A total of 32 specific pathogen-free normal female Wistar rats (weight, 120–140 g) were obtained from the Laboratory Animal Research Center in Shengjing Hospital of China Medical University (Shenyang, China). Rats were maintained in a 12 h light:dark cycle with access to food and water ad libitum. Initially, the rats were randomly divided into four groups: control, ovalbumin (OVA), NGF and anti-NGF (n=8 per group). All animal experiments were performed in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals (8th Edition, 2012). The animal-use protocol has been reviewed and approved by the Institutional Animal Care and Use Committee of the Shengjing Hospital of China Medical University (Shenyang, China).

Sensitization and challenge protocols were performed according to the methods of Li and Shang (11) and Vanacker et al (12) with certain modifications as described below. On days 0 and 7, all rats with the exception of those in the control group were actively sensitized with an intraperitoneal (i.p) injection of 1 mg OVA (Grade V; Sigma, St. Louis, MO, USA) and 200 μg aluminum hydroxide (Sigma) in 0.5 ml sterile phosphate-buffered saline (PBS). The OVA-sensitized rats were exposed to 1% aerosolized OVA (1 g OVA in 100 ml sterile PBS in a nebulizer) for 30 min, every two days from day 14 to day 70. As neurotrophic factors are highly conserved in different species, we used exogenous murine NGF (NGF-7S; Alomone Labs, Jerusalem, Israel) and blocked endogenous NGF activity using 100 μg/ml goat anti-rat-β-NGF antibody (R&D Systems, Minneapolis, MN, USA) for our rat model. The NGF and anti-NGF groups were administered an i.p injection of NGF-7S (80 ng/kg) or anti-NGF antibody (4 ml/kg) diluted at 1:1,000 in sterile PBS, respectively, 3 h prior to the OVA aerosol challenge. The administration route, timing and dose of the NGF and anti-NGF treatments were chosen based on a previous study (13). The OVA group received 4 ml/kg PBS 3 h prior to the OVA aerosol challenge. The control group was subjected to the same protocol using sterile PBS.

Airway reactivity to methacholine (MCH) was assessed in vivo 24 h following the last OVA challenge as previously described (14). Rats were anesthetized with 100 mg/kg pentobarbital sodium (i.p), a tracheal cannula was inserted via tracheotomy for mechanical ventilation and a small catheter (22G) was inserted into the external jugular vein for the administration of MCH (Sigma-Aldrich, Beijing, China). The rats were then placed in a sealed whole body plethysmograph and connected to a rodent ventilator (ML-V2; Shanghai Benda Biotechnology Co., Ltd., Shanghai, China). Ambient air was administered with a tidal volume of 8 ml/kg and a frequency of 90 strokes per min. Transducers (ML-AMP II; Shanghai Benda Biotechnology Co.,Ltd.) connected to the ventilatory circuit provided voltage signals of pressure and flow, which were amplified and transmitted to the analog/digital card (National Instruments, Austin, TX, USA) of a microcomputer running the AniRes2005 software (Beijing Bestlab High-Tech Co., Ltd., Beijing, China), which was used to calculate the inspiratory and expiratory resistances of the respiratory system from the digitized pressure and flow signals. Following stabilization of respiratory parameters (10–15 min) rats received MCH (dissolved in 0.9% sodium chloride) intravenously at an initial dose of 0.0625 mg/kg with the dose increasing 2-fold with each injection up to 1 mg/kg to obtain a response curve of lung resistance increase over baseline. Injections were administered at 5-min intervals. The MCH volume was 50 μl, which was administered over 3–4 sec based on the return of resistance curves to the pre-MCH level prior to the next MCH injection. Response was measured as the peak increase above the baseline immediately following MCH administration.

Following assessment of airway reactivity, the rats were bled and sacrificed via anesthetic overdose. BAL was performed in the left lungs. The left lungs were washed thrice with 1 ml saline. Lavage fluid was recovered by gentle manual aspiration with a syringe. The retrieved volume, which averaged 75–80% of the instilled saline was immediately centrifuged (10 min, 4°C, 1,000 × g) and the supernatant was stored at −70°C until the levels of IL-4 and IL-13 were measured. The pellet was kept on ice, washed twice with saline and resuspended in 1 ml saline. The total number of leukocytes in the BALF were determined with a Coulter counter (Coulter Electronics Ltd., Harpenden, UK). A differential cell count was performed on Cytospin (Thermo Shandon, Inc,, Pittsburgh, PA, USA) by Wright-Giemsa staining.

After measuring airway reactivity, serum was obtained by lethal cardiac puncture of anesthetized rats and stored at −70°C for measurement of OVA-specific IgE by ELISA.

Lung tissues were weighed (total lung weight) and then separated into individual lobes for hydroxyproline analysis, histological analysis, western blotting, quantitative polymerase chain reaction (qPCR) and MMP zymography.

The middle lobes of the right lung were fixed in 4% paraformaldehyde for 18–24 h, embedded in paraffin and then routinely processed. Serial 5-μm tissue sections were stained with hematoxylin and eosin (H&E), Masson’s trichrome and periodic acid-Schiff (PAS) for the assessment of peribronchial inflammation, collagen particles and goblet cells, respectively.

A minimum of five bronchi (luminal diameter, 150–350 μm) were analyzed per rat for various parameters using a Leica image analysis system (Leica, Cambridge, UK). Using a 10-fold magnification objective, four representative areas were chosen. Subsequently, with a 40-fold magnification corresponding to one microscopic field, hyperplasia of the goblet cells in the epithelial lining was recorded against a score based on the percentage of goblet cells observed in the epithelial cells. The length of the epithelial basement membrane of the bronchus of one area was ≥500 μm. To minimize sampling errors, a 5-point scoring system (grades 0–4) was adopted: grade 0, no goblet cells; grade 1, <25% goblet cells; grade 2, 25–50% goblet cells; grade 3, 50–75% goblet cells; and grade 4, ≥75% goblet cells (15). The mean score of the total epithelial cells in the four areas of one rat were counted. The mean score of hyperplasia of the goblet cells was calculated for 7–8 rats.

Masson’s trichrome-stained tissue sections were used for the assessment of subepithelial fibrosis using the Leica image analysis system. As described previously, three representative areas were chosen using a 10-fold magnification objective. Subsequently, with a 40-fold magnification, epithelial basement membrane areas of ≥250 μm were selected, and the thickness of the epithelial layer and fibrotic area (stained in blue), which were 30 μm beneath the basement membrane of the standardized sampling points, were measured (16). The mean fibrotic area divided by the length of the basement membrane were calculated for 7–8 rats.

Total collagen content in the fresh lung samples of all rats were determined by hydroxyproline assay. The hydroxyproline content was detected with a commercial hydroxyproline detection kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) following the manufacturer’s instructions.

Protein homogenates of lung tissue samples were prepared by rapid homogenization in 10 volumes of lysis buffer (2 mM EDTA, 10 mM EGTA, 0.4% NaF, 20 mM Tris-HCl, 1 mg/ml leupeptin, 1 mg/ml aprotinin and 1 mM Na₃VO₄ at pH 7.5). Samples were centrifuged at 12,000 × g for 1 h and the protein concentration of the soluble material was determined using the Coomassie Brilliant Blue G250 (Beyotime Institute of Biotechnology, Jiangsu, China) binding method (17). Proteins (10 μg) from each sample were loaded onto an 8% sodium dodecyl sulfate-polyacrylamide gel. Electroblotted proteins were transferred from the gel to nitro-cellulose membranes, which were incubated with 1:1,000 goat anti-NGF. The NGF band (140 kD) was visualized using an enhanced chemiluminescence kit (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). Integrated density values (IDV) were analyzed using a computerized image analysis system (Fluor Chen 2.0; Bio-Rad, Hercules, CA, USA) and normalized to those of β-actin.

Rat lung tissues were dissected and stored in TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA). The levels of MMP-9 mRNA were determined using the ABI PRISM 7500 Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). The RNA was extracted using TRIzol reagent in accordance with the manufacturer’s instructions. RNA purity was determined and cDNA synthesis was conducted using a SYBR^® PrimeScript™ RT-PCR kit (Takara Biotechnology, (Dalian) Co., Ltd., Dalian, China). The volume of rat MMP-9 mRNA was determined by qPCR. The primers for MMP-9 were 5′-CCCACTTACTTTGGAAACG-3′ (forward) and 5′-GAAGATGAATGGAAATACGC-3′ (reverse), and those for GAPDH were 5′-GCAAGTTCAACGGCACA-3′ (forward) and 5′-CATTTGATGTTAGCGGGAT-3′ (reverse) [Takara Biotechnology (Dalian) Co., Ltd]. Gene expression levels of MMP-9 were analyzed by the 2^−ΔΔCT method (18).

Total MMPs were extracted from a similar section of each lung tissue and analyzed by gelatin zymography (Genmed Scientifics Inc., Arlington, MA, USA) according to the manufacturer’s instructions. The resulting bands on the zymograph were analyzed by densitometry using a GS-710 densitometer (Bio-Rad Laboratories, Richmond, CA, USA) and Quantity-One software (Bio-Rad). The mean ± standard error density of each MMP was plotted in a graph and expressed as the relative ratio of the values in the control group, which were expressed as one.

Data are expressed as the mean ± standard error of the mean for each group. One-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test was used to compare group differences, whereas two-way ANOVA was used to assess differences in airway resistance. P<0.05 was considered to indicate a statistically significant difference.

All experimental rats exhibited dose-dependent augmentation of inspiratory and expiratory resistance in response to increasing doses of intravenously administered MCH. No statistically significant differences in baseline inspiratory and expiratory resistance values among all experimental groups was observed. The airway response to MCH significantly increased in the OVA and NGF groups compared with that of the control group (P<0.01). Treatment of OVA-sensitized rats with anti-NGF prior to the OVA challenge prevented the increase in airway reactivity, which was reflected by a shift in the dose-response curves (Fig. 1) and significantly decreased reactivity compared with that of the OVA group (P<0.01).

To investigate the function of NGF in antigen-induced inflammatory infiltration in the airways, we examined the effects of NGF and anti-NGF antibodies in the asthmatic rat model. As shown in Table I, the number of total cells, eosinophils, macrophages, lymphocytes and neutrophils increased in the BALF of the untreated OVA-sensitized and NGF-treated rats compared with those in the PBS-treated rats (controls) (Table I). By contrast, anti-NGF treatment significantly inhibited the increase in the number of total leukocytes, eosinophils and lymphocytes in the BALF following antigen inhalation.

To evaluate the effects of NGF on antigen-induced goblet cell hyperplasia in the airway epithelium, which is a cardinal feature of bronchial asthma, lung sections were stained with PAS for detection (Fig. 2). Goblet cell hyperplasia was then quantitatively estimated in terms of grade as described in Materials and methods (Fig. 2E). As shown in Fig. 2A, histological analyses of lungs from the PBS-treated rats showed normal lung histology. However, the number of goblet cells in the epithelium greatly increased following repeated antigen challenge (Fig. 2B), and goblet cells in the epithelium showed hypertrophic features. Antigen-induced goblet cell hyperplasia was greatly increased by NGF administration (Fig. 2C), but significantly reduced by anti-NGF treatment (Fig. 2D).

To determine whether NGF is involved in the development of airway remodeling, we evaluated the peribronchial cellular infiltration, airway smooth muscle thickness and lung collagen content in all experimental rats. Representative sections of each group were stained with either H&E (Fig. 3A–D) or Masson’s trichrome (Fig. 3E–H). As shown in Fig. 3A, no inflammation, mucosal edema and epithelial lesions were observed in the control group. Moderate inflammation, mucosal edema and epithelial lesions were observed in the anti-NGF group (Fig. 3B), whereas sensitization in the OVA group was associated with predominantly moderate to severe inflammation, mucosal edema and epithelial lesions (Fig. 3C). In particular, only NGF-treated rats developed severe inflammation, mucosal edema and epithelial lesions, which included interstitial infiltrates and large lymphoid aggregates (Fig. 3D). Peribronchial fibrosis (Fig. 3I) and lung collagen content (Fig. 3J) were quantitatively estimated. Exposure to OVA increased the peribronchial trichrome-stained area and lung collagen levels in the OVA (Fig. 3E) and NGF groups (Fig. 3F) when compared with those in the control group (Fig. 3G). Anti-NGF-treated rats showed reduced peribronchial collagen staining (Fig. 3H) compared with that in the OVA-treated rats. Lung collagen levels in the anti-NGF-treated rats were significantly reduced (57.6±1.19 μg collagen/g lung tissue) compared with those in the OVA-treated rats (75.23±3.64 μg collagen/g lung tissue) (P<0.01).

We analyzed the concentrations of Th2-associated cytokines (IL-4 and IL-13) in the BALF. As shown in Table II, the IL-4 and IL-13 levels were higher in the BALF of the untreated OVA-sensitized and NGF-treated rats compared with those of the PBS-treated rats (controls). Anti-NGF-treated rats showed reduced levels of IL-4 and IL-13 compared with those of the untreated OVA-sensitized rats.

To confirm that the presence of the OVA antigen resulted in immunological sensitization, OVA-specific IgE serum levels were measured. Serum IgE levels were increased in all OVA-sensitized groups compared with those of the control group. However, serum IgE levels were reduced in the anti-NGF group compared with those of the OVA group (Table II).

NGF protein levels in all groups were detected by western blotting. In the lung tissues of the control group, NGF was expressed at low levels, but these levels were markedly increased in the OVA and NGF groups. However, anti-NGF treatment markedly decreased the NGF upregulation caused by OVA and NGF (Fig. 4). The mean IDV ratio of NGF to β-actin was 0.36±0.01 in the control group, 0.62±0.02 in the OVA group, 0.50±0.02 in the anti-NGF group and 0.81±0.02 in the NGF group (Fig. 4).

qPCR was used to examine the expression levels of MMP-9 mRNA in the pulmonary tissues of all groups. MMP-9 mRNA expression levels were significantly higher in the OVA and NGF groups than in the control group, but anti-NGF treatment significantly decreased MMP-9 mRNA expression levels in the OVA-sensitized rats (Fig. 5).

Gelatin zymography of the lung tissues from the rats in all experimental groups was used to demonstrate changes in the activity of MMP-9. Densitometric analysis of the zymographs demonstrated that MMP-9 activity in the lung tissues of all OVA-sensitized rats was significantly increased compared with that of the controls (Fig. 6). By contrast, a significant reduction in MMP-9 activity in the lung tissues of the anti-NGF-treated rats compared with that in the OVA-treated rats was observed.