Synthesis of MUC1 Peptide and Glycopeptide Dendrimers

Article abstract of DOI:10.1071/CH09282

Several dendrimers possessing multiple copies of peptides and glycopeptides belonging to the MUC1 eicosapeptide tandem repeat sequence have been prepared. Fmoc-strategy solid-phase peptide synthesis was used to construct the peptides and glycopeptides, which were conjugated to suitably functionalized dendrimer cores using the copper-catalyzed azide-alkyne cycloaddition reaction to produce multivalent peptide and glycopeptide dendrimers.

Full text of DOI:10.1071/CH09282

RESEARCH FRONT
Communication
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2009, 62, 1339–1343
Synthesis of MUC1 Peptide and Glycopeptide Dendrimers
Candy K. Y. ChunA and Richard J. PayneA,B
ASchool of Chemistry, The University of Sydney, NSW 2006, Australia.
BCorresponding author. Email: payne@chem.usyd.edu.au
Several dendrimers possessing multiple copies of peptides and glycopeptides belonging to the MUC1 eicosapeptide tandem
repeat sequence have been prepared. Fmoc-strategy solid-phase peptide synthesis was used to construct the peptides and
glycopeptides, which were conjugated to suitably functionalized dendrimer cores using the copper-catalyzed azide-alkyne
cycloaddition reaction to produce multivalent peptide and glycopeptide dendrimers.
Manuscript received: 12 May 2009.
Manuscript accepted: 19 June 2009.
HO
HO
OH
O
HO
HO
OH
O
HO
O
OH
O
Introduction
The surfaces of human cells are decorated with glycoproteins
that function as important ligands and receptors for numerous
biological events, including cell adhesion, cell differentiation,
immunodifferentiation, and cell growth.[1–3] Significantly, the
progression of cancer is often associated with dysregulation of
the enzymes responsible for protein glycosylation, thus lead-
ing to altered carbohydrate patterns at the surface of a range
of tumour cell types. Specifically, in the case of several epithe-
lial tumours (including breast, prostate, ovarian, pancreatic, and
colorectal cancers), large MUC1 glycoproteins are drastically
overexpressed and aberrantly glycosylated.[4] Consequently,
there has been intense effort focussed on the development of
synthetic and semi-synthetic cancer vaccine candidates based
on MUC1 peptides and glycopeptides.[5–13]
AcNH
OH
AcNH
OR
OR
TN antigen 1
T antigen 2
OH
OH
HO
CO2H
HO
CO2H
O
OH
OH
HO
O
O
AcHN
AcHN
HO
HO
O
HO
OH
OH
OH
O
O
O
HO
O
AcNH
AcNH
OR
Sialyl TN antigen 3
OH
OR
α-2,6-Sialyl T antigen 4
The extracellular domain of MUC1 contains a variable
number tandem repeat (VNTR) domain of 20 amino acids
(GVTSAPDTRPAPGSTAPPAH), possessing five potential O-
glycosylation sites. Incomplete glycosylation and premature
sialylation of this repeat sequence leads to the presentation of
several well characterized carbohydrate tumour markers on the
cell surface. These include the TN (GalNAc, 1), T (Gal-β-1,3-
GalNAc, 2), sialyl TN (3), and sialyl T (4) antigens (Fig. 1).[14]
This notable difference in glycoprotein profile of tumour cells
constitutes an important basis for therapeutic intervention,
whereby a selective immunological attack of certain cancer cells
can be achieved.[5,13] In the case of MUC1, the truncated glycan
structures result in the exposure of additional peptide epitopes,
which become accessible to the immune system.[15,16]
Synthetic peptide and glycopeptide dendrimers provide an
avenue for the multivalent display of peptide and glycopeptide
epitopes, a desirable feature for a strong and sustained immuno-
logical response.[16,17] Additionally, the large size of dendrimers
has the potential to render them self-immunogenic, thereby pre-
cluding the need for carrier proteins traditionally employed for
vaccine constructs.[18–20] Given our interest in the development
of MUC1-based vaccine candidates, we set upon the synthesis
of multivalent dendrimers presenting several copies of immuno-
genic peptides and glycopeptides of the MUC1VNTR sequence.
Our synthetic strategy was to utilize the Cu(I)-catalyzed variant
Fig. 1. Tumour-associated glycans.
of the Huisgen 1,3-dipolar cycloaddition reaction,[21] namely
the Cu(I) catalyzed azide-alkyne cycloaddition (CuAAC),[22,23]
to generate these constructs in an efficient and concise manner.
In order to synthesize the desired MUC1-based peptide and
glycopeptide dendrimers, it was initially necessary to prepare
a suitable peptide substrate which could be coupled to a den-
drimer scaffold. We first chose to prepare the immunodominant
glycinylserinylthreoninylalanine (GSTA) tetrapeptide epitope of
MUC1, possessing an alkyne functionalized N-terminus. Treat-
ment of Fmoc-Ala-OH (5) with N,Nꢀ-diisopropylcarbodiimide
produced symmetrical anhydride 6, which was next coupled to
Wangresintogeneratethefunctionalizedsolidsupport7inquan-
titative yield (as determined by measuring the piperidine-fulvene
adduct at λ = 301 nm generated upon treatment of 7 with 10%
piperidine/DMF). Peptide assembly was then achieved by Fmoc-
strategysolid-phasepeptidesynthesis(SPPS)togiveresinbound
tetrapeptide 8 which, following Fmoc deprotection, was initially
treated with propiolic acid and 2-ethoxy-1-ethoxycarbonyl-1,2-
dihydroquinoline (EEDQ) followed by acidolytic deprotection
and cleavage from the resin to afford the desired alkynyl pep-
tide 9 in 73% yield after purification by preparative HPLC
(Scheme 1).
© CSIRO 2009
10.1071/CH09282
0004-9425/09/101339

RESEARCH FRONT
1340
C. K. Y. Chun and R. J. Payne
Wang resin,
DMAP, DMF
DIC,
DCM, 0ꢀC
OH
O
O
FmocHN
FmocHN
NHFmoc
quant.
FmocHN
O
O
O
O
7
6
5
(i) Fmoc removal:
10% piperidine in DMF
(ii) Amino acid coupling:
Fmoc-Xaa-OH, PyBOP,
Fmoc-SPPS
NMM in DMF
(i) Propiolic acid
EEDQ, DMF
(ii) 90:5:5 v/v/v
TFA/(iPr)3SiH/H2O
(iii) Capping: 10% Ac2O/
pyridine
HO
Ot-Bu
O
O
O
O
H
N
H
N
H
N
OH
H2N
O
N
N
H
O
N
H
N
H
73%
H
O
O
O
O
HO
t-BuO
9
8
Scheme 1.
NH2
N3
O
N
H2N
O
O
N3
O
O
N3
S
N
·HCl
O
NH
NH
O
O
NH
NH
O
HN
HN
HN
11
N
N
N
CuSO4, K2CO3
MeOH
N
HN
HO
25%
O
O
NH2
N3
H2N
N3
10
12
OH
NH
O
O
9, CuSO4 (0.1 eq.)
75% Sodium ascorbate (1 eq.)
t-butanol/H2O 1:1 v/v,
rt, 20 h
HN
HO
NH
O
HN
O
N
N
N
OH
O
NH
O
HO
HN
HN
O
N
O
NH
O
HO
N
N
HN
O
O
H
N
O
N
N
N
N
N
H
NH
O
N
OH
O
HN
O
OH
O
NH
NH
O
HN
O
HO
N
N
N
O
NH
OH
NH
O
HN
O
O
HO
HN
13
OH
O
Scheme 2.

RESEARCH FRONT
Synthesis of MUC1 Peptide and Glycopeptide Dendrimers
1341
OR
OR
O
O
O
O
O
Fmoc-SPPS
90:5:5 v/v/v
H
N
H
N
TFA/(iPr)3SiH/H2O
O
N3
OH
N3
O
FmocHN
N
H
N
H
N
H
N
H
O
(i) Fmoc removal:
10% piperidine in DMF
(ii) Amino acid coupling:
O
O
O
RO
RO
4
17: R ꢁ t-Bu
14: R ꢁ H (quant.)
Fmoc-Xaa-OH (incl. 15, 19, and 20),
21: R ꢁ
OAc
22: R ꢁ
OAc
OAc
PyBOP, NMM in DMF
OAc
(iii) Capping: 10% Ac2O/
pyridine
O
O
5% aq. H2NNH2,
40 min
AcO
AcO
AcNH
AcNH
OH
26%
OH
OH
OAc
OAc
N3
18: R ꢁ
O
O
O
HO
AcO
15
AcNH
AcNH
O
R
O
(i) NaN3, DMF
(ii) KOH, 1:1 v/v
THF/H2O
18% over two steps
OH
FmocHN
OMe
O
Br
19: R ꢁ CH3
20: R ꢁ H
16
Scheme 3.
HN
O
NH
O
O
O
O
N
O
NH
NH
HN
HN
5-Hexynoic acid, EEDQ
68%
N
PAMAM gen zero (10)
O
O
HN
NH
23
14 or 18, CuSO4 (0.4 eq.), sodium ascorbate (4 eq.)
t-butanol/H2O 1:1 v/v, rt, 48 h
RO
N
N
O
O
N
OR
H
O
O
N N
N
N
OH
H
N
N
H
N
H
HO
N
N
H
O
O
HN
HN
H
O
RO
O
NH
O
O
O
O
OR
O
O
O
NH
NH
HN
N
N
HN
O
RO
O
O
NH
H
N
OR
N
N
N N
OH
O
O
N
N
H
N
N
H
N
H
24: R ꢁ H (60%)
25: R ꢁ
HO
N
O
O
N
H
OH
OH
RO
H
O
O
O
OR
(20%)
HO
AcNH
Scheme 4.
Polyamidoamine (PAMAM) generation zero (10), possessing
an ethylenediamine core and four amino groups on the periph-
ery, was chosen as a suitable scaffold from which to construct
our multivalent MUC1-based dendrimers. To facilitate function-
alization of the chosen dendrimer core, it was initially necessary
to modify the surface amino groups by conversion to the
corresponding azides, thereby enabling subsequent conjugation
to alkynyl peptide 9 using CuAAC chemistry. In the event, con-
version to the corresponding azide moiety was readily achieved
by treating 10 with the imidazole-based Goddard–Borger–Stick
diazotransfer reagent 11 (Scheme 2).[24]
While it was possible to successfully prepare the desired
azido-PAMAM 12, we were disappointed to find that the iso-
lated yield (after preparative reverse-phase HPLC) was only

RESEARCH FRONT
1342
C. K. Y. Chun and R. J. Payne
Trt
N
N
O
FmocHN
O
27
(i) Fmoc removal:
10% piperidine in DMF
Fmoc-SPPS
(ii) Amino acid coupling:
Fmoc-Xaa-OH (incl. 15),
PyBOP, NMM in DMF
(iii) capping: 10% Ac2O/
pyridine
H
N
90:5:5 v/v/v TFA/(iPr)3SiH/H2O
HO
N
O
HO
HO
O
O
O
O
O
O
O
H
H
N
H
N
H
N
H
H
H
N
N
N
N
N
N
N
N
OH
N3
N
H
N
H
N
H
N
N
H
N
H
N
H
N
H
O
O
O
O
O
O
O
O
O
O
O
O
OH
COOH
HO
NH
H2N
NH
N3-GVTSAPDTRPAPGSTAPPAH-OH
26 (98%)
23, CuSO4 (0.4 eq.), sodium ascorbate (4 eq.)
t-butanol/H2O 1:1 v/v, rt, 48 h
N
N
N
N
GVTSAPDTRPAPGSTAPPAH-OH
N
N
HO-HAPPATSGPAPRTDPASTVG
HN
HN
NH
NH
O
O
O
O
O
O
O
O
NH
NH
HN
HN
N
N
N
N
N
GVTSAPDTRPAPGSTAPPAH-OH
N
N
N
HO-HAPPATSGPAPRTDPASTVG
28 (56%)
Scheme 5.
25%.This was attributed to the instability and consequent degra-
dation of the tetra-azide moiety during purification. Indeed, 12
also proved unstable to storage at low temperatures (−20◦C)
for extended periods. Despite these difficulties, 12 was success-
fully implemented as a scaffold in the preparation of peptide
dendrimer 13, bearing four copies of the immunogenic GSTA
segment of the MUC1 repeat. Treatment of PAMAM tetra-azide
12 with 6 molar equivalents of alkynylpeptide 9 (1.5 equiva-
lents per azide), 10 mol-% of CuSO4 and 1 molar equivalent of
sodium ascorbate in water and t-butanol[23] gave the desired pep-
tide dendrimer product 13 in 75% yield after HPLC purification
(Scheme 2).
While gratified to have successfully completed preparation
of the first member of our planned PAMAM dendrimer library,
the inherent instability of tetra-azide 12 prompted a revision of
our synthetic strategy. Thus, it was decided to alter the function-
ality on the PAMAM scaffold to incorporate terminal alkynes in
place of the azide moieties. It was anticipated that this revised
PAMAM core could then be reacted with MUC1-based pep-
tides and glycopeptides possessing N-terminal azide moieties
under analogous CuAAC conditions to generate multivalent
dendrimers.
To this end, azidopeptide 14 was synthesized by Fmoc-
strategy SPPS starting from preloaded Wang resin 7 (Scheme 3).
Azidoglycine 15 (synthesized in two steps from methylbromo-
acetate 16) was coupled to generate resin-bound tetrapeptidyl
azide 17. Side chain deprotection and cleavage from the resin
then furnished 14 in quantitative yield (as determined by the
Fmoc loading of the penultimate serine residue). A similar
sequence of reactions was followed for the preparation of gly-
copeptide 18, with glycosylthreonine 19 and glycosylserine 20
coupled to the growing chain in place of the standard side chain
protected Fmoc-amino acids. Incorporation of azidoglycine as
the final amino acid generated the resin bound glycopeptide 21
which, afteracidicdeprotectionandcleavagefromtheresin, gave
22, bearing two acetylated copies of the TN antigen. Treatment
of 22 with an aqueous hydrazine solution to remove the acetate
groups, followed by HPLC purification, then gave the desired
glycopeptide 18 in 26% yield based on the Fmoc loading of the
glycosylserine residue.

RESEARCH FRONT
Synthesis of MUC1 Peptide and Glycopeptide Dendrimers
1343
The next stage in the synthesis involved the preparation of
a suitably functionalized revised dendrimer core, which could
subsequently be conjugated to both 14 and 18. PAMAM core
23, bearing four terminal alkyne moieties, was generated in
68% yield by treatment of 10 with 5-hexynoic acid and EEDQ
(Scheme 4). Azidopeptide 14 was now reacted with 23 under
the CuAAC conditions described previously. In this instance,
however, these conditions returned only starting materials. Pos-
tulating that this may be a consequence of non-productive copper
chelates being formed with our new substrates, the reaction was
repeated with additional equivalents of both CuSO4 (40 mol-%)
and sodium ascorbate (4 eq.). Under these revised conditions, we
were delighted to find that dendrimer 24, bearing four copies of
the immunogenic GSTA peptide, was formed in 48 h at ambient
temperature and was isolated in 60% yield. In a similar man-
ner, glycopeptide dendrimer 25, possessing four copies of the
fully glycosylated GSTA fragment, was synthesized by reaction
between 23 and azidoglycopeptide 18. In this case, the desired
product 25 was isolated in a reduced yield of 20% yield, due in
major part to difficulties in separating the product from 18 by
reverse-phase HPLC.
Having established that PAMAM peptide and glycopeptide
dendrimers could be synthesized bearing four copies of ungly-
cosylated and glycosylated GSTA fragments, we next turned our
attention to the construction of multivalent dendrimers possess-
ing four copies of the full eicosapeptide MUC1 tandem repeat
sequence. To this end, we embarked on the synthesis of peptide
26 bearing an N-terminal azidoglycine residue. Starting from
Wang resin preloaded with Fmoc-His(Trt)-OH (27), the desired
peptide was assembled by SPPS according to the Fmoc-strategy.
Azidoglycine (15) was coupled as the final amino acid residue,
to afford azidopeptide 26 in 98% yield after side chain deprotec-
tion and cleavage from the resin followed by HPLC purification
(Scheme 5). Synthesis of dendrimer 28 bearing four copies of
the MUC1 eicosapeptide repeat was readily achieved under the
same conditions described for the shorter peptide analogues 24
and 25. We were delighted to find that the reaction proceeded
to completion after 48 h at ambient temperature and purification
by reverse HPLC gave the 8.5 kDa peptide dendrimer in 56%
yield.
Accessory Publication
For synthesis details please access the Accessory Publication
available on the Journal’s website.
Acknowledgements
We would like to thank the Australian Research Council for funding and
Yves Hsieh for assistance with MALDI-TOF mass spectrometry.
References
[1] A. Varki, Glycobiology 1993, 3, 97. doi:10.1093/GLYCOB/3.2.97
[2] R. G. Spiro, Glycobiology 2002, 12, 43R. doi:10.1093/GLYCOB/
12.4.43R
[3] R. A. Dwek, Chem. Rev. 1996, 96, 683. doi:10.1021/CR940283B
[4] F.-G. Hanisch, S. Muller, Glycobiology 2000, 10, 439. doi:10.1093/
GLYCOB/10.5.439
[5] H. Kunz, J. Pept. Sci. 2003, 9, 563. doi:10.1002/PSC.477
[6] H. Herzner, T. Reipen, M. Schultz, H. Kunz, Chem. Rev. 2000, 100,
4495. doi:10.1021/CR990308C
[7] T. Buskas, S. Ingale, G. J. Boons, Glycobiology 2006, 16, 113R.
doi:10.1093/GLYCOB/CWJ125
[8] S. J. Danishefsky, J. R. Allen, Angew. Chem. Int. Ed. 2000,
39, 836. doi:10.1002/(SICI)1521-3773(20000303)39:5<836::AID-
ANIE836>3.0.CO;2-I
[9] S. Dziadek, H. Kunz, Chem. Rec. 2004, 3, 308. doi:10.1002/
TCR.10074
[10] O. Ouerfelli, J. D. Warren, R. M. Wilson, S. J. Danishefsky, Expert
Rev. Vaccines 2005, 4, 677. doi:10.1586/14760584.4.5.677
[11] C. Brocke, H. Kunz, Bioorg. Med. Chem. 2002, 10, 3085.
doi:10.1016/S0968-0896(02)00135-9
[12] M. Meldal, P. M. St Hilaire, Curr. Opin. Chem. Biol. 1997, 1, 552.
doi:10.1016/S1367-5931(97)80052-X
[13] S. Dziadek, C. G. Espinola, H. Kunz, Aust. J. Chem. 2003, 56, 519.
doi:10.1071/CH02241
[14] G. F. Springer, Science 1984, 224, 1198. doi:10.1126/SCIENCE.
6729450
[15] S. Dziadek, C. Griesinger, H. Kunz, U. M. Reinscheid, Chem. Eur. J.
2006, 12, 4981. doi:10.1002/CHEM.200600144
[16] A. L. Sorensen, C. A. Reis, M. A. Tarp, U. Mandel, K. Ramachandran,
V. Sankaranarayanan, T. Schwientek, R. Graham, J. Taylor-
Papadimitriou, M. A. Hollingsworth, J. Burchell, H. Clausen,
Glycobiology 2006, 16, 96. doi:10.1093/GLYCOB/CWJ044
[17] A. Liakatos, H. Kunz, Curr. Opin. Mol. Ther. 2007, 9, 35.
[18] P. Niederhafner, M. Reinis, J. Sebestik, J. Jezek, J. Pept. Sci. 2008, 14,
556. doi:10.1002/PSC.1011
[19] C. Ozawa, H. Katayama, H. Hojo, Y. Nakahara, Org. Lett. 2008, 10,
3531. doi:10.1021/OL801340M
[20] C. Ozawa, H. Hojo,Y. Nakahara, H. Katayama, K. Nabeshima,T.Aka-
hane, Tetrahedron 2007, 63, 9685. doi:10.1016/J.TET.2007.07.023
[21] R. Huisgen, G. Szeimies, L. Mobius, Chem. Ber. 1967, 100, 2494.
doi:10.1002/CBER.19671000806
In summary, we have utilized solid- and solution-phase chem-
istry in combination with the CuAAC reaction to construct
a small library of peptide and glycopeptide dendrimers in a
fast and efficient manner. We found that the optimal strategy
involved coupling peptides and glycopeptides bearing an N-
terminal azide with a dendrimer scaffold presenting multiple
alkynes. The constructs synthesized in this study are currently
undergoing immunological evaluation for their potential to gen-
erate antibodies specific for peptide and glycopeptide fragments
of the MUC1 tandem repeat. Future directions within this lab-
oratory will also aim to utilize the synthetic strategy outlined
here for the synthesis of more complex MUC1-based dendrimers
possessing other cancer-associated glycans.
[22] C. W. Tornøe, C. Christensen, M. Meldal, J. Org. Chem. 2002, 67,
3057. doi:10.1021/JO011148J
[23] V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless,
Angew. Chem. Int. Ed. 2002, 41, 2596. doi:10.1002/1521-3773
(20020715)41:14<2596::AID-ANIE2596>3.0.CO;2-4
[24] E. D. Goddard-Borger, R. V. Stick, Org. Lett. 2007, 9, 3797.
doi:10.1021/OL701581G