Báo cáo khoa học: Molecular metamorphosis in polcalcin allergens by EF-hand rearrangements and domain swapping docx

Molecular metamorphosis in polcalcin allergens
by EF-hand rearrangements and domain swapping
Iris Magler
1
, Dorota Nu
¨
ss
1
, Michael Hauser
2
, Fatima Ferreira
2
and Hans Brandstetter
1
1 Division of Structural Biology, Department of Molecular Biology, University of Salzburg, Austria
2 Division of Allergy, Department of Molecular Biology, University of Salzburg, Austria
Introduction
Allergy is a health problem that is growing at an
almost epidemic rate, with approximately 20% of the
population being affected by type I allergy worldwide
[1–6]. Allergies appear in many versions, including pol-
len and food allergies, and mite dust and environmen-
tally caused allergies. Pollen allergens represent the
largest subgroup, and can be classified into 29 protein
families; most of them belong to the expansin, profilin
or calcium-binding protein families [7].
Massive efforts have been directed at elucidating the
characteristics and causative mechanisms underlying
the action of allergens. Among the biophysical proper-
ties shared by allergens with the ability to breach phys-
ical defense mechanisms in a susceptible host are: (a)

small size, typically ranging from 5 to 30 kDa; (b) high
effective concentration, implying high solubility and
stability; and (c) foreignness to the affected host [8].
Additionally, allergens elicit an IgE response and a
Keywords
covalently locked conformation; EF-hand
protein; protein engineering; structure;
temperature-dependent oligomerization
Correspondence
H. Brandstetter, Billrothstr. 11, 5020
Salzburg, Austria
Fax: +43 662 8044 7209
Tel: +43 662 8044 7270
E-mail:
(Received 5 February 2010, revised 17
March 2010, accepted 7 April 2010)
doi:10.1111/j.1742-4658.2010.07671.x
Polcalcins such as Bet v 4 and Phl p 7 are pollen allergens that are con-
structed from EF-hand motifs, which are very common and well character-
ized helix–loop–helix motifs with calcium-binding functions, as elementary
building blocks. Being members of an exceptionally well-characterized
protein superfamily, these allergens highlight the fundamental challenge in
explaining what features distinguish allergens from nonallergenic proteins.
We found that Bet v 4 and Phl p 7 undergo oligomerization transitions
with characteristics that are markedly different from those typically found
in proteins: transitions from monomers to dimers and to distinct higher
oligomers can be induced by increasing temperature; similarly, low concen-
trations of destabilizing agents, e.g. SDS, induce oligomerization transitions
of Bet v 4. The changes in the quaternary structure, termed molecular
metamorphosis, are induced and controlled by a combination of EF-hand

rearrangements and domain swapping rather than by the classical law of
mass action. Using an EF-hand-pairing model, we provide a two-step
model that consistently explains and substantiates the observed metamor-
phosis. Moreover, the unusual oligomerization behavior suggests a straight-
forward explanation of how allergens can accomplish the crosslinking of
IgE on mast cells, a hallmark of allergens.
Structured digital abstract
l
MINT-7718612: Bet v 4 (uniprotkb:Q39419) and Bet v 4 (uniprotkb:Q39419) bind (MI:0407)
by molecular sieving (
MI:0071)
l
MINT-7718648: Phl p 7 (uniprotkb:O82040) and Phl p 7 (uniprotkb:O82040) bind (MI:0407)
by molecular sieving (
MI:0071)
Abbreviations
GFP, green fluorescent protein; TEV, tobacco etch virus.
2598 FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS
clinical response, which may represent an immediate
and ⁄ or late-phase response [9]. Many allergens show
proteolytic activity, for which, in selected cases, a cau-
sal connection has been demonstrated [10–12]; addi-
tionally, surface-exposed hydrophobic patches have
been suggested to provide allergen-typical danger sig-
nals that are recognizable to the innate immune system
[13,14]; similarly, glycosylation patterns present on
allergen surfaces are believed to be involved in recogni-
tion and endocytotic internalization by innate immune
cells [15]. For a recent review, see [16]. The increased
biological knowledge is accompanied by an enormous

increase in the available structural database on aller-
gens, accomplished by crystallography and NMR
projects [17–32]. Despite this progress, our mechanistic
understanding of the molecular principles of allergen-
icity remains unsatisfactory. This is highlighted by the
fact that we are unable to predict the allergenic behav-
ior of a protein on the basis of biophysical properties,
such as its crystal structure [33,34]. We lack reliable
structural motifs that could serve as hallmarks of aller-
genicity – such as a catalytic triad and an oxyanion
hole that could identify proteolytic activity.
The investigations in the current study were aimed
at the identification of a biophysical hallmark that
could distinguish allergens from other proteins and
could ultimately reveal a causative mechanism acting
in a subfamily of allergens. To this end, we investi-
gated the hypothesis that the ability to undergo con-
formational changes represents a distinguishing feature
of allergens. The concept of molecular metamorphosis
is receiving increasing attention [35,36]. We have iden-
tified and characterized this unexpected molecular
metamorphosis in the pollen allergens Bet v 4 from the
white birch and Phl p 7 from timothy grass. These
allergens are built from EF-hand motifs, which are
exceptionally well-studied building blocks [37,38]. We
have identified physicochemical parameters that con-
trol the oligomerization transitions, and provide a
model relating the oligomerization to the ability of the
allergens to crosslink already synthesized IgE antibod-
ies on mast cells.

Results
Bet v 4 can be expressed in a soluble, SDS-stable
dimeric form
Bet v 4 and the related Phl p 7 were expressed in
Escherichia coli BL21(DE3) cells. Typically, SDS⁄
PAGE analysis of intact cells indicated the expression
of monomeric proteins with approximate sizes of
12.5 kDa and 11.7 kDa, as shown for Bet v 4 in
Fig. 1A and Phl p 7 in Fig. 1C, respectively. Purifica-
tion to almost homogeneity was achieved in a single
step by employing immobilized metal affinity chroma-
tography (Fig. 1B,C).
Under standard storage conditions, both Bet v 4
and Phl p 7 were also monomeric under native condi-
tions, as judged by gel filtration chromatography
(Fig 3A).
Surprisingly, we observed spontaneous dimerization
of Bet v 4 with a size of 25 kDa by SDS ⁄ PAGE
(Fig. 2A). Although we repeated the expression of
dimeric Bet v 4 more than 10 times, the underlying
mechanism of dimerization is partly statistical in nat-
ure, because we observed dimerization in $ 1–2% of
the expression trials only. However, when dimerization
72
55
43
72
55
43
72

55
43
12345 678 1 2 3 4 5 6 1 2 3 4 5 6 7 8910
34
26
34
26
34
26
17
17
10
17
10
10
ABC
Fig. 1. Bet v 4 and Phl p 7 protein samples appear exclusively as monomers on SDS ⁄ PAGE when expressed and purified under standard
conditions. (A) Expression of Bet v 4 under standard conditions. Lane 1: protein standard (Fermentas). Lane 2: sample before induction.
Lanes 3–8: Samples 4 h after induction. All samples were drawn from different expression flasks. (B) Bet v 4 purification by affinity chroma-
tography. Lane 1: protein standard. Lane 2: Bet v 4 cell lysate. Lane 3: flow-through. Lane 4: wash step. Lanes 5 and 6: eluted protein with-
out impurities. (C) Expression and purification of Phl p 7. Lane 1: protein standard. Lane 2: sample before induction. Lanes 3–6: samples
from different expression flasks 4 h after induction. Lane 7: unbound protein impurities (flow-through) after Ni
2+
–nitrilotriacetic acid treat-
ment. Lane 8: wash fraction. Lanes 9 and 10: purified Phl p 7 protein.
I. Magler et al. Molecular metamorphosis in allergens
FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS 2599
was observed at all, it was apparently 100% complete.
The statistical nature of the dimerization is puzzling,
and cannot be explained by obvious factors, e.g. the

presence or absence of metal factors such as Ca
2+
or
EDTA, as described in more detail below.
To exclude the possibility of artefacts and to confirm
the identity of the protein, we cleaved the N-terminal
His
6
-tag by utilizing the tobacco etch virus (TEV)
protease cleavage site. Figure 2B shows that, upon
removal of the N-terminal His
6
-tag plus linker
($ 3 kDa), the migration of the protein on SDS ⁄ PAGE
corresponds to a molecular mass reduced by approxi-
mately 6 kDa, as expected. As a consequence, we can
conclude that the dimer contact is not mediated by, but
is independent from, the N-terminus. The identity of the
Bet v 4 protein was unambiguously confirmed by
ESI-MS. The dimerization is reversible, because Bet v 4
monomers were observed by SDS ⁄ PAGE after several
weeks of storage at 4 °C. This finding, in particular,
shows that dimerization can take place at $ 37 °C.
Spontaneous in vitro dimerization of Bet v 4
When Bet v 4 was expressed as a monomer, it
remained in the monomeric state when stored at 4 °C
or 20 °C(Fig. 3A). By serendipity, we identified spon-
taneous dimerization of a Bet v 4 sample that was left
on the bench in the summer for weeks, as analyzed
by gel filtration chromatography. These findings

prompted us to systematically investigate possible
mechanisms that govern the unexpected and intriguing
oligomerization behavior of Bet v 4. Given the storage
at elevated temperatures over a very long time period,
we hypothesized that temperature and incubation time
may affect the oligomerization behavior.
Distinct oligomerization transitions in Bet v 4 and
Phl p 7 can be induced by temperature changes
We systematically studied the temperature dependence
of the oligomerization state of Bet v 4 under native
conditions by using gel filtration chromatography. Oligo-
merization was observed at $ 30 ° C, but only over
time intervals of several months. These long incubation
times effectively excluded the option to conduct sys-
tematic experiments at 30 °C. However, when heated
to 75 °C, a mixture of monomeric and dimeric proteins
appeared quite rapidly (Fig. 3B). When the tempera-
ture was further increased to 95 °C, the dimeric form
of the protein was observed exclusively (Fig. 3C). Con-
sequently, temperature is one key parameter that
induces Bet v 4 oligomerization transitions in vitro.
Naturally, the question arises of whether the structure
of Bet v 4 remains intact at high temperatures; we con-
firmed the structural integrity by performing overnight
CD measurements at 75 °C, as detailed below.
Oligomerization depends on incubation time
The oligomerization transitions did not occur instanta-
neously, but required some incubation time. To quan-
tify the required time scale, we analyzed protein
oligomerization after distinct incubation times.

We found approximately 75% of the protein to be
monomeric after 24 h at 75 °C and the rest of the
protein to be dimeric, whereas the situation was
12345678 1234567
43
55
72
72
55
43
34
26
34
17
26
17
11
11
AB
Fig. 2. Spontaneous and complete dimerization of Bet v 4 can be observed during protein expression. (A) Expression of the SDS-stable
dimer of Bet v 4 in E. coli BL21(DE3) cells. Lane 1: mass standard. Lane 2: cells before induction. Lanes 3–8: samples 4 h after induction.
Note that the samples in the different lanes were drawn from different expression flasks and showed complete dimerization in each case.
(B) Cleavage of the N-terminal His
6
-tag. Following the mass standard (lane 1), the His
6
-tagged Bet v 4 is shown before addition of the TEV
protease (lane 2). The protein migrates at an apparent size of 25 kDa ($ 2 · 12.5 kDa). Lanes 3–7: TEV-digested Bet v 4 at different time
points; TEV protease is visible at $ 30 kDa. Lane 3: TEV digest at time zero. Lane 4: digest after 1 h. Lane 5: digest after 6 h. Lane 6: digest
after 12 h. Lane 7: digest after 24 h. TEV protease cleavage releases the N-terminal His

6
-tag and thus shifts the size of the protein to about
19 kDa ($ 2 · 9.5 kDa).
Molecular metamorphosis in allergens I. Magler et al.
2600 FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS
reversed after 48 h (Fig. 3D–F). After 72 h, all of the
monomeric protein was converted to higher oligomers.
The process of oligomerization does not stop with
dimer formation. Instead, higher oligomeric forms
were observed by size exclusion chromatography after
72 h of incubation (Fig. 3F).
As mentioned earlier, we found dimerization of
Bet v 4 that was stored at room temperature for
approximately 4–10 weeks, whereas Bet v 4 stored at
4 °C was always found to be monomeric. Significantly,
the observed oligomerization does not correspond to
an unspecific aggregation phenomenon, but reflects a
reversible transition between distinct oligomerization
states; in particular, monomer formation can be
induced by lowering the temperature to 4 °C within
few days. Therefore, we conclude that Bet v 4 oligo-
merization depends on both incubation temperature
and time in a multiplicative manner.
SDS induces instantaneous monomer-to-dimer
transitions in Bet v 4 at room temperature
In contrast to monomer-to-dimer transitions in vitro,
dimerization in E. coli cells can take place rapidly,
within a few hours (Fig. 2A). We hypothesized that
dimerization can be efficiently catalyzed by compounds
that are presumably present in E. coli in trace

amounts. Therefore, we systematically screened a vari-
ety of chemicals, including Ca
2+
and other metals, for
their effect on oligomerization, both in expression con-
ditions and with purified protein.
Ca
2+
is known to affect the 3D structure of the
Bet v 4 monomer [32]. Interestingly, addition of neither
10 mm Ca
2+
nor EDTA had a direct effect on the
dimerization behavior, as judged by SDS⁄ PAGE and
gel filtration chromatography, which gave results iden-
tical to those shown in Fig. 3. These findings were
further corroborated by CD measurements, as
described below. Surprisingly, we found that SDS led
to partial dimer formation in Bet v 4 at 20 and 4 °C.
The addition of 0.05% SDS led to equal amounts of
the monomeric and dimeric states, as reflected by two
prominent peaks at approximately 13 and 11.4 mL
(Fig. 4A). To a lesser extent ($ 10%), a highly oligo-
meric species was observed at an elution volume of
$ 8 mL, represented by a broad peak. At an SDS con-
centration of 0.5%, nearly all of the protein aggregates
and only approximately 10% of the protein remained
in the monomeric or dimeric state, as shown by the
dashed line in Fig. 4A.
The bimodal oligomerization behavior of Bet v 4

contrasts with what is seen for most soluble proteins,
as confirmed by a control experiment with green fluo-
rescent protein (GFP). Whereas native GFP migrates
10.0
20.0
30.0
40.0
50.0
60.0
mAU
0.0 5.0 10.0 15.0 20.0 ml
12.15
13.56
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
mAU
0.0 5.0 10.0 15.0 20.0 25.0 30.0 ml
11.95
13.56
10.0
20.0
30.0
40.0
mAU

0.0 5.0 10.0 15.0 20.0 ml
9.19
11.63
24 h 48 h 72 h
0.0
20.0
40.0
60.0
80.0
mAU
0.0 5.0 10.0 15.0 20.0 25.0 ml
13.75
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
mAU
0.0 5.0 10.0 15.0 ml
11.97
13.58
5.0
10.0
15.0
20.0
25.0

30.0
35.0
mAU
0.0 5.0 10.0 15.0 20.0 ml
11.78
4 °C
A
75 °C 95 °C
Monomer Dimer Monomer Dimer
Dimer Monomer Dimer Monomer Tetramer Dimer
BC
DEF
Fig. 3. Temperature and time affect the dimerization of Bet v 4. (A–C) The temperature dependence of Bet v 4 oligomerization was analyzed
by gel filtration chromatography. Bet v 4 samples were incubated for 48 h at (A) 20 °C, (B) 75 °C, and (C) 95 °C. The experiments showed a
monomer-to-dimer transition as a function of incubation temperature. Further details are given in Experimental procedures. (D–F). The time
dependence of Bet v 4 oligomerization as analyzed by gel filtration chromatography. Bet v 4 was incubated at 75 °C and analyzed every 24 h
for 3 days.
I. Magler et al. Molecular metamorphosis in allergens
FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS 2601
exclusively as a monomer on gel filtration (Fig. 4B,
continuous line), addition of 0.05% SDS induced
aggregation of $ 10% of the protein (Fig. 4B, dashed
line). In summary, 0.05% SDS generates an equilib-
rium between native and aggregated GFP, resembling
the situation for Bet v 4, with the exception that
Bet v 4 can be monomeric or dimeric in its native
state.
Bet v 4 can be conformationally locked in the
monomeric state
The oligomerization properties of Bet v 4 revealed

unexpected and unique features, such as its dependence
on temperature and chemicals. Moreover, the dimer-
ization is apparently independent of protein concentra-
tion in the range from 0.1 to 25 mgÆmL
)1
. These
unique properties suggested that, in Bet v 4, oligomeri-
zation could involve not only intermolecular recogni-
tion events, governed by the law of mass action, but
also intramolecular conformational rearrangements.
To investigate this hypothesis, we constructed a
Bet v 4 variant containing a K25C ⁄ F60C double muta-
tion. On the basis of the NMR structure of monomeric
Bet v 4 [32], we devised these point mutations to form
an intramolecular disulfide bond that stabilizes the
conformation by covalently linking both EF-hand
motifs in Bet v 4 (Fig. 5).
This covalent linkage is absent in the presence of
dithiothreitol. If an intramolecular rearrangement does
indeed accompany the oligomerization of Bet v 4, the
oligomerization behavior of oxidized (disulfide-linked)
Bet v 4-K25C ⁄ F60C should deviate markedly from
that of the wild type. By contrast, reduced Bet
v 4-K25C ⁄ F60C should show oligomerization behavior
identical to that of the wild type.
We carried out experiments to test both the tem-
perature and time dependence of the oligomerization
by incubating the disulfide-linked Bet v 4 double
mutant at 20 °C for 7 days and at 75 °C for 24 h.
Under both conditions, the monomer was stable over

the observation period, as monitored by gel filtration
(Fig. 6A).
As a control experiment, we carried out similar
experiments under reducing conditions using 5 mm
dithiothreitol. The reduced Bet v 4 double mutant was
incubated at 75 °C for 24 h, and subsequently ana-
lyzed by gel filtration. The reduced protein revealed
the native-like induction of higher oligomer formation
(Fig. 6B, continuous line), clearly contrasting with
the behavior of the oxidized double mutant (Fig. 6B,
broken line).
0.0
20
40
60
80
mAU
0.0 2.0 4.0 6.0 8.0 10.0 12.0 ml
8.01
11.41
12.77
7.50
11.40
12.74
0.0
10.0
20.0
30.0
40.0
50.0

60.0
70.0
mAU
5.0 6.0 7.0 8.0 9.0 10.0 11.0 ml
8.71
10.67
10.70
AB
Fig. 4. SDS induces dimerization in Bet v 4. (A) Solid line: at 0.05% SDS and 20 °C, most of the Bet v 4 elutes at retention volumes corre-
sponding to monomers and dimers, with only a small ($ 10%) aggregated fraction eluting near the void volume. Dashed line: at 0.5% SDS,
most (90%) of the Bet v 4 aggregates (eluting at the void volume: 7.5 mL), and only 10% elutes at volumes corresponding to monomers
and dimers. (B) Solid line: Control experiment using GFP at 0.05% SDS and 20 °C reveals a predominantly native monomeric form, corre-
sponding to a retention volume of 10.67 mL, and a small ($ 10%) aggregated fraction eluting near the void volume (retention volume of
8.71 mL). Dashed line: at 0% SDS, GFP migrates exclusively as a monomer.
C25
C60
NH
2
COOH
Fig. 5. Engineering of a disulfide bridge intended to lock the mono-
mer conformation of Bet v 4. The introduced K25C ⁄ F60C double
mutation promotes disulfide bond formation between the two anti-
parallel b-strands, and thus crosslinks the first Ca
2+
-binding EF-hand
(shown in red) with the second EF-hand (blue).
Molecular metamorphosis in allergens I. Magler et al.
2602 FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS
Consequently, the formation of Bet v 4 dimers and
higher molecular mass oligomers does indeed involve

an intramolecular conformational rearrangement.
Temperature-induced oligomerization may be
universally conserved in polcalcins
Next, we tested whether the intriguing temperature-
dependent and time-dependent oligomerization is spe-
cific to Bet v 4 or could also be found in structurally
related proteins. We selected Phl p 7 as a further repre-
sentative of Ca
2+
-binding EF-hand proteins, and car-
ried out oligomerization analyses analogous to those
described for Bet v 4. We found that Phl p 7 is indeed
able to undergo temperature-dependent oligomeriza-
tion: At 4 °C, Phl p 7 formed monomers exclusively,
whereas significant amounts of dimeric Phl p 7 accumu-
lated after overnight incubation at 75 °C. Overnight
incubation at 95 °C completely converted the mono-
meric form of Phl p 7 to higher oligomer forms
(Fig. 7). The temperature dependence of the oligomeri-
zation state of Phl p 7 thus parallels the behavior
observed with Bet v 4. The broadness of the 95° C peak
may be partly related to heat-induced denaturation.
The secondary structure content is independent
of the oligomerization state and is conserved at
75 °C
We employed CD spectroscopy to investigate whether
the secondary structure of Bet v 4 at room temperature
was dependent on its oligomeric state. Furthermore, as
we used heating of Bet v 4 as a tool to speed up the
conformational transition, we wished to clarify

whether the structure becomes disrupted at 75 °C.
Finally, we used an engineered disulfide-containing
variant to test the nature of the conformer transforma-
tion, which raises the question of how well this mutant
resembles the wild type. CD is ideally suited for pro-
viding answers to these questions.
To investigate the first question, we used Bet v 4
stored at 4 °C, corresponding to the monomeric spe-
cies, and Bet v 4 that had been heated to 75 °C over-
night, corresponding to the dimeric species. The
0.0
20.0
40.0
60.0
80.0
mAU
6.0 8.0 10.0 12.0 14.0 16.0 18.0 ml
10.66
12.80
0.0
20.0
40.0
60.0
80.0
mAU
6.0 8.0 10.0 12.0 14.0 16.0 18.0 ml
10.66
12.80
AB
Fig. 6. Disulfide bond inhibits the monomer-dimer transformation. (A) The gel filtration chromatogram of the disulfide-containing (oxidizing)

Bet v 4 mutant; the protein was incubated at 4 °C (continuous line) and at 75 °C (dashed line) for 24 h. The chromatogram did not change
over an incubation period of up to 7 days, indicating an exclusively monomeric state. (B) The gel filtration chromatogram of the reduced
Bet v 4 mutant (no disulfide bond); the protein was incubated at 4 °C (continous line) and at 75 °C (dashed line) for 24 h. The heat-treated
protein eluted at a retention volume corresponding to the dimer, resembling the wild-type protein in this respect.
20
40
60
80
100
120
mAU
0.0 5.0 10.0 15.0 20.0 25.0 ml
12.62
0
100
200
300
400
500
600
mAU
0.0 5.0 10.0 15.0 20.0 ml
10.12
12.45
0
100
200
300
400
mAU

0.0 5.0 10.0 15.0 ml
10.08
11.50
Monomer Dimer Monomer Dimer
AB C
Fig. 7. Oligomerization of Phl p 7. Gel filtration chromatograms indicate the conversion from monomer to higher oligomerization states at
4 °C (A), 75 °C (B) and 95 °C (C) over an incubation period of 24 h, qualitatively resembling the behavior of Bet v 4.
I. Magler et al. Molecular metamorphosis in allergens
FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS 2603
oligomerization states were further confirmed by gel
filtration chromatography. Both protein samples
yielded CD spectra that revealed well-folded a-helical
proteins (Figs 8A and S1). Therefore, the secondary
structure in the monomeric and dimeric species is
qualitatively identical.
Next, we tested the a-helical content of Bet v 4 at
75 °C at three time points: after 15 min, after 16 h,
and after 20 h. The sample was kept at 75 °C. In all
three samples, the a-helical content was preserved, and
qualitatively coincided with that indicated by the spec-
tra measured at 20 °C (Fig. 8B). The double minimum
structure characteristic of a-helices was less pro-
nounced in the heated samples, however. Similarly, a
quantitatively reduced mean residual weight ellipticity
indicated increased flexibility in the secondary struc-
ture of the heated samples (Fig. S1).
As a third experiment, we tested the disulfide-engi-
neered Bet v 4 mutant (K25C ⁄ F60C). The disulfide
mutant stored at 4 C resembled the native monomer
and gave CD spectra qualitatively identical to those of

the unmodified protein, independently of whether the
disulfide bond was formed or reduced (Fig. 8C, CC oxi ⁄
red 4 °C). Interestingly, whereas the overall secondary
structure content was also conserved after heating at
75 °C, there appeared to be significant disorder in these
protein variants (Fig. 8C). The reduction in a-helix
content is most prominent in ‘CC oxi 75 °C’, in which
the monomeric state is enforced by the intact disulfide
bond. These findings are paralleled by the quantitative
representation of the ellipticity in Fig. S1.
Finally, we confirmed that Ca
2+
is tightly bound by
Bet v 4 and cannot be extracted by the addition of
10 mm EDTA, as demonstrated by the qualitatively
unchanged CD spectrum in the presence of EDTA
(Fig. S2).
Discussion
Bet v 4 forms monomers, dimers, and higher
oligomers
We identified several distinct oligomerization states for
Bet v 4. Although these findings appear to be in con-
flict with those from previous experiments, these dis-
agreements may be reconciled by considering the
settings of the particular experiments [32]. This is of
particular relevance for experiments with measurement
times of days, such as NMR and ultracentrifugation,
which were run at a constant temperature of 20 °Cor
4 °C, respectively. In fact, also in our hands, the pro-
tein’s oligomerization behavior over time was tempera-

ture-dependent. Even when stored for several months
at 4 °C, Bet v 4 remained in a monomeric conforma-
tion. If expressed in monomeric form, Bet v 4
remained monomeric over days to weeks at 20 °C.
However, after months, Bet v 4 adopted a dimeric con-
formation at room temperature.
Temperature is a universal inducer of
oligomerization transitions
The consistent observations made with Phl p 7 and
Bet v 4 suggest to us that temperature acts as an
important order parameter controlling oligomer forma-
tion in polcalcins. The fact that an increase in temper-
Fig. 8. CD measurements document the structural integrity of diverse Bet v 4 species. Data are presented as baseline-corrected mean resi-
due molar ellipticity [Q]
MRW
at a given wavelength. (A) The spectra of monomeric Bet v 4 protein stored 4 °C (continuous line) and after
heat-induced dimerization (dashed line) qualitatively coincide, indicating a near-identical secondary structure content. (B) Time series of CD
spectra of Bet v 4 kept and analyzed at 75 °C for 15 min (continuous line), 16 h (dashed line with dots), and 20 h (dashed line). The second-
ary structure is mostly conserved, and does not noticeably vary over time. (C) CD spectra of Bet v 4-K25C ⁄ F60C (CC) with the disulfide bond
formed (oxi) or reduced (red), each stored at either 4 °Cor75°C (overnight). When stored at 4 °C, the (monomeric) CC mutant adopted a
native-like ellipticity spectrum, independently of the status of the disulfide bond (oxidized or reduced), indicating a native like 3D structure.
After heat treatment, the qualitative form of the spectrum remained conserved, albeit with a significantly reduced amplitude.
Molecular metamorphosis in allergens I. Magler et al.
2604 FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS
ature induces transitions to high molecular mass oligo-
mers is surprising: with increasing temperature (T), the
entropy (S) of the protein becomes more important for
its Gibbs free energy (G) than the enthalpic contribu-
tion (E ): G = E ) TS. Thus, dissociation of oligomers
should be favored at high temperature, because the

degrees of freedom are maximized for monomers.
However, the observed behavior contrasts with these
fundamental physicochemical principles, and points to
the existence of temperature-induced intramolecular
rearrangements in Bet v 4. In other words, although
CD spectra indicate that the secondary structures of
monomers and dimers are alike (Fig. 8), their detailed
conformations differ in a subtle way (Fig. 10). Only
the excited conformation is able to form dimers; the
ground state conformation is monomeric. Importantly,
although, for practical reasons, we performed the
experiments shown in Figs 3, 6 and 7 at unphysiologi-
cally high temperatures, these transitions do also occur
at ambient temperatures. Additionally, the tempera-
ture-induced transitions may well be catalyzed by other
components present in the pollen, as discussed below.
As an additional cautionary remark, we must point
out that structural integrity could be demonstrated for
temperatures up to 75 °C only (Fig. 8); the sample at
95 °C may be partially unfolded.
Like temperature, chemical substances induce
metamorphosis by stabilizing or destabilizing
local free energy minima
SDS is known to destabilize the quaternary and or
ternary structure of proteins [39,40]. This effect is also
observed in Bet v 4 and our control protein, GFP,
leading to a broad peak near the void volume of the
gel filtration column (Fig. 4A,B). Significantly, how-
ever, we found that SDS induced specific dimerization
of Bet v 4, as reflected by a sharp elution peak

(Fig. 4A).
It is very likely that a number of other physiological
chemicals will affect the oligomerization of Bet v 4. In
fact, sodium chloride at 0.5 m favors the monomeric
state of Bet v 4 over the dimeric state. These findings
support the notion that oligomer transformations
are relevant in the physiological environment of the
pollen.
This observation can be explained by assuming a
multimodal free energy surface of Bet v 4 with several
distinct substates; in a simplified version, this surface
can be represented by two isothermal free energy
graphs (Fig. 9). The ratio of the free energy minima
representing the monomeric and dimeric states changes
with temperature. This property reflects the surprising
fact that dimers and higher oligomers are preferred
over monomers at high temperature.
The differences in free energy can be quantitatively
estimated by exploiting the fact that the statistics of
oligomer formation are governed by Boltzmann’s law
p ¼
1
Z
e
ÀDG=RT
where the probability p corresponds to the likelihood
of a dimer and is estimated to be $ 1%, reflecting the
frequency of observation of spontaneous dimerization
(see Results). Z represents the partition function,
which we roughly estimate, from the number of acces-

sible states, to be Z = 3 (monomer, dimer, and tetra-
mer). R is the gas constant [8.3 JÆ(mol K)
)1
], T is the
absolute temperature (300 K), and DG is the change in
free energy associated with the monomer-to-dimer
transition. Reformulation of Boltzmann’s law yields
DG ¼ÀRT ln ðZ Á 0:01Þ%RT ln ð3 Á 0:01Þ%9kJ
We admit that our estimations of both P and Z are
quite inaccurate; however, as both parameters (P, Z)
affect DG via logarithmic dependence, errors will trans-
late to the resulting DG value only in a dampened
manner.
In addition to this phenomenological consideration,
we tried to develop a structural model that could pro-
vide a mechanism that explains this counterintuitive
behavior. Such a model is developed below.
High temperature
Low temperature
12
Monomer Dimer
Oligomerization State
ΔG
Fig. 9. Schematic free energy diagram of Bet v 4 governing the
occupancy of its conformational substates. The free energy
depends most prominently on the temperature: at low temperature
(i.e. 4 °C), the monomer is preferred, whereas higher oligomers are
preferred at higher temperatures. Additionally, the free energy
depends on parameters such as ionic strength or SDS, which
together result in a more complex hypersurface, as illustrated.

I. Magler et al. Molecular metamorphosis in allergens
FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS 2605
Dimerization is a two-step process involving an
excited conformation of Bet v 4
The anomalous dependence of the oligomerization on
temperature and ⁄ or chemical substances had already
indicated that changes in the tertiary structure of the
Bet v 4 subunits precede the monomer-to-dimer transi-
tion. Proof for this hypothesis was obtained by engi-
neering a disulfide-containing variant that locks the
known 3D structure of the monomeric substate. This
protein is unable to undergo dimerization or higher
oligomer formation under oxidizing conditions in
which the disulfide bridge is conserved. This model
also explains the different modes of action of tempera-
ture and SDS. The latter slightly destabilizes the
monomeric state and effectively lowers the separating
energy barrier, leading to a population of both mono-
mers and dimers (Fig. 4A). The effect of temperature
is more sophisticated: although it also helps to over-
come the separating energy barrier, an additional
mechanism is required to explain why dimers are pre-
ferred over monomers at high temperature.
Our model involves a two-step process. First, elevated
temperatures will induce a conformational transition
within the Bet v 4 subunit from the ground state to an
excited state, whereby the ground state represents a
closed conformation and theexcited state an open con-
formation with no EF-hand pairing. Both the ground
state and excited state are monomeric (Fig. 5). Two

hydrogen bonds in the short antiparallel b-sheet have to
be broken in the excited state, representing an energetic
barrier that matches that derived previously from Boltz-
mann’s distribution ($ 9 kJ). In a second step, this
excited form is now able to gain enthalpy by intermolec-
ular pairing of the EF-hands, and thus forms dimers
(Fig. 10A,B). Such an extended conformation has been
observed in the crystal structure of Phl p 7 [41].
It should be noted that a straightforward extension
of the EF-hand-pairing model, shown in Fig. 10C,
explains the existence of higher order oligomers. Here,
we assume that an initial dimer is formed by monomer
pairing with one EF-hand rather than two. This gener-
ates a ‘sticky overhang’ dimer that provides two addi-
tional EF-hand docking sites. These docking sites can
attract additional Bet v 4 subunits, and thus provide a
mechanism to generate trimers, tetramers, and higher
oligomers. This model further explains why Bet v 4
proteins migrating as dimers during gel filtration may
differ on SDS ⁄ PAGE. We propose that single
EF-hand-paired dimers are less stable and dissociate to
form monomers on SDS ⁄ PAGE, whereas double
EF-hand-paired dimers stay intact on SDS ⁄ PAGE.
The proposed model (Fig. 10) has a qualitatively
unchanged secondary structure content in the mono-
meric and dimeric states, consistent with the recorded
CD spectra (Fig. 8).
Fig. 10. (A, B) EF-hand pairing as a mechanism for dimerization. Bet v 4 consists of two EF-hand motifs, EF1 (blue) and EF2 (red), which are
connected by a flexible connecting segment (green). A C-terminal helix (a5; gray) presumably contributes to stabilization of the EF-hand pair-
ing. In the experimentally determined monomer structure, intramolecular EF-hand pairing occurs via strands b1 and b2, forming a central

antiparallel b-sheet. This structure represents the ground state conformation. On the basis of the crystal structure of dimeric Phl p 7 [41],
we propose that dimerization is mediated by intermolecular EF-hand pairing via strands b1 and b2¢ and strands b2 and b1¢. For this dimeriza-
tion to occur, we propose the existence of an excited state intermediate (open form) that will be increasingly common at high temperature.
(C) Alternatively, a singly EF-hand-paired dimer may form, as shown here, via strands b1 and b1¢; possible alternative dimers would involve
strands b2 and b2¢, b1 and b2¢,orb2 and b1¢. Singly EF-hand-paired dimers will be less stable than doubly paired dimers.
Molecular metamorphosis in allergens I. Magler et al.
2606 FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS
CD measurements further showed that the second-
ary structure of Bet v 4 is mostly conserved during the
overnight heat treatments (Fig. 8). This lends support
to the notion that the heating protocol only accelerates
the transformation from monomer to dimer, and does
not change the reaction path of the transformation. In
particular, the heat-induced dimerization does not
occur via an unfolding–folding process.
Finally, the CD measurements of the disulfide-
containing variant indicate that the Bet v 4 dimer
structure will deviate in subtle details from the pro-
posed Phl p 7 dimer structure (Fig. 10).
A remaining puzzle is why Bet v 4 is mostly
expressed as a monomer in E. coli, but sometimes as a
dimer; moreover, if Bet v 4 is expressed as dimer, it is
exclusively dimeric. It seems quite plausible that the
difference in the observed oligomerization relates to
the presence of a dimerization catalyst. We propose
that a chaperone, such as a heat shock protein, could
account for the observed dimerization behavior. The
expression rate of heat shock proteins varies drastically
upon subtle and difficult-to-control changes.
Molecular metamorphosis provides a framework

to explain the ability of polcalcins to crosslink
IgE antibodies on mast cells
Naturally, the question arises of whether the intrigu-
ing biophysical behavior of Bet v 4 and Phl p 7
relates to their allergenic properties. The molecular
metamorphosis model provides a straightforward
explanation for one allergenic key feature, namely,
the ability to crosslink IgEs on mast cells. Addition-
ally, it is very plausible that the oligomerization status
of an allergen will affect its endocytosis and endoso-
mal processing. Significantly, dimerization or multi-
merization has been reported for numerous allergens
[42–51]. Clearly, IgE binding is necessary, but not suf-
ficient, to induce a Th2 immune response, which is
characteristic of allergy. In fact, there is conclusive
evidence for selected allergens that the allergenicity,
including antibody-binding capacity, differs for mono-
mers and dimers: Scho
¨
ll et al. have shown, for the
birch pollen allergen Bet v 1, that dimers (34 kDa),
and not monomers (17 kDa), represent the allergenic
Bet v 1 species [51]. The same basic mechanism has
been reported by Reese et al. for the carrot allergen
Dau c 1 [45]. According to the molecular metamor-
phosis hypothesis presented here, we suggest that con-
formationally locked allergens (e.g. disulfide-
stabilized) will cause drastically reduced allergic reac-
tions. Clearly, the validation of this hypothesis awaits
further experiments.

Experimental procedures
Materials
Plasmids coding for Bet v 4 and Phl p 7 (Uniprot Database
accession numbers are Q39419 for Bet v 4 and O82040 for
Phl p 7) were isolated from pollen, as described previously
[52,53]. Restriction enzymes and T4 ligase were obtained
from Fermentas (St Leon-Rot, Germany). Pfu Ultra II
Fusion HS DNA polymerase was obtained from Stratagene
(La Jolla, CA, USA). Custom-made primers were obtained
and sequence analyses were performed at Eurofins MWG
Operon (Germany). E. coli strain XL1 Blue (Stratagene)
was used for subcloning. Strain BL21(DE3) (Novagen,
Madison, WI, USA) was used as host strain for protein
expression. For expression, LB-Lennox (Roth, Karlsruhe,
Germany) was used. All reagents were of the highest stan-
dard available from Sigma-Aldrich (Mu
¨
nchen, Germany)
or AppliChem (Darmstadt, Germany).
Cloning
The plasmids were cloned in the pHIS parallel II vector
with an NcoI site at the 5¢-end and an EcoRI site at the
3¢-end [54]. To engineer the disulfide mutant of Bet v 4,
a double mutation K25C ⁄ F60C was constructed by site-
directed mutagenesis using the QickChange method [55].
The following primers were used: 5¢ -gccaatggcgatggt
TGCat-
ctcAgcagcagag-3¢ [K25C forward primer, bases exchanged
are underlined, silent control restriction (PstI) site is in
bold]; 5¢-ctctgctgcTgagat

GCAaccatcgccattggc-3¢ (K25C
reverse primer, bases exchanged are underlined, control
restriction site is in bold); 5¢-accgatggcgacggA
TGCatt-
tcgttccaagag-3¢ [F60C forward primer, bases exchanged are
underlined, control restriction site (NsiI) is in bold]; and
5¢-ctcttggaacgaaat
GCATccgtcgccatcggt-3¢ (F60C reverse
primer, bases exchanged are underlined, control restriction
site is in bold). The PCR product was digested with the meth-
ylation-sensitive enzyme DpnI for 1 h at 37 °C. Products
were purified (Qiagen, Hilden, Germany) and transformed
into XL1-blue cells by electroporation; cells were plated on
LB agar containing ampicillin. Plasmid Mini Preparation
(Qiagen) was performed and the obtained plasmids were
digested with the appropriate control restriction enzymes
(PstI and NsiI) for 2 h at 37 °C to screen for plasmids
with the correct mutation. The correctness of the restriction-
positive plasmids was finally confirmed by sequencing.
Protein expression
Plasmids were transformed into E. coli strain BL21(DE3)
via electroporation, and grown overnight in 100 mL of LB
medium containing 100 lgÆmL
)1
ampicillin. Large-scale
expression cultures (12 · 600 mL) were inoculated with
I. Magler et al. Molecular metamorphosis in allergens
FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS 2607
2 mL of preculture. The cells were grown at 37 °Ctoa
D

600 nm
of 0.8, or in the case of Phl p 7 to a D
600 nm
of 0.4,
when protein expression was induced by adding 1 mm iso-
propyl thio-b-d-galactoside. Cells were harvested 4 h after
induction by centrifugation (5000 g for 10 min.), resus-
pended in buffer A (50 mm NaH
2
PO
4
,10mm Tris, 150 mm
NaCl, 10 mm imidazole, pH 8.0), and sonicated (Sonicator,
Bandelin Sonopuls).
Purification
Immobilized metal affinity chromatography
Recombinant protein lysate carrying an N-terminal His
6
-
tag was purified using Ni
2+
–nitrilotriacetic acid resin
(Qiagen) [56]. Bound protein was washed twice (50 mm
NaH
2
PO
4
, pH 8.0, 300 mm NaCl, 20 mm imidazole). The
target protein was eluted with a highly concentrated imid-
azole buffer (50 mm NaH

2
PO
4
, pH 8.0, 300 mm NaCl,
250 mm imidazole).
TEV protease and target protein were used in a molar
ratio of 1 : 5 to remove the N-terminal His
6
-tag. The reac-
tion was carried out for 12–24 h at 4 °C; incompletely
digested target protein was separated by further chromato-
graphy on an Ni
2+
–nitrilotriacetic acid column.
Size exclusion chromatography
Prior to gel filtration chromatography, the protein was con-
centrated routinely to at least 10 mgÆmL
)1
using a 5 kDa
molecular mass cutoff (Centricon, Amicon-Ultra; Millipore,
Bedford, MA, USA). Samples were incubated at different
temperatures (4, 20, 75, and 95 °C) for defined time periods
(hours to days to weeks) and in the presence of different
detergents and buffer conditions (e.g. SDS, 0.05% final con-
centration; high salt, 500 mm NaCl; dithiothreitol, 5 mm
final concentration). For analysis, samples were loaded onto
a Superdex 75 10 ⁄ 300 GL (GE Healthcare, Munich, Ger-
many) column, using an A
¨
KTA FPLC system. If the sample

was pretreated with dithiothreitol, it was separated by using
a desalting column (illustra NAP-columns, GE Healthcare)
prior to the gel filtration run. Typical buffer conditions for
Bet v 4 samples were either 25 mm Tris (pH 7.5) and
50 mm NaCl, or 25 mm Hepes (pH 7.5) and 50 m m NaCl,
or 20 mm Mops (pH 7.5) and 10 mm CaSO
4
; for Phl p 7,
the running buffer contained 40 mm NH
4
Cl and 20 mm Tris
(pH 4.7). The observed retention times (corresponding to
distinct oligomeric states) were identical with sample
concentrations ranging from 1 to $ 80 mgÆmL
)1
.
Circular Dichroism
CD spectra of Bet v 4 were recorded in 4 mm Mops
(pH 7.5) and 2 mm CaSO
4
with a Jasco J-815 spectropola-
rimeter (Jasco, Tokyo, Japan), equipped with a Neslab
RTE-111M temperature control system (Thermo Fischer
Scientific, Waltham, MA, USA). The resulting curves were
baseline-corrected and presented as mean residue molar
elipticity [Q]
MRW
at a given wavelength. Protein concentra-
tions (typically at 20 lgÆmL
)1

) were detected using UV light
at 280 nm.
In order to have well-defined oligomerization states of
the individual protein samples, we employed gel filtration
as a preparative step, and selected fractions according
to their retention volumes. Reduction of the disulfide bond
in Bet v 4-K25C ⁄ F60C was accomplished with 5 mm
tris(2-carboxyethyl)phosphine.
Acknowledgements
We thank P. Briza for technical help with mass spec-
trometry, U. Brendel for help with preparing a figure,
D. Lamba for discussions, and the FWF (project
W_01213) for financial support.
References
1 Sicherer SH & Sampson HA (2007) Peanut allergy:
emerging concepts and approaches for an apparent
epidemic. J Allergy Clin Immunol 120, 491–503.
2 Warner JO (2007) Anaphylaxis; the latest allergy
epidemic. Pediatr Allergy Immunol 18, 1–2.
3 Kemp AS, Mullins RJ & Weiner JM (2006) The allergy
epidemic: what is the Australian response? Med J Aust
185, 226–227.
4 Garn H & Renz H (2007) Epidemiological and immu-
nological evidence for the hygiene hypothesis. Immuno-
biology 212, 441–452.
5 Romagnani S (2004) The increased prevalence of allergy
and the hygiene hypothesis: missing immune deviation,
reduced immune suppression, or both? Immunology 112,
352–363.
6 Holgate ST & Polosa R (2008) Treatment strategies for

allergy and asthma. Nat Rev Immunol 8, 218–230.
7 Radauer C & Breiteneder H (2006) Pollen allergens are
restricted to few protein families and show distinct pat-
terns of species distribution. J Allergy Clin Immunol
117, 141–147.
8 Holgate ST & C MK (1993) Allergy . Gower Medical
Publishing, New York.
9 Akdis CA (2006) Allergy and hypersensitivity: mecha-
nisms of allergic disease. Curr Opin Immunol 18, 718–726.
10 Ghaemmaghami AM, Gough L, Sewell HF & Shakib F
(2002) The proteolytic activity of the major dust mite
allergen Der p 1 conditions dendritic cells to produce less
interleukin-12: allergen-induced Th2 bias determined at
the dendritic cell level. Clin Exp Allergy 32, 1468–1475.
11 Ghaemmaghami AM, Robins A, Gough L, Sewell HF
& Shakib F (2001) Human T cell subset commitment
Molecular metamorphosis in allergens I. Magler et al.
2608 FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS
determined by the intrinsic property of antigen: the
proteolytic activity of the major mite allergen Der p 1
conditions T cells to produce more IL-4 and less
IFN-gamma. Eur J Immunol 31, 1211–1216.
12 Schulz O, Sewell HF & Shakib F (1998) Proteolytic cleav-
age of CD25, the alpha subunit of the human T cell inter-
leukin 2 receptor, by Der p 1, a major mite allergen with
cysteine protease activity. J Exp Med 187, 271–275.
13 Seong SY & Matzinger P (2004) Hydrophobicity: an
ancient damage-associated molecular pattern that initi-
ates innate immune responses. Nat Rev Immunol 4,
469–478.

14 Furmonaviciene R, Sutton BJ, Glaser F, Laughton CA,
Jones N, Sewell HF & Shakib F (2005) An attempt to
define allergen-specific molecular surface features: a bio-
informatic approach. Bioinformatics 21, 4201–4204.
15 Deslee G, Charbonnier AS, Hammad H, Angyalosi G,
Tillie-Leblond I, Mantovani A, Tonnel AB & Pestel J
(2002) Involvement of the mannose receptor in the
uptake of Der p 1, a major mite allergen, by human
dendritic cells. J Allergy Clin Immunol 110, 763–770.
16 Shakib F, Ghaemmaghami AM & Sewell HF (2008)
The molecular basis of allergenicity. Trends Immunol
29, 633–642.
17 Chruszcz M, Chapman MD, Vailes LD, Stura EA,
Saint-Remy JM, Minor W & Pomes A (2009) Crystal
structures of mite allergens Der f 1 and Der p 1 reveal
differences in surface-exposed residues that may influ-
ence antibody binding. J Mol Biol 386, 520–530.
18 Jin T, Guo F, Chen YW, Howard A & Zhang YZ
(2009) Crystal structure of Ara h 3, a major allergen in
peanut. Mol Immunol 46, 1796–1804.
19 Schweimer K, Petersen A, Suck R, Becker WM, Rosch
P & Matecko I (2008) Solution structure of Phl p 3, a
major allergen from timothy grass pollen. Biol Chem
389, 919–923.
20 Li M, Gustchina A, Alexandratos J, Wlodawer A,
Wunschmann S, Kepley CL, Chapman MD & Pomes A
(2008) Crystal structure of a dimerized cockroach aller-
gen Bla g 2 complexed with a monoclonal antibody.
J Biol Chem 283, 22806–22814.
21 Niemi M, Jylha S, Laukkanen ML, Soderlund H,

Makinen-Kiljunen S, Kallio JM, Hakulinen N, Haah-
tela T, Takkinen K & Rouvinen J (2007) Molecular
interactions between a recombinant IgE antibody and
the beta-lactoglobulin allergen. Structure 15, 1413–1421.
22 de Halleux S, Stura E, VanderElst L, Carlier V, Jacque-
min M & Saint-Remy JM (2006) Three-dimensional
structure and IgE-binding properties of mature fully
active Der p 1, a clinically relevant major allergen.
J Allergy Clin Immunol 117, 571–576.
23 Schirmer T, Hoffimann-Sommergrube K, Susani M,
Breiteneder H & Markovic-Housley Z (2005) Crystal
structure of the major celery allergen Api g 1: molecular
analysis of cross-reactivity. J Mol Biol 351, 1101–1109.
24 Gustchina A, Li M, Wunschmann S, Chapman MD,
Pomes A & Wlodawer A (2005) Crystal structure of
cockroach allergen Bla g 2, an unusual zinc binding
aspartic protease with a novel mode of self-inhibition.
J Mol Biol 348, 433–444.
25 Ichikawa S, Takai T, Inoue T, Yuuki T, Okumura Y,
Ogura K, Inagaki F & Hatanaka H (2005) NMR study
on the major mite allergen Der f 2: its refined tertiary
structure, epitopes for monoclonal antibodies and
characteristics shared by ML protein group members.
J Biochem 137, 255–263.
26 Kaiser L, Gronlund H, Sandalova T, Ljunggren HG,
van Hage-Hamsten M, Achour A & Schneider G (2003)
The crystal structure of the major cat allergen Fel d 1,
a member of the secretoglobin family. J Biol Chem 278,
37730–37735.
27 Markovic-Housley Z, Degano M, Lamba D, von

Roepenack-Lahaye E, Clemens S, Susani M, Ferreira
F, Scheiner O & Breiteneder H (2003) Crystal structure
of a hypoallergenic isoform of the major birch pollen
allergen Bet v 1 and its likely biological function as a
plant steroid carrier. J Mol Biol 325, 123–133.
28 Holm J, Baerentzen G, Gajhede M, Ipsen H, Larsen
JN, Lowenstein H, Wissenbach M & Spangfort MD
(2001) Molecular basis of allergic cross-reactivity
between group 1 major allergens from birch and
apple. J Chromatogr B Biomed Sci Appl 756, 307–
313.
29 Markovic-Housley Z, Miglierini G, Soldatova L,
Rizkallah PJ, Muller U & Schirmer T (2000) Crystal
structure of hyaluronidase, a major allergen of bee
venom. Structure 8, 1025–1035.
30 Chan SL, Ong TC, Gao YF, Tiong YS, Wang de Y,
Chew FT & Mok YK (2008) Nuclear magnetic reso-
nance structure and IgE epitopes of Blo t 5, a major
dust mite allergen. J Immunol 181, 2586–2596.
31 Lehmann K, Schweimer K, Reese G, Randow S, Suhr
M, Becker WM, Vieths S & Rosch P (2006) Structure
and stability of 2S albumin-type peanut allergens: impli-
cations for the severity of peanut allergic reactions. Bio-
chem J 395, 463–472.
32 Neudecker P, Nerkamp J, Eisenmann A, Nourse A,
Lauber T, Schweimer K, Lehmann K, Schwarzinger S,
Ferreira F & Rosch P (2004) Solution structure, dynam-
ics, and hydrodynamics of the calcium-bound cross-
reactive birch pollen allergen Bet v 4 reveal a canonical
monomeric two EF-hand assembly with a regulatory

function. J Mol Biol 336, 1141–1157.
33 Gendel SM & Jenkins JA (2006) Allergen sequence
databases. Mol Nutr Food Res 50, 633–637.
34 Jenkins JA, Griffiths-Jones S, Shewry PR, Breiteneder
H & Mills EN (2005) Structural relatedness of plant
food allergens with specific reference to cross-reactive
allergens: an in silico analysis. J Allergy Clin Immunol
115, 163–170.
I. Magler et al. Molecular metamorphosis in allergens
FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS 2609
35 Tokuriki N & Tawfik DS (2009) Protein dynamism and
evolvability. Science 324, 203–207.
36 Tuinstra RL, Peterson FC, Kutlesa S, Elgin ES, Kron
MA & Volkman BF (2008) Interconversion between
two unrelated protein folds in the lymphotactin native
state. Proc Natl Acad Sci USA 105, 5057–5062.
37 Herzberg O & James MN (1985) Common structural
framework of the two Ca2+ ⁄ Mg2+ binding loops of
troponin C and other Ca2+ binding proteins. Biochem-
istry 24, 5298–5302.
38 Herzberg O & James MN (1985) Structure of the cal-
cium regulatory muscle protein troponin-C at 2.8 A res-
olution. Nature 313, 653–659.
39 Chugh J, Chatterjee A, Kumar A, Mishra RK, Mittal
R & Hosur RV (2006) Structural characterization of the
large soluble oligomers of the GTPase effector domain
of dynamin. FEBS J 273, 388–397.
40 Wilk S, Chen WE & Magnusson RP (1998) Properties
of the proteasome activator subunit PA28 alpha and its
des-tyrosyl analog. Arch Biochem Biophys 359, 283–290.

41 Verdino P, Westritschnig K, Valenta R & Keller W
(2002) The cross-reactive calcium-binding pollen aller-
gen, Phl p 7, reveals a novel dimer assembly. EMBO J
21, 5007–5016.
42 Rajashankar K, Bufe A, Weber W, Eschenburg S,
Lindner B & Betzel C (2002) Structure of the functional
domain of the major grass-pollen allergen Phlp 5b. Acta
Crystallogr D Biol Crystallogr 58, 1175–1181.
43 Andersson K & Lidholm J (2003) Characteristics and
immunobiology of grass pollen allergens. Int Arch
Allergy Immunol 130, 87–107.
44 DeWitt AM, Niederberger V, Lehtonen P, Spitzauer S,
Sperr WR, Valent P, Valenta R & Lidholm J (2002)
Molecular and immunological characterization of a
novel timothy grass (Phleum pratense) pollen allergen,
Phl p 11. Clin Exp Allergy 32, 1329–1340.
45 Reese G, Ballmer-Weber BK, Wangorsch A, Randow S
& Vieths S (2007) Allergenicity and antigenicity of wild-
type and mutant, monomeric, and dimeric carrot major
allergen Dau c 1: destruction of conformation, not olig-
omerization, is the roadmap to save allergen vaccines.
J Allergy Clin Immunol 119, 944–951.
46 Gieras A, Focke-Tejkl M, Ball T, Verdino P, Hartl A,
Thalhamer J & Valenta R (2007) Molecular determi-
nants of allergen-induced effector cell degranulation.
J Allergy Clin Immunol 119, 384–390.
47 Das Dores S, Chopin C, Villaume C, Fleurence J &
Gueant JL (2002) A new oligomeric parvalbumin
allergen of Atlantic cod (Gad mI) encoded by a gene
distinct from that of Gad cI. Allergy 57, 79–83.

48 Gregoire C, Tavares GA, Lorenzo HK, Dandeu JP,
David B & Alzari PM (1999) Crystallization and preli-
minary crystallographic analysis of the major horse
allergen Equ c 1. Acta Crystallogr D Biol Crystallogr
55, 880–882.
49 Kristensen AK, Schou C & Roepstorff P (1997)
Determination of isoforms, n-linked glycan structure
and disulfide bond linkages of the major cat allergen
fel d1 by a mass spectrometric approach. Biol Chem
378, 899–908.
50 Kennedy MW, Brass A, McCruden AB, Price NC,
Kelly SM & Cooper A (1995) Aba-1 allergen of the
parasitic nematode ascaris suum – fatty acid and reti-
noid binding function and structural characterization.
Biochemistry 34, 6700–6710.
51 Scho
¨
ll I, Kalkura N, Shedziankova Y, Bergmann A,
Verdino P, Knittelfelder R, Kopp T, Hantusch B,
Betzel C, Dierks K et al. (2005) Dimerization of the
major birch pollen allergen Bet v 1 is important for its
in vivo IgE-cross-linking potential in mice. J Immunol
175, 6645–6650.
52 Engel E, Richter K, Obermeyer G, Briza P, Kungl AJ,
Simon B, Auer M, Ebner C, Rheinberger HJ,
Breitenbach M et al. (1997) Immunological and
biological properties of Bet v 4, a novel birch pollen
allergen with two EF-hand calcium-binding domains.
J Biol Chem 272, 28630–28637.
53 Niederberger V, Hayek B, Vrtala S, Laffer S, Twardosz

A, Vangelista L, Sperr WR, Valent P, Rumpold H,
Kraft D et al. (1999) Calcium-dependent immunoglobu-
lin E recognition of the apo- and calcium-bound form
of a cross-reactive two EF-hand timothy grass pollen
allergen, Phl p 7. FASEB J 13, 843–856.
54 Sheffield P, Garrard S & Derewenda Z (1999) Overcom-
ing expression and purification problems of RhoGDI
using a family of ‘parallel’ expression vectors. Protein
Expr Purif 15, 34–39.
55 Hogrefe HH, Cline J, Youngblood GL & Allen RM
(2002) Creating randomized amino acid libraries with
the QuikChange Multi Site-Directed Mutagenesis Kit.
BioTechniques 33, 1158–1164.
Supporting information
The following supplementary material is available:
Fig. S1. Mean residual weight ellipticity of Bet v 4
variants at 208 and 222 nm.
Fig. S2. The presence of 10 mm EDTA does not inter-
fere with Bet v 4 folding.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
Molecular metamorphosis in allergens I. Magler et al.
2610 FEBS Journal 277 (2010) 2598–2610 ª 2010 The Authors Journal compilation ª 2010 FEBS