Báo cáo khoa học: Odorant binding protein has the biochemical properties of a scavenger for 4-hydroxy-2-nonenal in mammalian nasal mucosa doc

Odorant binding protein has the biochemical properties
of a scavenger for 4-hydroxy-2-nonenal in mammalian
nasal mucosa
Stefano Grolli, Elisa Merli, Virna Conti, Erika Scaltriti and Roberto Ramoni
Dipartimento di Produzioni Animali, Biotecnologie Veterinarie, Qualita
`
e Sicurezza degli Alimenti, Universita
`
degli Studi di Parma, Italy
Reactive oxygen species (ROS) are short-lived radical
intermediates generated as a consequence of oxidative
metabolism. They can react with virtually all classes of
biological molecules and are responsible for most of
the cellular damage caused by oxidative stress. In the
case of reactions with membrane polyunsaturated fatty
acids (PUFA), ROS initiate a lipid peroxidation pro-
cess that gives rise to a large number of toxic low
molecular mass aldehydes as end products, including
the 4-hydroxyalkenals [1,2]. These molecules may be
responsible for significant loss of biological activity in
proteins and nucleic acids by reacting in a Michael-
type addition with the nucleophilic groups (-SH, -NH
2
and imidazole) of amino acids and nucleotides [1,2].
Inactivation of 4-hydroxyalkenals in vivo is achieved
by several enzymatic systems [3], with a predominant
role played by glutathione S-transferases (GSTs) which
catalyse a Michael addition of glutathione to the alde-
hyde double bond [4]. Furthermore, 4-hydroxyalkenal
cytotoxicity is possible if it is exported via the
Keywords

4-hydroxy-2-nonenal; lipocalins; odorant
binding protein; oxidative stress; reactive
oxygen species
Correspondence
R. Ramoni, Dipartimento di Produzioni
Animali, Biotecnologie Veterinarie, Qualita
`
e
Sicurezza degli Alimenti, Universita
`
di
Parma, Via del Taglio 8, 43100 Parma, Italy
Fax: +39 052 190 2770
Tel: +39 052 190 2767
E-mail:
(Received 7 June 2006, revised 18 August
2006, accepted 21 September 2006)
doi:10.1111/j.1742-4658.2006.05510.x
Odorant binding proteins (OBP) are soluble lipocalins produced in large
amounts in the nasal mucosa of several mammalian species. Although
OBPs can bind a large variety of odorous compounds, direct and exclusive
involvement of these proteins in olfactory perception has not been clearly
demonstrated. This study investigated the binding properties and chemical
resistance of OBP to the chemically reactive lipid peroxidation end-product
4-hydroxy-2-nonenal (HNE), in an attempt to establish a functional rela-
tionship between this protein and the molecular mechanisms combating
free radical cellular damage. Experiments were carried out on recombinant
porcine and bovine OBPs and results showed that both forms were able to
bind HNE with affinities comparable with those of typical OBP ligands
(K

d
¼ 4.9 and 9.0 lm for porcine and bovine OBP, respectively). Further-
more, OBP functionality, as determined by measuring the binding of the
fluorescent ligand 1-aminoanthracene, was partially lost only when incuba-
ting HNE levels and exposure time to HNE exceeded physiological values
in nasal mucosa. Finally, preliminary experiments in a simplified model
resembling nasal epithelium showed that extracellular OBP can preserve
the viability of an epithelial cell line derived from bovine turbinates
exposed to toxic amounts of the aldehyde. These results suggest that OBP,
which is expressed at millimolar levels, might reduce HNE toxicity by
removing from the nasal mucus a significant fraction of the aldehyde that
is produced as a consequence of direct exposure to the oxygen present in
inhaled air.
Abbreviations
AMA, 1-aminoanthracene; BT, bovine turbinate cells; GST, glutathione S-transferase; HNE, 4-hydroxy-2-nonenal; OBP, odorant binding
protein; PUFA, polyunsaturated fatty acids; ROS, reactive oxygen species; TTBS, 20 m
M Tris ⁄ HCl buffer, pH 7.8 containing 150 mM NaCl
and 0.01% (W ⁄ V) Tween 20.
FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS 5131
bloodstream to molecular targets distributed virtually
anywhere in the organism [1,4]. This is a consequence
of the small dimensions and moderate hydrophilicity
of these compounds, which, in contrast to their lipid
precursors, can be solubilized at millimolar levels in
the aqueous matrix of biological fluids. Therefore, a
protein scavenger that could trap, and eventually deli-
ver, 4-hydroxyalkenals to appropriate degradative
pathways, might aid other inactivating mechanisms
and prevent chemical modification by these molecules
in tissues where large-scale lipid peroxidation occurs.

The nasal mucosa is constantly exposed to the high
oxygen levels present in inhaled air and it has recently
been proposed that toxic aldehydes derived from lipid
peroxidation might be scavenged by odorant binding
proteins (OBPs) [5]. OBPs are soluble proteins present
in large amounts (mm levels) on the surface of the
nasal mucosa [6]. They can bind a broad spectrum
of odorous and nonodorous compounds with good
affinities (K
d
in the lm range), including some toxic 8–
11 carbon aldehydes [7–9] derived from the peroxida-
tion of PUFA. OBPs belong to the lipocalins, a family
of structurally related soluble proteins that bind differ-
ent types of small hydrophobic molecules [10,11]. In
general, OBPs are monomeric proteins with a molecu-
lar mass of 19 kDa. A nine-strand beta-barrel defines
the ligand-binding site, connected by a short linker
(hinge sequence) to a C-terminal a helix [12,13] of
unknown function. Dimeric OBPs have also been des-
cribed, although less frequently. For example, bovine
OBP is peculiar in that it is a dimer with domain
swapping [14,15]. There is currently a lack of informa-
tion on the in vivo binding properties of OBP, the low
binding specificity, however, suggests several hypothe-
ses for its role as a versatile carrier ⁄ scavenger involved
in different molecular mechanisms within the nasal
mucosa:
(a) OBP might be involved in olfactory perireceptor
events behaving either as carrier of odorous compounds

to their receptors or as a scavenger of excess odours;
(b) OBP, at least in ruminants, might have a protective
prophylactic role towards parasitosis and infectious
diseases carried by insects;
(c) OBP might protect against oxidative stress by
removing toxic compounds locally produced by lipid
peroxidation or inhalation.
Experimental evidence in favour of the first hypothe-
sis include the binding capacity of OBPs for odorous
compounds and the abundance of OBPs in nasal tissue
[6], and the recent report of an in vitro interaction
between porcine OBP and an olfactory receptor [16].
The second hypothesis is supported by the identifica-
tion of the natural ligand of bovine OBP, the insect
attractant 1-octen-3-ol, a component of bovine breath
that is produced by rumenal microflora [17]. OBP may
render animals less attractive for insects by trapping
1-octen-3-ol from expired breath as it passes through
nasal turbinates, resulting in a general decrease in
bites. This, in turn, would cause a reduction in vector-
mediated parasitosis and infectious diseases.
Evaluation of the third hypothesis, the molecular
basis of which is described above, is the aim of this
study. Here, we report the binding properties, chemical
modification and chemical resistance of recombinant
bovine and porcine OBPs with respect to 4-hydroxy-
2-nonenal (HNE), the most abundant and extensively
characterized toxic 4-hydroxyalkenal derived from per-
oxidation of x-6 PUFA [1,2]. Our results show that
OBP can bind HNE in a reversible equilibrium and

retain a relevant fraction of its binding capacity after
chemical modifications induced by the aldehyde. Fur-
thermore, preliminary experiments indicate that extra-
cellular OBP can prevent HNE-induced cytotoxicity in
an epithelial cell line derived from bovine nasal turbin-
ates. Taken together, the data suggest that, in vivo,
OBP may trap compounds derived from peroxidation
of PUFA and lead them from the mucus present on
nasal epithelia to the digestive tract for their chemical
inactivation.
Results
Protein purification and functional
characterization
SDS ⁄ PAGE of the purified forms of recombinant por-
cine and bovine OBP gave two single bands at the
expected molecular masses. Binding capacity was tes-
ted using the fluorescent ligand 1-aminoanthracene
(AMA). Hyperbolic titration curves (Fig. 1A,B) giving
K
d
values of 1.5 and 0.5 lm, and saturation levels of
0.9 and 1.83 for porcine and bovine OBP, respectively
were in agreement with functional preparations of
native and recombinant OBP [18].
Direct binding test to detect HNE–OBP reversible
binding complexes
The experiment, showing the formation of reversible
HNE–OBP binding complexes, was based on the
assumption that affinity (dissociation constants in the
lm range) and binding stoichiometry (1 mole of lig-

and ⁄ OBP equivalent) for HNE are similar to those of
the typical OBP ligands. In addition, the spectrophoto-
metric binding assay used here allowed us to discrimin-
ate, in the same test, between HNE molecules forming
OBP as a scavenger for HNE S. Grolli et al.
5132 FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS
reversible complexes with OBP and those irreversibly
bound as HNE–OBP covalent adducts.
The experiment, in which 90 lm HNE was incuba-
ted with a slight molar excess of OBP, showed that all
the aldehyde was reversibly complexed to the protein-
binding site, independent of incubation time. In fact,
HNE could be quantitatively displaced by a large
excess of undecanal, an OBP ligand whose binding
complexes with the protein have been resolved at the
structural level using X-ray crystallography [5,19].
Undecanal is known to bind within the barrel of OBP,
thus results from the displacement experiments indica-
ted that HNE was bound entirely within the barrel of
OBP.
The immediate and quantitative displacement of
HNE by undecanal indicates that, as expected in the
case of binding equilibria, the formation of reversible
HNE–OBP complexes is faster than the covalent
reaction between the aldehyde and protein nucleophilic
amino acids. The different rates of these two reactions
further indicate (see below) the potential efficacy of
HNE scavenging by OBP, preventing formation of
covalent adducts between the aldehyde and its cellular
targets. In fact, when OBP is in molar excess with

respect to HNE and its concentration is at least one
order of magnitude higher than the dissociation con-
stant of the reversible binding complex, it can be
expected that the amount of free aldehyde might be
negligible.
Determination of the dissociation constant
of HNE–OBP complexes
Affinities for HNE were determined by measuring the
progressive chasing of saturating amounts of AMA
bound to OBP in response to increasing concentrations
[AMA] µ
M
02468
Fluorescence intensity
0
20
40
60
80
100
120
140
160
180
[AMA] µ
M
02468
Fluorescence intensity
0
20

40
60
80
100
120
140
160
180
[HNE] µ
M
0 10203040
Relative fluorescence intensity
0.0
0.2
0.4
0.6
0.8
1.0
1.2
[HNE] µ
M
0204060
Relative fluorescence intensity
0.0
0.2
0.4
0.6
0.8
1.0
1.2

A
B
C
D
Fig. 1. Binding curves of the fluorescent ligand AMA to porcine (A) and bovine (B) OBP. The curves report the emission fluorescence inten-
sity at 480 nm of the AMA–OBP complex, upon excitation at 380 nm, versus the concentration of AMA. Competitions between HNE and
AMA are shown for porcine (C) and bovine (D) OBP. Protein samples were incubated with a fixed saturating amount of AMA and increasing
HNE. Each point on the y-axis shows the concentration of AMA still bound per OBP monomer relative to the initial value, on a scale of 0–1,
versus the concentration of HNE.
S. Grolli et al. OBP as a scavenger for HNE
FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS 5133
of the aldehyde (Fig. 1C,D). Competition curves,
matching a two-parameter hyperbolic decay model,
gave apparent dissociation constants of 11.0 and
21.0 lm for porcine and bovine OBP, respectively.
These values, given in Eqn (1) (see Experimental pro-
cedures), resulted in true K
d
values for HNE of 4.9
and 9.0 lm, which are comparable with those of other
typical OBP ligands determined under the same experi-
mental conditions [19]. Competitive titrations with
heat-denatured OBPs (overnight incubation at 90 °C)
were performed as negative controls, and showed that
formation of the binding complex between HNE and
OBP is strictly dependent on the structural integrity of
the protein (not shown).
Western blotting and ligand-binding assay
of HNE-modified OBP
The immunoblot analysis reported in Fig. 2A shows

the progressive formation of HNE–OBP adducts after
incubation (5 h at 37 °C) of a fixed amount of OBP
(26 and 52 lm for bovine and porcine OBP, respect-
ively) with increasing aldehyde concentrations (0.1–
2.5 mm). Molar HNE levels exceeded those of OBP
and covered the range reported in the literature for
this aldehyde in the case of lipid peroxidation in vivo
[1]. The labelling pattern showed that bovine OBP can
be chemically modified by lower amounts of HNE
than the porcine form. This is probably because of the
higher number of possibly HNE-reacting amino acid
side chains (His and Lys) in bovine OBP (13.2% of
the amino acidic residues) compared with the porcine
form (8.3%) [12], and because of their spatial arrange-
ment in the 3D structure of the proteins (see below).
The functionality of these HNE chemically modified
OBPs was determined by measuring residual binding
for the fluorescent ligand AMA. The plots in
Fig. 2B,C show that, even in the case of very high
incubating [HNE] ⁄ [OBP] molar ratios (50 : 1), both
forms lost only a fraction of their initial AMA-binding
capability.
Titration of HNE covalent adducts in OBP
These titration curves showed that formation of HNE–
OBP adducts was definitely time-dependent, passing
from 20 min, i.e. the time necessary for clearance of
substances dissolved in nasal mucus [20], to the arbi-
trary value of 16 h (Fig. 3A,B). This is particularly evi-
dent for porcine OBP, but similar behaviour was seen
[HNE] m

M
00.1 2.51.00.5 0 0.1 2.51.00.5
[HNE] μ
M
/ [OBP] μ
M
× 2
0 2 10 20 50 0 2 10 20 50
porcine OBP
bovine OBP
A
[HNE] µM / [pOBP] µM
0 102030405060
Residual AMA binding
0.0
0.2
0.4
0.6
0.8
1.0
1.2
B
[HNE] µ
M
/ [pOBP] µ
M
× 2
0 102030405060
Residual AMA binding
0.0

0.2
0.4
0.6
0.8
1.0
1.2
C
Fig. 2. (A) Immunoblotting of porcine and bovine OBP HNE-reacting groups. The immunostaining of OBP samples (1 mgÆmL
)1
), treated with
increasing amounts of HNE (0.1–2.5 m
M), was visualized after incubation with a rabbit serum raised against HNE–protein adducts. Ligand-
binding tests showing the functionality of the same HNE-treated porcine (B) and bovine (C) OBPs as in the immunostaining. The plots show
the residual AMA-binding capacity versus the ratios between incubating HNE and OBP equivalents. Single data points on the y-axes are
reported relative to the functionalities of OBP samples that had not been treated with HNE.
OBP as a scavenger for HNE S. Grolli et al.
5134 FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS
in the case of the bovine form. The divergence between
the different incubation times, occurring immediately
for porcine OBP and at molar [HNE] ⁄ [OBP] ratios
> 10 for bovine OBP, led us to hypothesize that,
in vivo, only a small fraction of the putative HNE-
reacting protein residues might be present as HNE-
covalent adducts.
In the case of bovine OBP, titrations did not reach
the putative end-point of 21 residues assumed from the
amino acid sequence. Indeed, titration could be almost
completed (9 HNE-covalent adducts ⁄ 13 putative HNE
targets) only with porcine OBP after 16 h in the pres-
ence of a very large excess of aldheyde. The curves,

which were steeper for [HNE] ⁄ [OBP] molar ratios
< 20, indicated facilitated reactivity of groups of
nucleophilic amino acids probably located on the
protein surfaces, which, on the basis of their 3D struc-
tures, are 5 and 11 for porcine and bovine OBP,
respectively [13,14].
The number of HNE-modified residues correlated
with a progressive decrease in protein functionality, as
shown in Fig. 3C,D which shows the residual AMA-
binding capacity versus [HNE] ⁄ [OBP] molar ratios.
Loss of protein functionality was more evident for the
16-h incubations, in which at least 50% of residual lig-
and-binding capacity was maintained at the highest
[HNE] ⁄ [OBP] values. In the case of 20-min incuba-
tions, HNE inactivation was even less efficient and the
fraction of functional protein remained > 70%. The
data suggest that HNE-covalent adducts do not com-
pletely disrupt the 3D arrangement of the native
proteins. Furthermore, they indicate the relevant
HNE reacting groupsResidual AMA binding
HNE reacting groups
Residual AMA binding
[HNE] µ
M
/ [pOBP] µ
M
[HNE] µ
M
/ [bOBP] µ
M × 2

0 10203040506070
0
2
4
6
8
10
0 1020304050
0
2
4
6
8
10
12
14
[HNE] µ
M
/ [pOBP] µ
M
0 10203040506070
0.0
0.2
0.4
0.6
0.8
1.0
1.2
[HNE] µ
M

/ [bOBP]
µ
M × 2
0 1020304050
0.0
0.2
0.4
0.6
0.8
1.0
1.2
A
B
CD
Fig. 3. (A,B) Spectrophotometric titrations of covalent adducts between HNE and porcine (A) and bovine (B) OBP, after 20 min (open sym-
bols) and 16 h incubation with increasing amounts of the aldheyde. Samples were concentrated using Centricon filters and the number of
HNE–OBP covalent adducts was determined by subtracting the amounts of HNE in the ultrafiltrate from the corresponding levels of incuba-
ting HNE. The number of HNE reacting residues was plotted versus the ratios between incubating HNE and OBP equivalents. Ligand-binding
tests showing the functionality of the same HNE-treated porcine (C) and bovine (D) OBP samples from the spectrophotometric titration. The
plots, showing the residual AMA-binding capacity versus the ratios between incubating HNE and OBP equivalents, are reported as in
Fig. 2B,C.
S. Grolli et al. OBP as a scavenger for HNE
FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS 5135
resistance of OBP to chemical attack by the aldehyde,
especially under experimental conditions in which incu-
bation times are closer to physiological values found
in vivo in nasal mucosa.
Biochemical characterization of OBP exposed to
HNE under conditions simulating oxidative stress
in nasal mucosa

In this experiment, we incubated HNE and OBP under
experimental conditions simulating those of the nasal
mucosa. OBP concentrations were those presumed for
this tissue (0.5 and 1 mm for bovine and porcine OBP,
respectively) [6]; HNE concentrations (2 mm) were in
the range reported for acute oxidative stress in vivo.
Incubation was carried out at 20 °C for 20 min, i.e.
the time necessary for the clearance of substances dis-
solved in nasal mucus [20]. Under these conditions,
formation of HNE–protein adducts was negligible and
changes in binding properties were not detected. The
results confirm those of the previous experiment in
which, at short incubation times and when the molar
ratio of HNE and OBP was < 5, the aldehyde did not
form a relevant number of covalent adducts with the
protein (Fig. 3A,B).
Evaluation of the protective role of OBP against
HNE cytotoxicity in a simplified model simulating
the nasal epithelium
The aim of this experiment was a preliminary evalua-
tion of the protective role of OBP against chemical
aggression by HNE on living cells. We incubated the
epithelial cell line BT, derived from bovine nasal tur-
binates [21], with a cytotoxic amount (20 lm) of HNE
in the absence and presence of a molar excess of
bovine OBP (40 lm). The protein was maintained in
the extracellular environment by immobilization to a
Ni-NTA affinity chromatography matrix, in order to
simulate the conditions in the nasal mucosa. We
employed a form of bOBP with an N-terminal 6· His-

tag (i-bOBP). The structural and functional properties
of this form (in terms of aggregation state and binding
with different ligands including AMA, HNE and unde-
canal) are analogous to those of the native form (data
not shown).
The cytotoxic effects of 20 lm HNE on BT cells
were clearly shown within few hours. Cells showed
changes in morphology and viability. Essentially, cells
treated with HNE alone became detached from culture
dishes or showed marked changes in cellular morphol-
ogy (i.e. shrinking and rounding) evident under phase-
contrast microscopy (Fig. 4B). By contrast, no signs of
cytotoxicity were present in control cultures or cells
treated with HNE in the presence of a molar excess of
i-bOBP (Fig. 4A,C). Trypan Blue exclusion test con-
firmed that cell viability was preserved in the presence
of i-bOBP (> 95% vital cells versus < 5% vital cells
in wells treated with HNE alone; data not shown).
Discussion
We investigated the hypothesis that the lipocalin odor-
ant binding protein might represent, in nasal tissue, a
carrier and ⁄ or a scavenger for toxic aldehydes derived
from the peroxidation of unsaturated fatty acids.
This study was realized through characterization of
the binding properties and chemical resistance of OBP
to HNE, an alken-aldehyde involved in the pathogene-
sis of several acute and chronic diseases. The hypothe-
sis that OBP might actually protect living cells against
HNE cytotoxicity was also evaluated. To verify whe-
ther the property of scavenging toxic aldehydes may

have general significance in mammals, experiments
were carried out using OBPs from bovine and porcine
species, two proteins with relevant differences in amino
acid sequence and 3D structure.
The experimental study was divided into four parts:
(a) the detection and characterization of the reversible
binding complex between OBP and HNE; (b) investi-
gation of the chemical modifications of OBP as a
result of incubation with large excesses of HNE and
evaluation of their effects on the binding properties of
the protein; (c) determination of OBP functionality
after incubation with HNE under experimental condi-
tions simulating the oxidative stress environment of
nasal mucosa; and (d) determination of the protective
role of extracellular OBP against HNE toxicity in
living cells.
Experimental results from the first two parts clearly
showed that the biochemical properties of OBP match
those of an efficient carrier ⁄ scavenger for HNE, and
eventually for other alken-aldehydes that the protein
could bind.
First, direct and competitive binding tests showed
that the formation of reversible OBP–HNE complexes,
whose K
d
values are in the lm range, match those of
most OBP ligands. This similarity suggests that the
binding of HNE might occur with the same features
determined by X-ray crystallography for other OBP
ligands of similar structure [5,19]. Hence HNE, like

undecanal and 1-octen-3-ol, must likely adapt its con-
formation to the shape of the binding site, establishing
very few specific interactions with the amino acid side
chains, which do not undergo significant motion. Fur-
ther structural studies of the OBP–HNE complexes in
OBP as a scavenger for HNE S. Grolli et al.
5136 FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS
the crystal and in solution will elucidate detailed struc-
tural and functional aspects of the binding process.
The reactivity of HNE towards OBP was investi-
gated using two different methodologies: western
blotting, which showed progressive formation of
HNE–OBP adducts after incubation with increasing
amounts of the aldehyde, and a spectrophotometric ti-
tration that allowed quantitative determination of the
HNE reactive amino acids in OBP. Results from these
experiments indicated that the formation of HNE–
OBP covalent adducts is largely dependent on the
incubation time and the molar ratio between the alde-
hyde and protein. In the case of the titrations, it could
be shown quantitatively that the number of modified
residues increased when incubation of the HNE–OBP
mixtures was increased from 20 min to 16 h, and when
the aldehyde was present in high molar excess (> 5).
A comparison between the molar [HNE] ⁄ [OBP] ratios
in these experiments and those presumed to be found
in nasal mucosa (< 5) [1,6] suggests that in this tissue
4 hours 24 hours
controlHNEHNE + i-bOBP
A

A
B
B
C
C
Fig. 4. Prevention of HNE-induced cytotoxicity by bOBP in BT cells. Cells were cultivated in a Costar Transwell plate (lower compartment)
for 72 h and then incubated with control medium (A), 20 l
M HNE (B) or 20 lM HNE in the presence of 40 lM i-bOBP (C), added to the upper
compartment. Phase-contrast photomicrography images, taken after 4 and 24 h, show that i-bOBP prevents HNE-induced cytotoxicity.
Original magnification: ·100.
S. Grolli et al. OBP as a scavenger for HNE
FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS 5137
OBP might be substantially free of chemical HNE
modifications. This is confirmed by MS mass determi-
nations of native porcine and bovine OBP samples [18]
(R Ramoni, unpublished results), whose values invaria-
bly matched those expected on the basis of the amino
acid sequences.
The biochemical characterization resulting from the
first two parts of this study indicates that this type of
protein might represent a general carrier ⁄ scavenger for
HNE. However, it must be considered that most of
these experiments were conducted under conditions
that, in terms of OBP and HNE levels, cannot be real-
istically compared with those present in vivo in nasal
mucosa. Hence, to further evaluate the credibility of
the ‘scavenger’ hypothesis, we designed the last two
parts of this study to simulate the physiological condi-
tions in this tissue. We evaluated the functional prop-
erties of the protein and in particular the chemical

resistance and binding capacity with respect to HNE.
The incubation time of 20 min is the time required for
the clearance of a molecule from the nasal mucosa
[20], while incubating OBP and HNE levels, both in
the mm range, were plausible physiological values [1,6].
This exposure to HNE did not give appreciable chem-
ical modification of OBP, further confirming that
OPB ⁄ HNE binding in vivo should be reversible and
should leave OBPs binding properties unmodified. We
also evaluated the capacity of OBP to protect living
cells from HNE chemical aggression. To better simu-
late the nasal mucosa environment, an epithelial cell
line derived from bovine nasal turbinates was used.
Incubating OBP was kept in the extracellular environ-
ment to simulate its tissue localization in the mucus
that covers the nasal epithelia. The experimental
results, although preliminary and not quantitative,
clearly show that OBP can protect living cells from
cytotoxic HNE.
These results can be considered as a further indica-
tion of the role of this protein as a plausible scavenger
for HNE in vivo. In particular, because the amount of
OBP in nasal tissue is at least two orders of magnitude
higher with respect to the dissociation constant of the
binding complex with HNE (and eventually other toxic
molecules with similar structure), it is possible that a
large fraction of the aldehyde produced from lipid
peroxidation might be trapped by this protein. Nasal
mucus, which flows from the nasal turbinates to the
pharynx with a clearance time of $ 20 min, might

drive the OBP–HNE complexes into the first tract of
the digestive system where they are inactivated. Hence,
this mechanism might be considered as an extracellular
counterpart of the chemical inactivation of HNE that
occurs intracellularly via GST and other enzymes that
are abundantly expressed in nasal mucosa [22]. Fur-
thermore, it must be stressed that OBP is a secreted
protein, and as such, might have particular relevance
in the protection of olfactory receptors, which are pro-
teins that play a crucial role in the behaviour and sur-
vival of most animal species. In fact, their extracellular
domains, which bear the binding sites for odorous
compounds, are completely embedded in nasal mucus
[23,24] and consequently exposed to the reactive mole-
cules rising from oxidative stress.
Importantly, this study shows that the OBPs from
two animal species, with 42% sequence homology and
relevant differences in 3D structure, display similar
binding modalities and chemical resistance to HNE.
This indicates that the protective role with respect to
HNE and other alken-aldehydes might be hypothesized
for most OBPs and thus might have a general signifi-
cance in mammals.
The same role that we hypothesize here for OBP has
been also proposed for human tear lipocalin (lcn-1), a
protein belonging to the same family. It is produced by
the lachrymal and lingual salivary glands, and has been
found to be expressed by several other secretory tissues
such as prostate, mucosal glands of the tracheobronchi-
al tree, nasal mucosa and sweat glands [25]. Human

tear lipocalin has significant sequence homology with
the human forms of OBP and, at least in humans, par-
tially shares a similar tissue distribution [26]. This pro-
tein can bind HNE, but compared with OBP, has a
binding spectrum that is particularly oriented to com-
pounds like 8-isoprostane or 7-b-hydroxycholesterol,
with higher molecular masses and hydrophobicity.
Despite the different binding spectrum, however, OBP
and lcn-1 may cooperate as complementary scavengers
in the removal of toxic compounds derived from lipid
peroxidation in tissues where they are both present.
There are several other lipocalins that bind small
hydrophobic molecules, but their physiological role has
not yet been univocally and unambiguously established
[11]. Further investigations, based on binding tests with
molecules derived from lipid peroxidation, might give
new insights into the ligand specificity and binding
affinity of these proteins and, in turn, their function in
lipocalin-driven molecular mechanisms for the removal
and inactivation of toxic compounds produced as a
consequence of oxidative stress in different tissues.
Experimental procedures
Materials
HNE (27 mm), in hexane, was purchased from Alexis Bio-
chemicals (Lausanne, Switzerland) and stored at )80 °C.
OBP as a scavenger for HNE S. Grolli et al.
5138 FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS
The concentration was controlled on the spectrophotometer
(e
223

¼ 13750 m
)1
Æcm
)1
) after dilution in water.
AMA was from Sigma Aldrich (Milan, Italy). All other
reagents, purchased from different companies, were of ana-
lytical grade.
Rabbit serum raised against HNE–protein adducts was
from Alpha Diagnostics International (San Antonio, TX),
poly(vinylidene difluoride) was from Millipore (Milan,
Italy) and the reagents for electrophoresis and western blot-
ting were from Sigma Aldrich. Media and reagents for cell
cultures were from Gibco (Milan, Italy). Costar Transwell
culture plates were from Corning (Schiphol-Rijk, the
Netherlands).
OBP purification and functionality test with AMA
Recombinant forms of porcine and bovine OBPs were
purified from BL21-DE3 Escherichia coli strains trans-
formed with the expression vector pT7-7 containing the
different OBP cDNAs as previously reported [18]. The
purity of each OBP preparation was determined by
SDS ⁄ PAGE and protein concentrations were calculated
based on the absorbance values at 280 nm (13 000 and
48 000 m
)1
Æcm
)1
for porcine and bovine OBP, respectively).
The functionality of the different OBP forms was deter-

mined by direct titrations using the fluorescent ligand
AMA as reported previously [18]. Briefly, 1 mL samples of
1 lm OBP, in 20 mm Tris ⁄ HCl buffer pH 7.8, were incu-
bated overnight at 4 °C in the presence of increasing con-
centrations of AMA (0.156–10 lm). Fluorescence emission
spectra between 450 and 550 nm were recorded with a
Perkin-Elmer LS 50 luminescence spectrometer (excitation
and emission slits of 5 nm) at a fixed excitation wavelength
of 380 nm and the formation of the AMA–OBP complex
was followed as an increase in the fluorescence emission
intensity at 480 nm. Dissociation constants for the AMA–
OBP complexes were determined from the hyperbolic titra-
tion curves using the nonlinear fitting program sigma
plot 5.0 (Cambridge Software Corp., Cambridge, MA).
The concentration of the AMA–OBP complex was deter-
mined on the basis of emission spectra obtained when
incubating AMA (0.1–10 lm) with saturating amounts of
both OBP forms.
Spectrophotometric assay for the detection
of HNE–OBP binding complexes
Two 0.4 lL aliquots of HNE were taken from a 27 mm
stock solution in hexane and dried under a nitrogen stream
to remove the organic solvent. The samples were resuspend-
ed in 100 lL of different solutions containing, respectively,
120 lm porcine OBP and 60 lm bovine OBP in 20 mm
Tris ⁄ HCl buffer pH 7.8. In both samples the concentration
of HNE was 90 lm. After two different incubation times,
20 min at room temperature and 16 h at 4 °C, one half
(50 lL) of each OBP–HNE sample was concentrated to a
final volume of 5 lL using an Ultrafree-MC, 10 000 Da

cut-off, Centrifugal Filter (Millipore, Medford, MA). The
HNE eventually not bound to OBP was recovered in the
ultrafiltrate and quantified spectrophotometrically after
dilution in 1 mL of 20 mm Tris ⁄ HCl buffer pH 7.8. The
HNE molecules specifically interacting with the binding
sites of the two OBP forms were then displaced by the
addition of a large excess of the OBP ligand undecanal
(730 lL of a 5 mm solution in Tris ⁄ HCl pH 7.8) and separ-
ated from the proteins with a second concentration step.
Because no other component in the ultrafiltrate had an
absorption band in the UV region, the amount of HNE
released from OBP could be determined spectrophotome-
trically. A 100 lL sample of 90 lm HNE alone in Tris
buffer was processed in parallel to evaluate the amount of
HNE eventually retained by nonspecific interactions with
the membrane of the Ultrafree-MC Centrifugal Filters.
Competitive binding test to determine the
dissociation constant for HNE–OBP complexes
The dissociation constants for the complexes between the
different OBP forms and HNE were determined, as des-
cribed previously for other ligands [17,19], by competitive
binding tests with the fluorescent ligand AMA. Recombin-
ant porcine (1 lm) and bovine OBP (0.5 lm), dissolved in
20 mm Tris ⁄ HCl buffer pH 7.8, were preincubated at room
temperature for 20 min with AMA (1.5 and 1.0 lm for por-
cine and bovine OBP, respectively). Samples (1 mL) of
these solutions were then poured into different tubes con-
taining increasing amounts of HNE (1.0–70.0 lm) in which
the hexane of the HNE stock solution had been previously
removed under a nitrogen stream. The samples were incuba-

ted for another 30 min at room temperature and the binding
of HNE was followed as a decrease of the fluorescence
emission of the AMA–OBP complex at 480 nm upon excita-
tion at 380 nm (excitation and emission slits 5 nm). Titra-
tion in the absence of OBP allowed us to exclude that HNE
might lead to changes of the fluorescence intensity of AMA.
The apparent K
d
values for the HNE–OBP complexes
were determined from the competition curves analysed as
two parameters hyperbolic decays using the nonlinear fit-
ting function of sigma plot 5.0 (Cambridge Software
Corp.). The true K
d
values were then calculated from
Eqn (1) [17]:
K
true
d
¼ K
app
d
Â
1
1 þð1=K
AMA
d
½AMAÞ
ð1Þ
where K

AMA
d
is the dissociation constant of the AMA–OBP
complex.
Competitive binding tests with heat-denatured OBPs
(incubated overnight at 90 °C) were performed to determine
the specificity of the interaction between the protein and
HNE.
S. Grolli et al. OBP as a scavenger for HNE
FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS 5139
Immunoblotting of HNE-modified OBP
Formation of HNE–OBP covalent adducts was induced by
incubating 0.5 mL of 1 mgÆmL
)1
porcine and bovine OBP
(52 and 26 lm respectively) for 4 h at 37 °C with increasing
super-saturating amounts of HNE (0.1–2.5 mm). The high-
est levels of HNE were chosen to exceed, in terms of molar
ratio, the number of protein amino acid residues (13 and 21
for porcine and bovine OBP, respectively) that could react
with the aldehyde. Aliquots (10 lL) were run on
SDS ⁄ PAGE [27] while the remaining volume of each sam-
ple, after dialysis in 20 mm Tris ⁄ HCl buffer pH 7.8, was
stored at 4 °C for a functionality assay with the OBP fluor-
escent ligand AMA (see below). After electrophoresis, the
protein bands were transferred onto a poly(vinylidene diflu-
oride) membrane [28] that was blocked with 5% skimmed
dry milk dissolved in 20 mm Tris ⁄ HCl buffer, pH 7.8 con-
taining 150 mm NaCl, and 0.01% (w ⁄ v) Tween 20 (TTBS).
After washing with TTBS the membrane was incubated

with an anti-(HNE–protein adducts) serum raised in rabbit
and diluted (·1000) in TTBS ⁄ containing 1% (w ⁄ v) skim-
med dry milk. These two steps were repeated with the
secondary antibody conjugated to horseradish peroxidase
(dilution factor 2000). HNE–chemically modified OBP
bands were visualized with diamino-benzidine.
The functionality of the OBP samples employed in the
immunoblotting experiment was tested with the fluorescent
ligand AMA as described above. The residual AMA bind-
ing of each sample was normalized to reference binding val-
ues obtained with OBP samples which had not been treated
with HNE.
Titration of HNE-reacting residues in OBP
The titration was specifically designed for the quantitative
determination of the OBP HNE-reacting residues that had
been displayed, at the qualitative level, throughout the
immunoblotting experiment described above.
Briefly, samples of porcine and bovine OBP (120 and
60 lm, respectively) in 0.2 mL of 20 mm Tris ⁄ HCl pH 7.8
buffer were incubated with increasing amounts of HNE
(0.09–5 mm). One half of each sample was stored at 4 °C
for an overnight incubation (16 h) while the remaining part,
after 20 min, was treated with 0.4 mL of a solution con-
taining a large excess (5 mm) of the OBP ligand undecanal
[5–19], to displace the aliquot of HNE present in the bind-
ing site. The solutions were then concentrated on Ultrafree-
MC, 10 000 Da cut-off, centrifugal filters and the amount
of HNE not associated with OBP was determined spectro-
photometrically in the ultrafiltrates on the basis of the
absorbance at 223 nm. The number of HNE covalent

adducts ⁄ equivalent of OBP was finally calculated by sub-
tracting the values determined in the ultrafiltrates from the
initial amounts of HNE. The second aliquot of each sample
that had been stored at 4 °C was than treated with the
same procedure to determine if the number of covalent
adducts increased with time (16 h).
Biochemical characterization of OBP exposed
to HNE in a test tube assay simulating conditions
of oxidative stress in nasal mucosa
These experiments were carried out incubating porcine and
bovine OBP at their presumed physiological concentrations
in nasal mucosa (1 and 0.5 mm, respectively, for pOBP and
bOBP) [6], with HNE at the levels reported in case of acute
oxidative stress in vivo (2.0 mm) [1]. Samples (50 lLin
20 mm Tris ⁄ HCl buffer pH 7.8) were incubated for 20 min
while the temperature of 25 °C is that of inhaled air in
favourable environments of most mammalian species [22].
The solutions were then concentrated on Ultrafree-MC,
10 000 Da cut-off, centrifugal filters after the addition of
0.95 mL of 5 mm undecanal in the same Tris buffer. The
amount of HNE-reacting groups was determined as repor-
ted above. The functionality of the OBP samples was finally
determined with AMA direct binding tests.
Evaluation of the protective role of OBP against
HNE cytotoxicity in a simplified model simulating
nasal mucosa
Preparation of immobilized bovine OBP (i-bOBP)
A6· His affinity tag was placed at the N-terminus of
bovine OBP by PCR using a specific primer. The fused
cDNA was subcloned in the expression vector pT7-7 and

the expression of the protein was realized as reported above
for the recombinant forms of porcine and bovine OBP. The
purification of the protein was obtained by affinity chroma-
tography with a Ni-NTA Agarose (Qiagen, Hilden,
Germany) according to the manufacturer’s instructions.
The aggregation state was determined by gel permeation by
a Superdex 70 column (Amersham Pharmacia, Milano
Italy) in FPLC, while the binding tests with AMA, HNE
and Undecanal were carried out as described above for the
recombinant OBPs. Following purification, the protein was
immobilized again to the Ni-NTA-Agarose (i-bOBP) and
was employed for the tests on living cells described below.
Cell culture and treatment
BT epithelial cells were maintained in Dulbecco’s modified
Eagle’s medium supplemented with 10% fetal bovine serum
(v ⁄ v) and 50 lgÆmL
)1
gentamycin. Cells were grown at
37 °C, 5% CO
2
in a humidified incubator.
Exponentially growing cells were seeded in 2 mL of
reduced serum medium (2.5% fetal bovine serum) at a den-
sity of 40 000 cellsÆwell
)1
in the lower compartment of a
six-well Costar Transwell plate. After 72 h, the following
treatments were added in the upper compartment of the
OBP as a scavenger for HNE S. Grolli et al.
5140 FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS

wells: 20 lm HNE, 20 lm HNE in association with a molar
excess of i-bOBP (40 lm), 40 lm i-bOBP alone. Since HNE
can diffuse throughout the microporous membrane of the
upper insert, its concentration is relative to the total volume
of medium in the well.
Cell viability was determined by Trypan Blue exclusion
test. Changes in cell number and morphology were evalu-
ated by phase contrast microscopy.
Acknowledgments
This work was supported by a grant FIRB-
RBAU0142ZF, from the Italian Ministry for Educa-
tion, University and Research. EM is the recipient of a
postdoctoral fellowship from the University of Parma.
We would like to thank Dr Mariella Tegoni (AFMB-
CNRS, Marseille, France), Dr Christan Cambillau
(AFMB-CNRS, Marseille, France) and Prof. Emanuele
Albano (Universita
`
del Piemonte Orientale, Novara,
Italy) for helpful discussions, and Debra Mc Millen
(OHSU, Portland, OR, USA) and Prof. Laura Kramer
(University of Parma) for reading the manuscript.
References
1 Esterbauer H, Schaur RJ & Zollner H (1991) Chemistry
and biochemistry of 4-hydroxynonenal, malonaldehyde
and related aldehydes (review). Free Radical Biol Med
11, 81–128.
2 Uchida K (2003) 4-Hydroxy-2-nonenal: a product and
mediator of oxidative stress (review). Prog Lipid Res 42,
318–343.

3 Alary J, Gueraud F & Cravedi JP (2003) Fate of
4-hydroxynonenal in vivo: disposition and metabolic
pathways (review). Mol Aspects Med 24, 177–187.
4 Engle MR, Singh SP, Czernik PJ, Gaddy D, Montague
DC, Ceci JD, Yang Y, Awasthi S, Awasthi YC & Zim-
niak P (2004) Physiological role of mGSTA4-4, a glu-
tathione S-transferase metabolizing 4-hydroxynonenal:
generation and analysis of mGsta4 null mouse. Toxicol
Appl Pharmacol 194, 296–308.
5 Vincent F, Spinelli S, Ramoni R, Grolli S, Pelosi P,
Cambillau C & Tegoni M (2000) Complexes of porcine
odorant binding protein with odorant molecules belong-
ing to different chemical classes. J Mol Biol 300, 127–
139.
6 Avanzini F, Bignetti E, Bordi C, Carfagna G, Cavagg-
ioni A, Ferrari G, Sorbi RT & Tirindelli R (1987)
Immunocytochemical localization of pyrazine-binding
protein in bovine nasal mucosa. Cell Tissue Res 247,
461–464.
7 Pelosi P & Tirindelli R (1989) Structure ⁄ activity studies
and characterization of an odorant-binding protein. In
Receptor Events and Transduction in Taste and Olfaction,
Chemical Senses, Vol. 1 (Brand JG, Teeter JH, Cagan
RH & Kare MR, eds), pp. 207–226. Marcel Dekker,
New York, NY.
8 Pevsner J, Hou V, Snowman AM & Snyder SH (1990)
Odorant-binding protein, characterization of ligand
binding. J Biol Chem 265, 6118–6125.
9 Dal Monte M, Centini M, Anselmi C & Pelosi P (1993)
Binding of selected odorants to bovine and porcine

odorant-binding proteins. Chem Senses 18, 713–721.
10 Flower DR (1996) The lipocalin protein family: struc-
ture and function. Biochem J 318, 1–14.
11 Flower DR, North AC & Sansom CE (2000) The lipo-
calin protein family: structural and sequence overview
(review). Biochim Biophys Acta 1482, 9–24.
12 Tegoni M, Pelosi P, Vincent F, Spinelli S, Campanacci
V, Grolli S, Ramoni R & Cambillau C (2000) Mamma-
lian odorant binding proteins (review). Biochim Biophys
Acta 1482, 229–240.
13 Spinelli S, Ramoni R, Grolli S, Bonicel J, Cambillau C
& Tegoni M (1998) The structure of the monomeric
porcine odorant binding protein sheds light on the
domain swapping mechanism. Biochemistry 37, 7913–
7918.
14 Tegoni M, Ramoni R, Bignetti E, Spinelli S & Cambil-
lau C (1996) Domain swapping creates a third putative
combining site in bovine odorant binding protein dimer.
Nat Struct Biol 3, 863–867.
15 Bianchet MA, Bains G, Pelosi P, Pevsner J, Snyder SH,
Monaco HL & Amzel LM (1996) The three-dimensional
structure of bovine odorant binding protein and its
mechanism of odor recognition. Nat Struct Biol 3, 934–
939.
16 Matarazzo V, Zsurger N, Guillemot JC, Clot-Faybesse
O, Botto JM, Dal Farra C, Crowe M, Demaille J, Vin-
cent JP, Mazella J et al. (2002) Porcine odorant-binding
protein selectively binds to a human olfactory receptor.
Chem Senses 8, 691–701.
17 Ramoni R, Vincent F, Grolli S, Conti V, Malosse C,

Boyer FD, Nagnan-Le Meillour P, Spinelli S, Cambillau
C & Tegoni M (2001) The insect attractant 1-octen-3-ol
is the natural ligand of bovine odorant-binding protein.
J Biol Chem 9, 7150–71555.
18 Ramoni R, Vincent F, Ashcroft AE, Accornero P,
Grolli S, Valencia C, Tegoni M & Cambillau C (2002)
Control of domain swapping in bovine odorant-binding
protein. Biochem J 365, 739–748.
19 Vincent F, Ramoni R, Spinelli S, Grolli. S, Tegoni M &
Cambillau C (2004) Crystal structures of bovine odor-
ant-binding protein in complex with odorant molecules.
Eur J Biochem 271, 3832–3842.
20 Rutland J & Cole PJ (1981) Nasal mucociliary clear-
ance and ciliary beat frequency in cystic fibrosis com-
pared with sinusitis and bronchiectasis. Thorax 36,
654–658.
S. Grolli et al. OBP as a scavenger for HNE
FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS 5141
21 Tahir RA & Goyal SM (1995) Rapid detection of pseu-
dorabies virus by the shell vial technique. J Vet Diagn
Invest 7, 173–176.
22 Reed CJ, Robinson DA & Lock EA (2003) Antioxidant
status of the tat nasal cavity. Free Radical Biol Med 34,
607–615.
23 Buck L & Axel R (1991) A novel multigene family may
encode odorant receptors: a molecular basis for odor
recognition. Cell 65, 175–187.
24 Paysan J & Breer H (2001) Molecular physiology of
odor detection: current views. Pflugers Arch 441, 579–
586.

25 Lechner M, Wojnar P & Redl B (2001) Human tear
lipocalin acts as an oxidative-stress-induced scavenger of
potentially harmful lipid peroxidation products in a cell
culture system. Biochem J 356, 129–135.
26 Lacazette E, Gachon AM & Pitiot G (2000) A novel
human odorant-binding protein gene family resulting
from genomic duplicons at 9q34: differential expression
in the oral and genital spheres. Hum Mol Genet 9, 289–
301.
27 Laemmli UK (1970) Cleavage of structural proteins dur-
ing the assembly of the head of the bacteriophage T4.
Nature 227, 680–685.
28 Towbin H, Staehelin T & Gordon J (1979) Electro-
phoretic transfer of proteins from polyacrylamide gels
to nitrocellulose sheets: procedure and some applica-
tions. Proc Natl Acad Sci USA 76, 4350–4354.
OBP as a scavenger for HNE S. Grolli et al.
5142 FEBS Journal 273 (2006) 5131–5142 ª 2006 The Authors Journal compilation ª 2006 FEBS