Download Manipulation of Epitope Function by Modification of Peptide

Biologicals (2001) 29, 197–207
doi:10.1006/biol.2001.0305, available online at http://www.idealibrary.com on
Manipulation of Epitope Function by Modification
of Peptide Structure: A Minireview
Ferenc Hudecz
Research Group of Peptide Chemistry, Hungarian Academy of Science, Eötvös L. University, P.O. Box 32,
Budapest 112, Hungary, H-1518
Abstract. We have explored various approaches to modify the immunrecognition of linear peptides
representing sequential or continuous topographic B-cell or T-cell epitopes. For these studies, epitopes
from herpes simplex virus (HSV) glycoprotein D (gD) and from mucin 1 and mucin 2 glycoproteins or
T-cell epitopes from 16 kDa and 38 kDa proteins of Mycobacterium tuberculosis were selected. To increase
antigenicity and immunogenicity we have prepared cyclic and chimaeric peptide variants as well as
epitope peptides with altered flanking regions and epitope-carrier conjugates containing multiple epitope
© 2001 The International Association for Biologicals
copies.
Key words: epitope peptide-conjugates, epitope-chimaera, epitope flanking, synthetic peptide antigens,
mucin antibody epitopes, protection against HSV infection, epitopes of M. tuberculosis proteins.
Introduction
Linear peptides representing sequential or continuous topographic B-cell or T-cell epitopes could be
poorly recognised by antibodies or T-cells specific
for the protein. To increase immunoreactivity,
including antigenicity and immunogenicity of peptides belonging to the above classes of B-cell or
T-cell epitopes, several experimental approaches
have been investigated.
Since B-cell epitope sequences are frequently
localised in -turns or loop regions of a protein, the
corresponding cyclic peptide could be a logical and
better mimicry of the native secondary structure
than the linear oligopeptide. Similarly, the stabilisation of steric, mainly secondary structure of a
B-cell epitope could be achieved by insertion of the
sequence into an appropriate site of a ‘‘host’’ protein sca#old. Thus the ‘‘guest’’ B-cell epitope in
chimaeric peptide could be more e#iciently recognised by epitope-specific antibodies compared with
the linear oligopeptide. Specific T-cell responses
induced by peptides containing a minimal size
functional T-cell epitope could be modulated by the
Address all correspondence to: Professor Ferenc
Hudecz, Research Group of Peptide Chemistry,
Hungarian Academy of Sciences, Eötvös L. University,
Budapest, Pázmány P. sétány. 2., H-1117, Hungary. Fax:
(36)-1-372-2620. E-mail: [email protected]
1045–1056/01/090197+11 $35.00/0
appropriate replacement of amino acid residues in
the epitope core and/or by the alteration of the
flanking regions connected to the N- and/or
C-terminal of the core. This approach could lead to
the development of T-cell antagonist/agonist compounds. Another strategy for increasing the sensitivity of antigen binding or the immunogenic
properties is the multiplication of copies of the same
or defined number of di#erent B- or T-cell epitopes
of microbial or tumour origin. To achieve this
polymerised epitope peptides could be prepared.
Alternatively, covalent epitope-carrier conjugates
could be synthesised using optimal size oligopeptides representing functional epitopes and protein or synthetic carriers (e.g. KLH, BSA, branched
chain polymeric polypeptides, multiple antigenic
peptides (MAP), a sequential oligopepide carrier
(SOC), oligotuftsin).
These approaches not only provide a better
understanding of the antigenic structure of proteins, but also contribute to the development of
synthetic antigens as artificial vaccines or diagnostic reagents. In this communication a brief overview will describe our recent results with (i)
epitopes from herpes simplex virus (HSV) glycoprotein D (gD); (ii) epitopes from mucin 1 and
mucin 2 glycoproteins and (iii) with T-cell epitopes
from 16 kDa and 38 kDa proteins of Mycobacterium
tuberculosis.
2001 The International Association for Biologicals
198
F. Hudecz
SALLEDPVG-NH2
CSALLEDPVG-NH2
I
CSALLEDPVG-NH2
II
H-YCCNPVACGRHYSC-NH2
III
H-SALLEDPVGK-NH2
H-CSALLEDPVGK-NH2
IV
V
1
α-[Tyr ]-conotoxin
H-YCCNPVACGDPVGC-NH2
1
bicyclic HSV-α-[Tyr ]-conotoxin
H-YCCNPVACGPDTRC-NH2
1
bicyclic MUC1-α-[Tyr ]-conotoxin
Figure 1. Schematic representation of cyclopeptides
containing the 281DPVG284 epitope sequence from glycoprotein D of herpes simplex virus in endo (IV, V) or in exo
(I–III) position.
Figure 2. Primary structure of -[Tyr1]-conotoxin and its
chimaera derivatives containing epitope sequences from
glycoprotein D of herpes simplex virus (281DPVG284) or
from mucin 1 glycoprotein (PDTR).
Cyclic and chimaeric peptides
IgG type antibody responses showed that the
bi-cyclic HSV--[Tyr1]-conotoxin chimaera induced
strong antibody responses in C57/Bl/6 mice but was
poorly immunogenic in CBA and BALB/c mice
(Fig. 3). Data obtained with the C57/Bl/6 serum
indicate that the polyclonal antibodies recognise
the DPVG motif presented in the bi-cyclic HSV-[Tyr1]-conotoxin and some reactivity was also found
with the monocyclic but not with the linear form of
the chimaera. We also found that the IgM monoclonal antibodies are able to recognise the linear
DPVG sequence, while the majority of the IgG
antibodies are directed to the same motif in a
conformation stabilised by double cyclisation. It is
interesting to note that the bi-cyclic HSV--[Tyr1]conotoxin chimaeric peptide and native -conotoxin
GI showed similar CD spectra in PBS, which might
suggest that these compounds also share similar
secondary structures.
Using the same -conotoxin GI sca#old we have
also synthesised a chimaeric peptide with an
inserted Pro-Asp-Thr-Arg (PDTR) epitope of the
mucin 1 glycoprotein (MUC1) instead of the native
8
Arg-His-Tyr-Ser12 tetramer (Fig. 2). MUC1 is
expressed and often secreted by epithelial cells
and contains a polypeptide core consisting of a
variable number (30–100) of repeats of a 20 amino
acid
sequence,
APDTRPAPGSTAPPAHGVTS.
Carcinoma cells produce MUC1 that is underglycosylated when compared to the version expressed by
normal resting cells. The incomplete glycosylation
exposes a normally cryptic polypeptide core whose
amino acid sequences could be recognised by the
immune system as epitopes. The appearance of these
epitope domains is largely restricted to cancer cells;
therefore MUC1 is an attractive target for cancer
immunodiagnosis and immunotherapy. It has been
demonstrated earlier that the PDTR sequence comprises the minimal epitope for MUC1 specific monoclonal antibodies HMFG1 (PDTR) and HMFG2
(DTR). Hydropathicity and secondary structure
HSV, with its two closely related serotypes
(HSV-1 and HSV-2), is one of the most common
infectious agents in humans. Glycoprotein D represents a major immunogenic component of the virion
envelope. Using prediction analysis of the sequence
of gD from HSV-1 and synthetic peptide-conjugates
with branched polypeptide poly[Lys(DL-Alam)],
(AK) where m3) the 281DPVG284 tetramer
sequence has been identified as the core of the
epitope within the highly hydrophilic 276–284
region possessing a -turn. Several cyclic versions
of S276ALLEDPVG284 nonapeptide were prepared
(Fig. 1) consisting of either (a) a head-to-side-chain
lactam ring, where the DPVG motif is situated
essentially outside the ring (exo-form) (I–III) or (b) a
side-chain-to-side-chain lactam ring between the
-carboxyl group of Glu and the -amino group of a
Lys residue attached to the C-terminus of the
sequence. In the latter case the DPVG motif is a part
of the six-residue lactam ring (endo-form) (IV–V).
The CD studies together with NMR data have indicated that the conformation of the peptides is highly
dependent on the relative position of the DPVG
epitope in the cyclic HSV peptides.1,2 Antibody
binding studies are in progress and the results will
be published elsewhere.
The DPVG core epitope has been inserted
as ‘‘guest’’ sequence in the ‘‘host’’ structure
of -conotoxin GI, a 13 residue peptide
(ECCNPACGRHYSC) isolated from the venom of
Conus geographus. The -conotoxin GI was selected
as a presenting sca#old since it also contains a
-turn in the 8–12 region which is stabilised by two
disulphide bridges in positions 2–7 and 3–13. Thus
the tetramer sequence of -conotoxin, 8Arg-His-TyrSer12 has been replaced by DPVG from the gD of
HSV (Fig. 2). The linear, the monocyclic and the
bi-cyclic forms of the chimaera have been synthesised and their immunogenicity was compared.3 The
characteristics of the primary and memory IgM and
Immunoreactivity of modified peptide epitopes
100
1200
(A)
(A)
CBA
BALB/c
C57B1/6
1000
80
800
60
Inhibition %
OD492
199
600
400
40
20
200
0
0
102
103
104
105
lg serum dilution
106
107
–20
0.0001
1200
(B)
60
Inhibition %
OD492
10
80
600
400
200
102
0.001
0.01
0.1
1
Peptide concentration (mmol/l)
(B)
800
0
10
100
CBA
BALB/c
C57B1/6
1000
0.001
0.01
0.1
1
Peptide concentration (mg/ml)
40
20
103
104
105
lg serum dilution
106
107
Figure 3. Antibody response induced by bi-cyclic HSV-Tyr1]-conotoxin chimaera in C57/Bl/6, CBA and BALB/c
mice. Binding of polyclonal antibodies to the bi-cyclic
HSV--[Tyr1]-conotoxin chimaera (A) and to [DPVG]-AK
conjugate (B) target.
prediction analysis confirmed by CD and NMR
experiments and supported by independent computational non-restrained study have identified a type
I -turn in the PDTR region. Based on these considerations we have prepared all three forms of the
chimaera (linear, monocyclic, bicylic) and compared
their antibody binding properties.4 As documented
by the CD spectra, the bi-cyclic MUC1--[Tyr1]conotoxin chimaera peptide showed a partially
ordered conformation with a -turn character. In
antibody binding studies the RIA data showed
that the MUC1--[Tyr1]-conotoxin chimaera was
recognised by monoclonal antibody HMFG1 specific
for the PDTR sequence, while no binding was
observed between monoclonal antibody HMFG2 and
0
–20
0.0001
Figure 4. Inhibition of the binding of monoclonal antibodies to [CAPDTRPAPG]-AK target antigen (A) or of
HMFG1 monoclonal antibody to [KAPDTRPAPG]-BSA
(B) with MUC1--[Tyr1]-conotoxin chimaera peptides in
competition RIA. Monoclonal antibody HMFG2 with
linear peptide MCnTx1 (), or with bi-cyclic peptide ().
Monoclonal antibody HMFG1 with linear peptide
MCnTx1 (), monocyclic (), or with bi-cyclic peptide
() and control peptide APDTRPAPG ().
various forms of the chimaera (Fig. 4A). Significant
di#erences were found in the HMFG1 recognition of
the PDTR epitope among the three forms of the
chimaera (Fig. 4B). HMFG1 using two di#erent target antigens (synthetic epitope-conjugate or native
MUC1) recognises the PDTR more e#iciently in
the linear than in the bi-cyclic compound, but no
reactivity was found with the monocyclic forms
200
F. Hudecz
Table 1. Amide I peaks in the FT-IR spectrum and monoclonal antibody 996 binding of
MUC2 peptide
Peptide
16
PTPTGTQ22
pTPTGTQ22
16
ptPTGTQ22
16
ptpTGTQ22
16
ptptGTQ22
16
High frequency
region
cm 1 (%)
Solvated amides
cm 1 (%)
-turns
cm 1 (%)
-turns
cm 1 (%)
1631 (4)
1630 (1)
1661 (5)
1660 (11)
1646 (15)
1644 (14)
1643 (17)
1643 (8)
1641 (8)
1673 (43)
1674 (46)
1674 (46)
1675 (37)
1675 (36)
1624 (3)
1625 (3)
IC50*
(mol/l)
3·8
6·5
33
160
6400
*IC50 is the peptide concentration required for 50% inhibition of 996 antibody binding to
BSA-[K12VTPTPTPTGTQTPT25] target antigen.
of MUC1--[Tyr1]-conotoxin chimaera, underlining
the importance of certain conformers stabilised by
double cyclisation.
Examples outlined above might indicate that both
the antibody binding and immunogenic properties
of the linear B-cell epitope could be improved by
restriction of the number of conformers present
in their linear form. However it is also clear that
there is a need for an appropriate design of the
cyclic or chimaeric version of the epitope. Small
or even minute changes in the position of the
epitope in the ‘‘presenting’’ sca#old could lead to
a dramatic alteration of antibody binding and/or
immunogenicity.
Replacement of amino acid residues in the epitope
flanking regions
Effect of D-amino acid substitution in a mucin 2 glycoproteion (MUC2) epitope on MUC2 specific monoclonal
antibody recognition. The majority of published work
reports on the replacement of all the -amino acid
residues in a peptide by the -enantiomers leading
to normal (in all--isomer peptides) or reversed
(in retro-all--isomer peptides) amide linkage.
Relatively little has been communicated on the
antibody recognition of peptides partially substituted by -amino acid residue(s) and to the best
of our knowledge no data are available on the
antibody binding to epitope peptides containing -amino acid residues in the flanking region.
In model experiments we aimed to perform a
systematic analysis to what extent could we substitute the N-terminal flanking region of the epitope by
-amino acids without significant decrease in
antibody recognition.
For this we have selected a B-cell epitope from
MUC2 whose applicability as a tumour marker in
colon carcinoma patients is under investigation.
Our previous studies showed that the core of the
epitope recognised by the protein specific monoclonal antibody (MAb 996) within the PTTTP
ITTTTTVTPTPTPTGTQT tandem repeat unit of
MUC2 glycoprotein is the 19TGTQ22 sequence. However, for optimal binding the 16PTPTGTQ22
sequence is required.5
We have studied the influence of -amino acid
substitution in the flanking region on the antibody recognition of the 19TGTQ22 epitope core.6
Analogue peptides corresponding to the optimal
epitope sequence (16PTPTGTQ22) have been prepared by the replacement of single or multiple
-amino acid residues at the N-terminal part of the
molecule (Table 1). According to previous studies
this portion of the all- 16PTPTGTQ22 peptide
possesses a -turn secondary structure important for
e#icient monoclonal antibody interaction.7 The
binding properties of sequentially modified peptides
(pTPTGTQ, ptPTGTQ, ptpTGTQ and ptptGTQ)
have been analysed by a MUC2 glycoprotein
specific monoclonal antibody (MAb 996) using an
RIA inhibition assay and characterised by IC50
values. At the same time we have investigated the
secondary structure of the compounds by CD and
Fourier-transform infrared spectroscopy in solution. Our data showed that the presence of amino acid residue(s) at position(s) 16P, 16PT17 or
16
PTP18 resulted in gradually decreasing antibody
binding, but the replacement of the -Thr at position
19 almost abolished the activity (Table 1). Parallel
with this reduction, changes in the conformer population have been detected. The propensity of the
pTPTGTQ peptide to adopt a folded, most probably
-turn, structure in water can be correlated with its
essentially preserved antibody recognition. After
Immunoreactivity of modified peptide epitopes
Effect of non-native flanking regions in Mycobacterium
tuberculosis epitope peptides on 38 kDa glycoproteion
specific T-cell recognition. T-cells play a critical role
in the development of protective and pathogenic
immune responses against M. tuberculosis infection.
Identification of immunodominant epitopes could
provide a rational basis for the construction of
synthetic peptide-based diagnostic reagents. It
is known that the T-cell stimulatory activity of
naturally processed proteins or synthetic peptides
depends not only on the sequence of the T-cell
epitope core, but also on its flanking residues.
The 38 kDa glycosylated lipoprotein, a secreted
constituent of M. tuberculosis, is immunogenic in
active tuberculosis. A peptide representing the
65–83 (FNLWGPAFHERYPNVTITA) region was
found to be one of the immunodominant T-cell
stimulatory domains in humans and in 57CBL/10
mice pre-sensitised with either live or killed organisms of the M. tuberculosis complex or with recombinant 38 kDa. The murine CD4 + T cell epitope core
of this region was localised to amino acid residues
75–81 (RYPNVTI) by deduction from PEPSCAN
analysis using overlapping 15-mer peptides.
To clarify the role of flanking regions adjacent to
the epitope core a peptide representing the deduced
core was prepared with extensions at both N- and
C-termini.8 These flanks are composed of either
amino acid residues from the native sequence and
terminated by Ala and/or Ser residues or oligopeptides of Ala or Ser exclusively. Their binding to
isolated H-2-Ab MHC glycoprotein as well as their
T-cell stimulatory capacity were assayed using a
specific murine hybridoma T cell line [38.H6], lymph
node cells from the native 20-mer peptide primed
57
CBL/10 mice, and human PBMCs from sensitised
individuals.9
The stimulation of the 65–83 specific T-cell
hybridoma was induced by the full-length FNLWG
PAFHERYPNVTITA peptide, but not by the
1200
1000
800
S.I.
further substitution the peptide still contained and/or -turn folded secondary structural elements,
but in a significantly smaller conformer population
and built up from -residues.
These data suggest that for significant MAb 996
recognition, the N-terminal flanking region next to
the core epitope could contain at least one -amino
acid without significant loss of binding activity and
folded conformation required. These findings might
be useful for the design of artificial epitopes,
peptide-vaccines with increased enzymatic stability
and extended biological half-life.
201
600
400
65–83
AASA(75–81)AAAA
AAAA(75–81)AAAA
AAA(74–82)AAA
74–81
200
0
0.1
1
10
100
c (µM)
Figure 5. The e#ect of non-native flanking regions on the
T-cell recognition of peptides containing the deduced
core. 65FNLWGPAFHERYPNVTITA83 specific hybridoma cells [38.H6] were incubated with peptides at
0·210 6 to 5010 6  in the presence of irradiated
57
CBL/10 spleen cells. Culture supernatants were tested
with the cell line HT2. The results are expressed as
stimulation indices (SI=cpm with peptide/cpm without
peptide). The average background value was 555·4 cpm.
eight-mer peptide 74ERYPNVTI81, being only one
amino acid longer at the N-terminus than the
deduced murine T-cell epitope core (Fig. 5). Elongation of the deduced core by four Ala residues at both
N- and C-termini in AAAARYPNVTIAAAA (A4-7581-A4) induced significant stimulation throughout
the concentration range studied. Peptide AAAERYPNVTITAAA (A3-74-82-A3), having an extended
core from the native protein, stimulated somewhat
less than that of peptide 65–83 (Fig. 5).
Interestingly a substitution of one Ala to Ser in
the N-terminal flank increased the T cell stimulatory capability significantly above that stimulated
by the native 20-mer (65–83). Peptide AASARYPNVTIAAAA (A2SA-75-81-A4) containing only the
deduced core was more potent than its counterpart
(Fig. 5). It should be noted that this construct
was even more active than peptide 65–83, which
contains the whole immunodominant domain.
To test the in vivo immunogenicity of peptides
65–83 and A2SA-75-81-A4, 57CBL/10 mice were immunised and LN cells were then stimulated in vitro
by the homologous and heterologous peptides.
Although both peptides were immunogenic in vivo,
immunisation with peptide A2SA-75-81-A4 resulted
in stronger immune responses when LN cells were
challenged in vitro with either the native 65–83
202
F. Hudecz
capacity of the deduced, and non-functional epitope
core.
100
Conjugation to branched chain polypeptide
Inhibition %
80
60
40
CLIP
AAAA(75–81)AAAA
AASA(75–81)AAAA
AA(73–83)AA
65–83
20
0
0.1
1
10
100
c (µM)
Figure 6. The e#ect of non-native flanking regions on the
binding to the H-2-Ab molecule. Inhibition of binding
of the biotinylated CLIP peptide (0·7 ) to H-2-Ab by
65
FNLWGPAFHERYPNVTITA83 and its analogues. Inhibition with the unlabelled CLIP peptide is shown as a
reference. The values represent the means of triplicate
wells from a single experiment, which was repeated three
times. Variations were <20% of the mean.
peptide or its non-native flanked analogue, A2SA-7581-A4. These results demonstrate that peptide A2SA75-81-A4 is not only more e#iciently recognised by T
cells in vitro, but it also has improved in vivo
immungenicity compared with the native 20-mer
peptide 65FNLWGPAFHERYPNVTITA83.
The peptides were tested for binding to isolated
H-2-Ab glycoproteins using VSKMRMATPLLM
QALP (CLIP) as a reference peptide. As shown in
Figure 6, peptide 65–83 binds to H-2-Ab with somewhat lower a#inity than CLIP. Replacement of
native flank region with alanine as in the A4-7581-A4 peptide resulted in a relatively small decrease
in binding. The peptide with Ser introduced into the
N-terminal flank (A2SA-75-81-A4) showed binding
similar to that of the native peptide (IC50 =0·9 ).
Therefore, when normalised for the experimental
conditions, the a#inity (Ki) of the native 65–83
peptide is 0·158 , whereas a#inities of peptides
with artificial flanks are in the range 0·07–0·29 .
Taken together, this is the first study in the
literature in which a synthetic peptide constructed
by elongation of a non-functioning deduced epitope
core with short, simple flanking segments proved to
be a stronger immunogen than the peptide which
contains the natural adjacent amino acid residues.
These findings suggest that non-native flanking
regions are able to enhance the T-cell stimulatory
New groups of branched chain polymeric polypeptides were developed in our laboratory with
the general formula poly[Lys(Xi--Alam)] (XAK),
poly[Lys(Xi--Serm) (XSK), or poly[Lys(-Alam-Xi)]
(AXK), where i<1, m3, and X represent an
additional optically active amino acid residue.10–12
These compounds are poly[-Lys] derivatives substituted by short (three to six amino acid residues)
branches at the -amino groups. The side-chains are
composed of about three -Ala (Fig. 7) or -Ser
residues and one other amino acid residue (X) at the
N-terminus of the branches (XAK or XSK) or next to
the poly[-Lys] backbone (AXK). These compounds
were characterised by their size, chemical structure
(primary structure, solution conformation) and
their biological properties (in vitro cytotoxicity,
pyrogenicity, biodegradation, immunoreactivity
and biodistribution). Under physiological conditions (pH 7·3 in 0·15  NaCl) depending on the
identity of amino acid X, branched polypeptides
could possess polycationic (e.g. poly[Lys(Leui-Alam)], (LAK) or poly[Lys(Orni--Alam)], (OAK)),
amphoteric (e.g. poly[Lys(Glui--Alam)], (EAK)
or polyanionic (e.g. poly[Lys(Ac-Glui--Alam)],
(Ac-EAK) or poly[Lys(Suc-Glui--Alam)]) (SucEAK) character. We have demonstrated that the
composition of the side chains and the charge
properties of these polypeptides determine their
solution conformation, phospholipid membrane
interaction and various biological e#ects (e.g.
cytotoxicity, blood clearance, tissue distribution,
immunreactivity).
These macromolecules were used for the synthesis
of B-cell epitope peptide conjugates to be used as
target antigens for the specific and sensitive detection of MUC1 glycoprotein-specific antibodies.13
More recently this class of polymeric polypeptides
was conjugated with several epitope peptides
derived from glycoprotein D of HSV or from M.
tuberculosis. We have shown that the composition of
the polymeric component has a marked influence on
the immuno-recognition of covalently attached
epitope peptide, and also on the interaction
between phospholipid mono- or bilayers and epitope
conjugates.14
Carrier-dependent induction of HSV gD epitopespecific, protective immune response. To investigate
the roleof a macromolecular carrier in inducing an
Immunoreactivity of modified peptide epitopes
+
NH3
+
NH3
- CH3
Leu
CH2-CH-(CH3)2
- CH3
Ser/Thr
+
NH3
203
- CH3
Orn/Lys
- CH3
+
NH3
+
NH3
OH
+
NH3
+
- CH3
- CH3
- CH3
poly[Lys(Leui-DL-Alam)]
LAK
- CH3
NH3
- CH3
poly[Lys(Xi-DL-Alam)]
SAK (X = Ser), TAK (X = Thr)
poly[Lys(Xi-DL-Alam)]
KAK (X = Lys), OAK (X = Orn)
+
NH3
+
NH3
- CH3
oligo(DL-Ala)
B
X
- CH3
- CH3
poly[L-Lys]
–
+
+
NH3
CO-(CH2)2-COO
poly[Lys(Xi-DL-Alam)]
XAK
Glu
+
NH3
- CH3
NH3
–
COO
- CH3
- CH3
- CH3
- CH3
+
NH3
NH-CO-CH3
–
COO
Suc-Glu
- CH3
–
- CH3
poly[Lys(Glui-DL-Alam)]
EAK
NH
COO
- CH3
Ac-Glu
- CH3
poly[Lys(SucGlui-DL-Alam)]
Suc-EAK
poly[Lys(AcGlui-DL-Alam)]
Ac-EAK
Figure 7. Schematic representation of the branched chain polymeric polypeptides.
epitope-specific, protective immune response
against viral infection, artificial antigens with
branched polypeptide carrier and B-cell epitopes
have been designed (Fig. 8). Peptides corresponding
to two di#erent epitope regions (KYALADASLKMADPNRFRG KDLP, 1–23 and SALLEDPVG-NH2,
276–284) of glycoprotein D of HSV type 1 were
conjugated with two representatives of branched
polypeptides (AK and LAK) and KLH.13
Under the conditions used for coupling the sidechains of Asp and Glu could be involved in amide
bond formation, but only one carboxyl group of the
gD-1 peptides was linked to the -amino group of
the terminal alanine (AK) or leucine (LAK) residue.
204
F. Hudecz
oligo(DL-Ala)
poly[L-Lys]
276
SALLE...284
1
K....P23
276
Figure 8. Schematic representation of the branched polypeptide conjugate of
SALLEDPVG284-NH2]-AK and
KYALADASLKMADPNRFRGKDLP23]-AK containing epitope sequence from glycoprotein D of herpes simplex virus.
1
CBA or BALB/c mice were immunised with conjugates and relevant controls in CFA. The magnitude
of the peptide-, conjugate- and carrier-specific Ab
responses were then analysed and the survival of
animals infected with a lethal dose of 50 HSV-1 was
investigated.
AK conjugates ([1–23]-AK and [276–284]-AK)
induced significant and comparable gD-epitopespecific IgG responses accompanied by the appearance of a low level of carrier-specific antibodies. In
contrast, negligible epitope-specific IgG responses
were elicited with [1–23]-LAK or [276–284]-LAK.
Immunisation with the respective peptide-KLH conjugates induced intense carrier-specific response
without measurable peptide specificity.
In a protection experiment (Fig. 9) carried out in
BALB/c mice, repeated administration of LAKbased conjugates ([276–284]-LAK or [1–23]-LAK)
were not able to prolong survival significantly compared to the free peptide or untreated control mice.
In sharp contrast, pre-immunisation with the [1–23]AK or [276–284]-AK conjugate resulted in complete
protection of a considerable proportion (50%) of
animals against a 100-fold lethal dose of HSV-1.
These results demonstrated that it is feasible to
construct synthetic immunogens with synthetic
branched polypeptide carriers which possess
higher e#icacy to induce epitope-specific antibody responses than KLH-based conjugates.
Comparative studies with branched polypeptide
carriers and two epitope peptides from gD suggest
that the capability of eliciting protective immune
response against HSV-1 infection is dependent on
the carrier molecule.
Enhancement of the T-cell response to a mycobacterial
peptide by conjugation to a synthetic branched polypeptide. T-cell epitopes containing peptides covalently
attached to macromolecular carriers can be considered as synthetic immunogens for the development
of skin-test diagnostics and of vaccines. To investigate the role of the carrier on the recognition
of an attached T-cell epitope, we have prepared
conjugates containing peptide 350DQVHFQPLPP
AVVKLSDALI369 representing a T-cell epitope
domain of 38 kDa protein of M. tuberculosis and
branched chain polypeptides EAK, Ac-EAK or SucEAK (Fig. 7). In vitro T-cell immunogenicity of these
three conjugates together with relevant free peptide, free branched polypeptide as well as their
mixture was studied using human peripheral blood
mononuclear cell (PBMC) cultures from healthy
subjects and from tuberculosis patients.15 We
found that interferon gamma production as well as
T-cell proliferation increase was dependent on
the carrier compound. A conjugate containing
Suc-EAK enhanced IFN- production more than
13-fold (from 22·6 to 294 pg/ml, P=0·001) in PBMCs
from healthy individuals, and 8·7-fold (P=0·012) in
cells from tuberculosis patients. In a proliferation
assay the stimulation index was elevated, when
Ac-EAK or Suc-EAK was present in the conjugate
(Fig. 10). These data clearly suggest that the selection of the carrier (Suc-EAK>Ac-EAK) greatly
Immunoreactivity of modified peptide epitopes
205
20
(A)
100
15
60
[1–23]-AK
40
S.I.
Survival %
80
1–23
20
control
2
0
10
[1–23]-LAK
4
6
8
10
12
14
Time after HSV-1 infection (days)
30
5
(B)
100
Survival %
80
0
0.1
60
[276–284]-AK
40
273–284
20
control
0
2
[276–284]-LAK
4
6
8
10
12
14
Time after HSV-1 infection (days)
30
Figure 9. Survival of HSV-1 infected BALB/c mice preimmunised with gD-1 epitope peptide-branched polypeptide conjugates showing the e#ect of the carrier on the
protection induced by [1–23]-AK, [1–23]-LAK conjugates
or [1–23] peptide (A) or [276–284]-AK, [276–284]-LAK
conjugates or [273–284] peptide (B). Control mice were
injected with PBS.
influences the T-cell
immune response.
epitope
peptide-specific
Stimulation of T-cell responses with a conjugate containing two distinct epitopes attached covalently to a
branched polypeptide carrier. Recently a fully syn-
thetic prototype conjugate with two independent
T-cell peptide epitopes of M. tuberculosis proteins
(16 kDa and 38 kDa) was produced (Fig. 11). As a
carrier, an amphoteric branched chain polypeptide,
EAK with poly[-lysine] backbone, has been used.
This polypeptide with free -amino and -carboxyl
groups at the end of the side chains was conjugated
with peptides representing two immunodominant
1
10
–6
Concentration × 10
100
Figure 10. E#ect of a polypeptide carrier on the reaction
of human peripheral blood mononuclear cells to peptide
350
DQVHFQPLPPAVVKLSDALI369 from 38 kDa protein
of Mycobacterium tuberculosis and its conjugates with
Ac-EAK or Suc-EAK polyanionic branched polypeptide
carrier (average of four experiments performed with four
donors). The results are expressed as stimulation indices
(SI=cpm with peptide/cpm without peptide). Positive
response: SI>3.
regions of the 16 kDa and 38 kDa proteins of M.
tuberculosis, respectively.16 Peptide 91CSEFAYG
SFVRTVSLPVGADE110 was elongated by Cys at the
N-terminus and attached to the carrier containing
protected SH groups to form disulphide bridges.
Peptide 65FNLWGPAFHERYPNVTITA83 was conjugated to the 3-(2-pyridyldithio)propionic acid
N-hydroxy-succinimide ester (SPDP) modified and
acetylated EAK by introducing an amide bond
between the free -amino group of the peptide and
the free -COOH group of Glu at the terminal
position of the branches. This strategy lead to
chemically well defined synthetic immunogens
that contain two di#erent epitopes in multiple
copies covalently linked to a synthetic branched
polypeptide carrier. In this conjugate peptide
91
SEFAYGSFVRTVSLPVGADE110 from the 16 kDa
protein and peptide 65FNLWGPAFHERYPNV
TITA83 from the 38 kDa protein were conjugated.
In vitro T-cell immunogenicity of a prototype
conjugate together with relevant Ac-EAK-based
206
F. Hudecz
NH-
-S-S
H2-NCO-
CH3-CO-
poly[L-Lys]
40
(A)
5 × 10–6 M
25 × 10–6 M
Glu
NH- oligo(DL-Ala)
-CONH2
S.I.
CH3-CO-
·S-S-
H2NCO-
CH3-CO-
H2-Peptide1
20
150
AcEAK
(93–110)
(65–83)
(B)
5 × 10–6 M
25 × 10–6 M
100
S.I.
AcEAK
(65–83)
(91–110)
0
(91–110)-AcEAK
50
K2
conjugates was studied using T-cell hybridomas,
lymph node cells from immunised mice and human
PBMC cultures from PPD-positive individuals.
Conjugate K2, the free peptides (91SEFAYG
SFVRTVSLPVGADE110 and 65FNLWGPAFHERY
PNVTITA83), their Ac-EAK conjugates and unsubstituted branched polypeptides (Ac-EAK and
EAK) were tested using the murine T-cell hybridomas 38.I.6 and 16.10.7 (Fig. 12). The concentration
of conjugates was calculated according to the
respective peptide contents.
The 65–83 specific T-cell hybridoma was recognised by conjugates K2 and 38.I-AcEAK as well as
the free 65FNLWGPAFHERYPNVTITA83 peptide;
as expected, peptide 91SEFAYGSFVRTVSLPVG
ADE110 corresponding to the 16 kDa protein did not
induce a response (Fig. 12A). The activity of peptide
91
SEFAYGSFVRTVSLPVGADE110 from the 16 kDa
protein was investigated using T-cell hybridomas
obtained from peptide 91–110 immunised mice. Peptide 91SEFAYGSFVRTVSLPVGADE110 conjugated
either to EAK through a disulphide bond (in conjugate K2) or to Ac-EAK with an amide bond was able
to induce T-cell stimulation (Fig. 10B). Specificity in
both sets of experiments has been proved by the lack
of e#ect on these cells of free 65–83 (Fig. 12B) or
91–110 (Fig. 10A) peptides or branched polypeptides
(Ac-EAK or EAK).
An alternative strategy for disulphide-based
conjugation is the application of Cys(Npys) as
a coupling agent containing Npys (3-nitro-2pyridinesulphenyl) group in the side-chain of
cysteine. Cys(Npys) residues were used to modify
branched polypeptides at the end of their branches
and HS-peptide epitopes.17
0
(65–83)-AcEAK
Figure 11. Schematic representation of conjugate K2
containing two di#erent epitope peptides from M.
tuberculosis proteins 65FNLWGPAFHERYPNVTITA83
and 91CSEFAYGSFVRTVSLPV-GADE110 attached to
branched polypeptide EAK by an amide or disulphide
bond, respectively.
K2
Cys-Peptide2
Figure 12. T-cell recognition of conjugate K2 by hybridoma [38.H6/1] specific for 65FNLWGPAFHERYPNVTI
TA83 (A) and by hybridoma [16.10.7] specific for
91
SEFAYGSFVRTVSLPV-GADE110. Cells were incubated
with conjugate K2 and control compounds at 510 6
and 2510 6  in the presence of irradiated 57CBL/10
spleen cells. Culture supernatants were tested with the
cell line HT2. Results are expressed as stimulation indices
(SI=cpm with peptide/cpm without peptide).
These findings suggest that the conjugate K2
containing two distinct T-cell epitopes are able to
induce epitope specific responses in hybridoma
cells. Thus, the capability of epitope peptides to
initiate specific T cell stimulation was preserved
Immunoreactivity of modified peptide epitopes
after their well-defined covalent attachment to a
synthetic polymeric polypeptide.
Acknowledgements
Experimental work summarised in this paper
was supported by grants from the from WHO
(T9/181/133), Hungarian-Spanish Intergovernmental
Programme (5/1998,3/2001), from the Hungarian
Research Fund (OTKA No. T-3024, T-4217, T-014964,
T-03838) and from the Hungarian Ministry of
Education (FKFP No. 0101/97).
9.
10.
11.
References
1. Mezö G, Majer Zs, Valero ML, Andreu D, Hudecz F.
Synthesis of cyclic herpes simplex virus peptides containing 281–284 epitope of glycoprotein D–1 in endo or
exo-position. J Peptide Sci 1999; 5: 272–282.
2. Mezö G, Majer Zs, Jimenez MA, Vass E, Carreño C,
Andreu D, Hudecz F. Conformational studies of
linear and cyclic HSV epitope peptides. In: Bajusz S,
Hudecz F (eds) Peptides 1998. Proc. 25th European
Peptide Symposium. Budapest, Akadémiai Kiadó,
1999: pp. 506–507.
3. Mezö G, Drakopoulou E, Paál V, Rajnavölgyi E
u , Vita
C, Hudecz F. Synthesis and immunological studies
of a-conotoxin chimaera containing an immunodominant epitope from the 268–284 region of HSV gD
protein. J Peptide Res 2000; 55: 7–17.
4. Drakopoulou E, Uray K, Mezö G, Price MR, Vita C,
Hudecz F. Synthesis and antibody recognition of
mucin 1–-conotoxin chimaera. J Peptide Sci 2000; 6:
175–185.
5. Uray K, Price MR, Hudecz F. Localisation of a protein core specific epitope from gastrointestinal mucin
(MUC2). The e#ect of epitope immobilisation on
antibody recognition. J Peptide Sci 1998; 4: 319–326.
6. Uray K, Kajtár J, Vass E, Price MR, Hollósi M,
Hudecz F. E#ect of D-amino acid substitution in a
Mucin 2 epitope on mucin-specific monoclonal antibody recognition. Archives of Biochemistry and
Biophysics 2000; 378: 25–32.
7. Uray K, Kajtár J, Vass E, Price MR, Hollósi M,
Hudecz F. E#ect of solution conformation on antibody
recognition of a protein core epitope from gastrointestinal mucin (MUC2). Archives of Biochemistry and
Biophysics 1999; 361: 65–74.
8. Bogdán K, Vordermeier HM, Kajtár J, Mák M, Ivanyi
J, Hudecz F. Influence of peptide length, nature of
12.
13.
14.
15.
16.
17.
207
flanking sequences and fidelity of critical residues on
T cell recognition of an immunodominant epitope
from the M.tuberculosis 38 kDa protein. In: Schneider
CH (ed.) Peptides in Immunology. London, J. Wiley,
1996: pp. 85–92.
Wilkinson KA, Vordermeier MH, Kajtár J, Jurcevič
S, Wilkinson R, Ivanyi J, Hudecz F. Modulation of
peptide specific T cell responses by non-native
flanking regions. Mol Immunol 1997; 34: 1237–1246.
Hudecz F. Design of synthetic branched-chain
polypeptides as carriers for bioactive molecules.
Anti-Cancer Drugs 1995; 6: 171–193.
Mezö G, Kajtár J, Nagy I, Szekerke M, Hudecz F.
Carrier design: Synthesis and conformational studies
of poly[L-lysine] based branched polypeptides with
hydroxyl groups. Biopolymers 1997; 42: 719–730.
Hudecz F, Pimm MV, Rajnavölgyi E
u , Mezö G, Fabra
A, Gaál D, Kovács AL, Horváth A, Szekerke M.
Carrier design: New generation of polycationic
branched polypeptides containing OH groups with
prolonged blood survival and diminished in vitro
cytotoxicity. Bioconjugate Chemistry 1999; 10: 781–
790.
Hudecz F. Alteration of immunogenicity and antibody
recognition of B-cell epitopes by synthetic branched
polypeptide carriers with poly[L-lysine], backbone.
Biomedical Peptides, Proteins and Nucleic Acids
1995; 1: 213-220.
Hudecz F, Nagy IB, Kóczán Gy, Alsina MA, Reig F.
Carrier design: influence of charge on interaction of
branched polymeric polypeptides with phospholipid
model membranes. In: Chiellini E, Sunamoto J,
Migliaresi C, Ottenbrite RM, Cohn D (eds) Biomedical
Polymers and Polymer Therapeutics. New York,
Kluwer Academic/Plenum Publishers, 2001: pp. 103–
120.
Wilkinson KA, Hudecz F, Vordermeier HM, Ivanyi J,
Wilkinson RJ. Enhancement of the T cell response to
a mycobacterial peptide by conjugation to synthetic
branched polypeptide. Eur J Immunol 1999; 29: 2788–
2796.
Wilkinson KA, Vordermeier MH, Wilkinson R,
Ivanyi J, Hudecz F. Synthesis and in vitro T cell
immunogenicity of conjugates with dual specificities:
attachment of epitope peptides of 16 kDa and
38 kDa proteins from M.tuberculosis to branched
polypeptide. Bioconjugate Chemistry, 1998; 9: 539–547.
Mezö G, Mihala N, Andreu D, Hudecz F. Conjugation
of epitope peptides to branched chain polypeptides via
Cys(Npys). Bioconjugate Chemistry 2000; 11: 484–491.