Systematic dissection of an aminopyrrolic cage receptor for β-glucopyranosides reveals the essentials for effective recognition

Article abstract of DOI:10.1002/chem.201400365

A set of structures designed for the recognition of glucosides has been obtained by systematically destructuring a tripodal aminopyrrolic cage receptor that selectively recognizes octyl-β-D-glucopyranoside (OctβGlc). NMR spectroscopy and isothermal titration calorimetry binding measurements showed that cleavage of one pillar of the cage was beneficial to the binding properties of the receptor, as long as two residual amino groups of the cleaved pillar were present. Removal of these two residual amino groups produced a dramatic loss of affinity for OctβGlc of the resulting monocyclic analogue of the parent cage receptor. A significant improvement in the binding ability was achieved by replacing one pillar with two aminopyrrolic hydrogen-bonding arms, despite the loss of a preorganized structure. In contrast to the cage receptor, recognition of OctβGlc was observed, even in a competitive medium (30% DMF in chloroform). Structural studies in solution, carried out through NMR spectroscopy and molecular modeling calculations, led to the elucidation of the 3D binding modes of the side-armed monocyclic receptors; this highlighted the key role of the amino groups and demonstrated the occurrence of a rotaxane-like complex, which featured the octyl chain of the glucoside threaded through the macrocyclic ring. Comfort is paramount! A set of structures designed for the recognition of glucosides has been obtained by systematically destructuring a tripodal aminopyrrolic cage receptor. A β-glucoside is more effectively recognized by an adaptive cleft than by a preorganized cage, if a more comfortable fit can be achieved (see picture).

Full text of DOI:10.1002/chem.201400365

DOI: 10.1002/chem.201400365
Full Paper
&
Host–Guest Systems
Systematic Dissection of an Aminopyrrolic Cage Receptor for b-
Glucopyranosides Reveals the Essentials for Effective Recognition
Oscar Francesconi,[b] Matteo Gentili,[b] Cristina Nativi,[b, c] Ana Ardꢀ,[d] F. Javier CaÇada,[d]
Jesffls JimØnez-Barbero,[d] and Stefano Roelens*[a]
Abstract: A set of structures designed for the recognition of
glucosides has been obtained by systematically destructur-
ing a tripodal aminopyrrolic cage receptor that selectively
recognizes octyl-b-d-glucopyranoside (OctbGlc). NMR spec-
troscopy and isothermal titration calorimetry binding mea-
surements showed that cleavage of one pillar of the cage
was beneficial to the binding properties of the receptor, as
long as two residual amino groups of the cleaved pillar were
present. Removal of these two residual amino groups pro-
duced a dramatic loss of affinity for OctbGlc of the resulting
monocyclic analogue of the parent cage receptor. A signifi-
cant improvement in the binding ability was achieved by re-
placing one pillar with two aminopyrrolic hydrogen-bonding
arms, despite the loss of a preorganized structure. In con-
trast to the cage receptor, recognition of OctbGlc was ob-
served, even in a competitive medium (30% DMF in chloro-
form). Structural studies in solution, carried out through
NMR spectroscopy and molecular modeling calculations, led
to the elucidation of the 3D binding modes of the side-
armed monocyclic receptors; this highlighted the key role of
the amino groups and demonstrated the occurrence of a ro-
taxane-like complex, which featured the octyl chain of the
glucoside threaded through the macrocyclic ring.
Introduction
recognition of specific carbohydrates of biological interest
through noncovalent interactions, mimicking natural lectins,[4]
have been developed to elucidate structural and functional re-
quirements for effective recognition.[5]
Molecular recognition of carbohydrates is a topic of growing
interest due to the crucial role exerted by saccharides in bio-
logical systems.[1] Exposed on the glycocalix as protein and
lipid glycoconjugates, carbohydrates are involved in cell-to-
cell, cell-to-bacteria, and cell-to-virus adhesion processes,
which are governed by selective recognition of specific saccha-
ridic epitopes.[1d] An increasing number of studies highlighted
the importance of carbohydrates in pathogen recognition, in
modulation of the immune system, and in the control of ho-
meostasis and inflammation;[2] this explains why a molecular-
level understanding of these processes is the main goal of cur-
rent research.[3] To this end, synthetic receptors for selective
In this context, in the last few years, we have developed
a family of synthetic receptors for carbohydrates (cyclic and
acyclic) that successfully recognize mono- and disaccharides in
polar organic media.[6] The architecture of these receptors,
common to the whole family, was based on a hexasubstituted
aromatic scaffold, implemented with side arms featuring hy-
drogen-bonding groups, of which the aminopyrrolic chelating
arrangement was the most effective binding combination. The
idea behind this structural design was to combine hydrogen-
bonding interactions between saccharidic hydroxyl groups and
complementary hydrogen-bonding groups on the receptor,
with CH–p and van der Waals interactions between the aro-
matic moiety and the aliphatic backbone of the carbohydrate,
to construct a biomimetic receptor capable of recognizing sac-
charidic ligands through noncovalent interactions. Although
recognition in water has not yet been achieved, the binding
properties measured in organic solvents for the receptors
developed so far shed light on the requirements for effective
recognition.
[a] Dr. S. Roelens
Istituto di Metodologie Chimiche (IMC)
Consiglio Nazionale delle Ricerche (CNR), Dipartimento di Chimica
Polo Scientifico e Tecnologico, 50019 Sesto Fiorentino, Firenze (Italy)
E-mail: stefano.roelens@unifi.it
[b] Dr. O. Francesconi, M. Gentili, Prof. C. Nativi
Dipartimento di Chimica, Universitꢀ di Firenze
Polo Scientifico e Tecnologico, 50019 Sesto Fiorentino, Firenze (Italy)
[c] Prof. C. Nativi
Within this family of structures, bicyclic cage receptor
1 showed remarkable recognition properties, in that it selec-
tively distinguished a b-glucoside (OctbGlc) from glycopyrano-
sides of the manno and galacto series, featuring 20 mm affinity
and complete b/a selectivity in chloroform.[7a] An investigation
into the structure and binding properties of 1 by NMR spec-
troscopy and isothermal titration calorimetry (ITC) techniques,
Centro Risonanze Magnetiche (CERM)
Polo Scientifico e Tecnologico, 50019 Sesto Fiorentino, Firenze (Italy)
[d] Dr. A. Ardꢁ, Prof. F. J. CaÇada, Prof. J. Jimꢂnez-Barbero
Chemical and Physical Biology
Centro de Investigaciones Biolꢃgicas, CSIC
Ramiro de Maeztu 9, 28040 Madrid (Spain)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201400365.
Chem. Eur. J. 2014, 20, 6081 – 6091
6081
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
calculations, in addition to providing a 3D view of the recogni-
tion modes in solution, led to the quantification of 1) the large
affinity difference for the glucoside between the constitutional-
ly analogous mono- and bicyclic receptors; 2) the recognition
improvement achieved by an adaptive structure, compared
with the preorganized parent cage receptor; and 3) the sub-
stantial contribution to glucoside recognition from the hydro-
gen-bonding amino groups, both in terms of binding affinity
and conformation-directing ability.
combined with molecular modeling calculations,[7b] revealed
that the recognition ability of 1 resulted from the interplay of
several factors, including the size of the bicyclic cage, the cor-
rect binding geometry, and the efficacy of the pyrrolic hydro-
gen-bonding groups, which cooperate for an energetically fa-
vorable binding of the all-equatorial glucoside inside the re-
ceptor cavity. Yet, when switching to more polar media, such
as acetonitrile, receptor 1 unex-
Results and Discussion
Design and synthesis
The systematic modifications designed for the structure of re-
ceptor 1 are outlined in Scheme 1. When cleaving one pillar of
the cage, the least perturbing synthetically feasible modifica-
pectedly failed to recognize the
b-glucoside. This evidence led us
to suspect that the shape-persis-
tent cage might be somewhat
too tight or too rigid to accom-
modate the saccharidic ligand
when contending with a compet-
itive solvent. This hypothesis
was based on the observation
that the CPK models of the com-
plex showed the glucoside com-
pletely filling the cavity of the
receptor, with little or no free-
dom of the guest to adjust to
the cavity itself. Following this
hypothesis, we thought that
cleaving a bond in one of the
pillars of the cage might relieve
congestion and provide the ap-
Scheme 1. Systematic modification of the structure of receptor 1.
propriate adaptivity, while bring-
ing minimal perturbation to the
constitution of the receptor, pro-
viding that a cage-like geometry could be achieved by the re-
sulting side-armed monocyclic structure upon binding. Further
dissection of the cleaved pillar, destructuring the cage architec-
ture down to the monocyclic core through a stepwise removal
of all constituent elements of the side arm, would show the ef-
fectiveness of the bicyclic receptor with respect to the mono-
cyclic counterpart and would shed light on the contribution to
binding from the side chain. On the other hand, converting
the monoarmed receptor into a double-armed chelating
monocyclic receptor by implementing an additional binding
arm would show whether recognition ability could be further
improved.
tion consists of removing one of the two methylene groups
adjacent to the pyrrolic moiety to give monocyclic structure 2.
Although devoid of the original bicyclic structure, the resulting
receptor preserves all binding groups of the parent architec-
ture. Subsequent removal of the pyrrole group from the side
arm affords symmetric diamino compound 3, which features
a reduced number of hydrogen-bonding groups. Stepwise re-
moval of the residual amino groups leads to compounds 4
and 5, which, through loss of one entire pillar, containing
three out of nine hydrogen-bonding groups of the parent
structure, affords the monocyclic analogue of 1. As an alterna-
tive to the described destructuring pathway, cleaved receptor
2 can be implemented with an additional pyrrole ring on the
primary amino group, giving rise to the symmetrical, double-
armed monocyclic structure 6. In addition to the increased
number of hydrogen-bonding groups, receptor 6 may endow
the monocyclic structure with a chelating arrangement of
Herein, we report the details of a study on the structure, dy-
namics, and binding properties of the receptors obtained by
the aforementioned systematic modification of the parent re-
ceptor architecture. The investigation, carried out by NMR
spectroscopy and ITC techniques, and by molecular modeling
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6082
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
binding groups, which may im-
prove affinity toward the gluco-
sidic ligand.
Receptor 2 could be prepared
through the convergent synthet-
ic
pathway
described
in
Scheme 2. Compounds 7 and 8,
prepared from 1,3,5-tri(bromo-
methyl)-2,4,6-triethylbenzene,
were treated with the sodium
salt of trifluoroacetamide to give
the corresponding trifluoroacet-
amides 9 and 10, which under-
went alkaline hydrolysis to afford
the corresponding amines 11
and 12. Monoamine 11 was con-
densed with pyrrole-2-carbox-
aldehyde to give the corre-
sponding Schiff base, which was
reduced to the aminopyrrolic
compound 13. On the other
hand, diamine 12 was con-
densed with 5-(5,5-dimethyl-1,3-
dioxan-2-yl)pyrrole-2-carbalde-
hyde, the monoprotected pyr-
role dicarbaldehyde, and then
reduced to compound 14, which
was deprotected to give dialde-
hyde 15. Diamine 16, which was
obtained from 13 by Staudinger
Scheme 2. Synthesis of receptor 2. Reagents and conditions: a) trifluoroacetamide, NaH, DMF, RT, overnight, 88%;
b) NaOH, CH3OH/H2O, RT, overnight, 98%; c) pyrrole-2-carbaldehyde, AcOH, CHCl3, RT, overnight, then NaBH4,
CH3OH, RT, 1 h, 84%; d) PPh3, H2O/THF, RT, overnight, 97%; e) trifluoroacetamide, NaH, DMF, RT, overnight, 84%;
f) NaOH, CH3OH/H2O, RT, overnight, 97%; g) 5-(5,5-dimethyl-1,3-dioxan-2-yl)-1H-pyrrole-2-carbaldehyde, CHCl3, RT,
overnight, then NaBH4, CH3OH, RT, 1 h, 79%; h) HCl, H2O/THF, RT, 2 h, >99%; i) CHCl3, RT, overnight, then NaBH4,
CH3OH, RT, 1 h, 96%; j) PPh3, H2O/THF, RT, overnight, 84%.
reduction, was condensed with 15 and the resulting cyclic
Schiff base was reduced by sodium borohydride to give com-
pound 17. Staudinger reduction of 17 finally afforded receptor
2.
Receptors 3 and 4 were prepared from 12 and 15, respec-
tively, through the synthetic sequence outlined in Scheme 3,
by condensation with the appropriate reagent, reduction of
the imines to azides 18 and 19, and subsequent Staudinger re-
duction. Receptor 5 was prepared by condensation of 1,3-bis-
(aminomethyl)-2,4,6-triethylbenzene with pyrrole-2,5-dicarbox-
aldehyde and subsequent reduction of the resulting cyclic
Schiff base with sodium borohydride (Scheme 4).
Finally, receptor 6 was conveniently synthesized by condens-
ing receptor 3 with pyrrole-2-carboxaldeyde (2 equiv), followed
by reduction with sodium borohydride (Scheme 4).
Binding studies
Scheme 3. Synthesis of receptors 3 and 4. Reagents and conditions: a) pyr-
role-2,5-dicarbaldehyde, CHCl3, RT, overnight, then NaBH4, CH3OH, RT, 1 h,
87%; b) PPh3, H2O/THF, RT, overnight, 84%; c) 1,3-bis(aminomethyl)-2,4,6-tri-
ethylbenzene, AcOH, CHCl3, RT, overnight, then NaBH4, CH3OH, RT, 1 h, 49%;
d) PPh3, H2O/THF, RT, overnight, 93%.
To homogeneously compare the binding properties of the new
set of monocyclic receptors with those of the parent cage re-
ceptor 1, the binding affinities toward OctbGlc were assessed
1
through H NMR spectroscopic titrations in CDCl3 at 298 K. As-
sociation constants were measured according to a previously
described protocol,[6k] which consisted of the simultaneous fit
of the complexation-induced shifts of all available signals from
both the receptor and the glycoside to the appropriate associ-
ation model by nonlinear regression analysis.
Analogously to receptor 1, upon addition of receptors 2, 3,
and 6, the H NMR spectra of OctbGlc showed the appearance
of a new set of separated signals from those of the free gluco-
side; this revealed the formation of a 1:1 species under a slow-
1
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6083
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 4. Synthesis of receptors 5 and 6. Reagents and conditions: a) pyr-
role-2,5-dicarbaldehyde, CHCl3, RT, overnight, then NaBH4, CH3OH, RT, 1 h,
57%; b) pyrrole-2-carbaldehyde, CHCl3, RT, overnight, then NaBH4, CH3OH,
RT, 1 h, 91%.
1
Figure 1. Selected spectral regions of the H NMR spectroscopic titration
(400 MHz, CDCl3, 298 K) of OctbGlc (0.992 mm) with increasing concentra-
tions of receptor 3: a) 0, b) 0.429, c) 1.00, d) 2.00 mm. Left: pyrrolic NH sig-
nals; right: terminal octyl chain signals of OctbGlc. The signals of the Fast (*)
exchange regime on the NMR spectroscopy timescale (Slow;
Figure 1 and Figures S1 and S2 in the Supporting Information).
However, although this was true for receptors 3 and 6, recep-
tor 2 revealed the presence of two different slow-exchanging
1:1 complexes, the signals of which increased along the titra-
tion. This behavior, which is clearly appreciated from the HSQC
map (Figure S3 in the Supporting Information), can be ex-
plained by considering that the two halves of receptor 2, with
respect to the plane bisecting the pyrrole rings, are not sym-
metrical. Therefore, two distinct isomeric complexes are
formed upon binding OctbGlc, namely, Slow(a) and Slow(b),
which are caused by inclusion of the glucoside inside the
receptor cleft with either the a or b face pointing toward the
half of the receptor with the primary amine,
*
and Slow ( ) 1:1 complexes are also indicated.
of the Fast complexes, which allowed the determination of the
cumulative binding constants reported in Table 1. Because
multiple complex species of different stoichiometries were
found in some cases, affinities were assessed on a common
scale from the measured binding constants through the BC500
parameter, which is a binding descriptor that quantitatively de-
fines the overall affinity of a receptor for a ligand.[8] The BC500
(intrinsic median binding concentration) descriptor is defined
as the total concentration of receptor necessary to bind 50%
respectively.
Table 1. Cumulative formation constants (logbn) and intrinsic median binding con-
centrations (BC050, [mm])[a] for receptor to glycoside (R/G) complexes of OctbGlc in
CDCl3.
In addition to slow-exchanging complexes, all
three receptors formed complex species in a fast-ex-
change regime on the NMR spectroscopy timescale
that were not detected for cage receptor 1. The fast-
exchanging complexes (Fast) were revealed from the
observed chemical shift variations of the OctbGlc sig-
nals along the titration with receptors 2, 3, and 6.
These chemical shift perturbations reflected the time-
averaging of signals from the bound and unbound
glucoside. Concomitantly, the receptor signals under-
went significant variations of chemical shift as well,
which further supported the formation of Fast com-
plexes in solution. Although this evidence did not
question the formation of the Slow complexes, it un-
avoidably affected the equilibrium concentration of
free glucoside required for a quantitative calculation
of binding constants, which, therefore, could not be
assessed by NMR spectroscopic titrations. In contrast,
receptors 4 and 5 featured the exclusive formation
Receptor
NMR[b]
ITC[c]
logb (R/G)
BC050
logb (R/G)
BC050
1
4.684Æ0.008 (1:1)
20.7Æ0.4
4.60Æ0.01 (1:1)
7.57Æ0.04 (1:2)
5.09Æ0.05 (1:1)
5.02Æ0.03 (1:1)
8.4Æ0.1 (1:2)
25.1Æ0.6
2
3
n.a.[d]
n.a.[d]
n.a.[d]
n.a.[d]
8.1Æ0.9
9.2Æ0.7
4
5
6
4.36Æ0.07 (1:1)
6.9Æ0.3 (2:1)
3.11Æ0.03 (1:1)
5.23Æ0.05 (1:2)
n.a.[d]
42.5Æ6.6
674Æ32
n.a.[d]
4.41Æ0.02 (1:1)
38.9Æ2.0
831Æ27
3.5Æ0.4
3.08Æ0.01 (1:1)
5.46Æ0.05 (1:1)
[a] Calculated from the logb values by using the “BC50 Calculator” program.[9] [b] Mea-
sured by 1H NMR spectroscopy (400 MHz) from titration experiments at T=298 K.
[c] Measured by ITC from titration experiments at T=298 K. [d] Not available: simulta-
neous presence of fast- and slow-exchanging complexes on the NMR spectroscopy
timescale prevents calculation of the formation constants from NMR spectroscopy
data.
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6084
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
of the ligand when the fraction of bound receptor is zero, that
is, when forming the first complex molecule. This value coin-
cides with the dissociation constant, Kd, for 1:1 complexes,
whereas it can be viewed as a “global” dissociation constant
when more than one complex is present in solution. The BC500
values calculated for the investigated systems through the
BC50 Calculator[9] are also reported in Table 1 and compared
with values previously obtained for receptor 1.
To correctly assess the affinities of receptors 2, 3, and 6, and
to confirm those obtained for receptors 4 and 5, calorimetric
ITC titrations were also performed under the same conditions
(CHCl3, 298 K). When necessary, the presence of multiple com-
plex species was addressed by combining independent titra-
tions run at different reactant concentrations into a simultane-
ous fit of all data to remove ambiguities in the definition of
the model. The results are reported in Table 1, in Figure 2 and
in Figures S4 and S5 in the Supporting Information, together
with previous data for receptor 1, as cumulative binding con-
stants and BC050 values. The overall 1:1 complex macrospecies,
irrespective of the exchanging kinetics, were detected by the
calorimetric technique; thus overcoming the issue of fast and
slow exchange observed in the NMR spectroscopic titrations.
The ITC results for receptors 4 and 5 were in good agreement
with those observed by NMR spectroscopy, whereby the slight
differences between the BC050 values were ascribed to the low
sensitivity of the ITC technique to higher stoichiometry
species.
From the results reported in Table 1, it can be noted that
cleavage of one pillar of the cage in 2 gave a threefold in-
crease in affinity, which confirmed the hypothesis that convert-
ing the cage into an adaptive structure was beneficial to the
receptor binding properties. Analysis of the thermodynamic
parameters (Table S1 in the Supporting Information) showed
that the affinity increase was enthalpic in origin, which com-
pensated for the entropic loss caused by the larger loss of de-
grees of freedom experienced by the flexible receptor, with re-
spect to the preorganized structure, upon binding. The
marked enthalpic gain can be understood in terms of im-
proved adaptivity of the armed monocyclic structure, which
allows for an improved hydrogen-bonding match between the
receptor and saccharidic ligand.
Figure 2. ITC titrations of OctbGlc with 2 in CHCl3 at T=298 K. a) Titration of
OctbGlc (0.407 mm) with 2 (2.80 mm). b) Titration of OctbGlc (0.204 mm)
with 2 (2.80 mm). c) Plot of the experimental and calculated data points ob-
Interestingly, removal of the pyrrolic hydrogen-bonding
group from the side arm in receptor 3 does not affect the af-
finity with respect to 2. This unexpected result, apparently in
contrast to the well-established contribution of the pyrrolic
groups to binding,[6c,e,h,j,7b] may be due to either marginal par-
ticipation of this pyrrole in binding for conformational reasons
or because its contribution is counterbalanced by an overall
weakening of the interaction, which is likely to be caused by
the geometry imposed by the binding arm. This latter explana-
tion is supported by molecular modeling studies (see below).
Remarkably, monocyclic diamino receptor 3 exhibits an im-
proved affinity with respect to the bicyclic cage receptor 1,
demonstrating that effective recognition is not necessarily as-
cribed to the cage structure.
&
*
tained for titrations a) ( ) and b) ( ). Crosshair symbols (+) are calculated
points obtained by simultaneous fit of all data through nonlinear regression.
and 5, which showed a dramatic drop of affinity toward
OctbGlc with respect to 3. Removal of one of the two amine
groups caused a fourfold affinity decrease, whereas a two
orders of magnitude drop was observed for the bare mono-
cyclic receptor when removing both amino groups from the
structure. Furthermore, Slow complexes were absent for both
receptors, which displayed Fast complexes exclusively; this in-
dicated that the diamino compound enforced a binding geom-
etry with common structural/kinetic features to those of the
parent cage structure.
The substantial contribution from the amine binding groups
clearly emerged from the results obtained with receptors 4
Finally, the active contribution from the pyrrolic binding
arms was demonstrated by bibracchial receptor 6, the affinity
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6085
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
of which for OctbGlc was the highest of the whole set of struc-
tures: three times larger than that of 2 and 3, and eight times
larger than that of 1. As expected, pyrrolic groups contribute
substantially to the binding ability of the receptor, whenever
binding arms are not affected by conformational restrictions.
The results obtained for the monocyclic receptors show that
the cage structure is not a prerequisite for effective recogni-
tion. Superior recognition properties could be achieved by a re-
lated monocyclic structure, when side arms with effective hy-
drogen-bonding groups were implemented in the architecture,
providing that the appropriate binding geometry could be
achieved, despite the unavoidable entropic loss upon complex-
ation with respect to the preorganized bicyclic structure.
in chloroform. Thermodynamic parameters, reported in
Table S2 in the Supporting Information, confirmed the enthal-
pic origin of binding, as measured in chloroform, together
with a similar adverse entropic contribution. In contrast to the
parent cage receptor, hydrogen-bonding interactions estab-
lished by the adaptive, side-armed monocyclic structures could
favorably compete with the solvent for the glucoside, even in
a largely polar medium.
Structural studies
The intriguing occurrence of concomitant Fast and Slow com-
plexes in the recognition of OctbGlc prompted us to investi-
gate the structural features of the recognition modes of the
monocyclic receptors. Unfortunately, crystals suitable for X-ray
structure determination could not be obtained for any of the
complexes under study. Solution studies are more informative
than solid-state studies when exploring dynamic phenomena,
because they provide a closer description of dynamic process-
es. Elucidation of the 3D binding modes in solution for the
most relevant complexes, in particular, of receptors 2, 3, and 6,
could be obtained by combining NMR spectroscopy data with
molecular modeling calculations, following a well-established
protocol.[10]
Encouraged by the excellent affinities obtained in chloro-
form, we challenged the recognition abilities of the investigat-
ed receptors in a more competitive medium to check whether
evidence of binding could be detected where the cage recep-
tor failed. The medium of choice was 30% DMF in chloroform,
which provided good solubility of the reactants and hydrogen-
bonding competitiveness comparable to that of acetonitrile.[6a]
Cumulative binding constants toward OctbGlc, together with
BC050 values, measured by NMR spectroscopy in CDCl3/DMF
(70:30) for the set of receptors are presented in Table 2. Unex-
pectedly, receptors 2, 3, and 6, together with Fast complexes,
still showed the presence of Slow complexes, although less
Basic information about the conformational features of com-
plexes was retrieved from the chemical shift differences (CSD)
observed between bound and unbound reactants.
Chemical shifts for bound glucoside were obtained
Table 2. Cumulative formation constants (logbn) and intrinsic median binding con-
centration (BC050, [mm])[a] for receptor to glycoside (R/G) complexes of OctbGlc in
CDCl3/DMF 70:30.
by adding an amount of receptor necessary to fully
complex OctbGlc, as calculated from the measured
binding constants. The results for the Fast and Slow
complexes of 2, 3, and 6 are reported in Table 3 and
Tables S3 and S4 in the Supporting Information, re-
spectively, and depicted in Figure 3 and Figures S6
and S7 in the Supporting Information, respectively.
The determination of the chemical shifts of the
bound species, which was prevented by severe over-
lapping of signals in the 1H spectra, was accom-
plished through 2D HSQC experiments that provided
unambiguous shift values for all resonances.
Receptor
NMR[b]
ITC[c]
logb (R/G)
BC050
logb (R/G)
BC050
2
3
4
5
6
n.a.[d]
n.a.[d]
4.371Æ0.006 (1:1)
0.9Æ0.2 (1:1)
n.a.[d]
n.a.[d]
2.797Æ0.007 (1:1)
2.698Æ0.006 (1:1)
2.425Æ0.001 (1:1)
n.d.[e]
1.59Æ0.03
2.01Æ0.03
3.76Æ0.01
n.d.[e]
n.a.[d]
4.25Æ0.06
114Æ41
n.a.[d]
3.008Æ0.008 (1:1)
0.98Æ0.02
[a] Calculated from the logb values by using the BC50 Calculator program.[9] [b] Mea-
sured by 1H NMR spectroscopy (400 MHz) from titration experiments at T=298 K.
[c] Measured by ITC from titration experiments at T=298 K. [d] Not available: simulta-
neous presence of fast- and slow-exchanging complexes on the NMR timescale pre-
vents calculation of the formation constants from the NMR spectroscopy data. [e] Not
detectable.
The remarkable upfield shift of the glucoside sig-
nals observed in all three cases for both the Fast and
Slow complexes clearly showed that the sugar experi-
enced a shielding effect from the aromatic rings of
abundant; this indicated that binding was still substantial,
Table 3. Chemical shifts (d, [ppm]) and chemical shift differences (CSD,
[ppm]) of the proton signals of OctbGlc, free and bound to 2 (OctbGlc
1.59 mm, 2 3.18 mm), for the Slow(a), Slow(b), and Fast complexes in
CDCl3 at 298 K.
even in the competitive solvent. Although the concomitant
presence of Fast and Slow complexes did not allow the deter-
mination of the binding constants by NMR spectroscopic titra-
tions, it clearly appeared that, in contrast to receptor 1, recog-
nition was not depleted by the competitive medium.
Nucleus
Free
Slow(a)
Slow(b)
Fast
Bound
Bound
CSD
Bound
CSD
CSD
Binding affinities could instead be assessed by ITC titrations
for all receptors, with the exception of 5, the affinity of which
was negligible in this medium. Predictably, binding constants
were depressed by two to three orders of magnitude with re-
spect to those in chloroform; nevertheless, millimolar affinities
could still be measured, for all but 5, which featured very shal-
low selectivity that followed the same order as that observed
H-1
H-2
H-3
H-4
H-5
H-6
H-6’
4.309
3.353
3.609
3.592
3.405
3.925
3.825
2.305
0.518
2.287
2.612
1.966
3.788
3.571
2.004
2.835
1.322
0.980
1.439
0.137
0.254
2.305
0.418
2.349
2.537
2.075
3.856
3.571
2.004
2.935
1.260
1.055
1.330
0.069
0.254
3.141
0.756
2.127
2.534
2.061
3.309
3.309
1.168
2.597
1.482
1.058
1.344
0.616
0.516
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6086
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ing different dynamics, and that
the slow kinetics, in analogy to
those observed for the bicyclic
structure 1, may be due to hin-
drance from the cyclic part of
the structure, causing slow
motion of the complexation
process.
Complementary information
on the structures of the studied
complexes was obtained by
NOESY and ROESY experiments
performed on mixtures of
OctbGlc and receptors 2, 3, and
6, from which several intermo-
lecular NOE/ROE contacts were
detected. A reasonable number
of contacts could be identified
for each of the three systems,
the most significant of which
have been reported in Table S5
in the Supporting Information
and graphically depicted in
Figure 4 for receptor 2, and in
Figures S8 and S9 in the Sup-
porting Information for receptors
3 and 6, respectively.
The observed NOE/ROE con-
tacts for all Slow complexes ex-
hibit close similarities, which
suggest a common binding ge-
ometry. For example, the H-5
proton of glucose shows a ROE
contact with the H-2 proton of
the ethyl groups of the receptor;
moreover, both the NH protons
of the endocyclic pyrroles show
intermolecular contacts with
either the anomeric proton or
the H-7 protons of the octyl
chain. These proximities impose
an orientation of the sugar in
which the octyl chain threads
the receptor macrocycle. Such
a rotaxane-like geometry is also
supported by the ROE contacts
between the H-8 protons of the
sugar and the H-24/30 protons
Figure 3. a) 500 MHz 1H NMR spectra (CDCl3, 298 K) of receptor 2 (top); OctbGlc (center); OctbGlc after the addi-
tion of two equivalents of receptor 2 (bottom). The arrows indicate chemical shift changes, as detected from 2D
HSQC spectra (Figure S3 in the Supporting Information), of the ring protons of the sugar due to the formation of
Fast (g) and Slow (a) complexes. b) Plot of the observed CSD between OctbGlc, free and in the presence of
2 (OctbGlc 1.59 mm, 2 3.18 mm), for the Slow(a) (black), Slow(b) (white) and Fast (gray) complexes in CDCl3 at
298 K.
the receptor. Furthermore, because the shielding affects hydro-
gen atoms on both sides of the sugar ring, it can be concluded
that the receptor is bent into a cleft conformation and that
OctbGlc lies inside the cleft between the two aromatic rings.
The largest upfield shift displayed by the glucose H-1 and H-2
signals reveals that this portion of glucose is deeply located
inside the cleft, more closely facing the aromatic rings. The
similar CSD exhibited by the Fast and Slow complexes suggest
that they may share similar binding modes, although display-
of the ethyl groups residing within the macrocyclic environ-
ment of 2 and 3, and with the H-13* methylene proton of 6.
This conformation may explain the slow kinetics observed for
the Slow complexes, which may be caused by the sugar octyl
chain threading the macrocycle to achieve the binding geome-
try. It should be noted that, in the case of receptor 2, the oc-
currence of the two isomeric Slow(a) and Slow(b) complexes is
demonstrated by the fact that the sugar H-8 proton simultane-
ously shows two different ROE contacts with the ethyl proton
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6087
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
calculation returned three different families of structures for re-
ceptor 2, each family was selected to include the structures
found within an energy window of 10 kJmolÀ1 from the mini-
mum (Figure S10 in the Supporting Information), and two dif-
ferent families for receptors 3 and 6 (Figures S11 and S12 in
the Supporting Information), which were in agreement with
the NMR spectroscopy data; all structures found showed
a well-defined cleft conformation for the receptor enclosing
the bound glucoside. The results from calculations indicate
that glucoside recognition drives the flexible, unbiased struc-
ture of the receptor to achieve a cleft conformation upon
binding.
In Figures 5, 6 and 7, the minimum energy structures of the
families obtained for the complexes of receptors 2, 3, and 6
are depicted, respectively, showing two different types of geo-
metries: a) a “threaded” geometry, featuring the octyl chain
threaded through the macrocycle; and b) an “unthreaded” ge-
ometry, featuring the octyl chain located outside the macrocy-
cle. Both geometries feature the glucose moiety nested inside
the receptor cleft, roughly orthogonal to each other, and paral-
lel to the aromatic rings. Notably, both geometries allow the
formation of several hydrogen-bonding interactions, which in-
volve not only the pyrrolic rings, but also the aminic groups.
Figure 4. Top: Schematic representation of the intermolecular ROEs found
between 2 and OctbGlc. ROE contacts unambiguously assigned to Slow(b)
and Fast complexes are indicated by dashed and dotted lines, respectively.
Bottom: 600 MHz ROESY spectrum of a 1:2 mixture of OctbGlc and 2 in
CDCl3 at 298 K. Intermolecular ROE cross-peaks assigned to Fast (F) and
Slow (S) complexes are indicated by squares.
H-30 and with the opposite ethyl proton H-24; this confirms
the two different orientations of the sugar.
Concerning the Fast complexes, because of severe overlap
of the signals, only a few of the several ROE cross-peaks could
be unambiguously assigned to intermolecular contacts. Never-
theless, in the case of receptor 2, the contact between the H-5
proton of the sugar and the H-2* proton of the ethyl group of
the receptor remote from the macrocycle, together with the
two contacts of the H-8 proton of the octyl chain with the pyr-
rolic H-9 and H-10 protons, impose a binding geometry in
which the octyl chain of the sugar resides outside the receptor
macrocycle; this is in agreement with the fast kinetics ob-
served. Analogous contacts involving the octyl chain of the
sugar were found in the complexes of 3 and 6, which con-
firmed a binding geometry similar to that of 2.
Because CSD and NOE/ROE contacts cannot fully define the
binding geometries of the investigated complexes, a well-
tested molecular mechanics computational protocol was ap-
plied to model the three-dimensional binding modes of
OctbGlc with 2, 3, and 6. The protocol consists of exploring
the conformational space of the complex to find the families
of structures in agreement with the experimental NMR spec-
troscopy data. In the first step, a conformational search was
run on the free receptors in the absence of the glucoside. The
result of the calculation was a pool of conformations that
showed no significant energy minima. This behavior indicated
that the monocyclic architecture was endowed with high flexi-
bility, as opposed to the parent receptor 1, which exhibited
a shape-persistent preorganized cavity. In contrast, when the
conformational search was run on the OctbGlc complex, the
Figure 5. Minimum energy structures of the three families of complexes be-
tween 2 and OctbGlc obtained from the conformational search and compat-
ible with the NMR spectroscopy data. Hydrogen bonds between the recep-
tor and carbohydrate are depicted as dashed lines. Selected hydrogen-bond
lengths: a) NÀH···OH-3, 2.09 Š; NÀH···OH-3, 2.10 Š; NÀH···OH-6, 2.00 Š.; b) NÀ
H···OH-3, 2.25 Š; OÀH-3···NH, 2.17 Š; NÀH···OH-4, 2.19 Š; NÀH···OH-6, 2.01 Š.
c) NÀH···OH-3, 1.99 Š; NÀH···O-5, 2.06 Š; NÀH···OH-6, 2.06 Š.
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6088
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 7. Minimum energy structures of the two families of complexes be-
tween 6 and OctbGlc obtained from the conformational search and compat-
ible with the NMR spectroscopy data. Hydrogen bonds between the recep-
tor and carbohydrate are depicted as dashed lines. Selected hydrogen-bond
lengths: a) NÀH···OH-3, 2.12 Š; OÀH-3···NH, 2.15 Š; NÀH···OH-6, 2.04 Š; b) NÀ
H···OH-1, 2.03 Š; NÀH···OH-2, 2.26 Š; NÀH···OH-3, 1.94 Š; NÀH···OH-3, 2.02 Š;
NÀH···O-5, 2.03 Š; OÀH-6···NH, 2.05 Š.
Figure 6. Minimum energy structures of the two families of complexes be-
tween 3 and OctbGlc obtained from the conformational search and compat-
ible with the NMR spectroscopy data. Hydrogen bonds between the recep-
tor and carbohydrate are depicted as dashed lines. Selected hydrogen-bond
lengths: a) NÀH···OH-3, 2.12 Š; OÀH-4···NH, 2.09 Š; NÀH···O-5, 2.04 Š; OÀH-
6···NH, 2.12 Š; b) NÀH···OH-3, 2.11 Š; NÀH···OH-4, 2.04 Š; NÀH···O-5, 2.00 Š;
OÀH-6···NH, 2.18 Š.
ometry for the two structures, with the exception of a 1808 ro-
tation of the glucose moiety with respect to the C1–C4 axis.
Unambiguous support for the above interpretation of
threading as the cause of slow kinetics was obtained by replac-
ing the octyl chain with a methyl group in the aglycon moiety.
Titration of 3 with increasing amounts of methyl-b-glucoside
(MebGlc) in CDCl3/DMF 70:30 showed the exclusive formation
of Fast complexes, with one single set of signals shifting upon
complexation and no evidence of Slow adducts (Figure S13 in
the Supporting Information), even though the affinities mea-
sured were comparable to those obtained for OctbGlc (NMR
Even in the absence of pyrrolic rings, amine groups are re-
sponsible for bending the structure into the cleft conforma-
tion. Good agreement between the experimental NOE/ROE
contacts and the corresponding distances obtained from the
calculated structures can be appreciated from Table 4, in which
interatomic distances for Slow and Fast complexes of receptors
2, 3, and 6 are reported.
Thus, the observed Slow and the Fast complexes, that is, ad-
ducts showing slow and fast kinetics, are associated with
threaded and unthreaded geometries, respectively. This may
account for the kinetic phenomenology by considering that
threading necessarily requires a conformational reorganization
of the octyl chain, which slows down the binding process,
whereas no reorganization is required for binding in the un-
threaded geometry. Furthermore, the difference between the
two isomeric Slow(a) and Slow(b) complexes of 2 can be per-
ceived from Figure 5a and b, which shows a closely related ge-
spectroscopy:
BC050 =(3.8Æ0.1) mm;
ITC:
BC050 =(3.69Æ
0.03) mm; cf. Table 2). Clearly, without the octyl chain, the
threaded conformation does not occur and fast kinetics are
restored.
An inventory of the hydrogen-bonding interactions has
been obtained by choosing, from the calculated minimum
energy structures, all O···N interatomic distances shorter than
Table 4. Intermolecular most intense NOE/ROE cross-peaks found in NOESY and ROESY spectra for complexes of OctbGlc with 2, 3, and 6, and correspond-
ing distances [Š] obtained from the calculated minimum energy structures.[a]
OctbGlc·2
Slow(b)
OctbGlc·3
OctbGlc·6
Slow(a)
Fast
Slow
Fast
Slow
Fast
H-5/H-2 (2.87)
H-7/NH-12 (2.57)
H-7’/NH-12* (2.04)
H-8/H-24 (2.48)
H-5/H-2* (2.89)
H-8/H-9 (3.31)
H-8/H-10 (3.59)
H-1/NH-12 (3.49)
H-1/NH-12 (3.47)
H-5/H-2 (2.76)
H-8/H-13* (4.42)
H-1/NH-12* (3.05)
H-2/H-18* (3.05)
H-4/NH-34 (3.74)
H-5/H-2 (3.12)
H-7’/H-7* (2.93)
H-8/H-18 (2.67)
H-7/NH-12 (2.10)
H-7’/NH-12* (2.55)
H-8/H-30 (2.43)
H-8/H-30 (2.39)
H-7/NH-12 (2.17)
H-8’/H-13* (3.17)
[a] Nuclei that became nonequivalent upon complexation are marked with an asterisk.
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6089
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the sum of the van der Waals radii, and selecting those distan-
ces that comply with hydrogen-bonding criteria. The results of
such an inventory are collected in Table 5, from which it is evi-
dent that the observed complex stabilities are determined by
a significant number of hydrogen-bonding interactions. In par-
ticular, it can be appreciated that the contribution from the
aminic groups to the hydrogen-bonding network is prevalent,
involving 16 out of a total of 27 interactions. The most dramat-
ture endowed with improved binding properties, despite the
loss of the preorganization advantage. The improvement relies
on the ability of the flexible structure to establish a larger
number of, or stronger, hydrogen-bonding interactions by
more closely matching the binding requirements of the gluco-
sidic guest. A further advantage of the side-armed flexible ar-
chitecture is the conveniently accessible modification of the
core structure to optimize recognition properties.
The systematic dissection of
the architecture of the parent
Table 5. Hydrogen-bonding interactions and corresponding distances [Š][a] found in the calculated minimum
energy structures of the threaded and unthreaded complexes of OctbGlc with receptors 2, 3, and 6.[b]
cage receptor revealed the con-
tribution from the aminopyrrolic
binding arrangement to the rec-
ognition properties of the recep-
tor, as well as to the conforma-
tional behavior of the structure
upon binding. In particular, it
has been shown that amino
groups provide not only a sub-
stantial contribution to the over-
all affinity, but also the driving
force for bending the monocy-
OctbGlc
OctbGlc·2
OctbGlc·3
OctbGlc·6
unthreaded
threaded (a) threaded (b) unthreaded threaded
unthreaded threaded
O-1
OH-2
OH-3
NH-12 (3.07)
NH-6 (3.05)
NH-34 (2.91) NH-34 (2.94) NH-12* (2.86) NH2-26 (3.01) NH2-26 (2.98) NH-37 (3.03) NH-12* (2.83)
NH-26 (3.05) NH2-20 (3.04)
NH-26 (3.00)
NH-26 (3.01) NH-14* (2.96)
OH-4
O-5
NH2-20 (3.04) NH-12 (2.83)
NH-34 (3.03) NH-12 (3.03) NH-12* (3.00)
NH-34 (3.03)
OH-6
NH-14 (3.07) NH-6* (3.02) NH-26 (3.05) NH-14 (3.03) NH-14* (3.07) NH-14 (3.04) NH-26 (2.98)
[a] Interatomic distances shorter than the sum of van der Waals radii and complying with hydrogen-bonding
criteria. [b] Nuclei that became nonequivalent upon complexation are marked with an asterisk.
clic structure into
conformation.
a
cleft
A rationale for the intriguing
ic drop in affinity is observed for the stepwise loss of two
amino groups in 4 and 5. The anticipated conformational
origin of the very similar affinities exhibited by 2 and 3 finds
a rationale in the changes induced by the pyrrolic binding arm
on the hydrogen-bonding network. Indeed, in both the Slow
and Fast complexes, hydrogen bonds established by this pyr-
role pulls glucose slightly out of the cleft, which causes the
loss of a hydrogen-bonding interaction of the sugar with one
of the endocyclic pyrroles. This feature compensates for the
advantage of the new interaction and leaves the affinity un-
changed with respect to 3. In contrast, the introduction of
a second pyrrolic binding arm into the architecture significant-
ly increases the number of hydrogen-bonding interactions by
pushing glucose closer to the macrocycle, with which it can es-
tablish an additional hydrogen bond. Because the added bind-
ing arm does not itself participate in hydrogen bonding, its
contribution most likely consists of sealing the receptor’s
cavity through nonbonding interactions. Interestingly, whereas
for 2 and 3 the contribution of the Slow and Fast complexes
to the overall conformer population is comparable, counterin-
tuitively, for 6, the contribution of the Fast complex becomes
prevalent, so that the increase in affinity for OctbGlc can be as-
cribed mainly to the unthreaded structure. This evidence can,
however, be easily explained on the basis of the six hydrogen-
bonding interactions, provided by three chelating aminopyrrol-
ic units, established by the unthreaded structure, which is the
largest number of interactions of the whole set.
occurrence of concomitant slow and fast exchanging 1:1 com-
plexes on the NMR spectroscopy timescale was provided by
combining experimental NMR spectroscopy data and molecu-
lar modeling calculations, which, besides giving a description
of the 3D binding modes in solution, revealed the existence of
two distinct types of glucoside complexes—a rotaxane-like
threaded adduct and an unthreaded structure—that account
for the observed kinetics, while cooperatively contributing to
the overall binding affinity.
The most interesting conclusion appears to be the finding
that effective recognition in a polar competitive medium can
indeed be achieved by carefully designed adaptive structures
endowed with aminopyrrolic hydrogen-bonding arrangements.
This finding is a step forward in the design of synthetic recep-
tors for the recognition of carbohydrates in a physiological
medium.
Acknowledgements
This work has been carried out in the frame of the COST
Action CM1102. High-field NMR spectra were acquired at the
Magnetic Resonance Center (CERM) facilities of the Università
di Firenze. We thank M. Lucci for technical assistance. Dr. A.
Minoja from Bruker Italia s.r.l. is gratefully acknowledged for
kind assistance. Ente Cassa di Risparmio di Firenze (Italy) is ac-
knowledged for granting an ITC nano calorimeter (grant no.
2009.0576).
Conclusion
Keywords: carbohydrates
· hydrogen bonds · molecular
recognition · receptors · structure elucidation
We have shown that converting a preorganized cage receptor
into a closely related adaptive structure results in an architec-
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6090
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[1] a) H.-J. Gabius, S. AndrØ, J. JimØnez-Barbero, A. Romero, D. Solís, Trends
Biochem. Sci. 2011, 36, 298–313; b) The Sugar Code: Fundamentals of
Glycosciences (Ed.: H.-J. Gabius), Wiley-VCH, Weinheim, 2009; c) H. E.
Murrey, L. C. Hsieh-Wilson, Chem. Rev. 2008, 108, 1708–1731; d) T. K.
Lindhorst, Essentials of Carbohydrate Chemistry and Biochemistry, Wiley-
VCH, Weinheim, 3rd ed., 2007; e) Carbohydrates in Chemistry and Biol-
ogy (Eds.: B. Ernst, G. W. Hart, P. Sinay¨), Wiley-VCH, Weinheim, 2000,
Part I, Vol. 2 and Part II, Vol. 4.
[2] a) Y. van Kooyk, G. A. Rabinovich, Nat. Immunol. 2008, 9, 593–601;
b) C. N. Scanlan, J. Offer, N. Zitzmann, R. A. Dwek, Nature 2007, 446,
1038–1045; c) D. R. Burton, R. A. Dwek, Science 2006, 313, 627–628;
d) N. L. Letvin, Nat. Rev. Immunol. 2006, 6, 930–939; e) D. H. Dube, C. R.
Bertozzi, Nat. Rev. Drug Discovery 2005, 4, 477–487; f) R. Medzhitov,
C. A. Janeway, Jr., Science 2002, 296, 298–300.
5064–5072; c) C. Nativi, O. Francesconi, G. Gabrielli, A. Vacca, S. Roelens,
Chem. Eur. J. 2011, 17, 4814–4820; d) A. Ardꢀ, F. J. CaÇada, C. Nativi, O.
Francesconi, G. Gabrielli, A. Ienco, J. JimØnez-Barbero, S. Roelens, Chem.
Eur. J. 2011, 17, 4821–4829; e) M. Cacciarini, C. Nativi, M. Norcini, S. Sta-
derini, O. Francesconi, S. Roelens, Org. Biomol. Chem. 2011, 9, 1085–
1091; f) A. Ardꢀ, C. Venturi, C. Nativi, O. Francesconi, G. Gabrielli, F. J.
CaÇada, J. JimØnez-Barbero, S. Roelens, Chem. Eur. J. 2010, 16, 414–
418; g) A. Ardꢀ, C. Venturi, C. Nativi, O. Francesconi, F. J. CaÇada, J. JimØ-
nez-Barbero, S. Roelens, Eur. J. Org. Chem. 2010, 64–71; h) C. Nativi, M.
Cacciarini, O. Francesconi, G. Moneti, S. Roelens, Org. Lett. 2007, 9,
4685–4688; i) M. Cacciarini, E. Cordiano, C. Nativi, S. Roelens, J. Org.
Chem. 2007, 72, 3933–3936; j) C. Nativi, M. Cacciarini, O. Francesconi, A.
Vacca, G. Moneti, A. Ienco, S. Roelens, J. Am. Chem. Soc. 2007, 129,
4377–4385; k) A. Vacca, C. Nativi, M. Cacciarini, R. Pergoli, S. Roelens, J.
Am. Chem. Soc. 2004, 126, 16456–16465.
[3] a) J. Arnaud, A. Audfray, A. Imberty, Chem. Soc. Rev. 2013, 42, 4798–
4813; b) J. Balzarini, Nat. Rev. Microbiol. 2007, 5, 583–597; c) A. Imberty,
E. P. Mitchell, M. Wimmerovꢀ, Curr. Opin. Struc. Biol. 2005, 15, 525–534;
d) M. R. Pratt, C. R. Bertozzi, Chem. Soc Rev. 2005, 34, 58–68; e) C. R. Ber-
tozzi, L. L. Kiessling, Science 2001, 291, 2357–2363.
[7] a) O. Francesconi, A. Ienco, G. Moneti, C. Nativi, S. Roelens, Angew.
Chem. 2006, 118, 6845–6848; Angew. Chem. Int. Ed. 2006, 45, 6693–
6696; b) O. Francesconi, M. Gentili, S. Roelens, J. Org. Chem. 2012, 77,
7548–7554.
[4] a) C. Ke, H. Destecroix, M. P. Crump, A. P. Davis, Nat. Chem. 2012, 4,
718–723; b) B. Sookcharoenpinyo, E. Klein, Y. Ferrand, D. B. Walker, P. R.
Brotherhood, C. Ke, M. P. Crump, A. P. Davis, Angew. Chem. 2012, 124,
4664–4668; Angew. Chem. Int. Ed. 2012, 51, 4586–4590; c) N. P. Barwell,
M. P. Crump, A. P. Davis, Angew. Chem. 2009, 121, 7809–7812; Angew.
Chem. Int. Ed. 2009, 48, 7673–7676; d) Y. Ferrand, M. P. Crump, A. P.
Davis, Science 2007, 318, 619–622.
[5] For recent reviews, see: a) A. P. Davis, Nature 2010, 464, 169–170; b) S.
Jin, Y. Cheng, S. Reid, M. Li, B. Wang, Med. Res. Rev. 2010, 30, 171–257;
c) D. B. Walker, G. Joshi, A. P. Davis, Cell. Mol. Life Sci. 2009, 66, 3177–
3191; d) A. P. Davis, Org. Biomol. Chem. 2009, 7, 3629–3638; e) M.
Mazik, Chem. Soc. Rev. 2009, 38, 935–956; f) S. Kubik, Angew. Chem.
2009, 121, 1750–1753; Angew. Chem. Int. Ed. 2009, 48, 1722–1725;
g) A. P. Davis, T. D. James, in Functional Synthetic Receptors (Eds.: T.
Schrader, A. D. Hamilton), Wiley-VCH, Weinheim, 2005, pp. 45–109.
[6] a) O. Francesconi, C. Nativi, G. Gabrielli, M. Gentili, M. Palchetti, B.
Bonora, S. Roelens, Chem. Eur. J. 2013, 19, 11742–11752; b) C. Nativi, O.
Francesconi, G. Gabrielli, I. De Simone, B. Turchetti, T. Mello, L. Di Cesare
Mannelli, C. Ghelardini, P. Buzzini, S. Roelens, Chem. Eur. J. 2012, 18,
[8] A. Vacca, O. Francesconi, S. Roelens, Chem. Rec. 2012, 12, 544–566.
[9] BC50 Calculator is a computer program developed for PC platforms and
is available for free upon request from the corresponding author. A de-
tailed description of the equations used by the program for computing
BC50 and BC050 is reported in previous papers, see refs. [6c], [6j], and [8].
[10] For examples of the application of NMR spectroscopy techniques assist-
ed by molecular modeling, see: a) J. JimØnez-Barbero, E. Dragoni, C.
Venturi, F. Nannucci, A. Ardꢀ, M. Fontanella, S. AndrØ, F. J. CaÇada, H. J.
Gabius, C. Nativi, Chem. Eur. J. 2009, 15, 10423–10431; b) M. Kolympadi,
M. Fontanella, C. Venturi, S. AndrØ, H. J. Gabius, J. JimØnez-Barbero, P.
Vogel, Chem. Eur. J. 2009, 15, 2861–2873; c) S. Mari, F. J. CaÇada, J. Ji-
mØnez-Barbero, A. Bernardi, G. Marcou, I. Motto, I. Velter, F. Nicotra, B.
La Ferla, Eur. J. Org. Chem. 2006, 2925–2933; d) S. Mari, H. Posteri, G.
Marcou, D. Potenza, F. Micheli, F. J. CaÇada, J. JimØnez-Barbero, A. Ber-
nardi, Eur. J. Org. Chem. 2004, 5119–5225.
Received: January 28, 2014
Published online on April 3, 2014
Chem. Eur. J. 2014, 20, 6081 – 6091
www.chemeurj.org
6091
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim