Facile synthesis of cyclopeptide-centered multivalent glycoclusters with 'click chemistry' and molecular recognition study by surface plasmon resonance

Article abstract of DOI:10.1016/j.bmcl.2009.04.090

A facile synthesis of cyclopeptide-centered multivalent glycoclusters using Cu(I) catalyzed Huisgen 1,3-dipolar cycloaddition of azides and terminal alkynes, so called 'click chemistry', has been developed. The affinities of mannose-specific protein Conca

Full text of DOI:10.1016/j.bmcl.2009.04.090

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3775–3778
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Facile synthesis of cyclopeptide-centered multivalent glycoclusters with
‘click chemistry’ and molecular recognition study by surface plasmon resonance
*
Yong-Xiang Chen, Lei Zhao, Zhi-Ping Huang, Yu-Fen Zhao, Yan-Mei Li
Key Laboratory of Bioorganic Phosphorous Chemistry & Chemical Biology (Ministry of Education), Department of Chemistry, Tsinghua University, Beijing 100084, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A facile synthesis of cyclopeptide-centered multivalent glycoclusters using Cu(I) catalyzed Huisgen 1,3-
dipolar cycloaddition of azides and terminal alkynes, so called ‘click chemistry’, has been developed. The
affinities of mannose-specific protein Concanavalin A (Con A) toward two synthetic glycoclusters respec-
tively bearing divalent or tetravalent mannoses were investigated by surface plasmon resonance (SPR). It
is founded that the tetravalent glycocluster has 3.0-fold increase in binding affinity relative to the diva-
lent glycoluster (valency-corrected values), which indicates the potential of this system in investigating
carbohydrate–protein interactions.
Received 2 February 2009
Revised 19 April 2009
Accepted 21 April 2009
Available online 24 April 2009
Keywords:
Glycocluster
Cyclopeptide
Ó 2009 Elsevier Ltd. All rights reserved.
Carbohydrate–protein interaction
Click chemistry
SPR
The interactions between carbohydrates and proteins immedi-
ate many important biological processes, such as: fertilization, dif-
ferentiation, development, tissue formation, cell adhesion, antigen/
antibody interactions, cancer metastasis, and infection of viruses or
bacteria.1,2 However, monovalent carbohydrate–protein interac-
tions have been characterized by very weak binding affinities, with
the Kd value in 10ꢀ3–10ꢀ6 M range.2 In nature, multivalent presen-
tation of sugar ligands frequently enhances the affinity by simulta-
neous formation of multiple carbohydrate–protein complexes,
which has been referred as ‘glycoside cluster effects’.3,4 In order
to facilitate multivalent carbohydrate protein interactions study,
multivalent glycoclusters presented on a great variety of well-de-
fined scaffolds such as cyclopeptides,5–11 polymers,12 dentrimers,13
cyclodextrins,14–18 calix[4]arene, adamantane and benzene scaf-
folds,19 monosaccharide,20,21 and self-assemble supermoleculars22
have been synthesized. Cyclopeptides might be excellent cores of
multivalent glycoclusters, due to the flexibility of the synthetic
strategy, easy control of size and spatial orientation, the ease of
incorporating amino acids with different functionalities for attach-
ment of linkers or ligands, and their biological compatibility.
It is well known that the efficient attachment of sugar ligands to
scaffolds is pivotal for preparation of multivalent glycoclusters. For
cyclopeptide scaffolds, different coupling methods have been
developed. Dumy’s group6,10 has selected the oxime bond forma-
tion for the ligation of carbohydrates and cyclopeptides. Ohta
and co-workers5 have used enzymic method to complete the con-
struction of multivalent glycoconjugates. Wittmann’s group11 has
used the formation of urethane bond for the construction of multi-
valent neoglycopeptides. In recent years, a highly chemoselective
and efficient ligation method ‘Cu(I) catalyzed Huisgen 1,3-dipolar
cycloaddition of azides and terminal alkynes’, so called ‘click chem-
istry’, has been developed by Sharpless23 and Meldal24 group and
exploited for the assembling of different glycosyls and various scaf-
folds.19,20,25–29 Therefore, we planed to utilize this highly efficient
ligation method for assembling cyclopeptide scaffolds and carbo-
hydrate ligands. Since the oligosaccharides containing azido
groups are readily accessible,26,30 the installation of multi alkynes
in cyclopeptide side chains has to be considered carefully.
In recent years, Dumy’s group10 has developed lysine-contain-
ing cyclodecapeptides called ‘regioselectively addressable func-
tionalized templates’ (RAFTs) for multivalent presentation of
carbohydrates. Based on this cyclic decapeptide scaffold presenting
tetra-aldehydes for ‘oxime bond formation’ ligation strategy, a new
cyclopeptide scaffold 1 bearing alkyne groups instead of aldehydes
in the side chains of lysine residues was designed (Scheme 1). The
amine in the side chain of lysine 9 that has been used to conjugate
oligonucleotide in Singh’s work10 would be kept free as further
derivation site. In addition, to determine the affinity efficiency of
multivalent glycocluster with carbohydrate-binding protein, a
cyclopeptide scaffold with di-alkynes 2 would be synthesized.
The side chain amines of lysine 5 and 10 in scaffold 2 were capped
by acetyl groups.
In this Letter, we report on a facile synthesis of two cyclopep-
tide scaffolds presenting di- or tetra-alkynes and the chemoselec-
* Corresponding author. Fax: +86 10 6278 1695.
E-mail address: liym@mail.tsinghua.edu.cn (Y.-M. Li).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.04.090

3776
Y.-X. Chen et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3775–3778
R4
R4
HN
R1
HN
R3
R1
HN
R3
R2
R2
NH
HN
HN
NH
HN
OH
OH
K
4
A
O
K
K
4
A
K
5
6
HO
HO
3
G
5
6
3
P
2
G
P
2
N3
O
9
K
K
K
1
K
9
K
7
K
P
1
8
7
10
G
P
8
10
G
Cu(I)
NH2
NH2
O
N
N
R3, R4 =
N
O
N
R1, R2 =
R1 =
1
8
9
sugar
R4 =
O
O
R2 =
O
2
R3 =
O
N
N
sugar
Scheme 1. The assembly of cyclopeptide scaffolds 1 and 2 with azido sugar.
tive and efficient assembly of unprotected
a
-mannose unit con-
cent, radioactive, or enzymatic reporter groups labeling and the
ability to evaluate interactions under dynamic conditions.
taining azido group to cyclopeptide scaffolds through Cu (I) cata-
lyzed Huisgen 1, 3-dipolar cycloaddition (Scheme 1).
Furthermore, these synthesized glycoclusters’ binding abilities to
the mannose-specific lectin Concanavalin A (Con A) have been
evaluated by surface plasmon resonance (SPR), which was proven
particularly suitable for the study of carbohydrate–protein interac-
tions31–33 due to low sample demands, being in needless of fluores-
The cyclopeptide scaffold 1 has been prepared following the
facile synthetic route in Scheme 2. Firstly, the linear peptide has
been assembled on a highly acid labile 2-chlorotritylchloride resin
by standard Fmoc solid phase peptide chemistry with commer-
cially available amino acids referred to Dumy’s strategy.10 One of
the advantages of cyclopeptide scaffold for glycoclusters is the ease
Cl
Cl
2-Chlorotrityl Chloride resin
Fmoc-Gly-OH/DIPEA/DCM
Fmoc-Gly-linker
1.CH3OH/DIPEA/DCM
2.Piperidine/DMF
3.Fmoc-AA-OH/HOBt/HBTU/DIPEA
Fmoc-Lys(Aloc)-Lys(Boc)-Lys(Aloc)-Pro-Gly-Lys(Aloc)-Ala-Lys(Aloc)-Pro-Gly-linker
1.Pd(Ph3P)4/PhSiH3/DCM
2.CHCCOOH/DCC/DCM
Fmoc-Lys(R)-Lys(Boc)-Lys(R)-Pro-Gly-Lys(R)-Ala-Lys(R)-Pro-Gly-linker
1.DBU/Piperidine/DMF
2.HFIP/DCM
H-Lys(R)-Lys(Boc)-Lys(R)-Pro-Gly-Lys(R)-Ala-Lys(R)-Pro-Gly-OH
PyBOP/DIPEA/DMF
3
Lys(R)-Lys(Boc)-Lys(R)-Pro-Gly-Lys(R)-Ala-Lys(R)-Pro-Gly
O
TFA/DCM
R:
Lys(R)-Lys-Lys(R)-Pro-Gly-Lys(R)-Ala-Lys(R)-Pro-Gly
1
Scheme 2. Synthesis route of cyclopeptide scaffold 1 presenting four alkynes.

Y.-X. Chen et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3775–3778
3777
of incorporating amino acids with orthogonally protected side
chains. Therefore, four lysine residues at the positions of 3, 5, 8
and 10 with Aloc protected side chains served as carbohydrates
attachment points and the lysine 9 bearing Boc protection group
could provide a free amine for further derivation. After the comple-
tion of peptide chain elongation, the Aloc groups were removed
with a catalytic amount of Pd(Ph3P)4 and 48 equiv PhSiH3 in
dichloromethane (DCM) to give four free amines. Under this neu-
tral condition, both of the Boc group and the linker between pep-
tide and resin were intact. Afterwards, propiolic acids were
efficiently coupled with four amines of the peptide on solid sup-
port using N,N0-dicyclohexylcarbodiimide (DCC) coupling method
in DCM, monitored by Kaiser’s test. After the deprotection of Fmoc
at the N-terminal of peptide by 2% 1,8-diazabicyclo[5.4.0]undec-7-
ene (DBU)/2% piperidine in dimethylformamide (DMF), the linear
peptide 3 was released from resin using 20% hexafluoroisopropa-
nol (HFIP) in DCM instead of the normal cleavage agent (1/1/8
AcOH/TFE/DCM) for this resin, due to trouble in removing acetic
acid. Under this cleavage condition, Boc group in the side chain
of lysine was intact. After precipitation from diethyl ether, fol-
lowed by purification with semi-preparative HPLC, the pure linear
peptide 3 was obtained. Subsequently, the ‘head to tail’ cyclization
of peptide 3 was performed in DMF at low concentration (less than
0.5 mM) using 1.0 equiv PyBOP and 5.0 equiv diisopropylethyl-
amine (DIPEA). Further deprotection of Boc at lysine 9 by treat-
ment with 50% trifluoroacetic acid (TFA) in DCM followed by
purification with semi-preparative HPLC afforded the final cyclo-
peptide scaffold 1 bearing four alkynes in the ‘arms’, confirmed
by ESI-MS (in the Supplementary data). In addition, the cyclopep-
tide scaffold 2 bearing di-alkynes was also synthesized following
the similar procedures as scaffold 1. Only the lysine residues with
acetyl capped side chains were substituted of lysine residues with
Aloc protected at positions of 5 and 10 during the assembling of
the corresponding linear peptide 4. The acetyl groups kept stable
in the following synthetic steps.
MS, which indicated that the tris–triazole adduct 9 was formed
quantitatively within 1 h. The yield after semi-preparative HPLC
purification was 53%. The analytical HPLC and ESI-MS spectra of
9 are shown in Figures S5 and S6 (Supplementary data). This con-
dition was also suitable for the tetra-alkynyl cyclopeptides 1 to
give the corresponding tetravalent glycoclusters 8. The yield after
semi-preparative HPLC purification was 44%. The analytical HPLC
and MALDI-TOF-MS spectra of 8 are shown in Figures S7 and S8
(Supplementary data).
The affinities of the prepared cyclopeptide-centered glycoclus-
ters 8 and 9 towards Con A were measured by surface plasmon res-
onance (SPR) on a Biacore 3000 instrument with CM5 sensor chip.
Since the SPR response is proportional to the accumulation of mass
on the sensor surface, in an initial attempt we would like to immo-
bilize the synthetic glycoclusters on the sensor chip through the
free amine at lysine 9 of cyclopeptide scaffolds. However, the ob-
served non-specific binding between Con A and CM5 sensor chip
in the test measurement resulted in the switch to immobilize
Con A on the chip instead of glycoconjugates which were found
no non-specific binding to chip surface. After Con A was immobi-
lized on one channel of a CM5 sensor chip (3600 RU), each cyclo-
peptide-centered glycocluster at a series of concentrations was
respectively injected over this channel to quantitatively character-
Tetravalent glycocluster 8
RU
64 uM
32 uM
16 uM
8 uM
4 uM
2 uM
1 uM
The unprotected a-mannose unit containing azido group 5 was
easily generated (Scheme 3) starting with 2,3,4,6-tetra-O-benzoyl-
mannopyranosyl trichloroacetimidate 634, which was coupled to
the spacer 2-bromoethan-1-ol in DCM using trimethylsilyl trifluo-
romethanesulfonate (TMSOTf) as a catalyst. Then the substitution
of bromo group by azido group from sodium azide in DMF followed
-50
0
50
Time
100
150
s
RU
Divalent glycocluster 9
by deacetylation afforded 2-azidoethyl a-D-mannopyranoside 5.
512 uM
256 uM
128 uM
With cyclopeptides bearing alkynes and the azido sugar in
hand, we explored the conjugation conditions. Cu(OAc)2 and the
reducing agent sodium ascorbate were chosen for the generation
of Cu(I) in situ. Cyclopeptide 2 containing two alkynes was firstly
investigated. To a solution of 2 and 5 in water and methanol
(2:1, v/v) were added 0.4 equiv Cu(OAc)2 and 0.8 equiv sodium
ascorbate. To accelerate the conjugation rate, the cycloaddition
reaction was carried at 37 °C by ultrasonic irradiation, which is
an important technique in organic synthesis in terms of reductions
in reaction times, improved yields, relative to traditional thermal
heating.35,36 Recently, Raghunathan37 group has also reported on
the application of ultrasonic irradiation to 1,3-dipolar cycloaddi-
tion reactions of azomethine ylides with high efficiency under mild
reaction. Our reaction was monitored by analytical HPLC and ESI-
64 uM
32 uM
16 uM
8 uM
-50
0
50
Time
100
150
s
Figure 1. Sensorgrams of binding of glycoclusters at different concentrations over
immobilized Con A. (a) for tetravalent glycocluster 8; (b) for divalent glycocluster 9.
OH
OBz
OBz
OH
1. BrCH2CH2OH,
TMSOTf, CH2Cl2
OBz
O
OBz
O
MeONa
MeOH
O
HO
HO
BzO
BzO
NH
BzO
BzO
2. NaN3, DMF
CCl3
O
O
O
N3
N3
5
7
6
Scheme 3. Synthesis route of 2-azidoethyl a-D-mannopyranoside 5.

3778
Y.-X. Chen et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3775–3778
system in investigating carbohydrate–protein interactions. The
multivalent cyclopeptide-centered glycoclusters bearing more
complex oligosaccharides have been also prepared through this
strategy in our lab using the readily accessible azido oligosaccha-
rides, which will be reported in due time. In addition, the flexible
peptide synthesis provides the chance to develop analogous cyclo-
peptide scaffolds with various size and spatial orientation for pre-
senting multivalent carbohydrates to be suitable for the
corresponding carbohydrate-binding proteins.
Acknowledgements
This work was supported by the National Natural Science Foun-
dation of China (Projects 20532020). We would thank Ms. Chen
Yuanyuan (Institution of Biophysics, Chinese Academic of Science)
for SPR support.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.04.090.
References and notes
1. Bertozzi, C. R.; Kiessling, L. L. Science 2001, 291, 2357.
2. Lis, H.; Sharon, N. Chem. Rev. 1998, 98, 637.
3. Lee, Y. C.; Lee, R. T. Acc. Chem. Res. 1995, 28, 321.
4. Mammen, M.; Choi, S. K.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37,
2755.
5. Ohta, T.; Miura, N.; Fujitani, N.; Nakajima, F.; Niikura, K.; Sadamoto, R.; Guo, C.
T.; Suzuki, T.; Suzuki, Y.; Monde, K.; Nishimura, S. I. Angew. Chem., Int. Ed. 2003,
42, 5186.
6. Renaudet, O.; Dumy, P. Org. Lett. 2003, 5, 243.
Figure 2. Equilibrium response as a function of glycoclusters concentration. (a) for
tetravalent glycocluster 8; (b) for divalent glycocluster 9. Experimental data (black
squares) were fitted by using the steady state model to obtain values for the
thermodynamic association constants.
7. Renaudet, O.; Dumy, P. Tetrahedron Lett. 2004, 45, 65.
8. Renaudet, O.; Dumy, P. Bioorg. Med. Chem. Lett. 2005, 15, 3619.
9. Renaudet, O.; Dumy, P. Org. Biomol. Chem. 2006, 4, 2628.
10. Singh, Y.; Renaudet, O.; Defrancq, E.; Dumy, P. Org. Lett. 2005, 7, 1359.
11. Wittmann, V.; Seeberger, S. Angew. Chem., Int. Ed. 2000, 39, 4348.
12. Ambrosi, M.; Cameron, N. R.; Davis, B. G.; Stolnik, S. Org. Biomol. Chem. 2005, 3,
1476.
ize its interaction with Con A. The concentration-dependent over-
lay sensorgrams of 8 and 9 were recorded in Figure 1. It is shown
that the mannose—Con A interactions are rapid association and
dissociation processes and a plateau was reached before the disso-
ciation phase for each injection. Thus, affinity constants for both of
two glycocluster were calculated by using a general steady state
model that assumes the system achieved equilibrium during sam-
ple injections.31 A plot of the response signal at equilibrium as a
function of concentration for each of glycoclusters is adjusted to
a hyperbolic equation from which the thermodynamic affinity con-
stants can be obtained as described in Figure 2. The KA values of
divalent and tetravalent glycoclusters were respectively
1.35 ꢁ 104 Mꢀ1 and 8.01 ꢁ 104 Mꢀ1, indicative of 3.0-fold increases
in binding affinity for tetravalent 8 relative to divalent 9 (valency-
corrected values). So, it can be inferred that this tetravalent glyco-
cluster is effective in improving the weak carbohydrate–protein
interaction through ‘glycoside cluster effects’.
In summary, we have developed a facile strategy for the synthe-
sis of cyclopeptide-centered glycoclusters. The cyclopeptide scaf-
folds bearing di- or tetra-alkynes have been synthesized for
highly chemoselective and efficient attachment of azido mono-
mannose through Cu(I) catalyzed Huisgen 1,3-dipolar cycloaddi-
tion. Furthermore, the affinity of these two synthetic multivalent
glycoclusters 8 and 9 to Con A have been determined by SPR tech-
nique. It is observed that tetravalent glycocluster 8 has 3.0-fold in-
crease in binding affinity relative to divalent glycocluster 9
(valency-corrected values), which indicates the potential of this
13. Thoma, G.; Streiff, M. B.; Katopodis, A. G.; Duthaler, R. O.; Voelcker, N. H.;
Ehrhardt, C.; Masson, C. Chem. Eur. J. 2005, 12, 99.
14. Andre, S.; Kaltner, H.; Furuike, T.; Nishimura, S. I.; Gabius, H. J. Bioconjugate.
Chem. 2004, 15, 87.
15. Carpenter, C.; Nepogodiev, S. A. Eur. J. Org. Chem. 2005, 3286.
16. Fulton, D. A.; Pease, A. R.; Stoddart, J. F. Isr. J. Chem. 2000, 40, 325.
17. Fulton, D. A.; Stoddart, J. F. J. Org. Chem. 2001, 66, 8309.
18. Fulton, D. A.; Stoddart, J. F. Bioconjugate Chem. 2001, 12, 655.
19. Dondoni, A.; Marra, A. J. Org. Chem. 2006, 71, 7546.
20. Gao, Y. J.; Eguchi, A.; Kakehi, K.; Lee, Y. C. Bioorg. Med. Chem. 2005, 13,
6151.
21. Kohn, M.; Benito, J. M.; Mellet, C. O.; Lindhorst, T. K.; Fernandez, J. M. G. Chem.
Biochem. 2004, 5, 771.
22. Kamiya, N.; Tominaga, M.; Sato, S.; Fujita, M. J. Am. Chem. Soc. 2007, 129, 3816.
23. Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int. Ed.
2002, 41, 2596.
24. Tornoe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67, 3057.
25. Dondoni, A.; Giovannini, P. P.; Massi, A. Org. Lett. 2004, 6, 2929.
26. Lin, H. N.; Walsh, C. T. J. Am. Chem. Soc. 2004, 126, 13998.
27. Joosten, J. A. F.; Tholen, N. T. H.; El Maate, F. A.; Brouwer, A. J.; van Esse, G. W.;
Rijkers, D. T. S.; Liskamp, R. M. J.; Pieters, R. J. Eur. J. Org. Chem. 2005, 3182.
28. Cheshev, P.; Marra, A.; Dondoni, A. Org. Biomol. Chem. 2006, 4, 3225.
29. Gouin, S. G.; Vanquelef, E.; Fernandez, J. M. G.; Mellet, C. O.; Dupradeau, F. Y.;
Kovensky, J. J. Org. Chem. 2007, 72, 9032.
30. Fazio, F.; Bryan, M. C.; Blixt, O.; Paulson, J. C.; Wong, C. H. J. Am. Chem. Soc. 2002,
124, 14397.
31. Vila-Perello, M.; Gallego, R. G.; Andreu, D. Chem. Biochem. 2005, 6, 1831.
32. Smith, E. A.; Thomas, W. D.; Kiessling, L. L.; Corn, R. M. J. Am. Chem. Soc. 2003,
125, 6140.
33. Beccati, D.; Halkes, K. M.; Batema, G. D.; Guillena, G.; de Souza, A. C.; van Koten,
G.; Kamerling, J. P. Chem. Biochem. 2005, 6, 1196.
34. Al-Mughaid, H.; Grindley, T. B. J. Org. Chem. 2006, 71, 1390.
35. Cioffi, E. A. Curr. Top. Med. Chem. 2008, 8, 152.
36. Lepore, S. D.; He, Y. J. Org. Chem. 2003, 68, 8261.
37. Babu, A. R. S.; Raghunathan, R. Tetrahedron Lett. 2007, 48, 6809.