|
Originally published In Press as doi:10.1074/jbc.M201076200 on March 21, 2002
J. Biol. Chem., Vol. 277, Issue 25, 22656-22661, June 21, 2002
The Major Conformational IgE-binding Epitopes of
Hevein (Hev b6.02) Are Identified by a Novel Chimera-based Allergen
Epitope Mapping Strategy*
Piia
Karisola ,
Harri
Alenius§,
Jari
Mikkola§,
Nisse
Kalkkinen¶,
Jari
Helin¶,
Olli T.
Pentikäinen ,
Susanna
Repo,
Timo
Reunala**,
Kristiina
Turjanmaa**,
Mark S.
Johnson,
Timo
Palosuo, and
Markku S.
Kulomaa§§
From the Department of Biological and Environmental
Science, University of Jyväskylä, P. O. Box 35 (YAB),
FIN-40014 University of Jyv skyl, Finland,
§ Finnish Institute of Occupational Health, Topeliuksenkatu
42 aA, FIN-00250 Helsinki, Finland, ¶ Institute of
Biotechnology, University of Helsinki, P. O. Box 56, FIN-00014
University of Helsinki, Helsinki, Finland, Department of
Biochemistry and Pharmacy, Åbo Akademi University, P. O. Box 66, FIN-20521 Turku, Finland, ** University Hospital of Tampere,
P. O. Box 2000, FIN-33521 Tampere, Finland, and
National Public Health Institute,
Mannerheimintie 166, FIN-00300 Helsinki, Finland
Received for publication, February 1, 2002, and in revised form, March 18, 2002
 |
ABSTRACT |
A novel approach to localize and reconstruct
conformational IgE-binding epitope regions of hevein (Hev b6.02), a
major natural rubber latex allergen, is described. An antimicrobial
protein (AMP) from the amaranth Amaranthus caudatus was
used as an immunologically non-IgE-binding adaptor molecule to which
terminal or central parts of hevein were fused. Hevein and AMP share a
structurally identical core region but have different N-terminal and
C-terminal regions. Only 1 of 16 hevein-allergic patients showed weak
IgE binding to purified native or recombinant AMP. Chimeric AMP with the hevein N terminus was recognized by IgE from 14 (88%) patients, and chimeric AMP with the hevein C terminus was recognized by IgE from
6 (38%) patients. In contrast, chimeric AMP containing the hevein core
region was recognized by IgE from only two patients. When both
the N-terminal and C-terminal regions of hevein were fused with the AMP
core, IgE from all 16 patients bound to the chimera. This chimera was
also able to significantly inhibit (>70%) IgE binding to the native
hevein. On the contrary, linear synthetic peptides corresponding to
hevein regions in the AMP chimeras showed no significant IgE binding
capacity in either enzyme-linked immunosorbent assay or inhibition
enzyme-linked immunosorbent assay. These results suggest that the IgE
binding ability of hevein is essentially determined by its N-terminal
and C-terminal regions and that major IgE-binding epitopes of hevein
are conformational. The chimera-based epitope mapping strategy
described here provides a valuable tool for defining structural
epitopes and creating specific reagents for allergen immunotherapy.
| |
INTRODUCTION |
To help understand the molecular basis of Type I allergic immune
responses, identification of conformational IgE-binding epitopes on the
surface of allergen molecules is required. This knowledge is crucial
when designing tools and strategies for allergen-specific immunotherapy or vaccination.
Natural rubber latex (NRL)1
allergy has become increasingly common during the last decade,
and a large proportion of the general population is regularly exposed
to NRL. In particular, health care workers (up to 17%) and children
with a history of multiple surgeries (up to 60%) are reported to be
sensitized to NRL products (1, 2). Although several NRL allergens have
recently been characterized at the molecular level, consensus exists
that hevein (Hev b6.02) is one of the most important NRL allergens (3, 4). Hevein (4.7 kDa, 43 amino acids) is the most predominant protein in
liquid latex obtained from the rubber tree Hevea
brasiliensis. Hevein is involved in latex coagulation, and it may
play a key role as a defense protein in the protection of wound sites
by inhibiting the growth of chitin-containing fungi. Hevein has been found to elute in large quantities from NRL products (5). The importance of hevein as a latex allergen is supported by the findings that >70% of latex-allergic patients have hevein IgE antibodies (6)
and that hevein elicits positive skin prick test reactions in the great
majority of latex-allergic patients (1). Hevein is therefore a useful
tool in the diagnosis of latex allergy as well as a potential candidate
for immunotherapy.
Treatment options for allergic diseases are currently based mainly on
allergen avoidance and the use of corticosteroids and antihistamines to
control inflammation and asthma. Whereas allergen-specific immunotherapy using authentic IgE-binding allergens is effective in
controlling some allergies, the high risk of systemic reactions during
therapy remains a serious problem. Recently, attempts have been made to
eliminate IgE-binding sites of allergens while retaining their
T-cell-stimulating sites (7, 8). In order to design a safe reagent for
allergen-specific immunotherapy, it is crucial to characterize in
detail the IgE-binding epitopes of the allergen. To date, in most
studies, identification of IgE-binding epitopes has been performed
using short synthetic overlapping peptides that cover the whole
allergen sequence. Although this approach can be very helpful in
identifying linear T-cell epitopes, its value is limited in B-cell
epitope mapping. This is due to the fact that most, if not all, IgE
epitopes are reported to be conformational and thus cannot properly be
studied using linear peptides (9, 10). An inherent requirement for
studying conformational epitopes is knowledge of the three-dimensional
structure of the allergen.
In the present study, we describe a novel allergen epitope mapping
strategy for the characterization of conformational IgE-binding epitopes of hevein. By combining protein engineering and modeling-based design, we were able to transfer hevein-specific IgE binding activity to a nonallergenic adaptor protein. Finally, by using this approach, N-terminal and C-terminal regions of hevein were shown to contain the
major conformational IgE-binding epitopes of hevein.
| |
EXPERIMENTAL PROCEDURES |
Sequences, Structures, and Computational
Analyses--
Hevein-like domains in the Swiss-Prot Protein Sequence
Database were identified by using the advanced BLAST search
service provided by the National Center for Biotechnology Information. On the basis of the multiple sequence alignment by Clustal X (11), the
structures of hevein-like domains were also obtained from the National
Center for Biotechnology Information service and used for additional
studies. The three-dimensional structures of hevein from the rubber
tree H. brasiliensis (hevein, Protein Data Bank access code
1HEV, Ref. 12) and an antimicrobial protein (AMP) from the amaranth
Amaranthus caudatus (Protein Data Bank access code 1MMC,
Ref. 13), both determined using NMR, were obtained from the Protein
Data Bank (14). The most representative NMR structure in the ensemble
was selected using criteria from WHATCHECK (15).
Structural Modeling of AMP Chimera--
The structural
alignment of hevein and AMP was performed using VERTAA (16) in the
BODIL modeling package
(www.abo.fi/fak/mnf/bkf/research/johnson/bodil.html).2
VERTAA produces rapid automated structural comparisons based on C-
atoms. MODELLER 4.0 (17) was used to construct three-dimensional model
structures of the four chimeras. All four chimera models were
energy-minimized using the GROMOS96 43a1 force field and deepest
descent method in GROMACS2.0 (18). After minimization, the structures
were further refined with 400 ps of molecular dynamics simulation to
obtain better packing of the chimeras and to identify any gross
problems in the folded model structures. Structure representations were
prepared with InsightII 98.0 software (Molecular Simulations Inc., San
Diego, CA) and RASTER3D (19).
Synthetic Peptides--
Peptides spanning the entire hevein
protein were purchased from MedProbe (Oslo, Norway). They include three
peptides of 11 (linN), 21 (lincore), and 11 (linC) amino acids
in length, corresponding to the hevein N terminus EQCGRQAGGKL (residues
1-11), the hevein core CPNNLCCSQWGWCGSTDEYCS (residues 12-32), or the
hevein C terminus PDHNCQSNCKD (residues 33-43), and the corresponding
peptides in which each cysteine residue was replaced with a serine.
Production of Recombinant Proteins--
The recombinant form of
hevein, AMP, and AMP chimeras were produced as chicken avidin (Av)
fusion proteins for efficient purification from the insect cells using
a baculovirus expression system (Bac-To-BacTM Baculovirus
Expression System; Invitrogen) as described previously (20). To
introduce a cleavage site for thrombin and the selected region of
hevein (Table I), the PCR primers were
designed and checked with the Amplify2 program. Primers were purchased
from TAGC (TAG Copenhagen A/S, Copenhagen, Denmark).
View this table:
[in this window]
[in a new window]
|
Table I
Primers used
Forward primers are indicated by (F), and reverse primers are indicated
by (R). Nucleotides coding hevein residues are shown in lowercase
letters, restriction enzyme sites (BamHI,
HindIII, EcoRI, and SalI) are
underlined, and thrombin cleavage sites are in italic.
|
|
Recombinant AMP and AMPcore were constructed by PCR by multiplying the
two annealed oligonucleotides (AMP1/EcoRI and
AMP2/SalI, AMPcoreI and AMPcoreII) in the absence of a
template. Recombinant AMP was used as a template for the AMPN and AMPC
constructs. Recombinant hevein was amplified from hevein cDNA in
the PIIProhev vector with the forward Prohev6 and reverse Prohev2
primers. All PCR reactions were conducted in a volume of 50 µl
containing the template (10 ng), deoxynucleotide triphosphates (200 µM each), the primers (100 nmol each) (Table I), 10×
Pfu polymerase buffer, and 2 units of Pfu DNA
polymerase. The following cycling conditions were used in all cases:
initial denaturation at 94 °C for 5 min following 30 cycles (1 min
at 94 °C, 2 min at 50 °C, and 2 min at 72 °C) and final
extension at 72 °C for 10 min. After the PCR amplification, the
products were digested with BamHI and HindIII
restriction enzymes (Promega, Madison, WI). Digests were extracted from
1.5% agarose gel, ligated into the
BamHI/HindIII-treated pBacAVs+C vector (20), and
transformed into JM109 Escherichia coli cells. The
recombinant vectors were named AvHev, AvAMP, AvAMPN, AvAMPC, AvAMPN+C, AvAMPcore.
The nucleotide sequence of the constructs was always confirmed by
dideoxynucleotide sequencing with an automated Li-Cor DNA sequencer
(Li-Cor Biotechnology Division, Lincoln, NE). The vectors were
transformed into E. coli strain DHC10Bac to construct the recombinant baculo bacmids. After blue/white selection, positive colonies were grown in 5 ml of Luria-Bertani/gentamycin + kanamycin sulfate for bacmid isolation. The viruses were completed during transfection according to the manufacturer's instructions for the
Bac-To-BacTM Baculovirus Expression System (Invitrogen).
To produce recombinant proteins, about 4 × 108
Sf9 cells (ATCC CRL 1711) in SF-900 II SFM serum-free culture
medium (Invitrogen) depleted of biotin were seeded to a final volume of
200 ml in a 1- or 2-liter Erlenmeyer flask. Recombinant viruses were
added to give a multiplicity of infection of 0.1 plaque-forming
unit/cell. Infection was allowed to continue at 28 °C and 110 rpm
agitation for 72 h.
Purification of Recombinant Proteins--
Purification of fusion
proteins was performed in principle as described previously (20). The
cell pellet was resuspended in 80 ml of HilloI buffer (50 mM Tris-HCl, pH 8, 1% Triton X-100, 2 mM EDTA,
and 150 mM NaCl). During the purification, a 75-µl aliquot was taken from the sonicated total cell lysate (sample T), from
the residual cell pellet after resuspension in 20 ml of the same buffer
and sonication (S), before binding to resin (L1), after
adjusting the pH (L2), and after binding to resin (L3). The intracellular fusion proteins bound on the
2-iminobiotin-agarose (Sigma) were washed once with 20 ml of binding
buffer (50 mM Na2CO3, pH 11, and 1 M NaCl) and twice in the column with 10 ml of the same
buffer. 2-Iminobiotin-agarose was used instead of a biotin resin
because avidin binds to biotin in an irreversible manner (Ka, ~10 15
M1) (21) that prevents the release of the
intact fusion protein from the column. Avidin binds to 2-iminobiotin in
a pH-dependent manner and can thus be released from the
resin by lowering pH (22).
One-ml fractions of constructs were eluted from the column with the
elution buffer (50 mM ammonium acetate, pH 4, and 0.5 M NaCl), and 20-µl aliquots were analyzed in SDS-PAGE
gels. AvHev, AvAMP, and all AvChimer fusion proteins were run on 15%
denaturing SDS-PAGE and analyzed by immunoblot analysis using an avidin
antibody (1:4000) as a primary antibody and goat anti-rabbit antibody
(1:2000) as a secondary antibody.
Depending on the method to be used next, the buffer of the fusion
proteins was changed to fresh 25 or 50 mM Tris buffer (pH 8) with a desalting column HR10/10 (Amersham Biosciences) or Fast Desalting (Amersham Biosciences). Samples were concentrated with an Ultrafree-MC (30,000 NMWL filter unit; Millipore Corp., Bedford, MA)
vacuum evaporator or with a MICROSEP (Pall Filtron Corp.). All of the
fusion proteins were digested with an excess of thrombin (Amersham
Biosciences) overnight at 37 °C.
Characterization of the Recombinant
Proteins--
Thrombin-cleaved recombinant hevein, recombinant AMP,
and chimeras were analyzed by reversed-phase chromatography on a PepRP 5/5 column (Amersham Biosciences) using a linear gradient of
acetonitrile (0-40% in 40 min) in 0.1% trifluoroacetic acid at a
flow rate of 0.7 ml/min. Eluted fractions were monitored at 214 nm. The separated peptides were collected and further analyzed by mass spectrometry and N-terminal amino acid sequencing (ABI 494 A
ProciseTM Sequencer; PerkinElmer Life Sciences).
Matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF)
mass spectrometry was carried out using a Bruker Biflex II
(Bruker-Daltonic, Bremen, Germany) instrument. About 1 pmol (0.5 µl)
of the HPLC-purified peptides was applied to a thin-layer matrix
preparation (0.5 µl of saturated cyano-4-hydroxycinnamic acid in
acetone) and air-dried. External calibration was performed with insulin (Sigma).
Immunological Analyses with Sera from NRL-allergic Patients and
Controls--
Serum specimens were obtained from 16 NRL-allergic
patients (15 women and 1 man; mean age, 46 years; age range, 16-70
years), all of whom were previously shown to have IgE against hevein. All of the patients had positive skin prick test responses to natural
hevein. Sera from 19 control subjects without NRL allergy were used in
the ELISA studies.
IgE ELISA--
ELISA was performed as described previously (6).
Purified recombinant hevein, recombinant AMP, or chimeras were diluted in 50 mmol/liter carbonate buffer, pH 9.6, to a final concentration of
1 µg/ml. The proteins were placed onto a polystyrene microtiter plate
(100 µl/well; Nunc, Roskilde, Denmark) and incubated at room
temperature (20-23 °C) for 3 h or overnight at 4 °C. The wells were emptied, and the remaining protein binding sites were blocked with 1% human serum albumin (HAS; Finnish Red Cross Blood Transfusion Service, Helsinki, Finland) in 50 mM carbonate
buffer, pH 9.6, at room temperature for 1 h. After rinsing three
times, 100 µl of serum (diluted 1:20 with phosphate-buffered saline
containing 0.02% Tween 20 and 0.2% HSA) from a NRL-sensitive
individual or from a negative control person was added to the well and
incubated at room temperature for 2 h. The wells were washed
again, and biotinylated goat anti-human IgE (Vector Laboratories,
Burlingame, CA; diluted 1:1000 with phosphate-buffered saline
containing 0.02% Tween 20 and 0.2% HSA) was added and incubated at
room temperature for 1 h. After washing again,
streptavidin-conjugated alkaline phosphatase (Bio-Rad; diluted 1:3000)
was added and incubated for 1 h. The substrate,
p-nitro-phenyl phosphate (Sigma), was added, and the
absorbance was read at 405 nm by using an automated ELISA reader
(Titertek Multiskan, Eflab, Finland). The mean ± 3 S.D. of the
controls was chosen as the threshold value for a positive result.
Inhibition ELISA--
The inhibition ELISA was performed as
described previously (6), with the following modifications. Hevein or
chimeras (AMPN, AMPC, and AMPN+C) were applied to microtiter plates at
a concentration of 2 µg/ml. The ELISA inhibition was performed with
eight different inhibitors (AMPN, AMPC, AMPN+C, linN, lincore, linC,
pool of linN and linC, and hevein) at four different concentrations
(0.0001, 0.01, 1, and 10 µg/ml for all inhibitors and also 100 µg/ml for the linear peptides). Sera (diluted 1:10) were obtained
from six patients that recognized AMPN and/or AMPC chimeras.
Streptavidin-conjugated alkaline phosphatase (Zymed
Laboratories Inc.) was diluted 1:1000 (with phosphate-buffered
saline containing 0.02% Tween 20 and 0.2% HSA) and used as a
secondary antibody. Substrate-produced color was read at 405 nm using
an automated ELISA reader (Multiskan MS, Labsystems, Helsinki, Finland).
| |
RESULTS |
Hevein and the Amaranth AMP Have a Structurally Identical Core
Region but Different N and C Termini--
Allergen structures involved
in IgE binding can be indirectly analyzed by comparing the IgE binding
abilities of structurally related proteins. To find hevein homologues
that could be used in the analysis of conformational IgE-binding
regions, we first analyzed hevein-related proteins by a sequence
alignment program (Clustal X). The central area of hevein showed the
most extensive conservation between the homologues (data not shown).
Because knowledge of the three-dimensional structure of the allergen is essential for the study of conformational epitopes, we analyzed further
only those hevein homologues whose three-dimensional structure had been
solved (23).
Although hevein-like domains commonly occur in a variety of plant
proteins, AMP (30 amino acids) from the amaranth Amaranthus caudatus was found to be the only monomeric hevein homologue with a known three-dimensional structure and a size relatively close to that
of hevein (hevein, 43 amino acids). In the sequence comparison, the
core regions of AMP (residues 9-28) and hevein (residues 12-31) are
highly conserved (Fig. 1). In contrast,
both the N terminus (residues 1-8) and C terminus (residues 28-30) of
AMP differ markedly from the corresponding regions in hevein (residues
1-12 and 32-43, respectively).
View larger version (7K):
[in this window]
[in a new window]
|
Fig. 1.
Structure-based alignment of amino acid
sequences of AMP and hevein. Black shading indicates
identical residues. Gaps (dots) were introduced for optimal
alignment. The box identifies the core regions.
|
|
Because sequence comparison gives only indirect structural information,
we compared the three-dimensional structures of hevein and AMP. The
Protein Data Bank file contains six alternative NMR structures for
hevein (12). Structure 4 was used for all the comparisons because the
majority of its residues had acceptable conformations according to
WHATCHECK. The same procedure was used to select AMP NMR structure 16 (Protein Data Bank access code 1MMC) (13). The structural alignment of
hevein and AMP was performed using VERTAA in the BODIL modeling
environment (Fig. 2). The superimposed
three-dimensional structures show that the backbone -carbons of
hevein (between residues 12 and 31) and AMP (between residues 19 and
28) occupy very similar positions in space relative to each other (Fig.
2). However, substantial differences in the structures were observed at
the N-terminal and C-terminal ends of the two proteins, which is in
good agreement with the results obtained from the sequence
alignment.
View larger version (35K):
[in this window]
[in a new window]
|
Fig. 2.
Superimposition of the NMR structures of the
C- atoms of hevein (blue) on
AMP (green), showing the similarities of the core
region (red). Disulfide bonds: hevein,
yellow; AMP, gray.
|
|
Native AMP Is Not Recognized by Hevein-specific IgE--
To study
the immunological properties of AMP, it was extracted from amaranth
seeds and purified by affinity chromatography on a chitin column and
reversed-phase HPLC as described previously for hevein (24). The
molecular mass (3184.8 Da) of the purified protein corresponded to that
calculated from the AMP sequence with three disulfide bonds (3183.7 Da). The analyzed N-terminal sequence of 20 residues was identical to
the known AMP sequence (AMP2, Swiss-Prot: P27275).
The ability of purified AMP to bind hevein-specific IgE was analyzed
using sera obtained from NRL-allergic patients with hevein IgE
antibodies. Only 1 of 16 patients and 0 of 19 controls showed IgE
binding to native AMP in ELISA (data not shown).
Design and Molecular Modeling of Hevein-AMP Chimeras--
To test
the IgE binding capacity of the core and terminal regions of hevein,
four hevein-AMP chimeras (AMPN, AMPC, AMPN+C, and AMPcore) were
designed, as shown in Fig. 3. On the
basis of their structural alignment, three-dimensional models of the
chimeras were constructed, energy-minimized, and subjected to molecular dynamics simulation to improve side chain packing and identify bad
contacts. No bad contacts were observed, and residues with poor
geometry were identified with WHATCHECK (WHATIF, Ref. 25) and
corrected. The schematic representations of the model structures and
their surface models are shown in Fig. 3.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 3.
A, schematic presentation of the
designed chimeras. B, modeled structures of the chimeras.
Three-dimensional structures correspond to parts of the chimera derived
from AMP (yellow, gray, and pink) or
hevein (blue, green, and red). The
coloring is the same in both A and B.
|
|
The modeled structures were used to evaluate the viability of the
chimeric constructs. No problems with either packing or side chain
interactions were seen when the four chimeras were compared with hevein
and AMP. Major differences occur in the N-terminal sequences of hevein
and AMP, but these differences were not expected to interfere with the
stable folding or packing of the chimeric constructs. Both proteins
have a cysteine residue within this segment that forms a disulfide bond
with the core region. The first two amino acid residues of AMP are not
present in hevein; these residues do not seem to have any critical role
in the fold. The core region in each of the chimera models is nearly
identical to the structures of hevein and AMP, and moreover, it was
possible to build the disulfide bonds in each model in a way similar to those in the NMR structures of hevein and AMP (Fig. 2). The C-terminal part of hevein contains 10 extra amino acids as compared with AMP (Fig.
1). Hevein has one additional disulfide bond (Cys37 to
Cys41), and the amino acids
Gln38-Ser39-Asn40 contribute an
extra strand to the  -sheet (Fig. 2). In addition to the interactions
mentioned above, the C-terminal part of hevein forms no internal
interactions that would be affected by the chimeric constructs.
Production and Biochemical Characterization of Hevein-AMP
Chimeras--
All constructs (recombinant AMP, recombinant HEV, and
the four chimeras) were then cloned and produced in insect cells using the Bac-To-BacTM Baculovirus Expression System. Chicken
avidin was used as an N-terminal fusion partner for easier affinity
purification of the recombinant proteins. After purification, the
avidin tag was cleaved from the fusion proteins by protease thrombin.
To confirm the correct structure and disulfide bridges of the products,
all hevein-AMP chimeras were analyzed by N-terminal sequencing and MALDI-TOF mass spectroscopy. After thrombin cleavage, all recombinant proteins had the correct N-terminal amino acid sequence. The measured molecular masses of the recombinant proteins corresponded to the calculated ones, suggesting that the correct number of disulfide bridges are present in recombinant hevein and AMP as well as in all of
the hevein-AMP chimeras (data not shown).
The IgE Binding Ability of Hevein Is Determined Almost Exclusively
by Its N-terminal and C-terminal Regions--
To study the
immunological properties of the recombinant allergens, ELISA tests were
performed for recombinant hevein and AMP, for all of the chimeras, and
for the three synthetic linear peptides of hevein. The binding of IgE
to recombinant hevein corresponded to that of native hevein (data not
shown). In agreement with the results with native AMP, only 1 of the 16 sera from hevein-allergic patients showed weak IgE binding to
recombinant AMP (Fig. 4). These results
indicated that the IgE binding ability of recombinant hevein and AMP
was virtually identical to that of their native counterparts.
View larger version (10K):
[in this window]
[in a new window]
|
Fig. 4.
Occurrence of IgE antibodies to AMPN, AMPC,
AMPN+C, and AMPcore by ELISA in sera from NRL patients. It can be
seen that of the 16 patients, 88% showed IgE antibodies to AMPN, 38%
showed IgE antibodies to AMPC, and 100% showed IgE antibodies
to AMPN+C. The AMPcore chimera was recognized in 2 of the 16 NRL
patients. No significant binding of IgE antibodies against any of the
linear peptides corresponding to hevein regions in the AMP chimeras
(linN, lincore, and linC, indicated as open circles) was
found. The cut-off level for a positive response (mean optical
density ± 3 S.D. of the control value) is indicated by the
horizontal black line.
|
|
Two of the 16 patients (13%) with IgE antibodies against hevein showed
IgE binding to the AMP chimera containing hevein core region (AMPcore).
This suggests that only a very few hevein-allergic patients develop IgE
response to epitopes present in the hevein core region (Fig. 4). The
chimeric AMPN with the N terminus of hevein (residues 1-11) was
recognized by 14 of 16 (88%) patients, and the chimeric AMPC with
residues from hevein C terminus (residues 32-43) was recognized by 6 of 16 (38%) patients. However, when both the N-terminal and C-terminal
regions of hevein were fused with AMP (AMPN+C), all 16 patients (100%)
showed IgE binding against the chimera.
The IgE binding profiles at the individual level revealed that in nine
patients, only the N-terminal region of hevein was recognized, and in
one patient, only the C-terminal region of hevein was recognized. In
four patients, IgE bound simultaneously to both the N-terminal and
C-terminal regions. For one patient, the hevein core region was the
principal reaction target, and in another patient, IgE bound weakly to
both the hevein core and C-terminal regions. Interestingly, in one
patient, IgE binding was seen only against the chimera in which both
the N terminus and C terminus were present.
Synthetic linear peptides of hevein fragments corresponding to the
AMPN, AMPC, and AMPcore chimeras were tested in direct ELISA, but none
of them bound IgE from the sera of NRL-allergic patients
(linN, lincore, and linC in Fig.
4).
Inhibition ELISA Confirms the Conformational Nature of the IgE
Epitopes--
Six patients who recognized the N terminus and/or C
terminus of hevein in direct ELISA were used for ELISA inhibition
analysis. The binding of IgE to solid-phase AMPN, AMPC, or AMPN+C was
inhibited with the analogous linear peptides (linN, linC, and
linN+linC, respectively) to compare the IgE binding affinities of the
chimeras and peptides. No significant inhibition of IgE binding to the chimeras was observed with any of the linear peptides, even at a
peptide concentration as high as 100 µg/ml (data not shown). Furthermore, the pool of the peptides linN and linC was unable to
inhibit IgE binding to the solid-phase AMPN+C chimera. In contrast, native hevein completely inhibited the binding of IgE to all chimeras, as expected (data not shown).
To compare IgE binding affinities against native hevein, ELISA
inhibition was also performed with the chimeras and peptides as
inhibitors. Chimeric AMPN+C remarkably inhibited the binding of IgE to
solid-phase hevein in a dose-dependent manner (Fig. 5). The binding of IgE was inhibited
60-97% in five patients (10 µg/ml), but only 20% inhibition was
achieved in one patient. In contrast, AMPC showed only a modest
(~30%) inhibition of IgE binding, and no significant inhibition was
seen with AMPN (Fig. 5). ELISA inhibition was also performed with the
linear hevein peptides (linN, linC, and lincore). No significant
inhibition was seen with any of these three peptides, even at the
highest peptide concentration used (100 µg/ml; data not shown).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 5.
The average inhibition of IgE binding to
solid-phase hevein was detected with AMP chimeras and hevein in sera
from six individual NRL-allergic patients. An average of 74%
inhibition of IgE binding to solid-phase hevein was found with the
AMPN+C chimera ( ) at a concentration of 10 µg/ml. No inhibition
was seen with AMPN ( ), but a modest inhibition was noted with AMPC
( ) at the highest concentration. The inhibition curve with native
hevein ( ) as the inhibitor is shown for comparison (indicated as a
dotted line).
|
|
| |
DISCUSSION |
During the last decade, the IgE-binding epitopes of various
allergens have been described, as a rule, by overlapping peptide mapping. Because this method enables immunological properties to be
studied only in short peptides (usually 8-15 amino acids), it is
obvious that only linear IgE-binding epitopes can be identified. However, the currently available data from crystallographic studies suggest that most, if not all, B-cell epitopes on proteins are conformational (9, 10).
Recombinant DNA technology has enabled single in-frame deletions and
N-terminal or C-terminal truncations to reduce or hinder antibody-antigen recognition. Tang et al. (26) and Tamborini et al. (27) have used the deletion system in the
characterization of the major allergens from Aspergillus
fumigatus (Asp f 2) and Lolium perenne (Lol p 1).
Site-directed mutagenesis is a quite recently described technique in
the identification of IgE-binding epitopes (28, 29).
In the present study, a novel chimera-based allergen epitope mapping
strategy was developed. The N terminus, core, and C terminus of hevein
(Hev b6.02), a major latex allergen, were fused with a nonallergenic
hevein homologue (AMP) that served as an immunologically non-IgE-binding adaptor protein. Structure-based modeling confirmed that the designed chimeras could be expressed without structural constraints and with all the possible disulfide bridges formed. IgE from only 2 of 16 hevein-allergic patients bound to the AMP chimera
containing the hevein core region (AMPcore), whereas the N-terminal
chimera (AMPN) was recognized in 88% of the patients, and the
C-terminal chimera (AMPC) was recognized in 38% of the patients. When
both the N-terminal and C-terminal regions of hevein were fused with
the AMP core, IgE from all patients recognized this chimera (AMPN+C).
This chimera was also able to strongly inhibit IgE binding to native
hevein. AMPC showed 30% inhibition, most probably due to low affinity
binding because it was detected only with a concentration of AMPC 3 or
5 orders of magnitude higher than that of AMPN+C or hevein,
respectively. AMPN had no measurable effect in ELISA inhibition. The
linear peptides, which corresponded to hevein regions in AMP chimeras,
showed no significant IgE binding capacity in our study. These results
strongly suggest that the N-terminal and C-terminal regions of hevein
and their correct conformational folding are essential in the formation
of viable IgE-binding hevein epitopes.
Banerjee et al. (30) and Beezhold et al. (31)
have independently reported two reactive areas for IgE antibody
interactions with hevein, containing amino acid residues 19-24 and
25-37 and 13-24 and 29-36, respectively. These regions are located
mainly in the core region, which did not seem to be directly involved in IgE binding in our AMP-hevein chimeras. Differences in IgE epitope
mapping approaches may explain part of the discrepancy. Banerjee
et al. (30) and Beezhold et al. (31) used a
solid-phase system and an overlapping peptide mapping strategy,
whereas, in the present study, synthetic peptides were examined both in
a solid phase (direct ELISA) and in a liquid phase (inhibition ELISA). Moreover, in the present study, a chimera-based IgE epitope mapping strategy was used to identify conformational IgE-binding epitopes. Another reason for the discrepant results may be that patients from
different geographical areas may have developed variable IgE
specificities and recognize different epitope areas on the surface of a
single allergen. However, no data to reject or accept this hypothesis
are currently available.
Recombinant DNA technology can be used in the design of chimeric
proteins, including allergens containing structural and functional properties of parent proteins (related or unrelated) in a single molecule. Colombo et al. (32) and Simon-Nobbe et
al. (33) constructed truncated allergens fused with the
maltose-binding protein and glutathione S-transferase,
respectively, to identify the smallest part of the protein that is
recognized by IgE antibodies from allergic patients. We have also
produced a number of substitution and deletion mutants of
hevein.3 A deletion mutant
lacking the C terminus (amino acid residues 33-43) induced distortion
of the tertiary structure and precipitation of the protein, rendering
it inappropriate for adequate immunological or biochemical analyses.
Discontinuous B-cell epitopes in vespid allergens were recently studied
by King et al. (8), who examined hybrids containing different segments of Ves v5 from the yellow jacket (Vespula
vulgaris) and Pol a5 from paper wasp (Polistes
annularis). They had fused portions of the "guest" allergen
(Ves v5) to large portions of a structurally homologous but poorly
cross-reacting scaffold protein (Pol a5). Physiochemical
characterization suggested that the hybrids had retained their
three-dimensional structures close to the native proteins.
Allergenicity of some hybrids obviously lacking critical epitope
regions of the native allergen was shown to be drastically reduced.
Despite the similar objectives of these studies, there are also
differences. The rationale in our study was to use the "gain in
function" approach, i.e. we wanted to transfer areas of a
known allergen to a virtually nonallergenic adaptor to verify the
existence of functionally active B-cell epitopes. In addition, we
assessed the direct IgE binding reactivity of our chimeras using human
sera. The results showed that biologically viable IgE-binding allergens
were indeed produced by transferring allergenic regions into a
minimally IgE-binding protein homologue. In our system, the
three-dimensional structure of both counterparts, i.e.
hevein and the adaptor protein AMP, were known, which facilitated our
interpretation of the results and allowed us to conclude that relative
small conformational areas of the tightly packed hevein contain major
IgE epitopes.
In conclusion, we describe here a novel allergen epitope mapping
strategy for characterization of conformational IgE-binding epitope
areas. In this approach, short regions of hevein, a major NRL allergen,
were fused with a nonallergenic hevein homologue (AMP) that served as
an immunologically non-IgE-binding adaptor protein. Chimera-based
technology, together with molecular modeling facilitating viable
construct design, enabled us to locate IgE-binding epitope areas to the
N-terminal and C-terminal regions of hevein. This approach may provide
new possibilities for defining structural allergenic or
antigenic epitopes and creating specific reagents for
immunotherapy and vaccine development. Moreover, it should be possible
to create chimeric proteins in an even more general fashion for
studying the structure-function relationship of multifunctional and
multidomain proteins.
| |
ACKNOWLEDGEMENTS |
We thank Irene Helkala, Helena Honkasalo,
and Sari Tillander for excellent technical assistance, and we are
grateful to Dr. Hoong Y. Yeang for the gift of hevein cDNA.
| |
FOOTNOTES |
*
This work was supported by Grant 37852 from the Academy of
Finland.The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in
accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
§§
To whom correspondence should be addressed. Tel.:
358-14-260-2272 or 358-500-599-904; Fax: 358-14-260-2221;
E-mail: markku. kulomaa{at}csc.fi.
Published, JBC Papers in Press, March 21, 2002, DOI 10.1074/jbc.M201076200
2
J. V. Lehtonen, V.-V. Rantanen, D.-J.
Still, M. Gyllenberg, and M. S. Johnson, unpublished data.
3
P. Karisola, J. Mikkola, N. Kalkkinen, J. Helin, O. T. Pentikäinen, S. Repo, T. Reunala, K. Turjanmaa,
M. S. Johnson, T. Palosuo, M. S. Kulomaa, and H. Alenius,
unpublished data.
| |
ABBREVIATIONS |
The abbreviations used are:
NRL, natural rubber
latex;
AMP, antimicrobial protein;
ELISA, enzyme-linked immunosorbent
assay;
MALDI-TOF, matrix-assisted laser desorption ionization
time-of-flight;
HPLC, high pressure liquid chromatography;
HSA, human
serum albumin.
| |
REFERENCES |
| 1.
|
Palosuo, T.,
Mäkinen-Kiljunen, S.,
Alenius, H.,
Reunala, T.,
Yip, E.,
and Turjanmaa, K.
(1998)
Allergy
53,
59-67[Medline]
[Order article via Infotrieve]
|
| 2.
|
Ylitalo, L.,
Alenius, H.,
Turjanmaa, K.,
Palosuo, T.,
and Reunala, T.
(2000)
Clin. Exp. Allergy
30,
1611-1617[Medline]
[Order article via Infotrieve]
|
| 3.
|
Chen, Z.,
Duser, M.,
Flagge, A.,
Maryska, S.,
Sander, I.,
Raulf-Heimsoth, M.,
and Baur, X.
(2000)
Int. Arch. Allergy Immunol.
123,
291-298[Medline]
[Order article via Infotrieve]
|
| 4.
|
Alenius, H.,
Kalkkinen, N.,
Lukka, M.,
Reunala, T.,
Turjanmaa, K.,
Mäkinen-Kiljunen, S.,
Yip, E.,
and Palosuo, T.
(1995)
Clin. Exp. Allergy
25,
659-665[CrossRef][Medline]
[Order article via Infotrieve]
|
| 5.
|
Alenius, H.,
Mäkinen-Kiljunen, S.,
Turjanmaa, K.,
Palosuo, T.,
and Reunala, T.
(1994)
Ann. Allergy
73,
315-320[Medline]
[Order article via Infotrieve]
|
| 6.
|
Alenius, H.,
Kalkkinen, N.,
Reunala, T.,
Turjanmaa, K.,
and Palosuo, T.
(1996)
J. Immunol.
156,
1618-1625[Abstract]
|
| 7.
|
Vrtala, S.,
Akdis, C. A.,
Budak, F.,
Akdis, M.,
Blaser, K.,
Kraft, D.,
and Valenta, R.
(2000)
J. Immunol.
165,
6653-6659[Abstract/Free Full Text]
|
| 8.
|
King, T. P.,
Jim, S. Y.,
Monsalve, R. I.,
Kagey-Sobotka, A.,
Lichtenstein, L. M.,
and Spangfort, M. D.
(2001)
J. Immunol.
166,
6057-6065[Abstract/Free Full Text]
|
| 9.
|
Aalberse, R. C.
(2000)
J. Allergy Clin. Immunol.
106,
228-238[CrossRef][Medline]
[Order article via Infotrieve]
|
| 10.
|
Van Regenmortel, M. H. V.
(1996)
Methods
9,
465-472[CrossRef][Medline]
[Order article via Infotrieve]
|
| 11.
|
Thompson, J. D.,
Gibson, T. J.,
Plewniak, F.,
Jeanmougin, F.,
and Higgins, D. G.
(1997)
Nucleic Acids Res.
25,
4876-4882[Abstract/Free Full Text]
|
| 12.
|
Andersen, N. H.,
Cao, B.,
Rodriguez-Romero, A.,
and Arreguin, B.
(1993)
Biochemistry
32,
1407-1422[CrossRef][Medline]
[Order article via Infotrieve]
|
| 13.
|
Martins, J. C.,
Maes, D.,
Loris, R.,
Pepermans, H. A.,
Wyns, L.,
Willem, R.,
and Verheyden, P.
(1996)
J. Mol. Biol.
258,
322-333[CrossRef][Medline]
[Order article via Infotrieve]
|
| 14.
|
Berman, H. M.,
Bhat, T. N.,
Bourne, P. E.,
Feng, Z.,
Gilliland, G.,
Weissig, H.,
and Westbrook, J.
(2000)
Nat. Struct. Biol.
7,
957-959
|
| 15.
|
Wilson, K. S.,
Butterworth, S.,
Dauter, Z.,
Lamzin, V. S.,
Walsh, M.,
Wodak, S.,
Pontius, J.,
Richelle, J.,
Vaguine, A.,
Sander, C.,
Hooft, R. W. W.,
Vriend, G.,
Thornton, J. M.,
Laskowski, R. A.,
MacArthur, M. W.,
Dodson, E. J.,
Murshudov, G.,
Oldfield, T. J.,
Kaptein, R.,
and Rullmann, J. A. C.
(1998)
J. Mol. Biol.
276,
417-436[CrossRef][Medline]
[Order article via Infotrieve]
|
| 16.
|
Johnson, M. S.,
and Lehtonen, J. V.
(2000)
in
Bioinformatics
(Higgins, D.
, and Taylor, W., eds)
, pp. 15-50, Oxford University Press, Oxford, United Kingdom
|
| 17.
|
 ali, A.,
and Blundell, T. L.
(1993)
J. Mol. Biol.
234,
779-815[CrossRef][Medline]
[Order article via Infotrieve]
|
| 18.
|
Berendsen, H. J. C.,
Van der Spoel, D.,
and Van Drunen, R.
(1995)
Comp. Phys. Comm.
91,
43-56[CrossRef]
|
| 19.
|
Merritt, E. A.
(1997)
Methods Enzymol.
177,
505-524
|
| 20.
|
Airenne, K. J.,
Laitinen, O. H.,
Alenius, H.,
Mikkola, J.,
Kalkkinen, N.,
Arif, S. A.,
Yeang, H. Y.,
Palosuo, T.,
and Kulomaa, M. S.
(1999)
Protein Expression Purif.
17,
139-145[CrossRef][Medline]
[Order article via Infotrieve]
|
| 21.
|
Green, N. M.
(1975)
Adv. Protein Chem.
29,
85-133[Medline]
[Order article via Infotrieve]
|
| 22.
|
Orr, G. A.,
Heney, G. C.,
and Zeheb, R.
(1986)
Methods Enzymol.
122,
83-87[Medline]
[Order article via Infotrieve]
|
| 23.
|
Broekaert, W. F.,
Marien, W.,
Terras, F. R., De,
Bolle, M. F.,
Proost, P.,
Van Damme, J.,
Dillen, L.,
Claeys, M.,
Rees, S. B.,
Vanderleyden, J.,
et al..
(1992)
Biochemistry
31,
4308-4314[CrossRef][Medline]
[Order article via Infotrieve]
|
| 24.
|
Hänninen, A. R.,
Mikkola, J. H.,
Kalkkinen, N.,
Turjanmaa, K.,
Ylitalo, L.,
Reunala, T.,
and Palosuo, T.
(1999)
J. Allergy Clin. Immunol.
104,
194-201[CrossRef][Medline]
[Order article via Infotrieve]
|
| 25.
|
Vriend, G.
(1990)
J. Mol. Biol.
8 (29),
52-56
|
| 26.
|
Tang, B.,
Banerjee, B.,
Greenberger, P. A.,
Fink, J. N.,
Kelly, K. J.,
and Kurup, V. P.
(2000)
Biochem. Biophys. Res. Commun.
270,
1128-1135[Medline]
[Order article via Infotrieve]
|
| 27.
|
Tamborini, E.,
Faccini, S.,
Lidholm, J.,
Svensson, M.,
Brandazza, A.,
Longhi, R.,
Groenlund, H.,
Sidoli, A.,
and Arosio, P.
(1997)
Eur. J. Biochem.
249,
886-894[Medline]
[Order article via Infotrieve]
|
| 28.
|
Smith, A. M.,
and Chapman, M. D.
(1997)
Clin. Exp. Allergy
27,
593-599[CrossRef][Medline]
[Order article via Infotrieve]
|
| 29.
|
Hakkaart, G. A.,
Chapman, M. D.,
Aalberse, R. C.,
and van Ree, R.
(1998)
Clin. Exp. Allergy
28,
169-174[Medline]
[Order article via Infotrieve]
|
| 30.
|
Banerjee, B.,
Wang, X.,
Kelly, K. J.,
Fink, J. N.,
Sussman, G. L.,
and Kurup, V. P.
(1997)
J. Immunol.
159,
5724-5732[Abstract]
|
| 31.
|
Beezhold, D. H.,
Kostyal, D. A.,
and Sussman, G. L.
(1997)
Clin. Exp. Immunol.
108,
114-121[CrossRef][Medline]
[Order article via Infotrieve]
|
| 32.
|
Colombo, P.,
Kennedy, D.,
Ramsdale, T.,
Costa, M. A.,
Duro, G.,
Izzo, V.,
Salvadori, S.,
Guerrini, R.,
Cocchiara, R.,
Mirisola, M. G.,
Wood, S.,
and Geraci, D.
(1998)
J. Immunol.
160,
2780-2785[Abstract/Free Full Text]
|
| 33.
|
Simon-Nobbe, B.,
Probst, G.,
Kajava, A. V.,
Oberkofler, H.,
Susani, M.,
Crameri, R.,
Ferreira, F.,
Ebner, C.,
and Breitenbach, M.
(2000)
J. Allergy Clin. Immunol.
106,
887-895[CrossRef][Medline]
[Order article via Infotrieve]
|
Copyright © 2002 by The American Society for Biochemistry and Molecular Biology, Inc.
 CiteULike  Complore  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
A. C. Drew, N. P. Eusebius, L. Kenins, H. D. de Silva, C. Suphioglu, J. M. Rolland, and R. E. O'Hehir
Hypoallergenic Variants of the Major Latex Allergen Hev b 6.01 Retaining Human T Lymphocyte Reactivity
J. Immunol.,
November 1, 2004;
173(9):
5872 - 5879.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P. Karisola, J. Mikkola, N. Kalkkinen, K. J. Airenne, O. H. Laitinen, S. Repo, O. T. Pentikainen, T. Reunala, K. Turjanmaa, M. S. Johnson, et al.
Construction of Hevein (Hev b 6.02) with Reduced Allergenicity for Immunotherapy of Latex Allergy by Comutation of Six Amino Acid Residues on the Conformational IgE Epitopes
J. Immunol.,
February 15, 2004;
172(4):
2621 - 2628.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
O. H. Laitinen, H. R. Nordlund, V. P. Hytonen, S. T. H. Uotila, A. T. Marttila, J. Savolainen, K. J. Airenne, O. Livnah, E. A. Bayer, M. Wilchek, et al.
Rational Design of an Active Avidin Monomer
J. Biol. Chem.,
January 31, 2003;
278(6):
4010 - 4014.
[Abstract]
[Full Text]
[PDF]
|
|
|
Copyright © 2002 by the American Society for Biochemistry and Molecular Biology.
|
Advertisement
Advertisement
|
TOP
ABSTRACT
INTRODUCTION