Cyclohexane ring as a tool to select the presentation of the carbohydrate moiety in glycosyl amino acids

Article abstract of DOI:10.1002/chem.201103089

The design of mimic molecules that resemble natural products can be a useful tool to help understand the key aspects in molecular recognition processes that are difficult to access by using natural derivatives. We present the synthesis and the conformatio

Full text of DOI:10.1002/chem.201103089

DOI: 10.1002/chem.201103089
Cyclohexane Ring as a Tool to Select the Presentation of the Carbohydrate
Moiety in Glycosyl Amino Acids
Fernando Rodrꢀguez, Vꢀctor J. Somovilla, Francisco Corzana,* Jesffls H. Busto,
Alberto Avenoza, and Prof Jesffls M. Peregrina*[a]
Abstract: The design of mimic mole-
cules that resemble natural products
can be a useful tool to help understand
the key aspects in molecular recogni-
tion processes that are difficult to
access by using natural derivatives. We
present the synthesis and the confor-
mational analysis of different glucosy-
lated diamide amino acids that simu-
late glycopeptides with b-O-linked glu-
cose and contain the nonnatural b-hy-
droxycyclohexane-a-amino acid. The
study, using NMR experiments, X-ray
spectroscopy, and molecular dynamics
simulations, reveals that the cyclohex-
ane ring allows some naturally occur-
ring ways of presentation of the carbo-
hydrate to be fixed, or to stabilize
some novel conformations. In addition,
different chair conformations for the
cyclohexane-a-amino acid moiety can
be set, in particular, those with high
population of conformers in which the
bulky groups are located at axial posi-
tions. Moreover, to increase the scope
of these cyclohexane derivatives, two
dipeptides incorporating the glycomi-
mics have been synthesized and further
glycosylated to obtain the correspond-
ing a-O-glycopeptides. These features
can have important implications for the
design of new drugs and for under-
standing the complex molecular pro-
cesses that take place between glyco-
peptides and their biological targets.
Keywords: conformation analysis ·
glycopeptides · molecular modeling ·
molecular recognition · NMR spec-
troscopy
Introduction
A large number of proteins undergo post-translational
modification by glycosylation of amino acids such as serine
(Ser) and threonine (Thr). It is well-known that this glycosy-
lation influences some properties of the corresponding pro-
teins such as solubility, stability toward chemical and enzy-
matic degradation, as well as conformation and folding. It
can also affect the biological functions of proteins in events
of molecular recognition, for example, cell-to-cell communi-
cation and adhesion of bacteria or viruses to cell-surface
proteins.[1] In particular, b-O-linked attachment of d-glucose
(Glc) to Ser/Thr has been found in the epidermal growth
factor (EGF) domains of different serum proteins and in
Notch receptor (see Figure 1).[2] The role of the Glc moiety
in these systems is unknown and remains controversial.
Figure 1. Crystal structure of the complex of the blood coagulation factor
VII with a soluble tissue factor (1DAN pdb code)[2e] showing the d-glu-
cose residue attached to a serine of the coagulation factor.
From a structural viewpoint, both the spatial disposition
of the sugar moiety with respect to the amino acid and the
properties of the underlying amino acid are crucial for suc-
cessful interaction with the biological targets. In this regard,
we have recently reported the different sugar presentation
showed by Ser and Thr glycopeptides[3] and its influence on
the binding to a synthetically prepared lectin-like receptor.[4]
In addition, it is necessary to study in detail the interac-
tions between the sugar and the underlying amino acid be-
cause they play a crucial role in the presentation of the car-
bohydrate moiety and, consequently, in the interactions with
the corresponding biological receptors.
[a] Dr. F. Rodrꢀguez, V. J. Somovilla, Dr. F. Corzana, Dr. J. H. Busto,
Prof. A. Avenoza, P. J. M. Peregrina
Departamento de Quꢀmica
Universidad de La Rioja
Centro de Investigaciꢁn en Sꢀntesis Quꢀmica
UA-CSI, Madre de Dios, 51
26006 LogroÇo (Spain)
Fax : (+34)941-299-621
E-mail: jesusmanuel.peregrina@unirioja.es
francisco.corzana@unirioja.es
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103089.
5096
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5096 – 5104

FULL PAPER
Our group has been involved in the design of new glyco-
peptides with well-defined conformational preferences.[4,5] In
general, our studies conclude that whereas the substituents
at carbon Ca affect mainly the conformation of the back-
bone—forcing it to adopt a helix-like conformation—those
present at Cb modify the behavior of the glycosidic linkage,
particularly the ys dihedral angle, forcing it to adopt an
eclipsed conformation. Moreover, these substituents and, in
particular, their spatial dispositions, have an important
effect on the flexibility of the lateral chain. As a conse-
quence, these custom-made molecules can stabilize confor-
mations present in natural occurring molecules or exhibit
some atypical conformations. This feature should be a power-
ful tool that could be used to design and modulate the con-
formational space of synthesized glycopeptides and could be
useful in determining the structural elements of the natural
occurring glycopeptides that are necessary for binding with
their biological targets.
ACHTUNGTRENNUNG
With the intention of expanding this outlook to novel sys-
tems, we herein report the synthesis and conformational
analysis in aqueous solution of the four glucosylated dia-
mide amino acids shown in Figure 2, by combining NMR ex-
Figure 3. X-ray crystal structure of compound 9, together with atom la-
beling and definition of dihedral angles used in this work.
Compounds 3 and 4 were obtained following an identical
protocol (Scheme 1). In this case, we could not obtain a crys-
tal structure for any of the intermediates, and the absolute
configuration of the glucosylated diamide amino acids was
determined by starting the synthesis from the enantiomeri-
cally pure (1R,2R)-1-amino-2-hydroxycyclohexanecarboxylic
acid hydrochloride (1R,2R)-10, which led to compound 4 via
14.
Figure 2. Glucosylated diamide amino acids studied in this work.
periments with molecular dynamics (MD) simulations. In
these molecules, Ser or Thr moieties have been replaced by
all the possible stereoisomers of the nonnatural b-hydroxy-
cyclohexane-a-amino acid (c6Ser)[6] and the amine and car-
boxylic acid groups were transformed into amides to simu-
late a peptide backbone. Additionally, a pivaloyl group was
employed, which favors the formation of monocrystals and
makes the manipulation of the new compounds easier. It is
important to note that the substructure b-O-Glc linked to
a cyclohexane ring is found in steroidal saponins, which pos-
sess anti-inflammatory, hemolytic, cytotoxic, antifungal, and
antibacterial properties.[7]
Conformational study: Relevant torsion angles and labels of
the atoms used in this work for all the compounds are
shown in Figure 3.
In a first step, full assignment of the protons in all of the
compounds was carried out using COSY and HSQC experi-
ments. Selective 1D NOESY experiments in D2O, and 2D
NOESY experiments in H2O/D2O (9:1) were then carried
out for all compounds (see Figure 4 and the Supporting In-
formation).
The presence of medium NOE cross-peaks between the
NH protons of the backbone indicates that the helix-like
conformations are significantly populated in the four deriva-
tives.[9] In addition, the strong NOE signal between the
anomeric proton H1s and H2 implies the predominance of
the typical syn conformation [angle C1s-O1s-C2-H2 close to
08] for the ys glycosidic linkage for compounds 1–4.[5d,10]
The next step was to generate a theoretical ensemble that
could reproduce our experimental NMR data. To this end,
Results and Discussion
Synthesis: The synthesis of the final glucosylated diamide
amino acids 1–4 (Scheme 1) started from the hydrochloride
salts of the corresponding 1-amino-2-hydroxycyclohexane-
carboxylic acids previously prepared by our group.[6] Thus,
treatment of rac-5 with pivaloyl chloride and diisopropyl-
Chem. Eur. J. 2012, 18, 5096 – 5104
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5097

F. Corzana, J. M. Peregrina et al.
Scheme 1. Synthetic route to glucosylated diamide amino acids 1–4: Reagents and conditions: a) PivCl, DIEA, CH2Cl2, 08C to RT, 16 h, 42% (rac-11)
and 46% (rac-6). b) MeNH2·HCl, DIEA, CH3CN, 708C, 24 h, 59% (rac-7) and 61% (rac-12); c) 2,3,4,6-tetra-O-benzoyl-a-d-glucopyranosyl bromide,
AgTfO, CH2Cl2, À30 to 258C, molecular sieves 4 ꢃ, 16 h, 36% (8+9) and 14 h, 19% (13+14); d) MeONa/MeOH, 258C, 3 h, 82% (1), 80% (2), 83%
(3) and 80% (4).
we used our previously developed protocol,[5,10,11] which
combines NMR data with time-averaged restrained MD
simulations (MD-tar).[12] Proton–proton distances were ex-
perimentally determined from the corresponding NOE
build-up curves[13] (see the Supporting Information) and dis-
tances involving NH protons were semi-quantitatively deter-
mined by integrating the volume of the corresponding cross-
peaks. These data were then used as restraints in MD-tar
simulations. The distribution for the peptide backbone (fp/
yp), obtained from the MD-tar simulations, is shown in
Table 1.
According to the NMR data described above, the four
glucosylated diamide amino acids showed a significant popu-
lation of conformers with fp/yp dihedral values that are
characteristic of helix-like conformations. However, some
important differences can be established among them. Thus,
whereas for 1 the MD simulations suggest the existence of
a similar population for the two possible helix-like confor-
mations, the aL (left-handed helix-like) conformer was the
most populated in derivatives 2 and 4. The major conforma-
tion of 2 in aqueous solutions is the same as that observed
in the solid state for its precursor (9, Figure 3). In this case,
it is important to note that the crystal structure of 9 showed
an inverse g-turn conformation in the solid state, stabilized
by a hydrogen bond between the PivC=O···NHMe groups.
In contrast, derivative 3 showed both a PPII (polyproline II)
and a helix-like conformation for the peptide backbone.
Moreover, whereas derivatives with R configuration at Ca
(compounds 2 and 4) tend to adopt the aL conformation, de-
rivatives 1 and 3 exhibit the aD (right-handed helix-like)
conformer. It is important to mention that these major con-
Table 1. Population of the major conformers obtained from the MD-tar
simulations for the peptide backbone of glucosylated diamide amino
acids 1–4.
Compound
aD [%]
aL [%]
PPII [%]
1
2
3
4
48
5
54
6
47
89
3
3
4
42
3
91
5098
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5096 – 5104

Presentation of Glycosyl Amino Acids
FULL PAPER
Figure 5. Distribution of the glycosidic linkage (fs/ys) of the four glucosy-
lated diamide amino acids obtained from the MD-tar simulations:
a) Compound 1, b) Compound 2, c) Compound 3; d) Compound 4. For
compound 2, the conformation of the glycosidic linkage in the solid state
of its precursor (compound 9) is shown as a square.
Figure 4. a) 1H NMR (D2O); b) 2D-NOESY (H2O/D2O, 9:1) spectra
(400 MHz), and c) selected build-up curves (D2O) for compound 2. The
experiments were run at 258C and pH 5.8.
for the six-membered ring of the amino acid moiety
(Figure 6). In all cases, chair a refers to the conformer with
the NHPiv group located at an axial position.
From an experimental viewpoint, the equilibrium between
the two possible chair conformations can be determined by
examining the distances between the H2a and H6a protons
deduced from the build-up curves.
formers found for the backbone of the glucosylated diamide
amino acids lie at the local minima previously calculated for
other a,a-disubstituted amino acids.[5d]
Concerning the glycosidic linkage, fs is mainly determined
by the exo-anomeric effect,[14] which fixes its value at
around À608 in all the derivatives. In addition, neither the
stereochemistry at Ca nor the presence of a substituent
group at Ca significantly affects the conformation of the gly-
cosidic linkage (Figure 5). In contrast, when a Thr residue is
present, the existence of a substituent group at C2 (or Cb)
notably affects the ys torsion angle, which adopts a value
close to 1208 (compounds 1 and 4) or À1208 (compounds 2
and 3). This conformation, denoted as A or Aꢄ in Figure 5,
Interestingly, in all cases, the most populated chair confor-
mations are those that locates the bulky group (NHPiv) in
an axial position (chair a), in particular for derivatives 2 and
3. This result is, to some extent, expected. Indeed, the pref-
erence for the amide or amine groups to adopt axial posi-
tions in cyclohexane-a-amino acids has been previously re-
ported for the solid state.[15] We also have some crystal struc-
tures that corroborate this experimental finding.[16] On the
other hand, the results of MD simulations are in good agree-
ment with the experimental data (see the Supporting Infor-
mation). In compound 2, with the sugar unit and the NHPiv
group in a cis relative position, the sugar occupies the equa-
torial location, which is expected to avoid the 1,3-diaxial in-
teractions. Indeed, chair a was also found in the crystal
structure of the perbenzoylated compound 9, which is its
precursor. In contrast, for 1, both chair conformations coex-
ist in aqueous solution with a similar population. Notably, in
3 and 4 the major populated chair conformation (chair a) lo-
cates the bulky groups (NHPiv and sugar) in axial positions.
In this regard, there are some exceptions to the well-estab-
À
À
displays H2 C2 and O1s C1s bonds in an eclipsed confor-
mation to avoid a nonstabilizing interaction between the cy-
clohexane ring of the amino acid and the endocyclic oxygen
of the carbohydrate moiety. In addition, when C2 adopts the
(S) configuration (compounds 2 and 3), the glycosidic link-
age becomes more flexible, and a small population of an ad-
ditional conformer (denoted as B in Figure 5) with ys close
to À1808 and fs around À1208 is also evident. This confor-
mation has also been observed in other nonnatural glycosy-
lated diamide amino acids.[5d]
Inspection of the NOE contacts revealed an interesting
equilibrium between the two possible chair conformations
Chem. Eur. J. 2012, 18, 5096 – 5104
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5099

F. Corzana, J. M. Peregrina et al.
formations for derivative 1 and
g(À) and anti for compound 4.
In contrast, derivative 2 and es-
pecially compound 3, exhibited
c1 values corresponding to g(À)
and anti, respectively. In addi-
tion, as can be seen in Figure 7,
compounds
3 and 4 showed
a clear preference for the anti
conformation of the lateral
chain. An important conse-
quence of this feature is that
whereas derivatives
1 and 2
present the carbohydrate
moiety in a perpendicular dis-
position with respect to the
backbone, derivatives 3 and 4
present the sugar unit almost
parallel to the peptide sequence
(Figure 8).
Figure 6. Equilibrium observed in aqueous solution between the two possible chair conformations (denoted as
a and b) of the six-membered ring of the amino acid moiety for the studied glucosylated diamide amino acids.
lished rule that states that, in
a cyclohexane ring, the bulky
substituents prefer to adopt
equatorial positions. One ex-
ception is defined as axial/equa-
torial stability reversal.[17] For
example, the sterically crowded
all-trans-hexaalkylcyclohexane
prefers a conformation in which
the alkyl groups are located at
axial rather than equatorial po-
sitions. Moreover, increasing
the number and bulk of the
alkyl substituent results in a rel-
ative stabilization of the axial
conformer. The authors suggest
that the stability reversal is not
due to stabilization of the axial
conformer but due to destabili-
zation of the equatorial confor-
Figure 7. Lateral chain distribution (c1) of the four glucosylated diamide amino acids obtained from the MD-
tar simulations.
mer because of steric and tor-
sional contributions.
As far as the lateral chain is
concerned, its conformational
behavior is characterized by the c1 torsional angle, which
can adopt three lowest energy staggered rotamers, denoted
as g(À) (c1 ꢀÀ608), g(+) (c1 ꢀ +608) and anti (c1 ꢀ1808).
Figure 7 shows the population of c1 obtained from the MD-
tar simulations for all the glucosylated diamide amino acids
1–4. For these systems, one should expect a rigid c1 due to
the restriction imposed by the six-membered ring. The dif-
ferent c1 values observed for each molecule are related to
the equilibrium described above between the two possible
chair conformations of the amino acid ring.
In previous works, we have demonstrated that whereas
for the b-Glc-Ser derivatives the carbohydrate and the pep-
tide moieties exhibit a perpendicular disposition, it shifts to-
wards a parallel arrangement in the b-Glc-Thr compounds.
On this basis, we could state that cyclohexane derivatives
1 and 2 mimic the conformational behavior of the Ser glyco-
peptides, whereas compounds 3 and 4 imitate the behavior
of the Thr-containing glycopeptides.
To investigate whether these conformations were stabi-
lized by interactions between the peptide and sugar moiet-
ies, a hydrogen-bond analysis was carried out; however, in
all cases, very weak hydrogen bonds (less than 10% popula-
tion) were obtained between the two parts of the molecule.
Thus, for compounds 1 and 4, with a significant presence
of both chair conformations in solution, the lateral chain is
quite flexible, with two possible values: g(+) and g(À) con-
5100
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5096 – 5104

Presentation of Glycosyl Amino Acids
FULL PAPER
Figure 9. Two-dimensional radial pair distribution function for O6s and
O found for compound 1 in the 20 ns unrestrained MD simulation in ex-
plicit water.
Figure 8. Snapshots taken from the 80 ns MD-tar trajectories collected
for the different glucosylated diamide amino acids.
Therefore, taking into account the influence that the water
molecules could exhibit on the conformation of the glucosy-
lated diamide amino acids, the anisotropic hydration of the
solutes was inspected. To this end, the normalized, one-di-
mensional and two-dimensional radial pair distributions[18]
were calculated from the 20 ns unrestrained MD simulations
carried out in explicit solvent on the four compounds for all
possible shared water density sites. Notably, only for com-
pound 1 was a significant intra-residue water bridge found
between atoms O6s and the carbonyl oxygen of the underly-
ing amino acid (Figure 9). This water pocket was present for
about 50% of the total trajectory time, coinciding with the
g(+) conformation for the lateral chain, exhibits a very high
maximum density (about 7.6 times the bulk density), and
differs from that previously found for its natural analogue
(bGlc-Ser diamide).[10]
Figure 10. fs/ys/c1 values obtained from the MD-tar simulations for the
most populated conformers of the glucosylated diamide amino acids 1–4,
together with conformers previously deduced for the natural derivatives
with Ser and Thr. The configuration of the substituents at the cyclohex-
ane ring can modulate the orientation of the sugar moiety with respect to
the underlying amino acid (lower part of the molecules; shown in dark
grey). The cyclohexane ring has been removed for clarity.
amino acid are possible using the cyclohexane ring. Thus,
for example, compound 4 exhibits one of the major con-
formers occurring in the natural bGlc-Thr derivative (see
compound 4 in Figure 10) and stabilizes an unusual confor-
mation that is not observed in Nature. These geometries
Figure 10 shows the major conformations found for the
glucosylated diamide amino acids in aqueous solution stud-
ied in this work. As can be seen, several presentations of
the carbohydrate moiety with respect to the underlying
Chem. Eur. J. 2012, 18, 5096 – 5104
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5101

F. Corzana, J. M. Peregrina et al.
might well be used to enhance the binding between the car-
bohydrate and its biological target. Consequently, the di-
verse 3D disposition of the sugar could have important im-
plications for drug design. The conformers derived from
these systems could also be useful in defining the pivotal el-
ements that are necessary for molecular recognition process-
es.
To demonstrate the scope of the approach developed in
this work, we have incorporated some of the nonnatural
amino acids developed above into short peptides: (1R,2R)-
c6Ser-l-Pro and (1S,2S)-c6Ser-l-Pro. The terminal amino and
carboxylic acid groups of these dipeptides were transformed
into amides to simulate larger peptides. These dipeptides
were treated with tri-O-benzyl-2-nitro-d-galactal and potas-
sium tert-butoxide to obtain the corresponding a-anomers,
thus demonstrating that this kind of nonnatural amino acid
can be used to form both a- and b-glycosidic bonds in the
glycosylation reactions. Their syntheses and spectroscopic
data are collected in the Supporting Information. Fortunate-
ly, it was possible to obtain the crystal structure of com-
pound 15, which corresponded to PivNH-(1R,2R)-c6Ser*-l-
Pro-CONHMe, in which (*) refers to the glycan moiety indi-
cated in Figure 11. As can be seen in Figure 11, the back-
mide amino acids and a lectin that recognizes glucose. Al-
though there are different systems that recognize glucose,
such as lens culinaris lectin[19] or the E. coli glucose/galactose
chemoreceptor protein,[20] GRFT protein[21] was chosen for
this purpose because it exhibits potent antiviral activity
against human immunodeficiency virus (HIV) by binding to
viral envelope glycoproteins. Surprisingly, this high activity
can be inhibited by low concentrations of mannose, N-ace-
tylglucosamine (GlcNAc), and glucose, which makes it inter-
esting to study in detail the molecular recognition process of
these monosaccharides by GRFT.
All the simulations were started from the crystal structure
of the complex between GRFT and glucose (PDB code:
2NUO). In our case, and at least from a theoretical view-
point, all the glucosylated diamide amino acids exhibited
the same interactions with lectin (see Figure 12). This result
could be explained by the fact that O1 and O2 point away
from the protein chain and also corroborates the experimen-
tal finding that lectin recognizes, in a similar way, mannose,
GlcNAc, and glucose.
Figure 12. Details of the interactions between compound 4 and site 1 of
the antiviral protein griffithsin (GRFT), as deduced from unrestrained
20 ns MD simulations in explicit water.
Figure 11. X-ray crystal structure (left) and most populated conformers
found in solution (right) for compound 15.
bone adopts a type II’ b-turn and the glycosidic linkage ex-
hibits the typical eclipsed conformation, with ys around
1408. As also occurs for its analogue, compound 4, the cyclo-
hexane ring of 15 mainly adopts the chair a conformation,
with the bulky groups located at axial positions. Additional-
ly, the conformational behavior of 15 in CHCl3 solution was
analyzed by combining NMR data (NOE cross-peaks) and
MD-tar simulations (see the Supporting Information for de-
tails). Whereas the chair a conformation is entirely populat-
ed in solution, and the glycosidic linkage mainly shows the
eclipsed conformer, the b-turn structure of the peptide back-
bone coexisted with typical helix-like conformations forced
by the c6Ser residue in a 3:1 ratio, respectively.
Conclusion
The synthesis and conformational analysis of glycosylated
diamide amino acids with b-O-linked glucose and containing
the nonnatural b-hydroxycyclohexane-a-amino acid residue
have been investigated. The study, which combines NMR
experiments, X-ray spectroscopy, and molecular dynamics
simulations, reveals that the cyclohexane ring can effectively
modulate the presentation of the sugar moiety with respect
to the underlying amino acid residue. These features, togeth-
er with the fact that the amino acid moiety exhibits different
chair conformations for the six-membered ring, allow the
molecule to adopt novel conformations that are not found
in natural derivatives. Furthermore, two dipeptides incorpo-
rating the new derivatives have been synthesized and further
glycosylated to obtain the corresponding a-O-glycopeptides.
The development of these nonnatural derivatives could have
important implications for the design of new drugs and for
Finally, in future work, we would like to introduce all
these derivatives into larger glycopeptides with the aim of
enabling them to be recognized by biological targets. To test
the viability of this idea, MD simulations were run to inves-
tigate the interactions between these novel glucosylated dia-
5102
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5096 – 5104

Presentation of Glycosyl Amino Acids
FULL PAPER
were run using an exponential decay constant of 8 ns and a simulation
length of 80 ns.
understanding the complex molecular processes between
glycopeptides and their biological targets.
X-Ray diffraction analysis[29]: Crystal data for 9: C47H50N2O12; Mw =
834.89; colorless prism of 0.25ꢆ0.25ꢆ0.22 mm; T=173 K; monoclinic;
space group P21/c; Z=2; a=13.5848(4) ꢃ, b=11.7140(3) ꢃ, c=
14.9527(4) ꢃ, b=112.8678(10)8; V=2192.44(10) ꢃ3; dcalc =1.265 gcmÀ3
;
Experimental Section
FACTHNUTRGNEUNG
(000)=884; l=0.71073 ꢃ (MoKa); m=0.091 mmÀ1; Nonius kappa CCD
diffractometer, q range 3.43–28.158, 18857 collected reflections, 9792
unique, full-matrix least-squares (SHELXL97),[30] R1 =0.0535, wR2 =
0.1128, (R1 =0.0874, wR2 =0.1302 all data), goodness of fit=1.014, resid-
ual electron density between 0.23 and À0.245eꢃÀ3. Hydrogen atoms
fitted at theoretical positions.
General Procedures: Solvents were purified according to standard proce-
dures. Analytical TLC was performed using Polychrom SI F254 plates.
Column chromatography was performed using Silica gel 60 (230–400
mesh). 1H and 13C NMR spectra were recorded with Bruker ARX 300
and Bruker AVANCE 400 spectrometers. 1H and 13C NMR spectra were
recorded in CDCl3 with TMS as internal standard or in D2O (chemical
shifts referenced to the internal solvent signals and reported in ppm on
the d scale; coupling constants in Hz). Assignment of all separate signals
Crystal data for compound 15·Et2O: C49H68N4O11; Mw =889.08; colorless
prism of 0.27ꢆ0.15ꢆ0.12 mm; T=173 K; monoclinic; space group P21;
Z=2; a=12.2297(6) ꢃ, b=15.0023(8) ꢃ, c=13.3856(5) ꢃ, b=98.902(3)8;
V=2426.3(2) ꢃ3; dcalc =1.217 gcmÀ3; F
ACHTNUTRGNE(UNG 000)=956; l=0.71073 ꢃ (MoKa);
1
for the final compounds in the H NMR spectra was made on the basis of
m=0.086 mmÀ1; Nonius kappa CCD diffractometer, q range 2.05–28.088,
21769 collected reflections, 5820 unique, full-matrix least-squares
(SHELXL97),[30] R1 =0.0618, wR2 =0.0990, (R1 =0.1254, wR2 =0.1166 all
data), goodness of fit=1.047, residual electron density between 0.293 and
À0.179eꢃÀ3. Hydrogen atoms fitted at theoretical positions.
coupling constants, ge-COSY and ge-HSQC experiments with a Bruker
AVANCE 400 spectrometer. The NMR data were processed with Mestre
Nova software (Mestrelab Research, Spain). Melting points were deter-
mined with a Bꢅchi SMP-20 melting point apparatus. Microanalyses were
carried out with a CE Instruments EA-1110 analyzer and are in good
agreement with the calculated values. Optical rotations were measured
with a Perkin–Elmer 341 polarimeter. Electrospray mass spectra were re-
corded with a microTOF-Q-BRUKER connected to a Waters 996 photo-
diode array detector with H2O or MeOH as carrier solvents. The synthet-
ic procedures for all new compounds as well as their physical properties
are detailed in the Supporting Information.
Acknowledgements
We thank the Ministerio de Ciencia e Innovaciꢁn and FEDER (project
CTQ2009–13814/BQU and FPI grant of V. J. S.), the Universidad de La
Rioja (project EGI10/65), and the Gobierno de La Rioja (Colabora 2010/
05). F. R. thanks the CSIC for his JAE-Doc Program contract. We also
thank CESGA for computer support.
NMR experiments: NMR experiments were recorded with a Bruker
Avance 400 spectrometer at 298 K. Magnitude-mode ge-2D COSY spec-
tra were recorded with gradients and using the cosygpqf pulse program
with 90 degree pulse width. Phase-sensitive ge-2D HSQC spectra were
recorded using z-filter and selection before t1, removing the decoupling
during acquisition by use of invigpndph pulse program with CNST2
(JHC)=145. 2D NOESY experiments were made using phase-sensitive
ge-2D NOESY with WATERGATE for H2O/D2O (9:1) spectra. Selective
ge-1D NOESY experiments were carried out using the 1D-DPFGE NOE
pulse sequence. NOE intensities were normalized with respect to the di-
agonal peak at zero mixing time. Experimental NOEs were fitted to
[1] a) R. A. Dwek, Chem. Rev. 1996, 96, 683–720; b) A. Varki, Glycobi-
ology 1993, 3, 97–130; c) D. H. Williams, Nat. Prod. Rep. 1996, 13,
469–478; d) D. F. Wyss, G. Wagner, Curr. Opin. Biotechnol. 1996, 7,
409–416; e) T. Buskas, S. Ingale, G.-J. Boons, Glycobiology 2006, 16,
113R-136R; f) C. R. Bertozzi, L. L. Kiessling, Science 2001, 291,
2357–2364; g) M. R. Pratt, C. R. Bertozzi, Chem. Soc. Rev. 2005, 34,
58–68.
[2] a) L. Shao, Y. Luo, D. J. Moloney, R. S. Haltiwanger, Glycobiology
2002, 12, 763–770; b) R. J. Harris, M. W. Spellman, Glycobiology
1993, 3, 219–224; c) K. Brꢅckner, L. Perez, H. Clausen, S. Cohen,
Nature 2000, 406, 411–415; d) D. J. Moloney, L. H. Shair, F. M. Lu,
J. Xia, R. Locke, K. L. Matta, R. S. Haltiwanger, J. Biol. Chem.
2000, 275, 9604–9611; e) D. W. Banner, A. D’Arcy, C. Chene, F. K.
Winkler, A. Guha, W. H. Konigsberg, Y. Nemerson, D. Kirchhofer,
Nature, 1996, 380, 41–46.
[3] F. Corzana, J. H. Busto, G. Jimꢇnez-Osꢇs, M. Garcꢀa de Luis, J. L.
Asensio, J. Jimꢇnez-Barbero, J. M. Peregrina, A. Avenoza, J. Am.
Chem. Soc. 2007, 129, 9458–9467.
[4] F. Corzana, A. Fernꢈndez-Tejada, J. H. Busto, G. Joshi, A. P. Davis,
J. Jimꢇnez-Barbero, A. Avenoza, J. M. Peregrina, ChemBioChem
2011, 12, 110–117.
[5] a) A. Fernꢈndez-Tejada, F. Corzana, J. H. Busto, G. Jimꢇnez-Osꢇs, J.
Jimꢇnez-Barbero, A. Avenoza, J. M. Peregrina, Chem. Eur. J. 2009,
15, 7297–7301; b) A. Fernꢈndez-Tejada, F. Corzana, J. H. Busto, A.
Avenoza, J. M. Peregrina, Org. Biomol. Chem. 2009, 7, 2885–2893;
c) A. Fernꢈndez-Tejada, F. Corzana, J. H. Busto, A. Avenoza, J. M.
Peregrina, J. Org. Chem. 2009, 74, 9305–9313; d) A. Fernꢈndez-
Tejada, F. Corzana, J. H. Busto, G. Jimꢇnez-Osꢇs, J. M. Peregrina, A.
Avenoza, Chem. Eur. J. 2008, 14, 7042–7058; e) F. Corzana, J. H.
Busto, M. Garcꢀa de Luis, A. Fernꢈndez-Tejada, F. Rodrꢀguez, J. Ji-
mꢇnez-Barbero, A. Avenoza, J. M. Peregrina, Eur. J. Org. Chem.
2010, 3525–3532; f) F. Corzana, J. H. Busto, F. Marcelo, M. Garcꢀa
de Luis, J. L. Asensio, S. Martꢀn-Santamarꢀa, J. Jimꢇnez-Barbero, A.
Avenoza, J. M. Peregrina, Chem. Eur. J. 2011, 17, 3105–3110; g) F.
Corzana, J. H. Busto, F. Marcelo, M. Garcꢀa de Luis, J. L. Asensio, S.
a double exponential function, f(t)=p0 (eÀp1t
)ACTHNGUTERNNUG
being adjustable parameters.[13] The initial slope was determined from the
first derivative at time t=0, f’(0)=p0p2. From the initial slopes, interpro-
ton distances were obtained by employing the isolated spin pair approxi-
mation.
MD simulations in explicit water: Simulations were performed by using
the AMBER 9 program package (parm99),[22] which was implemented
with GLYCAM 06 parameters.[23] Unrestrained MD simulations were car-
ried out in the presence of explicit TIP3P water molecules.[24] RESP
atomic charges[25] for the glycopeptides were derived by applying the
RESP module of AMBER to the HF/6–31G(d) ESP charges calculated
with Gaussian 08.[26] The simulations were run with the PMEMD module
of AMBER with SHAKE algorithm,[27] using periodic boundary condi-
tions, a 2 fs time step, a temperature of 300 K, a Langevin type thermo-
stat for temperature control, and constant pressure of 1 atm. A 9 ꢃ cut-
off was applied to the Lennard-Jones interactions, and Ewald sums for
the treatment of the electrostatic interactions.[28] An initial 2500 cycles of
minimization (combining steepest descent with conjugate gradient) were
run on both systems by first restraining the atoms of the complex. The
whole system was then minimized using 2000 cycles. This first step was
followed by 200 ps of dynamics at constant volume with weak positional
restrains on the complex (10 kcalmolÀ1·ꢃ2). In this step, the system was
heated from 100 to 300 K. The restrains on the solute were removed and
200 ps MD simulations were run at 300 K and 1 atm to obtain the appro-
priate density. Finally, 20 ns MD simulations were run using the condi-
tions mentioned above.
MD simulations: NOE-derived distances were included as time-averaged
distance restraints, and scalar coupling constants (J) as time-averaged
À1/6
coupling restraints. A <rÀ6
>
average was used for the distances and
a linear average was used for the coupling constants. Final trajectories
Chem. Eur. J. 2012, 18, 5096 – 5104
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5103

F. Corzana, J. M. Peregrina et al.
Martꢀn-Santamarꢀa, Y. Sꢈenz, C. Torres, J. Jimꢇnez-Barbero, A. Ave-
noza, J. M. Peregrina, Chem. Commun. 2011, 47, 5319–5321.
[6] A. Avenoza, J. I. Barriobero, C. Cativiela, M. A. Fernꢈndez-Recio,
J. M. Peregrina, F. Rodrꢀguez, Tetrahedron 2001, 57, 2745–2755.
[7] a) V. L. Challinor, P. Y. Hayes, P. V. Bernhardt, W. Kitching, R. P.
Lehmann, J. J. De Voss, J. Org. Chem. 2011, 76, 7275–7280; b) S. G.
Sparg, M. E. Light, J. J. vanStaden, J. Ethnopharmacol. 2004, 94,
219–243.
[8] S. Hanessian, J. Banoub, Carbohydr. Res. 1977, 53, C13-C16.
[9] H. J. Dyson, P. E. Wright, Annu. Rev. Biophys. Biophys. Chem. 1991,
20, 519–538.
[10] F. Corzana, J. H. Busto, S. B. Engelsen, J. Jimꢇnez-Barbero, J. L.
Asensio, J. M. Peregrina, A. Avenoza, Chem. Eur. J. 2006, 12, 7864–
7871.
B. R. OꢄKeefe, J. B. McMahon, K. E. Palmer, R. A. Dwek, M. R.
Wormald, A. Wlodawer, Proteins Struct. Funct. Bioinf. 2007, 67,
661–670.
[22] AMBER 9, 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. Roitberg, 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, University of California, San Francisco, 2006.
[23] K. N. Kirschner, A. B. Yongye, S. M. Tschampel, J. Gonzꢈlez-Outeir-
iÇo, C. R. Daniels, B. L. Foley, R. J. Woods, J. Comput. Chem. 2008,
29, 622–655.
[24] W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey,
M. L. Klein, J. Chem. Phys. 1983, 79, 926–935.
[25] C. I. Bayly, P. Cieplak, W. D. Cornell, P. A. Kollman, J. Phys. Chem.
1993, 97, 10269–10280.
[11] F. Corzana, J. H. Busto, G. Jimꢇnez-Osꢇs, J. L. Asensio, J. Jimꢇnez-
Barbero, J. M. Peregrina, A. Avenoza, J. Am. Chem. Soc. 2006, 128,
14640–14648.
[12] a) A. P. Nanzer, T. Huber, A. E. Torda, W. F. van Gunsteren, J.
Biomol. NMR 1996, 8, 285–291; b) A. E. Torda, R. M. Scheek, W. F.
van Gunsteren, J. Mol. Biol. 1990, 214, 223–235; c) D. A. Pearlman,
P. A. Kollman, J. Mol. Biol. 1991, 220, 457–479.
[13] T. Haselhorst, T. Weimar, T. Peters, J. Am. Chem. Soc. 2001, 123,
10705–10714.
[14] G. R. J. Thatcher, The Anomeric Effect and Associated Stereoelec-
tronic Effects, American Chemical Society, Washington, DC 1993.
[15] a) P. G. Vasudev, R. Rai, N. Shamala, P. Balaram, Biopolymers 2008,
90, 138–150; b) M. Lasa, A. I. Jimꢇnez, M. M. Zurbano, C. Cativie-
la, Tetrahedron Lett. 2005, 46, 8377–8380.
[16] a) A. Avenoza, J. H. Busto, C. Cativiela, J. M. Peregrina, F. Rodrıꢄ-
guez, Tetrahedron Lett. 2002, 43, 1429–1432; b) A. Avenoza, C. Cat-
iviela, M. A. Fernꢈndez-Recio, J. M. Peregrina, J. Chem. Soc. Perkin
Trans. 1 1999, 3375–3379.
[26] 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. Mo-
rokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, 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.
[27] J. P. Ryckaert, G. Ciccoti, H. J. C. Berendsen, J. Comput. Phys. 1977,
23, 327–341.
[17] a) O. Golan, Z. Goren, S. E. Biali, J. Am. Chem. Soc. 1990, 112,
9300–9307; b) Z. Goren, S. E. Biali, J. Am. Chem. Soc. 1990, 112,
893–894.
[18] a) C. Andersson, S. B. Engelsen, J. Mol. Graphics Modell. 1999, 17,
101–105. b) F. Corzana, M. S. Motawia, C. H. Du Penhoat, S. Perez,
S. M. Tschampel, R. J. Woods, S. B. Engelsen, J. Comput. Chem.
2004, 25, 573–586.
[19] R. Loris, F. Casset, J. Bouckaert, J. Pletinckx, M. H. Dao-Thi, F.
Poortmans, A. Imberty, S. Perez, L. Wyns, Glycoconjugate J. 1994,
11, 507–517.
[28] T. A. Darden, D. York, L. G. Pedersen, J. Chem. Phys. 1993, 98,
10089–10092.
[29] CCDC-846723 (9) and CCDC-854932 (15) contain the supplementa-
ry crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
[30] G. M. Sheldrick, SHELXL97. Program for the refinement of crystal
structures, University of Gçttingen, Germany, 1997.
[20] N. K. Vyas, M. N. Vyas, F. A. Quiocho, Science 1988, 242, 1290–
1295.
[21] a) N. E. Ziolkowska, S. R. Shenoy, B. R. OꢄKeefe, A. Wlodawer,
Protein Sci. 2007, 16, 1485–1489; b) N. E. Ziolkowska, S. R. Shenoy,
Received: October 3, 2011
Revised: January 19, 2012
Published online: March 1, 2012
5104
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5096 – 5104