6, 2011
Chem. Eur. J. 2012, 18, 641 – 651
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
651