Synthetic anticancer vaccine candidates: Rational design of antigenic peptide mimetics that activate tumor-specific T-cells

Article abstract of DOI:10.1021/jm0613368

A rational design approach was followed to develop peptidomimetic analogues of a cytotoxic T-cell epitope capable of stimulating T-cell responses as strong as or stronger (heteroclytic) than those of parental antigenic peptides. The work described herein focused on structural alterations of the central amino acids of the melanoma tumor-associated antigenic peptide Melan-A/MART-1 _26-35 using nonpeptidic units. A screening was first realized in silico to select altered peptides potentially capable of fitting at the interface between the major histocompatibilty complex (MHC) class-I HLA-A2 molecule and T-cell receptors (TCRs). Two compounds appeared to be high-affinity ligands to the HLA-A2 molecule and stimulated several Melan-A/MART-1 specific T-cell clones. Most remarkably, one of them even managed to amplify the response of one clone. Together, these results indicate that central TCR-contact residues of antigenic peptides can be replaced by nonpeptidic motifs without loss of binding affinity to MHC class-I molecules and T-cell triggering capacity.

Full text of DOI:10.1021/jm0613368

1598
J. Med. Chem. 2007, 50, 1598-1609
Synthetic Anticancer Vaccine Candidates: Rational Design of Antigenic Peptide Mimetics That
Activate Tumor-Specific T-Cells
Ce´line Douat-Casassus,†,‡ Nathalie Marchand-Geneste,‡ Elisabeth Diez,§ Nadine Gervois,§,∇ Francine Jotereau,*,§,∇ and
Ste´phane Quideau*,†,‡
Institut Europe´en de Chimie et Biologie, 2 rue Robert Escarpit, 33607 Pessac Cedex, France, Institut des Sciences Mole´culaires
(UMR 5255 CNRS), UniVersite´ Bordeaux 1, 351 cours de la Libe´ration, 33405 Talence Cedex, France, INSERM, U601, 44093 Nantes, France,
and UniVersite´ de Nantes, 44322 Nantes, France
ReceiVed NoVember 17, 2006
A rational design approach was followed to develop peptidomimetic analogues of a cytotoxic T-cell epitope
capable of stimulating T-cell responses as strong as or stronger (heteroclytic) than those of parental antigenic
peptides. The work described herein focused on structural alterations of the central amino acids of the
melanoma tumor-associated antigenic peptide Melan-A/MART-126-35 using nonpeptidic units. A screening
was first realized in silico to select altered peptides potentially capable of fitting at the interface between
the major histocompatibilty complex (MHC) class-I HLA-A2 molecule and T-cell receptors (TCRs). Two
compounds appeared to be high-affinity ligands to the HLA-A2 molecule and stimulated several Melan-
A/MART-1 specific T-cell clones. Most remarkably, one of them even managed to amplify the response of
one clone. Together, these results indicate that central TCR-contact residues of antigenic peptides can be
replaced by nonpeptidic motifs without loss of binding affinity to MHC class-I molecules and T-cell triggering
capacity.
Introduction
Considerable efforts have therefore been done to derive
epitope analogues of enhanced antigenicity or half-life in vivo.
Because most tumor antigens are self-proteins, their limited
antigenicity often results from central (intrathymic) tolerance
mechanisms that eliminate the high-avidity self-specific TCR
repertoire during T-cell development.7 Therefore, the function
of spared tumor associated antigen (TAA)-specific T-cells may
remain poorly effective due to their limited avidity. In some
cases, low avidity of TAA-specific T-cells result from the
cognate AP exhibiting low affinity/stability for the self-MHC
presenting molecule. The best known example of this is the
Melan-A/MART-126-35 AP expressed on the MHC class-I HLA-
A2 molecule by most melanoma tumors.8 In this case, amino
acid substitutions enhancing HLA binding may provide peptide
analogues of increased antigenicity capable of, at least some of
them, efficiently triggering a T-cell repertoire that could cross-
react with the natural peptide,9 which is of course mandatory
for vaccine use. High susceptibility to proteases, the other
probable major cause of the limited in vivo antigenicity of APs
in general and tumor APs in particular, can conceivably be
counteracted by substituting standard R-amino acid residues by
nonpeptidic groups. For example, previous attempts that were
aimed at replacing N- and/or C-terminal residues to reduce AP
susceptibility to amino- and carboxypeptidases succeeded in
producing analogues with high-affinity binding to the MHC
molecule and efficient TCR triggering.10,11 Other studies ad-
dressed, also with some success, the compatibility of performing
more central modifications of MHC-binding AP residues to
increase AP binding affinity to the MHC pocket and, hence,
T-cell reactivity.12
Cytotoxic T-lymphocytes (CTL) play a major role in the
immune defense against infections and tumors.1,2 They recognize
and kill, in the context of appropriate immune responses, target
cells expressing ligands of their receptors (TCRs). These ligands
are made of short (8-10 amino acids long) peptides that are
derived from antigenic proteins through limited intracellular
proteolysis. They are bound to MHC class-I molecules and are
thus exposed at the target cell surface as peptide-MHC (pMHC)
complexes.3 The functional consequences of these TCR/pMHC
interactions for T-cells are highly variable, ranging from full
activation to inactivation, depending in part on the binding
stability of the peptide to the MHC and, in a poorly understood
manner, on the molecular recognition characteristics and kinetic
aspects of the TCR interacting with the pMHC complex. TCR/
peptide contacts are made primarily through the complementary-
determining region 3 (CDR3) loops of the TCR, which exhibit
the greatest degree of genetic variability.4 Therefore, different
TCRs interact in a different way with a given pMHC. The
identification of antigenic peptides (APs) raised the hope that
these TCR ligands could be used to stimulate specifically T-cell
growth and function in vivo and hence to enhance protective
immune responses.5 This question was, in particular, addressed
in the context of therapeutic cancer vaccine development. Yet,
infusion of tumor APs to cancer patients, alone or in association
with adjuvants, was so far poorly effective in inducing tumor
specific T-cell expansion and tumor regression.6 Tumor AP
characteristics that could contribute to this lack of efficacy are:
a limited antigenicity, revealed by the poor triggering of specific
T-cell growth in vitro, and/or a short bioavailability due to high
susceptibility to serum or tissue proteases.
Another yet unexplored possibility is to substitute the central
AP residues specifically interacting with the TCR by nonpeptidic
motifs. Here we examined the feasibility of this approach, using
as a model the aforementioned epitope of the melanoma-
associated Melan-A/MART-1 antigen. The protein Melan-A/
MART-1 (referred to as MART-1 hereafter) is a melanocyte
lineage specific protein expressed in about 90% of primary and
* Address correspondence to either author. Phone: 33-540-00-2215. Fax:
33-540-00-2215. E-mail: s.quideau@iecb.u-bordeaux.fr (S.Q.); Phone: 33-
240-08-4720. Fax: 33-240-08-4082. E-mail: jotereau@nantes.inserm.fr (F.J.).
† Institut Europe´en de Chimie et Biologie.
‡ Institut des Sciences Mole´culaires (UMR 5255 CNRS).
§ INSERM.
∇ Universite´ de Nantes.
10.1021/jm0613368 CCC: $37.00 © 2007 American Chemical Society
Published on Web 03/01/2007

Synthetic Anticancer Vaccine Candidates
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1599
potentially heteroclytic tumor-specific antigenic peptidomimetics
based on the substitution of AP central TCR-contact amino acids
by nonpeptidic moieties susceptible to interact efficiently with
most TCRs. The crystallographic data available on the
ELAGIGILTV/HLA-A2 binary complex18 and others reported
on TCR/AP/HLA-A2 ternary complexes,20,21 as well as recent
results reported by us and others,22 all underlie the critical role
played by hydrophobic structural elements on AP central regions
in eliciting CD8+ T-cell responses. In this article, we thus
describe the rational design and synthesis approaches we
followed with this premise in mind to generate novel and
functional ELAGIGILTV-derived peptidomimetics.
Results and Discussion
Rational Design Strategy. The major difficulty encountered
in the design of biologically active AP analogues is that their
activity depends upon the accomplishment of two critical
events: (1) binding to the MHC class-I molecule and (2)
recognition by a TCR for signal triggering. To direct the design
of our peptidomimetics, we relied on a shape-based docking
engine, LigandFit from the Accelrys Cerius2 software package
(see Experimental Section).23 Since no ternary complex structure
involving ELAGIGILTV (referred to as ELA hereafter) has been
reported as of today, we decided to base our docking studies
on the 2.6 Å structure of the A6-TCR/Tax/HLA-A2 complex
(PDB 1AO7, vide infra).21 Notwithstanding the fact that the
A6-TCR is not specific of ELA recognition, we surmised from
literature data that conformations of any AP presented in the
HLA-A2 binding groove would be quite similar.24 The selected
crystal structure shows that the HTLV-derived Tax AP
(LLFGYPVYV) is oriented diagonally over the HLA-A2
molecule with the variable loops CDR1 and CDR3 of both VR
and Vâ TCR subunits contacting the peptide. The central Y
residue protrudes out from the center of the HLA-A2 binding
site and is completely surrounded by a deep and hydrophobic
pocket delimited by the two TCR CDR3 loops (Figure 1). In
order to gain some insights into the binding features of ELA
into the HLA-A2 binding groove, while justifying the choice
of the Tax-including ternary complex 1AO7 as the reference
object for this study, we first docked ELA in place of the Tax
peptide (Figure 1). The ELA conformation thus predicted is very
similar to that of the Tax peptide with an almost perfect
superimposition of their anchor residue side chains. Moreover,
this docked conformation is quasi identical to that observed in
the crystal structure of the ELA/HLA-A2 binary complex (PDB
1JF1).18
A virtual screening of several series of conceivable ELA
analogues was then achieved using the LigandFit program before
considering their synthesis. Every first proposed docked con-
formation (having the best docked energy) was compared to
that of ELA by superimposition in Insight II again using 1AO7
structure as the reference object. Since the activity of a given
AP essentially depends on its docked structure in the ternary
complex, three main checkpoints were chosen as a mean of
selecting the most appropriate central region modifications: (1)
orientation of anchor residues (L2 and V10); (2) conservation
of peptide backbone folding; (3) orientation of putative TCR-
contact side chains. One of the benefits of this analysis in silico
was to gain diagnostic structural information to guide our
peptidic backbone modifications toward a better targeting of
the TCR binding groove. Different generations of ELA ana-
logues were iteratively designed from the simplest ones, bearing
only different R-amino acid residues in their central region, to
peptidomimetics in which standard peptidic features were
gradually eliminated.
Figure 1. InsightII superimposition of the LigandFit-docked ELA
peptide (green) in the Tax peptide (orange) binding site of the A6-
TCR/HLA-A2 complex (PDB 1AO7). Positioning of main anchor
residues [i.e., L27 (P2) and V35 (P10)] is almost identical to that of
the Tax peptide with the ELA central region also bulging out from the
center of the HLA-A2 binding domain toward the CDR3 TCR variable
loops.
metastatic melanoma tumors.13 MART-1-specific CTLs have
been isolated from both peripheral blood mononuclear cells
(PBMC) and tumor infiltrating lymphocytes (TILs) of HLA-
A2 melanoma patients.14,15 Later, the screening of peptides
derived from MART-1 has permitted the identification of two
APs: the nonapeptide AAGIGILTV (MART-127-35) and the
decapeptide EAAGIGILTV (MART-126-35). More recently, it
was reported that the MART-126-35 decapeptide was better
recognized by most tumor-reactive T-cells than the nonapep-
tide.16 However, a drawback of both peptides was their low
binding affinity to the HLA-A2 molecule. It is established that
most human APs bind to MHC class-I molecules through 2 or
3 conserved anchoring residues, such as leucine at the 2-position
in HLA-A2 binding APs.17 Hence, synthetic analogues of the
MART-1 epitopes having an alanine replaced by a leucine
residue were produced and showed considerably increased HLA-
A2 binding affinity (i.e., ALGIGILTV and ELAGIGILTV).9
The ELAGIGILTV analogue was found to be more im-
munogenic than the natural peptide,9 whereas ALGIGILTV was
not recognized anymore by most Melan-A tumor-reactive CTL
lines and clones. Drastic conformational differences were indeed
observed between the central regions of the two bound peptides,
and showed that the decapeptide ELAGIGILTV central region
(i.e., GIG29-31) adopts a “zigzag” structure, in which the
hydrophobic I30 side chain bulges out from the center of the
decamer.18 Hence, the HLA-bound decapeptide would be better
suited than its nonameric analogue to engage in more intimate
contacts with TCRs (see Figure 1) for enhanced CTL activation.
In any event, the heteroclytic properties observed in vitro for
the modified decapeptide ELAGIGILTV led to its selection for
vaccination clinical trials on melanoma patients.19
Such an AP analogue, however, still lacks acute immunoge-
nicity. This disappointing clinical observation may be due again,
at least in part, to the low stability of ELAGIGILTV in
biological fluids.3 Nevertheless, the considerable amount of work
that has been performed on ELAGIGILTV both at the medicinal
chemistry and clinical levels thus led us to select it as a first
AP analogue model for developing our approach to nonpeptidic,

1600 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Table 1. HLA-A2 Binding Affinities of the 23 ELA Analogues
Douat-Casassus et al.
Scheme 1. Structural Representation of the Synthesized ELA
Mimetics of First and Second Generationsa
RFI
compd
sequencea
R group
ratiob
ELA ELAGIGILTV
EAA EAAGIGILTV
1.0
0.63
0.80
0.97
0.73
1.71
1.61
1.34
0.85
0.90
1
ELAGYGILTV
2
ELAGFGILTV
3
ELAGWGILTV
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
ELAG(ΨCH2NH)YG(ΨCH2NH)ILTV
ELAG(ΨCH2NH)WG(ΨCH2NH)ILTV
ELAG(ΨCH2NH)W-Amba-LTV
ELA-NH-(CH2)2-(R)G-GILTV
ELA-NH-(CH2)2-(R)G-GILTV
ELA-NH-(CH2)2-(R)G-GILTV
ELA-NH-(CH2)2-(R)G-GILTV
ELA-NH-(CH2)2-(R)G-GILTV
ELA-NH-(CH2)2-(R)G-GILTV
âALA-NH-(CH2)2-(R)G-GILTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-NH-(CH2)2-(R)G-Amba-LTV
âALA-G(ΨCH2NH)Tpi-Amba-LTV
âALA-G(ΨCH2NH)Tic-Amba-LTV
CH2-C6H4-NO2
CH2-C6H5
CH2-naphtalene 0.70
CH2-quinoline
CH2-C6H4-OH
CH2-C6H11
0.79
0.83
0.86
0.96
0.86
1.31
0.95
0.25
0.86
CH2-C6H5
CH2-C6H4-NO2
CH2-quinoline
CH2-C6H5
(CH2)2-C6H5
(CH2)2-indole
(CH2)2-C6H4-OH 0.42
CO-CH2-C6H5 0.54
CO-CH2-indole 1.02
0.81
0.64
a For chemical structures, see Schemes 1 and 2 and Supporting
Information. b Normalized HLA-A2 binding affinities of ELA analogues
relative to that of ELA expressed as relative fluorescence intensities (RFI).
a
NB: for 13, the N-terminal glutamic acid (E) residue was replaced by
a â-alanine residue (see text and Table 1).
Generation and HLA-A2 Binding Affinity of ELA Ana-
logues. Thus, our preliminary modifications of ELA were simply
based on the substitution of I30 (P5) by aromatic amino acids
in the aim of assessing the role played by central aromatic pegs
in the HLA-A2-restricted AP recognition process by TCRs.
Three peptides having Y, F, or W at P5 (1-3, Table 1) were
synthesized and assayed for their capacity to maintain binding
to the HLA-A2 molecule. All three bind with a relative affinity
comparable to that of ELA, but better than that of EAA, the
natural antigenic peptide.
The introduction of functional spacers was then envisioned
in the ELA central region. To mimic the central tetrapeptidic
fragment GIGI (P4-P7), nonpeptidic spacer needs to reproduce
as closely as possible the conformation of this central “zigzag”
region, while allowing adequate orientation of the anchor
residues L2 and V10. This is the reason why we initially
decreased as little as possible the central peptide features and
worked our way up stepwise to more drastic structural changes.
The first classical backbone modification implemented was the
introduction of reduced peptide bonds.25 Notwithstanding previ-
ous studies that have shown that such alterations are often
detrimental to the binding affinity to the MHC class-I mol-
ecules,26 not only one but two of these reduced bonds were
envisaged between each glycine residue and its right-adjacent
residue in peptides 1 and 3 to give 4 and 5 (Table 1). Docking
of the corresponding altered peptides in the A6-TCR/Tax/HLA-
A2 complex indicated backbone geometries similar to that of
the ELA peptide. These two modified ELA peptides 4 and 5
were then synthesized and evaluated for their HLA-A2 binding
affinity. Strikingly, they turned out to be better ligands than
the ELA peptide, showing a 2-fold increase in relative affinity
as compared to that of the natural antigenic peptide EAA (Table
1).
thought of pursuing our modification by replacing the dipeptide
GI29,30 by such a type of spacer and exploiting the secondary
amine function to attach various organic motifs that could mimic
the central TCR-oriented side chain functionalities of the natural
aromatic amino acids F, Y, and W used in peptides 1-5. We
were further inclined to do so on the basis of previous studies
that have shown that the central part of an AP can be replaced
either by poly-N-acylated amines or by aliphatic spacers without
loss of binding affinity to MHC molecules.12,27 Among the
various spacers we considered, N-(2-aminoethyl)glycine, which
is classically used in peptide nucleic acid chemistry,28 was
identified as the unit responding the best to the structural criteria
imposed by the docking procedure, the most important of which
being the positioning of the central tether in the same vicinity
as the side chain of ordinary residues at P5 (Scheme 1, first
generation). When this spacer, bearing different organic motifs,
was introduced in the peptide sequence in place of GI29,30, the
resulting altered peptides 7-12 displayed docked conformations
similar to that of ELA. The L2 and V10 anchor side chains
were positioned as in ELA, and the modified central parts bulged
out of the HLA-A2 binding groove, hence presenting their
organic tether toward the CDR3R loop of the TCR (not shown).
An analogue of the N-benzylated construct 8 (i.e., 13, Table 1)
was also made by replacing its N-terminal glutamic acid (E)
residue by a â-alanine residue, for such a modification of the
ELA N-terminus has been shown to render the molecule much
more resistant to peptidase-mediated degradation.10,29 The seven
pseudo-peptides of this series (i.e., 7-13) expressed affinities
to the HLA-A2 molecule at a level comparable to that of the
ELA peptide (Table 1). Moreover, the presence of the N-terminal
â-alanine in 13 did not alter its HLA-A2 binding affinity (Table
1). Again, all of these pseudo-peptides were of course much
better ligands than the natural antigenic peptide EAA (Table
1).
Since the incorporation of 1,2-diaminoethyl units (i.e., reduced
G residue) in the peptide backbone was found to have a positive
influence on binding affinity to the HLA-A2 molecule, we

Synthetic Anticancer Vaccine Candidates
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1601
Scheme 2. Examples of ELA Mimetics of Third Generation
these ELA mimetics to penetrate deeper in between the TCR
CDR3 loops.22e Such transformations were found to be highly
detrimental to the binding affinity to the HLA-A2 molecule,
except for the indole-bearing compounds (see RFI values for
17 and 19-20 vs 18 and 21 in Table 1). The docking in the
A6-TCR/Tax/HLA-A2 complex structure of the indole-bearing
construct 21 illustrates particularly well the adequacy of the
modifications made on ELA, as it shows an overall backbone
conformation and a hydrogen bonding connection with the HLA-
A2 molecule identical to those of the ELA peptide (Figure 2).
Both ELA and 21 express the same level of binding affinity to
the HLA-A2 molecule (Table 1).
The satisfactory level of HLA-A2 affinity observed with the
AMBA-containing compounds of second generation led us to
examine an additional type of transformation aimed at rigidify-
ing the peptidomimetic backbone, while fixing the orientation
of central aromatic motifs toward the hydrophobic cavity of
the TCR CDR3 loops. In agreement with the results of our in
silico docking analysis, two additional analogues were thus
envisaged bearing either the cyclic Phe mimic N-Fmoc-L-
tetrahydroisoquinoline-3-carboxylic acid (i.e., Tic)31 or the cyclic
Trp mimic N-Fmoc-L-tetrahydronorharman-3-carboxylic acid
(Tpi)32 in place of the first GI fragment. These building blocks
were coupled to a reduced glycine residue at P4 to constitute
CR-cyclized variants of the N-(2-aminoethyl)glycine spacer. This
last series of AMBA-containing ELA mimetics again featured
a bioresistant â-alanine residue at P1 (i.e., 22 and 23, Table 1
and Scheme 2). Unfortunately, these modifications did not lead
to any improvement of the binding affinity to the HLA-A2
molecule (Table 1).
Overall, our binding affinity results demonstrate that the
central GIGI part of ELA can be replaced by nonpeptidic
elements on which can be attached various organic motifs
without drastic changes in binding affinity to the HLA-A2
molecule. It is admittedly difficult to predict each time the
affinity behavior of a given peptidomimetic from its docking
in silico using the LigandFit protocol. All designed ELA
mimetics do not experimentally respond to the expectations
drawn from the preliminary docking analysis, but all peptido-
mimetics having a high affinity to the HLA-A2 molecule feature
structural modifications that were shown by LigandFit to allow
docking in between the HLA-A2 and the TCR molecules in a
manner similar to that of the starting peptide ELA. Thus, the
LigandFit program constitutes a valuable tool to screen potential
peptidomimetic structures prior to their synthesis.
Figure 2. LigandFit docking of ELA and its mimetic 21 in the A6-
TCR/Tax/HLA-A2 complex. Top: InsightII superimposition of 21
(green) and ELA (orange) after docking in the Tax binding site; the
CDR3-R loop is in blue, the CDR3-â is in gray, and the MHC class-I
HLA-A2 molecule is in yellow. Bottom: InsightII representation of
the H-bonding network between 21 and the HLA-A2 molecule.
In addition to this spacer, the replacement of the second GI31,32
was envisaged. For the sake of simplicity of synthesis, com-
monly used building blocks were screened. Among the different
hydrophobic dipeptide mimetics described in the literature and
commercially available,30 the only one that gave satisfactory
docking results was the 3-aminomethyl benzoic acid (AMBA,
see Scheme 1, second generation). When AMBA was thus
introduced in place of GI31,32, the peptide backbone folding in
the HLA-A2 binding cleft was unchanged. Remarkably, the
phenyl ring of AMBA was oriented in a manner similar to that
of the I32 side chain at P7 (see AMBA positioning in Figure
2), filling the hydrophobic HLA-A2 pocket with no perturbation
of the positioning of the main anchor residue side chains at P2
and P10. This AMBA spacer was thus combined either with a
reduced peptide bond at P4 (i.e., 6, Table 1) or with the N-(2-
aminoethyl)glycine at P4 and P5. In this later case, the
N-terminal glutamic acid residue was also replaced by a
â-alanine residue (vide supra). A second generation of eight
peptidomimetics was thus synthesized and evaluated for their
binding affinity to the HLA-A2 molecule (i.e., 14-21, Table 1
and Scheme 1). In this series, the length of the central aromatic
tether was also increased by one carbon member either as a
methylene unit (i.e., 17-19) or as a carbonyl unit (i.e., 20-
21). These homologations were made in the aim of enabling

1602 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Douat-Casassus et al.
Scheme 3. Solid-Phase Synthesis of Peptidomimetics 7-10 and
12/13 and the AMBA-Containing Peptidomimetics 14-17a
Scheme 4. Synthesis of N-Fmoc-N-(2-Aloc)-aminoethylglycine
27 and Peptidomimetic 11a
a Key: (a) standard SPPS; (b) RCHO (0.25 M), 2% AcOH-DCE, 2 h,
filtration, NaBH(OAc)3, DCE, 3 h; (c) Fmoc-Gly-H, NaBH3CN, 1%
AcOH-DMF, 10 h; (d) SPPS; (e) final deprotection and cleavage from
resin, TFA-TIS-H2O (95:2.5:2.5). NB: t-butyl protecting groups on Thr
(T) and Glu (E) side chains are omitted for clarity; for 13 to 17, the
N-terminal glutamic acid (E) residue was replaced by a â-alanine residue
(see text).
a Keys: (a) Boc2O, THF, 0 °C, 95%; (b) BrCH2CO2Et, TEA, THF, 56%;
(c) (i) Aloc-Cl, Na2CO3, H2O, dioxane, (ii) LiOH, H2O, 76%; (d) (i) TFA,
(ii) Fmoc-OSu, NaOH, H2O, CH3CN, 91%; (e) SPPS; (f) 20% piperidine-
DMF; (g) 27, HBTU, DIEA, DMF; (h) SPPS; (i) Pd(PPh3)4, PhSiH3, DCM;
(j) RCHO (i.e., hydroxybenzaldehyde), NaBH3CN, 1% AcOH-DMF, 10
h; (k) final deprotection and cleavage from resin, TFA-TIS-H2O (95:2.5:
2.5).
Chemical Synthesis. All modified peptides and peptidomi-
metics 1-23 (Table 1) were manually prepared via classical
solid-phase peptide synthesis (SPPS) on a Wang resin using
the Fmoc/t-butyl strategy. N-Fmoc-protected amino acids were
condensed using HBTUa and DIEA as coupling agents in DMF
for 45 min. N-Fmoc deprotections were carried out in 4:1
DMF-piperidine solvent mixture. The N-terminal residue was
introduced as an N-Boc-protected amino acid. All compounds
were then simultaneously cleaved from the resin and fully
deprotected by the addition of TFA in the presence of TIS and
water (EDT was added to this cleavage cocktail in the case of
the indole-containing compounds 18 and 21). Purification of
all compounds was accomplished by semipreparative RP-HPLC,
and their identification was carried out by positive-mode
electrospray ionization (ESI) mass spectrometry (MS) (see Table
1, Experimental Section and Supporting Information).
The synthesis of the peptide analogues containing reduced
peptide bonds (Table 1, 4-6) required a reductive amination
between the N-R-amino function of the peptide fragments
anchored to the resin and N-Fmoc-glycinal (i.e., Fmoc-Gly-H),
preliminary generated in solution by reduction of its corre-
sponding morpholine amide using LiAlH4.33 Reduction of the
imine intermediate was carried out using NaBH3CN in DMF
containing 1% (v/v) of AcOH.25 This procedure was also applied
to the synthesis of the rigidified peptidomimetics 22 and 23
(Table 1 and Scheme 2). These peptides were obtained in
satisfactory yields (54-68%) and purities, after purification by
semipreparative RP-HPLC (see Experimental Section and
Supporting Information, Table S1).
anchored to the resin and the corresponding aldehyde RCHO
using NaBH(OAc)3 as reducing agent.34 The resulting secondary
amine was then subjected to a second reductive amination
reaction with Fmoc-Gly-H using NaBH3CN as reducing agent.27
Following condensation with the N-terminal ELA or â-ALA
(i.e., for 13-17) fragment, final deprotection, and cleavage from
the resin, the peptidomimetics 7-10 and 12-17 were obtained
in yields ranging from 27 to 55%, after semipreparative RP-
HPLC purification (see Experimental Section and Supporting
Information, Table S1).
The second strategy consisted in preparing first the entire
peptide sequence using the required amino acid building blocks
and the N-Fmoc-N-(2-Aloc)-aminoethylglycine spacer 27, the
synthesis of which is depicted in Scheme 4, and then in attaching
any organic side chain after release of the Aloc-protected
secondary amine. This strategy was thought to be advantageous
when using aldehydes RCHO bearing a second function
sensitive to SPPS conditions, for they could thus potentially be
attached via reductive amination as an unprotected entity just
before the release of the construct from the support. This turned
out to be successful only for the attachment of the 4-hydroxy-
benzaldehyde (Scheme 4), which led to the formation of the
first-generation construct 11 in a yield of 46%, after RP-HPLC
purification (see Experimental Section and Supporting Informa-
tion, Table S1). All attempts to introduce the N-unprotected
3-formylindole group by this approach failed.
The third strategy was based on an approach previously used
for peptoid synthesis35,36 and was applied here for the synthesis
of the second-generation AMBA-containing peptidomimetics
18 and 19, for which central side chains (RCH2-) could be
attached as free primary amines via nucleophilic substitution
reactions. Bromoacetic acid served to introduce the comple-
mentary electrophilic function and was directly condensed with
the AMBA amino group of the solid-supported peptidomimetic
precursor (Scheme 5). The resulting R-bromo acetyl amide 28
was then reacted with either Nin-Boc-tryptamine (29)37 or O-tBu-
tyramine (30),38 and each product was submitted to reductive
amination with Fmoc-Gly-H as before. Completion of the
synthesis via SPPS furnished the desired peptidomimetics 18
Four different strategies were followed for the synthesis of
ELA mimetics of first and second generations (Schemes 3-6).
For incorporation of N-linked central side chains (RCH2-) that
are nonsensitive to the SPPS conditions used (Scheme 3), a first
reductive amination was performed between the free N-terminal
R-amino function of the glycinyl residue of the fragments
a Abbreviations: HBTU, 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyl-
uroniumhexafluorophosphate; DIEA, N,N′-diisopropylethylamine; DMF,
dimethylformamide; DCE, 1,2-dichloroethane, DIC, 1,3-di-iso-propylcar-
bodiimide; TFA, trifluoroacetic acid; TIS, triisopropylsilane; EDT, ethanedi-
thiol.

Synthetic Anticancer Vaccine Candidates
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1603
Scheme 5. Synthesis of AMBA-containing Peptidomimetics 18
and 19a
20 and 21 in satisfactory yields and purities (see Experimental
Section and Supporting Information, Table S1).
T-Cell Recognition. All 23 modified ELA peptides and
peptidomimetics were assayed for their capacity to stimulate a
panel of seven different Melan-A-specific T-cell clones (ex-
pressing different TCR Vâ segments) previously obtained from
melanoma infiltrating T-cells (TILs) or from peptide-stimulated
PBLs. HLA-A2+ human mutant T2 cells pulsed with each of
the 23 compounds in 10 µM solutions were used to stimulate
T-cell clones (see Experimental Section). Both the natural
MART-126-35 epitope EAA, naturally expressed by tumor cells,
and its analogue ELA were used as positive controls. Since the
23 ELA analogues and mimetics were synthesized at different
periods of our investigation, they were assayed in two separate
series of experiments (Table 2). Their capacity to induce T-cell
responses was evaluated by determining the fraction of clone
cells producing IFN-γ by intracellular labeling. Cytokine
production by CTL clones correlates with the level of TCR
engagement and intracellular cytokine labeling of the fraction
of activated T-cells clones provides a fine quantitative assess-
ment of the relative T-cell activation.39 Results obtained with
all T-cell-stimulating antigenic peptide analogues and peptido-
mimetics are reported in Table 2 and are compared with
responses obtained in the same experiment with the control
peptides EAA and ELA. The simple replacement of the TCR-
contacting I residue at P5 by Y, F, or W (i.e., peptides 1-3)
resulted in full conservation of the recognition by several
T-cell clones. All three peptides stimulated the same four clones,
although at a different level, and one (i.e., 3) also activated
another clone (M77.84). Therefore, the nature of the aromatic
residue thus introduced at P5 of the ELA sequence influences
both the specificity and the level of T-cell recognition. These
first results confirmed those of previous studies showing that
the introduction of an aromatic side chain at the center of the
ELA sequence was not detrimental to its recognition by TCRs,
even though the corresponding peptides were not as good HLA-
A2 ligands as ELA (Table 1).22d The next and first series of
pseudo-peptides (i.e., 4-6), still bearing the same types of
central aromatic amino acid residues (i.e., Y or W) was credited
with a high potential for efficient T-cell stimulation, because
of their high relative binding affinity to the HLA-A2 molecule
(Table 1). We were then obviously disappointed by the total
lack of activity of the two doubly reduced peptide analogues 4
and 5. However, the replacement of the second reduced GI
fragment of 5 by the AMBA unit led to the high-affinity ligand
6 that was capable of stimulating four different Melan-A-specific
CTL clones with similar efficacy than the ELA peptide. Indeed,
6 induced a response quite similar to that of ELA for two clones
(MEL1-37 and M17.29 ELA1) and slightly lower for two others
(M199.2.7 and CDM41.ELA1) (Table 2, second series).
a Keys: (a) Fmoc-AMBA-OH, HBTU, DIEA, DMF; (b) 20% piperi-
dine-DMF; (c) bromoacetic acid (0.4 M), DIC (2 M), DMF, 5 min, rt; (d)
29 or 30, DMSO (see text); (e) Fmoc-Gly-H, NaBH3CN, 1% AcOH-DMF,
10 h; (f) SPPS; (g) final deprotection and cleavage from resin using TFA-
EDT-TIS-H2O (92.5:2.5:2.5:2.5) for 18 and TFA-TIS-H2O (95:2.5:2.5)
for 19.
Scheme 6. Synthesis of AMBA-Containing Peptidomimetics 20
and 21a
a Keys: (a) Fmoc-Gly-OH, HBTU, DIEA, DMF; (b) 20% piperidine-
DMF; (c) Fmoc-Gly-H, NaBH3CN, 1% AcOH-DMF, 10 h; (d) 31 or 32
(see text), DIC, DMF; (e) SPPS; (f) final deprotection and cleavage from
resin using TFA-TIS-H2O (95:2.5:2.5) for 20 and TFA-EDT-TIS-H2O
(92.5:2.5:2.5:2.5) for 21.
These last and very encouraging results prompted us to know
whether or not our more drastically modified ELA mimetics of
first, second, and/or third generations could mimic as well the
activity of the ELA peptide. Among the 17 compounds of these
series (i.e., 7-23), it is another AMBA-containing structure that
revealed itself as a particularly interesting construct. This
peptidomimetic 21 features four modifications of the starting
ELA peptide sequence. The N-terminal residue has been
replaced by a â-alanine, the first GI fragment by the N-(2-
aminoethyl)glycine spacer, the side chain of the central isoleu-
cine residue at P5 by a N-carbonylmethylene-linked indole, and
the second GI fragment by the AMBA unit. Gratifyingly, this
heavily modified structure turned out to be the best elicitor of
IFN-γ secretion, as it stimulated at significant levels four out
and 19 in satisfactory yields and purities (see Experimental
Section and Supporting Information, Table S1).
The last strategy was followed to build peptidomimetics 20
and 21 bearing side chains acylated to the N-(2-aminoethyl)-
glycine spacer. To the N-terminal amino group of their AMBA-
containing peptidomimetic precursor was successively attached
Fmoc-Gly-OH via standard SPPS coupling and Fmoc-Gly-H
via reductive amination (Scheme 6). The resulting free secondary
amine function was then acylated with either phenylacetic acid
(31) or indolacetic acid (32) using DIC in DMF. Completion
of the synthesis via SPPS furnished the desired peptidomimetics

1604 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Douat-Casassus et al.
Table 2. IFN-γ Production of T-Cell Clones after Stimulation by T2 Cells Pulsed with Different ELA-Derived Peptides and Mimeticsa
First Series
compd
M77.84
M199.2.7
M199.3.2
MEL1.37
CDM41 ELA1
M17.29 ELA1
10C10
EAA
ELA
1
2
3
7
10
11
87
90
4
32
90
0
41
38
0
7
15
0
85
88
78
83
54
0
83
86
84
71
81
0
73
77
63
58
47
64
69
46
75
65
32
48
55
0
50
71
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
Second Series
compd
M77.84
M199.2.7
MEL1.38
MEL1.37
CDM41
M17.29
10C10
EAA
ELA
6
40
53
0
27
30
22
25
80
86
26
44
82
84
82
89
35
33
28
31
76
78
84
27
60
76
0
21
12
0
a Results are expressed as percentages of IFN-γ producing cells measured by flow cytometry upon intracellular IFN-γ labeling (see Figure 4 and Experimental
Section).
Figure 3. Space filling representations of 21 and ELA docked in the
A6-TCR/Tax/HLA-A2 complex. ELA mimetic 21 (A) and ELA (B)
viewed from above the HLA-A2 molecule. The R-helical domains of
HLA-A2 are shown in blue, and its â-sheet floor is shown in magenta.
Peptidic structures are shown in Corey-Pauling-Kultun representations
(yellow) with their central P5 position in red. Both structures display
a similar global binding mode with an identical orientation of their
HLA-A2 anchoring side chains. The main difference resides in the
orientation of the indole side chain of 21 as compared to that of I5 of
ELA.
Figure 4. Melan-A/MART-1-specific clone Mel1-37 was stimulated
for 6 h with T2 cells, pulsed with EAA, ELA, 6, or 21 or without (as
negative control). T-cells were fixed at the end of the stimulation,
permeabilized, stained, and analyzed on a FACScan. Percentages of
IFN-γ-labeled cells are indicated, as well as mean fluorescence
intensities in parentheses.
of the seven clones tested. This compound was even capable
of activating the clone MEL1-37 at a level superior to that
reached with ELA, as shown by the significantly higher mean
fluorescence intensity of T-cell clones stimulated by this
peptidomimetic 21 (1425) than by ELA (1019) (Table 2 and
Figure 4). It is worth noting that the HLA-A2 binding affinity
of 21 is just about equal to that of ELA (Table 1). Therefore,
the higher capacity of this analogue essentially relies on the
TCR/peptidomimetic interaction. Although the fine character-
istics of TCR/peptide interactions that condition T-cell activation
and response remain unclear, Davies et al.40,41 recently proposed,
in their two-step model for T-cell recognition, that it is the final
folding of the two CDR3 loops of the TCR over the antigenic
peptide that is crucial for T-cell activation outcome. In the case
of 21, its docking study revealed a binding mode to the MHC
class-I HLA-A2 molecule highly similar to that of ELA (Figures
2 and 3). The first step of T-cell recognition should therefore
probably be identical for both the peptide ELA and its mimetic
21. The capacity of 21 to induce stronger IFN-γ secretion than
ELA could then be a consequence of stronger and/or more
appropriate contacts of its indole acetyl side chain with the
MEL1-37 TCR CDR3 loops. The fact that 18, the analogue of
21 lacking the side chain carbonyl group, was unable to activate
T-cell clones could be due to its higher flexibility, whereas the
amide link of 21 could better constrain its indolic side chain to
orient itself toward the TCR binding zone. In any event, these
results and those obtained here with other peptidomimetics are
new experimental illustrations to the fact that certain T-cells
are very sensitive to even minor structural changes of the central
part of their ligand. Nevertheless, this fine structural specificity
is not a characteristic common to all T-cells. This is clearly
illustrated by our observation that the CDM41-ELA1 PBL clone
responded in a similar fashion to several ELA analogues (i.e.,
6, 7, 10, 11, and 21), all bearing polar aromatic units in their
central part (i.e., indole, p-nitroxyphenyl and quinoline, Table
2). One might thus argue that the CDR3 loops of the CDM41-
ELA1 receptor are flexible enough to accommodate sterically
and electronically different organic motifs of the same general
type during the second step of their recognition of these
peptidomimetics.42,43 In sharp contrast, one of the clones used
in this study (i.e., 10C10) did not tolerate any of the alterations
we performed on its ligand including the simple substitution of
Ile at P5 by other standard aromatic residues (i.e., in 1 to 3).
Interestingly, another clone (M199.2.7) not tolerating either these
simple modifications was, however, capable of recognizing the
ELA mimetics 6 and 21. This suggests that nonpeptidic
alterations could be perhaps more efficient than peptidic ones

Synthetic Anticancer Vaccine Candidates
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1605
the parental peptide structure from the 1A07 PDB file enables us
to gain insight of the backbone modification consequence on the
interaction with the protein.
in generating active analogues of wild-type epitopes, as a
plausible consequence of their higher structural diversity.
Conclusion
Materials and Methods. All reagents were either purchased
from Aldrich, Acros, or Fluka. Amino acids and their derivatives
were purchased from Advanced Chem. Tech. (U.K.), Bachem, or
Novabiochem (Switzerland). The Wang resin and HBTU were
purchased from Novabiochem. All organic solvents were of
analytical quality and Milli-Q (Millipore) water was used for reverse
phase (RP) HPLC analyses and purifications. Peptide and pepti-
domimetic syntheses were performed manually in a glass reactor
(vide infra). RP-HPLC analyses were performed on a Thermo
system using a Chromolith performance RP-18e column (4.6 ×
100 mm, 5 µm) with P1000 XR pumps at a flow rate of 3 mL
min-1 (see Table 1 and Supporting Information). Semipreparative
purifications were performed on the same Thermo system using a
Waters semi-prep deltapack RP-C18 column (3.9 mm × 300 mm,
100 Å pore size, 15 µm). The mobile phase was composed of 0.1%
(v/v) TFA/H2O (Solvent A) and 0.1% TFA/CH3CN (Solvent B).
A gradient elution (0-40 min: 90% to 50% A) was applied at a
flow rate of 4 mL min-1. Column effluent was monitored by UV
detection at 214 and 254 nm using a Thermo UV 6000 LP diode
array detector. Flash column chromatography was carried out under
positive pressure using 40-60 µm silica gel (Merck) and the
indicated solvents. Evaporations were conducted under reduced
pressure at temperatures less than 45 °C unless otherwise noted.
Further drying of the residues was accomplished under high
vacuum. NMR spectra of samples in the indicated solvent were
run at 400 MHz on a Bruker DPX spectrometer. Electrospray
ionization mass spectrometric low- and high-resolution data
(ESIMS, HRMS) were obtained from the Mass Spectrometry
Laboratory at the European Institute of Chemistry and Biology
(IECB), Pessac, France, and from the Centre Re´gional de Mesures
Physiques de l’Ouest (CRMPO), Universite´ Rennes 1, Rennes,
France.
N-tert-Butoxycarbonyl-1,2-diaminoethane (25).46 To an ice-
cooled solution of 1,2-diaminoethane (24, 2.23 g, 30 mmol) in dry
THF (30 mL) was added dropwise a solution of Boc2O (1.96 g, 9
mmol) in THF (15 mL). The reaction mixture was stirred at 0 °C
for 30 min, then at room temperature overnight. After evaporation
of the solvent, the solid mixture was dissolved in ethyl acetate (200
mL) and washed with brine (3 × 30 mL). The organic layer was
dried over Na2SO4, filtered, and evaporated to give 25 as a white
solid (2.73 g, 97%): 1H NMR (DMSO-d6, 400 MHz) δ 1.38 (s,
9H), 2.53 (m, 2H), 2.91 (m, 2H), 6.75 (bs, 1H, NH); 13C NMR
(DMSO-d6, 100 MHz) δ 156.5, 78.5, 44.2, 42.3, 28.1.
N-(N′-tert-Butoxycarbonyl-2-aminoethyl)glycine ethyl ester
(26).47 To an ice-cooled solution of 25 (1 g, 6.25 mmol) and
diisopropylethylamine (0.215 mL, 1.25 mmol) in dry THF (20 mL)
was slowly added a solution of ethylbromoacetate (0.206 mL, 1.85
mmol) in dry THF (20 mL). The reaction mixture was stirred 1 h
at 0 °C and then 3 h at room temperature. The mixture was then
poured into brine and extracted with ethyl acetate (3 × 20 mL).
The combined organic layers were washed again with brine, dried
over Na2SO4, filtered, and evaporated to give a residue, which was
submitted to flash column chromatography, eluting with CH2Cl2-
MeOH (10:1), to furnish pure 26 as a colorless oil (0.26 g, 56%):
1H NMR (CDCl3, 400 MHz) δ 1.26 (t, J ) 7.1 Hz, 3H), 1.44 (s,
9H), 2.83 (m, 2H), 3.27 (m, 2H), 3.46 (s, 2H), 4.19 (q, J ) 7.1 Hz,
3H), 5.22 (bs, 1H, NH); 13C NMR (CDCl3, 100 MHz): 177.8,
161.2, 83.0, 65.5, 55.6, 53.9, 33.8, 19.7; ESIMS (positive mode):
m/z (rel. intensity) 247 (MH+, 100%).
Replacing the central part of MHC class-I bound antigenic
peptides (APs) by nonpeptidic residues is an approach to
peptidomimetics that has previously been followed in attempts
essentially aimed at increasing the binding affinity of the altered
peptides to MHC molecules, while improving their bioresistance
in applications in vivo. To the best of our knowledge, none of
these modifications were specifically implemented with the
possibility of altering and modulating TCR/peptide interactions
in mind. The experimental findings described therein demon-
strate the value of our LigandFit-based rational design of
antigenic peptidomimetics in which the central part is replaced
by a pseudo-peptidic unit that can be equipped with a side chain
capable of interacting with the CDR3 loops of TCRs. Most of
the 20 altered peptides, thus designed, retained or even exceeded
the binding affinity of the parent ELA peptide to the MHC
class-I HLA-A2 molecule. Furthermore, five of them (i.e., 6,
7, 10, 11, and 21) were capable of inducing T-cell responses.
The Melan-A-specific T-cell clones used in this study appeared
particularly sensitive to polar aromatic side chains and, more
especially, to indole-containing motifs. The two most potent
compounds (i.e., 6 and 21) feature such a motif. The most
spectacular results of this study are those obtained with the
second-generation peptidomimetic 21. Although this heavily
modified peptide structure is not the best HLA-A2 ligand of
those we designed therein, it is capable of inducing IFN-γ
secretion by one of the T-cell clones tested at a level higher
than that reached by the already optimized analogue ELA
peptide. This compound undeniably constitutes a valuable lead
for the development of peptidomimetic-based vaccine-type
immunotherapies against melanoma. Future work will address
the pursuit of the immunological evaluation of this peptidomi-
metic and of its structural modification aimed at improving its
bioresistance while maintaining or increasing its immuno-
genecity. Results will be reported in due course.
Experimental Section
Docking Methodology. All molecular modeling calculations
were performed on a Silicon Graphics Octane workstation using
Catalyst 4.7 and Cerius2 4.7 molecular modeling softwares. The
peptidomimetics used in this docking study are listed in Table 1.
The following procedure was used to perform the docking study.
First, the three-dimensional structure of the A6-TCR/Tax/HLA-
A2 complex was taken from the PDB file 1A07.21 The Tax ligand
and all the water molecules were removed. Hydrogen atoms were
added using the Cerius2 templates for the protein residues. The 23
peptide analogue and mimetic structures were constructed in
Catalyst. Partial charges were assigned using the Gasteiger method44
as implemented in Cerius2. A site model based on the Tax ligand
docked into the protein was identified by LigandFit. Then, the
docking45 of the 23 ligands employed the following protocol: (a)
a Monte Carlo conformational search for generating a candidate
ligand conformation, (b) selection of a ligand position and orienta-
tion based on comparing the shape of the binding site model with
that of the ligand conformation, and (c) evaluation of the goodness
of docking by computing the dock energies using a grid-based
energy calculation. The dock energy is expressed as the sum of
the ligand internal strain energy and the interaction energy of the
ligand with the protein. Then, the position and the orientation of
the ligand are optimized by minimizing the dock energy, with
respect to rigid body translations and rotations of the ligand using
a steepest descent method. The docked conformations found are
then clustered, and the first 20 are saved. A superimposition of the
highest docked structure obtained for each peptidomimetic with
N-Aloc-N-(N′-Fmoc-2-aminoethyl)glycine (27).48 To a solution
of 26 (250 mg, 1.0 mmol) in a 2:3 water-dioxane solvent mixture
(10 mL) and Na2CO3 (534 mg, 5.0 mmol) was added dropwise a
solution of allylchloroformate (0.118 mL, 1.12 mmol) in dioxane
(2 mL). This reaction mixture was stirred at room temperature for
90 min, after which time it was poured into water (5 mL) and
extracted with ethyl acetate (2 × 20 mL). The combined organic
layers were washed twice with 1 N HCl (10 mL) and brine (10
mL), dried over Na2SO4, filtered, and evaporated. The resulting

1606 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Douat-Casassus et al.
residue was then submitted to flash column chromatography, eluting
Formation of the Resin-Bound Peptidomimetics 7-10 and
12-17 by N-Alkylation on Solid Support. To the free resin-bound
glycine amine (Scheme 3) suspended in a solution of 2% (v/v)
AcOH in DCE was added the appropriate aldehyde RCHO in a
quantity sufficient to reach a concentration of 0.25 M in the reaction
medium. After being mixed for 2 h, the reaction mixture was
filtered, and a freshly prepared 0.25 M solution of NaBH(OAc)3
in DCE was added and allowed to react for 3 h. The reaction
mixture was then quenched with MeOH (3 times) and filtered, and
the resin was washed twice with CH2Cl2, MeOH, and again CH2-
Cl2. The next reductive amination with the secondary amine
anchored to the support was then carried out overnight under
shaking by using a solution of N-Fmoc-Gly-H (10 equiv) and
NaBH3CN (10 equiv) in a 1% AcOH-DMF solvent mixture. After
filtration and washings with MeOH and CH2Cl2, the last coupling
reactions were performed as before according to the general
procedure described above. For the peptidomimetic 17, the same
procedure was applied except that the N-alkylation step was carried
out overnight using phenylacetaldehyde (20 equiv) in the presence
of NaBH3CN (20 equiv) in a solution of 1% AcOH in DMF.
Formation of the Resin-Bound Peptidomimetic 11. The
construction of its peptide fragment was done as described above
via standard SPPS protocols, and N-Aloc-N-(N′-Fmoc-2-amino-
ethyl)glycine (27, 3 equiv) was introduced using the coupling
reagents HBTU (3 equiv) and DIEA (5 equiv). The resin-supported
Aloc-containing peptide (see Scheme 4) was then treated at room
temperature with Pd(PPh3)4 (0.1 equiv) and PhSiH3 (20 equiv) in
dry CH2Cl2 with occasional stirring for 20 min at room temperature.
This deprotection step was repeated once. The completion of the
Aloc group removal was controlled via the chloranil test. After
filtration, the resin was washed twice with CH2Cl2, MeOH, and
again CH2Cl2. To the resin (50 mg, 23 µmol) then suspended in a
solution of 2% ACOH in DCE (2 mL) was added 4-hydroxybenz-
aldehyde (53.5 mg, 46 µmol, 20 equiv) and NaBH(OAc)3 (93.5
mg, 46 µmol, 20 equiv). The resin was gently shaken overnight,
and the reaction mixture was quenched three times with MeOH
and washed twice with CH2Cl2, MeOH, and again CH2Cl2. The
completion of this alkylation step was checked via the chloranil
test.
Formation of the Resin-Bound Peptidomimetics 18 and 19.
To the free resin-bound AMBA-derived primary amine (100 mg,
0.556 mmol, see Scheme 5) were added bromoacetic acid (111 mg,
0.80 mmol), as a 0.4 M solution in DMF, and DIC (0.156 mL, 1.0
mmol) as a 2.0 M solution in DMF. The resin was allowed to react
in this media for 5 min, after which time the solution of reagents
was drained, and the resulting brominated resin 28 was washed
with DMF (5 × 5 mL). To this functionalized resin was then added
a 1.0 M DMSO solution of either Nin-Boc-tryptamine (29, 260 mg,
1 mmol) or O-tBu-tyramine (30, 193 mg, 1 mmol).50 These mixtures
were allowed to react for 3 h, after which time they were filtered
and the resins were washed twice with DMF, MeOH, CH2Cl2, and
again DMF (ca. 5 mL of each solvent), before being submitted to
the same reductive amination conditions as those described for
7-10, using N-Fmoc-Gly-H. The final amino acid coupling
reactions were then performed as before according to the general
SPPS procedure described above.
with ethyl acetate-cyclohexane (1:1), to furnish the desired N-Aloc
1
derivative of 26 as a colorless oil (242 mg, 76%): H NMR (CDCl3,
400 MHz) δ 1.28 (t, J ) 7.0 Hz, 3H), 1.43 (s, 9H), 3.27 (m, 2H),
3.46 (t, J ) 5.7, 2H), 3.97 (d, J ) 6.4 Hz, 2H), 4.21 (m, 2H), 4.60
(dd, J ) 5.4, 16.4 Hz, 2H), 5.35 (m, 2H), 5.90 (m, 1H); 13C NMR
(CDCl3, 100 MHz) δ 170.5, 156.5, 133.0, 117.8, 67.0, 61.9, 50.3,
49.1, 39.7, 28.8, 14.6; HRMS (ESIMS, positive mode) calcd for
C13H22N2O6Na 325.1375, found 325.1367.
To a solution of the N-Aloc derivative of 26 (200 mg, 0.63 mmol)
in THF (5 mL) was added a solution of LiOH (46 mg, 1.90 mmol)
in water (5 mL). This mixture was stirred overnight at room
temperature, evaporated to remove THF, and lyophilized. The
resulting crude carboxylic acid was then dissolved in TFA (5 mL)
and let standing without stirring for 30 min. After removal of TFA
by evaporation, the resulting TFA salt was dissolved in water and
treated with 1 N NaOH until pH 9. A solution of Fmoc-OSu (213
mg, 0.63 mmol) in CH3CN (10 mL) was then added, and the
mixture was stirred at room temperature overnight, after which time
it was poured into water (20 mL) and extracted once with Et2O
(10 mL). The aqueous phase was acidified with 1 N HCl until pH
1 and extracted twice with CH2Cl2 (2 × 20 mL). The combined
organic layers were washed twice with brine (2 × 10 mL), dried
over Na2SO4, filtered, and evaporated to furnish 27 as a yellow
powder (260 mg, 91%): IR (CHCl3) 2360, 1697, 1540, 1474, 1450,
1141, 740 cm-1; 1H NMR (CDCl3, 400 MHz) δ 3.15 (s, 2H), 3.32
(m, 2H), 3.92 (d, J ) 12.7 Hz, 2H), 4.28 (d, J ) 7.33 Hz, 2H),
4.49 (m, 2H), 4.83 (d, J ) 6.0 Hz, 1H), 5.15 (m, 1H), 5.25 (m,
1H), 5.88 (m, 1H); 13C NMR (CDCl3, 100 MHz) δ 172.1, 170.6,
157.0, 144.7, 143.5, 141.6, 134.0, 128.8, 128.5, 128.2, 126.0, 125.8,
121.2, 121.0, 117.6, 66.2, 61.9, 49.7, 47.6, 46.8; ESIMS (positive
mode) m/z (rel. intensity) 425.5 (MH+, 48); HRMS (ESIMS,
positive mode) calcd for C23H24N2O6Na 447.1532, found 447.1537.
General Procedure for Solid-Phase Peptidomimetic Synthesis.
The synthesis of all peptidomimetics and peptide fragments thereof
was carried out on a Wang resin (0.89, 0.91, or 0.93 mmol/g) via
standard SPPS protocols.49 To a solution of N-Fmoc-Val-OH (10
equiv relative to resin loading) in dry CH2Cl2 (a few drops of DMF
were required to ensure complete dissolution) under N2 was added
di-iso-propylcarbodiimide (DIC, 5 equiv). The reaction mixture was
stirred for 20 min at 0 °C. In the meantime, the resin was suspended
in dry CH2Cl2 and allowed to swell for 20 min. After filtration, the
symmetrical N-Fmoc-Val anhydride solution and DMAP (0.1 equiv)
were added, and the resin was shaken 4 h. After filtration, the resin
was successively washed twice with DMF, MeOH, and CH2Cl2
and again DMF. The Fmoc protecting group was removed by using
twice a solution of piperidine in DMF (1:4, v/v) for 3 min then for
7 min. After filtration, the resin was successively washed as before.
The SPPS was then continued using N-Fmoc amino acids (3 equiv)
or N-Fmoc-AMBA-OH (3 equiv) in the presence of HBTU (3
equiv) and DIEA (5 equiv) in DMF. Each coupling reaction was
performed for 45 min, after which time the resin was washed as
before. The last amino acid (i.e., Glu or âAla) was introduced as
an N-Boc-protected residue. Completion of each coupling reaction
was monitored via the trinitrobenzenesulfonic acid (TNBS) test.
After final deprotection and cleavage from the resin, all compounds
were checked for purity by RP-HPLC and purified by semiprepara-
tive RP-HPLC (see Materials and Methods). Compound identifica-
tion was accomplished by ESIMS analysis in positive mode.
Formation of the Resin-Bound Peptidomimetics 20 and 21.
The free resin-bound AMBA-derived primary amine (100 mg, 0.556
mmol see Scheme 6) was first coupled with Fmoc-Gly-OH using
standard SPPS conditions according to the general procedure
described above. After removal of the Fmoc-protecting group, the
resulting free amine was condensed with Fmoc-Gly-H, under the
same reductive amination conditions as those described for 4-6,
to furnish the AMBA-containing resin-bound N-Fmoc-aminoethyl
glycine construct. Its free secondary amine function was then
acylated with either indolacetic acid (101 mg, 0.56 mmol, 10 equiv)
or phenylacetic acid (87 mg, 0.56 mmol, 10 equiv) in the presence
of DIC (90 µL, 0.56 mmol, 10 equiv) in DMF (5 mL). The resin
was shaken for 1 h, after which time the mixture was filtered and
the resin was washed as before. Both acylation reactions were
repeated once. The final amino acid coupling reactions were then
Formation of the Resin-Bound Peptidomimetics 4-6 by
Reductive Amination on Solid Support. The ELA mimetics 4-6
were obtained by replacing one or two peptide bonds by the
aminomethylene (CH2-NH) surrogate. To the N-terminal R-amino
group of the resin-bound peptide was added N-Fmoc-Gly-H (2.5
equiv relative to the resin loading) and NaBH3CN (7.5 equiv) in a
solution of 1% (v/v) AcOH in DMF for 2 h, as previously
described.25 The completion of the reaction was monitored by the
TNBS test. All coupling reactions that were performed after the
introduction of a reduced peptide bond were carried out as described
above.

Synthetic Anticancer Vaccine Candidates
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1607
performed as before according to the general SPPS procedure
described above.
the presence of ELA analogues or mimetics (10 µM). After being
washed, 2 × 105
peptide-pulsed T2 cells were incubated with 105
Formation of the Resin-Bound Peptidomimetics 22 and 23.
Peptidomimetics 22 and 23 were synthesized according to the
general SPPS procedure described above, using either N-Fmoc-
Tic-OH (3 equiv) or N-Fmoc-Tpi-OH (3 equiv). The introduction
of the reduced peptide bond was performed again via condensation
reactions with N-Fmoc-Gly-H (3 equiv) using NaBH3CN (10 equiv).
These reactions were carried out overnight. The last coupling
reactions were then performed as before according to the general
SPPS procedure described above.
General Procedure for TFA Cleavage. Final deprotection and
cleavage from the resin were carried out using a TFA-H2O-TIS
mixture (95:2.5:2.5) for 4 h, except for the indole-containing
constructs 17 and 21, for which EDT was added to the cleavage
cocktail [i.e., TFA-H2O-TIS-EDT (92.5:2.5:2.5:2.5)]. All the
peptidomimetic constructs were then precipitated using cold Et2O,
filtered, dissolved in a water-acetonitrile-TFA mixture and
lyophilized. They were then purified by semipreparative RP-HPLC
(see Materials and Methods).
Cell Lines. The TAP-deficient human mutant cell line CEMx721
T2 (T2) used as presenting cell was a generous gift from T. Boon
(Ludwig Institute for Cancer Research, Brussels, Belgium). The
EBV-B-transformed cell line LAZ-338 used as feeder cell was a
gift from T. Hercend (Vertex Pharmaceutical, Abingdon, England).
These two cell lines were cultured in RPMI-1640 (Sigma-Aldrich,
Saint Quentin Fallavier, France) containing 10% FCS (Biowest,
Nuaille, France), penicillin-streptomycin (100U PEN/mL-100 µg
STREP/mL (Sigma) and glutamine (1 nM) (Sigma).
T-Cell Clones. Melanoma-reactive CD8 T-cell clones (M77.84,
M199.2.7, M199.3.2, 10C10, M17.29ELA1, MEL1.37, and
CDM41ELA1), specific for Melan-A/MART-1 antigen, were
obtained from TILs or from PBLs stimulated ex vivo by peptide
pulsed antigen presenting cells (APC).51 Melanoma-specific clones
were expanded using a polyclonal T-cell stimulation protocol, as
previously described.52 Briefly, 2000 T-cell clones were distributed
in 96-well plates with 150 µL of culture medium (RPMI with 8%
human serum and IL-2 150 U/mL) and irradiated feeder cells: LAZ
EBV-B cells (104/well) and allogenic PBLs (105/well). Polyclonal
T-cell amplification was performed using 15 µg/mL of PHA-L
(Sigma) at the onset of the culture. Culture medium was changed
2 days later to remove PHA-L and then twice a week. T-cell clones
were split when their number exceeded 2 × 105/well.
T-cell clones in 200 µL of RPMI-1640/10% FCS, in the presence
of 10 µg/mL of brefeldin A (Sigma). The cultures were realized in
round-bottom 96-well plates and incubated for 6 h at 37 °C in 5%
CO2 humidified atmosphere. Cells were then fixed 10 min at room
temperature in a solution of PBS 4% paraformaldehyde (Electron
Microscopy Sciences), washed, and stored at 4 °C until labeling.
Flow Cytometric Analysis of Intracellular IFN-γ. Stimulated
T-cell clones were stained for IFN-γ using the method described
by Jung and co-workers.53 Briefly, fixed cells were washed in PBS
containing 0.1% BSA and 0.1% saponin (Sigma) and stained for
30 min at room temperature with the PE-conjugated mouse anti-
human IFN-γ MAb (BD Pharmingen) at a concentration of 5 µg/
mL, which was shown to give optimal staining, and then washed
twice. Reagent dilutions and washes were made with PBS contain-
ing 0.1% BSA and 0.1% saponin. After being stained, cells were
resuspended in PBS and 1 × 104 events were analyzed on a
FACScan flow cytometer using Cell Quest software (Beckton
Dickinson). Results are expressed by the fraction of stimulated
T-cells secreting IFN-γ and by the mean fluorescence intensity of
IFNγ labeled cells which correlates with the intensity of clone
activation.
Acknowledgment. Financial support from the University of
Bordeaux 1 (Interdepartmental competitive grant, 2001-2004),
the European Institute of Chemistry and Biology (Pessac,
France, 2003-2005), and La Ligue Contre le Cancer (Label-
lisation 2002 and 2005) is gratefully acknowledged. The authors
also thank the Conseil Re´gional d’Aquitaine for a postdoctoral
fellowship for C.D.-C.
Supporting Information Available: 1H and 13C NMR spectra
of compounds 25-27, RP-HPLC chromatograms and low-resolu-
tion ESI+ mass spectra of ELA peptide analogues 1-3 and
peptidomimetics 4-23, HPLC-determined purities of all analogues
and peptidomimetics 1-23 (Table S1), and high-resolution mass
1
spectrometric data and TOCSY-based H NMR chemical shifts
assignments of 21 (Table S2). This material is available free of
charge via the Internet at http://pubs.acs.org.
References
HLA-A*0201 Binding Assay. The relative peptidomimetic
binding capacity to HLA-A*0201 was assessed using the TAP-
deficient HLA-A2+ T2 cells. These cells were loaded with saturating
concentrations (100 µM) of the different peptidomimetics and, in
parallel, of the natural Melan-A/MART-126-35 peptide (EAAGIG-
ILTV) and of its peptide analogue A27L (ELAGIGILTV), with
human â2-microglobulin (100 ng/mL, Sigma), by overnight incuba-
tion at 37 °C in serum-free medium (RPMI-1640). Cells were
washed twice before staining with mouse mAb HB54 (HLA-A2
specific) (ATCC) to measure HLA-A2 Ag expression. Briefly, cells
were stained with the specific or isotype control mAb for 30 min
at 4 °C. After being washed, cells were incubated for 30 min with
the secondary goat F(ab′)2 fragment anti-mouse IgG-PE (Beckman
Coulter, Marseille, France) and analyzed on a FACScan flow
cytometer using Cellquest software (Becton Dickinson, Grenoble,
France). Increased fluorescence intensity, compared to fluorescence
in the absence of peptide or peptidomimetic is expressed as the
ratio fluorescence intensity (RFI) for HLA-A2 expression calculated
as follows: mean fluorescence intensity with the specific mAb
divided by mean fluorescence intensity with the negative control.
This RFI correlates with the HLA-A2 binding affinity. The relative
HLA-A2 affinity of each peptide analogue or peptidomimetic
construct is expressed as the ratio between the RFI induced by the
optimized analogue ELA and the tested peptide analogues and
peptidomimetics. RFI ratios superior or inferior to 1 respectively
indicate a higher or lower HLA-A2 binding affinity of the tested
peptide analogues and peptidomimetics compared to that of ELA.
T-Cell Stimulation. T2 cells were incubated for 1 h at 37 °C in
(1) Kagi, D.; Hengartner, H. Different roles for cytotoxic T cells in the
control of infections with cytopathic versus noncytopathic viruses.
Curr. Opin. Immunol. 1996, 8, 472-477.
(2) Greenberg, P. D. Adoptive T cell therapy of tumors: mechanisms
operative in the recognition and elimination of tumor cells. AdV.
Immunol. 1991, 49, 281-355.
(3) Cresswell, P. Antigen processing and presentation. Immunol. ReV.
2005, 205, 5-7.
(4) Davis, M. M.; Boniface, J. J.; Reich, Z.; Lyons, D. S.; Hampl, J.;
Arden, B.; Chien, Y. Ligand recognition by alpha beta T cell
receptors. Annu. ReV. Immunol. 1998, 16, 523-544.
(5) (a) Larche, M.; Wraith, D. C. Peptide-based therapeutic vaccines for
allergic and autoimmune diseases. Nat. Med. 2005, 11, 69-76. (b)
Boon, T.; Coulie, P. G.; Van den Eynde, B. J.; van der Bruggen, P.
Human T cell responses against melanoma. Annu. ReV. Immunol.
2006, 24, 175-208.
(6) Rosenberg, S. A.; Yang, J. C.; Restifo, N. P. Cancer immuno-
therapy: moving beyond current vaccines. Nat. Med. 2004, 10, 909-
915.
(7) Starr, T. K.; Jameson, S. C.; Hogquist, K. A. Positive and negative
T cell selection. Annu. ReV. Immunol. 2003, 21, 139-176.
(8) (a) Kawakami, Y.; Eliyahu, S.; Sakaguchi, K.; Robbins, P. F.;
Rivoltini, L.; Yannelli, J. R.; Appella, E.; Rosenberg, S. A. Identifica-
tion of the immunodominant peptides of the MART-1 human
melanoma antigen recognized by the majority of HLA-A2-restricted
tumor infiltrating lymphocytes. J. Exp. Med. 1994, 180, 347-352.
(b) Coulie, P. G.; Brichard, V.; Van Pel, A.; Wo¨lfel, T.; Schneider,
J.; Traversari, C.; Mattei, E.; De Plaen, C.; Lurquin, J.; Szikora, P.;
Renauld, J. C.; Boon, T. A new gene coding for a differentiation
antigen recognized by autologous cytolytic T lymphocytes on HLA-
A2 melanomas. J. Exp. Med. 1994, 180, 35-42.

1608 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Douat-Casassus et al.
(9) Valmori, D.; Fonteneau, J. F.; Maranon Lizana, C.; Gervois, N.;
Lie´nard, D.; Rimoldi, D.; Jongeneel, C. V.; Jotereau, F.; Cerottini,
J. C.; Romero, P. Enhanced generation of specific tumor-reactive
CTL in vitro by selected Melan-A/MART-1 immunodominant peptide
analogs. J. Immunol. 1998, 160, 1750-1758.
(10) Blanchet, J. S.; Valmori, D.; Dufau, I.; Ayyoub, M.; Nguyen, C.;
Guillaume, P.; Monsarrat, B.; Cerottini, J. C.; Romero, P.; Gairin, J.
E. A new generation of Melan-A/MART-1 peptides that fulfill both
increased immunogenicity and high resistance to biodegradation:
implication for molecular anti-melanoma immunotherapy. J. Immunol.
2001, 167, 5852-5861.
(11) Brinckerhoff, L. H.; Kalashnikov, V. V.; Thompson, L. W.; Yam-
shchikov, G. V.; Pierce, R. A.; Galavotti, H. S.; Engelhard, V. H.;
Slingluff, C. L. Terminal modifications inhibit proteolytic degradation
of an immunogenic MART-127-35 peptide: Implications for peptide
vaccines. Int. J. Cancer 1999, 83, 326-334.
(12) (a) Hin, S.; Bianco, A.; Zabel, C.; Jung, G.; Walden, P. Mimetics of
a T cell epitope based on poly-N-acylated amine backbone structures
induce T cells in vitro and in vivo. J. Biol. Chem. 2001, 276, 48790-
48796. (b) Bianco, A.; Brock, C.; Zabel, C.; Walk, T.; Walden, P.;
Jung, G. New synthetic non-peptide ligands for classical major
histocompatibility complex class I molecules. J. Biol. Chem. 1998,
273, 28759-28765.
Douat-Casassus, C.; Marchand Geneste, N.; Diez, E.; Aznar, C.;
Picard, P.; Geoffre, S.; Huet, A.; Bourguet-Kondracki, M.-L.; Gervois,
N.; Jotereau, F.; Quideau, S. Covalent modification of a melanoma-
derived antigenic peptide with a natural quinone methide. Preliminary
chemical, molecular modelling and immunological evaluation studies.
Mol. BioSyst. 2006, 2, 240-249. (e) Gagnon, S. J.; Borbulevych, O.
Y.; Davis-Harrison, R. L.; Turner, R. V.; Damirjian, M.; Wojnaro-
wicz, A.; Biddison, W. E.; Baker, B. M. T cell receptor recognition
via cooperative conformational plasticity. J. Mol. Biol. 2006, 363,
228-243.
(23) Venkatachalam, C. M.; Jiang, X.; Oldfield, T.; Waldman, M.
LigandFit: a novel method for the shape-directed rapid docking of
ligands to protein active sites. J. Mol. Graph. Model. 2003, 21, 289-
307.
(24) (a) Apostolopoulos, V.; Yu, M.; Corper, A. L.; Li, W.; McKenzie, I.
F.; Teyton, L.; Wilson, I. A.; Plebanski, M. Crystal structure of a
non-canonical high affinity peptide complexed with MHC class I: a
novel use of alternative anchors. J. Mol. Biol. 2002, 318, 1307-
1316. (b) Apostolopoulos, V.; Yu, M.; Corper, A. L.; Teyton, L.;
Pietersz, G. A.; McKenzie, I. F.; Wilson, I. A.; Plebanski, M. Crystal
structure of a non-canonical low-affinity peptide complexed with
MHC class I: A new approach for vaccine design. J. Mol. Biol. 2002,
318, 1293-1305. (c) Saper, M. A.; Bjorkman, P. J.; Wiley, D. C.
Refined structure of the human histocompatibility antigen HLA-A2
at 2.6 Å resolution. J. Mol. Biol. 1991, 219, 277-319. (d) Hillig, R.
C.; Coulie, P. G.; Stroobant, V.; Saenger, W.; Ziegler, A.; Hulsmeyer,
M. High-resolution structure of HLA-A*0201 in complex with a
tumour-specific antigenic peptide encoded by the MAGE-A4 gene.
J. Mol. Biol. 2001, 310, 1167-1176.
(13) Hofbauer, G. F.; Kamarashev, J.; Geertsen, R.; Boni, R.; Dummer,
R. Melan A/MART-1 immunoreactivity in formalin-fixed paraffin-
embedded primary and metastatic melanoma: frequency and distribu-
tion. Melanoma Res. 1998, 8, 337-343.
(14) Kawakami, Y.; Eliyahu, S.; Delgado, C. H.; Robbins, P. F.;
Sakaguchi, K.; Appella, E.; Yannelli, J. R.; Adema, G. J.; Miki, T.;
Rosenberg, S. A. Identification of a human melanoma antigen
recognized by tumor-infiltrating lymphocytes associated with in vivo
tumor rejection. Proc. Natl. Acad. Sci. 1994, 91, 6458-6462.
(15) Pittet, M. J.; Valmori, D.; Dunbar, P. R.; Speiser, D. E.; Lienard,
D.; Lejeune, F.; Fleischhauer, K.; Cerundolo, V.; Cerottini, J. C.;
Romero, P. High frequencies of naive Melan-A/MART-1-specific
CD8+ T cells in a large proportion of human histocompatibility
leukocyte antigen (HLA)-A2 individuals. J. Exp. Med 1999, 190,
705-716.
(16) Romero, P.; Gervois, N.; Schneider, J.; Escobar, P.; Valmori, D.;
Pannetier, C.; Steinle, A.; Wolfel, T.; Lienard, D.; Brichard, V.; van
Pel, A.; Jotereau, F.; Cerottini, J. C. Cytolytic T lymphocyte
recognition of the immunodominant HLA-A*0201-restricted Melan-
A/MART-1 antigenic peptide in melanoma. J. Immunol. 1997, 159,
2366-2374.
(17) Ruppert, J.; Sidney, J.; Celis, E.; Kubo, R. T.; Grey, H. M.; Sette,
A. Prominent role of secondary anchor residues in peptide binding
to HLA-A2.1 molecules. Cell, 1993, 74, 929-937.
(25) Quesnel, A.; Zerbib, A.; Connan, F.; Guillet, J. G.; Briand, J. P.;
Choppin, J. Synthesis and antigenic properties of reduced peptide
bond analogues of an immunodominant epitope of the melanoma
MART-1 protein. J. Pept. Sci. 2001, 7, 157-165.
(26) (a) Calbo, S.; Guichard, G.; Bousso, P.; Muller, S.; Kourilsky, P.;
Briand, J. P.; Abastado, J. P. Role of peptide backbone in T cell
recognition. J. Immunol. 1999, 162, 4657-4662. (b) Stemmer, C.;
Quesnel, A.; Pre´vost-Blondel, A.; Zimmermann, C.; Muller, S.;
Briand, J. P.; Pircher, H. Protection against lymphocytic choriomen-
ingitis virus infection induced by a reduced peptide bond analogue
of the H-2Db-restricted CD8+ T cell epitope GP33. J. Biol. Chem.
1999, 274, 5550-5556. (c) Ayyoub, M.; Mazarguil, H.; Monsarrat,
B.; Van den Eynde, B.; Gairin, J. E. A structure-based approach to
designing non-natural peptides that can activate anti-melanoma
cytotoxic T cells. J. Biol. Chem. 1999, 274, 10227-10234.
(27) (a) Krebs, S.; Lamas, J. R.; Rognan, D. Substituting nonpeptidic
spacers for the T cell receptor-binding part of class I major
histocompatibility complex-binding peptides. J. Biol. Chem. 1998,
273, 19072-19079. (b) Rognan, D.; Scapozza, L.; Folkers, G.; Daser,
A. Rational design of nonnatural peptides as high-affinty ligands for
the HLA-B*2705 human leukocyte antigen. Proc. Natl. Acad. Sci.
1995, 92, 753-747.
(18) Sliz, P.; Michielin, O.; Cerottini, J. C.; Luescher, I.; Romero, P.;
Karplus, M.; Wiley, D. C. Structures of two closely related but
antigenically distinct HLA-A2/melanocyte-melanoma tumor-antigen
peptide complexes. J. Immunol. 2001, 167, 3276-3284.
(19) (a) Ayyoub, M.; Zippelius, A.; Pittet, M. J.; Rimoldi, D.; Valmori,
D.; Cerottini, J. C.; Romero, P.; Lejeune, F.; Lienard, D.; Speiser,
D. E. Activation of human melanoma reactive CD8+ T cells by
vaccination with an immunogenic peptide analog derived from Melan-
A/melanoma antigen recognized by T cells-1. Clin. Cancer Res. 2003,
9, 669-677. (b) Lienard, D.; Rimoldi, D.; Marchand, M.; Dietrich,
P. Y.; van Baren, N.; Geldhof, C.; Batard, P.; Guillaume, P.; Ayyoub,
M.; Pittet, M. J.; Zippelius, A.; Fleischhauer, K.; Lejeune, F.;
Cerottini, J. C.; Romero, P.; Speiser, D. E. Ex vivo detectable
activation of Melan-A-specific T cells correlating with inflammatory
skin reactions in melanoma patients vaccinated with peptides in IFA.
Cancer Immun. 2004, 4, 1-19.
(20) (a) Rudolph, M. G.; Wilson, I. A. The specificity of TCR/pMHC
interaction. Curr. Opin. Immunol. 2002, 14, 52-65. (b) Rudolph,
M. G.; Stanfield, R. L.; Wilson, I. A. Annu. ReV. Immunol. 2006,
24, 419-466.
(21) Garboczi, D. N.; Ghosh, P.; Utz, U.; Fan, Q. R.; Biddison, W. E.;
Wiley, D. C. Structure of the complex between human T-cell receptor,
viral peptide and HLA-A2. Nature 1996, 384, 134-140.
(22) (a) Pinilla, C.; Martin, R.; Gran, B.; Appel, J. R.; Boggiano, C.;
Wilson, D. B.; Houghten, R. A. Exploring immunological specificity
using synthetic peptide combinatorial libraries. Curr. Opin. Immunol.
1999, 11, 193-202. (b) Rubio-Godoy, V.; Dutoit, V.; Zhao, Y.;
Simon, R.; Guillaume, P.; Houghten, R.; Romero, P.; Cerottini, J.
C.; Pinilla, C.; Valmori, D. Positional scanning-synthetic peptide
library-based analysis of self- and pathogen-derived peptide cross-
reactivity with tumor-reactive Melan-A-specific CTL. J. Immunol.
2002, 169, 5696-5707. (c) Pinilla, C.; Rubio-Godoy, V.; Dutoit, V.;
Guillaume, P.; Simon, R.; Zhao, Y.; Houghten, R. A.; Cerottini, J.
C.; Romero, P.; Valmori, D. Combinatorial peptide libraries as an
alternative approach to the identification of ligands for tumor-reactive
cytolytic T lymphocytes. Cancer Res. 2001, 61, 5153-5160. (d)
(28) (a) Nielsen, P. E. Peptide Nucleic Acids: Protocols and Applications;
Horizon Scientific Press: Portland, U.K., 1999. (b) Baker, E. S.;
Hong, J. W.; Gaylord, B. S.; Bazan, G. C.; Bowers, M. T. PNA/
dsDNA complexes: site specific binding and dsDNA biosensor
applications. J. Am. Chem. Soc. 2006, 128, 8484-8492.
(29) Guichard, G.; Zerbib, A.; Le Gal, F.-A.; Hoebeke, J.; Connan, F.;
Choppin, J.; Briand, J.-P.; Guillet, J.-G. Melanoma Peptide MART-
1(27-35.) Analogues with Enhanced Binding Capacity to the Human
Class I Histocompatibility Molecule HLA-A2 by Introduction of
â-Amino Acid Residue: Implications for Recognition by Tumor-
Infiltrating Lymphocytes. J. Med. Chem. 2000, 43, 3803-3808.
(30) (a) Weerapana, E.; Imperiali, B. Peptides to peptidomimetics: towards
the design and synthesis of bioavailable inhibitors of oligosaccharyl
transferase. Org. Biomol. Chem 2003, 1, 93-99. (b) Qian, Y.;
Blaskovich, M.; Saleem, M.; Seong, M.; Wathen, S. P.; Hamilton,
A. D.; Sebti, S. M. Design and structural requirements of potent
peptidomimetic inhibitors of p21ras farnesyltransferase. J. Biol. Chem.
1994, 269, 12410-12413. (c) Cristau, M.; Devin, C.; Oiry, C.;
Chaloin, O.; Amblard, M.; Bernad, N.; Heitz, A.; Fehrentz, J.-A.;
Martinez, J. Synthesis and biological evaluation of bombesin
constrained analogues. J. Med. Chem. 2000, 43, 2356-2361. (d)
Lunney, E. A.; Para, K. S.; Rubin, J. R.; Humblet, C.; Fergus, J. H.;
Marks, J. S.; Sawyer, T. K. Structure-based design of a novel Series
of nonpeptide ligands that bind to the pp60src SH2 domain. J. Am.
Chem. Soc. 1997, 119, 12471-12476.
(31) Bolin, D. R.; Swain, A. L.; Sarabu, R.; Berthel, S. J.; Gillespie, P.;
Huby, N. J.; Makofske, R.; Orzechowski, L.; Perrotta, A.; Toth, K.;
Cooper, J. P.; Jiang, N.; Falcioni, F.; Campbell, R.; Cox, D.;
Gaizband, D.; Belunis, C. J.; Vidovic, D.; Ito, K.; Crowther, R.;
Kammlott, U.; Zhang, X.; Palermo, R.; Weber, D.; Guenot, J.; Nagy,

Synthetic Anticancer Vaccine Candidates
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1609
Z.; Olson, G. L. Peptide and peptide mimetic inhibitors of antigen
presentation by HLA-DR class II MHC molecules. Design, structure-
activity relationships, and X-ray crystal structures. J. Med. Chem.
2000, 43, 2135-2148.
(44) Gasteiger, J.; Marsili, M. A new model for calculating atomic charges
in molecules. Tetrahedron Lett. 1978, 19, 3181-3184.
(45) Vieth, M.; Hirst, J. D.; Dominy, B. N.; Daigler, H.; Brooks, C. L.
Assessing search strategies for flexible docking. J. Comput. Chem.
1998, 19, 1623-1631.
(46) (a) Krapcho, A. P.; Kuell, C. S. Mono-protected diamines N-tert-
Butoxycarbonyl-R,ω-alkanediamines from R,ω-alkanediamines. Synth.
Commun. 1990, 20, 2559-2564. (b) Kurtan, T.; Nesnas, N.; Li, Y.-
Q.; Huang, X.; Nakanishi, K.; Berova, N. chiral recognition by CD-
sensitive dimeric zinc porphyrin host. 1. chiroptical protocol for
absolute configurational assignments of monoalcohols and primary
monoamines. J. Am. Chem. Soc. 2001, 123, 5962-5973.
(47) (a) Aldrian-Herrada, G.; Rabie´, A.; Wintersteiger, R.; Brugidou, J.
Solid-phase synthesis of peptide nucleic acid (PNA) monomers and
their oligomerization using disulphide anchoring linkers. J. Pept. Sci.
1998, 4, 266-281.
(48) Jackson, R. W.; Tabone, J. C.; Howbert, J. J. Identification of TNF-R
inhibitors from a split-Pool library based on a tyrosine-proline
peptidomimetic scaffold. Biol. Med. Chem. Lett. 2003, 13, 205-
208.
(49) (a) Lloyd-Williams, P.; Albericio, F.; Giralt, E. Chemical Approaches
to the Synthesis of Peptides and Proteins; CRC Press: Boca Raton,
FL, 1997. (b) Chan, C.; White, P. D. Fmoc Solid Phase Peptide
Synthesis: a practical approach; Oxford Press: New York, 2000.
(50) Uno, T.; Beausoleil, E.; Goldsmith, R. A.; Levine, B. H.; Zuckermann,
R. N. New submonomers for poly N-substituted glycines (peptoids).
Tetrahedron Lett. 1999, 40, 1475-1478.
(51) (a) Gervois, N.; Guilloux, Y.; Diez, E.; Jotereau, F. Suboptimal
activation of melanoma infiltrating lymphocytes (TIL) due to low
avidity of TCR/MHC-tumor peptide interactions. J. Exp. Med. 1996,
183, 2403-2407. (b) Vignard, V.; Lemercier, B.; Lim, A.; Pan-
dolfino, M. C.; Guilloux, Y.; Khammari, A.; Rabu, C.; Echasserieau,
K.; Lang, F.; Gougeon, M. L.; Dreno, B.; Jotereau, F.; Labarriere,
N. Adoptive transfer of tumor-reactive Melan-A-specific CTL clones
in melanoma patients is followed by increased frequencies of
additional Melan-A-specific T cells. J. Immunol. 2005, 175, 4797-
4805.
(52) Jotereau, F.; Pandolfino, M. C.; Boudart, D.; Diez, E.; Dreno, B.;
Douillard, J. Y.; Muller, J. Y.; LeMevel, B. High-fold expansion of
human cytotoxic T-lymphocytes specific for autologous melanoma
cells for use in immunotherapy. J. Immunother. 1991, 10, 405-411.
(53) Jung, T.; Schauer, U.; Heusser, C.; Neumann, C.; Rieger, C. Detection
of intracellular cytokines by flow cytometry. J. Immunol. Methods
1993, 159, 197-207.
(32) Nagy, A.; Armatis, P.; Cai, R. Z.; Szepeshazi, K.; Halmos, G.;
Schally, A. V. Design, synthesis, and in vitro evaluation of cytotoxic
analogs of bombesin-like peptides containing doxorubicin or its
intensely potent derivative, 2-pyrrolinodoxorubicin. Proc. Natl. Acad.
Sci. 1997, 94, 652-656.
(33) Douat, C.; Heitz, A.; Martinez, J.; Fehrentz, J. A. Synthesis of
N-protected-amino aldehydes from their morpholine amide deriva-
tives. Tetrahedron Lett. 2000, 41, 37-40.
(34) Shannon, S. K.; Peacock, M. J.; Kates, S. A.; Barany, G. Solid-phase
synthesis of lidocaine and procainamide analogues using backbone
amide linker (BAL) anchoring. J. Comb. Chem. 2003, 5, 860-868.
(35) Burkoth, T. S.; Fafarman, A. T.; Charych, D. H.; Connolly, M. D.;
Zuckermann, R. N. Incorporation of unprotected heterocyclic side
chains into peptoid oligomers via solid-phase submonomer synthesis.
J. Am. Chem. Soc. 2003, 125, 8841-8845.
(36) Olivos, H. J.; Alluri, P. G.; Reddy, M. M.; Salony, D.; Kodadek, T.
Microwave-assisted solid-phase synthesis of peptoids. Org. Lett. 2002,
4, 4057-4059.
(37) Uno, T.; Beausoleil, E.; Goldsmith, R. A.; Levine, B. H.; Zuckermann,
R. N. New submonomers for poly N-substituted glycines (peptoids).
Tetrahedron Lett. 1999, 40, 1475-1478.
(38) Kruijtzer, J. A. W.; Hofmeyer, L. J. F.; Heerma, W.; Versluis, C.;
Liskamp, R. M. J. Solid-phase syntheses of peptoids using Fmoc-
protected N-substituted glycines: The synthesis of (retro)peptoids
of Leu-enkephalin and substance P. Chem. Eur. J. 1998, 4, 1570-
1580.
(39) Fonteneau, J. F.; Le Drean, E.; Le Guiner, S.; Gervois, N.; Diez, E.;
Jotereau, F. Heterogeneity of biologic responses of melanoma-specific
CTL. J. Immunol. 1997, 159, 2831-2839.
(40) Wu, L. C.; Tuot, D. S.; Lyons, D. S.; Garcia, K. C.; Davis, M. M.
Two-step binding mechanism for T-cell receptor recognition of
peptide-MHC. Nature 2002, 418, 552-558.
(41) Krogsgaard, M.; Davis, M. M. How T cells “see” antigen. Nature
Immunol. 2005, 6, 239-245.
(42) Reiser, J. B.; Darnault, C.; Gregoire, C.; Mosser, T.; Mazza, G.;
Kearney, A.; Van der Merwe, P. A.; Fontecilla-Camps, J. C.; Housset,
D.; Malissen, B. CDR3 loop flexibility contributes to the degeneracy
of TCR recognition. Nature Immunol. 2003, 4, 241-247.
(43) Valmori, D.; Gervois, N.; Rimoldi, D.; Fonteneau, J. F.; Bonelo, A.;
Lienard, D.; Rivoltini, L.; Jotereau, F.; Cerottini, J. C.; Romero, P.
Diversity of the fine specificity displayed by HLA-A0201-restricted
CTL specific for the immunodominant Melan-A/MART-1 antigenic
peptide. J. Immunol. 1998, 161, 6956-6962.
JM0613368