Self-adjuvanting multicomponent cancer vaccine candidates combining per-glycosylated muc1 glycopeptides and the toll-like receptor 2 agonist pam3CysSer

Article abstract of DOI:10.1002/anie.201006115

More sugar with your "T": The title candidates comprising per-glycosylated MUC1 peptides, a T-cell helper peptide, and an immunoadjuvant were assembled in high yields using iterative pentafluorophenyl ester-mediated fragment condensations. Following immun

Full text of DOI:10.1002/anie.201006115

DOI: 10.1002/anie.201006115
Glycopeptide Vaccines
Self-Adjuvanting Multicomponent Cancer Vaccine Candidates
Combining Per-Glycosylated MUC1 Glycopeptides and the Toll-like
Receptor 2 Agonist Pam3CysSer**
Brendan L. Wilkinson, Stephanie Day, Lara R. Malins, Vasso Apostolopoulos, and
Richard J. Payne*
The aberrant glycosylation of cell surface mucin glycoproteins
such as MUC1 and the overexpression of highly truncated,
tumor-associated carbohydrate antigens (TACAs) are
strongly correlated with tumor metastasis and poor progno-
sis.[1] Such TACAs, including the well-studied TN and T
antigens and their sialylated derivatives, are important targets
for the development of antigen-specific carbohydrate vac-
cines.[2] However, these structures are self antigens that
generally elicit T-cell independent immune responses which
are tolerated by the immune system and lead to poor
immunogenicity. One approach aimed at enhancing the
immune recognition of TACAs utilizes conjugation of
tumor-associated MUC1 glycopeptides or TACAs alone to
foreign carrier proteins such as keyhole limpet hemocyanin
(KLH) or bovine serum albumin (BSA).[3] Kunz and co-
workers have described several impressive examples of
MUC1 glycopeptide carrier protein vaccine candidates
which have elicited significant antibody responses in murine
models.[3d–h] While conjugate vaccines have shown promise in
tumor vaccinology, the inherent immunogenicity of the
carrier protein has been associated with suppression of the
immune response towards the desired carbohydrate/glyco-
peptide epitope.[4]
cell epitope.[6] These multicomponent vaccines elicited high
titers of IgG antibodies which were able to recognize epitopes
on MCF7 cancer cells.[6a] It was reasoned that multicompo-
nent vaccines incorporating the TLR2 ligand enhance local
production of inflammatory mediators, leading to the upre-
gulation of co-stimulatory proteins and the activation and
regulation of the adaptive component of the immune
system.[5,6,9]
Glycosylation of the MUC1 variable number tandem
repeat (VNTR) sequence (GVTSAPDTRPAPGSTAPPAH)
is an important feature for modulating the immunogenicity of
MUC1 peptide epitopes with the aim of breaking self-
tolerance and inducing humoral immunity.[3,10,11] Previous
studies from the laboratories of Clausen and Livingston have
shown that per-glycosylated MUC1 glycopeptides are able to
elicit high IgG antibody titers when used in conjunction with a
carrier protein and an external adjuvant.[10a,11] In addition,
Kunz and co-workers have demonstrated that the peptide
fragment (YSYFPSV) of the tetanus toxoid protein, when
conjugated to MUC1 glycopeptide antigens, is capable of
inducing proliferation of CD3 + and CD8 + cells.[13]
With these observations in mind, we were interested in
exploring the use of synthetic multicomponent vaccines
incorporating a per-glycosylated MUC1 peptide as the B-
cell epitope covalently linked to an immunostimulating
lipopeptide (Pam3CysSer), with or without the tetanus toxin
peptide as the helper T-cell epitope. We recently reported an
efficient fragment condensation strategy for the construction
of MUC1–lipopeptide conjugates employing pentafluoro-
phenyl esters as the N-acyl donor.[12] Here, we exploit this
strategy for the convergent assembly and immunological
investigation of dicomponent vaccines 1a–c, comprised of
multiple copies of TN or T antigens conjugated to the TLR2
ligand Pam3CysSer, and tricomponent vaccines 2a–c incor-
porating the tetanus toxin-derived helper T-cell epitope[13]
Self-adjuvanting, multicomponent vaccines however,
avoid anti-carrier immune responses while incorporating the
necessary structural features for evoking an essential class
switch from low-affinity and short lived immunoglobulin M
(IgM) antibodies, to high-affinity immunoglobulin G (IgG)
antibodies.[5–7] Recently, Boons and co-workers reported the
synthesis and immunological activity of multicomponent
vaccines incorporating the immunostimulating Toll-like
[8]
receptor 2 (TLR2) ligand, Pam3CysSerK4 and a 12-amino
acid peptide bearing a single copy of the TN antigen as the B-
[*] Dr. B. L. Wilkinson, L. R. Malins, Dr. R. J. Payne
School of Chemistry, The University of Sydney
New South Wales 2006 (Australia)
between
a per-glycosylated MUC1 B-cell epitope and
Pam3CysSer (Scheme 1). We chose to install flexible, immu-
nogenically silent ethylene glycol spacer units between each
component to minimize conformational distortion and to
allow for optimal display of each recognition element to the
immune system.[7a,13]
The synthesis of the target vaccine candidates began with
the preparation of the three requisite peptide fragments,
namely the B-cell (glyco)peptide epitopes 3a–c, tetanus toxin
peptide 4, and lipopeptide 5. Synthesis of the completely
deprotected MUC1 peptide 3a and glycopeptides 3b and 3c
bearing a per-glycosylated array of the TN and T antigens,
Fax: (+61)2-9351-3329
E-mail: richard.payne@sydney.edu.au
Homepage: http://www.chem.usyd.edu.au/~payne/
Dr. S. Day, Prof. Dr. V. Apostolopoulos
Immunology and Vaccine Laboratory, Burnet Institute
Melbourne 3004 (Australia)
[**] We would like to thank the Australian Research Council for funding
(ARC Discovery Project: 0986632) and Dr. Kelvin Picker and Dr.
Keith Fisher for technical support.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201006115.
Angew. Chem. Int. Ed. 2011, 50, 1635 –1639
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1635

Communications
Following elongation, cleavage of the side-chain protected
peptide from the resin using 30% hexafluoroisopropyl
alcohol (HFIP) in DCM and subsequent precipitation
afforded pure peptide 4 in 48% yield. Likewise, the lipopep-
tide fragment was assembled using Fmoc-strategy SPPS on 2-
chlorotrityl chloride resin pre-loaded with Fmoc-protected
triethylene glycolic acid. Mild acidic cleavage then furnished
the protected lipopeptide 5 in 95% yield following purifica-
tion by flash chromatography.
With the desired peptide fragments in hand, we next
assembled the proposed vaccine candidates by pentafluoro-
phenyl ester-mediated fragment condensations.[12,15] We have
previously employed such an approach in the preparation of
dicomponent vaccine candidates (1a–c) lacking the helper T-
cell epitope. We now sought to examine the utility of this
approach for the construction of tricomponent vaccine
candidates 2a–c employing two consecutive pentafluoro-
phenyl ester-mediated fragment condensations (Scheme 3).
To this end, treatment of the tetanus toxin fragment 4 with
equimolar amounts of pentafluorophenol and N,N’-diiso-
propyl carbodiimide (DIC) afforded the crude pentafluoro-
phenyl ester after 1 h (Scheme 3). Treatment of the crude
ester with a solution of the respective MUC1 (glyco)peptides
3a–c (0.5 equiv), HOBt (0.5 equiv), and DIPEA (1.0 equiv)
in DMF furnished the partially protected conjugates. Sub-
sequent deprotection and purification by semi-preparative
reverse-phase HPLC furnished the MUC1–tetanus toxin
conjugates 6a–c in 40-91% yields. The homogeneous, com-
pletely deprotected fragments 6a–c were subsequently con-
jugated to the lipopeptide 5 which was pre-activated as the
pentafluorophenyl ester to afford tricomponent vaccine
candidates 2a–c in excellent yields (84–92%) after HPLC
purification.
Scheme 1. Di- and tricomponent MUC1-based vaccine candidates
1a–c and 2a–c.
respectively, was achieved by Fmoc-strategy solid-phase
peptide synthesis (SPPS) employing 2-chlorotrityl chloride
resin preloaded with Fmoc-His(Trt)-OH (see Supporting
Information and Scheme 2).[14] Cleavage from the resin and
removal of the side-chain protecting groups was achieved
using an acidic cocktail of TFA/TIS/thioanisole/water
(85:5:5:5, v/v/v/v). The O-acetate protecting groups on the
glycans of glycopeptides 3b and 3c were subsequently
removed by hydrazinolysis. Purification by preparative
reverse-phase HPLC afforded the target (glyco)peptides
3a–c in 45%, 22%, and 14% yield, respectively, based on
the original resin loading. Protected tetanus toxin T-cell
epitope fragment 4 was also prepared by Fmoc-strategy SPPS.
Scheme 2. Solid-phase peptide synthesis (SPPS) of peptide, glycopeptide, and lipopeptide fragments. Fmoc=9-fluorenylmethoxycarbonyl; DMF=dimethyl-
formamide; PyBOP=benzotriazolyl-1-oxy-tripyrrolidinophosphonium hexafluorophosphate; NMM=N-methylmorpholine; HATU=O-(7-azabenzotriazol-1-
yl)-tetramethyluronium hexafluorophosphate; DIPEA=N,N-diisopropylethylamine; TFA=trifluoroacetic acid; TIS=triisopropylsilane; TA=thioanisole;
HFIP=hexafluoroisopropyl alcohol; DCM=dichloromethane.
1636
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 1635 –1639

Scheme 3. Synthesis of MUC1 vaccine candidates 2a–c through a fragment condensation approach. HOBt=1-hydroxybenzotriazole.
Having successfully prepared di- and tricomponent vac-
cines 1a–c and 2a–c, respectively, we next evaluated their
capacity to induce humoral immunity in murine models.[16] To
this end, four C57BL/6 mice were immunized with 20 mg of
each vaccine candidate in the absence of an external adjuvant.
Booster immunizations were administered after days 10 and
17 before the mice were euthanized on day 31. Serum was
taken after the second and third boosters for determination of
IgG antibody titers, measured by their ability to bind
immobilized MUC1 (glyco)peptides 3a–c in an ELISA
assay (see Supporting Information). Low-level antibody
titers were obtained for the dicomponent vaccines 1a–c
following the second immunization. In the case of the
unglycosylated construct 1a, IgG titers were considerably
increased after the third immunization (Figure 1). The
analogous glycosylated vaccine constructs, 1b and 1c, also
elicited significant IgG levels after the third immunization
(reciprocal endpoint IgG titers were 1250 for 1b and 937 for
1c). However, IgG titers for 1b and 1c were significantly
lower than the unglycosylated counterpart 1a, highlighting
the relative importance of exposed peptide epitopes com-
bined with the TLR2 ligand for generating an immune
response. It is proposed that the clustered arrangement of the
TN and T antigens on constructs 1b and 1c, respectively, may
impede epitope recognition, leading to reduced immunoge-
Figure 1. ELISA anti-MUC1 IgG reciprocal antibody titers after three
immunizations for constructs 1a–c and 2a–c. Histograms represent
the mean endpoint titers of the sera from four C57BL/6 mice.
Angew. Chem. Int. Ed. 2011, 50, 1635 –1639
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1637

Communications
nicity. Nonetheless, these constructs provided comparable
adjuvanting, multicomponent MUC1-based vaccine candi-
dates. Immunological evaluation in murine models identified
a number of key structural features required to invoke strong
humoral immune responses which should greatly assist in the
rational design of synthetic MUC1-based cancer vaccines in
the future. Specifically, vaccines incorporating per-glycosy-
lated MUC1 VNTR glycopeptides in combination with a T-
cell helper peptide and the TLR2 agonist Pam3CysSer were
able to elicit high levels of IgG antibodies in murine models in
the absence of an external adjuvant, liposomal preparation or
carrier protein. The anti-sera from tricomponent vaccines 2a–
c were shown to possess some selectivity for the (glyco)pep-
tides to which they were raised and were capable of binding to
epitopes presented on the surface of the MCF7 breast cancer
cell line. Future work in our laboratories is focused on the
evaluation of these vaccine constructs in MUC1-transgenic
mice. In addition, the preparation of MUC1-based tricompo-
nent cancer vaccine candidates bearing other TACAs, T-cell
helper peptides, and synthetic immunoadjuvants is currently
underway.
titers to previously reported MUC1-based glycopeptide
vaccine candidates using carrier proteins or immunoadju-
vants, without the use of a liposomal formulation or an
external adjuvant such as complete Freundꢀs adjuvant
(CFA).[3f,g,7a] This indicates that the Pam3CysSer adjuvant
within these multicomponent vaccines is capable of inducing
strong humoral immunity, resulting in significant IgG anti-
body titers. We were surprised to find that no IgM antibodies
could be detected in the sera for vaccines 1a–c. This suggests
the establishment of immunological memory, consistent with
a recent report from Kunz and co-workers with tetanus
toxoid–MUC1 vaccines.[3h]
The tricomponent vaccine constructs 2a and 2b, incorpo-
rating the tetanus toxin helper T-cell epitope, proved to be
significantly more immunogenic than the dicomponent vac-
cines following the second immunization and maintained
higher IgG levels after the third immunization (reciprocal
endpoint IgG titers were 52000 for 2a and 2600 for 2b).
Tricomponent vaccine 2c, bearing multiple copies of the T
antigen, elicited impressive levels of IgG antibodies (recip-
rocal endpoint IgG titer 54600), higher than 2a containing a
completely exposed peptide epitope. Compared with the
endpoint titer for the corresponding dicomponent vaccine 1c,
this result clearly demonstrates that, in the presence of a
TLR2 ligand and a helper T-cell epitope, tricomponent
MUC1-based glycopeptide vaccines bearing clustered glyco-
sylation with the T-antigen can elicit high IgG antibody titers.
The serum antibodies induced by vaccines 2a–c also
exhibited selectivity for the (glyco)peptides to which they
were raised (see Supporting Information). Unglycosylated
MUC1 vaccine 2a did not cross-react with the per-glycosy-
lated glycopeptide epitopes 3b and 3c, suggesting a peptide-
specific antibody response. In addition, antibodies generated
from per-glycosylated tricomponent TN antigen-containing
vaccine 2b did not bind to glycopeptide 3c, bearing the T
antigen. Antibodies elicited from vaccine 2c, presenting
multiple copies of the T antigen, exhibited weak binding to
the unglycosylated peptide 3a (reciprocal endpoint IgG titer
5400 as compared to 52000 for 3c). Interestingly, anti-sera did
show appreciable binding to glycopeptide 3b (reciprocal
endpoint IgG titer 18400), presenting multiple copies of the
TN antigen, suggesting that IgG antibodies can recognize the
GalNAc moiety of these two TACAs, an observation also
made by Franco and co-workers.[17] Finally, to ensure that the
antibodies generated from tricomponent vaccines 2a–c were
capable of recognizing MUC1 epitopes expressed on tumor
cells, binding of the anti-sera to the MUC1 expressing human
breast cancer cell line MCF7 was investigated. Gratifyingly,
sera from all three constructs exhibited significant binding to
MCF7 (22–48% of cells bound) as determined by fluores-
cence-activated cell sorting (FACS) analysis (see Supporting
Information). In contrast, anti-sera from 2a–c did not bind to
native MUC1 isolated from human breast milk,[18] suggesting
that antibodies raised from these vaccines are tumor-selective
and do not recognize the more heavily glycosylated MUC1
epitopes displayed on normal cells.
Received: September 30, 2010
Revised: October 27, 2010
Published online: January 7, 2011
Keywords: antibodies · glycopeptides · mucin 1 ·
.
solid-phase peptide synthesis · vaccines
[1] a) D. S. A. Sanders, M. A. Kerr, J. Clin. Pathol. Mol. Pathol.
1999, 52, 174; b) J. Taylor-Papadimitriou, J. M. Burchell, D. W.
Miles, M. Dalziel, Biochim. Biophys. Acta Mol. Basis Dis. 1999,
1455, 301.
[2] a) T. Buskas, P. Thompson, G. J. Boons, Chem. Commun. 2009,
5335; b) R. D. Astronomo, D. R. Burton, Nat. Rev. Drug
Discovery 2010, 9, 308.
[3] a) G. D. MacLean, M. A. Reddish, R. R. Koganty, B. M. Longe-
necker, J. Immunother. 1996, 19, 59; b) S. J. Danishefsky, J. A.
Allen, Angew. Chem. 2000, 112, 882; Angew. Chem. Int. Ed.
2000, 39, 836; c) B. G. Davis, Chem. Rev. 2002, 102, 579; d) S.
Dziadek, D. Kowalczyk, H. Kunz, Angew. Chem. 2005, 117, 7798;
Angew. Chem. Int. Ed. 2005, 44, 7624; e) S. Wittrock, T. Becker,
H. Kunz, Angew. Chem. 2007, 119, 5319; Angew. Chem. Int. Ed.
2007, 46, 5226; f) U. Westerlind, A. Hobel, N. Gaidzik, E.
Schmitt, H. Kunz, Angew. Chem. 2008, 120, 7662; Angew. Chem.
Int. Ed. 2008, 47, 7551; g) A. Kaiser, N. Gaidzik, U. Westerlind,
D. Kowalczyk, A. Hobel, E. Schmitt, H. Kunz, Angew. Chem.
2009, 121, 7688; Angew. Chem. Int. Ed. 2009, 48, 7551; h) A.
Hoffmann-Rꢁder, A. Kaiser, S. Wagner, N. Gaidzik, D. Kowalc-
zyk, U. Westerlind, B. Gerlitzki, E. Schmitt, H. Kunz, Angew.
Chem. 2010, 122, 8676; Angew. Chem. Int. Ed. 2010, 49, 8498.
[4] a) T. Buskas, Y. H. Li, G. J. Boons, Chem. Eur. J. 2004, 10, 3517;
b) J. Ni, H. Song, Y. Wang, N. M. Stamatos, L. X. Wang,
Bioconjugate Chem. 2006, 17, 493.
[5] a) K. H. Wiesmꢂller, G. Jung, G. Hess, Vaccine 1989, 7, 29;
b) W. G. Bessler, L. Heinevetter, K. H. Wiesmꢂller, G. Jung, W.
Baier, M. Huber, A. R. Lorenz, U. V. Esche, K. Mittenbuhler, P.
Hoffmann, Int. J. Immunopharmacol. 1997, 19, 547; c) M. Obert,
H. Pleuger, H. G. Hanagarth, J. Schulte-Mꢁnting, K. H.
Wiesmꢂller, H. D. G. Braun, G. Brandner, R. D. Hess, Vaccine
1998, 16, 161.
[6] a) S. Ingale, M. A. Wolfert, T. Buskas, G. J. Boons, ChemBio-
Chem 2009, 10, 455; b) S. Ingale, M. A. Wolfert, J. Gaekwad, T.
Buskas, G. J. Boons, Nat. Chem. Biol. 2007, 3, 663; c) T. Buskas,
In summary, we have utilized a fragment condensation
strategy for the high yielding assembly of a number of self-
1638
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 1635 –1639

S. Ingale, G. J. Boons, Angew. Chem. 2005, 117, 6139; Angew.
Chem. Int. Ed. 2005, 44, 5985.
[11] E. Kagan, G. Ragupathi, S. S. Yi, C. A. Reis, J. Gildersleeve, D.
Kahne, H. Clausen, S. J. Danishefsky, P. O. Livingston, Cancer
Immunol. Immunother. 2005, 54, 424.
[12] B. L. Wilkinson, L. R. Malins, C. K. Y. Chun, R. J. Payne, Chem.
Commun. 2010, 46 , 6249.
[13] S. Keil, C. Claus, W. Dippold, H. Kunz, Angew. Chem. 2001, 113,
379; Angew. Chem. Int. Ed. 2001, 40, 366.
[14] K. Barlos, D. Gatos, S. Kapolos, C. Poulos, W. Schafer, W. Q.
Yao, Int. J. Pept. Protein Res. 1991, 38, 555.
[7] a) A. Kaiser, N. Gaidzik, T. Becker, C. Menge, K. Groh, H. Cai,
Y.-M. Li, B. Gerlitzki, E. Schmitt, H. Kunz, Angew. Chem. 2010,
122, 3772; Angew. Chem. Int. Ed. 2010, 49, 3688; b) D. Krikorian,
E. Panou-Pomonis, C. Voitharou, C. Sakarellos, M. Sakarellos-
Daitsiotis, Bioconjugate Chem. 2005, 16, 812; c) O. Renaudet, L.
BenMohamed, G. Dasgupta, I. Bettahi, P. Dumy, ChemMed-
Chem 2008, 3, 737.
[8] G. Jung, C. Carrera, H. Brꢂckner, W. G. Bessler, Liebigs Ann.
Chem. 1983, 1608.
[15] E. Atherton, L. R. Cameron, R. C. Sheppard, Tetrahedron 1988,
44, 843.
[9] R. Spohn, U. Buwitt-Beckmann, R. Brock, G. Jung, A. J. Ulmer,
K. H. Wiesmꢂller, Vaccine 2004, 22, 2494.
[16] Animal experiments were approved by AMREP Animal Ethics
Committee, Melbourne, VIC, Australia. AEC Numbers, E/0950/
2010/F, E/0893/2009/F.
[17] Y. Xu, S. J. Gendler, A. Franco, J. Exp. Med. 2004, 199, 707.
[18] G. A. Pietersz, W. Li, C. Osinski, V. Apostolopoulos, I. F. C.
McKenzie, Vaccine 2000, 18, 2059.
[10] a) A. L. Sꢁrensen, 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; b) G. A.
Cremer, N. Bureaud, V. Piller, H. Kunz, H. Piller, A. F.
Delmas, ChemMedChem 2006, 1, 965.
Angew. Chem. Int. Ed. 2011, 50, 1635 –1639
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1639