Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Allergic bronchial asthma is characterized by the development of a T-helper (Th) 2 immune response with elevated immunoglobulin (Ig)E levels; airway inflammation; influx of activated T cells, monocytes, and eosinophils; goblet-cell hyperplasia; and airway hyperresponsiveness (AHR). (For review, see References 1-4.) Although the immunologic network in bronchial asthma has been extensively studied during recent years, the underlying cellular and molecular mechanisms leading from inflammatory events to AHR still remain enigmatic.

In vivo models of allergic bronchial asthma have been reported in which repeated exposures of sensitized mice to allergen-aerosol recruited eosinophils and lymphocytes into the airways, resulting in the development of a Th2 immune response, changes in airway epithelium, and AHR (5). In the 1990s, growing evidence has emerged indicating neuronal disregulation in allergic asthma (8). Substance P (SP) and neurokinin A (NKA) are proinflammatory neuropeptides, belonging to the family of tachykinins, that can mimic asthma-like responses including smooth-muscle contraction, dilatation of vascular tone, increased vascular permeability, and enhanced mucus production in the lung of asthmatic patients (9). The levels of neuropeptides were shown to be increased in the lungs of asthmatic patients (10). In a guinea-pig model, it has been shown that the increase of SP and NKA in the lung in response to allergen challenge is due to an induction of these neuropeptides in neurons of the nodose ganglion (11).

NGF, a member of the neurotrophin family, was recently shown to play a critical role in allergy and asthma (12, 13). We could show that NGF production in the lung is elevated following allergen challenge in sensitized mice as well as in allergic asthma patients (12, 14). We identified T cells and macrophages as a source of NGF. Blocking of NGF activity by administration of anti-NGF antibodies almost entirely prevented development of AHR in mice (12). Brain-derived neurotrophic factor (BDNF) is known to induce neuropeptide synthesis in neurons (15). BDNF has been shown to support survival, differentiation, and function of a broad number of neurons of the central as well as the peripheral nervous system. BDNF modulates neuronal activity and enhances tachykinin neuropeptide production (16). Detection of BDNF messenger RNA (mRNA) and protein have been reported in extracts of the lung (17) and other non-neuronal tissues (18). The tyrosine kinase receptor trkB was identified as a high-affinity receptor and p75 (NTR) as a low-affinity receptor for BDNF, playing a central role in cellular signaling of this neurotrophin. Expression of trkB has been reported on neuronal and non-neuronal cells including lymphocytes (20, 23). BDNF and trkB knockout mice showed that a substantial number of visceral sensory neurons, localized in the nodose and dorsal root ganglia, require BDNF for survival (24). BDNF knockout mice demonstrate severe deficits in breathing control due to a great loss of viscerosensory neurons (25). Therefore, BDNF is a candidate for mediating functional neuronal changes in asthma. The aim of the present study was to investigate the production and the cellular sources of BDNF in allergic immune responses in lung and airways using a mouse model.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

BALB/c mice were obtained from Harlan Winkelmann (Borchen, Germany) and maintained under controlled conditions.

Protocol of Allergic Sensitization

Ovalbumin (OVA)-sensitized mice were sensitized to OVA (10 µg/injection) (Sigma, Deisenhofen, Germany) adsorbed to 1.5 mg Al(OH)₃ (Pierce, Rockford, IL) or vehicle alone by intraperitoneal injections on Days 1, 14, and 21. Before analysis, animals received two consecutive local aerosol challenges of 1% OVA (wt/vol) diluted in phosphate-buffered saline (PBS) or of PBS alone, delivered by aerosolization of 20 min on Days 26 and 27 (2× challenge) as previously described (26). A second group was additionally challenged on Days 33, 34, 40, 41, 47, and 48 (8× challenge). All animals were analyzed 24 h after the last challenge (Day 28 or 49 for the first or second group, respectively).

Determination of Allergen-Specific Antibodies

Blood was sampled from the lateral caudal veins. Blood was clotted at room temperature and centrifuged (1,200 × g, 10 min, room temperature [RT]). Total IgE and allergen-specific IgE, IgG₁, and IgG_2a antibody titers were measured by enzyme-linked immunosorbent assay (ELISA) as described previously (26).

Bronchoalveolar Lavage

Animals were killed and the tracheas cannulated. Airways were lavaged twice with 0.8 ml ice-cold PBS, and cell numbers and cytokine contents were determined as described previously (26). Cell-free bronchoalveolar lavage fluid (BALF) was frozen at 20°C until analysis.

Determination of BDNF by ELISA

BDNF was measured in cell-free BALF or cell-culture supernatants with commercial ELISA kits according to the manufacturer's instructions (Promega, Madison, WI). U-bottom 96-well plates (Nunc, Wiesbaden, Germany) were coated with monoclonal anti-BDNF antibodies. Captured neurotrophins were detected with secondary chicken polyclonal anti-BDNF and tertiary antibodies conjugated to horseradish peroxidase. Wells were developed with tetramethylbenzidine and measured at 450 nm. Neurotrophin content was quantified against a standard curve generated with known amounts of protein. The range of the linear standard curve was 3.9 to 500 pg/ml. The detection limits were 4 pg/ml for BDNF. Measurements were performed in duplicate and are expressed as means.

Histologic Evaluation of Lung Tissue

For analysis of lung inflammation, lungs were fixed with 4% formaldehyde (wt/vol) via the trachea. The lung was removed and stored in 4% formaldehyde. Paraffin-embedded sections, 3 µm, were stained with hematoxylin and eosin (H&E) for structural evaluation and with periodic acid-Schiff (PAS) for mucus production. For in situ hybridization, 14-µm sections of cryofixed lungs were made, dried for 30 min, and frozen at 80°C until staining.

Electron Microscopy

Lungs were fixed in 5% glutaraldehyde in 0.06 M phosphate buffer, pH 7.3; postfixed in 2% OsO₄ for 2 h; washed three times in phosphate buffer; dehydrated in a graded alcohol series; and embedded in Araldite. Polymerization was carried out at 60°C. Ultrathin sections (70 nm) were stained with uranyl acetate and investigated with a LEO EM 906.

In Situ Hybridization

Preparation of riboprobe. The riboprobe for BDNF was prepared as described elsewhere (27). Briefly, for in vitro transcription, 1 µg of linearized plasmid containing approximately 500 base pairs of the BDNF coding sequence (nucleotides 224-734) was used as a template. The reaction was performed in a 50-µl volume using the DIG-RNA labeling mix from Boehringer Mannheim (Mannheim, Germany) and a T7 (riboprobe) or T3 (sense control) polymerase (Promega). After a 3-h incubation, the reaction was stopped by adding deoxyribonuclease. The probe was hydrolyzed by adding 2 volumes of carbonate buffer (60 mM Na₂CO₃ and 40 mM NaHCO₃, pH 10.2), followed by a 45-min incubation at 60°C. After neutralization with an equal volume of neutralization buffer (200 mM Na-acetate and 1% acetic acid, pH 6), the probe was purified by ethanol precipitation. To estimate the concentration of the probe, a dot blot was performed as described in the manufacturer's manual (Boehringer Mannheim) for nonradioactive in situ hybridization. Probes were stored at 80°C until use.

In situ hybridization on tissue sections. Tissue was prepared and quickly frozen in Tissue Tek (Miles, Elkhart, IN). Slides were coated with 2% TESPA (Sigma). Cryosections, 14 µm, were allowed to dry for 30 min and sections were fixed in 4% cold paraformaldehyde for 10 min, then washed in cold PBS. Acetylation sections were then washed again in PBS. For prehybridization, 500 µl of hybridization buffer (50% formamide, 4× saline sodium citrate [SSC], 2× Denhardt's solution, and 50 µg/ml transfer RNA) were added to each slide. The slides were placed in a humid chamber containing a 50% formamide/4× SSC mix on the bottom. Prehybridization lasted for 1 to 2 h at RT. The hybridization mixture was prepared by adding 150 ng DIG-labeled cellular RNA per milliliter of hybridization buffer. Hybridization was performed overnight at 56°C. Post-hybridization washes were carried out in the following sequence: four times for 10 min in 2× SSC at 67°C, 45 min in 2× SSC at 67°C, 60 min in 0.1× SSC at 67°C, and 10 min in 0.2× SSC at RT. Detection of the DIG-labeled probe was performed as described in the manufacturer's instructions, with the only modification being that the antibody incubation was done overnight at 4°C. Color development was allowed to proceed in the dark for 1 to 2 h. The reaction was terminated by immersing the slides in PBS.

T- and B-Cell Separation by Magnet-Activated Cell Separation Technique

Splenic mononuclear cells (MNC) from control or OVA-sensitized mice were prepared by density centrifugation. MNC from the lung were prepared by digestion with collagenase followed by density centrifugation as previously described (28). T cells were then purified by negative selection with CD11- and CD19-coated magnetic beads by a Super-MACS system (Miltenyi Biotec, Bergisch Gladbach, Germany) as described elsewhere (5). B cells were purified by positive selection with CD19-coated microbeads. T-cell fraction contained > 90% CD3-positive cells as determined by fluorescence-activated cell sorter (FACS) analysis. B-cell fraction contained > 85% CD19-positive cells. Purified cells were washed with PBS and used for cell-culture experiments.

Flow Cytometric Analysis of Cell Surface Marker Expression

MNC were pretreated with lysing buffer (0.15 M NH₄Cl, 0.01 M KHCO₃, and 0.1 mM ethylenediaminetetraacetic acid [EDTA] in distilled water) at RT. After washing with staining buffer (PBS, 0.1 mM EDTA, and 0.02% wt/vol NaN₃), 2 × 10⁵ cells were incubated with the respective antibodies (1.25 µg/ml) (PharMingen, Hamburg, Germany) for 20 min at 4°C in the dark and analyzed on a FACScan analyzer (Becton Dickinson, Mountain View, CA). The following fluorescein isothiocyanate-labeled monoclonal antibodies were used: rat antimouse CD19, CD45, CD- 45RO(B220), Pan NK, and hamster antimouse CD3. Dead cells were detected by propidium iodide (1 µg/ml).

Cell Preparation and Culture Conditions

Purified T and B cells were suspended in RPMI 1640 culture medium supplemented with 10% heat-inactivated fetal calf serum (GIBCO BRL, Eggenstein, Germany), 2 mM L-glutamine (Biochrom, Berlin, Germany), 1.25 µg/ml amphotericin B (GIBCO BRL), 100 U/ml penicillin (Biochrom), and 100 µg/ml streptomycin (Biochrom). Cells were counted in a Coulter counter (Coulter Electronics, Krefeld, Germany). For neurotrophin production, 3 × 10⁶ MNC/well were incubated in 24-well tissue culture plates (Falcon; Becton Dickinson Labware, Le Pont De Claix, France) at 37°C in a 95% air/5% CO₂ humidified atmosphere. Cell samples were stimulated with phorbol ester plus ionomycin (PI) (PDB 10⁷ M, ionomycin 0.5 µg/ml) for 7 d for neurotrophin measurement.

Airway Reactivity

Airway reactivity was assessed by head-out body plethysmography as described elsewhere (29, 30). Briefly, four mice were placed in four body plethysmographs attached to an exposure chamber (Crown Glass, Somerville, NJ). Airflow was measured with a PTM 378/1.2 pneumotachograph (Hugo Sachs Electronics, March-Hugstetten, Germany) and an 8-T2 differential pressure transducer (Gaeltec, Dunvegan, UK). Mid-expiratory airflow (EF₅₀), i.e., the expiratory airflow when 50% of the tidal volume is exhaled, was determined. Changes of EF₅₀ in response to various concentrations of methacholine (2, 4, 6, 8, 12, 16, and 20 mg/ml, 5 min) delivered by a jet nebulizer (Pari-Boy; Pari-Werke, Starnberg, Germany) were assessed. The concentration of methacholine that caused a 50% reduction in mid-expiratory airflow (MCh₅₀) was determined.

Statistical Analysis

Results are presented as mean values ± standard deviation. Means were compared by using analysis of variance (ANOVA) coupled with a Bonferroni correction (ANOVA with SPSS, Version 7.5.2G [SPSS, Inc., Chicago, IL]). P values < 0.05 were regarded as significant differences.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

OVA Sensitization and Ig Synthesis

BALB/c mice sensitized to OVA by intraperitoneal injections developed high levels of OVA-specific IgE and IgG₁ antibody titers in the serum, regardless of whether mice were 2× OVA-aerosol or PBS-aerosol challenged. However, prolonged and repeated exposure to aerosolized OVA twice a week for 4 wk (8×) further enhanced total IgE and anti-OVA IgG₁ levels in the serum in comparison with OVA-sensitized mice exposed to PBS-aerosol (P < 0.05) (Figure 1). In nonsensitized mice (PBS intraperitoneally), anti-OVA antibodies could not be detected in serum.

View larger version (34K): [in this window] [in a new window]   Figure 1.   Ig levels in serum. Total IgE and anti- OVA IgE, IgG1, and IgG2a were measured in serum from nonsensitized or OVA-sensitized mice 24 h after aerosol challenge as described in MATERIALS AND METHODS. Open bars: PBS intraperitoneal/PBS aerosol challenge, lightly shaded bars: PBS intraperitoneal/OVA aerosol challenge, medium-shaded bars: OVA intraperitoneal/PBS aerosol challenge, filled bars: OVA intraperitoneal/OVA aerosol challenge. Asterisks indicate statistically significant difference (P < 0.05). Ten mice were analyzed in each group.

Airway Inflammation

In sensitized mice, the OVA challenge triggered inflammation of the airways that was characterized by influx of inflammatory cells into the airways and an inflammatory cell infiltrate around the airways and blood vessels. After 2× OVA-aerosol challenge, the number of inflammatory cells in BALF was significantly (P < 0.01) higher than in control groups. Repeated challenges in the 8× OVA group resulted in a further increase in the number of inflammatory cells compared with the 2× OVA group. In particular, increased numbers of lymphocytes, eosinophils, and neutrophils were observed after 8× OVA challenge (Figure 2). The number of macrophages was not affected by the OVA challenge; their activation, however, as demonstrated by changes in granulation and size, was increased. More than 95% of cells in BALF of control groups (OVA intraperitoneal/PBS challenge and PBS intraperitoneal/OVA challenge) (Figure 2) were macrophages. In control groups, H&E staining revealed no evidence for airway and parenchymal inflammation. In contrast, a significant eosinophilic and mononuclear peribronchial cell infiltrate was apparent after 2× OVA challenge via the airways in OVA-sensitized mice. This cell infiltrate was augmented after 8× challenge (Figure 3). PAS staining of mucus was negative in control groups, whereas few PAS-positive cells were detected in the airway epithelium of 2× OVA-challenged mice and a strongly hypertrophic, mucus-producing epithelium was present after 8× OVA challenge. Electron microscopic analysis revealed signs of airway remodeling as defined by thickening of the respiratory epithelium and goblet-cell hypertrophy as well as hyperplasia (Figure 3).

View larger version (31K): [in this window] [in a new window]   Figure 2.   Cell differentiation in BALF. Lymphocytes, macrophages, eosinophils, and neutrophils were differentiated according to morphologic criteria in BALF from nonsensitized or OVA-sensitized mice 24 h after the last aerosol challenge as described in MATERIALS AND METHODS. Open bars: PBS intraperitoneal/PBS aerosol challenge, lightly shaded bars: PBS intraperitoneal/OVA aerosol challenge, medium-shaded bars: OVA intraperitoneal/ PBS aerosol challenge, filled bars: OVA intraperitoneal/OVA aerosol challenge. Asterisks indicate statistically significant difference (P < 0.05). Eight mice were analyzed in each group.

View larger version (94K): [in this window] [in a new window]   Figure 3.   Histology of the lung (PAS staining and electron microscopy). Upper panels: Lungs from nonsensitized and OVA-sensitized mice were fixed with paraformaldehyde. Paraffin sections, 3 µm, were stained with PAS. Bars represent 25 µm. A: airway, V: vein. Lower panels: Lungs were fixed in glutaraldehyde and stained with OsO4. Sections, 70 nm, were analyzed by electron microscopy with a LEO EM 906 as described in MATERIALS AND METHODS. Bars represent 2.2 µm. A: airway, BM: basal membrane. Arrowheads indicate mucus-containing vesicles.

To characterize the inflammation, cytokines were measured in BALF by ELISA. The levels of Th2 cytokines interleukin (IL)-4 and IL-5 were increased in OVA-sensitized mice after 2× OVA-aerosol challenge, but not in nonsensitized control animals (Figure 4). In contrast, the level of the Th1 cytokine interferon (IFN)- remained low. After 8× OVA challenge, IL-4 levels returned to baseline levels whereas IL-5 levels remained high and IFN- levels low.

View larger version (19K): [in this window] [in a new window]   Figure 4.   Cytokine levels in BALF. IL-4, IL-5, and IFN- were measured in BALF from nonsensitized or OVA-sensitized mice 24 h after aerosol challenge by ELISA. Asterisks indicate statistically significant difference (P < 0.05). Eight mice were analyzed in each group.

Development of AHR

To correlate the inflammatory response with changes of the breathing pattern as they occur in asthma, the airway responsiveness was measured by body plethysmography (29, 30) in response to methacholine and expressed as MCh₅₀ (31). AHR was already detected after 2× OVA-aerosol challenge and remained on this level after 8×. The airway inflammation and the histologic changes in sensitized mice were accompanied by a decrease in MCh₅₀ after repeated OVA-aerosol challenge (Figure 5).

View larger version (31K): [in this window] [in a new window]   Figure 5.   Airway responsiveness to methacholine. Methacholine challenges were performed with increasing concentrations 24 h after the last OVA challenge. MCh50 was calculated. C: nonsensitized, PBS-challenged mice; 8×: OVA-sensitized, OVA-challenged animals. Asterisks indicate statistically significant difference (P < 0.05). Six mice were analyzed in each group.

Production and Source of BDNF in the Inflamed Lung

Concentrations of BDNF were measured in cell-free BALF. After 2× challenge there was a 2-fold increase in BDNF levels in OVA-sensitized mice in comparison with control groups (Figure 6). However, highest BDNF levels were detected after 8× challenge. For stimulation of BDNF production, OVA sensitization plus OVA-aerosol challenges were required; each event by itself had no effect on BDNF levels in BALF (Figure 6). BDNF was not detectable in serum of any mouse group (data not shown).

View larger version (21K): [in this window] [in a new window]   Figure 6.   BDNF levels in BALF. BDNF was measured in BALF from nonsensitized or OVA-sensitized mice 24 h after aerosol challenge by ELISA. Asterisks indicate statistically significant difference (P < 0.05). Eight mice were analyzed in each group.

To localize the cellular source of BDNF, cytospin preparations were stained for BDNF mRNA by in situ hybridization. Cells from control animals were negative for BDNF mRNA (Figure 7). In contrast, BDNF mRNA- positive macrophages were detected in BALF from OVA-sensitized and -challenged mice with strongest signals after 8× challenge. Except for macrophages, all other cell types were BDNF mRNA-negative.

View larger version (90K): [in this window] [in a new window]   Figure 7.   BDNF mRNA staining by in situ hybridization in cytospin preparations of BALF. Bronchoalveolar lavage cells from nonsensitized and OVA-sensitized mice were stained by nonradioactive in situ hybridization technique with a specific antisense-BDNF probe (nucleotides 224-734). Nonbinding sense probe was used as a control. Cells were counterstained with H&E. Bars represent 25 µm. Arrows indicate macrophages.

Cryosections of the lung (14 µm) were analyzed for BDNF mRNA expression by in situ hybridization. Constitutive expression of BDNF mRNA was detected in airway epithelium (Figure 8). In inflamed lungs, BDNF mRNA- positive peribronchial cellular infiltrates were observed that were not present in control animals. Sense controls were always negative.

View larger version (132K): [in this window] [in a new window]   Figure 8.   BDNF mRNA staining by in situ hybridization in lung sections. Shown are lung sections (14 µm) from nonsensitized mice after PBS challenge (A), OVA-sensitized mice after 2× OVA challenge (B), and OVA-sensitized mice after 8× OVA challenge (C). Sections were stained by nonradioactive in situ hybridization with a specific antisense-BDNF probe (nucleotides 224-738). A nonbinding sense probe was used as control. Shown is a sense staining of a sequential section of the 8× OVA-challenged lung (D). Bars represent 50 µm. Arrowheads indicate epithelial cells; arrows show cells of the peribronchial inflammatory infiltrate.

Because the peribronchial cell infiltrate consists mainly of lymphocytes, we investigated whether lymphocytes could be an additional source of BDNF. BDNF production was measured in cell-culture supernatants of MNC and purified splenic T or B cells. Unstimulated or lipopolysaccharide (LPS)-stimulated lymphocytes did not produce detectable amounts of BDNF. Stimulation with PI induced BDNF production in cells from OVA-sensitized mice (Figure 9A), but not in cells from control mice (data not shown). Only T cells, but not B cells, could be stimulated to produce BDNF (Figure 9A).

View larger version (16K): [in this window] [in a new window]   Figure 9.   BDNF production in separated T and B cells from OVA-sensitized and -challenged mice. T and B cells were separated by magnet-activated cell separation (MACS) technology as described in MATERIALS AND METHODS. (A) Not-separated splenic MNC; T or B cells were incubated with medium alone or stimulated with PI. (B) T and B cells were isolated from the inflamed lung, separated by MACS, and incubated with medium for 7 d. It was not possible to isolate a sufficient number of lymphocytes from not-inflamed lungs of control mice. BDNF was measured in cell culture supernatants by ELISA after 7 d of culture.

To assess whether lymphocytes from the lung produce BDNF, T and B cells were isolated from lung digests of OVA-sensitized and -challenged mice. T cells released BDNF into the cell-culture supernatants without further stimulation, whereas BDNF could not be detected in B-cell cultures (Figure 9B). In vitro activation of MNC with LPS or PI did not further increase BDNF synthesis in this condition (data not shown).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Neurotrophins have recently been described as playing a critical role in the pathogenesis of allergy and asthma (12, 13). Because these reports focused on NGF, little is known about the role of other neurotrophins in this condition. We have demonstrated previously that in addition to NGF, BDNF and NT-3 production are enhanced in human lung after allergen challenge in patients with allergic asthma (14). Therefore, we investigated the source of enhanced BDNF production in a murine model of allergic inflammation. As shown in humans (14), BDNF levels were upregulated in BALF from mice after allergen challenge. Possible sources of enhanced BDNF production are activated epithelial cells as well as infiltrating immune cells.

This is the first report showing BDNF synthesis in activated immune cells in allergic inflammation. Beyond a production of BDNF by epithelial cells, we demonstrated BDNF production by activated macrophages and T cells. Resting macrophages did not express detectable levels of BDNF mRNA. In contrast, activation in response to allergen induced BDNF expression in these cells. T cells were identified as an additional source of BDNF in the inflamed lung; however, there were insufficient numbers of T cells in control lungs to analyze their BDNF production. Therefore, it cannot be excluded that BDNF is physiologically produced by lung T cells. However, unstimulated T cells from the spleen did not produce detectable amounts of BDNF. These results suggest that BDNF is induced by inflammatory stimuli in distinct immune cells. These findings have been confirmed by a recent report showing BDNF production in human T cells, B cells, and macrophages in multiple sclerosis (32). The previously reported BDNF production in mast cells (22) underlines a possible role for BDNF in allergic diseases. A wide range of hematopoetic cells, including mast cells (33), macrophages (34), T cells (35), and B cells (36), has been shown to produce the neurotrophin NGF. We demonstrate here that the closely related neurotrophin BDNF is produced in allergic inflammation by T cells and macrophages, but not by B cells. These results do not exclude BDNF production by murine B cells under other conditions, as has recently been described for human B cells (32), but activation of murine B cells by allergen, PI, or LPS is not sufficient to induce significant BDNF production. Because lymphocytes and macrophages were also identified as sources of NGF in allergic airway inflammation (12), we suggest a similar regulation of NGF and BDNF production in this condition.

To investigate the production of the neurotrophin BDNF in allergic immune responses, we used a murine model of chronic and severe airway inflammation. This type of inflammation was induced by repeated allergen challenges of sensitized BALB/c mice. After 8× OVA-aerosol challenge, the high levels of IL-5 were measured in BALF together with a marked lymphocytic and eosinophilic infiltration. In addition, the airway epithelium was thickened and showed remodeling into mainly mucus-producing tissue. In addition, these mice demonstrated AHR. However, AHR in response to methacholine was not further enhanced by repeated allergen challenges. These results suggest that, at least in this model, the degree of airway inflammation is not necessarily correlated to the extent of AHR. One reason for this observation may be biologic saturation effects. However, our murine model of allergic airway inflammation mirrors several of the characteristic features of allergic bronchial asthma (4, 37, 38).

Airway smooth-muscle tone is controlled by sympathetic and parasympathetic nerves as well as by the nonadrenergic noncholinergic (NANC) neurons. Several studies have emphasized the importance and the plasticity of the NANC system in allergic diseases (8, 39). The axon reflex of sensory neurons especially has been described as playing a crucial role in this condition (9). Possible mediators linking inflammation and nerve function are neurotrophic factors (40). NGF has been shown to play a critical role in the neuroimmunologic crosstalk occurring in allergic asthma (12, 13). Recently, an upregulation of neurotrophins including BDNF was shown in asthmatics following segmental allergen provocation (14). A possible role of BDNF in neuropeptide upregulation has been described in the central as well as the peripheral nervous system (15, 41, 42). In addition, BDNF was shown to increase excitability of motor neurons (43). BDNF expression after peripheral inflammation was previously described in neurons of dorsal root ganglia (44). Production of BDNF was also shown in neurons and Schwann cells after peripheral nerve lesion. Interestingly, the upregulation of BDNF displayed a delayed time course with maximal levels 3 to 4 wk after lesion (45). An increase of BDNF mRNA has been reported in bladder extracts after turpentine oil induced inflammation (46).

Because most of the vagal neurons innervating the lung are BDNF-dependent (24, 47), local production of BDNF in the allergic inflammation may be involved in neuronal changes occurring in allergic airway inflammation and bronchial asthma. Those neuronal changes include an altered neuropeptide (tachykinin) expression in lung-innervating sensory neurons (11, 48). Neuropeptides have many effects that could be relevant for the pathogenesis of bronchial asthma, including airway smooth-muscle contraction, mucus hypersecretion, facilitation of acetylcholine release, plasma extravasation, and inflammation (9). Our data suggest that BDNF may serve as a mediator linking airway inflammation with neuronal changes occurring in allergic asthma.