The nature and sequence of the amino acid aglycone strongly modulates the conformation and dynamics effects of Tn antigen's clusters

Article abstract of DOI:10.1002/chem.200801777

Synthetic oligosaccharide vaccines based on carbohydrate epitopes are currently being evaluated as potential immunotherapeutics in the treatment of cancer. In an effort to study the role that the amino acid moiety (L-serine and/or L-threonine residues) pl

Full text of DOI:10.1002/chem.200801777

FULL PAPER
DOI: 10.1002/chem.200801777
The Nature and Sequence of the Amino Acid Aglycone Strongly Modulates
the Conformation and Dynamics Effects of Tn Antigenꢀs Clusters
Francisco Corzana,*[a] Jesffls H. Busto,[a] Marisa García de Luis,[a]
Jesffls JimØnez-Barbero,[b] Alberto Avenoza,[a] and Jesffls M. Peregrina*[a]
Abstract: Synthetic oligosaccharide
vaccines based on carbohydrate epi-
topes are currently being evaluated as
potential immunotherapeutics in the
treatment of cancer. In an effort to
study the role that the amino acid
moiety (l-serine and/or l-threonine
residues) plays on the global shape of
the resulting glycopeptides and on the
dynamics of the carbohydrate moiety,
diverse glycopeptides based on the
Tn antigen have been synthesized and
studied in aqueous solution by combin-
ing NMR spectroscopic experiments
and molecular dynamics simulations.
Our results demonstrate that although
the effect of the clustering of Tn on the
peptide backbone is not remarkable, it
substantially modifies the dynamics,
and thus, the presentation features of
the carbohydrate moiety. In fact, the
selected sequence has a crucial influ-
ence on both the orientation and flexi-
bility of the sugar region. Thus, al-
though a serine–threonine pair shows a
well-defined spatial disposition of the
Tn epitopes, its analogue sequence
threonine–serine allows
a
certain
degree of mobility that could favor the
interaction with a diversity of receptors
without a major energy penalty. These
features can be explained by attending
to the different conformational behav-
ior of the glycosidic linkage of threo-
nine-containing glycopeptides when
compared with those of the serine ana-
logues. On this basis, and taking into
account that these carbohydrates inter-
act with components of the immune
system, these findings could have im-
plications for further design of new
cancer vaccines.
Keywords: antigens · conformation
analysis · glycopeptides · molecular
dynamics · NMR spectroscopy
Introduction
are now accessible[2] and may be associated with cancer pro-
gression and metastasis.[3,4] Therefore, mucins and their de-
rivatives are attracting real interest as potential targets for
immunotherapy in the development of vaccines for the
treatment of cancer.[5]
Two main approaches have been taken for carbohydrate-
based vaccines designs: the monomer approach, using isolat-
ed carbohydrate-based antigens, and the cluster approach,
achieved by glycosylation of contiguous l-serine (S or Ser)
and/or l-threonine (T or Thr) residues. Thus, dimeric (or
even trimeric) clusters of carbohydrate–antigens have been
attached to a Ser or Thr peptide scaffold.[6]
From a structure-based design perspective, and to discern
how mucins interact with their biological targets, it is essen-
tial to study in detail their conformational preferences and
the factors that can modify these preferences. At the present
moment, it is well known that a-O-glycosylation forces the
underlying peptide into an extended conformation,[7] which
minimizes the steric crowding of the proximal carbohydrate
structures and could favor the binding between the sugar
and the components of the immune system.[8] This finding
has been explained by forming specific hydrogen bonds be-
Mucin MUC1 is an extensively O-glycosylated protein that
is largely composed of an expressed variable number of
tandem repeats of the sequence PDTRPAPGSTAP-
PAHGVTSA.[1] Expression of this mucin MUC1 is drastical-
ly increased in tumor cells. This over-expression is accompa-
nied by down-regulation of glucosaminyltransferase activity.
As a result of this incomplete glycosylation, the normally
hidden core Tn (a-O-GalNAc-Ser/Thr), sialyl-Tn (STn), and
Thomsen–Friedenreich (TF) carbohydrate-based antigens
[a] Dr. F. Corzana, Dr. J. H. Busto, M. Garcꢀa de Luis, Prof. A. Avenoza,
Dr. J. M. Peregrina
Departamento de Quꢀmica, Universidad de La Rioja
Grupo de Sꢀntesis Quꢀmica de La Rioja
UA-CSIC. Madre de Dios 51, 2006 LogroÇo (Spain)
Fax : (+34)941-299-621
E-mail: jesusmanuel.peregrina@unirioja.es
[b] Prof. J. Jimꢁnez-Barbero
Centro de Investigaciones Biolꢂgicas (CSIC)
Ramiro Maeztu 9, 2840 Madrid (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801777.
Chem. Eur. J. 2009, 15, 3863 – 3874
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3863

tween the peptide moiety and the first a-d-N-acetylgalactos-
amine (GalNAc) unit, which “locks” the orientation of the
sugar with respect to the peptide backbone.[7,9] In this con-
text, we have recently reported that the existence of bridg-
ing-water molecules between the sugar and the peptide moi-
eties could also explain the stabilization of the extended
conformation of the backbone.[10a] Interestingly, the confor-
mational analyses of LI-cadherin glycopeptides reported by
Kuhn and Kunz not only confirm the effect of the carbohy-
drate moiety on the peptide backbone but also show that
this effect is dependent on the carbohydrate type.[11]
On the other hand, although an important number of
studies have been reported on the preferred conformations
of structurally complex and relevant mucin-related glyco-
peptides,[7,11] few conformational studies have been carried
out on Tn glycopeptides containing “clustered” Ser/Thr se-
quences. An excellent example was reported by Kunz and
co-workers.[12] The study revealed that the glycosylation of
the STAPPA and PAPGSTAPPA peptides with aGalNAc
results in a stiffening effect at
sidic linkages of GalNAc-Ser and GalNAc-Thr adopt com-
pletely different 3D orientations.[15] These distinct spatial ar-
rangements of the carbohydrates could have important in-
ferences on the global shape and presentation of the epi-
tope, which could be of paramount importance for a proper
interaction between the sugars and the components of the
immune system.
Importantly, although great efforts have been made to un-
derstand the influence of the carbohydrate moiety on the
peptide backbone, not enough information has been report-
ed concerning the dynamics and conformational preferences
of the carbohydrate, particularly, with regard to the confor-
mational behavior of the glycosidic linkage.
Taking into account the above comments and to gain
detail on the influence that the underlying amino acid (Ser
or Thr) has on the global structure as well as on the dynam-
ics of the carbohydrate moiety, we have synthesized and car-
ried out the conformational study in an aqueous solution of
the model glycopeptides as shown in Scheme 1. The confor-
the site of glycosylation. The
authors conclude that this fea-
ture could have important im-
plications on the binding affini-
ty between the antibody and
the MUC1-core-related glyco-
peptides.
From a biological point of
view, some important features
have been observed. First, the
amino acids flanking the car-
bohydrate antigen have been
shown to modulate antibody
recognition.[13] Second, recent
studies have experimentally
shown that the aglyconic part
of the Tn structure (Ser or Thr
residues) could also play a key
role for antibody recognition.
Thus, the STT backbone was
found to be the most permis-
sive scaffold for induction of
anti-Tn antibodies.[14] More-
over, in recent work[5m] con-
cerning the study of synthetic
MUC1-based
glycopeptides
used as antigen-presenting
cells, only the processed glyco-
peptides that carried GalNAc
a-O-linked to the GSTA motif
of the peptide were capable of
activating hybridoma T-cells.
Glycopeptides with glycosyla-
tion at the VTSA motif were
inactive.
In this context, we have re-
cently reported that the glyco-
Scheme 1. Glycopeptides studied in this work.
3864
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3863 – 3874

Conformation and Dynamics of the of Tn Antigen
FULL PAPER
mational analysis of all these
compounds has been carried
out by using NMR spectrosco-
py. In particular, NOE and
coupling constants have been
interpreted with the assistance
of molecular dynamics (MD)
simulations, following a proto-
col previously described by
our group.[10]
Results and Discussion
Synthesis: As key examples,
Scheme 2 and Scheme 3 show
the synthesis of T*T* and SS*
glycopeptides. The rest of the
molecules were synthesized as
described in the Supporting In-
formation and follow a similar
methodology.
The synthesis starts with l-
Thr, whose carboxylic acid
group was protected as an allyl
ester, a selectively removable
carboxy-protecting function,[16]
and the amino group as a tert-
butyloxycarbonyl (Boc) carba-
mate, giving compound 1. The
treatment of the protected Thr
derivative 1 with 3,4,6-tri-O-
benzyl-2-nitrogalactal, follow-
ing the procedure described by
Schmidt and co-workers,[17a–c]
gave the linked 2-nitroglyco-
side 2 in good yield. Then, se-
lective deprotection of the
amino or acid groups in this
compound gave, almost quan-
titatively, derivative 3 or 4, re-
spectively. Further treatment
of a mixture of 3 and 4 with
N,N,N’,N’-tetramethyl-O-(ben-
zotriazol-1-yl)uronium tetra-
fluoroborate (TBTU) as a cou-
pling agent and diisopropyl-
Scheme 2. Synthesis of T*T*. a) i) Allylic alcohol, pTsOH, toluene, reflux, 24 h. ii) (Boc)2O, Na2CO3, THF/
H2O (5:1), 258C, 12 h, 71% overall yield; b) 3,4,6-Tri-O-benzyl-2-nitrogalactal, tBuOK, THF, 258C, 12 h,
68%; c) TFA/CH2Cl2 (2:5), 0 to 258C, 3 h, quantitative; d) [Pd
ACHTUNGTERN(NUNG PPh3)4], morpholine, THF, 258C, 3 h, 87 %;
e) TBTU, DIEA, CH3CN, 258C, 10 h, 67%; f) i) [Pd(PPh3)4], morpholine, THF, 258C, 3 h. ii) MeNH2·HCl,
AHCTUNGTRENNUNG
DIEA, CH3CN, 258C, 10 h, 64% overall yield; g) i) TFA/CH2Cl2 (1:2), 0 to 258C, 3 h. ii) Ac2O, Py, 25 8C, 3 h,
80% overall yield; h) i) H2/Ni Raney (T4), EtOH, 258C, 7 h. ii) Ac2O, Py, 258C, 3 h. iii) H2/Pd-C, MeOH/ethyl
acetate (2:1), 258C, 12 h, 33% overall yield.
ACHTUNGTRENNUNGethylamine (DIEA) as a base
led to the doubly glycosylated
compound 5. Selective removal
of the allyl group was carried
out with Pd0-catalyzed allyl
transfer to morpholine. Under
these conditions, the deprotec-
tion proceeded without affect-
Scheme 3. Synthesis of SS*. a) 3,4,6-Tri-O-benzyl-2-nitrogalactal, tBuOK, THF, 258C, 12 h, 74%; b) TFA/
CH2Cl2 (2:5), 0 to 258C, 3 h, quantitative; c) TBTU, DIEA, CH3CN, 258C, 10 h, 69%; d) i) [PdACHTNUTRGNE(UNG PPh3)4], mor-
pholine, THF, 258C, 3 h. ii) MeNH2·HCl, DIEA, CH3CN, 258C, 10 h, 65% overall yield; e) i) TFA/CH2Cl2
(1:2), 0 to 258C. ii) Ac2O, Py, 258C, 3 h, 83% overall yield; f) i) H2/Ni Raney (T4), EtOH, 258C, 7 h. ii) Ac2O,
Py, 258C, 3 h. iii) H2/Pd-C, 258C, MeOH/ethyl acetate (2:1), 12 h, 46% overall yield.
ing other labile groups.[18] The subsequent transformation of
tection of the amino group and its further acetylation with
the acid group into methyl amide gave derivative 6. Depro-
acetic anhydride (Ac2O) and pyridine (Py) led to 7 in a
Chem. Eur. J. 2009, 15, 3863 – 3874
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3865

F. Corzana, J. M. Peregrina et al.
good yield. The reduction of the nitro groups was carried
out by hydrogenation of 7 with nickel Raney (T4). Finally,
acetylation of the amino groups and deprotection of the hy-
droxy groups of the sugar moieties gave the desired model
glycopeptide T*T* with a moderate yield (Scheme 2).
Glycopeptide SS* was obtained as follows. First, 2-nitro-
glycoside 9 was synthesized starting from the Ser derivative
8[17d] and following the methodology described above for its
analogue, 6. Then, selective deprotection of the amino
group afforded compound 10, which was treated with the
commercially available Ser derivative 11 in the presence of
TBTU and DIEA to give 12. Finally, this compound was
transformed into glycopeptide SS* by using the strategy de-
scribed above for the synthesis of T*T* (Scheme 3).
Figure 2. Chemical-shift deviations (CSD) of Ha of different glycopepti-
des. The CSD for Ser and Thr residues was calculated taking dHarandomcoil
as 4.47 and 4.35, respectively (see reference [20]).
NMR spectroscopic studies: In a first step, full assignment
1
of H NMR spectra of all of the compounds was carried out
by using standard COSY, HSQC, and NOESY experiments.
This assignment is displayed in the Supporting Information.
The torsional angles and the numbering used in this work
for the key compounds are shown in Figure 1.
the Ha of the glycosylated residues showed a significant
downfield shift. Therefore, although our model glycopepti-
des are too short to adopt a well-defined secondary struc-
ture, this result is in agreement with the increase in the pop-
ulation of extended peptide structures upon glycosylation.[10]
Chemical shift deviation: Generally, in proteins and large
peptides, Ha chemical-shift deviations (CSD, DdHa=
2D NOESY cross-peaks: 2D NOESY experiments were
then carried out in H2O/D2O (9:1) (208C, pH 5.2) for the
model glycopeptides (see the Supporting Information and
Figure 3 and Figure 4). In all cases, strong consecutive
Ha,NH (i,i+1) connectivities were found (Figure 3). This
finding, together with the weak or absence of consecutive
NH,NH (i,i+1) NOE values (Figure 4) indicated the pre-
dominance of extended backbone conformations.[21] On the
other hand, when a Thr residue was present, a small–
medium NOE value between the NH hydrogen of the corre-
sponding GalNAc (NHs) and the NH of the attached Thr
was observed, indicating the presence of an eclipsed confor-
mation[15] for the ys dihedral angle of the GalNAc-Thr
moiety (Figure 4). Furthermore, no significant NOE values
were detected between the methyl or the NHs proton of the
N-acetyl group of the GalNAc and the NH (i+1) proton of
the peptide backbone. This suggests the absence of relevant
hydrogen bonds between the sugar and the peptide moiet-
ies.[10a]
dHaobservedÀdHarandomcoil
)
exhibit an average value of
À0.39 ppm when a residue is placed in a helical conforma-
tion. In contrast, a mean shift of +0.37 ppm is observed
when the residue is found in an extended conformation.[19]
As can be seen in Figure 2, the obtained CSD values[20] of
As a next step, distances involving NH protons were
semi-quantitatively determined by integrating the corre-
sponding cross-peak volumes in the 2D NOESY spectrum
(Table 1).
3J coupling constants: In addition, 3J coupling constants
were measured from the splitting of the resonance signals in
the 1D spectrum (Table 2). The relatively high values of the
3
peptide backbone JNH,Ha coupling constants for all the resi-
dues also suggest the existence of a preferred extended
3
structure for all the glycopeptides.[22] The JNHs,H2s values in
the GalNAc residues were all close to 9 Hz, thus suggesting
that the torsion angle between the H2 and NH protons to
be ꢀ1808.[23] Accordingly, the orientation of the N-acetyl
group relative to the sugar framework seems essentially
Figure 1. Molecular structure of compound SS* with definitions of the
torsional angles and the numbering of the atoms. The same definitions
were used for the other model glycopeptides.
3866
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3863 – 3874

Conformation and Dynamics of the of Tn Antigen
FULL PAPER
Figure 4. Section of the 800 ms NOESY spectrum (400 MHz) of com-
pounds T*S* in H2O/D2O (9/1) at 208C showing the amide–amide cross-
peaks. The NOE contact between NH1-NHs(I) is representative of an
eclipsed conformation of the glycosidic linkage (ys =1208). Diagonal
peaks are negative and the NOE contacts are represented as positive
cross-peaks.
Figure 3. Sections of the 800 ms NOESY spectra (400 MHz) of com-
pounds S*T* and T*S* in H2O/D2O (9:1) at 208C showing amide–ali-
phatic cross-peaks. The NOE contacts are represented as positive cross-
peaks.
formers showed fp,yp dihedral values corresponding to
helix-like conformations (Figure 5 and the Supporting Infor-
mation). This result is in good agreement with the observed
pattern of NOE values described above.
Figure 6 shows the side-chain distribution (c1 dihedral
angle) obtained from the MD-tar simulations. Important
conclusions can be drawn from these data. First, the side
chain of the non-glycosylated residues was more flexible
than those of their glycosylated counterparts, showing a sig-
nificant population of the three possible staggered conform-
ers [anti, with c1 around 1808; gauche (+) or g(+), with c1
around 608; and gauche (À) or g(À), with c1 around À608].
Second, the rotation around c1 for the glycosylated Thr resi-
dues was significantly restricted, with a clear preference for
a g(+) conformation, except for the derivative T*S*, for
which both the anti and g(+) conformers seem to be similar-
ly populated. These findings are all consistent with the small
3JHa,Hb coupling constant values for Thr residues.
On the other hand, no significant hydrogen bonding was
detected over the course of MD-tar simulations between the
N-acetyl group of the GalNAc residue and the carbonyl
group of the underlying amino acid in any of the studied
glycopeptides. Indeed, if any, this inter-residual hydrogen
bonding was present in less than 1% of the total trajectory
for all glycosylated Ser residues. In contrast, in the case of
Thr residues, this hydrogen bonding was present about 10%
of time. This result is in accordance with the experimental
3
fixed. Additionally, JHa,Hb values were also determined to
provide insights into the c1 angles for the Thr and Ser side
chains. Strikingly, the coupling constants for the Thr residues
had remarkably small values, falling between 2.4 and 2.7 Hz.
These low values preclude any significant torsional averag-
ing and strongly indicate that the rotation around c1 is
rather restricted.[24]
MD simulations: To get an experimentally derived ensem-
ble, MD-tar (MD with time-averaged restraints)[25] simula-
tions were carried out by the inclusion of the experimental
3
distances and J coupling constants (Table 1 and Table 2) as
time-averaged restraints. The results obtained from the MD-
tar simulations (see the Supporting Information) showed a
close agreement, in numerical terms, between the calculated
3
and the experimental proton–proton distances and J cou-
pling constants.
The calculated fp,yp distribution maps for the peptide
backbone of both the singly and doubly Tn-glycosylated de-
rivatives were similar to the typical values expected for ex-
tended conformations, such as polyproline II and the b-
sheet conformations. Indeed, only a small number of con-
Chem. Eur. J. 2009, 15, 3863 – 3874
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3867

F. Corzana, J. M. Peregrina et al.
Table 1. Calculated and experimental (in brackets) distances obtained for the glycopeptides.[a]
orientation. In fact, although
the GalNAc moiety in Thr is
almost perpendicular to the
peptide backbone, in Ser resi-
dues, the sugar moiety adopts
a parallel disposition. It is evi-
dent that these differences
must also have implications in
the global shape of the mole-
cules, and also in the flexibility
of the systems. Figure 8 shows
the superimposition of 20 con-
formers randomly taken from
the MD-tar simulations. As
can be seen, both S*T* and
T*T* can properly accommo-
date the two GalNAc residues,
giving a rather rigid structure.
However, in the case of T*S*
and S*S*, a larger degree of
flexibility was observed. Addi-
tionally, the global shape of
T*S* differs completely from
that derived for the rest of the
model glycopeptides. In fact,
the sugar moieties are located
at different regions of the
plane defined by the peptide.
On the other hand, the S*T*
glycopeptides exhibits one of
the lowest root mean square
Distance
SS*
S*S
S*S*
S*T*
T*S*
T*T*[b]
NH1–NH2
>3.0
>3.0
>3.0
>3.0
>3.0
>3.0
A
E
E
G
U
ACHTUNGTRENNUNG
NH2–NH3
NHs(I)–NH1
NHs(II)–NH2
Ha1–NH1
G
G
G
N
U
ACHTUNGTRENNUNG
–
(–)
–
ACHTUNGTRENNUNG
E
ACHTUNGTRENNUNG
A
E
E
G
N
ACHTUNGTRENNUNG
[c]
Ha1–NH2
A
ACHTUNGTRENNUNG
Ha2–NH2
A
E
E
G
U
ACHTUNGTRENNUNG
Ha2–NH3
A
G
G
N
G
ACHTUNGTRENNUNG
[c]
[c]
NH1–Hb1[d]
NH2–Hb1[d]
NH2–Hb2[d]
NH3–Hb2[d]
2.8/2.8[d]
A
ACHTUNGTRENNUNG
[c]
A
T
ACHTUNGTRENNUNG
[c]
A
U
ACHTUNGTRENNUNG
[c]
A
E
E
N
R
ACHTUNGTRENNUNG
[a] The distances are given in ꢄ. The maximum estimated experimental error is 10%. [b] In the case of the
T*T* sequence, NOE build-up curves were carried out to give H1s(I)–Hb1 (2.1 ꢄ) and H1s(II)–Hb2 (2.2 ꢄ)
distances. [c] Not determined owing to an overcrowded NMR spectrum. [d] The first value is referred to
HbproR and the second one to HbproS.
3
Table 2. Calculated and experimental (in brackets) J coupling constants obtained for the glycopeptides.[a]
3J
SS*
S*S
S*S*
S*T*
T*S*
T*T*
Ha1, Hb1[b]
5.8[c]
5.1[c]
5.3[c]
6.0/4.5
2.5
2.7
ACHTUNGTRENNUNG
A
G
G
N
E
Ha2, Hb2[b]
NH1,Ha1
deviation
(RMSD)
values
E
C
ACHTUNGTRENNUNG
(2.2 ꢄ), whereas the T*S* de-
rivative is rather flexible with
an RMSD close to 2.6 ꢄ.
A
E
E
G
U
ACHTUNGTRENNUNG
NH2,Ha2
A
G
N
G
N
ACHTUNGTRENNUNG
To assess the flexibility of
this type of molecule, the fol-
lowing experimental approach
was considered. Selective NOE
values were obtained upon in-
version of the Hb proton of
the Thr residue in T*S* and
S*T* compounds at different
NHs(I),H2s(I)
NHs(II),H2s(II)
–
(–)
9.8
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
[a] 3J coupling constants are given in Hz. The estimated experimental error is 10%. [b] The first value is re-
ferred to HbproR and the second one to HbproS, which were extracted from J resolved experiments (see the
Supporting Information). [c] Experimentally observed as a pseudotriplet.
findings reported by Gururaja and co-workers,[26] who previ-
ously stated that, in mucin derivatives, this hydrogen bond
was only present when a Thr residue was incorporated in
the glycopeptide.
As we have recently reported,[15] one of the most impor-
tant differences between the glycosidic linkages of Ser and
Thr concerns the ys dihedral angle value (C1s-O1s-Cb-Ca).
In general, when a Thr is present, this angle adopts a value
close to 1208 (showing an eclipsed conformation around
O1s-Cb as shown in Figure 4). In contrast, for Ser deriva-
tives a value around 1808 is preferred. The different behav-
ior of both glycosidic linkages can be seen in Figure 7. As a
consequence of the different ys values in Ser and Thr deriv-
atives, the carbohydrate moiety adopts a complete different
3
3868
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3863 – 3874

Conformation and Dynamics of the of Tn Antigen
FULL PAPER
Figure 6. a) More stable conformers of the side chain (c1) for a glycosy-
lated Ser residue. b) Distributions for the side chain (c1) obtained from
the MD-tar simulations for the glycopeptides.
Figure 5. Distributions obtained from the MD-tar simulations for the
peptide backbone (fp/yp) of SS* and S*S* glycopeptides.
In summary, the results obtained in this work demonstrate
that the peptide sequence used for clustering of Tn antigens
plays a significant role, not only on the flexibility of the sys-
tems, but also, and more interestingly, on their different
global shape. This result is, to some extend, in good accord-
ance with the results previously obtained by Lo-Man and
co-workers.[14] In this study, the STT backbone was found to
be the most satisfactory of all those tested for the induction
of anti-Tn antibodies.
well as some key NOE contacts (Thr2 NH-Ser1 Ha and Ala3
NH-Thr2 Ha) indicate that the peptide adopts an extended
conformation in this region of the peptide chain. Moreover,
the presence of a NOE contact between Thr2 NH and the
NH proton of the GalNAc attached to this Thr corroborates
the eclipsed conformation of the ys torsion angle. As ex-
pected, this key NOE contact was not observed in the Ser
residue. In addition, all these experimental data and the
3JHa,Hb coupling constants were used as restraints in the MD-
tar simulations carried out on this glycohexapeptide. As can
be seen in Figure 10, the dynamics and the global shape of
the carbohydrate moieties are quite similar to those de-
scribed in this work for the S*T* species.
Conclusion
Our results demonstrate that, although the effect of cluster-
ing of Tn on the peptide backbone is not remarkable, the
clustering strongly modifies the dynamics of the carbohy-
On the other hand, the eclipsed conformation of the ys
glycosidic linkage of the aGalNAc-Thr residue can be ex-
perimentally
observed
in
recent detailed studies on gly-
copeptides, such as the glyco-
sylated MUC1 eicosapeptide,
studied by Kunzꢅs group,[7a]
and the triglycosylated se-
quence STTAV reported by
Danishefsky and co-workers.[7b]
Furthermore, the rigidity ex-
hibited for S*T* and T*T* gly-
copeptides is in accordance
with the conformationally
highly stable structure predict-
ed for the STTAV molecule.
Figure 7. fs/ys distributions obtained from the MD-tar simulations for the glycosidic linkage of residue I of
S*S* and T*T* glycopeptides, together with the histograms of ys dihedral angle.
Chem. Eur. J. 2009, 15, 3863 – 3874
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3869

F. Corzana, J. M. Peregrina et al.
Figure 9. Selective 1D NOESY spectra of compounds T*S* and S*T* at
different temperatures.
Figure 8. Calculated ensembles obtained from the MD-tar simulations for
the glycopeptides. The numbers indicate the RMSDs (ꢄ) for heavy-atom
superimposition with respect to the average structure of the trajectory.
drate moiety and thus, its presentation ability. In fact, the se-
lected sequence has a crucial influence on both the orienta-
tion and flexibility of the carbohydrate moiety. Consequent-
ly, although the ST sequence somehow fixes the spatial dis-
position of the Tn epitopes, the alternative analogue se-
quence, TS, allows for a significant degree of mobility for
the carbohydrate moiety. This feature can be explained by
considering the different conformational behavior of the gly-
cosidic linkage of the Thr residue when compared with that
of the Ser one. On this basis, and taking into account that
these carbohydrates presumably interact with components
of the immune system, these findings could be implications
for designing new cancer vaccines.
Figure 10. a) fs/ys distribution obtained from the 80 ns MD-tar simula-
tions for S*T*APPA glycopeptide, showing the eclipsed conformation of
the glycosidic linkage in the Thr residue. b) Calculated ensembles ob-
tained from the MD-tar simulations for S*T*APPA glycopeptide. The
RMSDs (ꢄ) was calculated for heavy-atom superimposition of the aGal-
NAc-Ser and aGalNAc-Thr residues with respect to the average struc-
ture of the trajectory. c) Distributions for the side chain (c1) obtained
from the MD-tar simulations for the glycopeptide.
Experimental Section
General procedures: Solvents were purified according to standard proce-
dures. Analytical TLC was performed by using Polychrom SI F254 plates.
Column chromatography was performed by using silica gel 60 (230–400
mesh). H and 13C NMR spectra were recorded on Bruker ARX 300 and
1
3870
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3863 – 3874

Conformation and Dynamics of the of Tn Antigen
FULL PAPER
1
Bruker Avance 400 spectrometers. H and 13C NMR spectra were record-
3.40–3.52 (m, 2H), 3.88–4.00 (m, 2H), 4.22–4.46 (m, 6H), 4.57–4.66 (m,
2H), 4.73 (d, 1H, J=10.7 Hz), 4.80–4.88 (m, 1H), 5.08 (d, 1H, J=
8.5 Hz), 5.30–5.39 (m, 1H), 7.12–7.30 ppm (m, 15H); 13C NMR
(100 MHz, CDCl3): d=18.0, 28.3, 57.7, 68.2, 69.9, 72.9, 73.0, 73.5, 75.0,
75.0, 77.1, 80.2, 84.2, 97.3, 127.7, 127.8, 127.9, 128.0, 128.1, 128.3, 128.4,
128.6, 137.3, 137.6, 137.9, 156.0, 173.7 ppm.
ed in CDCl3, CD3OD, and D2O with TMS as the external standard by
using a coaxial microtube (chemical shifts are reported in ppm on the d
scale, coupling constants in Hz). Melting points were determined on a
Bꢆchi B-545 melting-point apparatus and are uncorrected. Optical rota-
tions were measured on a Perkin-Elmer 341 polarimeter. Microanalyses
were carried out on a CE Instruments EA-1110 analyser and are in good
agreement with the calculated values.
Synthesis of compound 5: A solution of acid 4 (681 mg, 1.0 mmol) in ace-
tonitrile (30 mL) was treated with DIEA (0.7 mL, 4.2 mmol), compound
3 (735 mg, 1.0 mmol), and (benzotriazol-1-yl)-1,1,3,3-tetramethyluronium
TBTU (0.42 g, 1.3 mmol). The reaction mixture was stirred at 258C for
10 h and then partitioned between brine (20 mL) and ethyl acetate
(50 mL). The organic layer was washed with 0.1n HCl (2ꢇ20 mL) and
5% NaHCO3 (2ꢇ15 mL). Finally, the organic layer was dried, filtered,
and evaporated to give a residue that was purified by a silica gel column
chromatography (hexane/ethyl acetate, 4:1) to give 5 as a colorless oil
(860 mg, 67%). [a]2D5 =+108.0 (c=1.30, CHCl3); 1H NMR (400 MHz,
CDCl3): d=1.19 (d, 3H, J=5.8 Hz), 1.29 (d, 3H, J=6.0 Hz), 1.44 (s,
9H), 3.45–3.55 (m, 4H), 3.94–4.01 (m, 2H), 4.10–4.25 (m, 4H), 4.32–4.51
(m, 7H), 4.53–4.58 (m, 2H), 4.61–4.69 (m, 2H), 4.70–4.81 (m, 4H), 4.83–
4.97 (m, 4H), 5.10 (dd, 1H, J1 =10.6 Hz, J2 =3.2 Hz), 5.23–5.39 (m, 2H),
5.41 (d, 1H, J=3.7 Hz), 5.55 (d, 1H, J=6.2 Hz), 5.59 (d, 1H, J=3.0 Hz),
5.90–6.03 (m, 1H), 6.87 (d, 1H, J=8.7 Hz), 7.13–7.40 ppm (m, 30H);
13C NMR (100 MHz, CDCl3): d=16.2, 18.0, 28.3, 56.6, 57.5, 66.7, 68.3,
68.5, 70.0, 70.1, 72.7, 72.8, 73.4, 73.5, 73.5, 73.6, 74.8, 75.0, 75.0, 75.7, 77.9,
80.0, 84.3, 85.3, 97.4, 97.6, 119.1, 127.6, 127.7, 127.8, 127.9, 128.0, 128.1,
128.2, 128.3, 128.3, 128.4, 131.7, 137.5, 137.7, 137.8, 137.9, 138.0, 138.0,
155.1, 168.7, 169.0 ppm; elemental analysis calcd for C70H82N4O19: C
65.51, H 6.44, N 4.37; found: C 65.41, H 6.42, N 4.39.
Synthesis of compound 1:
A mixture of allylic alcohol (20.0 mL,
0.296 mol), l-Thr (4.00 g, 33.6 mmol), p-TsOH (7.65 g, 40.4 mmol), and
toluene (120 mL) was refluxed for 24 h by using a Dean-Stark trap and
then concentrated. The crude salt was suspended in THF/H2O (5:1,
120 mL) and Na2CO3·10H2O (28.6 g, 0.10 mol) was added. The resulting
suspension was cooled to 08C and
a solution of Boc2O (7.33 g,
33.6 mmol) in THF (50 mL) was added. The mixture was allowed to
warm to 258C and stirred for 12 h. The solvent was then removed and
the mixture partitioned between brine (40 mL) and ethyl acetate
(100 mL). The organic layer was washed with 0.1m HCl (2ꢇ20 mL) and
5% NaHCO3 (2ꢇ20 mL), dried, filtered, and evaporated to give a resi-
due, which was purified by silica gel column chromatography (hexane/
ethyl acetate, 3:2) to give 1 (6.22 g, 71%) as a colorless oil. [a]2D5 =À13.2
(c=1.06, CHCl3); 1H NMR (300 MHz, CDCl3): d=1.20 (d, 3H, J=
6.3 Hz), 1.40 (s, 9H), 2.89 (brs, 1H), 4.16–4.34 (m, 2H), 4.61 (d, 2H, J=
5.6 Hz), 5.15–5.36 (m, 2H), 5.46 (d, 1H, J=9.0 Hz), 5.75–5.97 ppm (m,
1H); 13C NMR (100 MHz, CDCl3): d=19.8, 28.2, 58.8, 65.9, 67.9, 79.9,
118.6, 131.4, 156.1, 171.2 ppm; elemental analysis calcd for C12H21NO5: C
55.58, H 8.16, N 5.40; found: C 55.17, H 8.11, N 5.39.
Synthesis of compound 2: 3,4,6-Tri-O-benzyl-2-nitrogalactal (1.10 g,
2.65 mmol), which was prepared as described in the literature,[17b] and
Thr derivative 1 (529 mg, 2.04 mmol) were dissolved in THF (40 mL)
under an argon atmosphere, and molecular sieves were then added. The
reaction mixture was stirred at 258C for 15 min, and then a 1m potassium
tert-butoxide solution in THF (0.2 mL, 0.20 mmol) was added. The reac-
tion was stirred for 12 h, the molecular sieves were then filtered off, and
all solvents were removed by evaporation. The residue was purified by a
silica gel column chromatography (hexane/ethyl acetate, 4:1) to give 2
(1.00 g, 68%) as a colorless oil. [a]2D5 =+71.5 (c=1.20, CHCl3); 1H NMR
(400 MHz, CDCl3): d=1.23 (d, 3H, J=6.3 Hz), 1.40 (s, 9H), 3.41–3.52
(m, 2H), 3.91–3.99 (m, 2H), 4.19–4.44 (m, 6H), 4.58 (d, 2H, J=5.7 Hz),
4.61–4.69 (m, 2H), 4.75 (d, 1H, J=11.2 Hz), 4.85 (dd, 1H, J1 =10.7 Hz,
J2 =4.2 Hz), 4.99 (d, 1H, J=9.8 Hz), 5.18 (d, 1H, J=10.4 Hz), 5.24–5.32
(m, 2H), 5.81–5.92 (m, 1H), 7.10–7.31 ppm (m, 15H); 13C NMR
(100 MHz, CDCl3): d=18.2, 28.3, 58.2, 66.5, 68.2, 69.9, 73.0, 73.6, 75.0,
75.0, 77.5, 80.1, 84.2, 97.5, 118.9, 127.7, 127.8, 127.9, 128.0, 128.1, 128.3,
128.4, 128.5, 131.7, 137.2, 137.6, 137.8, 156.1, 169.6 ppm; elemental analy-
sis calcd for C39H48N2O11: C 64.99, H 6.71, N 3.89; found: C 64.86, H 6.68,
N 3.86.
Synthesis of compound 6: The cleavage of the allyl ester of 5 was ach-
ieved by using the same procedure described for the preparation of de-
rivative 4 with the following amounts: compound 5 (821 mg, 0.64 mmol),
[PdACHTNUTGRNEUNG
(PPh3)4] (7 mg, 6.4ꢇ10À3 mmol), and morpholine (0.37 mL,
4.48 mmol) to give the desired acid (663 mg, 83%) as a colorless oil, fol-
lowing the workup described for compound 4. A solution of this acid
(665 mg, 0.53 mmol) in acetonitrile (25 mL) was treated with DIEA
(0.35 mL, 2.1 mmol), methylamine hydrochloride (71 mg, 1.06 mmol),
and (benzotriazol-1-yl)-1,1,3,3-tetramethyluronium TBTU (204 mg,
0.64 mmol). The reaction mixture was stirred at 258C for 10 h, and then
partitioned between brine (15 mL) and ethyl acetate (50 mL). The organ-
ic layer was washed with 0.1n HCl (2ꢇ15 mL) and 5% NaHCO3 (2ꢇ
15 mL). Finally, the organic layer was dried, filtered, and evaporated to
give a residue that was purified by a silica gel column chromatography
(hexane/ethyl acetate, 1:1) to give 6 as a colorless oil (514 mg, 77%).
[a]2D5 =+ 105.4 (c=1.10, CHCl3); 1H NMR (400 MHz, CDCl3): d=1.15
(d, 6H, J=5.9 Hz), 1.39 (s, 9H), 2.72 (d, 3H, J=4.1 Hz), 3.41–3.56 (m,
4H), 3.92–4.06 (m, 5H), 4.20–4.48 (m, 10H), 4.59–4.78 (m, 7H), 4.91–
5.01 (m, 2H), 5.28 (d, 1H, J=5.7 Hz), 5.38–5.44 (m, 1H), 5.48 (d, 1H,
J=2.0 Hz), 6.02–6.11 (m, 1H), 7.06–7.33 ppm (m, 31H); 13C NMR
(100 MHz, CDCl3): d=16.9, 17.6, 26.4, 28.3, 56.1, 58.5, 68.1, 68.3, 70.1,
70.3, 72.8, 73.0, 73.0, 73.4, 73.5, 73.5, 74.2, 75.0, 75.0, 75.1, 75.2, 80.7, 85.0,
96.5, 97.3, 127.7, 127.8, 127.8, 128.0, 128.0, 128.1, 128.3, 128.4, 128.5,
137.3, 137.7, 137.9, 156.1, 168.2, 169.1 ppm; elemental analysis calcd for
C68H81N5O18: C 65.01, H 6.50, N 5.57; found: C 64.83, H 6.47, N 5.61.
Synthesis of compound 3: Trifluoroacetic acid (TFA; 2 mL) was added to
a solution of compound 2 (0.50 g, 0.69 mmol) in CH2Cl2 (5 mL) at 08C.
The reaction was maintained at 08C for 30 min, at 258C for 2.5 h, and
concentrated. The crude compound 3 was quantitatively obtained and
used without further purification. 1H NMR (400 MHz, CD3OD): d=
1.41–1.49 (m, 3H), 3.51–3.59 (m, 2H), 4.11–4.21 (m, 3H), 4.39–4.57 (m,
4H), 4.54–4.62 (m, 1H), 4.64–4.85 (m, 5H), 4.86–4.95 (m, 1H), 5.25–5.33
(m, 1H), 5.35–5.46 (m, 1H), 5.46–5.53 (m, 1H), 5.93–6.09 (m, 1H), 7.15–
7.38 ppm (m, 15H); 13C NMR (100 MHz, CD3OD): d=18.3, 58.7, 68.7,
70.1, 72.0, 73.8, 74.6, 74.8, 76.0, 76.3, 76.5, 85.5, 99.2, 120.2, 128.9, 129.0,
129.1, 129.2, 129.3, 129.4, 129.5, 129.5, 132.7, 139.1, 139.3, 139.5,
167.7 ppm.
Synthesis of compound T*T*: TFA (3 mL) was added to a solution of
compound 6 (0.50 g, 0.40 mmol) in CH2Cl2 (6 mL) at 08C. The reaction
was maintained at 08C for 30 min, at 258C for 2.5 h and concentrated.
The crude product was dissolved in pyridine (5 mL) and acetic anhydride
(2 mL) was added. The resulting mixture was stirred for 3 h at 258C. The
solvent was evaporated and the crude product purified by silica gel
column chromatography (hexane/ethyl acetate, 1:4) to give 7, as a color-
less oil (382 mg, 80%). Then, Platinized Raney-nickel (T4) catalyst was
freshly prepared as described in the literature.[27] The catalyst obtained
by using 2 g of Raney nickel/aluminum alloy was suspended in ethanol
(10 mL) and pre-hydrogenated for 10 min before the addition of com-
pound 7 (150 mg, 0.12 mmol) in ethanol (5 mL). The reaction mixture
was shaken under H2 (1 atm) for 7 h at 258C. The catalyst was filtered
off and the solvent evaporated. The residue was dissolved in pyridine/
acetic anhydride (2:1, 6 mL) and stirred at 258C for 3 h. After removing
Synthesis of compound 4: A solution of derivative 2 (500 mg, 0.69 mmol)
in THF (15 mL) was stirred under an argon atmosphere at 258C, [Pd-
ACHTUNGTRENNUNG
(PPh3)4] (8 mg, 7.1ꢇ10À3 mmol) and morpholine (0.4 mL, 4.87 mmol)
were subsequently added. After stirring for 3 h, the solvent was evaporat-
ed and the residue was taken up in 50 mL of CH2Cl2. The resulting solu-
tion was extracted with 1n HCl (3ꢇ20 mL), dried, and concentrated in
vacuo. The desired acid 4 was obtained as a pallid yellow oil (408 mg,
87%), which was then used directly used for the subsequent transforma-
tions. 1H NMR (400 MHz, CDCl3): d=1.16–1.25 (m, 3H), 1.38 (s, 9H),
Chem. Eur. J. 2009, 15, 3863 – 3874
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3871

F. Corzana, J. M. Peregrina et al.
the volatile products, a solution of the crude in MeOH/ethyl acetate (2:1,
6 mL) was treated with 10% palladium–carbon (15 mg) as a catalyst. The
reaction mixture was shaken under H2 (1 atm) for 12 h at 258C. Removal
of the catalyst and a further purification of the residue first by a silica gel
column chromatography (aq. NH3/EtOH/butanol/CHCl3, 8:5:4:2) and
then with a C18 reverse-phase sep-pak cartridge gave T*T* as a colorless
oil (27 mg, 33%). [a]2D5 =+ 96.0 (c=0.82, H2O); 1H NMR (400 MHz,
D2O): d=1.25 (d, 3H, J=6.2 Hz), 1.36 (d, 3H, J=6.2 Hz), 2.03 (s, 3H),
2.10 (s, 3H), 2.14 (s, 3H), 2.72 (s, 3H), 3.72–3.80 (m, 4H), 3.85–3.95 (m,
2H), 3.97–4.01 (m, 2H), 4.01–4.08 (m, 2H), 4.08–4.15 (m, 2H), 4.24–4.32
(m, 1H), 4.34–4.42 (m, 1H), 4.55 (brs, 1H), 4.68 (brs, 1H), 4.89 (d, 1H,
J=3.3 Hz), 4.95 (d, 1H, J=3.3 Hz) ppm; 1H NMR (400 MHz, H2O/D2O):
d=7.51 (d, 1H, J=9.6 Hz, NHs(I)), 7.91–8.00 (m, 2H, NHs(II) + NH3),
8.39 (d, 1H, J=8.5 Hz, NH1), 8.62ppm (d, 1H, J=9.1 Hz, NH3);
13C NMR (100 MHz, D2O): d=18.1, 18.3, 21.7, 22.1, 22.3, 25.8, 49.8, 49.8,
57.2, 57.7, 61.3, 67.6, 68.0, 68.5, 68.6, 71.3, 76.2, 76.8, 98.8, 99.3, 171.1,
Synthesis of compound 13: The cleavage of the allyl ester of 12 was ach-
ieved by using the same procedure described for the preparation of de-
rivative
0.64 mmol), [PdAHCTUNGTRENNUNG
4
with the following amounts: compound 12 (566 mg,
(PPh3)4] (7 mg, 6.4ꢇ10À3 mmol), and morpholine
(0.37 mL, 4.48 mmol) to give the desired acid (469 mg, 87%) as a color-
less oil, following the workup described for compound 4. A solution of
this acid (447 mg, 0.53 mmol) in acetonitrile (25 mL) was treated with
DIEA (0.35 mL, 2.1 mmol), methylamine hydrochloride (71 mg,
1.06 mmol), and (benzotriazol-1-yl)-1,1,3,3-tetramethyluronium (TBTU)
(204 mg, 0.64 mmol). The reaction mixture was stirred at 258C for 10 h,
and then partitioned between brine (15 mL) and ethyl acetate (50 mL).
The organic layer was washed with 0.1n HCl (2ꢇ15 mL) and 5%
NaHCO3 (2ꢇ15 mL). Finally, the organic layer was dried, filtered, and
evaporated to give a residue that was purified by a silica gel column
chromatography (hexane/ethyl acetate, 1:1) to give compound 13 as a
colorless oil (359 mg, 65%). [a]2D5 =+ 74.6 (c=1.00, CHCl3); 1H NMR
(400 MHz, CDCl3): d=1.46 (s, 9H), 2.63 (d, 3H, J=4.7 Hz), 3.51–3.61
(m, 2H), 3.68 (dd, 1H, J1 =9.1 Hz, J2 =6.0 Hz), 3.74 (dd, 1H, J1 =
11.0 Hz, J2 =4.8 Hz), 3.84 (dd, 1H, J1 =9.5 Hz, J2 =5.1 Hz), 3.95 (t, 1H,
J=6.6 Hz), 3.99–4.04 (m, 1H), 4.08–4.17 (m, 1H), 4.26–4.33 (m, 1H),
4.35 (dd, 1H, J1 =10.7 Hz, J2 =2.8 Hz), 4.41 (d, 1H, J=11.6 Hz), 4.44–
4.52 (m, 2H), 4.52–4.70 (m, 5H), 4.81 (d, 1H, J=11.2 Hz), 4.97 (dd, 1H,
J1 =10.6 Hz, J2 =4.1 Hz), 5.34 (d, 1H, J=4.1 Hz), 5.40 (d, 1H, J=
4.1 Hz), 6.54–6.66 (m, 1H), 6.71 (d, 1H, J=8.2 Hz), 7.18–7.39 ppm (m,
20H); 13C NMR (100 MHz, CDCl3): d=26.4, 28.3, 52.9, 55.2, 67.4, 68.2,
69.8, 70.1, 72.8, 73.6, 73.7, 75.1, 80.9, 84.6, 96.7, 127.8, 127.9, 128.0, 128.1,
128.1, 128.4, 128.5, 128.7, 137.2, 137.6, 137.9, 156.1, 169.0, 170.1 ppm; ele-
mental analysis calcd for C46H56N4O12: C 64.47, H 6.59, N 6.54; found: C
64.32, H 6.58, N 6.57.
172.0, 173.9, 174.3, 174.5 ppm; elemental analysis calcd for C27H47N5O15
C 47.57, H 6.95, N 10.27; found: C 48.98, H 6.91, N 10.23.
:
Synthesis of compound 9: 3,4,6-Tri-O-benzyl-2-nitrogalactal (1.10 g,
2.65 mmol), which was prepared as described in the literature,[17b] and Ser
derivative 8[17d] (500 mg, 2.04 mmol) were dissolved in THF (35 mL)
under an argon atmosphere, and molecular sieve was then added. The re-
action mixture was stirred at 258C for 15 min and a 1m potassium tert-
butoxide solution in THF (0.2 mL, 0.20 mmol) was then added. The reac-
tion was stirred for 12 h, the molecular sieve was then filtered off and all
solvents were removed by evaporation. The residue was purified by a
silica gel column chromatography (hexane/ethyl acetate, 4:1) to give 9
(1.07 g, 74%) as a colorless oil. [a]2D5 =+62.4 (c=1.30, CHCl3); 1H NMR
(400 MHz, CDCl3): d=1.45 (s, 9H), 3.48–3.63 (m, 2H), 3.87–3.96 (m,
2H), 3.96–4.05 (m, 2H), 4.34–4.54 (m, 5H), 4.59–4.67 (m, 2H), 4.69–4.76
(m, 2H), 4.82 (d, 1H, J=11.2 Hz), 4.95 (dd, 1H, J1 =10.6 Hz, J2 =
4.2 Hz), 5.23–5.38 (m, 3H), 5.47 (d, 1H, J=8.2 Hz), 5.83–6.00 (m, 1H),
7.17–7.39 ppm (m, 15H); 13C NMR (100 MHz, CDCl3): d=28.3, 53.9,
66.6, 68.1, 70.0, 72.9, 73.1, 73.6, 75.0, 75.2, 80.3, 84.1, 97.1, 119.2, 127.9,
128.1, 128.2, 128.4, 128.6, 131.6, 137.3, 137.7, 137.9, 155.4, 169.4 ppm; ele-
mental analysis calcd for C38H46N2O11 : C 64.58, H 6.56, N 3.96; found: C
64.42, H 6.58, N 3.95.
Synthesis of compound 14: TFA (3 mL) was added to a solution of com-
pound 13 (0.50 g, 0.58 mmol) in CH2Cl2 (6 mL) at 08C. The reaction was
maintained at 08C for 30 min and then warmed to 258C for 2.5 h and
concentrated. The crude was dissolved in pyridine (5 mL) and acetic an-
hydride (2 mL) was added. The resulting mixture was stirred for 3 h at
258C. The solvent was evaporated and the crude purified by a silica gel
column chromatography (CH2Cl2/MeOH, 15:1) to give 14, as a colorless
oil (386 mg, 83%). [a]2D5 =+54.3 (c=1.00, CHCl3); 1H NMR (400 MHz,
CDCl3): d=1.99 (s, 3H), 2.53 (d, 3H, J=4.4 Hz), 3.44–3.52 (m, 2H),
3.58–3.68 (m, 1H), 3.68 (dd, 1H, J1 =11.0 Hz, J2 =4.3 Hz), 3.80–3.91 (m,
2H), 3.93–3.98 (m, 1H), 4.06 (dd, 1H, J1 =11.0 Hz, J2 =3.2 Hz), 4.27–4.64
(m, 10H), 4.74 (d, 1H, J=11.2 Hz), 4.90 (dd, 1H, J1 =10.6 Hz, J2 =
3.8 Hz), 5.25 (d, 1H, J=3.6 Hz), 6.33 (d, 1H, J=6.1 Hz), 6.46–6.53 (m,
1H), 6.60 (d, 1H, J=8.3 Hz), 7.16–7.33 ppm (m, 20H); 13C NMR
(100 MHz, CDCl3): d=23.2, 26.3, 53.0, 53.6, 67.1, 68.1, 69.3, 70.1, 72.7,
72.7, 73.6, 73.7, 75.0, 75.0, 84.6, 96.5, 127.8, 127.9, 127.9, 128.0, 128.1,
128.1, 128.2, 128.3, 128.5, 128.5, 128.6, 137.0, 137.1, 137.5, 137.8, 168.9,
169.7, 171.1 ppm; elemental analysis calcd for C43H50N4O11: C 64.65, H
6.31, N 7.01; found: C 64.25, H 6.28, N 6.98.
Synthesis of compound 10: TFA (2 mL) was added to a solution of com-
pound 9 (0.50 g, 0.71 mmol) in CH2Cl2 (5 mL) at 08C. The reaction was
maintained at 08C for 30 min, at 258C for 2.5 h and concentrated. The
crude compound 10 was quantitatively obtained and used without further
purification. 1H NMR (400 MHz, CD3OD): d=3.44–3.56 (m, 2H), 3.83–
3.92 (m, 1H), 3.95–4.09 (m, 3H), 4.25–4.45 (m, 4H), 4.49–4.73 (m, 6H),
4.83–4.91 (m, 1H), 5.16–5.34 (m, 3H), 5.80–5.95 (m, 1H), 7.06–7.29 ppm
(m, 15H); 13C NMR (100 MHz, CD3OD): d=54.1, 67.6, 68.6, 69.9, 71.8,
73.8, 74.6, 74.7, 75.9, 76.3, 85.4, 98.4, 120.2, 128.9, 129.0, 129.1, 129.2,
129.3, 129.4, 129.5, 132.5, 139.0, 139.1, 139.4, 167.7 ppm.
Synthesis of compound 12: A solution of the commercially available acid
11 (384 mg, 1.3 mmol) in acetonitrile (30 mL) was treated with DIEA
(0.7 mL, 4.2 mmol), compound 10 (721 mg, 1.0 mmol), and (benzotriazol-
1-yl)-1,1,3,3-tetramethyluronium TBTU (0.42 g, 1.3 mmol). The reaction
mixture was stirred at 258C for 10 h and then partitioned between brine
(20 mL) and ethyl acetate (50 mL). The organic layer was washed with
0.1n HCl (2ꢇ20 mL) and 5% NaHCO3 (2ꢇ15 mL). Finally, the organic
layer was dried, filtered, and evaporated to give a residue that was puri-
fied by a silica gel column chromatography (hexane/ethyl acetate, 4:1) to
give 12 as a colorless oil (860 mg, 69%). [a]2D5 =+63.0 (c=1.40, CHCl3);
1H NMR (400 MHz, CDCl3): d=1.44 (s, 9H), 3.46–3.55 (m, 2H), 3.59
(dd, 1H, J1 =9.2 Hz, J2 =6.6 Hz), 3.82–3.89 (m, 1H), 3.89–4.00 (m, 4H),
4.30 (dd, 1H, J1 =10.6 Hz, J2 =1.9 Hz), 4.34–4.45 (m, 3H), 4.49 (d, 1H,
J=11.8 Hz), 4.53–4.71 (m, 6H), 4.71–4.82 (m, 2H), 4.91 (dd, 1H, J1 =
10.6 Hz, J2 =4.2 Hz), 5.24–5.37 (m, 3H), 5.44–5.55 (m, 1H), 5.83–5.98 (m,
1H), 7.16–7.37 ppm (m, 21H); 13C NMR (100 MHz, CDCl3): d=28.3,
52.7, 53.9, 66.6, 68.2, 69.3, 69.9, 70.2, 72.9, 73.3, 73.5, 74.9, 75.1, 80.1, 84.0,
97.0, 119.3, 127.6, 127.8, 127.8, 127.9, 128.0, 128.1, 128.3, 128.4, 128.5,
128.5, 131.4, 137.2, 137.5, 137.6, 137.8, 155.5, 168.6, 170.3 ppm; elemental
analysis calcd for C48H57N3O13: C 65.22, H 6.50, N 4.75; found: C 65.02,
H 6.51, N 4.73.
Synthesis of compound SS*: Platinized Raney-nickel (T4) catalyst was
freshly prepared as described in the literature.[27] The catalyst obtained
by using 2 g of Raney nickel/aluminum alloy was suspended in ethanol
(10 mL) and pre-hydrogenated for 10 min before the addition of com-
pound 14 (200 mg, 0.25 mmol) in ethanol (5 mL). The reaction mixture
was shaken under H2 (1 atm) for 7 h at 258C. The catalyst was filtered
off and the solvent evaporated. The residue was dissolved in pyridine/
acetic anhydride (2:1, 6 mL) and stirred at 258C for 3 h. After removing
the volatile products, a solution of the crude products in MeOH/ethyl
acetate (2:1, 6 mL) was treated with 10% palladium–carbon (15 mg) as a
catalyst. The reaction mixture was shaken under H2 (1 atm) for 12 h at
258C. Removal of the catalyst and a further purification of the residue
first by a silica gel column chromatography (aqueous NH3/EtOH/buta-
nol/CHCl3, 8:5:4:2) and then with a C18 reverse-phase sep-pak cartridge
gave SS* as a colorless oil (52 mg, 46%). [a]2D5 =+ 51.6 (c=0.76, H2O/
MeOH, 1:1); 1H NMR (400 MHz, D2O): d=2.05 (s, 3H), 2.08 (s, 3H),
2.77 (s, 3H), 3.73–3.78 (m, 2H), 3.81–3.93 (m, 5H), 3.96 (dd, 1H, J1 =
10.8 Hz, J2 =4.6 Hz), 3.98–4.01 (m, 1H), 4.17 (dd, 1H, J1 =11.0 Hz, J2 =
3.5 Hz), 4.50 (t, 1H, J=5.7 Hz), 4.60–4.65 (m, 1H), 4.91 (d, 1H, J=
3.5 Hz); 1H NMR (400 MHz, H2O/D2O): d=7.98–8.04 (m, 2H, NHs(II)
+ NH3), 8.35 (d, 1H, J=6.6 Hz, NH1), 8.55 (d, 1H, J=7.2 Hz, NH2),
3872
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3863 – 3874

Conformation and Dynamics of the of Tn Antigen
FULL PAPER
8.62 ppm (d, 1H, J=9.1 Hz, NH3); 13C NMR (100 MHz, D2O): d=24.1,
24.4, 28.3, 52.1, 56.2, 57.8, 63.4, 63.6, 69.2, 69.9, 70.8, 73.7, 100.1, 173.9,
174.7, 176.9, 176.9 ppm; elemental analysis calcd for C17H30N4O10: C
45.33, H 6.71, N 12.44; found: C 45.29, H 6.68, N 12.42.
Acknowledgements
We thank the Ministerio de Educaciꢂn y Ciencia and FEDER (project
CTQ2006–05825/BQU and Ramꢂn y Cajal contract to F. C.) and the Uni-
versidad de La Rioja (grant to M.G.L.). We also thank CESGA for com-
puter support. We appreciate the comments of the referees, which have
resulted in an improvement in the manuscript.
NMR spectroscopic experiments: All the NMR spectroscopic experi-
ments were recorded on a Bruker Avance 400 spectrometer at 293 K. 1H
and 13C NMR spectra were recorded in CDCl3 and CD3OD with TMS as
the internal standard and in D2O (chemical shifts are reported in ppm on
the d scale). Magnitude-mode ge-2D COSY spectra were recorded with
gradients and by using the cosygpqf pulse program with 90 degree pulse
width. Phase-sensitive ge-2D 1H-13C edited-HSQC spectra[28] were record-
ed by 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 by using phase-sensi-
tive ge-2D NOESY for CDCl3 spectra and phase-sensitive ge-2D
NOESY with WATERGATE for H2O/D2O (9:1) spectra. Selective ge-
1D NOESY experiments were carried out by using the 1D-DPFGE NOE
pulse sequence.
[1] a) F.-G. Hanisch, S. Mꢆller, Glycobiology 2000, 10, 439–449; b) S. J.
Gendler, J. Mammary Gland Biol. Neoplasia 2001, 6, 339–353.
[2] S. H. Itzkowitz, M. Yuan, C. K. Montgomery, T. Kjeldsen, H. K. Ta-
kahashi, W. L. Bigbee, Y. S. Kim, Cancer Res. 1989, 49, 197–204.
[3] G. F. Springer, J. Mol. Med. 1997, 75, 594–602.
[4] a) K. Terasawa, H. Furumoto, M. Kamada, T. Aono, Cancer Res.
1996, 56, 2229–2232; b) J. Taylor-Papadimitriou, J. Burchell, D. W.
Miles, M. Dalziel, Biochim. Biophys. Acta 1999, 1455, 301–313.
[5] a) J. N. Blattman, P. D. Greenberg, Science 2004, 305, 200–205; b) P.
Moingeon, Vaccine 2001, 19, 1305–1326; c) S. J. Danishefsky, J. R.
Allen, Angew. Chem. 2000, 112, 882–912; Angew. Chem. Int. Ed.
2000, 39, 836–863; d) S. Jarcz, J. Chen, L. V. Kuznetsova, I. Ojima,
Bioorg. Med. Chem. 2005, 13, 5043–5054; e) S. Dziadek, A. Hobel,
E. Schmitt, H. Kunz, Angew. Chem. 2005, 117, 7803–7808; Angew.
Chem. Int. Ed. 2005, 44, 7630–7635; f) T. Buskas, S. Ingale, G.-J.
Boons, Glycobiology 2006, 16, 113R–136R; g) D. H. Dube, C. R.
Bertozzi, Nat. Rev. 2005, 4, 477–488; h) M. A. Hollingsworth, B. J.
Swanson, Nat. Rev. Cancer 2004, 4, 45–60; i) C. Brocke, H. Kunz,
Bioorg. Med. Chem. 2002, 10, 3085–3112; j) Y. J. Kim, A. Varki,
Glycoconjugate J. 1997, 14, 569–576; k) M. A. Tarp, H. Clausen,
Biochim. Biophys. Acta 2008, 1780, 546–563; l) B. Acres, J.-M. Li-
macher, Expert Rev. Vaccines 2005, 4, 493–502; m) F.-G. Hanisch,
Biochem. Soc. Trans. 2005, 33, 705–708; n) F.-G. Hanisch, T. Nin-
kovic, Curr. Protein Pept. Sci. 2006, 7, 307–315; o) T. Becker, S.
Dziadek, S. Wittrock, H. Kunz, Curr. Cancer Drug Targets 2006, 6,
491–517.
MD-tar simulations: MD-tar simulations were performed with
AMBER 6.0[29] (parm94),[30] which was implemented with GLYCAM 04
parameters[31] to accurately simulate the conformational behavior of the
sugar moiety. NOE-derived distances were included as time-averaged dis-
tance constraints and scalar coupling constants 3J as time-averaged cou-
pling constraints.
For each restraint (distance or coupling constant), six main constants (r1
to r4, rk2 and rk3) have to be defined. These parameters define the
ꢅshapeꢅ of the restraining potential (Figure 11): R is the experimental dis-
tance or 3J couplings. The values of r1 to r4 have to be specified in ꢄ or
Hz, depending on the restraint type. rk2 and rk3 are defined in kcalmo-
lÀ1 ꢄÀ2 for distance restraints and in kcalmolÀ1 radÀ2 for dihedral re-
straints. The experimental key proton-pair distances and 3J couplings
were implemented as structural restraints with a 10% margin by using a
flat well potential as described above.
[6] G. Ragupathi, L. Howard, S. Cappello, R. R. Koganty, D. X. Qiu,
B. M. Longenecker, M. A. Reddish, K. O. Lloyd, P. O. Livingston,
Cancer Immunol. Immunother. 1999, 48, 1–8.
[7] a) S. Dziadek, C. Griesinger, H. Kunz, U. M. Reinscheid, Chem.
Eur. J. 2006, 12, 4981–4993; b) D. M. Coltart, A. K. Royyuru, L. J.
Williams, P. W. Glunz, D. Sames, S. D. Kuduk, J. B. Schwarz, X. T.
Chen, S. J. Danishefsky, D. H. Live, J. Am. Chem. Soc. 2002, 124,
9833–9844; c) J. Schuman, D. X. Qiu, R. R. Koganty, B. M. Longe-
necker, A. P. Campbell, Glycoconjugate J. 2000, 17, 835–848; d) Y.
Mimura, Y. Inoue, N. J. Maeji, R. Chujo, Int. J. Pept. Protein Res.
1989, 34, 363–368.
[8] D. I. R. Spencer, M. R. Price, S. J. B. Tendler, C. I. DeMatteis, T.
Stadie, F.-G. Hanisch, Cancer Lett. 1996, 100, 11–15.
[9] G. I. Csonka, G. A. Schubert, A. Perczel, C. P. Sosa, I. G. Csizmadia,
Chem. Eur. J. 2002, 8, 4718–4733.
[10] a) F. Corzana, J. H. Busto, G. Jimꢁnez-Osꢁs, J. L. Asensio, J. Jimꢁ-
nez-Barbero, J. M. Peregrina, A. Avenoza, J. Am. Chem. Soc. 2006,
128, 14640–14648; b) F. Corzana, J. H. Busto, S. B. Engelsen, J. L.
Asensio, J. Jimꢁnez-Barbero, J. M. Peregrina, A. Avenoza, Chem.
Eur. J. 2006, 12, 7864–7871.
[11] A. Kuhn, H. Kunz, Angew. Chem. 2007, 119, 458–462; Angew.
Chem. Int. Ed. 2007, 46, 454–458.
[12] P. Braun, G. M. Davies, M. R. Price, P. M. Williams, S. J. B. Tendler,
H. Kunz, Bioorg. Med. Chem. 1998, 6, 1531–1545.
Figure 11. The violation energy is a well with a square bottom between r2
and r3, with parabolic sides out to a defined distance or J (r1 and r4 for
lower and upper bounds, respectively).
3
The input-file used to start a MD-tar simulation consists of two parts.
The first part is an unrestrained energy minimization of 1000 steps of
steepest-descents minimization followed by 1000 steps of conjugate-gradi-
ent minimization. The second part contains a 80 ns molecular dynamics
simulation at 300 K in which the distance restraints are time-averaged ac-
cording to RÀ6 by using a exponential memory function with t equal to
8 ns. Similarly, scalar coupling restraints are time averaged according to J
by using the same memory function. To assure a smooth run of the calcu-
lation, both types of restraints are introduced slowly over the first 1 ns of
the calculation.
All the simulations were performed by using a dielectric constant of 80
to simulate the water environment. Additionally, a 20 ns MD-tar simula-
tion in explicit water (TIP3P water molecules[32]) on glycopeptide S*S
was carried out and obtained a similar result (see the Supporting Infor-
mation).
3JHa,Hb and 3JNH,Ha coupling constants were estimated by using the Karplus
equations given in reference [33a,b], respectively.
[13] E. Osinaga, S. Bay, D. Tello, A. Babino, O. Pritsch, K. Assemat, D.
Cantacuzene, H. Nakada, P. Alzari, FEBS Lett. 2000, 469, 24–28.
[14] R. Lo-Man, S. Vichier-Guerre, R. Perraut, E. Dꢁriaud, V. Huteau,
L. BenMohamed, O. M. Diop, P. O. Livingston, S. Bay, C. Leclerc,
Cancer Res. 2004, 64, 4987–4994.
[15] F. Corzana, J. H. Busto, G. Jimꢁnez-Osꢁs, M. Garcꢀa de Luis, J. L.
Asensio, J. Jimꢁnez-Barbero, J. M. Peregrina, A. Avenoza, J. Am.
Chem. Soc. 2007, 129, 9458–9467.
[16] H. Waldmann, H. Kunz, Liebigs Ann. Chem. 1983, 1712–1715.
Chem. Eur. J. 2009, 15, 3863 – 3874
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3873

F. Corzana, J. M. Peregrina et al.
[17] a) G. A. Winterfeld, Y. Ito, T. Ogawa, R. R. Schmidt, Eur. J. Org.
Chem. 1999, 1167–1171; b) G. A. Winterfeld, R. R. Schmidt, Angew.
Chem. 2001, 113, 2718–2721; Angew. Chem. Int. Ed. 2001, 40, 2654–
2657; c) R. R. Schmidt, Y. D. Vankar, Acc. Chem. Res. 2008, 41,
1059–1073; d) P. Baumhof, R. Mazitschek, A. Giannis, Angew.
Chem. 2001, 113, 3784–3786; Angew. Chem. Int. Ed. 2001, 40, 3672–
3674.
[18] a) H. Kunz, H. Waldmann, Angew. Chem. 1984, 96, 49–50; Angew.
Chem. Int. Ed. Engl. 1984, 23, 71–72; b) H. Kunz, C. Unverzagt,
Angew. Chem. 1984, 96, 426–427; Angew. Chem. Int. Ed. Engl.
1984, 23, 436–437.
[19] D. S. Wishart, B. D. Sykes, F. M. Richards, J. Mol. Biol. 1991, 222,
311–333.
[20] D. S. Wishart, C. G. Bigam, A. Holm, R. S. Hodges, B. D. Sykes, J.
Biol. NMR 1995, 5, 67–81.
[21] H. J. Dyson, P. E. Wright, Annu. Rev. Biophys. Biophys. Chem. 1991,
20, 519–538.
[22] A. Pardi, M. Billeter, K. Wꢆthrich, J. Mol. Biol. 1984, 180, 741–751.
[23] K. Wuthrich, NMR of Proteins and Nucleic Acids, Wiley, New York,
1986.
[26] G. A. Naganagowda, T. L. Gururaja, J. Satyanarayana, M. J. Levine,
J. Pept. Res. 1999, 54, 290–310.
[27] S. Nishimura, Bull. Chem. Soc. Jpn. 1959, 32, 61–64.
[28] T. Parella, F. Sanchez-Ferrando, A. Virgili, J. Magn. Reson. 1997,
126, 274–277.
[29] a) D. A. Pearlman, D. A. Case, J. W. Caldwell, W. R. Ross, T. E.
Cheatham III, S. DeBolt, D. Ferguson, G. Seibel, P. A. Kollman,
Comput. Phys. Commun. 1995, 91, 1–41; b) P. A. Kollman,
AMBER 6, University of California, San Francisco 1999.
[30] W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould, K. M. Merz, Jr.,
D. M. Ferguson, D. C. Spellmeyer, T. Fox, J. W. Caldwell, P. A. Koll-
man, J. Am. Chem. Soc. 1995, 117, 5179–5197.
[31] R. J. Woods, R. A. Dwek, C. J. Edge, B. Fraser-Reid, J. Phys.
Chem. B. 1995, 99, 3832–3846.
[32] W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey,
M. L. Klein, J. Chem. Phys. 1983, 79, 926–935.
[33] a) A. Marco, M. Llinas, K. Wuthrich, Biopolymers 1978, 17, 617–
636; b) G. W. Vuister, A. Bax, J. Am. Chem. Soc. 1993, 115, 7772–
7777.
Received: August 28, 2008
[24] V. F. Bystrov, Prog. Nucl. Magn. Reson. Spectrosc. 1976, 10, 41–81.
[25] D. A. Pearlman, J. Biomol. NMR 1994, 4, 1–16.
Revised: December 10, 2008
Published online: February 19, 2009
3874
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3863 – 3874