Synthesis, biological evaluation, wac and NMR studies of s-galactosides and non-carbohydrate ligands of cholera toxin based on polyhydroxyalkylfuroate moieties

Article abstract of DOI:10.1002/chem.201302786

The synthesis of several non-carbohydrate ligands of cholera toxin based on polyhydroxyalkylfuroate moieties is reported. Some of them have been linked to D-galactose through a stable and well-tolerated S-glycosidic bond. They represent a novel type of non-hydrolyzable bidentate ligand featuring galactose and polyhydroxyalkylfuroic esters as pharmacophoric residues, thus mimicking the GM1 ganglioside. The affinity of the new compounds towards cholera toxin was measured by weak affinity chromatography (WAC). The interaction of the best candidates with this toxin was also studied by saturation transfer difference NMR experiments, which allowed identification of the binding epitopes of the ligands interacting with the protein. Interestingly, the highest affinity was shown by non-carbohydrate mimics based on a polyhydroxyalkylfuroic ester structure. No carbs here: Saturation transfer difference (STD) NMR studies of bidentate ligands of cholera toxin (see figure, WAC = weak affinity chromatography) show the methylfuran moiety as the main contact point in the interaction with the toxin. Several polyhydroxyalkylfuroate-based structures are synthesized and analyzed and show similar or even better affinity than the bidentate ligands. They constitute the first examples of non-carbohydrate ligands for cholera toxin. Copyright

Full text of DOI:10.1002/chem.201302786

FULL PAPER
DOI: 10.1002/chem.201302786
Synthesis, Biological Evaluation, WAC and NMR Studies of S-Galactosides
and Non-Carbohydrate Ligands of Cholera Toxin Based on
Polyhydroxyalkylfuroate Moieties
Javier Ramos-Soriano,[a] Ulf Niss,[b] Jesffls Angulo,[c, e] Manuel Angulo,[d]
Antonio J. Moreno-Vargas,*[a] Ana T. Carmona,[a] Sten Ohlson,[b] and
Inmaculada Robina*[a]
Abstract: The synthesis of several non-
carbohydrate ligands of cholera toxin
based on polyhydroxyalkylfuroate moi-
eties is reported. Some of them have
been linked to d-galactose through
a stable and well-tolerated S-glycosidic
bond. They represent a novel type of
non-hydrolyzable bidentate ligand fea-
turing galactose and polyhydroxyalkyl-
furoic esters as pharmacophoric resi-
dues, thus mimicking the GM1 ganglio-
side. The affinity of the new com-
pounds towards cholera toxin was mea-
sured by weak affinity chromatography
(WAC). The interaction of the best
candidates with this toxin was also
studied by saturation transfer differ-
ence NMR experiments, which allowed
identification of the binding epitopes
of the ligands interacting with the pro-
tein. Interestingly, the highest affinity
was shown by non-carbohydrate
mimics based on a polyhydroxyalkyl-
furoic ester structure.
Keywords: weak affinity chroma-
tography (WAC) · biomimetic syn-
thesis · carbohydrates · cholera tox-
in · NMR spectroscopy
Introduction
responsible for binding to ganglioside GM1 (Galb1-3Gal-
NAcb1-4[NeuAca2-3]Galb1-4Glcb1,1-ceramide) on the sur-
face of the membrane of intestinal cells.[1] These toxins are
responsible for diarrheal diseases and several other disor-
ders. Due to the importance of the processes that they pro-
mote, the study of the complexes between gangliosides and
AB5 toxins is very relevant. The B pentamer of CT interacts
with the monovalent oligosaccharide fragment of GM1 gan-
glioside with a strong affinity (Kd =43 nm, as measured by
isothermal calorimetry). This interaction represents one of
the strongest carbohydrate–protein interactions, the complex
GM1:CT being an extremely well characterized pair in the
field of sugar–protein interactions. In fact, the interaction
between GM1 and CT has been widely studied through cal-
orimetry,[2] X-ray diffraction,[3] and computational meth-
ods.[4] It is known that the two pharmacophoric units of
GM1 are the terminal galactose and the sialic acid residues.
The conformational preorganization of these pharma-
Cholera toxin (CT) from Vibrio cholerae, and the closely re-
lated heat-labile toxin (LT) of Escherichia coli, belong to
the bacterial AB5 holotoxin family, which comprises a single
catalytically active component A subunit and five identical
B subunits forming a regular pentamer B5. This pentamer is
[a] J. Ramos-Soriano, Dr. A. J. Moreno-Vargas, Dr. A. T. Carmona,
Prof. Dr. I. Robina
Department of Organic Chemistry, Faculty of Chemistry
University of Seville
Prof. Garcꢀa Gonzꢁlez 1, 41012 Seville (Spain)
Fax : (+34)954624960
E-mail: ajmoreno@us.es
robina@us.es
[b] U. Niss, Prof. Dr. S. Ohlson
Department of Chemistry and Biomedical Sciences
Linnaeus University
391 82 Kalmar (Sweden)
AHCTUNGTREGcNNNU ophores and the distance between them is essential to in-
teract efficiently with CT.[5] The heat-labile enterotoxin (LT)
has a homology of about 80% with CT and is basically iden-
tical in binding with GM1 and in the toxic effect, although
LT causes less severe symptoms than CT.[6] In consequence,
the study of the interaction CT:GM1 is a valid model for
the interaction with LT. Several GM1 derivatives or mimet-
ics containing terminal galactose residues and/or sialic acid
have been described to have affinity towards CT and LT.
They bind reversibly to the B subunit (forming an associa-
tion complex) and act as inhibitors of the recognition be-
tween toxins and host mucins, thus preventing infec-
[c] Dr. J. Angulo
Glycosystems Laboratory
Instituto de Investigaciones Quꢀmicas (IIQ)
CSIC and University of Seville
Americo Vespucio 49, 41092 Seville (Spain)
[d] Dr. M. Angulo
Centro de Investigaciꢂn Tecnolꢂgica e Innovaciꢂn (CITIUS)
University of Seville
Avda Reina Mercedes s/n, 41012 Seville (Spain)
[e] Dr. J. Angulo
Current address: School of Pharmacy, University of East Anglia
Norwich Research Park, Norwich NR4 7TJ (UK)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201302786.
AHCTUNGTRENNUNG
tions.[1a,5,7]
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
17989

Figure 2. General structure for E- and P-selectin ligands.
Based on our previous results on polyhydroxyalkylfuryl
thiofucosides, which show affinity towards selectins,[15c] we
hypothesized that S-galactosides bearing a polyhydroxyalkyl-
furan aminoester moiety as a non-proteinogenic polyhydrox-
ylated amino acid could interact with the enterotoxins CT
and LT, thus acting as GM1 glycomimetics. We report
herein the preparation and biological study by WAC of new
S-galactosides of Generations I and II (Figure 3). Both gen-
Figure 1. General structures for GM1 mimetics.
Bernardi and co-workers have reported the synthesis of
a series of pseudo-tetrasaccharides bearing a cyclohexane
diol moiety, such as 1 (Figure 1), as GM1 mimetics, which
displayed the same affinity as the natural ligand.[8] Other
pseudo-oligosaccharides obtained through replacing the
sialic acid moiety by simpler hydroxyacids presented dissoci-
ation constants in the millimolar to micromolar range.[9]
This group has also reported metabolically stable bifunction-
al compounds 2,[10] based on C-galactosides having galactose
and sialic acid moieties connected through a triazole moiety
as linker. The affinity of the compounds towards CT was
tested by weak affinity chromatography (WAC). Affinity
could be enhanced by up to one or two orders of magnitude
over those of the individual pharmacophoric sugar residues.
Verlinde and co-workers[11] first described that m-nitro-
phenyl-a-d-galactopyranoside (MNPG) 3 is an inhibitor of
the GM1–enterotoxin interaction. On its side, Fan and co-
workers[7b,12] have described that 3 exhibits an IC50 of
600 mm. Galactose itself displays a high specificity but low
affinity relative to CT (IC50 =45 mm). Ohlson and co-work-
ers[13] have determined for MNPG 3 an affinity constant
Kd =1.1 mm, by means of the WAC method based on immo-
bilized cholera toxin subunit B (CTB). Derivatives such as 4
are also CTB5 antagonists in the low millimolar range
(Figure 1).[14]
Our group has reported the preparation and biological
studies of novel thiofucosides incorporating hydroxylated
non-proteinogenic amino acids ending in carboxylate groups
as aglycons (Figure 2). These compounds have shown affini-
ty towards E- and P-selectins in the millimolar range.[15] Se-
lectins are glycoproteins that present heterophilic binding
with their receptors,[16] thereby mediating cellular interac-
tions through the lectin domain and the cell surface carbo-
hydrate ligands.[17] Another group of proteins that, like se-
lectins, presents heterophilic binding with their receptors are
pathogen-secreted enterotoxins.[18]
Figure 3. General structures of new S-linked galactosides and fragments.
AA=amino acid, Bz=benzoyl, Bn=benzyl.
erations of compounds are S-galactosides containing a poly-
hydroxyalkylfuran moiety in the aglycon, which is linked
through a spacer to the galactose moiety. For compounds of
Generation I, the polyhydroxyalkylfuran moiety is attached
to the spacer by the furan unit (Generation I: “head connec-
tion”). In the case of compounds of Generation II, a polyhy-
droxyalkylfuran moiety is attached to the spacer by a polyol
chain (Generation II: “tail connection”).[22]
These compounds are structurally simpler and hydro-
AHCTUNGTREGlNNNU ytically more stable than the natural ganglioside GM1 and
have appropriate groups to generate structural diversity and
to further assemble them into multivalent structures. We
demonstrate that the polyhydroxyalkylfuran moiety contrib-
17990
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17989 – 18003

Non-Carbohydrate Ligands of Cholera Toxin
FULL PAPER
utes effectively to the affinity, and that the type of connec-
tion to galactose, head or tail connection, is crucial for the
binding affinity. To establish which part of the molecule is
more important for binding affinity, fragment compounds
have also been prepared and evaluated by WAC. Thus, Gen-
eration III is constituted by differently configured and sub-
stituted polyhydroxyalkylfurans.
oxygen counterparts is an important advantage for our com-
pounds with regard to tolerance by most biological sys-
tems.[20]
Results and Discussion
We show herein that some of these compounds present
binding affinities in the same order as that of known sugar-
derived MNPG 3. Their binding affinity depends on the
type and site of substitution, that is, in the furan ring (“head
substitution”) or in the polyolic side chain (“tail substitu-
tion”). Some of the prepared compounds represent the first
examples of non-sugar-based antagonists of enterotoxins.
With the aim of accurately obtaining structural information
on the interactions of this family of ligands with the CTB,
saturation transfer difference (STD) NMR experiments are
also presented.
Compounds of Generation IV (Figure 3) present the
sugar fragments of compounds of Generations I and II,
which additionally are the thio analogues of the reported
MNPG series.[12,13] Notably, the introduction of a sulfur
group at the anomeric position provides good acid and
enzyme-mediated hydrolytic stability.[19] In addition, the
higher water solubility of sulfur derivatives relative to their
Synthesis of compounds from Generations I and II: The syn-
theses of compounds of Generations I and II were carried
out by coupling different thioglycosides with conveniently
protected polyhydroxyalkylfuroates. Thus, the preparation
of compounds 16–19 (Generation I) was carried out starting
from penta-O-acetyl-b-d-galactopyranoside (5) by applying
two main strategies. Direct glycosylation of a thiol-contain-
ing compound such as Fmoc-cysteamine in the presence of
TMSOTf (method A) gave aminoalkyl b-d-thiogalacto-
AHCTUNGTREGpNNNU yranoside 6 in 72% yield. Transformation of 5 into aceto-
bromogalactose and glycosylation with Boc-cysteine methyl
ester (method B) afforded b-d-thiogalactosyl aminoester 7
in 57% yield (Scheme 1). 1’-Acetamido-2’,3’,4’-tri-O-acetyl-
1-yl-furancarboxylic acid 10 and per-O-acetyl-1-yl-furancar-
boxylic acid 11 were obtained by hydrogenation of com-
pounds 9 and 8, respectively, followed by acetylation. The
introduction of the azido moiety at C-1’ in compound 8 was
carried out as described previously by our group,[21] through
Scheme 1. Reagents and conditions: a) Method A: TMSOTf, CH2Cl2, FmocNHCH2CH2SH, 72% for 6; b) Method B: 1. HBr/AcOH 33%, 2. Na2CO3,
TBAHS, AcOEt/H2O, Boc-Cys-OMe, 57% for 7; c) 1. H2, Pd-C, 2. Ac2O, Py, DMAP, 80% for 10, 60% for 11; d) 1. For Pg:Fmoc, PIP/DMF (20%) and
for Pg:Boc, TFA/CH2Cl2 (20%), 2. PyBOP, DIPEA, DMF, 56% for 12, 42% for 13, 89% for 14, 89% for 15; e) NaOMe, MeOH, 100% for 16–19, 10a,
and 11a; f) 1. PIP/DMF (20%), 2. Ac2O/Py, 3. NaMeO/MeOH, 63% (three steps) for 20; 1. NaMeO/MeOH, 2. TFA/CH2Cl2 (20%), 94% (two steps) for
21. TMSOTf=trimethylsilyl triflate, Fmoc=9-fluorenylmethoxycarbonyl, TBAHS=tetrabutyl ammonium hydrogenosulfate, Boc=tert-butoxycarbonyl,
Cys=cysteine, Py=pyridine, DMAP=4-dimethylaminopyridine, Pg=protecting group, PIP=piperidine, TFA=trifluoroacetic acid, PyBOP=(benzotria-
zol-1-yloxy)trispyrolidinophosphonium hexafluorophosphate, DIPEA=N,N-diisopropylethylamine.
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17991

A. J. Moreno-Vargas, I. Robina et al.
the formation of a cyclic sulfite as intermediate. Removal of
the Fmoc protecting group in 6 with piperidine/DMF (20%)
followed by coupling with furancarboxylic acids 10 and 11
by using PyBOP and DIPEA as condensing agents furnished
adducts 12 and 13 in 56 and 42% yield, respectively. On the
other hand, acid treatment of 7 with TFA/CH2Cl2 (20%) fol-
lowed by coupling with 10 and 11, under the same reaction
conditions, afforded 14 and 15 in 89% yield in both cases.
Final deprotection under Zemplꢄn conditions gave com-
pounds 16–19 in quantitative yield. With the purpose of
comparing the biological activities of the compounds of
Generation I with their corresponding fragments, deprotec-
tion of 6, 7, 10, and 11 under standard conditions gave 20,
21, 10a, and 11a, respectively, in good overall yields
(Scheme 1).
Compounds from Generation II (26, 30) bearing an alkyl
polyhydroxyalkylfuroate moiety were prepared through two
synthetic sequences. The synthesis of compound 26 was car-
ried out from known ethyl 4’-aminotriol-1-yl furoate 24,
which was obtained from the corresponding azido derivative
23 as described previously by our group[22] (Scheme 2). Simi-
larly, compound 30 was prepared from benzyl 4’-azidotriol-
1-yl furoate 28, obtained from compound 8[21] in good over-
all yield, after chemoselective tosylation of the hydroxyl pri-
mary position and conventional azido displacement. Reac-
tion of 24 with carboxyalkyl-b-d-thiogalactoside 22,[23] under
peptide coupling conditions, with PyBOP and DIPEA gave
25 in 49% yield, which was deprotected under Zemplꢄn
conditions furnishing 26 in quantitative yield. “Click chemis-
try” coupling between azido derivative 28 and b-galactosyl
alkyne 27 furnished triazole 29 in 82% yield, which was fi-
nally deprotected to give 30 in quantitative yield. b-Galacto-
syl alkyne 27 was obtained from per-O-acetylated b-galacto-
pyranoside through Lewis acid catalyzed (BF3·Et2O) glyco-
sylation with thiourea, according to the procedure reported
by Ibatullin and co-workers,[24] followed by treatment with
Et3N[25] and reaction with propargyl bromide (Scheme 2).
Scheme 2. Reagents and conditions: a) DIPEA, PyBOP, DMF, 49%;
b) NaOMe, MeOH, 100%; c) 1. tosyl chloride (TsCl), Py, ꢀ158C, 69%,
2. NaN3, DMF, 808C, 85%; d) CuSO4, sodium ascorbate, MeOH/H2O,
82%.
Ligand evaluation of compounds from Generations I and II
by weak affinity chromatography: To evaluate the binding
affinity of these ligands and their corresponding fragments
to CT, the method of WAC, based on immobilized CTB, was
used. The method is mostly appropriate for quantifying tran-
sient binding events, such as sugar–protein interactions. In
our case, it was found to be the method of choice, because
for our ligands the binding affinity was suspected to be
rather weak, in the millimolar range. This method has previ-
ously been used for the evaluation of CTB binders, both in
terms of small compounds and galactosyl derivatives as bi-
dentate ligands.[10,13] A HPLC column with immobilized re-
combinant CTB (rCTB) was used and small amounts of the
synthesized ligands were injected. An elution solution was
subsequently applied to the column to make each ligand run
through the column to be released from the entrapment.
The retardation (expressed as the retention factor k’) is di-
rectly related to the affinity of the interaction.[26] A com-
pound of known affinity was used as a reference, and the
dissociation constant of the ligand, Kd, was calculated from
k’ by calibrating the retention with the Kd of the reference
compound. In our case, MNPG, a well-known CT ligand,
was used as reference to calibrate our experiments. The re-
sults of the WAC analysis of our ligand library and the cor-
responding fragments compared with MNPG are summa-
AHCTUNGTRENNUNG
tion I) did not show binding affinity to CTB (Table 1, en-
tries 2–5), and neither did N,O-unprotected b-d-thiogalacto-
pyranoside fragments 20 and 21 and polyhydroxyalkylfuran
carboxylic acids 10a and 11a (Table 1, entries 8–11). In con-
trast, compounds bearing a “tail connection” (Genera-
AHCTUNGTREGtNNNU ion II) and their corresponding fragments showed signifi-
cant activities, and some interesting candidates were identi-
fied. Thus, the best result for Generation II was obtained for
compound 30 with Kd =1.05 mm. Compound 26 showed
17992
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17989 – 18003

Non-Carbohydrate Ligands of Cholera Toxin
FULL PAPER
Table 1. Retention factor (k’) and dissociation constant (Kd) of CT li-
gands and their fragments at 228C determined with WAC at pH 7.
Entry Compound Generation k’
Kd [mm] KMNPG/Kd
Ligands
1
2
3
4
5
6
7
8
MNPG
16
17
18
19
26
30
20
21
–
I
I
I
I
II
II
I
I
I
1.58 1.1[a]
1
0.19 NB
0.18 NB
0.14 NB
0.12 NB
0.36 4.7
1.63 1.05
0.13 NB
0.15 NB
0.03 NB
0.20 NB
0.52 3.3
2.63 0.6
–
–
–
–
0.23
1.05
–
Fragments
9
–
–
–
10
11
12
13
10a
11a
24
I
II
II
0.33
1.83
28
[a] Ref. [13]. NB=no binding.
Figure 4. STD NMR (a) and reference (b) spectra of a sample containing
1 mm ligand 30 in the presence of 5 mm subunit B of cholera toxin (CTB;
25 mm in binding sites). The spectra were obtained at 700 MHz and 278 K
(saturation time, tsat =2 s). Key signals are labeled on the difference spec-
trum (a) .
Kd =4.7 mm. Interestingly, the aminopolyhydroxyalkylfuran
ethyl ester 24 and the azidopolyhydroxyalkylfuran benzyl
ester 28 exhibited remarkable binding affinities of Kd =
3.3 mm and 0.6 mm, respectively. The latter is almost twice
as active as MNPG under the same conditions.[13]
other hand, Figure 6 shows the relative saturation values for
each ligand proton, obtained from normalization of the STD
initial growth rates (see Experimental Section), which gives
an accurate picture of the binding epitope of 30 for its inter-
action with CTB. Onto this quantitative map of ligand bind-
ing, it is clear that the entire molecule makes contacts with
the surface of CTB in the bound state. Interestingly, the sat-
uration level decreases from both terminal ends of 30 to-
wards the triazole ring, indicative of a bidentate-like binding
to CTB. What is more, the furan and carboxybenzyl-contain-
ing end of the molecule received the largest amount of satu-
ration (Figure 5, bottom right, and Figure 6). This is a key
result, as it explains the previously observed contribution of
the polyhydroxyalkylfuran moiety to the binding affinity to
CTB, by WAC experiments. The protein indeed closely rec-
ognizes this moiety, most likely enhancing the enthalpy of
the interaction due to the increase in the number of contacts
of the ligand with the protein surface.
Interaction studies by NMR spectroscopy: binding epitopes:
To obtain structural information on the interactions of this
family of ligands with the CTB, we carried out STD NMR
experiments.[27] With these experiments we identified the
main contact points of the ligands with the protein in the
binding pocket (the so-called ligand binding epitope).[28]
This is possible because the binding process transfers
1H magnetization, created selectively on the protein, onto
1
the ligand molecule due to intermolecular H–1H NOE pro-
cesses in the bound state and the fast exchange with the free
state.[27–29] The presence of signals in the resulting STD
NMR spectrum of the ligand reveals the existence of bind-
ing in solution and, in addition, the intensities of the STD
signals of the ligand are related to ligand–protein spatial
proximity; that is, the stronger the signal, the closer the con-
tact of that part of the ligand to the protein surface.
The STD NMR spectrum of 30 in the presence of rCTB is
shown in Figure 4. This compound was selected for a detailed
structural study by NMR spectroscopy as the best represent-
In addition, we also carried out transferred NOESY ex-
periments, to elucidate whether the binding process involves
conformational selection on the ligand, which would con-
tribute negatively to the free energy of the interaction pro-
cess by an entropic penalty. Yet, the similarity of the
NOESY experiments in the presence and in the absence of
CTB (see Supporting Information) indicated that the pro-
tein does not select any particular conformer of 30, or that
it simply recognizes the most populated one in the free
state.
The binding epitope of 30 from STD NMR spectroscopy
(Figure 6), and the affinity studies, were both supportive of
a bidentate binding of the ligand in which both terminal
ends, the galactose and the furan-carboxybenzyl residues,
are strongly involved in establishing contacts with the pro-
tein. Nevertheless, to confirm this issue and to rule out non-
specific interactions, we formally “dissected” the ligand into
the two constituent moieties. We then carried out STD
ACHTUNGTRENNUNGative of the diverse set of synthesized molecules, as it con-
tains different chemical elements sought to afford affinity
for CTB. The experiments were carried out at 700 MHz on
a sample containing 1 mm ligand and 5 mm rCTB, which, due
to its pentameric nature, corresponds to a total concentra-
tion of 25 mm in binding sites (i.e., ligand-to-receptor ratio of
40:1). The spectra at different saturation times (see Experi-
mental Section) showed clear signals (in Figure 4, tsat =2 s),
thus confirming that we were able to detect the binding of
30 to rCTB in solution.
The evolution of the STD signals with saturation time
(buildup curves) is shown in Figure 5. It shows that 30 re-
ceived saturation all along the molecule, so that the binding
is not simply through one end of the molecule, which is fre-
quently the case in lectin–carbohydrate interactions. On the
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17993

A. J. Moreno-Vargas, I. Robina et al.
Figure 5. STD buildup curves of 30 interacting with the CTB. The signals have been split for simplicity into four graphs (four different regions of the
ligand 30). The most intense STD responses came from the galactose and aromatic ends.
NMR experiments on a ligand constituted by the galactose
terminal end 20 (see Scheme 1) and another one constituted
by the polyhydroxyalkylfuran moiety 33 (see Scheme 3,
Generation III). Their STD NMR spectra in the presence of
CTB are shown in Figure 7. Both molecules showed signifi-
cant STD NMR signals in their spectra, which indicates that
both are binding to CTB in solution.
It is worth mentioning the observation of STD signals
from 20, which highlights the
outstanding sensitivity of STD
NMR spectroscopy for weak
protein–ligand interactions, as
20 binding was not detectable
by the WAC experiments. The
comparison of global STD in-
tensities of 30, 20, and 33 (see
Supporting Information) clearly
indicated that 20 is the weakest
binder, in agreement with the
WAC data (Tables 1 and 2).
The binding epitopes from the
analysis of the corresponding
STD buildup curves are shown
in Figure 8.
The resulting binding epi-
topes confirm that ligands 20
Figure 6. Binding epitope of 30 for its interaction with the CTB. The numbers represent relative values of satu-
ration after their normalization related to the most intense one (assigned 100%), obtained from STD initial
slopes (see Experimental Section). Both molecular ends receive the largest amounts of saturation from the
protein, supportive of a bidentate binding.
17994
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17989 – 18003

Non-Carbohydrate Ligands of Cholera Toxin
FULL PAPER
Scheme 3. Synthesis of compounds of Generation III (1). Reagents and conditions: a) H2S(g), Py/H2O (1:1), 08C!RT; b) acetic anhydride, MeOH, RT;
c) BzCl, Et3N, dry DMF, 58C!RT; d) propargyl alcohol, CuSO4 (aq), sodium ascorbate (aq), MeOH, RT; e) CeCl3·7H2O, NaI, SiO2, MeCN; f) TsCl, Py,
ꢀ158C; g) NaN3, DMF, 808C.
Table 2. Retention factor (k’) and dissociation constant (Kd) of CT li-
gands from Generation III at 228C determined with WAC at pH 7.
and 33 bind to CTB in solution with binding modes
(Figure 8) that involve very close contacts with the protein
through the galactose residue and the furan-carboxybenzyl
end, respectively, similar to those observed for the binding
of 30. Therefore, with these two extra binding experiments
we were able to demonstrate that both terminal ends of 30
are indeed specific binders of CTB, as independent mole-
cules. The protein recognizes both structures on ligands 20
and 33 and when linked on the same molecule, 30, demon-
strates that both the sugar and the non-sugar moieties are
contributing to the binding affinity of the ligand 30 for the
CTB.
It is well known that a terminal galactose residue consti-
tutes a pharmacophoric unit of the natural ligand GM1 gan-
glioside. Our NMR data on 30 and 20 are in very good
agreement with this, showing that the region of the sugar
ring outlined by carbon atoms C4, C5, and C6 is involved in
close contacts with the protein (Figure 6 and Figure 8a),
thus strongly supporting the participation of this part of the
hexopyranose ring in the stacking interactions with the side
chain of the amino acid Trp-88, as reported in the case of
GM1.[3b] The NMR results also highlight the relevance of
the furan moiety of 33 in the binding process (Figure 6 and
Figure 8b), and the importance of the distance between the
two pharmacophoric moieties is demonstrated by the dra-
matic reduction in binding affinity to CTB in the case of li-
gands of Generation I (head-connected).
Entry
Compound
k’
Kd [mm])
KMNPG/Kd
1
2
3
4
5
6
7
8
MNPG
23
24
34
32
36
28
31
48
33
35
37
40
45d-arab
47
49d-arab
49l-arab
49d-xyl
49l-xyl
50d-arab
50l-arab
50d-xyl
50l-xyl
51d-arab
51l-arab
51d-xyl
51l-xyl
52
1.58
0.43
0.52
0.15
0.14
1.20
2.63
3.32
0.16
1.16
1.16
0.31
7.6
1.1[a]
3.6
3.3
NB
NB
1.3
1
0.3
0.33
–
–
0.85
1.83
2.34
–
0.8
0.8
0.2
5.5
0.24
–
0.25
0.27
0.23
0.23
2.2
2.2
2.2
2.2
2.16
1.83
1.57
1.37
–
0.6
0.47
NB
1.35
1.35
5.1
0.2
4.6
NB
4.4
4.1
4.1
4.1
0.5
0.5
0.5
0.5
0.51
0.6
0.7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
0.34
0
0.36
0.38
0.38
0.38
3.1
3.2
3.2
3.1
3.06
2.5
2.1
1.9
0.16
0.14
0.8
NB
NB
32
–
[a] Ref. [13]. NB=no binding.
Synthesis of compounds from Generation III: towards
novel, easily available, and better alternatives to MNPG:
This intriguing result obtained with both subunits prompted
us to synthesize libraries of a panel of differently configured
and substituted polyhydroxyalkylfurans (Generation III)
and thio analogues of MNPG (Generation IV; see Support-
ing Information). The polyhydroxyalkylfuran skeleton is
easily obtained from sugars by condensation with b-keto-
AHCTUNGTRENNUNG
esters.[30] The synthesis of compounds of Generation III is
outlined in Schemes 3 and 4. From azido derivatives 23 and
28, different functionalities were attached at the end of the
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17995

A. J. Moreno-Vargas, I. Robina et al.
preparation of compounds 41 and 42 in moderate to good
yields. Chemoselective tosylation of the primary hydroxy
group followed by azido displacement afforded compounds
45 and 46 in good overall yields. Ester hydrolysis on 45d-arab
furnished 47 in good yield. Additionally, click reaction of
azido derivative 45d-arab with propargyl alcohol gave furan-
triazole 48 in good yield. Hydrogenation of compounds 45
or reduction with H2S of derivatives 46 afforded aminoesters
49 and 50. Chemoselective benzoylation of compounds 49
furnished derivatives 51. Similarly, chemoselective acetyla-
tion of 49d-arab gave 52.
Ligand evaluation by weak affinity chromatography of com-
pounds from Generation III: All these compounds were
evaluated by WAC for their affinity to rCTB, and the results
are summarized in Table 2. Several structure–activity rela-
tionships could be drawn. Thus, d-glucose-derived polyhy-
droxyalkylfuran derivatives bearing a COOEt moiety (alkyl
head substitution; compounds 23 and 24, Table 2, entries 2
and 3) present worse activity than MNPG. An acetamido or
a hydroxymethyltriazole moiety at the end of the side chain
(compounds 34 and 32, Table 2, entries 4 and 5) abolishes
the binding affinity. On the contrary, the presence of a benz-
amido moiety at the same position (compound 36, Table 2,
entry 6) slightly improves the activity, being in the same
order as MNPG. An aromatic moiety contributes to the
binding affinity with slighter efficacy than an alkyl moiety,
especially when it is linked to the furan ring (aryl head sub-
stitution). Thus, benzyl ester analogues of the above com-
pounds show better affinity properties: 23 (Kd =3.6)/28
(Kd =0.6), Table 2, entries 2/7; 24 (Kd =3.3)/31 (Kd =0.47),
Table 2, entries 3/8; 34 (NB)/35 (Kd =1.35), Table 2, en-
tries 4/11. However, two aromatic moieties at both ends in
this type of compound are not beneficial for binding (37,
Table 2, entry 12). An acetamido or a hydroxymethyltriazole
moiety, instead of an amino group, also diminishes the affin-
ity in this series: 31 (Kd =0.47)/33 (Kd =1.35)/35 (Kd =1.35),
Table 2, entries 8/10/11). The best compound in this series
bears a benzyloxymethyl group at the furan ring (40 (Kd =
0.2), Table 2, entry 13), being 5.5 times more active than
MNPG itself (see Supporting Information for a typical weak
affinity chromatogram of compound 33).
Shortening the side chain slightly diminishes the binding
affinity: 23 (Kd =3.6)/45d-arab (Kd =4.6), Table 2, entries 2/
14). For ethyl(3-amino-1,2-dihydroxypropyl)furoates (diaste-
reoisomeric compounds 49, Table 2, entries 16–19) the same
Kd was observed, thus indicating no influence of the stereo-
chemistry in the affinity. On changing to benzyl esters (aryl
head substitution, compounds 50, Table 2, entries 20–23) the
affinity improved considerably, but no influence of the ste-
reochemistry was observed.
On the contrary, for aryl-tail-substituted compounds (51,
Table 2, entries 24–27), appreciable slight differences in the
binding affinities could be observed among the four diaste-
reoisomers, which diminished the affinity in the order: 51d-
arab >51l-arab >51d-xyl >51l-xyl. It could be speculated that for
compounds 50, the aminodihydroxypropyl side chain has
Figure 7. STD NMR (a,c) and reference (b,d) spectra of the interactions
of 20 and 33, respectively, with the CTB. Each sample contained 1 mm
ligand in the presence of 5 mm CTB (25 mm in binding sites). The spectra
were obtained at 700 MHz and 278 K (saturation time, 2 s). Key signals
are labeled on the difference spectra (a and c).
side chain on ethyl and benzyl tetritol-1-yl furoates. Thus,
transformation of the 4’-amino or 4’-azido groups into acet-
amido, benzamido, or triazole moieties under standard con-
ditions gave rise to derivatives 32–37 (Scheme 3). Reaction
of d-glucose with ethyl benzyloxymethylcarbonyl acetate[31]
following Bartoliꢅs modification of the Garcꢀa Gonzꢁlez re-
action,[32] furnished ethyl tetritol-1-yl furoate 38 bearing
a benzyloxy appendage in 60% yield. Conventional tosyla-
tion and azido displacement gave azido derivative 40 in
moderate to good overall yield.
Other ethyl and benzyl polyhydroxyalkylfuroates (41–52)
with shorter side chains, presumably with least conforma-
tional freedom, were also prepared starting from l- and d-
arabinose and l- and d-xylose (Scheme 4). Compounds 42d-
and 42d-xyl have been obtained previously[33] in our group
arab
by reaction of the corresponding aldose with benzyl aceto-
ACHTUNGTRENNUNG
17996
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17989 – 18003

Non-Carbohydrate Ligands of Cholera Toxin
FULL PAPER
hydroxyalkylfuran 33 (see Fig-
ure 8b). This is not the case for
compounds 51 (“tail-substitut-
ed”, see Figure 3), in which the
interaction with the binding
pocket presumably takes place
through the benzoyl moiety.
The four diastereoisomers pres-
ent different retention factors,
thus showing differences in
binding affinities (see Support-
ing Information for a typical
weak affinity chromatogram of
compounds 51d-arab, 51l-arab, 51d-
xyl, and 51l-xyl).
Finally,
moiety at the furan ring abol-
ishes the affinity (47/45d-arab
a
carboxylic acid
,
Table 2, entries 15/14; 10a/11a,
Table 1, entries 10/11). As sus-
pected, amide 52 and triazole
derivative 32 did not show
binding affinity (Table 2, en-
tries 28 and 29).
Figure 8. Binding epitopes of a) 20 and b) 33 for their interactions with the CTB. The values represent relative
values of saturation after their normalization related to the most intense one (assigned 100%), obtained from
STD initial slopes (see Experimental Section).
a lot of conformational freedom, once the benzyl moiety
(“head-substituted”, see Figure 3) interacts with the CTB
binding pocket as was observed in the NMR study of poly-
Synthesis of thio analogues of MNPG Generation IV: The
synthesis of thio analogues of MNPG 5 (Generation IV) is
Scheme 4. Synthesis of compounds of Generation III (2). Reagents and Conditions: a) CeCl3·7H2O, NaI, silica gel, MeCN, 508C; b) TsCl, dry Py, ꢀ158C;
c) NaN3, DMF, 808C; d) H2, Pd-C, EtOH, RT for 45 or e) H2S(g), Py/H2O (1:1) 08C!RT for 46; f) Ac2O, MeOH, RT; g) BzCl, Et3N, dry DMF, 58C!
RT; h) propargyl alcohol, CuSO4 (aq), sodium ascorbate (aq), MeOH, RT; i) 1m NaOH, EtOH, 608C.
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17997

A. J. Moreno-Vargas, I. Robina et al.
these compounds was carried out by the WAC method as in-
dicated above. Compounds of Generation IV are S ana-
logues of the reported MNPG series. They all showed lower
affinity than MNPG itself, although it can be concluded that
the presence of a m-NO2 group is necessary for the affinity.
m-Nitrophenyl a-d-thiogalactoside 58a (Table 3, entry 10)
Table 3. Retention factor (k’) and dissociation constant (Kd) of CT li-
gands at 228C determined with WAC at pH 7 for Generation IV.
Entry
Compound
k’
Kd [mm]
KMNPG/Kd
1
2
3
4
5
6
7
8
MNPG
56b
57b
58b
62b
63b
64b
56a
57a
58a
62a
63a
64a
1.58
0.16
0.19
0.33
0.07
0.11
0.08
0.19
<0.2
0.63
0.07
0.06
0.28
1.1[a]
NB
NB
4.7
NB
NB
NB
NB
NB
2.5
1
–
–
0.21
–
–
–
–
–
0.44
–
9
10
11
12
13
NB
NB
5.5
–
0.20
[a] Ref. [13]. NB=no binding.
showed better affinity than its corresponding b-d-thiogalac-
toside 58b (Table 3, entry 4). Oxidation of the sulfur to sul-
fone in these compounds did not improve their binding af-
finity. Although aryl O-galactoside MNPG[12] is a slightly
better CTB binder than its S analogue, the introduction of
a sulfur group at the anomeric position provides better acid
and enzyme-mediated hydrolytic stability. In addition, the
higher water solubility of sulfur derivatives relative to their
oxygen counterparts is an important advantage for our com-
pounds with regard to tolerance by most biological sys-
tems.[20]
Scheme 5. Reagents and conditions: a) ArSH, TMSOTf, 1,2-dichloro-
ethane, 08C!RT, 86% for 53b; b) 1. HBr/AcOH 33%, 2. ArSH,
Bu4NHSO4, NaOH(aq)/CH2Cl2, 83% for 54b, 71% for 55b; c) 1. PCl5,
BF3ACHTUNGTRENNUNG(Et2O), CH2Cl2, 2. ArSNa, DMF, 508C, 82% for 53a, 66% for 54a,
73% for 55a; d) m-chloroperoxybenzoic acid, anhydrous CH2Cl2, 08C!
RT, 100% for 59b and 61b, 76% for 60b, 100% for 59a, 83% for 60a,
95% for 61a; e) NaOMe/MeOH, 100% for 56b, 74% for 57b, 84% for
58b, 100% for 56a, 57a, 58a, 94% for 62b, 77% for 63b, 88% for 64b,
98% for 62a and 63a, 82% for 64a.
outlined in Scheme 5. The preparation of the aryl b-thioga-
lactopyranosides was accomplished by following two meth-
odologies. Direct glycosylation of penta-O-acetyl-b-d-galac-
topyranose with thiophenol, in the presence of TMSOTf, af-
forded 53b in 86% yield. Para- and meta-NO2-aryl thioga-
lactopyranosides 54b and 55b were obtained in 83 and 71%
yield, respectively, by phase-transfer catalysis from aceto-
bromogalactose and the corresponding aryl thiols by using
a mixture of NaOHaq/CH2Cl2 and tetrabutylammonium hy-
drogenosulfate as catalyst. The a series were obtained by re-
action of penta-O-acetyl-b-d-galactopyranose with PCl5 in
the presence of BF3·Et2O to give highly reactive tetra-O-
acetyl-b-d-galactopyranosyl chloride,[34] which reacts with
the appropriate aryl thiols furnishing 53a, 54a, and 55a in
82, 66, and 73% yield, respectively.
Conclusion
The synthesis of a library of novel nonhydrolyzable com-
pounds that act as rCTB ligands has been developed. An
important point to note in this approach is that poly-
hydroxyACHTNUTRGNEaNUG lkylfuroate derivatives presented the highest affini-
ties, being the first examples of non-carbohydrate ligands of
CT. They constitute a new type of ligand apart from the
well-known MNPG derivatives and the structural mimics of
GM1 reported previously.[8–10] These compounds are better
ligands themselves than when attached to d-galactose.
Biological evaluation has been carried out by WAC.
NMR experiments in the presence of rCTB indicated that
the furan moiety and the spacer to thiogalactose are crucial
for the ligand–CT interaction. Tail-connected benzyl polyhy-
droxyalkylfuroates (Generation II) interact along the whole
molecule but less tightly in the middle of it, thus indicating
a bidentate ligand. However, with head-connected com-
pounds (Generation I), the affinity is dramatically reduced
probably because the furan moiety is placed under the thio-
Oxidation of the aryl b- and a-tetra-O-acetyl-d-thiogalac-
tosides gave the corresponding sulfones 59–61 (a and b) in
good yields. Deprotection under Zemplꢄn conditions of
both series of compounds gave 56–58 (a and b) and 62–64
(a and b) in excellent yields.
Ligand evaluation by weak affinity chromatography of com-
pounds from Generation IV: The biological evaluation of
17998
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17989 – 18003

Non-Carbohydrate Ligands of Cholera Toxin
FULL PAPER
performed on an Agilent-1100/Agilent-1200 HPLC system. The mobile
phase was 10 mm sodium phosphate, 0.15 mm sodium chloride, pH 7. The
flow rate was 0.1 mLminꢀ1 and the temperature 228C. The sample
volume was 5 mL and the sample concentration 35 mm. Detection was per-
formed at 220 nm. The retention factor (k’) and the affinity (Kd) of the
derivatives were calculated as described previously[13] by using MNPG as
reference compound.
galactose ring by means of stabilizing carbohydrate–p inter-
actions.[35]
For polyhydroxyalkylfuroates (Generation III), besides
the furan, the presence of another aromatic moiety is bene-
ficial for the binding affinity, that is, ethyl benzyloxymethyl
furoate (40) is 5.5 times more active than MNPG itself.
Shortening the side chain slightly diminishes the binding af-
finity. For this type of aryl-tail-substituted compound, with
less conformational freedom, a relationship between binding
affinity and stereogenic centers was observed.
The presence of S at the anomeric position of d-galactose
in the thio analogues of MNPG, although increasing the hy-
drolytic stability of the compounds, is not beneficial for the
binding affinity. Finally, the compounds presented herein
represent an important step towards low-cost potent ligands
for CT with applications in therapy and detection. The good
hydrolytic stability and their easy preparation from carbohy-
drates that are readily available on a large scale, by using
high-performance reactions, will provide new possibilities
for the discovery of low-molecular-weight blockers of rCTB.
Progress in this area, to find more efficient CTB antagonists,
is currently under way in our laboratory.
Compounds from Generation I
2-(Fluoren-9-yl-methoxycarbonylamino)ethyl
2,3,4,6-tetra-O-acetyl-1-
thio-b-d-galactopyranoside (6): Molecular sieve (4 ꢆ) was added to a so-
lution of N-Fmoc-aminoethanothiol (430 mg, 1.44 mmol) and penta-O-
acetyl-b-d-galactopyranose 5 (375 mg, 0.96 mmol) in dry dichlorome-
thane (10 mL), and the mixture was stirred for 15 min. Then, trimethyl-
silyl trifluoromethanesulfonate (265 mL, 1.44 mmol) was added, and the
solution was stirred at RT overnight. After this period, the mixture was
washed with a saturated aqueous solution of NaHCO3, water, and brine,
dried with Na2SO4, filtered, and concentrated. The resulting residue was
purified by column chromatography (AcOEt/petroleum ether, 1:2!1:1)
to give 6 (432 mg, 1.22 mmol, 72%) as a white solid. [a]D27 =ꢀ17 (c=
1
0.58 in CH2Cl2); H NMR (300 MHz, CDCl3, 258C, mixture of rotamers):
d=7.80–7.29 (m, 8H; H-Ar), 5.62 (br t, 1H; NHFmoc), 5.25 (d, J4,3
=
3 Hz, 1H; H-4), 5.12 (t, J2,1 =J2,3 =9.9 Hz, 1H; H-2), 4.94 (dd, 1H; H-3),
4.51 (d, 1H; H-1), 4.75 (dd, 2JHꢀH =10.8, 3JHꢀH =5.4 Hz, 1H; CHH of
Fmoc), 4.55 (dd, 3JHꢀH =5.1 Hz, 1H; CHH of Fmoc), 4.28–4.14 (m, 2H;
H-1, CH of Fmoc), 3.90 (dd, J6a,5 =8.1 Hz, 1H; H-6a), 3.80 (dd, J6b,6a
=
11.4, J6b,5 =4.2 Hz, 1H; H-6b), 3.69–3.53 (m, 1H; H-2a’), 3.28–3.10 (m,
2H; H-2b’, H-5), 3.80–3.67 (m, 2H; H-1a’, H-1b’), 2.13, 2.07, 2.01,
1.69 ppm (4s, 12H; Me of OAc); 13C NMR (75.5 MHz, CDCl3, 258C, mix-
ture of rotamers): d=170.1–169.5 (4 C=O of OAc), 156.6 (C=O of
Fmoc), 146.8, 141.3, 127.8, 127.2, 124.7, 120.0 (C-Ar), 85.1 (C-1), 74.1 (C-
5), 71.5 (C-3), 67.3 (C-4, C-2), 65.4 (CH2 of Fmoc), 62.4 (C-6), 47.3 (CH
of Fmoc), 42.3 (C-2’), 32.5 (C-1’), 20.8–20.6 ppm (4 Me of OAc); IR: n˜ =
1743 (C=O), 1216, 1037 cmꢀ1; MS (FAB): m/z (%): 652 (20) [M+Na+];
HRMS (FAB): m/z calcd for C31H35NO11NaS (M+Na)+: 652.1829; found:
652.1832.
Experimental Section
General procedures: Optical rotations were measured in a 1.0 cm or
1.0 dm tube with a Jasco P-2000 spectropolarimeter. 1H and 13C NMR
spectra were obtained for solutions in CD3OD, D2O, CDCl3, and
[D6]DMSO. All the assignments were confirmed by two-dimensional
NMR experiments. The fast atom bombardment (FAB)–liquid secondary
ion mass spectra were obtained by using glycerol or 3-nitrobenzyl alcohol
as the matrix. TLC was performed on silica gel HF254 (Merck), with de-
tection by UV light and charring with H2SO4, ninhidrin, or Pancaldi re-
(2S)-2-[2-(tert-Butoxycarbonylamino)-2-methoxycarbonyl]ethyl 2,3,4,6-
tetra-O-acetyl-1-thio-b-d-galactopyranoside (7):
A solution of HBr/
AcOH (33%, 3.5 mL) was added to a solution of penta-O-acetyl-b-d-gal-
actopyranose 5 (1.20 g, 3.08 mmol) in dry dichloromethane (20 mL)
cooled to 08C. The reaction mixture was stirred for 1 h at 08C, and then
at RT for 3.5 h. Then, the mixture was washed sequentially with cooled
aqueous solutions of 1% NaHCO3, saturated NaHCO3, and brine. The
organic layer was dried with Na2SO4, filtered, and concentrated. The re-
sulting crude was dissolved in AcOEt (25 mL) and added to a solution of
N-Boc-l-Cys-OMe (723 mg, 3.08 mmol) in a 10% aqueous solution of
Na2CO3 (25 mL). Then, TBAHS (4.20 g, 12.32 mmol) was added and the
reaction mixture was stirred vigorously at RT overnight. After this
period, the reaction mixture was diluted with AcOEt and washed with
a saturated aqueous solution of NaHCO3 and brine. The organic layer
was dried with Na2SO4, filtered, and concentrated. The resulting crude
was purified by column chromatography (AcOEt/petroleum ether, 1:2)
to give 7 (986 mg, 1.74 mmol, 57%) as a white solid. [a]D27 =ꢀ26 (c=
1.01 in CH2Cl2); 1H NMR (300 MHz, CDCl3, 258C): d=5.63 (br d,
agent [(NH4)6MoO4, CeACHTNUGTREN(NUG SO4)2, H2SO4, H2O]. Silica gel 60 (Merck,
230 mesh) was used for preparative chromatography.
STD NMR and transferred NOESY experiments: The NMR binding ex-
periments were carried out at 278 K on a Bruker Avance III 700 MHz
spectrometer equipped with a four-channel TCI cryoprobe. For STD
NMR experiments, a pulse sequence that included two trim pulses (2.5
and 5 ms) and 3 ms spoil gradient before the first 908 pulse was used, to
remove unwanted magnetization in the x,y plane between consecutive
scans. Water suppression was carried out by excitation sculpting, and the
broad signals from the protein were removed by using a 20 ms spin-lock
(T11) filter at a field strength of 6.9 kHz. Selective saturation was
achieved by using a train of 50 ms Gaussian pulses at 0 ppm (on-reso-
nance experiment), whereas 40 ppm was the frequency of low-power irra-
diation for the control (off-resonance) experiment. NOESY experiments
were carried out at a mixing time of 300 ms. Samples contained 5 mm pen-
tameric subunit B of cholera toxin (rCTB) and 1 mm ligand, on a D2O
pH 7 buffer solution of 20 mm phosphate and 150 mm NaCl, which led to
a ligand-to-receptor ratio of 40:1 (five binding sites per protein subunit).
To obtain the binding epitopes of the ligands, STD NMR experiments
were carried out at different saturation times (0.3, 0.5, 0.75, 1.0, 1.5, 2.0,
3.0, 4.0, 5.0, and 6.0 s) and the resulting building curves were fitted math-
ematically to a monoexponential equation, from which the initial slopes
were obtained. From these values, the binding epitope was obtained by
dividing all by the largest value, to which an arbitrary value of 100% was
assigned.
J
NH,2’ =7.5 Hz, 1H; N-H), 5.43 (dd, J4,3 =3.3, J4,5 =0.9 Hz, 1H; H-4), 5.22
(t, J2,1 =J2,3 =9.9 Hz, 1H; H-2), 5.04 (dd, 1H; H-3), 4.52 (m, 1H; H-2’),
4.50 (d, 1H; H-1), 4.16 (d, J6,5 =6.6 Hz, 2H; H-6), 3.95 (td, 1H; H-5),
3.75 (s, 3H; COOCH3), 3.22 (dd, 2J1a’,1b’ =14.4, J1a’,2’ =4.2 Hz, 1H; H-1a’),
3.03 (dd, J1b’,2’ =6.3 Hz, 1H; H-1b’), 2.16, 2.08, 2.06, 1.98 (4s, 12H; Me of
OAc), 1.45 ppm (s, 9H; C
d=171.4 (COOMe), 170.5, 170.3, 170.1, 169.8 (4C, C=O of OAc), 155.6
(C=O of Boc), 83.7 (C(CH3)3), 80.3 (C-1), 75.0 (C-5), 71.8 (C-3), 67.2,
66.8 (C-4, C-2), 61.4 (C-6), 53.5 (C-2’), 52.7 (COOCH3), 32.1 (C-1’), 28.5
(CH3)3); 13C NMR (75.4 MHz, CDCl3, 258C):
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
(CACHTNUGRTNEUNG(CH3)3), 20.9, 20.8, 20.7 ppm (4C, Me of OAc); IR: n˜ =1754, 1715 (C=
O), 1216, 1032 cmꢀ1; MS (FAB): m/z (%): 588 (100) [M+Na+]; HRMS
(FAB): m/z calcd for C23H35NO13SNa (M+Na)+: 588.1727; found:
588.1737.
Weak affinity chromatography: Recombinant CTB (Crucell Company,
Sweden) was immobilized into Nucleosil silica (10 mm, 300 ꢆ) and
packed into a 50ꢇ2.1 mm column. The number of active groups of CTB
was estimated to be 261 nmol. All chromatographic experiments were
5-(1-Acetamido-2,3,4-tetra-O-acetyl-1-deoxy-d-ribotetritol-1-yl)-2-
methyl-3-furoic acid (10): A solution of 9[21] (318 mg, 0.88 mmol) in anhy-
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17999

A. J. Moreno-Vargas, I. Robina et al.
drous MeOH (10 mL) was hydrogenated under atmospheric pressure for
3 h using Pd-C (10%) as catalyst. Then, the solution was filtered through
Celite and the catalyst was washed with MeOH. The filtered solution was
concentrated, and conventionally acetylated. Column chromatography
(CH2Cl2/MeOH, 25:1) gave pure 10 (287 mg, 0.70 mmol, 80%) as a white
solid. [a]D27 =ꢀ66 (c=0.62 in CH2Cl2); 1H NMR (300 MHz, CDCl3,
258C): d=6.50 (s, 1H; H-4), 6.49 (br d, JNH,1’ =9.3 Hz, 1H; NHAc), 5.43
except that acid derivative 11[22] was used as starting material. Purifica-
tion on silica gel (ether/acetone, 15:1) afforded 13 in 42% yield as
a
yellow oil. [a]D26 =ꢀ16 (c=0.90 in CH2Cl2); 1H NMR (300 MHz,
CDCl3, 258C): d=6.53 (s, 1H; H-4’’’), 6.23 (br t, 1H; N-H), 6.01 (d,
J
J
1’’,2’’ =5.1 Hz, 1H; H-1’’), 5.61 (dd, J2’’,3’’ =7.2 Hz, 1H; H-2’’), 5.44 (dd,
4,3 =3.3, J4,5 =0.9 Hz, 1H; H-4), 5.26 (t, J2,1 =J2,3 =9.9 Hz, 1H; H-2), 5.16
(ddd, J3’’,4b’’ =8.7, J3’’,4b’’ =3.3 Hz, 1H; H-3’’), 5.06 (dd, 1H; H-3), 4.51 (d,
1H; H-1), 4.25 (dd, 2J4a’’,4b’’ =12.6 Hz, 1H; H-4a’’), 4.18–4.06 (m, 3H; H-
4b’’, H-6a, H-6b), 3.96 (br td, J5,6 =5.8 Hz, 1H; H-5), 3.71 (m, 1H; H-2a’),
3.52 (m, 1H; H-2b’), 3.01 (m, 1H; H-1a’), 2.84 (m, 1H; H-1b’), 2.57 (s,
3H; Me), 2.17, 2.10, 2.08, 2.07, 2.05, 2.04, 2.03, 1.99 ppm (8s, 24H; Me of
OAc); 13C NMR (75.4 MHz, CDCl3, 258C): d=170.1–169.5 (8C, C=O de
OAc), 163.4 (CONH), 157.8 (C-2’’’), 146.7 (C-5’’’), 116.5 (C-3’’’), 109.0
(C-4’’’), 84.1 (C-1), 75.0 (C-5), 71.9 (C-3), 70.1 (C-2’’), 68.9 (C-3’’), 67.5
(C-4), 67.3 C-2), 66.0 (C-1’’), 61.7 (C-4’’, C-6), 38.9 (C-2’), 30.7 (C-1’),
20.9–20.7 (8C, Me of OAc), 13.7 ppm (Me); IR: n˜ =1739 (C=O), 1206,
1041 cmꢀ1; MS (FAB): m/z (%): 826 (35) [M+Na+]; HRMS (FAB): m/z
calcd for C34H45NO19NaS (M+Na)+: 826.2206; found: 826.2204.
(dd, J1’,2’ =5.1 Hz, 1H; H-1’), 3.34 (dd, J2’,3’ =6.0 Hz, 1H; H-2’), 3.17 (td,
2
J
3’,4b’ =6.0, J3’,4a’ =2.7 Hz, 1H; H-3’), 4.36 (dd, J4a’,4b’ =12.6 Hz, 1H; H-4a’),
4.20 (dd, 1H; H-4b’), 2.56 (s, 3H; Me), 2.07, 2.06, 2.05, 2.04 ppm (4s,
12H; Me of Ac); 13C NMR (75.4 MHz, CDCl3, 258C): d=170.9, 170.6,
170.3, 169.8 (C=O of Ac), 168.1 (COOH), 160.8 (C-2), 148.5 (C-5), 113.7
(C-3), 108.9 (C-4), 72.1 (C-2’), 69.8 (C-3’), 62.1 (C-4’), 47.5 (C-1’), 23.4,
21.0, 20.8 (Me of Ac), 14.0 ppm (Me); IR: n˜ =1743 (C=O), 1211,
1047 cmꢀ1; MS (CI): m/z (%): 414 (30) [M+H+]; HRMS (CI): m/z calcd
for C18H24NO10 (M+H)+: 414.1400; found: 414.1385.
5-(1-Acetamido-1-deoxy-d-ribotetritol-1-yl)-2-methyl-3-furoic acid (10a):
NaOMe in methanol (1m, 0.1 equiv per O-acetyl group) was added to
a solution of compound 10 (50 mg, 0.12 mmol) in anhydrous MeOH
(1 mL), and the reaction mixture was stirred for 1 h at RT. Then the mix-
ture was neutralized with Amberlite IR-120H+, filtered, and washed with
MeOH. The filtered solution was concentrated to give pure 10a (quant.)
as a white solid. [a]D27 = +48 (c=0.73 in MeOH); 1H NMR (300 MHz,
CD3OD, 258C): d=6.58 (s, 1H; H-4), 5.35 (d, J1’,2’ =3.9 Hz, 1H; H-1’),
3.81–3.69 (m, 2H; H-2’, H-4a’), 3.59 (dd, 2J4b’,4a’ =11.4, J4b’,3’ =5.9 Hz, 1H;
H-4b’), 3.43 (m, 1H; H-3’), 2.54 (s, 3H; Me), 1.99 ppm (s, 3H; Me of
NHAc); 13C NMR (75.4 MHz, CD3OD, 258C): d=172.4 (2C, C=O of
NHAc, COOH), 159.4 (C-2), 151.3 (C-5), 114.6 (C-3), 110.4 (C-4), 74.4
(C-2’), 73.3 (C-3’), 64.7 (C-4’), 50.4 (C-1’), 22.6 (Me of NHAc), 13.8 ppm
(2S)-2-{5-[(1-Acetamido-2,3,4-tri-O-acetyl-1-deoxy-d-ribotetritol-1-yl)-2-
methyl-3-furancarboxamido]-2-methoxycarbonyl}ethyl
acetyl-1-thio-b-d-galactopyranoside (14): Compound
2,3,4,6-tetra-O-
(100 mg,
7
0.18 mmol) was dissolved in TFA/CH2Cl2 (20%, 3 mL) and the mixture
was stirred at RT for 2 h. The resulting amine was dissolved in DMF
(4 mL) and acid derivative 10 (81 mg, 0.20 mmol), DIPEA (146 mL,
0.86 mmol), and PyBOP (105 mg, 0.20 mmol) were added sequentially.
The reaction mixture was stirred vigorously at RT overnight. Then the
solvent was removed, and the residue was diluted with AcOEt and
washed with 1m HCl, a saturated aqueous solution of NaHCO3, and
water. The organic layer was dried (Na2SO4), filtered, and concentrated.
Chromatographic purification on silica gel (ether/acetone, 6:1) afforded
the protected ligand 14 (134 mg, 0.16 mmol, 89%) as a white solid.
(Me); IR: n˜ =3306 (COOH, OH), 2927, 1649 (C=O), 1226, 1068 cmꢀ1
;
MS (FAB): m/z (%): 310 (27) [M+Na+]; HRMS (FAB): m/z calcd for
C12H17NO7Na (M+Na)+: 310.0903; found: 310.0901.
[a]D26 = +24 (c=1.01 in CH2Cl2); H NMR (300 MHz, CDCl3, 258C): d=
1
6.97 (br d, JNH,1’’ =9.0 Hz, 1H; NHAc), 6.80 (br d, JNH,2’ =7.2 Hz, 1H; N-
H), 6.41 (s, 1H; H-4’’’), 5.44 (dd, J1’’,2’’ =6.0 Hz, 1H; H-1’’), 5.40–5.36 (m,
2H; H-4, H-2’’), 5.21 (t, J2,1 =10.0 Hz, 1H; H-2), 5.17 (m, 1H; H-3’’), 5.02
(dd, J3,2 =10.0, J3,4 =3.3 Hz, 1H; H-3), 4.95 (m, 1H; H-2’), 4.55 (d, 1H;
2-[5-(1-Acetamido-2,3,4-tri-O-acetyl-1-deoxy-d-ribotetritol-1-yl)-2-
methyl-3-furancarboxamido]ethyl 2,3,4,6-tetra-O-acetyl-1-thio-b-d-galac-
topyranoside (12): Compound 6 (200 mg, 0.32 mmol) was dissolved in pi-
peridine/DMF (20%; 8.4 mL) and the mixture was stirred at RT for
15 min. Then, the solvent was removed and the residue was purified by
column chromatography (CH2Cl2/MeOH, 8:1) affording the unprotected
amine (83 mg, 0.20 mmol, 63%) as a yellow oil. The resulting amine
(72 mg, 0.18 mmol) was dissolved in DMF (4 mL), and acid derivative 10
(80 mg, 0.20 mmol), DIPEA (132 mL, 0.78 mmol), and PyBOP (105 mg,
0.20 mmol) were added sequentially. The reaction mixture was stirred
vigorously at RT overnight. Then the solvent was removed, and the resi-
due was diluted with AcOEt and washed with 1m HCl, a saturated aque-
ous solution of NaHCO3, and water. The organic phase was dried
(Na2SO4), filtered, and concentrated. Chromatographic purification on
silica gel (ether/acetone, 7:1!5:1) afforded the protected ligand 12
(76 mg, 0.10 mmol, 56%) as a yellow solid. [a]D27 = +29 (c=0.80 in
H-1), 4.34 (dd, 2J4a’’,4b’’ =12.4, J4a’’,3’’ =2.7 Hz, 1H; H-4a’’), 4.22 (dd, J4b’’,3’’
=
6.6 Hz, 1H; H-4b’’), 4.03 (dd, J6a,5 =7.8, J6b,5 =1.8 Hz, 2H; H-6a, H-6b),
3.88 (m, 1H; H-5), 3.79 (s, 3H; COOCH3), 3.34 (dd, 2J1a’,1b’ =14.4, J1a’,2’
=
4.5 Hz, 1H; H-1a’), 3.12 (dd, J1b’,2’ =6.0 Hz, 1H; H-1b’), 2.56 (s, 3H; Me),
2.15, 2.11, 2.06, 2.05, 2.04, 2.04, 2.03, 2.00 ppm (8s, 24H; Me of Ac);
13C NMR (75.4 MHz, CDCl3, 258C): d=171.0, 170.9, 170.7, 170.6, 170.4,
170.3, 170.2, 170.0, 169.8 (8C, C=O of ester), 163.2 (CONH), 157.6 (C-
2’’’), 149.1 (C-5’’’), 115.8 (C-3’’’), 106.2 (C-4’’’), 84.2 (C-1), 74.9 (C-5), 71.9,
71.8 (C-3, C-2’’), 70.1 (C-3’’), 67.4, 67.3 (C-4, C-2), 62.0 (C-4’’), 61.5 (C-6),
53.0 (COOCH3), 52.2 (C-2’), 47.2 (C-1’’), 31.8 (C-1’), 23.3 (Me of NHAc),
21.0–20.7 (7C, Me of Ac), 13.8 ppm (Me); IR: n˜ =1739 (C=O), 1211,
1046 cmꢀ1; MS (FAB): m/z (%): 883 (100) [M+Na+]; HRMS (FAB): m/z
calcd for C36H48N2O20NaS (M+Na)+: 883.2419; found: 883.2399.
CH2Cl2); 1H NMR (500 MHz, CDCl3, 258C): d=6.56 (br d, JNH,1’’
8.9 Hz, 1H; NHAc), 6.34 (s, 1H; H-4’’’), 6.29 (br t, 1H; N-H), 5.44 (dd,
=
(2S)-2-{5-[(1,2,3,4-Tetra-O-acetyl-d-arabinotetritol-1-yl)-2-methyl-3-fur-
ancarboxamido]-2-methoxycarbonyl}ethyl 2,3,4,6-tetra-O-acetyl-1-thio-b-
d-galactopyranoside (15): This compound was prepared following the
procedure described for 14 except that acid derivative 11 was used as
J
J
J
4,3 =3.0 Hz, 1H; H-4), 5.40 (dd, J1’’,2’’ =5.2 Hz, 1H; H-1’’), 5.34 (br t,
2’’,3’’ =5.8 Hz, 1H; H-2’’), 5.30 (t, J2,1 =J2,3 =9.9 Hz, 1H; H-2), 5.19 (ddd,
3’’,4b’’ =6.1, J3’’,4b’’ =2.6 Hz, 1H; H-3’’), 5.06 (dd, 1H; H-3), 4.52 (d, 1H; H-
starting material. Purification by column chromatography on silica gel
1), 4.36 (dd, 2J4a’’,4b’’ =12.1 Hz, 1H; H-4a’’), 4.20 (dd, 1H; H-4b’’), 4.14
(dd, 2J6a,6b =11.2, J6a,5 =6.9 Hz, 1H; H-6a) 4.05 (dd, J6b,5 =6.1 Hz, 1H; H-
6b), 3.96 (br td, 1H; H-5), 3.68 (m, 1H; H-2a’), 3.53 (m, 1H; H-2b’), 3.03
(m, 1H; H-1a’), 2.84 (m, 1H; H-1b’), 2.55 (s, 3H; Me), 2.19–1.96 ppm (8s,
24H; Me of Ac); 13C NMR (125.7 MHz, CDCl3, 258C): d=170.4–169.5
(8C, C=O of OAc), 163.4 (CONH), 156.8 (C-2’’’), 148.1 (C-5’’’), 116.2 (C-
3’’’), 106.5 (C-4’’’), 84.0 (C-1), 74.8 (C-5), 72.0 (C-2’’), 71.7 (C-3), 69.8 (C-
3’’), 67.3 (C-4), 67.1 (C-2), 61.9 (C-4’’), 61.7 (C-6), 47.3 (C-1’’), 38.9 (C-2’),
30.5 (C-1’), 23.2 (Me of NHAc), 20.9–20.5 (8C, Me of OAc), 13.7 ppm
(Me); IR: n˜ =1739 (C=O), 1211, 1046 cmꢀ1; MS (FAB): m/z (%): 825
(100) [M+Na+]; HRMS (FAB): m/z calcd for C34H46N2O18NaS (M+Na)+
: 825.2364; found: 825.2377.
26
(ether/acetone, 15:1) afforded 15 in 89% yield as a white solid. [a]D
=
ꢀ33 (c=1.04 in CH2Cl2); 1H NMR (300 MHz, CDCl3, 258C): d=6.61 (s,
1H; H-4’’’), 6.54 (br d, JNH,2’ =7.2 Hz, 1H; N-H), 6.01 (d, J1’’,2’’ =5.0 Hz,
1H; H-1’’), 5.59 (dd, J2’’,3’’ =7.2 Hz, 1H; H-2’’), 5.41 (dd, J4,3 =3.0, J4,5
=
0.6 Hz, 1H; H-4), 5.21 (t, J2,1 =J2,3 =9.9 Hz, 1H; H-2), 5.17 (m, 1H; H-
3’’), 5.05 (dd, 1H; H-3), 4.91 (td, 1H; H-2’), 4.52 (d, 1H; H-1), 4.25 (dd,
2J4a’’,4b’’ =12.3, J4a’’,3’’ =3 Hz, 1H; H-4a’’), 4.12 (dd, J4b’’,3’’ =5.4 Hz, 1H; H-
4b’’), 4.06 (dd, J6a,5 =7.2, J6b,5 =0.9 Hz, 2H; H-6a, H-6b), 3.91 (m, 1H; H-
5), 3.79 (s, 3H; COOCH3), 3.33 (dd, 2J1a’,1b’ =13.8, J1a’,2’ =4.8 Hz, 1H; H-
1a’), 3.08 (dd, J1b’,2’ =6.6 Hz, 1H; H-1b’), 2.58 (s, 3H; Me), 2.16, 2.11,
2.07, 2.06, 2.05, 2.05, 2.02, 1.98 ppm (8s, 24H; Me of OAc); 13C NMR
(75.4 MHz, CDCl3, 258C): d=170.1–169.5 (8C, C=O of ester), 163.0
(CONH), 158.4 (C-2’’’), 146.9 (C-5’’’), 115.9 (C-3’’’), 109.0 (C-4’’’), 83.7
(C-1), 74.8 (C-5), 71.7 (C-3), 70.1 (C-2’’), 68.8 (C-3’’), 67.3, 67.2 (C-4, C-
2), 66.0 (C-1’’), 1.8 (C-4’’), 61.4 (C-6), 53.0 (COOCH3), 51.8 (C-2’), 31.6
2-[5-(1,2,3,4-Tetra-O-acetyl-d-arabinotetritol-1-yl)-2-methyl-3-furancar-
boxamido]ethyl 2,3,4,6-tetra-O-acetyl-1-thio-b-d-galactopyranoside (13):
This compound was prepared following the procedure described for 12
18000
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17989 – 18003

Non-Carbohydrate Ligands of Cholera Toxin
FULL PAPER
(C-1’), 20.9–20.7 (8C, Me of OAc), 13.8 ppm (Me); IR: n˜ =1739 (C=O),
1211, 1046 cmꢀ1; MS (FAB): m/z (%): 884 (100) [M+Na+]; HRMS
(FAB): m/z calcd for C36H47NO21NaS (M+Na)+: 884.2259; found:
884.2268.
6.9 Hz, 1H; H-2’’), 3.79 (s, 3H; COOCH3), 3.78–3.58 (m, 7H; H-4a’’, H-
4b’’, H-3’’, H-6a, H-6b, H-5, H-3), 3.56 (t, J2,3 =9.3 Hz, 1H; H-2), 3.44
(dd, J1a’,1b’ =14.4, J1a’,2’ =5.1 Hz, 1H; H-1a’), 3.12 (dd, J1b’,2’ =9 Hz, 1H; H-
2
1b’), 2.50 ppm (s, 3H; Me); 13C NMR (75.4 MHz, D2O, 25 8C): d=173.3
(COOCH3), 167.3 (CONH), 158.1 (C-2’’’), 152.9 (C-5’’’), 115.7 (C-3’’’),
107.1 (C-4’’’), 86.1 (C-1), 79.7 (C-5), 74.5 (C-3), 73.3 (C-2’’), 71.8 (C-3’’),
70.3 (C-2), 69.3 (C-4), 67.2 (C-1’’), 63.2 (C-6), 61.6 (C-4’’), 53.8
(COOCH3), 53.5 (C-2’), 31.2 (C-1’), 13.5 ppm (Me); IR: n˜ =3284 (OH),
1735 (C=O), 1636, 1223, 1037 cmꢀ1; MS (FAB): m/z (%): 548 (15)
[M+Na+]; HRMS (FAB): m/z calcd for C20H31NO13NaS (M+Na)+:
548.1414; found: 548.1400.
General deacetylation procedures for the synthesis of 16–19: Sodium
methoxide in methanol (0.1m, 10% molar) was added to a solution of
acetylated ligand in anhydrous MeOH (ꢁ0.05m), and the reaction mix-
ture was stirred for 1–6 h at RT. Then, the mixture was neutralized with
Amberlite IR-120H+, filtered, and concentrated.
2-[5-(1-Acetamido-1-deoxy-d-ribotetritol-1-yl)-2-methyl-3-furancarbox-
ACHTUNGTRENNUNGamido]ethyl 1-thio-b-d-galactopyranoside (16): Deprotection of 12
(61 mg, 0.08 mmol), by following the general procedure described above,
afforded compound 16 (quant.) as a white solid. [a]D26 =+25 (c=0.44 in
MeOH); 1H NMR (300 MHz, D2O, 25 8C): d=6.55 (s, 1H; H-4’’’), 5.21
2-(Acetamido)ethyl 1-thio-b-d-galactopyranoside (20): Compound
6
(200 mg, 0.32 mmol) was dissolved in piperidine/DMF (20%, 8.4 mL),
and the mixture was stirred at RT overnight. Evaporation of the solvent
followed by conventional acetylation and purification by column chroma-
tography (CH2Cl2/MeOH, 8:1) afforded the deprotected amine (83 mg,
0.20 mmol, 63%) as a yellow oil. Deacetylation of the resulting amine
(108 mg, 0.13 mmol) by following the general procedure described above
(d, J1’’,2’’ =4.4 Hz, 1H; H-1’’), 4.49 (d, J1,2 =9.4 Hz, 1H; H-1), 3.95 (d, J4,3
4.1 Hz, 1H; H-4), 3.85 (dd, J2’’,3’’ =8.1 Hz, 1H; H-2’’), 3.73 (dd, J4a’’,4b’’
=
2
=
11.9, J4a’’,3’’ =2.7 Hz, 1H; H-4a’’), 3.70–3.46 (m, 9H; H-4b’’, H-6a, H-6b,
H-2, H-5, H-3, H-2a’, H-2b’, H-3’’), 3.05–2.80 (m, 2H; H-1a’, H-1b’), 2.47
(s, 3H; Me), 2.02 ppm (s, 3H; Me of NHAc); 13C NMR (75.4 MHz, D2O,
258C): d=173.3 (C=O of NHAc), 166.6 (CONH), 156.8 (C-2’’’), 148.8
(C-5’’’), 115.6 (C-3’’’), 107.5 (C-4’’’), 86.0 (C-1), 78.9 (C-5), 73.9 (C-3), 72.1
(C-2’’), 71.5 (C-3’’), 69.7 (C-2), 68.7 (C-4), 62.5 (C-4’’), 61.0 (C-6), 48.9
(C-1’’), 39.5 (C-2’), 29.5 (C-1’), 21.8 (Me of NHAc), 12.8 ppm (Me); IR:
n˜ =3313 (OH), 1626 (C=O), 1036 cmꢀ1; MS (FAB): m/z (%): 531 (25)
[M+Na+]; HRMS (FAB): m/z calcd for C20H32N2O11NaS (M+Na)+:
531.1625; found: 531.1616.
and purification of the residue by column chromatography (CH2Cl2/
26
MeOH/NH4OH, 2:1:0.35) afforded 20 (quant.) as a yellow solid. [a]D
=
ꢀ17 (c=0.50 in MeOH); 1H NMR (300 MHz, CD3OD, 258C): d=4.34
(d, J1,2 =9.4 Hz, 1H; H-1), 3.89 (dd, J4,3 =3.2, J4,5 =0.8 Hz, 1H; H-4), 3.77
(dd, 2J6a,6b =11.5, J6a,5 =7.1 Hz, 1H; H-6a), 3.69 (dd, J6b,5 =5.0 Hz, 1H; H-
6b), 3.56 (t, 1H; H-2), 3.56 (m, 1H; H-5), 3.51–3.39 (m, 3H; H-3, H-2a’,
H-2b’), 2.88 (m, 1H; H-1a’), 2.76 (m, 1H; H-1b’), 1.95 ppm (s, 3H; Me);
13C NMR (75.5 MHz, CD3OD, 258C): d=173.4 (C=O of NHAc), 87.8 (C-
1), 80.8 (C-5), 76.2 (C-3), 71.4 (C-2), 70.6 (C-2), 62.8 (C-6), 41.2 (C-2’),
30.7 (C-1’), 22.6 ppm (Me); IR: n˜ =3320 (OH), 1630 (C=O), 1225,
1004 cmꢀ1; MS (CI): m/z (%): 282 (4) [M+H+]; HRMS (CI): m/z calcd
for C10H20NO6S (M+Na)+: 282.1011; found: 282.1012.
2-[5-(d-Arabinotetritol-1-yl)-2-methyl-3-furancarboxamido]ethyl 1-thio-
b-d-galactopyranoside (17): Deprotection of 13 (56 mg, 0.07 mmol), by
following the general procedure described above, afforded compound 17
(quant.) as
a
white solid. [a]D26 =ꢀ12 (c=0.80 in H2O); 1H NMR
(2S)-2-(2-Amino-2-methoxycarbonyl)ethyl 1-thio-b-d-galactopyranoside
(21): Deacetylation of compound 7 (100 mg, 0.17 mmol), by following the
general procedure described above, and purification by column chroma-
tography (CH2Cl2/MeOH, 10:1) afforded the corresponding O-deprotect-
ed derivative (65 mg, 0.16 mmol, 94%). This compound was dissolved in
TFA/CH2Cl2 (20%, 4 mL) and the mixture was stirred at RT for 1.5 h.
The reaction mixture was concentrated and purified by column chroma-
tography (CH2Cl2/MeOH, 4:1) to give pure 21 (quant.) as its trifluoroace-
tate salt (yellow oil). [a]D26 =ꢀ15 (c=0.63 in MeOH); 1H NMR
(300 MHz, CD3OD, 258C): d=4.32 (d, J1,2 =9.4 Hz, 1H; H-1), 3.89 (m,
1H; H-4), 3.82–3.72 (m, 2H; H-6a, H-2’), 3.75 (s, 3H; COOMe), 3.69
(dd, 2J6b,6a =11.4, J6b,5 =4.9 Hz, 1H; H-6b), 3.62–3.52 (m, 2H; H-2, H-5),
(300 MHz, D2O, 25 8C): d=6.60 (s, 1H; H-4’’’), 4.89 (d, J1’’,2’’ =4.0 Hz, 1H;
H-1’’), 4.52 (d, J1,2 =9.3 Hz, 1H; H-1), 3.98 (d, J4,3 =3.3 Hz, 1H; H-4),
3.90 (dd, J2’’,3’’ =7.2 Hz, 1H; H-2’’), 3.79 (dd, 2J4a’’,4b’’ =11.4, J4a’’,3’’ =3.0 Hz,
1H; H-4a’’), 3.77–3.55 (m, 8H; H-4b’’, H-6a, H-6b, H-5, H-3, H-2a’, H-
2b’, H-3’’), 3.57 (t, 1H; H-2), 3.09–2.87 (m, 2H; H-1a’, H-1b’), 2.50 ppm
(s, 3H; Me); 13C NMR (75.4 MHz, D2O, 25 8C): d=167.4 (CONH), 157.3
(C-2’’’), 152.8 (C-5’’’), 116.3 (C-3’’’), 107.2 (C-4’’’), 86.7 (C-1), 79.6 (C-5),
74.6 (C-3), 73.3 (C-2’’), 71.8 (C-3’’), 70.3 (C-2), 69.4 (C-4), 67.2 (C-1’’),
63.2 (C-4’’), 61.7 (C-6), 40.1 (C-2’), 30.2 (C-1’), 12.4 ppm (Me); IR: n˜ =
3243 (OH), 1081, 1051, 838 cmꢀ1; MS (FAB): m/z (%): 490 (5) [M+Na+];
HRMS (FAB): m/z calcd for C18H29NO11NaS (M+Na)+: 490.1359; found:
490.1340.
3.48 (dd, J3,2 =9.4, J3,4 =3.3 Hz, 1H; H-3), 3.22 (dd, 2J1a’,1b’ =14.2, J1a’,2’
4.5 Hz, 1H; H-1a’), 2.88 (dd,
1b’,2’ =7.6 Hz, 1H; H-1b’); 13C NMR
=
(2S)-2-{5-[(1-Acetamido-1-deoxy-d-ribotetritol-1-yl)-2-methyl-3-furancar-
boxamido]-2-methoxycarbonyl}ethyl 1-thio-b-d-galactopyranoside (18):
Deprotection of 14 (111 mg, 0.13 mmol), by following the general proce-
dure described above, afforded compound 18 (quant.) as a white solid.
[a]D31 =ꢀ2 (c=1.38 in H2O); 1H NMR (300 MHz, CD3OD, 258C): d=
6.74 (s, 1H; H-4’’’), 5.36 (d, J1’’,2’’ =3.9 Hz, 1H; H-1’’), 4.84 (dd under
J
(125.7 MHz, CD3OD, 258C, acid–base equilibrium): d=175.0 (COOMe),
3
1
163.2 (c, JC,F =34.8 Hz, CF3COOꢀ), 116.8 (c, JC,F =292.7 Hz, CF3COOꢀ),
92.2 (C-2’), 88.5, 87.6 (C-1), 81.2, 80.8 (C-5), 76.3, 76.1 (C-3), 71.4, 71.0,
70.7, 70.6, 70.5, 70.3 (C-2, C-4, Me of COOMe), 62.8, 62.7 (C-6), 36.0,
35.0 ppm (C-1’); IR: n˜ =3368 (OH, NH), 1673 (C=O), 1196, 1132 cmꢀ1
MS (ESI): m/z calcd for C12H19F3NO9S (M+): 411.1; found: 412.1.
;
CD3OD, 1H; H-2’), 4.37 (d, J1,2 =9.6 Hz, 1H; H-1), 3.88 (br d, J4,3
=
2.7 Hz, 1H; H-4), 3.80 (dd, 2’’,3’’ =8.4 Hz, 1H; H-2’’), 3.75 (s, 3H;
J
Compounds from Generation II
COOCH3), 3.76–3.33 (m, 9H; H-4a’’, H-4b’’, H-3’’, H-1a’, H-6a, H-6b, H-
5, H-3, H-2), 3.08 (dd, 2J1b’,1a’ =14.1, J1b’,2’ =7.8 Hz, 1H; H-1b’), 2.53 (s,
3H; Me), 2.00 ppm (s, 3H; Me of NHAc); 13C NMR (75.4 MHz, CD3OD,
258C): d=172.9 (COOCH3), 172.5 (C=O of NHAc), 166.3 (CONH),
157.9 (C-2’’’), 151.3 (C-5’’’), 116.9 (C-3’’’), 108.5 (C-4’’’), 87.7 (C-1), 80.9
(C-5), 76.2 (C-3), 74.3 (C-2’’), 73.3 (C-3’’), 71.6 (C-2), 70.5 (C-4), 64.7 (C-
6), 62.7 (C-4’’), 54.2 (C-2’), 53.0 (COOCH3), 49.9 (C-1’’), 32.9 (C-1’), 22.7
(Me of NHAc), 13.7 ppm (Me); IR: n˜ =3315 (OH), 1737, 1638 (C=O),
1524, 1227, 1036 cmꢀ1; MS (FAB): m/z (%): 589 (31) [M+Na+]; HRMS
(FAB): m/z calcd for C22H34N2O13NaS (M+Na)+: 589.1679; found:
589.1700.
2-[4-Carbamoyl-1-(4-ethoxycarbonyl-5-methyl-furan-2-yl)-d-arabinotetri-
tol]ethyl 1-thio-b-d-galactopyranoside (26): Compound 22[23] (760 mg,
1.74 mmol) was dissolved in DMF (7 mL), and amine 24[22] (500 mg,
1.91 mmol) followed by DIPEA (299 mL, 1.74 mmol) and PyBOP
(934 mg, 1.74 mmol) were added. The reaction mixture was stirred vigor-
AHCTUNGERTGoNNUN usly at RT overnight. Then the solvent was removed, and the residue
was diluted with AcOEt and washed with 1m HCl, a saturated aqueous
solution of NaHCO3, and water. The organic layer was dried (Na2SO4),
filtered, and concentrated. The residue was purified by column chroma-
tography on silica gel (CH2Cl2/MeOH, 15:1) to give derivative 25
(586 mg, 0.85 mmol, 49%) as a yellow oil. Deacetylation of this com-
pound by following the general procedure afforded 26 (quant.) as a white
solid. [a]D27 =+3 (c=0.81 in MeOH); 1H NMR (500 MHz, CD3OD,
298 K): d=6.57 (s, 1H; H-4’’’), 4.87 (d, J1’’,2’’ =2.4 Hz, 1H; H-1’’), 4.35 (d,
(2S)-2-{5-[(d-arabinotetritol-1-yl)-2-methyl-3-furancarboxamido]-2-me-
thoxy-carbonyl}ethyl 1-thio-b-d-galactopyranoside (19): Deprotection of
15 (108 mg, 0.13 mmol), by following the general procedure described
above, afforded compound 19 (quant.) as a white solid. [a]D31 =ꢀ65 (c=
0.82 in H2O); 1H NMR (300 MHz, D2O, 25 8C): d=6.66 (s, 1H; H-4’’’),
J
1,2 =9.6 Hz, 1H; H-1), 4.27 (q, 3JH,H =7.1 Hz, 2H; CH2CH3), 3.86 (br d,
4.89 (d, J1’’,2’’ =3.9 Hz, 1H; H-1’’), 4.82 (dd, 1H; H-2’), 4.50 (d, J1,2
9.3 Hz, 1H; H-1), 3.96 (br d, J4,3 =3.0 Hz, 1H; H-4), 3.89 (dd, J2’’,3’’
=
1H; H-4), 3.82–3.74 (m, 2H; H-6a, H-3’’), 3.72–3.66 (m, 2H; H-6b, H-
2’’), 3.58–3.52 (m, 3H; H-2, H-5, H-4a’’), 3.46 (dd, J3,4 =9.6, J3,4 =3.3 Hz,
=
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
18001

A. J. Moreno-Vargas, I. Robina et al.
1H; H-3), 3.37 (dd, 2J4b’’,4a’’ =14.0, J4b’’,3’’ =6.5 Hz, 1H; H-4b’’), 3.03 (dt,
2J1a’,1b’ =13.7, J1a’,2’ =7.3 Hz, 1H; H-1a’), 2.92 (dt, J1b’,2’ =6.5 Hz, 1H; H-
1b’), 2.61 (m, 2H; H-2a’, H-2b’), 2.54 (s, 3H; Me), 1.33 ppm (t, 2H;
CH2CH3); 13C NMR (75.4 MHz, CD3OD, 258C): d=175.3 (CONH),
165.8 (COOEt), 159.6 (C-2’’’), 155.4 (C-5’’’), 115.1 (C-3’’’), 108.5 (C-4’’’),
88.2 (C-1), 80.8 (C-5), 76.2 (C-3), 74.7 (C-2’’), 71.2 (C-2, C-3’’), 70.7 (C-4),
67.6 (C-1’’), 62.9 (C-6), 61.3 (CH2CH3), 44.0 (C-4’’), 38.0 (C-2’), 27.9 (C-
1’), 14.7 (CH2CH3), 13.8 ppm (Me); IR: n˜ =3325 (OH), 2924, 1761, 1630,
1558 (C=O), 1227, 1081 cmꢀ1; MS (FAB): m/z (%): 546 (100) [M+Na+];
HRMS (FAB): m/z calcd for C21H33NO12NaS (M+Na)+: 546.1621; found:
546.1605.
Acknowledgements
This work was supported by the Ministerio de Economꢀa y Competitivi-
dad of Spain (CTQ2012-31247), CITIUS–University of Seville, and Lin-
naeus University, Sweden. J.A. acknowledges financial support from the
MICINN through the Ramꢂn y Cajal program. We also thank to Dr. Josꢄ
Francisco Pꢄrez Hernꢁndez from the Faculty of Veterinary Medicine,
Autonomous University of Barcelona, Spain, for his support within the
project AGL2009-07328.
[1] a) E. Fan, E. A. Merritt, C. L. M. J. Verlinde, W. G. Hol, Curr. Opin.
Struct. Biol. 2000, 10, 680–686; b) C. Podlipnik, A. Bernardi, Acta
Chim. Slov. 2007, 54, 425–436.
[2] a) J. B. Epp, T. S. Widlanski, J. Org. Chem. 1999, 64, 293–295;
b) W. B. Turnbull, B. L. Precious, S. W. Homans, J. Am. Chem. Soc.
2004, 126, 1047–1054.
[3] a) E. A. Merritt, P. Kuhn, S. Sarfaty, J. L. Erbe, R. K. Holmes,
W. G. J. Hol, J. Mol. Biol. 1998, 282, 1043–1059; b) E. A. Merritt, S.
Sarfaty, F. van den Akker, C. L’Hoir, J. A. Martial, W. G. J. Hol, Pro-
tein Sci. 2008, 3, 166–175.
[4] A. Bernardi, L. Raimondi, F. Zuccotto, J. Med. Chem. 1997, 40,
1855–1862.
Benzyl
5-(4-azido-4-deoxy-d-arabinotetritol-1-yl)-2-methyl-3-furoate
(28): A cooled solution of tosyl chloride (2.84 g, 14.89 mmol) in anhy-
drous pyridine (5 mL) was added to a solution of 8[21] (2 g, 5.95 mmol) in
anhydrous pyridine (16 mL) cooled to ꢀ158C. The reaction mixture was
stirred at ꢀ158C for 1.5 h and then water (1.5 mL) was added and the
solvent removed. The resulting residue was diluted with CH2Cl2 and se-
quentially washed with 1m HCl, saturated aqueous solution of NaHCO3
and brine. The organic layer was dried (Na2SO4), filtered and concentrat-
ed. The resulting residue was purified by column chromatography
(AcOEt/petroleum ether, 1:1) to give the corresponding tosyl derivative
(2.02 g, 4.12 mmol, 69%) as a yellow oil. NaN3 (160 mg, 2.46 mmol) was
added to a solution of this compound (604 mg, 1.23 mmol) in DMF
(10 mL), and the mixture was stirred at 808C for 1 h. After evaporation
of the solvent, the residue was diluted with AcOEt and washed with
water. The organic layer was then dried (Na2SO4), filtered, and concen-
trated. Chromatographic purification on silica gel (AcOEt/petroleum
ether, 1:1) afforded 28 (336 mg, 0.93 mmol, 76%) as a white solid.
[a]D27 =ꢀ11 (c=0.99 in CH2Cl2); 1H NMR (300 MHz, CD3OD, 258C):
d=7.45–7.38 (m, 5H; H-Ar), 6.61 (s, 1H; H-4), 5.26 (s, 2H; CH2Ph), 4.88
(d, J1’,2’ =2.1 Hz, 1H; H-1’), 3.84 (ddd, J3’,2’ =8.6, J3’,4b’ =6.3, J3’,4a’ =2.7 Hz,
1H; H-3’), 3.70 (dd, 1H; H-2’), 3.52 (dd, 2J4a’,4b’ =12.8 Hz, 1H; H-4a’),
3.38 (dd, 1H; H-4b’), 2.54 ppm (s, 3H; CH3); 13C NMR (75.4 MHz,
CD3OD, 258C): d=165.4 (COOBn), 159.9 (C-2), 155.6 (C-5), 137.8 (Cq-
Ar), 129.6, 129.2, 129.1 (C-Ar), 114.9 (C-3), 108.5 (C-4), 74.2 (C-2’), 71.6
(C-3’), 67.6 (C-1’), 66.9 (CH2Ph), 55.5 (C-4’), 13.8 ppm (CH3); IR: n˜ =
3298 (OH), 2084 (N3), 1696 (C=O), 1162 cmꢀ1; MS (FAB): m/z (%): 384
(36) [M+Na+]; HRMS (FAB): m/z calcd for C17H19N3O6Na (M+Na)+:
384.1172; found: 384.1163.
[5] A. Bernardi, C. Podlipnik, J. Jimꢄnez-Barbero, in Protein-Carbohy-
drate Interactions in Infectious Diseases (Ed.: C. A. Bewley), RSC
Publishing, Cambridge, 2006, pp. 73–91.
[6] B. D. Spangler, Microbiol. Rev. 1992, 56, 622–647.
[7] a) B. Ernst, J. Magnani, Nat. Rev. Drug Discovery 2009, 8, 661–677;
b) J. C. Pickens, E. A. Merritt, M. Ahn, C. L. M. J. Verlinde, W. G. J.
Hol, E. K. Fan, Chem. Biol. 2002, 9, 215–224; c) A. Bernardi, P.
ˇ
Cheshev, Chem. Eur. J. 2008, 14, 7434–7441; d) C. Podlipnik, I.
Velter, B. La Ferla, G. Marcou, L. Belvisi, F. Nicotra, A. Bernardi,
Carbohydr. Res. 2007, 342, 1651–1660.
[8] A. Bernardi, A. Checchia, P. Brocca, S. Sonnino, F. Zuccotto, J. Am.
Chem. Soc. 1999, 121, 2032–2036.
[9] a) A. Bernardi, D. Arosio, D. Potenza, I. Sanchez-Medina, S. Mari,
F. J. CaÇada, J. Jimꢄnez-Barbero, Chem. Eur. J. 2004, 10, 4395–
4406; b) D. Arosio, S. Baretti, S. Cattaldo, D. Potenza, A. Bernardi,
Bioorg. Med. Chem. Lett. 2003, 13, 3831–3834; c) A. Bernardi, L.
Carrettoni, A. Grosso Ciponte, D. Monti, S. Sonnino, Bioorg. Med.
Chem. Lett. 2000, 10, 2197–2200; d) A. Bernardi, D. Potenza, A. M.
Capelli, A. Garcꢀa-Herrero, F. J. CaÇada, J. Jimꢄnez-Barbero, Chem.
Eur. J. 2002, 8, 4597–4612; e) A. Bernardi, D. Arosio, S. Sonnino,
Neurochem. Res. 2002, 27, 539–545.
[10] P. Cheshev, L. Morelli, M. Marchesi, C. Podlipnik, M. Bergstrçm, A.
Bernardi, Chem. Eur. J. 2010, 16, 1951–1967.
[11] W. E. Minke, C. Roach, W. G. J. Hol, C. L. M. J. Verlinde, Biochem-
istry 1999, 38, 5684–5692.
[12] W. E. Minke, F. Hong, C. L. M. J. Verlinde, W. G. J. Hol, E. Fan, J.
Biol. Chem. 1999, 274, 33469–33473.
[13] M. Bergstrçm, S. Liu, K. L. Kiick, S. Ohlson, Chem. Biol. Drug Des.
2009, 73, 132–141.
[14] D. D. Mitchell, J. C. Pickens, K. Korotkov, E. Fan, H. W. J. Hol,
Bioorg. Med. Chem. 2004, 12, 907–920.
[15] a) A. J. Moreno-Vargas, L. Molina, A. T. Carmona, M. Lambelet, O.
Spertini, I. Robina, ES Patent, ES2325567B2, 2010; b) A. J.
Moreno-Vargas, L. Molina, A. T. Carmona, M. Lambelet, O. Sperti-
ni, I. Robina, ES Patent, ES2324137B2, 2010; c) A. J. Moreno-
Vargas, L. Molina, A. T. Carmona, M. Lambelet, O. Spertini, I.
Robina, Eur. J. Org. Chem. 2008, 2973–2982.
[16] a) S. M. Watt, A. M. Teixeira, G.-Q. Zhou, R. Doyonnas, Y. Zhang,
F. Grunert, R. S. Blumberg, M. Kuroki, K. M. Skubitz, P. A. Bates,
Blood 2001, 98, 1469–1479; b) S. C. Hsu-Lin, C. L. Berman, B. C.
Furie, D. August, B. Furie, J. Biol. Chem. 1984, 259, 9121–9126;
c) G. S. Kansas, G. S. Wood, D. M. Fishwild, E. G. Engleman, J. Im-
munol. 1985, 134, 2995–3002.
Triazole derivative 30: Aqueous sodium ascorbate (1.1 equiv, 1m) and an
aqueous solution of CuSO4 (0.11 equiv, 0.3m) were added to a solution of
alkyne 27[25] (110 mg, 0.27 mmol) and azide 28 (109 mg, 0.30 mmol) in
MeOH (2.75 mL). The reaction vessel was protected from light and the
reaction mixture was stirred vigorously at RT overnight. The mixture was
then diluted with MeOH and filtered. The solution was evaporated to
dryness, and the resulting crude was purified by column chromatography
(CH2Cl2/MeOH, 25:1) to give 29 (164 mg, 0.22 mmol, 82%) as a white
solid. This compound (114 mg, 0.15 mmol) was deacetylated by following
the general procedure to give 30 (quant.) as a white solid. [a]D27 =ꢀ38
(c=0.60 in H2O/DMSO); 1H NMR (300 MHz, CD3OD, 258C): d=7.96
(s, 1H; H-triazole), 7.45–7.28 (m, 5H; H-Ar), 6.61 (s, 1H; H-3’’’), 5.27 (s,
2H; CH2Ph), 4.88 (under water, 1H; H-1’’), 4.76 (dd, 2J4a’’,4b’’ =14.0,
J
J
J
4a’’,3’’ =2.7 Hz, 1H; H-4a’’), 4.39 (dd, J4b’’,3’’ =8.1 Hz, 1H; H-4b’’), 4.27 (d,
1,2 =9.6 Hz, 1H; H-1), 4.11 (d, 2J1a’,1b’ =14.6 Hz, 1H; H-1a’), 4.04 (td,
3’’,2’’ =8.1 Hz, 1H; H-3’’), 3.91 (d, 1H; H-1b’), 3.85 (br d, J4,3 =3.4 Hz,
1H; H-4), 3.78 (dd, 2J6a,6b =11.5, J6a,5 =7.3 Hz, 1H; H-6a), 3.71–3.61 (m,
2H; H-6b, H-2’’), 3.60 (t, 1H; J2,3 =9.6 Hz, H-2), 3.52 (m, 1H; H-5), 3.41
(dd, 1H; H-3), 2.54 ppm (s, 3H; Me); 13C NMR (75.4 MHz, CD3OD,
298 K): d=165.4 (COOBn), 160.0 (C-5’’’), 155.3 (C-2’’’), 146.2 (C-tria-
zole), 137.8 (Cq-Ar), 129.6, 129.2, 129.1 (C-Ar), 126.0 (CH-triazole),
115.0 (C-4’’’), 108.6 (C-3’’’), 86.4 (C-1), 80.8 (C-5), 76.2 (C-3), 75.0 (C-2’’),
71.4 (C-2), 71.3 (C-3’’), 70.7 (C-4), 67.6 (C-1’’), 67.0 (CH2Ph), 62.9 (C-6),
54.7 (C-4’’), 24.2 (C-1’), 13.8 ppm (Me); MS (FAB): m/z (%): 618 (32)
[M+Na+]; HRMS (FAB): m/z calcd for C26H33N3O11NaS (M+Na)+:
618.1734; found: 618.1763.
[17] a) S. R. Barthel, J. D. Gavino, L. Descheny, C. J. Dimitroff, Expert
Opin. Ther. Targets 2007, 11, 1473–1491; b) T. Satoh, M. Kaneko,
M.-H. Wu, H. Yokozeki, K. Nishioka, Eur. J. Immunol. 2002, 33,
18002
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17989 – 18003

Non-Carbohydrate Ligands of Cholera Toxin
FULL PAPER
1274–1281; c) R. Renkonen, Adv. Exp. Med. Biol. 1998, 435, 63–73;
d) S. D. Rosen, Annu. Rev. Immunol. 2004, 22, 129–156.
[18] T. D. Connell, Expert Rev. Vaccines 2007, 6, 821–834.
[19] For reviews on thio-oligosaccharides and thio-glycopeptides, see:
a) K. Pachamuthu, R. R. Schmidt, Chem. Rev. 2006, 106, 160–187;
b) Z. J. Witczak, J. M. Culhane, Appl. Microbiol. Biotechnol. 2005,
69, 237–244.
[20] P. M. Dey, Z. J. Witczak, Mini-Rev. Med. Chem. 2003, 3, 271–280.
[21] G. Coste, T. Horlacher, L. Molina, A. J. Moreno-Vargas, A. T. Car-
mona, I. Robina, P. H. Seeberger, S. Gerber-Lemaire, Synthesis
2011, 1759–1770.
[28] M. Mayer, B. Meyer, J. Am. Chem. Soc. 2001, 123, 6108–6117.
[29] a) V. Jayalakshmi, N. R. Krishna, J. Magn. Reson. 2002, 155, 106–
118; b) J. Angulo, C. Rademacher, T. Biet, A. J. Benie, A. Blume,
H. Peters, M. Palcic, F. Parra, T. Peters, Methods Enzymol. 2006,
416, 12–30; c) J. Angulo, P. Nieto, Eur. Biophys. J. 2011, 40, 1357–
1369.
[30] For a recent review of Garcꢀa-Gonzꢁlez reaction, see: M. C. Scherr-
mann, Top. Curr. Chem. 2010, 295, 1–18.
[31] D. J. Miller, F. Yu, D. W. Knight, R. K. Allemann, Org. Biomol.
Chem. 2009, 7, 962–975.
[32] G. Bartoli, J. G. Fernꢁndez-BolaÇos, G. Antonio, G. Foglia, S. Giuli,
R. Gunnella, M. Mancinelli, E. Marcantoni, M. Paoletti, J. Org.
Chem. 2007, 72, 6029–6036.
[22] A. J. Moreno-Vargas, J. Jimꢄnez-Barbero, I. Robina, J. Org. Chem.
2003, 68, 4138–4150.
[23] P. R. Ahston, S. E. Boyd, C. L. Brown, S. A. Nepogodiev, E. W.
Meijer, H. W. I. Peerlings, H. W. I. Peerlings, J. F. Stoddart, Chem.
Eur. J. 1997, 3, 974–984.
[24] F. M. Ibatullin, K. A. Shabalin, J. V. Jꢈnis, A. G. Shavvac, Tetrahe-
dron Lett. 2003, 44, 7961–7964.
[25] M. Singer, M. Lopez, L. F. Bornaghi, A. Innocenti, D. Vullo, C. T.
Supuran, S. A. Poulsen, Bioorg. Med. Chem. Lett. 2009, 19, 2273–
2276.
[33] a) L. Molina, A. J. Moreno-Vargas, A. T. Carmona, I. Robina, Syn-
lett 2006, 1327–1330; b) L. Molina, E. Moreno-Clavijo, A. J.
Moreno-Vargas, A. T. Carmona, I. Robina, Eur. J. Org. Chem. 2010,
3110–3119.
[34] F. M. Ibatullin, S. I. Selivanov, Tetrahedron Lett. 2002, 43, 9577–
9580.
[35] Z. R. Laughrey, S. E. Kiehna, A. J. Riemen, M. L. Waters, J. Am.
Chem. Soc. 2008, 130, 14625–146233.
[26] D. S. Hage, J. Chromatogr. B 2002, 768, 3–30.
[27] M. Mayer, B. Meyer, Angew. Chem. 1999, 111, 1902–1906; Angew.
Chem. Int. Ed. 1999, 38, 1784–1788.
Received: July 16, 2013
Revised: August 22, 2013
Published online: November 21, 2013
Chem. Eur. J. 2013, 19, 17989 – 18003
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
18003