Male A/J mice (23–27 g) were purchased from Harlan (Indianapolis, Indiana) and housed in a conventional animal facility at McGill University. Animals were treated according to guidelines of the Canadian Council for Animal Care and protocols were approved by the Animal Care Committee of McGill University.

Forty-eight mice were exposed to either room air (control) or 800 ppm Cl₂ gas diluted in room air for 5 minutes using a nose-only exposure chamber. This concentration of Cl₂ gas was chosen as it was previously shown to result in severe airway damage but with minimal animal mortality9. Mice exposed to Cl₂ were studied at 12 hrs, 24 hrs, 48 hrs, 5 days (d), or 10 d after Cl₂ exposure (n = 8 at each time point). The control mice were studied 24 hrs after exposure to room air (n = 8).

The chest was opened, the left main bronchus clamped, and 0.3 ml of sterile saline followed by four separate 0.5 ml instillations were washed into the right lung. Fluid recovered from the first wash was centrifuged at 1500 rpm for 5 minutes at 4°C and the supernatant used for protein quantification. The cell pellet was pooled with the remaining lavage samples and total live and dead cells were counted using trypan blue exclusion. Cytospin slides were prepared using a cytocentrifuge (Shandon, Pittsburgh, PA) and stained with Dip Quick (Jorgensen Labs Inc., Loveland, CO). Differential cell counts, including epithelial cells, were determined on 300 cells/slide. Total protein in the BAL supernatant was quantified using a dye-binding colorimetric assay (Bio-Rad, Hercules, CA), and determined by spectrophotometry at 620 nm and quantified using a bovine serum albumin standard curve.

Following BAL, the lungs were removed and the left lung was fixed with an intratracheal perfusion of 10% buffered formalin at a constant pressure of 25 cmH₂O for a period of 24 hrs. Histology and immunohistochemistry were performed on 5 μm thick paraffin-embedded sections taken from the parahilar region. Adjacent sections were either stained with hematoxylin-eosin (H&E), periodic acid Schiff (PAS), or processed for immunohistochemistry.

Cells undergoing proliferation were detected in tissue sections by immunostaining for proliferating cell associated nuclear antigen (PCNA. Following deparaffination in xylene and rehydration through graded ethanol solutions, the tissue sections underwent a high temperature epitope unmasking treatment by a modified version of the microwave boiling method. An acidic antigen retrieval buffer (Vector Laboratories, Burlingame, CA) was microwave pre-heated to 95°C, and the slides were incubated in it for 30 minutes using a pre-warmed coplin jar protected with styrofoam. After cooling for 20 minutes, a membrane permeabilization treatment was applied by immersing the slides for 20 minutes in a 0.2% dilution of Triton X-100 (Sigma Chemical Co., St. Louis, MO) in pH 7.6 Trizma base (Sigma) buffered saline. The tissues were then blocked for 1 hour using a blocking reagent designed for immunohistochemistry using mouse primary antibodies on mouse tissues (Vector Laboratories). Primary murine anti-PCNA antibody was applied at a concentration of 2.5 μg/ml and the sections were incubated for 30 min. at room temperature. A biotinylated anti-mouse antibody (1:250 dilution; Vector Laboratories) was applied for 10 min. followed by a 45-min. incubation with an avidin-biotin complex-alkaline phosphatase reagent (ABC-AP). Rat intestine was used as a positive control and mouse lung sections incubated with isotype control mouse IgG were used as a negative control. PCNA-positive cells were visualized with Vector Red chromogen (Vector Laboratories) and the tissue was counterstained using methyl green (Sigma). Finally, the sections were dehydrated and mounted under glass coverslips with VectaMount (Vector Laboratories).

To determine the amount of airway smooth muscle by morphometry, airway smooth muscle was detected by immunostaining for smooth muscle α-actin. The lung sections were prepared as described above with the exception of high temperature antigen unmasking, and incubated with monoclonal antibody to smooth muscle α-actin (1A4, 1:1000 dilution; Sigma) for 30 minutes followed by biotinylated anti-mouse IgG antibody and ABC-AP steps as above.

PCNA was colocalized with smooth muscle α-actin in order to detect cell proliferation in the airway smooth muscle. Immunohistochemistry for PCNA was done first as described above, and the signal developed with BCIP/NTB chromogen (Vector Laboratories) instead. The sections were then incubated with anti-smooth muscle α-actin antibody (1A4, 1:1000 dilution, Sigma) for 30 min. at 37°C, followed by the biotinylated anti-mouse antibody and ABC-AP steps as above. The smooth muscle α-actin signal was developed with Vector Red, and the tissues counterstained with methyl green.

To detect apoptotic cells in lung tissue sections we used a TUNEL technique (ApopTag peroxidase detection kit; Intergen, Purchase, NY). The sections were deparaffinized, pretreated with 20 μg/ml proteinase K (Intergen) for 15 min at 37°C, and endogenous peroxidase activity was quenched with 3% hydrogen peroxide for 5 min This was followed by polymerization of digoxigenin-labeled UTP on nicked DNA ends and application of anti-digoxigenin peroxidase conjugate, using ApopTag kit components as per manufacturer's instructions. The signal was developed with DAB chromogen, and the tissues counterstained with methyl green.

Quantification of PCNA-positive cells was performed on parahilar lung sections. Cross-sectioned airways, with a major/minor diameter ratio < 2.5, were selected for analysis. The number of PCNA⁺ cells in the epithelium and sub-epithelial layers were quantified under a light microscope using a 40× objective. The airway basement membrane length was measured by superimposing the image of the airway onto a calibrated digitizing tablet (Jandel Scientific, Chicago, IL), with a microscope equipped with a camera lucida projection system (Leica Microsystems, Richmond Hill, ON, Canada). The numbers of proliferating cells corrected for airway size were expressed as PCNA⁺ cells/mm of basement membrane perimeter (P_BM).

ASM mass was measured on control, 5 d, and 10 d post-exposure groups by tracing the ASM bundles, as defined by positive staining for smooth muscle α-actin, using a camera lucida and digitizing system. The sum of the ASM bundle areas was calculated for each airway and referenced to P_BM² for airway size correction. To determine if airway smooth muscle cells expressed PCNA, co-localization of PCNA with smooth muscle α-actin was done in a subset of animals. The number of PCNA+ cells in the epithelial and sub-epithelial layers of each airway with a major/minor diameter ratio < 2.5 was quantified and expressed per mm of P_BM for epithelium or P_BM² for subepithelial cells.

The number of goblet cells was assessed on PAS stained tissue sections. A total of 118 airways from 28 animals representing animals from the different exposure times was analyzed and cells were expressed as cell numbers per mm of P_BM.

To address whether chlorine exposure could affect the development of subepithelial fibrosis, lung sections were stained with Picrosirius red and collagen deposition scored in airways. Scoring by two blinded observers of collagen deposition in airways was performed independently using a scale from 1 to 3. The cumulative score for each mouse was averaged according to treatment group.

The quantity of airway smooth muscle (ASM) was quantified by the camera lucida technique. Images of the airways were traced using a microscope side arm attachment and areas of the α-actin positive smooth muscle bundles were digitized using commercial software. The area of ASM was standardized for airway size using the P_BM, with the quantity of ASM expressed as ASM/P_BM² (mm²). Morphometric assessments were made on all airways in the tissue section that met the above criterion for its aspect ratio.

In a separate group of sixty mice, airway responsiveness to methacholine was measured at similar time points after room air or Cl₂ exposure (n = 10 at each time point). Animals were sedated with xylazine hydrochloride (10 mg/kg i.p.) and anaesthetized with sodium pentobarbital (40 mg/kg i.p). A flexible, saline-filled cannula (PE-10 tubing) was inserted into the jugular vein for administration of drugs and the trachea was cannulated with a snug-fitting metal cannula. Animals were connected to a computer-controlled small animal ventilator (flexiVent, Scireq, Montreal, PQ, Canada) and paralysed using pancuronium chloride (0.8 mg/kg i.v.). Mice were ventilated in a quasi-sinusoidal fashion with a tidal volume of 0.18 ml at a rate of 150 breaths/min. A positive end-expiratory pressure (PEEP) of 1.5 cmH₂O was used. Measurements of pulmonary mechanics were made using a 2.5 Hz sinusoidal forcing function with an amplitude of 0.18 ml. The perturbation was applied after cessation of regular ventilation and expiration by the animal to functional residual capacity. Respiratory system resistance (Rrs) and dynamic elastance (Ers) was derived from the relationship between airway opening pressure, tidal flow and volume After initial baseline measurements of Rrs and Ers, doubling doses of methacholine chloride (Sigma;10 μg/kg to 320 μg/kg i.v.) were administered. Rrs and Ers were measured every 15 seconds after methacholine infusion until peak Rrs was reached. Thirty seconds after peak Rrs was reached, the next highest dose of methacholine was administered. The peak Rrs and Ers at each methacholine dose were used to construct a dose-response curve. After completion of all methacholine doses, animals were euthanized by i.v. pentobarbital overdose. Airway responses were evaluated as the difference between the peak in Ers after 160 μg/kg methacholine and baseline Ers (ΔErs). Changes in Ers rather than Rrs were chosen to represent airway responsiveness because methacholine-induced changes in elastance are affected to a greater degree in mice after Cl₂ exposure9.

One-way analysis of variance was used to determine the effect of time on the dependent variables except ASM/mm². The significance of the post-hoc comparisons was determined using Dunnett's test versus control at the p < 0.05 level. The effect of Cl₂ on ASM/P_BM² (in mm²) at different times after exposure was tested using the Kolmogorov-Smirnoff test.

Normal airway structure and basal levels of proliferation and apoptosis in airway epithelium are shown in Figures 1A, 2A, 3A. Histological examination from samples obtained 12 hrs after exposure showed severe injury to the bronchial epithelium with extensive detachment of the epithelium from the basement membrane and complete denudation of the epithelium in some airways (Figure 1B). Cell cycle was inhibited at this time point after chlorine exposure, as indicated by the virtual absence of positive staining for PCNA (Figure 2B). The TUNEL technique produced cytoplasmic staining of the injured epithelium, but not a signal conforming to usual histopathological criteria for the identification of apoptosis, suggesting that a mechanism other than apoptosis accounts for the rapid and massive epithelial disaggregation following Cl₂ gas exposure (Figure 3B). At 24 hrs after Cl₂ exposure, most of the detached airway epithelial cells were cleared and airway epithelial cell proliferation was re-established (Figure 3C). In this phase, some clusters of basal cells undergoing apoptosis alternated with proliferating cells, overlying a preserved basement membrane (Figure 3D). Epithelial regeneration was evident at 48 hrs with flattened cells with elongated nuclei lining the basement membrane and an increased frequency of PCNA positive cells. Co-localisation of PCNA and smooth muscle α-actin provided evidence of airway smooth muscle proliferation (Figure 2F). Five days following chlorine exposure, the airway epithelium was evenly re-populated with cells showing an intense proliferative activity, and the frequency of apoptotic cells was similar to baseline levels. Ten days after chlorine exposure, the epithelium was reconstituted and the airway wall was thickened (1 D). Cl₂ exposure did not induce goblet cell metaplasia as determined by PAS staining at any time point (data not shown). Only 4 of 118 airways analyzed from 28 mice, sampled at all time points showed any PAS positive cells and these were very infrequent.

Cl₂ exposure did affect the quantity of ASM as determined by morphometry (Figure 4). 10 days after Cl₂ exposure, a shift was observed in the distribution of airways with small amounts of ASM. For example, the proportion of airways with values of ASM area > 0.0015 (ASM/mm² of BM) was approximately 50% for control animals, but < 10% for the 10 day post-exposure group.

The number of PCNA+ cells in the airway epithelium and sub-epithelium is shown in Figure 5. A baseline frequency of epithelial and sub-epithelial proliferation was detectable in control animals. Twelve hours after Cl₂ exposure, epithelial PCNA expression tended to be lower than control values although the difference did not reach statistical significance. Epithelial PCNA expression was significantly elevated by 48 hrs after chlorine exposure, increasing approximately 14-fold from control levels (p < 0.05) and over 30-fold by 5 d post-exposure (p < 0.05). Although the majority of the PCNA+ cells in the airways were epithelial cells, a significant amount of sub-epithelial PCNA expression was also observed after Cl₂ exposure. Subepithelial PCNA expression was significantly elevated at 5 d post-exposure. By 10 d post-exposure, both epithelial and subepithelial PCNA immunoreactivity had returned to control levels. No significant correlation was found between airway size (as determined by basement membrane length) and PCNA index at any of the time points.

Assessment of collagen deposition using Picrosirius red staining demonstrated a significant increase in collagen in the airways 10 days following chlorine exposure (Figure 6). There was no significant difference in the amount of collagen at 24 hours or 5 days. Twenty nine animals were analyzed and assessed by two observers independently.

The recovery of BALF averaged 90% and did not differ significantly among groups. Total cell counts were significantly elevated at 5 d and remained elevated at 10 d post-exposure relative to controls (Table 1). Differential cell counts showed no significant change in eosinophils or lymphocytes after Cl₂ exposure (Figure 7), but neutrophils were significantly elevated relative to controls at 5 d post-exposure (0.02 ± 0.01 (SE) × 10⁴ cells in controls, 4.76 ± 1.94 at 5 d post-exposure; p < 0.05) and macrophages were significantly elevated at both 5 d and 10 d post-exposure (12.0 ± 1.9 × 10⁴ in controls, 32.2 ± 7.7 at 5 d, 33.7 ± 3.3 at 10 d, p < 0.05 versus controls). Dead cells in the BALF, identified by trypan blue, were markedly elevated from 12 hrs to 48 hrs post-exposure (Table 1); these cells were almost exclusively comprised of epithelial cells, identified by their cuboidal shape and cilia. Similarly, the number of epithelial cells counted during differential cell counting of cytospin slides was markedly elevated at 12 and 24 hr (p < 0.05) but had returned to control levels by 48 hr (Figure 7). The amount of total protein in BALF supernatant, a marker of airway microvascular permeability and epithelial damage, was significantly elevated 12 hrs after chlorine exposure, and remained elevated up to 5 d post-exposure (Table 1).

Cl₂ exposure altered respiratory mechanics as reflected by changes in baseline Ers and Rrs. The initial response to Cl₂ exposure was an elevation of Ers and Rrs, which persisted up to 48 hrs post-exposure (Ers = 51.1 ± 3.09 cmH₂O/ml in control mice vs. 70.9 ± 3.23, 67.5 ± 2.16, and 61.5 ± 1.67 cmH₂O/ml at 12, 24, and 48 hrs post-exposure respectively, p < 0.05; Rrs = 0.98 ± 0.05 cmH₂O/ml/sec in control mice vs. 1.32 ± 0.06 and 1.23 ± 0.05 cmH₂O/ml/sec at 12 and 24 hrs post-exposure respectively, p < 0.05) (Figure 8). Airway mechanics returned to baseline levels by 5 d, but at 10 d post-exposure, Ers levels fell significantly below control levels (Ers = 51.1 ± 3.09 cmH₂O/ml in control mice vs. 40.7 ± 0.97 cmH₂O/ml at 10 d post-exposure, p < 0.05). Airway responsiveness to methacholine, as determined by ΔErs, increased after Cl₂ exposure compared to control, and was significantly higher at 12 hrs and 5 d post exposure (ΔErs = 100 ± 19.7 in control mice vs. 257 ± 45.3 and 269 ± 34.0 at 12 hrs and 5 d post-exposure respectively, p < 0.05) (Figure 9). ΔRrs was not significantly altered at any time point after Cl₂ exposure, although a trend for ΔRrs to be lower 24 hrs after Cl₂ exposure was observed (p = 0.055).