ADAM-8, a metalloproteinase, drives acute allergen-induced airway inflammation.

Abstract

Asthma is a complex disease linked to various pathophysiological events including the activity of proteinases. The multifunctional A disintegrin and metalloproteinases (ADAMs) displaying the ability to cleave membrane-bound mediators or cytokines appear to be key mediators in various inflammatory processes. In the present study, we investigated ADAM-8 expression and production in a mouse model of allergen-induced airway inflammation. In allergen-exposed animals, increased expression of ADAM-8 was found in the lung parenchyma and in DC purified from the lungs. The potential role of ADAM-8 in the development of allergen-induced airway inflammation was further investigated by the use of an anti-ADAM-8 antibody and ADAM-8 knockout animals. We observed a decrease in allergen-induced acute inflammation both in BALF and the peribronchial area in anti-ADAM-8 antibody-treated mice and in ADAM-8-deficient mice (ADAM-8(-/-) ) after allergen exposure. ADAM-8 depletion led to a significant decrease of the CD11c(+) lung DC. We also report lower levels of CCL11 and CCL22 production in antibody-treated mice and ADAM-8- deficient mice that might be explained by decreased eosinophilic inflammation and lower numbers of DC, respectively. In conclusion, ADAM-8 appears to favour allergen-induced acute airway inflammation by promoting DC recruitment and CCL11 and CCL22 production.

Full text

Page 1

1??  
??  
ADAM-8, a metalloproteinase, drives acute allergen-induced airway inflammation
Geneviève PAULISSEN, Natacha ROCKS, Maud M. GUEDERS, Denis
BEDORET, Céline CRAHAY, Florence QUESADA-CALVO, Jonathan HACHA,
Sandrine BEKAERT, Christophe DESMET, Jean-Michel FOIDART, Fabrice
BUREAU, Agnès NOEL, Didier D. CATALDO
Laboratory of Tumor and Development Biology, GIGA-Research (GIGA-I³ and GIGA-
cancer), University of Liege and CHU of Liege, Sart-Tilman, Belgium.


Key words : ADAM, asthma, inflammation, CCL22, dendritic cells



Correspondence and reprint requests should be addressed to:

Professor Didier D. CATALDO
Tower of Pathology (B23)
GIGA-Research, University of Liege
4000 Liege, Belgium
Fax : +3243662939
Phone: +3243662521
E-mail: didier.cataldo@ulg.ac.be





















List of abbreviations:
ADAM: A Disintegrin And Metalloproteinase
ADAM-8-/-: ADAM-8 deficient mice
ADAM-8+/+: ADAM-8 wild-type mice
sADAM-8: soluble form of ADAM-8

Page 2

2??  
??  
Summary
Asthma is a complex disease linked to various pathophysiological events including
the activity of proteinases. The multifunctional A Disintegrin And Metalloproteinases
(ADAMs) displaying the ability to cleave membrane-bound mediators or cytokines
appear to be key mediators in various inflammatory processes. In the present study,
we have investigated ADAM-8 expression and production in a mouse model of
allergen-induced airway inflammation. In allergen-exposed animals, increased
expression of ADAM-8 was found in the lung parenchyma and in dendritic cells
purified from the lungs. The potential role of ADAM-8 in the development of allergen-
induced airway inflammation was further investigated by the use of an anti-ADAM-8
antibody and ADAM-8 knock-out animals. We observed a decrease in allergen-
induced acute inflammation both in BALF and the peribronchial area in anti-ADAM-8
antibody-treated mice and in ADAM-8 deficient mice (ADAM-8-/-) after allergen
exposure. ADAM-8 depletion led to a significant decrease of the CD11c+ lung
dendritic cells. We also report lower levels of CCL11 and CCL22 production in
antibody-treated mice and ADAM-8-/- mice that might be explained by decreased
eosinophilic inflammation and lower numbers of dendritic cells, respectively. In
conclusion, ADAM-8 appears to favour allergen-induced acute airway inflammation
by promoting dendritic cell recruitment and CCL11 and CCL22 production.

Page 3

3??  
??  
Introduction
Asthma is mainly characterized by chronic bronchial inflammation and
hyperresponsiveness [1, 2]. The disease phenotype in humans encompasses
generally repeated episodes of non permanent airway obstruction along with
wheezing, breathlessness and cough. Mouse models of asthma mimic most of these
human asthma features and provide valuable tools for deciphering disease
mechanisms [3]. Such mouse models have proven valuable for elucidating the
implication of specific gene products, such as matrix proteinases. Therefore, several
research groups have studied the asthmatic phenotype by using knock-out animals
and reported that MMP-9 depleted mice display a lower cell infiltration and bronchial
responsiveness upon allergen exposure [4]. In sharp contrast, MMP-8 deficiency
promotes a neutrophilic inflammation in the airways [5] and MMP-2 deficiency
induces inflammatory cell accumulation in lung parenchyma responsible for asphyxia
in some animals [6].
ADAMs (A Disintegrin And Metalloproteinases) belong to the metzincin superfamily
of proteinases. The typical structure of ADAM proteinases consists of
metalloproteinase, disintegrin, and cytoplasmic domains which endow the protein
with catalytic activity, adhesion properties, and potentially signaling functions [7-10].
Light was shed on the potential role of ADAM proteinases in asthma by a genetic
association study demonstrating a link between increased risk of asthma
development and polymorphisms in the ADAM-33 gene [11]. It was also reported that
ADAM-33 gene expression correlates with asthma severity [12]. ADAM-8 is another
member of the ADAM family displaying some structural similarities with ADAM-33.
ADAM-8 has been first cloned from mouse macrophages [13] and is expressed at
baseline or after stimulation by a wide range of cell types including neurons,

Page 4

4??  
??  
osteoclasts, bronchial epithelial cells and cells of the immune system (eosinophils,
polymorphonuclear leukocytes (PMN), monocytes, macrophages, dendritic cells and
B lymphocytes) [13-17]. ADAM-8 is implicated in various complex biological
processes such as neurodegenerative inflammation and osteoclastogenesis [18, 19].
Moreover, two distinct studies based on a wide genome microarray have shown that
ADAM-8 mRNA expression appears to be increased in asthma [20, 21]. More
recently, transgenic mice overexpressing a soluble form of ADAM-8 (sADAM-8) in
liver tissue showed less cellular infiltration in an allergen-induced model of lung
inflammation, suggesting that sADAM-8 might play a role as inflammation inhibitor
[22]. Known biological functions of ADAM-8 are to date restricted to cell fusion and
adhesion [23]. ADAM-8 also appears to be a sheddase and cleaves CD23, neural
cell adhesion molecule close homologue of L1 (CHL1), beta-amyloid precursor
protein (APP), L-selectin, vascular cell adhesion molecule-1 (VCAM-1), and myelin
basic protein (MBP) while there is also evidence for a possible cleavage of pro-tumor
necrosis factor (TNF) α and CD16 [24, 25] rendering plausible a role as inflammation
modulator for this proteinase [25-27]. However, the question of whether ADAM-8 is
involved in the development of asthma phenotype remains unanswered.
In this study, we demonstrate through both genetic and antibody-based approaches
that ADAM-8 is a key mediator for the development of asthma-linked inflammation.
We describe that ADAM-8 depletion (transgenic mice deficient for ADAM-8) and
injected antibodies raised against ADAM-8 impair the development of airway
inflammation after allergen exposure. We also explored the mechanisms implicating
ADAM-8 and report that ADAM-8 depletion modifies dendritic cell recruitment and
production of key cytokines contributing to the network of interactions leading to
eosinophilic inflammation.

Page 5

5??  
??  
Results
ADAM-8 expression and production in lungs of mice acutely exposed to allergens
In order to investigate ADAM-8 expression in an experimental model of asthma, male
Balb/c mice were sensitized and subsequently exposed to ovalbumin (OVA) for 5
days. As expected, mice exposed to OVA showed higher BALF eosinophil
percentages and peribronchial inflammation as compared to mice exposed to PBS
(eosinophil percentage: 1.700±1.361 in sham-exposed mice versus 56.187±6.255 in
allergen-exposed mice, p<0.0001; inflammation score: 0.325±0.067 in sham-exposed
mice versus 1.161±0.084 in allergen-exposed mice, p<0.0001). ADAM-8 was
significantly upregulated at mRNA level in lung parenchyma of mice acutely exposed
to allergens as compared to controls (figure 1A and C). Moreover, ADAM-8 protein
production was also significantly increased in OVA-exposed mice as assessed by
Western blotting and immunohistochemistry (figure 1B, D and E). Interestingly, the
immunohistochemical analysis revealed that ADAM-8 stained cells mostly
correspond to inflammatory cells such as eosinophils and macrophages located in
the peribronchial area (figure 1E).
ADAM-8 expression and production in chronic allergen-induced remodeling
In order to further investigate ADAM-8 expression and production in a chronic
remodeling model of asthma, male Balb/c mice were sensitized and subsequently
exposed to OVA for 5 days per week (odd weeks) from days 22 to 96 (eosinophil
percentage: 4.033±0.704 in sham-exposed mice versus 24.600±1.613 in allergen-
exposed mice, p<0.0001; inflammation score: 0.593±0.096 in sham-exposed mice
versus 1.129±0.061 in allergen-exposed mice, p<0.05; glandular hyperplasia
expressed in positive cells percentage: 0.062±0.062 in sham-exposed mice versus

Page 6

6??  
??  
33.640±4.363 in allergen-exposed mice, p<0.0001; basement membrane collagen:
0.654±0.046 in sham-exposed mice versus 1.444±0.070 in allergen-exposed mice,
p<0.0001). In this long-term model, mRNA levels of ADAM-8 were only poorly
detectable (figure 1A) and the ADAM-8 protein production was very weak (figure 1B).
Moreover, no significant difference between mice exposed to allergen or not was
observed by RT-PCR and Western blotting.
ADAM-8 blockade by antibody treatment
In order to assess the implication of ADAM-8 in asthma-related acute inflammation,
Balb/c mice were treated with an anti-ADAM-8 antibody for 3 days while a control
isotype was injected to control Balb/c mice. Two hours after antibody injection, mice
were exposed to OVA by aerosols. As shown in table 1 and figure 2A, eosinophil
percentages in BALF were significantly decreased upon anti-ADAM-8 antibody
treatment as compared to control mice treated with control isotype (p<0.05). In
peribronchial areas, acute allergen-induced inflammation was also decreased in anti-
ADAM-8-treated mice (inflammation score: 1.080±0.160 in mice treated by control
isotype versus 0.500±0.150 after anti-ADAM-8 antibody). Peribronchial eosinophilic
inflammation was also significantly lower in anti-ADAM-8-treated mice (p< 0.0005)
(figure 2B). Amounts of detectable ADAM-8 protein were strongly decreased in lung
parenchyma of anti-ADAM-8-treated mice as assessed by Western blot (figure 2C).
Moreover, decrease of eosinophilia BALF was assessed in both Th2-prone mice
(Balb/c) and non Th2-prone (C57Bl/6) and found to be similar (figure 2D).
Acute allergen-induced inflammation in ADAM-8-/- animals
We also applied the allergen-induced acute inflammation model to ADAM-8 deficient
mice (ADAM-8-/-) and wild-type counterparts (ADAM-8+/+). After allergen exposure,

Page 7

7??  
??  
BAL displayed an increase of eosinophil counts in ADAM-8+/+ while allergen-induced
eosinophil BAL infiltration expressed as absolute value or percentage was lower in
ADAM-8-/- mice (p<0.0005) (table 2). Bronchial inflammation was also studied by
histology. Upon allergen exposure, ADAM-8-/- mice displayed far less peribronchial
inflammation than ADAM-8+/+ mice (p<0.0005) (figure 2 E-H and K). Peribronchial
tissue was specifically evaluated regarding eosinophil infiltration by histology and
specific staining. After allergen exposure, eosinophil counts in airway walls were 5
times lower in ADAM-8-/- mice as compared to ADAM-8+/+ (p<0.0005) (figure 2 I-J
and L).
Cytokine measurements by protein arrays and ELISA
As cytokines are key mediators in the development of the asthmatic phenotype, a
cytokine array was performed on lung tissues in ADAM-8-/- mice and WT mice after
acute allergen exposure. Among measured cytokines, CCL11 and CCL22 levels
were found to be significantly lower after allergen exposure in ADAM-8-/- as
compared to ADAM-8+/+ (data not shown). These data were confirmed by ELISA
(p<0.0005) (figure 3A-B). In mice exposed to OVA, CCL22 levels were positively
correlated with eosinophil counts in the lung tissues (r=0.7598 and p=0.0004,
(figure3E)). The modulation of CCL11 and CCL22 by ADAM-8 depletion has also
been observed in our study based on ADAM-8 blockade by antibody injection
(p<0.05) (figure 3C-D). Prototypical Th2 cytokines IL-4, IL-5 and IL-13 have been
assessed by ELISA in whole lungs extracts and did not display significant changes
upon ADAM-8 depletion (figure 3 F-H).

Page 8

8??  
??  
Determination of dendritic cells recruitment in lung parenchyma in ADAM-8-/- animals
As CCL22 is mainly secreted by dendritic cells and macrophages, an
immunohistochemistry against CD11c and F4/80, respectively was performed on
lung tissues in ADAM-8-/- mice as compared to ADAM-8+/+ after allergen exposure.
Numbers of dendritic cells and alveolar macrophages positive for this staining were
significantly decreased in ADAM-8–/– mice as compared to ADAM-8+/+ after allergen
exposure (figure 4 A-D). This suggests that lower levels of CCL22 are the
consequence of a significant decrease of the two most important CCL22-producing
cells in lung parenchyma from ADAM-8 deficient mice exposed to allergens.
ADAM-8 expression in lung dendritic cells
In order to verify the potential implication of ADAM-8 in immunological processes
leading to asthma-related acute inflammation, we performed a RT-PCR on mRNA
from dendritic cells extracted from lungs of mice acutely exposed to allergens. Levels
of mRNA coding for ADAM-8 were higher in allergen-exposed mice as compared to
control mice (figure 4E-F), suggesting that ADAM-8 might be a key regulator of
dendritic cells behavior. Additionally, ADAM-8 expression was assessed by RT-PCR
in other major cellular sources in the lung such as epithelial cells, eosinophils,
alveolar macrophages and interstitial macrophages (Figure 4G).

Page 9

9??  
??  
Discussion
We designed the present study to investigate the potential role of the
metalloproteinase ADAM-8 in allergen-induced airway diseases. Through genetic
and pharmacological approaches in two different strains of mice (Balb/c, Th2-prone
and C57Bl/6, Th1-prone), we report that ADAM-8 significantly contributes to the
establishment of acute allergen-induced inflammation in asthma. This novel concept
is based on the findings that anti-ADAM-8 antibody-treated mice and ADAM-8-
deficient mice displayed significantly decreased levels of allergen-induced acute
inflammation in BAL and in peribronchial areas. We also report for the first time that
ADAM-8 is overexpressed in lung dendritic cells after allergen exposure. Moreover,
we found increased levels of CCL11 and CCL22 after allergen exposure in ADAM-
8+/+ mice while these two cytokines failed to increase in ADAM-8-/- animals displaying
therefore levels similar to sham-exposed mice for those two cytokines. We also
reported a decrease of CCL11 and CCL22 production in anti-ADAM-8-treated mice.
To explore the mechanism leading to a decrease of CCL22, we measured the
numbers of dendritic cells and alveolar macrophages in the lung parenchyma and
described a significant decrease of those two cell types in lungs of ADAM-8-/-
animals. Moreover, CCL22 levels were highly correlated to eosinophilic inflammation
in allergen-exposed animals. Finally, we also demonstrate that the expression and
production of ADAM-8 proteinase were not chronically modulated in chronic asthma
and therefore could not play significant roles in remodeling processes linked to
asthma.
These novel findings are in accordance with emerging data on the implication of
ADAM-8 in the pathogenesis of pulmonary diseases [28] and suggest that ADAM-8
might be a key regulator of dendritic cells and alveolar macrophages homeostasis.

Page 10

10??  
??  
From these data, one can speculate that ADAM-8 could play a significant role in
asthma. This assumption is supported by the observation of increased levels of
ADAM-8 mRNA expression in induced sputum cells and endobronchial biopsies from
human asthmatics [12, 29] and previous authors have reported that ADAM-8 is
overexpressed in mouse models of asthma. We demonstrate that ADAM-8 depletion
modifies asthma phenotype. However, the precise mechanisms that could account
for this ADAM-8 overexpression in asthma and its role as an effector are not yet
completely dissected. Our data suggest that ADAM-8 might be essential for dendritic
cells accumulation in lung parenchyma although other mechanisms such as a
relationship between ADAM-8 and several interleukins (IL-4, IL-13) or related
intracellular activation pathways (STAT-6) might also interfere with bronchial
inflammation [21]. A genomic study recently reported a link between ADAM-8 single
nucleotide polymorphisms and asthma in humans [30]. In addition, two separate
genome-wide searches performed on mice subjected to allergen exposure
discriminated ADAM-8 as a valid marker of allergen-induced inflammation [20, 21].
Interestingly, one of these studies compared different durations of exposure to
allergens and reported that ADAM-8 was overexpressed only in acute models while
ADAM-8 overexpression was not reported after chronic allergen exposures [20]. In
line with this recent study of Di Valentin et al who performed microarray studies, we
emphasized the role of ADAM-8 in acute inflammation and not in chronic remodeling
associated to asthma. Moreover, this is in keeping with the findings of Matsuno et al
who showed that ADAM-8 levels are increased in patients suffering from acute
eosinophilic pneumonia but not chronic idiopathic eosinophilic pneumonia [31]. In our
study, we report that there is less eosinophilia in the BAL in ADAM-8 KO mice. As a

Page 11

11??  
??  
consequence of these huge variations in eosinophil counts, we also found that total
cell counts were significantly decreased in ADAM-8 KO animals.
Since there is no difference in features of chronic remodeling established between
ADAM-8 deficient mice and wild-type mice exposed to allergens, we hypothesize that
ADAM-8 is only implicated in acute airway inflammation. ADAM-8 cleaves various
key molecules in inflammatory processes including CD23, pro-TNF-α and L-selectin
[25, 26]. From this substrate specificity, it can be deduced that ADAM-8 might be an
effector in immunological processes [22, 29]. However, ADAM-8-/- mice did not
display any developmental abnormalities and their immune system was previously
reported as not impaired by ADAM-8 gene deletion [32]. We verified that
unchallenged ADAM-8-/- mice bear normal eosinophil percentages in their blood by
studying blood smears (data not shown).
Our experiments using ADAM-8-/- animals and anti-ADAM-8 treatment demonstrate
that ADAM-8 is per se responsible for a significantly increased allergen-induced
acute inflammation both in lung tissue and BAL. In order to rule out the potential role
of mice genetic background in our results, we compared the Th2-prone strain Balb/c
with Th1-prone C57Bl/6 in the antibody-blocking experiment and we report similar
decreases of lung inflammation. This validates our study on ADAM-8 genetically
depleted mice performed on a Th1-prone background (129/Ola; C57Bl/6 mixed
genetic background as most of KO mice). These experiments allow us to
demonstrate that the effects of ADAM-8 depletion reported in the present study are
not restricted to a specific strain and probably correspond to a conserved
mechanism. In a first approach, our results could seem in apparent discrepancy with
those generated by Matsuno et al [22] using mice overexpressing soluble ADAM-8 in
the liver. Indeed, these authors reported that sADAM-8 overexpressing mice were

Page 12

12??  
??  
protected from the development of histological inflammation when subjected to
asthma induction (noteworthy, these authors did not find any variation of BAL
inflammation) [22]. However, transgenic mice used by these previous authors
express normal levels of membrane-bound ADAM-8 in their tissues and have an
excess of sADAM-8 in their plasma, released from the liver, rendering the final
picture far more complex than targeted gene deletion. Moreover, one cannot exclude
the possibility that, in this setting, sADAM-8 might display some specific biological
activities different from membrane-bound ADAM-8.
Mechanisms that could explain such a decrease in eosinophils as we observed in
ADAM-8-/- and anti-ADAM-8-treated mice could be linked to a decreased production
of CCL11 and CCL22, two chemo-attractants for eosinophils [33, 34]. CCL22,
referred to as macrophage-derived chemokine (MDC), also takes part to Th2
inflammation. Interestingly, immunostainings performed in humans and mice
suggested that cells producing ADAM-8 are mostly eosinophils [21, 29]. As CCL11
(eotaxin-1) is mainly secreted by eosinophils, a significant decrease of those cells, as
reported in the present study might account per se for lower levels of CCL11 found
after ADAM-8 inhibition. By contrast, CCL22 is mostly produced by macrophages and
dendritic cells [35, 36] and appears to play a central role in allergen-induced
inflammation since it is produced by Th2-induced dendritic cells [37]. As
demonstrated in this study, number of macrophages and dendritic cells present in
lung tissue in allergen-exposed ADAM-8-/- mice is lower than ADAM-8+/+ mice.
Very recently, other authors have reported that ADAM-8 is detectable in bronchial
epithelium and smooth muscle cells [38]. In contrast with these observations, our own
findings, in line with those of King et al who performed in situ hybridization, indicate
that ADAM-8–expressing cells are mostly eosinophils and macrophages [21]. This

Page 13

13??  
??  
might be the consequence of technical issues (e.g. antibodies used by previous
authors are different). However, one cannot exclude that ADAM-8 expressed by
structural cells might play a significant role in the modulation of inflammation. One
argument to state that stromal cells such as fibroblasts probably do not participate in
the ADAM-8 production in an important proportion is that immunohistochemistry
performed on lung sections do not display an important staining in these cells. In the
present study, we measured the ADAM-8 expression by alveolar macrophages and
interstitial macrophages since those two subtypes have recently been described to
play very different roles in asthma pathology. We found that these two subtypes
express ADAM-8 without obvious difference.
Our results are in accordance with a very recent paper published on-line during the
revision period of this manuscript and reporting that genetic depletion of ADAM-8 is
protective against main features of asthma [39].
We conclude that ADAM-8 is not only a marker of allergen-induced acute
inflammation but probably also an effector since its depletion impairs the
development of eosinophilic inflammation through decreasing lung dendritic cells and
alveolar macrophages and modulating CCL11 and CCL22 production.

Page 14

14??  
??  
Material and Methods
Experimental asthma protocol
Six to 8 weeks old Balb/c male mice were provided by the animal house of the
University of Liège (Belgium). ADAM-8 deficient mice and corresponding wild-type
mice were kindly provided by Dr. Andrew Docherty (UCB-Celltech, Slough, UK).
ADAM-8 knock-out (KO) mice display a 129/Ola; C57/BL/6 mixed genetic
background and only brethren were used. We used only male ADAM-8 KO mice or
controls in our experimental procedures. In order to perform a genotyping to ensure
that animals were homozygous KO or homozygous WT before entering our
experimental protocols, DNA was obtained from mice tails by “Machery-Nagel
Nucleospin®Tissue” kit. Primer pair used to distinguish KO and WT is:
CACTGTTGGACTGGCTAAGGTG and GACATCGGTAACATTGGTCAG. Protocols
used in this study were approved by the Ethical Committee (University of Liège,
Belgium). Concerning the acute inflammation model, male mice were sensitized by
intraperitoneal injections of 10 µg OVA (Sigma Aldrich, Schnelldorf, Germany)
complexed with aluminium hydroxide (Perbio, Erembodegem, Belgium) on days 0
and 7. From days 21 to 25, mice were challenged daily by OVA 1% aerosols for 30
minutes. In the control group, mice were exposed to PBS aerosols. Mice were
sacrificed 24 hours after the last allergen exposure. For the remodeling model, mice
were sensitized by intraperitoneal injections on days 0 and 11, and groups of mice
were exposed to OVA or PBS aerosols 5 days per week (odd weeks) from days 22 to
96. As described in the “acute inflammation model”, mice were sacrificed 24h after
the last allergen exposure. We used OVA from one single batch that has been
characterized regarding endotoxin content. We have performed endotoxin
measurement (Lonza, Braine-l’Alleud, Belgium). Mean measured endotoxin levels in

Page 15

15??  
??  
the current batch of OVA are 0.236 ng of endotoxin/gr of OVA. We also ensured that
PBS used in our experiment was endotoxin-free (manufacturer specifications indicate
no detectable levels of endotoxin).
ADAM-8 blockade by antibody treatment
On days 0 and 7, male Balb/c and C57Bl/6 mice were sensitized by intraperitoneal
injections of 10 µg OVA (Sigma Aldrich, Schnelldorf, Germany) complexed with
aluminium hydroxide (Perbio, Erembodegem, Belgium). From day 21 and for 3
consecutive days, mice were exposed to OVA for 30 minutes. Two hours before each
allergen challenge, intravenous injection was performed with either 25 µg anti-ADAM-
8 antibody (Santa Cruz Biotechnology Inc, Santa Cruz, CA, USA) or control IgG
antibody (R&D Systems, Minneapolis, MN, USA). Mice were sacrificed 24 h after the
last allergen exposure.
Bronchoalveolar lavage fluid (BALF)
After sacrifice, a canula was inserted in the trachea of the mouse and BALF was
performed manually by instillation of 4x1ml of PBS-EDTA 0.05mM (Calbiochem,
Darmstadt, Germany). BAL was subjected to centrifugation for 10 minutes at 4°C and
at 282 x g. Supernatants were stored at -80°C for future assessments while cell
pellets were resuspended in 1ml PBS-EDTA 0.05mM in order to proceed with total
and differential cell counts. Total cell number was measured by using a Z2 coulterΧ®
particle count and size analyzer (Beckman Coulter, Analis, Namur, Belgium) and
differential cell count was assessed by a skilled observer blinded to experimental
details, based on morphological criteria. For this purpose, cells were centrifuged
(cytospin) on a slide and stained with Diff Quick® (Dade, Brussels, Belgium).
Lung tissue processing