Synthesis of new peptidic glycoclusters derived from β-alanine

Article abstract of DOI:10.1016/j.tetlet.2003.10.073

The synthesis of an asymmetric glycocluster 1 has been achieved by coupling of a sugar unit with the β-alanine polypeptide, the principal chain, and combining a carbohydrate chain with the side chain causing it to branch from the N terminal. The synthesis

Full text of DOI:10.1016/j.tetlet.2003.10.073

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 9331–9335
Synthesis of new peptidic glycoclusters derived from b-alanine
Koji Sato, Noriyasu Hada and Tadahiro Takeda*
Kyoritsu College of Pharmacy, 1-5-30 Shibakoen, Minato-ku, Tokyo 105-8512, Japan
Received 9 July 2003; revised 10 October 2003; accepted 10 October 2003
Abstract—The synthesis of an asymmetric glycocluster 1 has been achieved by coupling of a sugar unit with the b-alanine
polypeptide, the principal chain, and combining a carbohydrate chain with the side chain causing it to branch from the N
terminal. The synthesis of this side chain multivalent ligands is based on the scaffolding of some v-amino acid (glycine, b-alanine,
and GABA) derivatives. This method facilitated the synthesis of the cluster, of which the length of each unit differs.
© 2003 Elsevier Ltd. All rights reserved.
The carbohydrate portions of glycoconjugates in the
cell surface participates at a macromolecular level in
many biological recognition processes, such as those
involving immune defense response, viral replication,
parasite infection, cell–cell adhesion and inflammation.1
The research of many glycobiologists is, therefore, often
concentrated on detailed investigation of carbohydrate–
protein interactions which might eventually lead to the
development of carbohydrate-based drugs and prove to
be an important tool for investigation of their function.
With the carbohydrate chain localized intracellularly to
form a microdomain, reciprocal interactions demon-
strate increased binding affinity with certain proteins
when compared with the individual monomeric unit.2
By using the cluster carbohydrate chain (glycocluster)
in a synthetic carbohydrate, it is reported that the
bioactivation is strengthened. Glycoclusers have proved
to be advantageous in many instances, as the multi-
presentation of a specific sugar epitope in one molecule
can result in remarkably increased adhesion. It is
expected that multivalent carbohydrates would bind to
the cell surface adhesion molecule more tightly than
monovalent ones. The past method for the synthesis of
glycoclusters3 has an advantage over ours in view of the
repeated simultaneous equivalent reactions needed in
ours for the polyvalence functional group to amplify
the carbohydrate chain exponentially. However, in the
previous method only the symmetrical structure could
be taken.
This paper will focus on a newly designed glycocluster
derived from b-alanine. Glycoclusters of this type are
now more than 10 years old, without considering the
pioneering work of Y. C. Lee et al.4 who made
analogous structures based on aspartic acid and that of
R. Roy et al.5 on lysine and glycines. The new glyco-
cluster which we synthesized involves making the b-ala-
Keywords: glycocluster; b-peptide;
hydrate–protein interaction.
D-galactose; unit synthesis; carbo-
* Corresponding author. Tel.: +81(3)54002696; fax: +81(3)54002556;
e-mail: takeda-td@kyoritsu-ph.ac.jp
Scheme 1.
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.10.073

9332
K. Sato et al. / Tetrahedron Letters 44 (2003) 9331–9335
nine polypeptide the principal chain, and combining a
carbohydrate chain with the side chain causing it to
branch from the N terminal. Unlike the cluster using
the usual amino acid, the problem of racemization in
the amide condensation was avoided by using the N
terminal b-alanine. Furthermore, by using some v-
amino acid (glycine, b-alanine, and GABA) it facili-
tated the synthesis of the cluster, of which the length of
each unit differs. We describe the synthesis of new
asymmetric peptide glycoclusters of which the length
partially differs.
ther, the amino group was protected with
p-nitrobenzenesulfonyl (pNBS) chloride to give 3. N-
Carboxymethylation of 3 was carried out by the Mit-
sunobu protocol6 with benzyl glycolate to give
compound 4. Removal of the pNBS group by PhSH
and K2CO3 in DMF at 25°C7 and subsequent t-butoxy-
carbonylation by (Boc)2O gave compound 6. This com-
pound 6 has two carboxylic acid derivatives, and it is a
chemical compound with the features which can expand
the sugar units by being, respectively, independent.
Condensation of 7 with sugar unit 8, which is a simple
galactose derivative, in the presence of typical peptide
coupling conditions [DEPC (diethyl cyanophosphonate)
and Et3N] was carried out for 16 h at 25°C to give the
coupled compound 9.8 On the other hand, coupling of
7 with each v-amino acid (glycine, b-alanine, and
GABA) benzyl ester under DEPC for 16 h at 25°C gave
dipeptide derivative 10, 11 and 12. Subsequent removal
of the benzyl ester of compound 10, 11 and 12 by
catalytic hydrogenolysis over 10% PdꢀC afforded com-
pounds 13, 14 and 15. Furthermore, coupling of 8 with
13, 14 and 15 in the presence of DEPC as described for
9 gave the different length of each side chain of deriva-
tives 16, 17 and 18.9
In this paper we report the synthesis of the glycocluster
1 of dancyl-labelled b-alanine trees, depicted in Scheme
1. The dancyl probe will allow flow cytometric analysis
of the interaction and the uptake of the constructions
by human dendritic cells.
First, the synthesis of the four desired compounds 9,
16, 17, and 18, which were selected as the glycocluster
units, is shown in Scheme 2. Preparation of the N-car-
boxymethyl b-alanine derivative 7 was prepared from
b-alanine by the following six-step procedure. b-Ala-
nine was protected with 2,2,2-trichloroethanol and fur-
Scheme 2. Reagents and conditions: (a) p-Nitrobenzenesulfonyl chloride, Et3N, CH2Cl2, 86%; (b) benzyl glycolate, DEAD, PPh3,
CH2Cl2, 90%; (c) PhSH, K2CO3, DMF, 81%; (d) (Boc)2O, Et3N, CH2Cl2, 94%; (e) Pd-C, H2, THF, 7: 84%, 13: 80%, 14: 99%, 15:
82%; (f) sugar unit 8, DEPC, Et3N, DMF, 9: 78%, 16: 80%, 17: 76%, 18: 80%; (g) v-amino acid benzylester, DEPC, Et3N, DMF,
10: 84%, 11: 82%, 12: 92%.

K. Sato et al. / Tetrahedron Letters 44 (2003) 9331–9335
9333
Scheme 3. Reagents and conditions: (a) Zn–AcOH, 19: 75%, 22: 75%, 23: 89%, 26: 76%; (b) 50% TFA, 20: 79%, 24: 76%, 28: 82%;
(c) DEPC, Et3N, DMF, 21: 98%, 25: 94%; (d) sugar unit 8, DEPC, Et3N, DMF, 88%.
Next, compounds 22 and 28 were selected as the N-
terminal block and C-terminal block, respectively, and
are shown in Scheme 3. Glycocluster unit 19 derived
from 9 by reductive removal of the trichloroechyl
group using Zn–AcOH was treated with another gly-
cocluster unit 20 derived from 16 by removal of the
Boc group using 50% TFA under coupling conditions
as described for 9 to afford cluster block 21. Then,
removal of the trichloroechyl group from 21 by Zn–
AcOH gave the N-terminal block compound 22. On
the other hand, compound 25 was obtained by cou-
pling of 23 from 17 with 24 from 18 in the presence
of DEPC as described for 21, and subsequent
removal of the trichloroechyl group from 25 gave
compound 26. Furthermore, coupling of 26 with
sugar unit 8 gave 27 and removal of the Boc group
using 50% TFA gave the target C-terminal block
compound 28.

9334
K. Sato et al. / Tetrahedron Letters 44 (2003) 9331–9335
Scheme 4. Reagents and conditions: (a) DEPC, Et3N, DMF, 80%; (b) 50% TFA, 83%; (c) dansyl glycine, DEPC, Et3N, DMF,
77%; (d) NaOMe, MeOH, 1,4-dioxane, 94%.
Finally, coupling of 22 with 28 in the presence of
DEPC as described for 21 gave the pentamer derivative
29. The Boc group of 29 was removed under acidic
conditions with 50% TFA giving a free secondary
amino group, which was subsequently treated with
dancyl glycine under DEPC giving 31. Then, complete
de-O-benzoylation of 31 afforded the target compound
1 with free hydroxyl groups on all the asymmetric
glycoclusters.10 (Scheme 4) At this time, each step gave,
respectively, a high yield and there were no by-products
generated in the peptide condensation.
tific Research (No. 14771250) from the Ministry of
Education, Science, Sports and Culture of Japan. We
gratefully acknowledge financial support in the form of
a Sasagawa Scientific Research Grant and the Tokyo
Biochemical Research Foundation Grant. The authors
are grateful to Ms. J. Hada for providing NMR and
MS data.
References
1. (a) Essential of glycobiology; Varki, A., Cummings, R.,
Esko, J., Freeze, H., Haet, G., Marth, J., Eds., Cold
Spring Harbor Laboratory Press: Cold Spring Harbor,
NY, 1991; (b) Molecular Glycobiology; Fukuda, M.,
Hindsgaul, O., Eds.; Oxford University Press: New York,
1994.
2. (a) Lee, Y. C.; Lee, R. T. Acc. Chem. Res. 1995, 28, 321;
(b) Lasky, L. A. Annu. Rev. Biochem. 1995, 64, 113.
3. (a) Roy, R. J. Carbohydr. Chem. 2002, 21, 769; (b)
Langer, P.; Ince, S. J.; Ley, S. V. J. Chem. Soc., Perkin
Trans. 1 1998, 3913; (c) Ashton, P. R.; Boyd, S. E.;
Brown, C. L.; Jayaraman, N.; Stoddart, J. F. Angew.
Chem., Int. Ed. Engl. 1997, 36, 732; (d) Lindhorst, T. K.;
Ludewig, M.; Thiem, J. J. Charbohydr. Chem. 1998, 17,
1131.
4. (a) Lee, Y. C.; Lee, R. T. The glycoconjugates; Horowits,
M. I., Ed.; Academic Press: New York, NY, 1982; Vol.
IV, p. 57; (b) Lee, R. T.; Lee, Y. C. Biochemistry 1984,
23, 4255.
5. (a) Khan, S. H.; O’Neill, R, A. Modern Methods in
Carbohydrate Synthesis; Harwood Academic Publishers;
1996, p. 378; (b) Roy, R. Curr. Opin. Struct. Biol. 1996,
6, 692.
In conclusion, efficient and widely applicable synthetic
strategies in glycoconjugate chemistry have given access
to a new glycocluster. The glycocluster was synthesized
this time in order to optionally change the distance
from the core to the carbohydrate chain and the dis-
tance between carbohydrate chains, thus becoming the
cluster with the optimum structure which fits the struc-
ture of the carbohydrate chain recognition domain, and
higher biological activities are expected.
Typical procedure: To a solution of compound 13 (81.2
mg, 0.19 mmol) and compound 8 (143.0 mg, 0.22
mmol) in DMF (2 mL) were added DEPC (34 mL, 0.22
mmol) and triethylamine (52 mL, 0.37 mmol). The
reaction mixture was stirred for 16 h at 25°C. After
completion of the reaction, the mixture was extracted
with ethyl acetate, washed with water, dried (Mg2SO4),
and concentrated. The product was purified by silica-
gel column chromatography (benzene: acetone=5:1) as
elute to give compound 16 (156.8 mg, 79.6%).
Acknowledgements
6. Mitsunobu, O. Synthesis 1981, 1.
7. Fukuyama, T.; Jow, C.-K.; Cheung, M. Tetrahedron Lett.
This work was supported by a Grant-in-Aid for Scien-
1995, 36, 6373.

K. Sato et al. / Tetrahedron Letters 44 (2003) 9331–9335
9335
8. Physical data for 9: [h]2D3=+59.0° (c=1.9, CHCl3) 1H
NMR (500 MHz, CDCl3): l 8.10–7.23 (20H, m, Ar-H),
6.25 (1H, s, NH), 6.00 (1H, d, H-4), 5.77 (1H, dd,
J=10.4 Hz, H-2), 5.63 (1H, dd, J=3.1 Hz, H-3), 4.87
(1H, d, J=7.9 Hz, H-1), 4.72 (2H, s, CH2CCl3), 4.67 (1H,
dd, H-6a), 4.45 (1H, dd, H-6b), 4.36 (1H, m, H-5), 3.98
and 3.76 (2H, m, OCH2 of sugar unit), 3.65 (2H, s,
NCH2CO), 3.55–3.49 (4H, m, NCH2 of sugar unit, NCH2
of b-alanine), 2.72 (2H, t, J=6.7 Hz, COCH2 of b-ala-
nine), 1.43 (9H, s, t-Bu). 13C NMR (125 MHz, CDCl3): l
169.3, 166.0, 165.5, 165.3, 133.7, 133.5, 133.4, 133.3,
130.0, 129.7, 129.3, 129.2, 129.0, 128.7, 128.6, 128.5,
128.3, 101.7, 94.8, 80.9, 74.1, 71.6, 71.5, 69.8, 68.9, 68.1,
62.1, 51.7, 44.9, 39.2, 33.0, 28.3. MALDI-TOFMS: calcd
for C48H49Cl3N2O15: 998, found 1021[M+Na]+.
(1H, dd, J2,3=10.4 Hz, H-2), 5.65 (1H, dd, J3,4=3.6 Hz,
H-2), 4.84 (1H, d, J1,2=7.9 Hz, H-1), 4.74 (2H, s,
CH2CCl3), 4.68 (1H, dd, H-6a), 4.44 (1H, dd, H-6b), 4.36
(1H, t, H-5), 4.01, 3.70 (2H, m, OCH2 of sugar unit), 3.85
(2H, s, COCH2N), 3.60 (2H, t, NCH2 of b-alanine),
3.57-3.34 (4H, m, NCH2 of sugar unit, NCH2 of b-ala-
nine), 2.78 (2H, t, CH2CO), 2.11 and 1.99 (2H, m,
CH2CO), 1.44 (9H, s, t-Bu). 13C-NMR (125 MHz,
CDCl3): l 171.4, 169.1, 166.0, 165.5, 165.4, 133.4, 130.0,
129.7, 129.7, 129.3, 128.9, 128.7, 128.5, 128.3, 101.6, 94.8,
81.0, 74.0, 71.6, 71.3, 69.9, 68.9, 68.0, 62.0, 52.0, 44.8,
39.0, 35.2, 34.7, 33.2, 28.2. HR-FABMS: calcd for
C51H54Cl3N3O16: 1069.2570, found 1070.2609[M+H]+.
1
Physical data 18 [h]D23=+53.5° (c=1.8, CHCl3) H NMR
(500 MHz, CDCl3): l 8.10–7.23 (20H, m, Ar-H), 6.69
(1H, s, NH), 6.10 (1H, s, NH), 6.01 (1H, d, H-4), 5.77
(1H, dd, H-2), 5.65 (1H, dd, H-3), 4.86 (1H, d, H-1), 4.73
(2H, s, CH2CCl3), 4.69 (1H, dd, H-6a), 4.44 (1H, m,
H-6b), 4.37 (1H, t, H-5), 4.01 and 3.72 (2H, OCH2 of
sugar unit), 3.85 (2H, s, COCH2N), 3.62 (2H, t, NCH2 of
b-alanine), 3.56-3.34 (2H, m, NCH2 of sugar unit), 3.18
(2H, m, NCH2 of GABA), 2.77 (2H, t, CH2CO of
b-alanine), 2.01–1.85 (2H, m, CH2CO of GABA), 1.67
(2H, m, CCH2C), 1.44 (9H, s, t-Bu). 13C NMR (125
MHz, CDCl3): l 172.6, 169.5, 166.0, 165.5, 165.4, 133.64,
133.61, 133.3, 120.0, 129.7, 129.3, 129.04, 128.95, 128.64,
128.61, 128.5, 128.3, 101.6, 94.7, 81.0, 74.1, 71.5, 71.4,
69.9, 69.0, 68.1, 62.0, 52.1, 44.9, 39.1, 39.0, 33.3, 28.3,
24.8. HR-FABMS: calcd for C52H56Cl3N3O16: 1083.2726,
found 1084.2780[M+H]+.
9. Physical data for 16 [h] 2D3=+57.5° (c=1.9, CHCl3) 1H
NMR (500 MHz, CDCl3): l 8.10–7.23 (20H, m, Ar-H),
6.42 (1H, s, NH), 6.01 (1H, s, NH), 6.00 (1H, d, H-4),
5.75 (1H, dd, J2,3=10.4 Hz, H-2), 5.64 (1H, dd, J3,4=3.1
Hz, H-3), 4.84 (1H, d, J1,2=7.9 Hz, H-1), 4.71 (2H, s,
CH2CCl3), 4.69 (1H, dd, H-6a), 4.44 (1H, dd, H-6b), 4.36
(1H, t, H-5), 3.99, 3.75 (2H, m, CH2O of sugar unit), 3.89
(2H, s, NCH2CO), 3.68–3.40 (6H, m, NHCH2CO,
NCH2C, CONHCH2), 2.79 (2H, t, COCH2C), 1.46 (9H,
s, t-Bu). 13C NMR (125 MHz, CDCl3): l 169.5, 166.1,
165.5, 165.5, 133.7, 133.4, 130.0, 129.7, 129.7, 129.3,
129.0, 128.7, 128.5, 128.3, 101.6, 94.7, 81.3, 74.1, 71.6,
71.4, 70.0, 68.6, 68.1, 52.3, 45.0, 42.5, 39.2, 33.4, 28.3.
HR-FABMS: calcd for C50H52Cl3N3O16: 1055.2413,
found 1056.2516[M+H]+.
1
Physical data 17 [h]D23=+53.0° (c=1.8, CHCl3) H NMR
10. MALDI-TOFMS of final compound 1: calcd for
(500 MHz, CDCl3): l 8.10–7.23 (20H, m, Ar-H), 6.79
(1H, s, NH), 6.00 (1H, d, H-4), 5.96 (1H, s, NH), 5.76
C83H134N14O44: 2063, found: 2086 [M+Na]+.