Mimicking chitin: Chemical synthesis, conformational analysis, and molecular recognition of the β(1→3) N-acetylchitopentaose analogue

Article abstract of DOI:10.1002/chem.200902860

Mimicking Nature by using synthetic molecules that resemble natural products may open avenues to key knowledge that is difficult to access by using substances from natural sources. In this context, a novel N- acetylchitooligosaccharide analogue, β-1,3-N-a

Full text of DOI:10.1002/chem.200902860

FULL PAPER
DOI: 10.1002/chem.200902860
Mimicking Chitin: Chemical Synthesis, Conformational Analysis, and
Molecular Recognition of the b
N
Maria Morando,[a] Yanping Yao,[b, c] Sonsoles Martꢀn-Santamarꢀa,[d] Zhenyuan Zhu,[b, c]
Tong Xu,[c] F. Javier CaÇada,[a] Yongmin Zhang,*[b, c] and Jesffls JimØnez-Barbero*[a]
Abstract: Mimicking Nature by using
synthetic molecules that resemble natu-
ral products may open avenues to key
knowledge that is difficult to access by
using substances from natural sources.
In this context, a novel N-acetylchito-
oligosaccharide analogue, b-1,3-N-acet-
amido-gluco-pentasaccharide, has been
designed and synthesized by using ami-
noglucose as the starting material. A
phthalic group has been employed as
the protecting group of the amine
moiety, whereas a thioalkyl was used as
the leaving group on the reducing end.
The conformational properties of this
new molecule have been explored and
compared to those of the its chito ana-
logue, with the b-1,3 linkages, by a
combined NMR spectroscopic/molecu-
lar modeling approach. Furthermore,
the study of its molecular recognition
properties towards two proteins,
a
lectin (wheat germ agglutinin) and one
enzyme (a chitinase) have also been
performed by using NMR spectroscopy
and docking protocols. There are
subtle differences in the conformation-
al behavior of the mimetic versus the
natural chitooligosaccharide, whereas
this mimetic is still recognized by these
two proteins and can act as a moderate
inhibitor of chitin hydrolysis.
Keywords: conformation analysis ·
glycomimetics · molecular recogni-
tion · NMR spectroscopy · synthetic
methods
Introduction
N-Acetylchitooligosaccharides are involved in a variety of
biological events. Apart from being part of the exoskeleton
of different insects and fungi, they can also act as chemical
signals in plant-induced resistance and are essential parts of
the lipopolysaccharides, which behave as nodulation fac-
tors.[1] The use of mimetics (in this case, glycomimetics) is of
paramount importance to further evaluate these key pro-
cesses, from understanding the recognition process, to ex-
ploring the mechanisms and finding molecules with en-
hanced activities. Chitin oligosaccharides (COs, N-acetylchi-
tooligosaccharides) and chitosan oligosaccharides (CSOs,
chitooligosaccharides) are ubiquitous potential external
chemical signals,[1,2] derived by many chitinases from the cell
walls of fungi or the degradation of insoluble chitin.
[a] M. Morando,+ F. J. CaÇada, J. Jimꢀnez-Barbero
Chemical and Physical Biology, Centro de Investigacione
Biolꢁgicas, CSIC, Ramiro de Maeztu 9
28040 Madrid (Spain)
Fax : (+34)915360432
E-mail: jjbarbero@cib.csic.es
[b] Y. Yao,+ Z. Zhu, Y. Zhang
Universitꢀ Pierre & Marie Curie-Paris 6
Institut Parisien de Chimie Molꢀculaire (UMR 7201)
C 181, 4 place Jussieu, 75005 Paris (France)
Fax : (+33)1-4427-5504
E-mail: yongmin.zhang@upmc.fr
[c] Y. Yao,+ Z. Zhu, T. Xu, Y. Zhang
Institute of Biotechnology, College of Agriculture Biotechnology
Zhejiang University, Hangzhou 310029 (China)
[d] S. Martꢂn-Santamarꢂa
Department of Chemistry, Faculty of Pharmacy
Universidad San Pablo CEU, Boadilla del Monte
28668 Madrid (Spain)
[+] These two authors have equally contributed to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200902860. It contains all the
details of the calculations for the molecular modeling, the F and Y
variations of every bGlcNAc1-3GlcNAc glycosidic linkage during the
vacuum (two force fields), solvated MD simulations, and sections of
NMR spectra.
Chem. Eur. J. 2010, 16, 4239 – 4249
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4239

N-Acetylchitooligosaccharides and chitooligosaccharides,
composed of 2-deoxy-2-acetamido-d-glucopyranosyl resi-
dues and 2-deoxy-2-amido-d-glucopyranosyl residues, re-
spectively, are important structural parts of glycans, glyco-
peptides, and glycoproteins. Chitin oligosaccharides and
their derivatives play an important role as signal molecules
in plant and animal processes, and they are involved in de-
velopmental and defense-related signaling pathways.[3] After
binding with members of the GhCTL group (a new group of
chitinase-like proteins), chitin oligosaccharides are also es-
sential for cellulose synthesis in primary and secondary cell
walls.[4]
cosamine pentasaccharide to be used as a tool for initial in-
teraction studies as a ligand or inhibitor for lectins and en-
zymes involved in the recognition and metabolism of chito-
AHCTUNGTREGoNNNU ligosacharides. The knowledge derived from this study
would be of importance for further investigation of the func-
tion of COs signals in plants and in defense mechanisms.
Results
Synthesis: A key building block in the synthesis of pentasac-
charide 1 was methyl (4,6-O-benzylidene-2-deoxy-2-phthal-
Although the defense mechanism against pathogens has
become increasingly apparent, little is known about the
mechanisms of the COs molec-
N
ular signals perceived by cells
and about the molecular struc-
tures of COs responsible for in-
duced resistance; even very
little is known about the molec-
ular basis of the signal trans-
duction pathways underlying
Scheme 1. Synthesis of the trisaccharide acceptor 2.
oligosaccharide recognition pro-
cesses. Even more, it is hard to
understand the determinant specificity of oligosaccharide
idene-2-deoxy-2-phthalimido-b-d-glucopyranoside (2), which
could be readily prepared from the known trisaccharide 3[7]
recognition processes.[5]
The induced activity of COs and CSOs is determined by
both the molecular weight and the degree of acylation/ace-
tylation, and the molecular weight is directly influenced by
the degree of polymerization (DP).[6] Besides, the main
backbone of COs, sugar residues, and the different substitu-
ents could all influence the plant-induced resistance of COs.
Mimicking nature by using synthetic molecules that re-
semble natural products may open avenues to key knowl-
edge that is difficult to access by using substances from nat-
ural sources. In particular, the oligosaccharides obtained
from natural sources are not suitable for studies on the
mechanism of induced resistance due to their complicated
compositions, varied molecular weight, and the difficulty in
isolating a pure component from natural products. Chemical
synthesis could be used to obtain pure oligosaccharides with
explicit structures, which are appropriate for the induced-re-
sistance study. To know wheth-
by deacetylation with a Mg
Scheme 1.
ACHTUNGRTEN(NUNG OMe)2 solution, as shown in
Compound 2 was obtained after stirring for 48 h in dry di-
chloromethane at room temperature. The trisaccharide 2
has a free hydroxyl group at C-3’’, plus a methyl group at C-
1, which can then be used as an acceptor in the next glycosy-
lation reaction.
For the synthesis of the pentasaccharide 1, a 2+3 block-
synthesis strategy was used. Treatment of the known com-
pound 4[8] with BF3·Et2O/HgO gave a hemiacetal 5, which
was used directly for the next step without further character-
ization. Reaction of
5
with Cl3CCN/1,8-diazabicyclo-
AHCTUNGTRENNUNG
40% yield (for two steps), as shown in Scheme 2. The
1H NMR spectrum showed that the imidate was formed es-
sentially in the b form (J1,2 =8.8 Hz).
er the b-1,4 linkage is essential
in keeping the biological prop-
erties of these molecules, it is
important to synthesize a series
of chitin oligosaccharide ana-
Scheme 2. Synthesis of the donor 6.
logues possessing
a different
backbone linkage (for example,
a b-1,3 linkage), to explore the
plant-induced resistance elicited by these chitin oligosac-
charide derivatives.
Condensation of 6 with the previously described 7,[9] in
the presence of trimethylsilyl triflate (TMSOTf) and di-
In our previous work, we have synthesized two chitin oli-
gosaccharide analogues, b-1,3-N-acetyl-glucosamine disac-
charide and b-1,3-N-acetyl-glucosamine trisaccharide.[7]
Herein, we report the first synthesis of b-1,3-N-acetyl-glu-
chloromethane, gave b
G
yield (Scheme 3). Its stereochemistry was determined to be
the desired b form on the basis of the H-1’,H-2’ coupling
constant (J1,2 =8.0 Hz).
4240
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4239 – 4249

Mimicking Chitin
FULL PAPER
was performed. Purification of
the product was performed by
chromatography on Sephadex
G25 (H2O), to provide the de-
sired
b-1!3-N-acetyl-glucos-
Scheme 3. Synthesis of the disaccharide donor 8.
AHCTUNGTERGaNNUN mine pentasaccharide (1) in
48% yield (for two steps), as
shown in Scheme 5.
The glycosylation of the trisaccharide 2 with donor 8 was
achieved in the presence of N-iodosuccinimide (NIS)-tri-
fluoromethanesulfonic acid (TfOH) in dry dichloromethane
(with 4 ꢄ ground molecular sieves) for 4 h at ꢀ308C, pro-
NMR spectroscopic data and assignment: All the details of
the NMR experiments are given in the material and meth-
ods sections. The combination of selective 1D-TOCSY, se-
lective 1D-NOESY, 2D-NOESY, 2D-TOCSY (in D2O and
H2O/D2O), and 2D-HSQC experiments[10] permitted assign-
ment of all the resonances of
viding the desired pentasaccharide
(Scheme 4).
9 in 90% yield
the pentasaccharide 1 (see the
Supporting Information). As
frequently found in carbohy-
drates, severe overlap within
the ring proton region was
found. Fortunately, there was a
Scheme 4. Synthesis of the protected pentasaccharide 9.
distinction between the signals
arising from the different resi-
The stereochemistry of the newly introduced linkage was
determined to be b, on the basis of the GlcN H-1,H-2 cou-
pling constant (J1,2 =8.4 Hz).
dues within the amide region, which was evaluated by run-
ning experiments in H2O.
As an example, sections of the TOCSY and HSQC spec-
tra recorded at 800 MHz and 298 K are presented in
Figure 1. The behavior of the amide protons with tempera-
ture is also given. Second-order effects were found in the
anomeric protons even at 800 MHz, as depicted in the Sup-
porting Information. However, the recording of the spectra
at different temperatures permitted us to evaluate the cou-
plings and to assign most of the signals. In any case, the
analysis of the coupling constants and the NOE patterns al-
lowed us to establish that the pyranoid rings adopt the ex-
Deprotection of the benzylidene groups of 9 was first per-
formed by the classical hydrogenolysis method. Surprisingly,
the desired compound could not be isolated from the reac-
tion mixture after testing a variety of reaction times, sol-
vents, and amounts of catalyst. An alternative method was
then used. Treatment of pentasaccharide 9 with p-toluene-
sulfonic acid (PTSA) in THF and methanol for 24 h at 558C
gave a mixture of two compounds 10-1 and 10-2 in 87%
yield and in a 3:1 ratio (Scheme 5).
4
pected C1 chair conformations.
The chemical shifts are given in
Table 1.
To properly evaluate the
NOEs, and to discard the possi-
bility of aggregates at the work-
ing concentration, NMR diffu-
sion
order
spectroscopy
(DOSY) experiments[11] were
used to confirm the aggregation
state of the molecules for the
employed experimental condi-
tions. It was confirmed that, at
Scheme 5. Deprotection and N-acetylation of the pentasaccharide.
the working concentration (ca.
1–2 mm), the molecule behaved
After separation, the MS analysis revealed that 10-1 is the
desired 4,6-O-debenzylidenation product, whereas 10-2 is a
derivative of 3e-deacetylated 10-1. These two compounds
were both used for the next reaction without further charac-
terization. Treatment of 10-1 and 10-2 with hydrazine hy-
drate and water in boiling ethanol, followed by acetylation
with acetic anhydride and NaHCO3 in water and methanol
as a monomer, since at these concentrations the diffusion
coefficient was basically identical to that previously reported
for the chitin pentasaccharide.[11] Thus, the experimental
data can be unambiguously correlated with a single species.
After assignment of the key protons, emphasis was placed
on the inter-residual NOEs and those connecting the differ-
ent fragments of the molecules. It was observed that every
Chem. Eur. J. 2010, 16, 4239 – 4249
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4241

J. Jimꢀnez-Barbero, Y. Zhang et al.
MM3*[13] as implemented in the
Macromodel/MAESTRO[14]
package, and as described in
the Experimental Section.
As the first step, the penta-
saccharide was built by setting
all the F and Y angles of every
glycosidic linkage to 60:08.
Then, the resulting geometry
was first extensively minimized
by using conjugate gradients
and then taken as the starting
structure for the MD simula-
tions by using AMBER* and
MM3*. Typical trajectories are
displayed in the Supporting In-
formation. In all cases, no chair-
to-chair or chair-to-boat inter-
conversions were observed. De-
spite extensive minimization of
the starting conformers, it is im-
portant to remark that, in both
cases, although the average
temperature was constant at
around 300 K, important fluctu-
ations (275–325 K) took place
during the simulations, inde-
pendently of the force field that
Figure 1. A) Sections of the NOESY experiment (300 ms, mixing time) from the anomeric region to the sugar
resonances. B) Section of HSQC for the ring sugar protons, with the exception of the anomeric region. C) Var-
iation of the shape of the amide protons of 1 with temperature.
Table 1. 1H NMR spectroscopic chemical shifts [ppm] deduced for the
was employed. This situation is known to introduce artefacts
in the dynamic properties of the system.[15] Although the re-
sults should be considered as merely qualitative, it can be
observed that, for both MM3* and AMBER* force fields
(see also the Supporting Information), the trajectory re-
mained in the corresponding low-energy region for the four
glycosidic linkages, with basically no interconversions be-
tween conformers at these points. Therefore, according to
these MD simulations, the exo-anomeric conformer[16] at
every glycosidic torsion was the most stable one, from a con-
formational point of view, together with the syn-Y confor-
mer for the aglyconic linkage. Average angles oscillate be-
tween 40–658 for F and 0–408 for Y.
Several transitions between the possible orientations of
the hydroxymethyl groups were also observed, especially be-
tween the gg and gt rotamers.[17] Important differences were
found for the torsion angle at the amide moiety. For the
AMBER* simulations, an average H2-C2-N-H torsion of
2408 was found, whereas for the MM3* protocol, the aver-
age value was 1808. These values were maintained even
when the temperature of the simulation was set to 288, 303,
or 313 K (6 ns simulation in each case with both force
fields). In both cases, and considering the major conforma-
tion around the glycosidic linkage, there is basically no pos-
sibility of establishing inter-residual hydrogen bonds.
pentasaccharide 1 in H2O/D2O at 298 K and 900 MHz.
Proton
Ring
H1
H2
H3
H4
H5
H6
NH
CH3
a
b
c
d
e
4.57
4.54
4.53
4.52
4.35
3.65
3.69
3.68
3.71
3.74
3.57
3.78
3.80
3.82
3.74
3.45
3.49
3.49
3.49
3.50
3.44
3.43
3.43
3.43
3.45
3.91, 3.73
3.91, 3.73
3.91, 3.73
3.91, 3.73
3.91, 3.73
8.02
8.15
8.15
8.15
8.10
1.99
2.01
2.02
2.00
2.05
anomeric proton produced one NOE contact with H-3 and
H-5 of its own residue and H-3 at the residue to the other
side of the glycosidic linkage. The intraresidual NOEs were
employed as internal references for estimation of the
proton–proton distances (Figure 1) and compared to those
calculated from molecular mechanics (MM) and molecular
dynamics (MD) calculations. Regarding the temperature co-
efficients for the amide protons, all of them oscillated be-
tween 8 and 10 ppbKꢀ1, which indicates that their degree of
protection or participation in the intramolecular hydrogen
bond is rather low. Indeed, they are in fast exchange with
bulk solvent H2O, and at 303 K, their signals are basically
lost (Figure 1).
Molecular dynamics simulations: Different protocols were
employed for accessing to the conformational and dynamic
information of compound 1. First, MD simulations (10 ns)
were performed by using a continuum solvent model (GB/
SA) and two different force fields, AMBER*[12] and
As further step, a 10 ns MD simulation in explicit water
was also performed with the AMBER 9.0 force field, as de-
scribed in the Experimental Section. The protocol included
a 100 ps period in which the system was heated (100–303 K),
4242
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4239 – 4249

Mimicking Chitin
FULL PAPER
followed by a 100 ps equilibration at 303 K. Positional re-
straints were applied to the heavy atoms and were gradually
lowered until no constrictions were applied. The unrestrain-
ed molecular dynamics simulation was continued during
10 ns at 303 K and 1 atm. Typical trajectories are displayed
in Figure 2 and in the Supporting Information. The stability
Discussion
Correlation between the NMR spectroscopic and MD data:
According to the MD simulations, the glycosidic linkages
adopt a well-defined conformation in the F/Y region
around 60:08. To demonstrate the existence of conformers
defined by these values, the ex-
perimental NOE data can be
analyzed. Indeed, for these geo-
metries, close distances between
H1 of a given residue and H3
of the following one should be
expected, with interproton dis-
tances around 2.4 ꢄ for every
glycosidic linkage.
These contacts were indeed
present in the NOE spectra,
with strong intensities, in the
range of those found for the in-
traresidual H1–H5 proton pair,
which, according to the simula-
tions, is defined by a very simi-
4
lar distance in a regular C1 (D)
chair. Thus, the experimental
NMR spectroscopic results vali-
dated the accuracy of the MD
simulations. Also, the global
minimum generated from the
MD simulations after extensive
energy minimization cannot
give any intraresidual hydrogen
bond, as also shown by the ex-
periments that followed the
chemical-shift variations of the
amide protons with tempera-
ture. Moreover, strong cross-
peaks between the amide pro-
tons and the water peak were
observed in the NOESY spec-
Figure 2. A) F/Y plot of the glycosidic torsions at the nonreducing end and its contiguous moiety of pentasac-
charide 1 during the solvated MD simulation (10 ns, AMBER) at 300 K. B) Trajectories of the C4-C5-C6-O6
w torsion angles at the nonreducing GlcNAc moiety during the 10 ns solvated simulation of 1 (AMBER,
300 K). C) Trajectory of the H2-C2-N-H torsion angle at the nonreducing end moiety during the 10 ns solvated
simulation of 1 (AMBER, 300 K). D) Trajectory of the key inter-residual distance [ꢄ] between the anomeric
H-1 proton at the nonreducing end and the H-3 proton at the contiguous GlcNAc unit. The MD simulation
spanned 10 ns with the AMBER force field at 300 K.
of the energy, pressure, and temperature was monitored
along the trajectory and is given in different plots in the
Supporting Information. The system was found to be fairly
stable under the employed conditions. Again, the glycosidic
linkages adopt the same conformation described above with
tra, thus indicating their availability to chemical exchange
processes. For the conformation around the amide region,
the experimentally measured coupling constants JNH,H2
(larger than 8 Hz) for all the residues are in agreement with
major anti-type conformations for the corresponding pro-
tons, thus supporting the conclusions of the MM3* and
AMBER 9.0 MD simulations. Also, the NHs gave NOE
cross-peaks with the intraresidual H-1 proton (correlated
with average MD HN–H1 distances of ca. 2.7 ꢄ), which
were stronger than those corresponding to the HN–H2 pairs
(average MD HN–H2 distances of ca. 2.9 ꢄ), again support-
ing a major anti-type geometry for the amide linkages.
In the obtained conformational families, the sugar glycosi-
dic linkages can be described by major conformers in which
F/Y are 60:08 with fluctuations around these values. Al-
though the geometry is well defined, the pentasaccharide is
not rigid at all, since the individual fluctuations at every
linkage permits the existence of a large degree of accessible
fluctuations around 30–658 for
F and 0–408 for Y
(Figure 2). Also, transitions between the gg and gt rotamers
of the hydroxymethyl groups were observed, with popula-
tions ranging between 60–80% for gg and 40–20% for gt,
depending on the position within the pentasaccharide se-
quence (Figure 2). Nevertheless, it is possible that considera-
bly longer simulation times[18] (~100 ns) may be necessary to
adequately sample the conformational space available to
these molecules. Under these conditions in explicit water,
the amide torsions adopted a major conformation with an
anti-like H2-C2-N-H geometry for every residue, similar to
that found in the MD simulation with the MM3* force field
and the continuum solvent model (Figure 2).
Chem. Eur. J. 2010, 16, 4239 – 4249
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4243

J. Jimꢀnez-Barbero, Y. Zhang et al.
conformational space. All the glycosidic torsions behave in
an independent manner, with no correlations between their
individual behavior and no contacts between residues that
are more than one unit apart in the sequence. The possible
conformations are fairly extended. In any case, the relative
degree of flexibility permits these molecules to interact with
a variety of receptors through similar or different presenta-
tion modes. Nevertheless, attending to the different situa-
tions, different entropic penalties will have to be paid for
the interaction to take place. When the major conformer of
this molecule is superimposed with that of the natural chitin
analogue,[19] it can be observed that the acetamide groups
and the C6-hydroxymethyl groups occupy different spatial
orientations. Also, the extension of the chain leads to subtle
differences in the distances between the two reducing and
nonreducing ends. These features may obviously imply dis-
tinct molecular recognition properties when the mimetic
and the analogue interact with biomolecular receptors or
specific enzymes.
tons of the pentasaccharide, especially for those of the non-
reducing end N-acetyl-d-glucosamine residue (e). For this e
residue, the observed STD effects (Table 2) were more than
double than those of the central residues (b–d) and three-
fold those observed for the reducing a end. Thus, the bind-
ing epitope was clearly characterized.
Table 2. STD percentages [%] observed on the different proton resonan-
ces of 1 upon saturation of the 1H NMR spectrum at ca. d=ꢀ1 ppm (pro-
tein envelope).
Ring[a]
H1
H2
H3
H4
H5
H6
COCH3
OCH3
ring a nonred
ring b
ring c+d
ring d
10
8
12
6
15
5
15
7
14
7
15
6
7
7
4
4
8
4
17
8
6
3
2
[b]
–
–
12
6
14
7
[b]
[b]
[b]
[b]
[b]
ring e red
5
–
4
–
–
–
2
[a] See Figure 3. [b] Not detected.
Finally, trNOE experiments were also performed to char-
acterize the bound geometry.[24] Very strong negative cross-
peaks were observed for the pentasaccharide at 298 K and
500 MHz, and even when using a mixing time of only 75 ms.
This fact is a clear indication of binding, because the
NOESY spectrum of the free pentasaccharide at 298 K at
this mixing time was basically devoid of cross-peaks. No
new peaks were observed in the trNOESY spectrum
(Figure 4), when compared with the NOESYs recorded for
the saccharide in the free state at 800 MHz, which indicates
that no lectin-induced conformational variations were taking
place in the pentasaccharide structure. Thus, the major con-
formation in solution is that bound by the lectin.
To rationalize the interaction on the molecular level, the
low-energy conformer of 1, as deduced by the NMR spec-
troscopic experiments, was docked into the different WGA
binding sites[25] by using two different docking programs,
AutoDock[26] and Glide.[27] Four hevein domains are known
in the lectin.[25] The binding sites for chitooligosaccharides
are very similar in the different domains and are defined by
one Ser residue (residue at relative position 62, from the N
terminus), which provides hydrogen bonds to the carbonyl
group of one acetamide residue, two aromatic residues (i.e.,
residues Tyr64 and His66), which make stacking interactions
to two consecutive GlcNAc units, and one tyrosine moiety
(Tyr 73), which provides stabilizing van der Waals interac-
tions to the acetamide methyl group and one additional hy-
drogen bond to one sugar hydroxyl group (Figure 5).[28] The
docking protocol (see the Experimental Section) permitted
the deduction that the N terminus was nicely accommodated
in the protein binding site, at Tyr73. The second and third
units provided minor contacts with the protein, whereas the
fourth one and the reducing end were basically in contact
with the solvent. It has been demonstrated that chitooligo-
saccharides may adopt different orientations at the binding
sites of hevein domains. In this case, given the different ori-
entations of the acetamide moieties for the GlcNAc residues
in the b-1!3 versus the b-1!4 linkages, Glide only led to
Molecular recognition. The interaction with a model lectin
and a model enzyme: WGA and chitinase: Since, in princi-
ple, the shape of the pentasaccharide is similar to that
adopted by the similar penta-N-acetylated chitopentaose an-
alogue, the possibility that the synthetic pentasaccharide
could also be recognized by a chitin-binding lectin was ana-
lyzed. As a model carbohydrate-binding protein, wheat
germ agglutinin was chosen, since it has been deeply studied
and its recognition mode by chitin fragments has been
firmly established.[20]
Thus, NMR spectroscopic-based binding experiments
were performed,[21] at different concentrations with penta-
saccharide/lectin ratios of 100:1, 40:1, and 20:1. The meas-
urements were performed at 298 K and STD experiments[22]
were employed to detect binding and to deduce the binding
epitope of the pentasaccharide (Figure 3).[23]
Very neat results were obtained with the sample at a ratio
L/P 20:1, which showed significant STD signals for the pro-
Figure 3. STD-NMR spectroscopic experiments on the WGA/pentasac-
charide system. The number of scans was 128 ns, with 1 mm of the penta-
saccharide at pH 7.02 (20 mm phosphate in D2O). The ligand to protein
ratio was 20:1. At the left-hand side, the STD on (above) and off-reso-
nance (below) spectra. To the right-hand side, the observed absolute en-
hancements for the nonreducing end, which shows the maximum re-
sponse after perturbation of the protein.
4244
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4239 – 4249

Mimicking Chitin
FULL PAPER
but keeping the overall shape and somehow, the molecular
recognition abilities.
Furthermore, we also explored the possibility of the enzy-
matic hydrolysis of this synthetic pentasaccharide by an N-
acetyl glycosaminidase enzyme. NMR spectroscopic experi-
ments were carried out on a pentasaccharide sample, by
using the chitinase from Streptomyces Grisues. This enzyme
was employed since it is able to readily hydrolyze b1-4 link-
ages in natural chitin oligosaccharides. The 1D 1H NMR
spectroscopic experiments were performed in buffer phos-
phate D2O (50 mm pH 6 at 298 K) with a sugar to enzyme
molar ratio of 20:1. After two hours, the NMR spectra did
not show any variation of the intensity of the sugar peaks,
especially at the structural reporter anomeric region. The
nonreducing-end proton kept the same intensity throughout
the whole two-hour experiment, and also after 24 h. A com-
pletely different behavior was observed for the reference ex-
periments with the natural chitopentaose substrate. Thus,
the change from b1-4 to a b1-3 linkage precludes the hydrol-
ysis, at least for this particular enzyme. However, slight line
broadening was observed for the signals of compound 1 and
also STD experiments gave rise to several signals at the
sugar region, thus assessing that interactions between the
synthetic pentasaccharide 1 with the enzyme were taking
place (Figure 6). DOSY experiments allowed the confirma-
tion that 1 is not degraded. Although merely speculative,
the observed interaction throughout the line broadening of
the NMR spectroscopic resonance signals and the STD en-
hancements could arise from interactions at the chitin bind-
ing domain of the enzyme and not at the intrinsic catalytic
site. Indeed, the inhibitory power of the synthetic pentasac-
charide to avoid chitopentaose degradation was very weak,
in the millimolar range.
Figure 4. A) Tr NOESY spectrum (200 ms mixing time) of 1 in the pres-
ence of WGA, a model lectin. Negative cross-peaks are observed, which
indicate lectin binding. In the views below, the recognition mode of the
chitin mimetic by the lectin is shown. B) Major contacts are observed for
the nonreducing end (as also assessed by the NMR STD experiments),
followed by the second residue. Only marginal contacts are observed for
the third residue, whereas the fourth and fifth ones are in contact with
the solution. C) Pentasaccharide 1 is superimposed with chitobiose, by
using the nonreducing end. The different orientations of the acetamide
and C6-hydroxymethyl groups are evidenced.
Figure 5. Typical binding site of chitooligosaccharides at hevein domains.
The case for one of the binding sites of WGA is shown. The binding is
defined by Ser62, which provides a hydrogen bond to the carbonyl group
of one acetamide residue, residues Tyr64 and His66, which make stacking
interactions to the two consecutive GlcNAc units, and Tyr73, which pro-
vides stabilizing van der Waals interactions to the acetamide methyl
group and one additional hydrogen bond to one sugar hydroxyl group.[28]
Figure 6. Hydrolysis of the natural penta-N-acetyl chitopentaose by the
chitinase enzyme. Clear variations at the sugar region (A) and at the ace-
tate methyl region (B) are observed, verifying hydrolysis. In contrast,
only line broadening is observed for pentasaccharide 1 in the presence of
the chitinase, both for the sugar (C) and methyl region (D). The data in-
dicate ligand binding to the protein, but not hydrolysis.
one major docking pose. The analogous analysis with Auto-
Dock provided the same results. Thus, the key conclusion is
that the docking analysis was in accordance with the NMR
spectroscopic-derived observations. The studies described
herein indicate that the chemical modifications from b
ACHTUNGTNER(NUNG 1-4)
to b(1-3) influence the spatial disposition of the sugar chain,
ACHTUNGTRENNUNG
Chem. Eur. J. 2010, 16, 4239 – 4249
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4245

J. Jimꢀnez-Barbero, Y. Zhang et al.
3-O-Acetyl-4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-glucopyranosyl
trichloroacetimidate (6): A mixture of 5 (690 mg, 1.57 mmol) and 4 ꢄ
molecular sieves (2.16 g) in dry dichloromethane (28 mL) was stirred at
room temperature for 30 min under argon. After the reaction mixture
had been cooled to 08C, trichloroacetonitrile (2.16 mL) and DBU
(279 mL) were added dropwise. The mixture was stirred at 08C for 5 h,
then filtered through a Celite bed. After concentration, the residue was
purified by flash chromatography with silica gel (dichloromethane/ethyl
acetate 40:1 with 0.1% triethylamine) to give the 6 as a white powder
(610 mg, 67%). Rf =0.50 (cyclohexane/ethyl acetate 1.5:1); [a]D =ꢀ24
(c=1 in chloroform); 1H NMR (400 MHz, CDCl3): d=7.90-7.20 (m, 9H;
Experimental Section
Synthesis: General methods: Optical rotations were measured at 20ꢁ
28C with a Perkin–Elmer Model 241 digital polarimeter, by using a
10 cm, 1 mL cell. Chemical ionization (CI) and Fast Atom Bombardment
(FAB) mass spectra were obtained with a JMS-700 spectrometer. Elec-
trospray ionization (ESI) mass spectra were recorded with a Q-TOF1
(Micromass) time-of-flight mass spectrometer. 1H NMR spectra were re-
corded with a Brꢅker DRX 400 spectrometer at ambient temperature.
Assignments were aided by COSY experiments. 13C NMR spectra were
recorded at 100.6 MHz with a Brꢅker DRX 400 for solutions in CDCl3
or D2O. CDCl3 was adopting a peak at d=77.00 ppm (for the central line
of CDCl3). Spectra in water were referenced by using DSS (dodecyl
sodium sulfate) as external standard. Assignments were aided by a J-mod
technique and proton–carbon correlation. Reactions were monitored by
TLC on a precoated plate of silica gel (60F254, layer thickness 0.2 mm, E.
Merck, Darmstadt, Germany) and detection by charring with sulfuric
acid. Flash-column chromatography was performed on silica gel 60 (230–
400 mesh, E. Merck).
arom.), 6.75 (d, 1H, J1,2 =8.78 Hz; H-1), 6.05 (dd, 1H, J2,3 =9.29, J3,4
10.04 Hz; H-3), 5.61 (s, 1H; PhCH), 4.64 (dd, 1H, J1,2 =8.79, J2,3
=
=
10.27 Hz; H-2), 4.53 (dd, 1H, J5,6b =4.17, J6a,6b =9.91 Hz; H-6a), 4.03-3.89
(m, 3H; H-4, H-5, H-6b), 1.96 ppm (s, 3H; OAc); 13C NMR (100.6 MHz,
CDCl3): d=170.14 (C=O, Ac), 167.49, 163.62 (C=O, NPhth), 160.55 (C=
NH), 136.62, 131.13, (arom. C), 134.41, 129.23, 128.24, 126.22, 123.64
(arom. CH), 101.74 (PhCH), 93.94 (C-1), 78.78 (C-4), 69.38 (C-3), 68.37
(C-6), 66.94 (C-5), 54.21 (C-2), 20.53 ppm (CH3C=O); HRMS (FAB+):
m/z: calcd for C25H21O8N2Cl3K: 621.0001 [M+K]+; found: 621.0016.
Methyl (4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-glucopyranosyl)-
(1!3)-(4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-glucopyranosyl)-
(1!3)-4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-glucopyranoside
(2): Magnesium (1.68 g, 0.07 mol) was added to dry methanol (50 mL);
then a few iodine crystals were added into the solution. The mixture was
Phenyl
3-O-(3-O-acetyl-4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-
glucopyranosyl)-4,6-O-benzylidene-2-deoxy-2-phthalimido-1-thio-b-d-glu-
copyranoside (8): A solution of 6 (360 mg, 0.62 mmol, 1.2 equiv) and 7
(250 mg, 0.52 mmol, 1 equiv) in dry dichloromethane (2.5 mL) was stirred
with ground 4 ꢄ (896 mg) molecular sieves for 40 min at room tempera-
ture under an argon atmosphere. TMSOTf (110 mL, 0.52 mmol, 1 equiv)
was added dropwise at 08C, and the mixture was stirred at that tempera-
ture for 3 h. The reaction mixture was then filtered through Celite and
the solid was washed with dichloromethane. The filtrate was washed with
saturated aqueous NaHCO3 solution and then with water, dried over
MgSO4, and concentrated. The residue was flash-chromatographed from
a column of silica gel (cyclohexane/ethyl acetate 2:1) to give 8 (420 mg,
79%). Rf =0.39 (cyclohexane/ethyl acetate 1.5:1); [a]D =+29 (c=1 in
chloroform); 1H NMR (400 MHz, CDCl3): d=7.8–7.20 (m, 23H; arom.),
5.62 (t, 1H, J2’,3’ =J3’,4’ =9.73 Hz; H-3’), 5.61, 5.45 (2s, 2H; 2ꢆPhCH),
5.60, 5.47 (2d, 2H, J=8.54, 8.05 Hz; H-1, H-1’), 4.89 (dd, 1H, J2,3 =8.88,
refluxed for 2 h to give a Mg
ACHTUNGRTENN(UNG OMe)2 solution. The freshly prepared Mg-
ACHTUNGTRENNUNG
0.316 mmol) in dry dichloromethane (12 mL). After the reaction mixture
had been stirred for 48 h at room temperature under argon, TLC showed
completion of the reaction. The mixture was neutralized to pH 7 with
acetic acid, filtered, and concentrated. The residue was dissolved in di-
chloromethane and washed with water, dried with MgSO4, and concen-
trated. The residue was flash-chromatographed with silica gel (dichloro-
methane/ethyl acetate 20:1) to give 2 (329 mg, 89%) as a white powder:
Rf =0.38 (cyclohexane/ethyl acetate 1:2); [a]D =ꢀ38 (c=1.0 in chloro-
form); 1H NMR (400 MHz, CDCl3): d=5.52 (s, 1H; PhCH), 5.47 (s, 1H;
PhCH), 5.42 (s, 1H; PhCH), 5.21 (d, 1H, J=8.40 Hz; H-1“), 5.10, 4.85
J
3,4 =9.84 Hz; H-3), 4.40 (dd, 1H, J5,6a =4.7, J6a,6b =10.47 Hz; H-6a), 4.35
(dd, 1H; H-2), 4.21 (dd, 1H; H-2’), 4.12 (dd, 1H; J5’,6’a =4.81, J6’a,6’b
=
(2d, 2H, J=8.5, 8.4 Hz; H-1, H-1’), 4.82, 4.66 (2dd, 2H, J2,3 =10.37, J3,4
=
10.47 Hz; H-6’a), 3.87 (t, 1H, J=10.10 Hz; H-6b), 3.78 (t, 1H, J=
9.05 Hz; H-4), 3.70 (m, 1H; H-5), 3.67 (t, 2H, J=9.65 Hz; H-4’, H-6’b),
3.48 (m, 1H; H-5’), 1.77 ppm (s, 3H; OAc); 13C NMR (100.6 MHz,
CDCl3): d=170.00 (C=O; Ac), 137.02, 136.85, 131.68 (arom. C), 134.08,
133.93, 132.31, 129.22, 129.09, 128.89, 128.29, 128.15, 128.02, 126.23,
126.00, 123.26 (arom. CH), 101.56, 101.32, 97.58 (2ꢆPhCH, C-l’), 84.72
(C-1), 80.11, 78.82 (C-4, C-4’), 75.81, 69.75 (C-3, C-3’), 70.51, 65.90 (C-5,
C-5’), 68.56, 68.50 (C-6, C-6’), 55.64, 54.54 (C-2, C-2’), 20.35 ppm (CH3C=
O); HRMS (FAB+): m/z: calcd for C50H42O13N2SNa: 933.2305 [M+Na]+;
found: 933.2318.
9.10, J2’,3’ =10.30, J3’,4’ =8.90 Hz; H-3, H-3’), 4.27 (dd, 1H, J2’’,3’’ =10.25,
J
3’’,4’’ =9.78 Hz; H-3’’), 4.35, 4.20 (2dd, 2H, J5,6b =4.81, J6a,6b =10.45, J5’,6’b
=
=
4.80, J6’a,6’b =10.44 Hz; H-6b, H-6’b), 4.14, 4.08 (2dd, 2H, J1,2 =8.45, J2,3
8.51, J1’2’ =8.32, J2’,3’ =10.30 Hz; H-2, H-2’), 4.11 (dd, 1H, J6’’a,6’’b =10.40,
5’’,6’’b =4.53 Hz; H-6’’b), 4.03 (dd, 1H, J1’’,2’’ =8.45, J2’’,3’’ =8.39 Hz; H-2’’),
J
3.81, 3.72 (2t, 2H, J5,6a =J6a,6b =10.25, J5’,6’a =J6’a,6’b =10.17 Hz; H-6a, H-
6’a), 3.65 (t, 2H, J=9.19, 9.01 Hz; H-4, H-4’), 3.57 (t, 1H, J=10.7 Hz; H-
6’’a), 3.55, 3.39 (2 m, 2H; H-5, H-5’), 3.39 (dd, 1H, J=9.13, 9.10 Hz; H-
4’’), 3.30 (s, 3H; OCH3), 3.27 (m, 1H; H-5’’), 2.16 ppm (brs, 1H; OH);
13C NMR (100.6 MHz, CDCl3): d=137.21, 137.11, 136.91, 131.37, 131.15,
130.92, (arom. C), 133.90, 133.62, 129.29, 129.27, 129.07, 129.01, 128.28,
128.24, 128.20, 128.15, 126.23, 126.04, 125.99, 123.48, 123.19, 123.07
(arom. CH), 101.76, 101.11 (3ꢆPhCH), 99.65, 97.43, 97.17 (C-1, C-1’, C-
1”), 81.79, 79.90, 79.78 (C-4, C-4’, C-4“), 74.22, 73.93, 68.34 (C-3, C-3’, C-
3”), 68.62, 68.54 (C-6, C-6’, C-6“), 66.29, 66.07, 65.72 (C-5, C-5’, C-5”),
56.84 (OCH3), 56.01, 55.58, 55.54 ppm (C-2, C-2’, C-2“); HRMS (CI+):
m/z: calcd for C64H55O19N3Na: 1192.3327 [M+Na]+; found: 1192.3298.
Methyl (3-O-acetyl-4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-gluco-
pyranosyl)-(1!3)-(4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-gluco-
pyranosyl)-(1!3)-(4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-gluco-
pyranosyl)-(1!3)-(4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-gluco-
pyranosyl)-(1!3)-4,6-O-benzylidene-2-deoxy-2-phthalimido-b-d-gluco-
pyranoside (9): A mixture of compound 2 (312 mg, 0.267 mmol, 1 equiv),
8 (243 mg, 0.267 mmol, 1 equiv), 4 ꢄ powered molecular sieves (1.2 g)
and dry dichloromethane (17 mL) was stirred at room temperature for
30 min under argon. NIS (140 mg, 0.61 mmol, 2.3 equiv) was added and
the reaction mixture was cooled to ꢀ308C; then triflic acid (2.36 mL,
0.027 mmol, 0.1 equiv) was introduced by dilution in dichloromethane.
After stirring at ꢀ308C for 4 h, the reaction mixture was neutralized with
Et3N, filtered through a Celite bed, washed with water, saturated aqueous
thiosulfate solution, and saturated brine, dried with MgSO4, and concen-
trated. The residue was flash-chromatographed with silica gel (cyclohex-
ane/acetone 4:3), to give 9 (472 mg, 90%) as an amorphous white
powder. Rf =0.32 (cyclohexane/acetone 1:1). [a]D =ꢀ49 (c=1 in chloro-
form); 1H NMR (400 MHz, CDCl3): d=7.80–7.10 (m, 45H; arom.), 5.49
(t, 1H, J=9.8 Hz; H-3e), 5.49 (s, 1H; PhCH), 5.41 (s, 1H; PhCH), 5.40
(s, 1H; PhCH), 5.38 (s, 1H; PhCH), 5.37 (s, 1H; PhCH), 5.35 (d, 1H, J=
8.35 Hz; H-1e), 4.99, 4.85, 4.82, 4.78 (4d, 4H, J=8.37, 8.33, 8.23, 7.39 Hz;
3-O-Acetyl-4,6-O-benzylidene-2-deoxy-2-phthalimido-d-glucopyranose
(5): A solution of compound 4 (3 g, 5.64 mmol, 1 equiv), THF, (18 mL),
and water (3.6 mL) was stirred for 30 min at room temperature. Then the
reaction mixture was cooled to 08C and a mixture of HgO (1.8 g,
8.42 mmol, 1.5 equiv), THF (8 mL), and BF3·Et2O (1.4 mL, 2.0 equiv)
was added slowly. After stirring at room temperature for 32 h, the reac-
tion mixture was neutralized with Et3N and concentrated. The residue
was dissolved in dichloromethane, washed with saturated aqueous
NaHCO3 solution, aqueous KI solution (10%), water, and saturated
brine. The resulting mixture was dried with MgSO4 and concentrated.
The residue was flash-chromatographed with silica gel (dichloromethane/
ethyl acetate 15:1, Et3N, 0.1%), to give 5 (1.39 g, 56%), which was en-
gaged directly to the next reaction.
4246
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4239 – 4249

Mimicking Chitin
FULL PAPER
H-1a, H-1b, H-1c, H-1d), 4.79 (m, 1H; H-3c), 4.57, 4.53, 4.49 (3dd, 3H,
pH 5.7). Spectra in H2O 85/D2O 15% (0.5 mL) were also recorded to
look at the exchangeable amide protons.
J
2a,3a =9.07, J3a,4a =10.3, J2b,3b =9.07, J3b,4b =10.39, J2d,3d =9.12, J3d,4d =
10.21 Hz; H-3a, H-3b, H-3d), 4.33 (dd, 1H, J5c,6c =4.74, J6c,6’c =10.46 Hz;
H-6c), 4.18, 4.13 (3dd, 3H, J5a,6a =4.72, J6a,6’a =10.3, J5b,6b =5.34, J6b,6’b =9.4,
2D-NOESY and TOCSY spectra were obtained by using the standard
pulse sequences provided by the manufacturer in different spectrometers:
BRUKER AVANCE spectrometer operating at a frequency of 800 MHz
J
5d,6d =5.34, J6d,6’d =9.4 Hz; H-6a, H-6b, H-6d), 4.07-3.94 (m, 5H; H-6e, H-
2a, H-2b, H-2c, H-2e), 3.90 (dd, 1H, J1d,2d =8.37, J2d,3d =10.37 Hz; H-2d),
3.79 (t, 1H, J=10.11 Hz; H-6’c), 3.70-3.43 (m, 10H; H-6’a, H-6’b, H-6’d,
H-6’e, H-4a, H-4b, H-4c, H-4d, H-4e, H-5c), 3.30 (ddd, 2H; H-5d, H-5e),
3.28 (s, 3H; OCH3), 3.25-3.15 (m, 2H; H-5a, H-5b), 1.70 ppm (s, 3H;
OAc); 13C NMR (100.6 MHz, CDCl3): d=169.84 (C=O, Ac), 167.18,
167.15, 167.01 (C=O, NPhth), 137.11, 137.08, 137.04, 137.02, 136.79,
131.26, 130.81, 130.72 (arom. C), 133.89, 133.66, 133.62, 133.46, 129.01,
128.96, 128.89, 128.87, 128.14, 128.10, 128.07, 126.14, 125.96, 125.91,
125.88 (arom. CH), 101.41, 101.03, 100.95, 100.87 (5ꢆPhCH), 99.60,
97.32, 96.78 (C-1a, C-1b, C-1c, C-1d, C-1e), 79.75, 79.72, 79.32, 79.28,
78.67 (C-4a, C-4b, C-4c, C-4d, C-4e), 74.09, 73.71, 73.15, 73.06, 69.59 (C-
3a, C-3b, C-3c, C-3d, C-3e), 68.43 (C-6a, C-6b, C-6c, C-6d, C-6e), 66.20,
65.99, 65.86, 65.75, 65.63 (C-5a, C-5b, C-5c, C-5d, C-5e), 56.79 (OCH3),
55.63, 55.53, 55.49, 55.42, 55.34 (C-2a, C-2b, C-2c, C-2d, C-2e), 20.24 ppm
(CH3C=O); HRMS (FAB+): m/z: calcd for C108H91O32N5Na: 1992.5545
[M+Na]+; found: 1992.5571.
and
a VARIAN NMR spectrometer operating at a frequency of
900 MHz. NOESY spectra were collected with mixing times ranging be-
tween 60 and 300 ms at 278, 288, 298 and 308 K. For DOSY experiments,
the samples were prepared in D2O and the standard BRUKER DOSY
protocol was used at 298 K on an AVANCE 500 MHz equipped with a
broad-band z-gradient probe. Thirty-two 1D 1H spectra were collected
with a gradient duration of d=2 ms and an echo delay of D=100 ms. Ac-
quisition times of 8—15 min (8–16 scans) were required for the samples.
The ledbpg2s pulse sequence, with stimulated echo, longitudinal eddy
current compensation, bipolar gradient pulses, and two spoil gradients,
was run with a linear gradient (53.5 Gcmꢀ1) stepped between 2 and 95%.
The 1D 1H spectra were processed and automatically baseline corrected.
The diffusion dimension, zero-filled to 1k, was exponentially fitted ac-
cording to preset windows for the diffusion dimension (ꢀ8.5<logD<
ꢀ10.0).
Interaction studies with WGA lectin: Commercial WGA was purchased
from Sigma. The binding of the pentasaccharide was evaluated by STD
experiments performed with 20:1, 40:1, and 100:1 molar ratios of the
sugar/WGA mixture. The concentration of the protein was ca. 150 mm. A
series of Gaussian-shaped pulses of 50 ms each was employed with a
total saturation time for the protein envelope of 2 s and a maximum B1
field strength of 60 Hz. An off-resonance frequency of d=40 ppm and
on-resonance frequency of d=ꢀ1.0 ppm (protein aliphatic signals region)
were applied.
Methyl (3-O-acetyl-2-deoxy-2-phthalimido-b-d-glucopyranosyl)-(1!3)-
(2-deoxy-2-phthalimido-b-d-glucopyranosyl)-(1!3)-(2-deoxy-2-phthali-
mido-b-d-glucopyranosyl)-(1!3)-(2-deoxy-2-phthalimido-b-d-glucopyra-
nosyl)-(1!3)-2-deoxy-2-phthalimido-b-d-glucopyranoside (10-1) and
methyl (2-deoxy-2-phthalimido-b-d-glucopyranosyl)-(1!3)-(2-deoxy-2-
phthalimido-b-d-glucopyranosyl)-(1!3)-(2-deoxy-2-phthalimido-b-d-glu-
copyranosyl)-(1!3)-(2-deoxy-2-phthalimido-b-d-glucopyranosyl)-(1!3)-
2-deoxy-2-phthalimido-b-d-glucopyranoside (10-2): p-Toluenesulfonic
acid (2.3 mg) was added to
a solution of compound 9 (235 mg,
The exchange transferred NOE experiments (trNOE) were performed by
using regular 2D-NOESY experiments. Measurements were done with a
freshly prepared ligand/lectin mixture, with mixing times of 75 and
150 ms, by using a 20:1 molar ratio of ligand/protein. A concentration of
1 mm of the ligand was employed. No purging spin-lock period was em-
ployed to remove the NMR spectroscopic signals of the macromolecule
background. Strong negative NOE cross-peaks were observed, in contrast
to the free state, which indicates binding of the sugars to the lectin prepa-
ration.
0.12 mmol) in THF (2.8 mL) and methanol (11.9 mL), the mixture was
then stirred and heated at 558C for 24 h. After neutralization with satu-
rated aqueous NaHCO3 solution, the mixture was dried over MgSO4 and
concentrated. The residue was flash-chromatographed by using a column
of silica gel (dichloromethane/methanol 7:1). Compound 10-1 eluted first
(122 mg, 66%; MS (FAB+): m/z: C73H71O32N5Na: 1552.32 [M+Na]+) fol-
lowed by compound 10-2, which eluted second (38 mg, 21%; MS
ACHTUNGTRENNUNG
(FAB+): m/z: C71H69O31N5Na: 1510.60 [M+Na]+.
Methyl (2-acetamido-2-deoxy-b-d-glucopyranosyl)-(1!3)-(2-acetamido-
2-deoxy-b-d-glucopyranosyl)-(1!3)-(2-acetamido-2-deoxy-b-d-glucopyra-
nosyl)-(1!3)-(2-acetamido-2-deoxy-b-d-glucopyranosyl)-(1!3)-2-acet-
amido-2-deoxy-b-d-glucopyranoside (1): Hydrazine hydrate (0.25 mL)
and water (0.25 mL) were added to a stirred solution of compound 10-1
(20 mg, 0.013 mmol) in ethanol (4.2 mL), and the mixture was refluxed
for 12 h at 808C. The reaction mixture was concentrated and the residue
was used in the next reaction directly.
Interaction studies with the chitinase: Commercial chitinase (from Strep-
tomyces Grisues) was purchased from Sigma (C6137-5UN), and its inter-
action with pentasaccharide 1 was monitored by 1D 1H NMR spectro-
scopic experiments. NMR spectra were performed in deuterated buffer
phosphate (50 mm, pH 6) at 298 K, the concentration of the pentasacchar-
ide was 1 mm, and the ratio between the sugar and enzyme was 20:1.
1H NMR spectroscopic experiments were recorded for 2 h (for the initial
30 min, the 1H NMR spectroscopic experiments were recorded every
2 min; then at regular intervals of 5 min for 1.30 h) after the addition of
the enzyme. One additional experiment was carried out 24 h later. STD
experiments were recorded on the same samples, a series of Gaussian-
shaped pulsed of 50 ms each was employed with a total saturation time
for the protein envelope of 2 s and a maximum B1 field strength of
60 Hz. An off-resonance frequency of d=40 ppm and an on resonance
frequency of d=ꢀ1.0 ppm (protein aliphatic signals region) were applied.
The DOSY experiments were performed with the standard BRUKER
DOSY protocol at 298 K on an AVANCE 600 MHz, equipped with a
broad-band z-gradient probe. No change in the diffusion coefficient of
the sample was observed after 24 h.
Water (0.4 mL), methanol (3.6 mL), and NaHCO3 (300 mg, 7.5 equiv)
were added to the residue obtained above. The mixture was stirred at
08C and acetic anhydride (1 mL) was added dropwise. After stirring at
room temperature for 4 h, the mixture was evaporated under reduced
pressure. The residue was chromatographed by a column of Sephadex
G25 (distilled water) to give the pentasaccharide 1 (6.5 mg, 48% for two
steps) as an amorphous white powder. Rf =0.21 (ethyl acetate/isopropa-
nol/H2O 1:1:1); [a]D =ꢀ10 (c=0.8 in H2O); 1H NMR (400 MHz, D2O):
d=4.57, 4.56, 4.54, 4.52, 4.35 (5d, 5H, J=8.2, 7.2, 8.4, 8.7, 8.3 Hz; H-1a,
H-1b, H-1c, H-1d, H-1e), 3.47 (s, 3H; OCH3), 2.05, 2.02, 2.02, 2.01,
1.99 ppm (5s, 15H; CH3, 5ꢆAc); 13C NMR (100.6 MHz, D2O): d=
174.88, 174.52, 174.25, 174.22, 174.15 (C=O, 5ꢆAc), 102.57, 101.34,
101.19, 100.92, 100.89 (5ꢆC-1), 81.22, 80.77, 80.36, 80.17, 76.12, 75.73,
75.63, 75.58, 75.56, 73.55, 70.15, 68.86, 68.77, 68.67, 68.66 (5ꢆC-3, 5ꢆC-4,
5ꢆC-5), 61.12, 61.00, 60.94 (5ꢆC-6), 57.55 (OMe), 56.02, 55.23, 55.22,
55.01, 54.83 (5ꢆC-2), 22.86, 22.84, 22.76, 22.73, 22.63 ppm (CH3, 5ꢆAc);
MS (FAB+): m/z: 1086.37 [M+K]+; HRMS (ESI+): m/z: calcd for
C41H69O26N5Na: 1070.4128 [M+Na]+; found: 1070.4175.
Molecular dynamics simulations
In vacuum: As the first step, pentasaccharide 1 was built by setting all
the F and Y angles of every glycosidic linkage to 60:08. Then, the result-
ing geometry was extensively minimized by using conjugate gradients
and then taken as the starting structure for the MD simulations by using
AMBER* and MM3* force fields. The desired temperature (300 K) was
obtained by raising the temperature from 0 to 300 K in 10 K increments
every ps. This heating period was followed by a 170 ps equilibration
period and a 10 ns trajectory. The temperature was controlled during the
equilibration and simulation periods by coupling to a temperature bath,
with an exponential decay constant of 0.1 ps. During the equilibration
By using the same procedure, compound 10-2 was deprotected and N-
acetylated to also afford the target 1.
NMR spectroscopy: For the NMR spectroscopic measurements, the pen-
tasaccharide (~2 mg) was dissolved in D2O (0.5 mL, phosphate buffer,
Chem. Eur. J. 2010, 16, 4239 – 4249
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4247

J. Jimꢀnez-Barbero, Y. Zhang et al.
period, the velocities were scaled when the difference between the actual
and the required temperature was higher than 108. Trajectory frames
were saved every 5 ps.
Acknowledgements
This work was partially supported by the Programme franco-chinois de
Recherches Avancꢀes (PRA B07-06). Y. Yao thanks the French Embassy
in China for a Ph.D fellowship. We also thank the EU for financial sup-
port (MRTN CT2006-035546 and MICINN; Spain) and for grants
CTQ2006-10874-C02-01 and CTQ2009-8536.
In explicit solvent: A 10 ns MD simulation was carried out by using the
AMBER 9 package.[29] Initial structures were built by using Sybyl 7.3 and
their initial coordinates were based on geometries taken from the previ-
ous results obtained in vacuum. Partial atomic charges were obtained by
using the restricted electrostatic potential (RESP) method:[30] for this
purpose all molecules were first subjected to a single-point calculation
with the HF/6-31G* basis set by using Gaussian 03.[31] Glycam 04 atomic
types were assigned to the carbohydrate moiety.[32] Compound 1 was
placed in a 10 ꢄ depth truncated octahedral box of explicit TIP3P
waters. The equilibration phase consisted of energy minimization of the
solvent followed by an energy minimization of the entire system without
restraints. The system was then heated up to 288 K during 100 ps, fol-
lowed by 100 ps at constant temperature and constant pressure of 1 atm.
The unrestrained MD simulation was continued during 10 ns under a
constant pressure of 1 atm. and constant temperature of 288 K controlled
by the Langevin thermostat with a collision frequency of 1.0 psꢀ1. During
the simulation, the SHAKE algorithm[33] was applied to all hydrogen
atoms. A cut-off of 10 ꢄ for all nonbonded interactions was adopted. An
integration time step of 2 fs was employed and periodic boundary condi-
tions were applied throughout. The particle mesh Ewald (PME) method
was used to compute long-range electrostatic interactions.[34] Minimiza-
tion, equilibration, and production phases were carried out by the
SANDER module, whereas the analyses of the simulations were per-
formed by using the Ptraj module of AMBER 9. The visualization of the
trajectories was performed by using VMD software. Data processing and
a 2D plot were created by using Scilab and Sigmaplot software.
[1] a) G. Felix, M. Regenass, Plant J. 1993, 4, 307–316; b) R. F. Dalisay,
J. A. Kuc, Physiol. Mol. Plant Pathol. 1995, 47, 315–327; c) T. Yama-
guchi, Y. Ito, Trends Glycosci. Glycotechnol. 2000, 12, 113–120;
d) P. G. Patrick, R. M. S. Helmi, P. S. Herman, Curr. Opin. Struct.
Biol. 2001, 11, 608–616; e) L. L. Walling, Trends Plant Sci. 2001, 6,
445–447; f) N. Ben-Shalom, C. Aki, Isr. J. Plant Sci. 2002, 50, 199–
206; g) Z.-Y. Zhu, Y.-M. Zhang, T. XU, Acta Phytopathol. Sin. 2004,
34, 231–236.
[2] a) D. Roby, A. Gadelle, A. Toppan, Biochem. Biophys. Res.
Commun. 1987, 143, 885–892; b) R. B. Day, M. Okada, Y. Ito, Plant
Physiol. 2001, 126, 1162–1173; c) K. Tsukada, M. Ishizaka, Y. Fuji-
sawa, Y. Iwasaki, T. Yamaguchi, E. Minami, N. Shibuya, Physiol.
Plant. 2002, 116, 373–382.
[3] P. Vander, K. M. Varum, A. Domard, Plant Physiol. 1998, 118,
1353–1359.
[4] D. Zhang, M. Hrmova, Plant Mol. Biol. 2004, 54, 353–372.
[5] P. P. G. van der Holst, H. R. M. Schlaman, Curr. Opin. Struct. Biol.
2001, 11, 608–616.
[6] N. Ben-Shalom, N. Kudabeava, R. Pinto, Isr. J. Plant Sci. 2002, 50,
259–263.
Docking calculations: The major conformer of the pentasaccharide in the
free state (as deduced by the combined NMR spectroscopic and molecu-
lar modeling approach) was docked into the carbohydrate binding sites
of WGA (PDB code 2UVO). Indeed, this conformer was used as input
geometry for the docking calculations with AutoDock 3.0[26] and Glide.[27]
The ligand charges were those estimated above. For the docking studies
with AutoDock, the multiple Lamarckian Genetic Algorithm was chosen.
Only local searches were performed centered in the experimental chitin-
specific binding sites. Grids of probe atom interaction energies and elec-
trostatic potential were generated by the AutoGrid program present in
AutoDock 3.0. A grid spacing of 0.375 ꢄ was used for the local search.
For each calculation, 100 docking runs were performed by using a popu-
lation of 200 individuals and an energy evaluation number of 3ꢆ106. For
the Glide-based analysis, the protein structure was prepared by using the
Maestro 8.5 protein preparation wizard (Schrodinger, LLC, 2008, New
York, NY); water molecules were deleted, bond orders assigned, and hy-
drogen atoms added. Next, the orientation of hydroxyl groups, amide
groups of Asn and Gln, and the charge state of the His residues were op-
timized. Finally, a restrained minimization of the protein structure was
performed by using the default constraint of 0.30 ꢄ RMSD and the
OPLS 2001 force field. The prepared protein structure was used for the
subsequent docking calculations. A grid box of default size (20ꢆ20ꢆ
20 ꢄ3) was centered on the key binding site defined by Tyr73. Default pa-
rameters were used and no constraints were included during grid genera-
tion. The extra-precision (XP) docking protocol was employed.[35] Three
docking solutions were found by using this protocol with GlideScores be-
tween ꢀ3.34 and ꢀ7.17 kcalmolꢀ1. For comparison purposes, in the Auto-
Dock runs, four docking solutions were found, within a 15 kJmolꢀ1
energy threshold. Nevertheless, in both AutoDock and Glide protocols,
the best solution showed the typical interactions for the terminal GlcNAc
moiety in hevein domains.
[7] Y.-P. Yao, Z.-Y. Zhu, T. Xu, Y.-M. Zhang, Chem. J. Chinese Univer-
sities 2007, 28, 265–269.
[8] a) J. O. Kihlberg, D. A. Leigh, D. Bundle, J. Org. Chem. 1990, 55,
2860–2863; b) R. Liang, L. Yan, J. Loebach, Science 1996, 274,
1520–1522.
[9] R. K. Jain, K. L. Matta, Carbohydr. Res. 1992, 226, 91–100.
[10] NMR Spectroscopy of Glycoconjugates (Eds: J. Jimꢀnez-Barbero, T.
Peters), Wiley-VCH, Weinheim, 2002.
[11] P. Groves, M. Rasmussen, M. D. Molero, E. Samain, F. J. CaÇada ,
H. Driguez, J. Jimꢀnez-Barbero, Glycobiology 2004, 14, 451–456.
[12] D. A. Case, T. E. Cheatham III, T. Darden, H. Gohlke, R. Luo,
K. M., Jr. Merz, A. Onufriev, C. Simmerling, B. Wang, R. J. Woods,
J. Comput. Chem. 2005, 26, 1668-.
[13] N. L. Allinger, Y. H. Yuh, J. H. Lii, J. Am. Chem. Soc. 1989, 111,
8551–8566.
[14] Maestro, A powerful, all-purpose molecular-modeling environment,
Schrodinger LLC, 2005.
[15] H. J. C. Berendsen, J. P. M. Postma, W. F. van Gunsteren, A.
Di Nola, J. R. J. Haak, Chem. Phys. 1984, 83–91, 3684–3690.
[16] J. L. Asensio, F. J. Canada, A. Garcia-Herrero, M. T. Murillo, A.
Fernꢇndez-Mayoralas, B. A. Johns, K. Janusz, Z. Zhu, C. R. Johnson,
J. Jimꢀnez-Barbero, J. Am. Chem. Soc. 1999, 121, 11318–11329.
[17] K. Bock, J. O. Duus, J. Carbohydr. Chem. 1994, 13, 513–543.
[18] K. N. Kirschner, R. J. Woods, Proc. Natl. Acad. Sci. USA 2001, 98,
10541–10545.
[19] J. L. Asensio, F. J. Canada, H. C. Siebert, J. Laynez, A. Poveda, P. M.
Nieto, U. M. Soedjanaamadja, H. J. Gabius, J. Jimenez-Barbero,
Chem. Biol. 2000, 7, 529–543.
[20] See, for instance: J. F. Espinosa, J. L. Asensio, J. L. Garcia, J.
Laynez, M. Bruix, C. Wright, H. C. Siebert, H. J. Gabius, F. J.
Canada, J. Jimenez-Barbero, Eur. J. Biochem. 2000, 267, 3965–3978.
[21] As a recent example, see, for instance: M. Kolympadi, M. Fontanel-
la, C. Venturi, S. Andrꢀ, H.-J. Gabius, J. Jimꢀnez-Barbero, P. Vogel,
Chem. Eur. J. 2009, 15, 2861–2873.
Molecular modeling: All the details of the calculations are given in the
Supporting Information.
[22] B. Meyer, T. Peters, Angew. Chem. 2003, 115, 890–918; Angew.
Chem. Int. Ed. 2003, 42, 864–890.
[23] M. Mayer, B. Meyer, J. Am. Chem. Soc. 2001, 123, 6108–6117.
[24] For a typical application of trnOE experiments in the carbohydrate
field, see, for instance: J. L. Asensio, J. F. Espinosa, H. Dietrich, F. J.
CaÇada, R. R. Schmidt, M. Martꢂn-Lomas, S. Andre, H. J. Gabius, J.
4248
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4239 – 4249

Mimicking Chitin
FULL PAPER
Jimꢀnez-Barbero, J Am Chem. Soc. 1999, 121, 8995–9000; for fur-
ther recent examples, see: a) R. S. Houliston, N. Yuki, T. Hirama,
N. H. Khieu, J.-R. Brisson, M. Gilbert, H. C. Jarrell, Biochemistry
2007, 46, 36–44; b) T. Haselhorst, H. Blanchard, M. Frank, M. J.
Kraschnefski, M. J. Kiefel, A. J. Szyczew, J. C. Dyason, F. Fleming,
G. Holloway, B. S. Coulson, M. von Itzstein, Glycobiology 2006, 17,
68–81.
[30] C. I. Bayly, P. Cieplak, W. D. Cornell, P. A. Kollman, J. Phys. Chem.
1993, 97, 10269–10280.
[31] Gaussian 98, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuse-
ria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Mont-
gomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q.
Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W.
Chen, M. W. Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle,
J. A. Pople, Gaussian, Inc., Pittsburgh, PA, 1998.
[25] H. C. Siebert, C. W. von der Lieth, R. Kaptein, J. J. Beintema, K.
Dijkstra, N. van Nuland, U. M. S. Soedjanaatmadja, A. Rice, J. F. G.
Vliegenthart, C. S. Wright, H. J. Gabius, Proteins 1997, 28, 268–284.
[26] G. M. Morris, D. S. Goodsell, R. S. Halliday, R. Huey, W. E. Hart,
R. K. Belew, A. J. Olson, J. Comput. Chem. 1998, 19, 1639–1647.
[27] a) R. A. Friesner, J. L. Banks, R. B. Murphy, T. A. Halgren, J. J.
Klicic, D. T. Mainz, M. P. Repasky, E. H. Knoll, M. Shelley, J. K.
Perry, D. E. Shaw, P. Francis, P. S. Shenkin, J. Med. Chem. 2004, 47,
1739–1749; b) T. A. Halgren, R. B. Murphy R. A. Friesner, H. S.
Beard, L. L. Frye, W. T. Pollard, J. L. Banks, J. Med. Chem. 2004, 47,
1750–1759.
[32] K. N. Kirschner, A. B. Yongye, S. M. Tschampel, J. Gonzꢇlez-Outeir-
iÇo, C. R. Daniels, B. L. Foley, R. J. Woods, J. Comput. Chem. 2008,
29, 622–655.
[28] J. Jimenez-Barbero, F. J. Canada, J. L. Asensio, N. Aboitiz, P. Vidal,
A. Canales, P. Groves, H. J. Gabius, H. C. Siebert, Adv. Carbohydr.
Chem. Biochem. 2006, 60, 303–354.
[29] D. A. Case, T. A. Darden, T. E. Cheatham III, C. L. Simmerling, J.
Wang, R. E. Duke, R. Luo, K. M. Merz, Jr. D. A. Pearlman, M.
Crowley, R. C. Walker, W. Zhang, B. Wang, S. Hayik, A. E. Roit-
berg, G. Seabra, K. F. Wong, F. Paesani, X. Wu, S. Brozell, V. Tsui,
H. Gohlke, L. Yang, C. Tan, J. Mongan, V. Hornak, G. Cui, P.
Beroza, D. H. Mathews, C. E. A. F. Schafmeister, W. S. Ross, P. A.
Kollman, AMBER 9, University of California, San Francisco, 2006.
[33] J. P. Ryckaert, G. Ciccoti, H. J. C. Berendsen, J. Comput. Phys. 1977,
23, 327–341.
[34] T. A. Darden, D. York, L. G. Pedersen, J. Chem. Phys. 1993, 98,
10089–10092.
[35] R. A. Friesner, R. B. Murphy, M. P. Repasky, L. L. Frye, J. R. Green-
wood, T. A. Halgren, P. C. Sanschagrin, D. T. Mainz, J. Med. Chem.
2006, 49, 6177–6196.
Received: October 16, 2009
Published online: March 12, 2010
Chem. Eur. J. 2010, 16, 4239 – 4249
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4249