Sustained Club Cell Injury in Mice Induces Histopathologic Features of Deployment-Related Constrictive Bronchiolitis

Abstract

Histopathologic evidence of deployment-related constrictive bronchiolitis (DRCB) has been identified in soldiers deployed to Southwest Asia. While inhalational injury to the airway epithelium is suspected, relatively little is known about the pathogenesis underlying this disabling disorder. Club cells are local progenitors critical for repairing the airway epithelium after exposure to various airborne toxins, and a prior study using an inducible transgenic murine model reported that 10 days of sustained targeted club cell injury causes constrictive bronchiolitis. To further understand the mechanisms leading to small airway fibrosis, a murine model was employed to show that sustained club cell injury elicited acute weight loss, caused increased local production of proinflammatory cytokines, and promoted accumulation of numerous myeloid cell subsets in the lung. Transition to a chronic phase was characterized by up-regulated expression of oxidative stress–associated genes, increased activation of transforming growth factor-β, accumulation of alternatively activated macrophages, and enhanced peribronchiolar collagen deposition. Comparative histopathologic analysis demonstrated that sustained club cell injury was sufficient to induce epithelial metaplasia, airway wall thickening, peribronchiolar infiltrates, and clusters of intraluminal airway macrophages that recapitulated key abnormalities observed in DRCB. Depletion of alveolar macrophages in mice decreased activation of transforming growth factor-β and ameliorated constrictive bronchiolitis. Collectively, these findings implicate sustained club cell injury in the development of DRCB and delineate pathways that may yield biomarkers and treatment targets for this disorder.

Numerous reports document increased incidence of respiratory symptoms in soldiers returning from the Gulf War (reviewed in two studies^1,2) and post-9/11 conflicts (reviewed in three studies3, 4, 5) in Southwest Asia. A review of US Department of Veterans Affairs medical encounters from 2003 to 2011 identified an increase in the prevalence of chronic lung disease in veterans deployed to this region.⁶ The first histopathologic correlate of a chronic deployment-related lung disease was reported by King et al⁷ following analysis of surgical lung biopsies obtained from 49 soldiers with unexplained dyspnea on exertion and impaired exercise tolerance. Most biopsies identified small airway disease, with a subset of 38 soldiers exhibiting histopathologic evidence of deployment-related constrictive bronchiolitis (DRCB), defined as a 20% increase in airway wall thickness with associated findings of chronic inflammation and peribronchiolar collagen deposition.

Reports linking the development of constrictive bronchiolitis (CB) to military conflict date back to the use of mustard gas during World War I⁸ and the exposure of 30,000 Iranian civilians to this agent during the Iraq-Iran War (1980 to 1988). Case reports of CB have also been documented in veterans diagnosed with Gulf War Veterans Illness⁹ and following the World Trade Center attack.¹⁰ The incidence of CB among soldiers is likely underestimated as even significant abnormalities in this anatomic region of the lung may often not be detected using conventional pulmonary function testing (spirometry and whole-body plethysmography) and radiographic imaging (chest X-rays and computed tomography scans). Biomarkers of DRCB have not been identified. Thus, at present, definitive diagnosis of this condition requires a surgical lung biopsy, which is infrequently performed because of the risks associated with this invasive procedure and the current lack of proven treatments for this condition.

The cause of lung injury in soldiers diagnosed with DRCB on return from conflicts in Southwest Asia remains unknown. Although exposure to a sulfur dioxide mine fire was common in the cohort described by King et al,⁷ it was not a universal finding, raising concerns that other deployment-related exposures, including burn pit smoke, particulate matter from dust storms, or chemical weapons, might also play a causative role. The composition of burn pit smoke is complex, but known to include the following: i) sulfur dioxide, ii) volatile particle- and vapor-phase organic compounds, iii) particulate matter containing geological materials, iv) carbon from combustion sources, v) metals from soils, and vi) industrial emissions.⁵ More importantly, many of these toxins have been previously implicated in the development of bronchiolitis in nonmilitary populations.11, 12, 13, 14, 15 Chemical weapons known to cause small airway injury include sulfur and nitrogen mustard (mustard gas), nerve gas (including sarin), and chlorine gas.16, 17, 18

The current understanding of DRCB remains limited because of the logistical difficulties and dangers associated with performing research in active combat zones and the challenges inherent to generating animal models that mimic the numerous potential deployment-related inhalational exposures. On the basis of the descriptions of lung histopathology observed in patients with DRCB,^7,19 we hypothesized that injury to the small airways' epithelium is critical to development of this disorder. The studies particularly focused on club cell injury as these cells reside in terminal bronchioles and serve as local progenitors for injured ciliated cells.^20,21 In addition, club cells produce numerous molecules, including club cell secretory protein (CCSP) that helps maintain local airway integrity,^21,22 protects against oxidative stress,^23,24 and modulates host immune responses, including those mediated by lung macrophages.^25,26 Furthermore, abundant evidence links club cell injury and alteration in CCSP levels with inhalational exposures in nonmilitary personnel, such as exposures to smoke from burning buildings,²⁷ nitric oxides,²⁸ chlorine,²⁹ sulfur dioxide,¹¹ silica,¹⁴ and particulate matter.30, 31, 32 Additional experimental evidence obtained from studies in rodents demonstrates an association between club cell injury, the development of chronic bronchiolitis, and exposure to sulfur dioxide,^33,34 diacetyl,³⁵ naphthalene,³⁶ and particulate matter.³⁷ In light of these findings, we propose that sustained airway club cell injury is a common injury pathway downstream of exposure to various inhalational toxins.

The current study utilized a murine model developed by Perl et al,³⁸ in which inducible, targeted, and sustained injury of club cells induces persistent airway wall thickening consistent with CB. Pathologic pathways triggered by sustained club cell injury were investigated, and a two-phase response was identified. The initial, acute injury was characterized by CCSP release and cellular and molecular features of a proinflammatory response. The subsequent chronic response consisted of up-regulated expression of genes associated with oxidative stress, activation of transforming growth factor (TGF)-β, accumulation of alternatively activated macrophages, and deposition of peribronchiolar collagen. The chronic histopathologic features elicited by sustained club cell injury in mice recapitulated key findings identified in soldiers diagnosed with DRCB. Lastly, depletion of alveolar macrophages in mice subjected to sustained club cell injury decreased activation of TGF-β and ameliorated CB. Collectively, these findings implicate sustained club cell injury and alveolar macrophages in the pathophysiology of DRCB and identify specific pathways that could be further investigated to establish diagnostic biomarkers and potential treatments for this condition.

Materials and Methods

Mice and Doxycycline Exposure

Triple transgenic Scgb1a1rtTA/tetOCre/R26:lacZ/DT-A (CC-DTA) mice, described by Perl et al,³⁸ have a reverse tetracycline transactivator gene driven by the CCSP promoter, a Cre recombinase gene under the control of a tet operator sequence, and a lox-P activated diphtheria toxin-A gene. Ingestion of doxycycline by CC-DTA mice activates diphtheria toxin-A expression only in club cells, leading to targeted autonomous cell death. CC-DTA and single-transgenic CRE littermate control mice were bred and housed under specific pathogen-free conditions. Studies were approved by the Veterans Affairs Institutional Animal Care and Use Committee. Sustained club cell injury, leading to murine CB, was induced by exposure of CC-DTA mice to doxycycline via their food (625 mg doxycycline/kg chow; Teklad; Envigo, Madison, WI) for 10 consecutive days.

Tissue Collection and Processing

Blood samples were obtained from mice immediately after euthanasia via intracardiac puncture. After clotting at room temperature for 2 hours, samples were spun at 2000 × g for 10 minutes at 4°C. Serum samples were collected and frozen at −80°C.

Bronchoalveolar lavage fluid (BALF) was generated immediately after mouse euthanasia by instilling twice 1.0 mL of sterile 0.5 mmol/L EDTA in phosphate-buffered saline into the lungs via the trachea and aspirating the fluid back. BALF samples were then spun down at 5000 × g for 10 minutes at 4°C, and supernatants were collected and frozen at −80°C.

For flow cytometric analysis, lungs were perfused in situ via the right heart using phosphate-buffered saline (approximately 10 mL) until pulmonary vessels became clear. Lung lobes were then harvested, minced, and placed in digestion buffer (5 mL/lung) containing 5% complete medium [RPMI 1640 medium, fetal bovine serum (5%), penicillin-streptomycin (1%), MEM Non-Essential Amino Acids Solution (1%), and sodium pyruvate (1%); all from Gibco by Life Technologies, Ann Arbor, MI], deoxyribonuclease I (250 Kunitz units/lung; Sigma-Aldrich, St. Louis, MO), and collagenase type I (0.1%; Gibco by Life Technologies). Lungs were mechanically homogenized twice using a gentleMacs dissociator (Miltenyi Biotec, Auborn, CA), and enzymatically digested between homogenization cycles at 37°C for 35 minutes on a rocker. After erythrocyte lysis by ACK (KD Medical, Columbia, MD), cells were washed and filtered over a 100-μm mesh. Dead cells were removed by centrifugation over a Percoll (Sigma-Aldrich) gradient. Viable, lung-derived cells in each sample were enumerated in the presence of Trypan Blue (Gibco by Life Technologies) using a hemocytometer. Each lung was processed and analyzed individually.

Quantifying Gene Expression

RNA was extracted from lung homogenates using TRIzol reagent (Invitrogen by Life Technologies) and DNase treated using the Turbo DNA-free Kit (Ambion by Life Technologies, Waltham, MA). Transcript levels were quantified using a one-step quantitative RT-PCR (QuantiTect Sybr Green Kit; Qiagen, Valencia, CA) with glyceraldehyde-3-phosphate dehydrogenase serving as an endogenous reference. Reactions were run in triplicates using a StepOne Plus real-time PCR system (Applied Biosystems by Life Technologies, Waltham, MA). QuantiTect Primer Assays were obtained from Qiagen. Data were analyzed using the $2^{-\Delta \Delta {\text{C}}_{\text{T}}}$ method.

Measurement of Protein Concentration

CCSP concentration in serum and BALF samples was determined in duplicate using an enzyme-linked immunosorbent assay (Novus Biologicals, Centennial, CO) following the manufacturer's protocol and using a Synergy microplate reader (BioTek, Winooski, VT).

Cytokine concentrations in serum and BALF samples were quantified using a cytometric bead array (LEGENDplex; BioLegend, San Diego, CA), according to the manufacturer's instructions, with data acquired on an LSRFortessa flow cytometer equipped with FACSdiva software version 9.0 (both from BD Biosciences, San Jose, CA).

Cell Staining and Flow Cytometry Analysis

Fluorochrome-conjugated antibodies used for cell staining are listed in Table 1. Cells were first stained with a fixable viability dye (Zombie aqua; BioLegend) following the manufacturer's protocol. After blocking Fc receptors using anti-CD16/32 antibody (clone 93; BioLegend), cells were stained for cell surface markers and then fixed with 2% formaldehyde (Thermo Fisher Scientific, Waltham, MA) in phosphate-buffered saline. Data were acquired using an LSRFortessa flow cytometer (BD Biosciences) and analyzed using FlowJo software version 10.0 (Treestar, Ashland, OR). At least 100,000 events in the CD45⁺ gate were acquired per lung sample. To determine the number of cells in each population of interest in each sample, the corresponding percentage was multiplied by the total number of viable CD45⁺ cells in that sample. The latter value was calculated for each sample as the product of the percentage of viable CD45⁺ cells and the original hemocytometer count of total viable cells identified within that sample.

Table 1.

Antibodies Used for Staining in Flow Cytometric Analysis

Target	Clone	Fluorochrome	Manufacturer
CD45	30-F11	PerCP-Cy5.5	BioLegend
Ly6G	1A8	APC	BioLegend
CD11b	M1/70	APC-Cy7	BioLegend
CD11c	N418	Brilliant Violet 421	BioLegend
Siglec F	E50-2440	PE-CF594	BD Biosciences
CD24	M1/69	Brilliant Violet 650	BD Biosciences
CD103	2E7	PE	BioLegend
MHC II (I-A/I-E)	M5/114.15.2	PE-Cy5	BioLegend
Ly6C	HK1.4	Alexa Fluor 700	BioLegend
CD206	C068C2	PE-Cy7	BioLegend
TCRβ chain	H57-597	PE-Cy5	BioLegend
CD8a	53-6.7	FITC	BioLegend
CD4	GK1.5	APC	BioLegend
CD19	1D3	PE-CF594	BD Biosciences

APC, allophycocyanin; Cy, cyanine; FITC, fluorescein isothiocyanate; PE, R-phycoerythrin.

Hydroxyproline Assay

After euthanasia of mice, all five lung lobes were harvested from each experimental animal and frozen at −20°C. Individual lungs were then assessed for hydroxyproline content, as previously described by Sisson et al.³⁹ Briefly, lungs were homogenized and then hydrolyzed in 6 N HCl at 120°C for 24 hours. Aliquots of each sample were mixed with citrate/acetate buffer and incubated with chloramine T solution. Ehrlich's reagent was added, and samples were incubated for 30 minutes at 65°C. The absorbance of each sample was measured at 540 nm, and hydroxyproline concentration was determined using a standard curve.

Histology and Fluorescent Immunohistochemistry

Staining of murine lung sections was performed, as previously described.⁴⁰ Briefly, lungs were inflation fixed in situ via the trachea with 10% neutral-buffered formalin (Thermo Fisher Scientific). Lung lobes were then harvested, further fixed in 10% neutral-buffered formalin, processed, and embedded in paraffin. Paraffin sections (4 μm thick) were stained with hematoxylin and eosin, picrosirius red, trichrome, or, for fluorescence immunohistochemistry, according to standard laboratory procedures using the following primary antibodies: rabbit anti-CCSP (1:100; Abcam, Cambridge, MA) and mouse anti–α-smooth muscle actin (1:20; MilliporeSigma, Burlington, MA). Sections were mounted using ProLong Diamond Antifade Mountant with DAPI (Life Technologies, Grand Island, NY). Imaging was performed with a Zeiss Apotome microscope connected to a Zeiss AxioCam MRm camera and using the AxioVision software (Zeiss, Wetzlar, Germany).

Staining of frozen murine lung sections was performed, as previously described.⁴¹ Briefly, lungs were inflated first with 1 mL air followed by 1 mL of a 4:1 mixture of RPMI 1640 medium to optimal cutting temperature compound (Tissue Tek; Sakura Finetek, Torrance, CA). Lungs were embedded in optimal cutting temperature compound, snap frozen in liquid N₂, and stored overnight at −20°C. Lung sections (5 μm thick) were blocked with hamster serum, and antigens were visualized using the following biotinylated primary antibodies: anti-CD64, anti-CD11c, anti-F4/80 (BioLegend) followed by an aminoethyl carbazole peroxidase reaction. Slides were counterstained in hematoxylin and coverslipped using aqueous mounting media (Biomeda).

Human surgical lung biopsies were formalin fixed, paraffin embedded, and stained with hematoxylin and eosin or with anti-CD68 antibody (using a diaminobenzidine peroxidase reaction) per clinical laboratory protocols at the VA Ann Arbor Health System. The diagnosis of DRCB was established by histopathologic review at Vanderbilt University Medical Center.

Macrophage Depletion Using Oropharyngeal Aspiration of Clodronate Liposomes

To deplete lung alveolar macrophages, mice were briefly anesthetized using isoflurane (Baxter, Deerfield, IL) inhalation and positioned vertically using an RIS 100 rodent intubation stand (Braintree Scientific, Braintree, MA). After lateral retraction of the tongue, anionic clodronate or empty liposomes (Formumax Scientific, Sunnyvale, CA) were inoculated into the posterior oropharynx using a micropipette, and nostrils were closed to ensure aspiration. Anionic clodronate or empty liposomes (50 μL/20 to 25 g mouse body weight per dose) were administered via oropharyngeal aspiration on days 0, 3, 7, 10, 13, and 17. BALF and lungs were harvested on day 20.

Statistical Analysis

Data in graphs are presented as means ± SEM. Data were evaluated by unpaired, two-tailed t-test, corrected for multiple comparisons using the Holm-Sidak method when appropriate or one-way analysis of variance followed by the Tukey multiple comparisons test. Statistical analysis was performed using GraphPad Prism software version 9.0 (San Diego, CA). P < 0.05 was considered statistically significant.

Results

Exposure of CC-DTA Mice to Doxycycline for 10 Consecutive Days Induces Sustained Club Cell Injury and Alters CCSP Gene Expression and Protein Level

To investigate the effects of sustained club cell injury, transgenic CC-DTA mice, developed by Perl et al,³⁸ were used, in which the expression of diphtheria toxin-A is restricted to club cells and induced by doxycycline. Exposure of CC-DTA mice to doxycycline for 10 consecutive days (Figure 1A) led to a substantial depletion of airway club cells at day 10 (Figure 1B), using fluorescent immunohistochemistry for CCSP. At day 20, club cell depletion was still readily evident in CC-DTA mice but observed to a lesser extent. In contrast, control mice exposed to doxycycline had no club cell depletion. The degree of depletion observed here in CC-DTA mice was comparable to that described by Perl et al.³⁸

Figure 1.

Exposure of CC-DTA mice to doxycycline for 10 consecutive days leads to sustained club cell injury. A: Schematic representation of exposure protocol. Cohorts of CC-DTA and control mice were exposed to doxycycline on days 0 to 9 of the protocol to induce club cell injury. B: Representative lung sections of control (top panels) and CC-DTA (bottom panels) mice at the indicated protocol time points were stained using fluorescent immunohistochemistry for club cell secretory protein (CCSP; green), α-smooth muscle actin (α-SMA; red), and DAPI (blue). C:Scgb1a1 (CCSP; left panel) and Foxa2 (right panel) relative gene expression, as measured by quantitative RT-PCR analysis in total lung homogenates at the indicated time points. Expression level was normalized to glyceraldehyde-3-phosphate dehydrogenase and expressed relative to the mean expression of control mice at day 10. D: CCSP protein concentrations measured in bronchoalveolar lavage fluid (BALF; left panel) and serum (right panel) at the indicated time points. C and D: Cumulative data shown are from two independent experiments and present data from 8 to 10 mice. CC-DTA mice, open bars; control mice, closed bars. Data are presented as means ± SEM (C and D). ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001 (t-test corrected for multiple comparisons using the Holm-Sidak method). Scale bars = 50 μm (B). AW, airway; V, blood vessel.

Sustained club cell injury led to reduced CCSP gene expression in total lung homogenates of CC-DTA versus control mice at day 10, but transcript levels reverted to normal by day 20 (Figure 1C). Increased expression of Foxa2, a transcription factor known to regulate CCSP gene expression,⁴² was observed at day 20 in CC-DTA versus control mice (Figure 1C) and may have contributed to the normalization in CCSP transcript level.

To determine how sustained club cell injury alters local or systemic levels of CCSP, BALF and serum samples from CC-DTA and control mice were analyzed. CCSP levels were elevated in both the BALF (Figure 1D) and serum (Figure 1D) of CC-DTA mice (relative to controls) at day 10, but reverted to baseline thereafter. Thus, these results confirm that exposure of CC-DTA mice to doxycycline for 10 consecutive days induces airway club cell depletion and provides new insights by demonstrating that sustained club cell injury results in detectable alterations in CCSP gene expression and protein levels.

Sustained Club Cell Injury Results in Weight Loss, Altered Expression of Oxidative Stress–Related Genes, and Modulation of Cytokine Production

To further investigate the response to sustained club cell injury, doxycycline-exposed CC-DTA mice were assessed for local and (or) systemic signs of inflammation. CC-DTA mice lost about 10% of their body weight between day 0 and day 10, but gradually regained weight thereafter (Figure 2A). Next, the expression of genes associated with oxidative stress within the lung tissue was evaluated (Figure 2B). At day 10, expression of Hmox1 (heme oxygenase 1) was increased in the lungs of CC-DTA mice (relative to controls), whereas expression of Cat (catalase), Sod1 (superoxide dismutase), and Gpx1 (glutathione peroxidase) was reduced. At day 20, however, expression of all four genes was increased.

Figure 2.

Exposure of CC-DTA mice to doxycycline for 10 consecutive days modulates body weight, oxidative stress, and cytokine production. A: Percentage change in body weight from baseline (day 0) in CC-DTA (black squares) and control (black circles) mice at the indicated time points. Data represent at least three independent experiments. B: Relative gene expression measured by quantitative RT-PCR analysis in total lung homogenates for genes associated with oxidative stress: Hmox1 (heme oxygenase 1), Cat (catalase), Gpx1 (glutathione peroxidase), and Sod1 (superoxide dismutase) at the indicated time points. C: Cytokine concentrations were measured by a cytometric bead array in bronchoalveolar lavage fluid for tumor necrosis factor (TNF)-α, IL-6, B-cell activating factor (BAFF), and free active transforming growth factor (TGF)-β1 at the indicated time points. B and C: Cumulative data are from two independent experiments and are presented as data from 8 to 10 mice. CC-DTA mice, open bars; control mice, closed bars. Data are presented as means ± SEM (B and C). n = 5 (A). ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001 (t-test corrected for multiple comparisons using the Holm-Sidak method).

To examine whether sustained club cell injury modulated local production of inflammatory, regulatory, and (or) profibrotic cytokines, cytokine levels were quantified in BALF using a cytometric bead array. Increased levels of tumor necrosis factor-α, IL-6, and B-cell activating factor (BAFF) were detected in CC-DTA relative to control mice at protocol day 10; BAFF remained elevated at day 20 (Figure 2C). Of note, IL-4, IL-10, IL-12p70, and IL-13 were undetectable in BALF of all mice tested at both day 10 and day 20 (data not shown). More importantly, the concentration of free active TGF-β1 was increased at protocol day 20. To determine whether club cell injury altered systemic cytokine levels, serum samples from CC-DTA and control mice were assayed. An increase in BAFF in CC-DTA relative to control mice was identified at both day 10 (104.4 ± 5.1 versus 82.3 ± 7.3 ng/mL, respectively; P < 0.05) and day 20 (97.1 ± 6.5 versus 79.3 ± 3.8 ng/mL, respectively; P < 0.05), whereas levels of all other cytokines tested were either undetectable or did not differ significantly between the two groups. Thus, these findings demonstrate that sustained club cell injury significantly alters molecular pathways associated with oxidative stress and skews the inflammatory and profibrotic cytokine environment within the lung of exposed mice.

Sustained Club Cell Injury Promotes Chronic Accumulation of Alternatively Activated Lung Macrophages

To determine the cellular immune response to sustained club cell injury, a multiparameter flow cytometric analysis and an established gating scheme⁴³ were used to identify and enumerate total CD45⁺ lung leukocytes and specific subsets of myeloid (gating shown in Supplemental Figure S1) and lymphoid (gating not shown) cells in the lungs of CC-DTA versus control mice. In doxycycline-exposed CC-DTA mice, an increase in the total number of CD45⁺ lung leukocytes was observed that differed from control mice by day 10 and persisted at day 20 (Figure 3A). Among granulocyte populations, there was a nonsignificant trend toward increased neutrophils at day 5 and day 10 in CC-DTA mice, whereas no difference was observed at day 20. In contrast, eosinophils were increased at both day 5 and day 10. Within the two populations of conventional dendritic cells, a reduction in CD103⁺ conventional dendritic cells was observed at day 5 in the lungs of CC-DTA mice and an increase in CD11b⁺ conventional dendritic cells was detected at day 10; no difference in either population was noted at day 20. Both monocyte-derived dendritic cells and Ly6C^high monocytes were increased in CC-DTA mice at day 10 and returned to baseline by day 20. Following sustained club cell injury, two populations of lung macrophages were identified in the lungs: resident alveolar macrophages and monocyte-derived exudate macrophages. Both alveolar macrophages and exudate macrophages were increased in CC-DTA mice at day 10 and were the only myeloid cell population that remained increased at day 20. No differences in lymphoid subsets (B cells, CD4⁺ T cells, and CD8⁺ T cells) were identified between CC-DTA and control mice at any time point evaluated (data not shown).

Figure 3.

Exposure of CC-DTA mice to doxycycline for 10 consecutive days induces chronic accumulation of alternatively activated lung macrophages. A: Total numbers of CD45⁺ leukocytes and each myeloid subset per mouse lung were calculated at the indicated time points. B: Representative cell surface CD206 expression on alveolar macrophages (left panel) and exudate macrophages (right panel) obtained at protocol day 20 from a CC-DTA (thick black line; open histogram) and control mouse (thin black line; gray histogram) relative to isotype-matched control staining (thin orange line; open histogram). C: CD206 cell surface expression [Δ geometric mean fluorescence intensity (GMFI) relative to isotype-matched control antibody staining] on alveolar macrophages (AMs) and exudate macrophages (ExMs) at protocol day 20. D: Relative gene expression measured by quantitative RT-PCR analysis in total lung homogenates for genes associated with alternative macrophage activation: Lgals3 (galectin-3), Retnla (resistin-like α; Fizz1), Chil3 (chitinase-like 3; YM1), and Mrc1 (mannose receptor C type 1; CD206) at protocol day 20. A, C, and D: Cumulative data are from two independent experiments and presented as data from 10 mice. CC-DTA mice, open bars; control mice, closed bars. Data are presented as means ± SEM (A, C, and D). ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001 (t-test corrected for multiple comparisons using the Holm-Sidak method). cDC, conventional dendritic cell; max, maximum; moDC, monocyte-derived dendritic cell.

Prior studies^44,45 have implicated persistent accumulations of alternatively activated macrophages in the development of interstitial lung fibrosis in response to alveolar epithelial injury in multiple murine models. To assess the activation status of lung macrophages accumulating in response to sustained club cell injury, cell surface expression of CD206, a marker associated with alternative activation, was evaluated. At day 20, CD206 cell surface expression was increased in both alveolar macrophages and exudate macrophages from CC-DTA versus control mice (Figure 3, B and C). To further characterize the macrophage activation profile at day 20, the expression of multiple genes associated with alternative activation was assessed, including Arg1 (arginase), Lgals3 (galectin 3), Retnla (resistin-like α; Fizz1), Chil3 (chitinase-like protein 3; YM1), and Mrc1 (mannose receptor C type 1; CD206) as well as Nos2 (inducible nitric oxide synthase), which is commonly associated with classic macrophage activation. Results show that expression of Lgals3, chil3, and MRC1 was increased in the lungs of CC-DTA versus that in control mice (Figure 3D), whereas gene expression of both Arg1 and Nos2 was low or undetectable in both CC-DTA and control mice (data not shown). Collectively, these findings identify persistent accumulation of macrophages with a predominantly alternatively activated phenotype following sustained club cell injury.

Sustained Club Cell Injury Causes Chronic Peribronchiolar Fibrosis

In the initial study by Perl et al,³⁸ extensive morphometric histologic analysis and trichrome staining, performed on lung sections obtained from CC-DTA mice, demonstrated that exposure of CC-DTA mice to doxycycline for 10 days led to a twofold increase in airway wall thickness and peribronchiolar collagen deposition. In the current study, these features of CB were further evaluated using picrosirius red staining for collagen on lung sections derived from day 20 CC-DTA versus control mice (Figure 4A). In CC-DTA mice, airway wall thickening associated with prominent subepithelial collagen deposition was identified relative to control mice. The next objective was to determine how these histopathologic features of CB were correlated with quantitative changes in collagen gene and (or) protein expression in total lungs. Quantitative RT-PCR analysis performed on whole lung homogenates demonstrated increased Col1a1 gene expression in CC-DTA relative to control mice at day 10 and day 20 (Figure 4B). Moreover, these findings were associated with increased total lung collagen at both day 20 and day 30, as assessed by a hydroxyproline assay (Figure 4C). These findings show that sustained club cell injury in mice leads to the development of chronic peribronchiolar fibrosis that can be readily quantified.

Figure 4.

Exposure of CC-DTA mice to doxycycline for 10 consecutive days induces chronic peribronchiolar fibrosis. A: Representative lung sections stained using picrosirius red (to identify collagen) from control and CC-DTA mice at protocol day 20. B: Relative gene expression measured by quantitative RT-PCR analysis in total lung homogenates for Col1a1 at the indicated time points. C: Hydroxyproline assay for total lung collagen content quantification at the indicated time points. B and C: Cumulative data are from two independent experiments and are presented as data from 8 to 10 mice. C: The two graphs represent separate, independent experiments. CC-DTA mice, open bars; control mice, closed bars. Data are presented as means ± SEM (B and C). ∗∗P < 0.01, ∗∗∗P < 0.001 (t-test corrected for multiple comparisons using the Holm-Sidak method). Scale bars: 100 μm (A, left panels); 10 μm (A, right panels). AW, airway; V, blood vessel.

Sustained Club Cell Injury–Induced Murine CB Recapitulates Major Histopathologic Features of DRCB

The next objective was to compare histopathologic features of murine CB induced by sustained club cell injury with findings identified in soldiers with DRCB. In lung sections obtained from CC-DTA mice at protocol day 20 (Figure 5, A and B), features of constrictive bronchiolitis were observed, including epithelial thinning and squamous metaplasia with evidence of thickening of airway walls due to subepithelial matrix deposition. The lung sections from CC-DTA mice were compared with lung sections obtained from a surgical lung biopsy of a veteran evaluated at the VA Ann Arbor Health System whose diagnosis of DRCB was confirmed following expert clinical and pathologic review at Vanderbilt Medical Center (R. Miller and J. Johnson, personal communication, 2012). The veteran's biopsy was obtained 9 years after his deployment to Iraq following a workup for persistent unexplained dyspnea. He reported exposures to burn pits, the Mosul sulfur mine fire, and sandstorms. Histopathologic analysis of this biopsy revealed findings of squamous metaplasia and airway wall thickening (Figure 5, C and D) similar to those observed in CC-DTA mice on protocol day 20.

Figure 5.

Airway wall thickening and epithelial metaplasia in murine and deployment-related constrictive bronchiolitis (DRCB). Representative images of lung sections (hematoxylin and eosin stained) obtained from a CC-DTA mouse (A and B) at protocol day 20 and a veteran diagnosed with DRCB (C and D). Note thickening of the airway wall with subepithelial matrix deposition (blue brackets; B and D) and squamous metaplasia (red arrows; B and D) present in both specimens. Scale bars: 100 μm (A and C); 40 μm (B and D). AW, airway; CB, constrictive bronchiolitis.

Next, the cellular components of the inflammatory response in murine and DRCB were compared. In murine CB, peribronchiolar infiltrates containing a mix of small mononuclear cells along with cells displaying macrophage morphology were evident at protocol day 10 (data not shown) and persisted at day 20 (Figure 6, A and B). In addition, clusters of larger cells displaying macrophage morphology were observed within the airway lumens (Figure 6, C and D). Peribronchiolar infiltrates were also present in DRCB (Figure 6, E and F), as were numerous collections of large intraluminal foamy macrophages (Figure 6, G and H).

Figure 6.

Chronic airway inflammation in murine and deployment-related constrictive bronchiolitis (DRCB). Representative images of lung sections (hematoxylin and eosin stained) obtained from a CC-DTA mouse (A–D) at protocol day 20 and a veteran diagnosed with DRCB (E–H). Note the peribronchiolar mononuclear cell infiltrates (orange brackets; B and F) and collections of intraluminal cells with the appearance of foamy macrophages (blue arrows; D and H) present in both specimens. Scale bars: 100 μm (A, C, E, and G); 40 μm (B, D, F, and H). AW, airway; CB, constrictive bronchiolitis; MCI, mononuclear cell infiltrate.

To verify that the accumulations of peribronchiolar and intraluminal airway cells observed in murine CB were of a myeloid/macrophage lineage, immunohistochemistry was performed using a panel of control (Figure 7, A and B) and specific antibodies identifying proteins expressed by lung monocytes and macrophages, including CD68 (Figure 7, C–F), F4/80, and CD11c (Supplemental Figure S2). Results show numerous smaller cells expressing CD68 (Figure 7, B and C) and F4/80 (Supplemental Figure S2), displaying monocyte morphology in subepithelial infiltrates, and collections of larger cells expressing CD68 (Figure 7, C and D), and CD11c (Supplemental Figure S2), displaying macrophage morphology within airways. Similarly, an abundance of smaller CD68⁺ cells, displaying monocyte morphology localized beneath the airway epithelium, as well as more substantial clusters of larger CD68⁺ cells, displaying macrophage morphology aggregating within the airways, was observed in DRCB (Figure 7, G–J).

Figure 7.

Peribronchiolar myeloid cell infiltrates and large intraluminal macrophages characterize both murine and deployment-related constrictive bronchiolitis (DRCB). Representative images of lung sections obtained from a CC-DTA mouse (A–F) at protocol day 20 and a veteran diagnosed with DRCB (G–J). Murine frozen sections (A–J) were stained with isotype control antibody (A and B) or anti-CD68 (C–F) using an aminoethyl carbazole peroxidase reaction (red staining), whereas sections from the veteran with DRCB were paraffin embedded and stained with anti-CD68 (G–J) using a 3,3′-diaminobenzidine peroxidase reaction (brown staining). Both specimens were counterstained with hematoxylin. Note the presence of smaller CD68⁺ mononuclear cells localized to the subepithelial infiltrates (red arrows; D, F, and H) and the larger CD68⁺ cells with macrophage morphology residing within airway lumens (blue arrows; F and J) present in both specimens. Scale bars: 100 μm (A, C, E, and G–J); 40 μm (B, D, and F). AW, airway; CB, constrictive bronchiolitis.

Altogether, this histopathologic analysis verifies that the small airways are the microanatomic site of chronic inflammation in mice with sustained club cell injury. Results further show that the pattern of epithelial cell injury and chronic macrophage-enriched inflammation evident in murine CB that results from club cell injury recapitulates many of the histopathologic abnormalities characteristic of DRCB.

Depletion of Alveolar Macrophages in Mice Subjected to Sustained Club Cell Injury Ameliorates CB

Accumulation of alternatively activated macrophages in response to lung injury has been implicated in the pathogenesis of alveolar fibrosis.^44,46,47 Because lung macrophages remained persistently elevated during the chronic, fibrotic phase of our model of sustained club cell injury, we hypothesized that they may contribute to the development of CB. To test this hypothesis, lung macrophages were depleted by administering clodronate liposomes (CLs; via oropharyngeal aspiration) to doxycycline-exposed CC-DTA mice on protocol days 0, 3, 7, 10, 13, and 17 of the 20-day protocol (Figure 8A). Cohorts of doxycycline-exposed control and CC-DTA mice receiving empty liposomes served as controls. As expected, short-term administration of CLs to doxycycline-exposed CC-DTA mice depleted alveolar macrophages (by 86%) and reduced the numbers of exudate macrophages (by 60%) (Supplemental Figure S3A). Alveolar macrophage depletion persisted at day 20 (by 91%), whereas exudate macrophages were no longer decreased but rather increased in numbers (by 7.3-fold) (Supplemental Figure S3B). Results demonstrated that doxycycline-exposed CC-DTA mice receiving CLs versus empty liposomes lost less weight and reverted faster to baseline (Figure 8B). More importantly, CL- versus empty liposome–treated doxycycline-exposed CC-DTA mice demonstrated a striking reduction in peribronchiolar collagen deposition, as assessed by picrosirius red and trichrome staining at protocol day 20 (Figure 8C) and quantified by total lung hydroxyproline assay (Figure 8D). Lastly, reduced peribronchiolar fibrosis in CL-treated doxycycline-exposed CC-DTA mice was associated with substantial decreased levels of free activated TGF-β1 in BALF (relative to doxycycline-exposed CC-DTA mice receiving Els) (Figure 8E). These findings implicate alveolar macrophages in promoting murine CB.

Figure 8.

Depletion of alveolar macrophages ameliorates murine constrictive bronchiolitis in CC-DTA mice exposed to doxycycline for 10 consecutive days. A: Schematic representation of macrophage depletion protocol. Cohorts of CC-DTA and control mice were exposed to doxycycline on days 0 to 9 of the protocol to induce club cell injury. Mice also received either clodronate liposomes (CLs) or empty liposomes (ELs) via oropharyngeal aspiration on protocol days 0, 3, 7, 10, 13, and 17. B: Percentage change in body weight from baseline (day 0) in doxycycline-exposed CC-DTA mice treated with CLs (red squares) or ELs (black circles) relative to doxycycline-exposed control mice treated with ELs (black triangles) at the indicated time points. Data are representative of one of three experiments. C: Representative lung sections from doxycycline-exposed control or CC-DTA mice given the indicated treatments stained using picrosirius red (top panels) or trichrome (bottom panels) to identify peribronchiolar collagen deposition at protocol day 20. Right panels: Note the reduction in collagen staining in doxycycline-exposed CC-DTA mice treated with CLs. D: Hydroxyproline assay for total lung collagen content quantification at protocol day 20. E: Free active transforming growth factor (TGF)-β1 concentrations measured by a cytometric bead array in bronchoalveolar lavage fluid at protocol day 20. D and E: Cumulative data shown are from two independent experiments and presented as data from 9 to 10 mice. CC-DTA mice, open bars; control mice, closed bars. Data are presented as means ± SEM (D and E). n = 5 mice per experimental condition (B). ∗P < 0.05, ∗∗P < 0.01 (one-way analysis of variance followed by the Tukey multiple comparisons test). Scale bars = 100 μm (C).

Discussion

In response to mounting concerns about potential adverse health effects associated with exposures to airborne hazards during military deployment in Southwest Asia, there is an urgent need to better understand the pathogenesis of DRCB. Results from the current study show that sustained club cell injury in mice leads to an acute local and systemic response identified by weight loss, CCSP release, enhanced expression of proinflammatory cytokines, and an influx of numerous subsets of myeloid cells. Although sustained club cell injury for 10 consecutive days caused only transient depletion of club cells, it induced a chronic fibrotic phase characterized by increased oxidative stress and TGF-β1 activation, persistent accumulation of alternatively activated macrophages, and peribronchiolar collagen deposition. The histopathologic features resulting from sustained club cell injury in mice recapitulated key chronic findings identified in DRCB, including peribronchial and intraluminal macrophage accumulation. Results demonstrating that depletion of alveolar macrophages ameliorated murine CB directly implicate these cells in the pathogenesis of this disorder.

A recent report by the National Academy of Sciences Engineering and Medicine highlighted the numerous challenges associated with efforts to identify a single or multiple specific exposures leading to the development of respiratory symptoms and lung disease after deployment in Southwest Asia.⁴⁸ Many of these challenges reflect the dangers and logistical obstacles inherent to investigating airborne exposures in combat zones. Given the uncertainty regarding the type(s), length, and dose of exposure capable of causing DRCB, the study focused on studying sustained club cell injury in mice as a common pathway of injury inflicted by numerous potential airborne hazards. A 10-day injury model was chosen, as Perl et al³⁸ had shown that exposing CC-DTA mice to doxycycline for 48 hours resulted in transient club cell depletion, but failed to induce features of CB. In contrast, a 10-day injury model resulted in airway wall thickening that was readily identifiable 10 days after injury (protocol day 20) and persisted out to 19 weeks, as previously shown. Herein, the findings of that study were extended by demonstrating that sustained 10-day club cell injury results in enhanced Col1a1 gene expression at protocol day 10 and day 20. Furthermore, this increase in Col1a1 gene expression was associated with increased total lung collagen at day 20 that persisted out to day 30 (20 days after injury was terminated). This long-term persistence in augmented collagen production and deposition is important to note as it is concordant with the histopathologic features observed in DRCB, which often are identified using lung biopsies obtained years after deployment. In addition, the observed differences in quantitative measures of lung collagen content between control and CC-DTA mice provide objective end points for future evaluations of potential treatments for CB that could be tested using this model.

Additional evidence that epithelial cell injury plays a role in the pathogenesis of DRCB stems from the observation that in both murine CB caused by sustained club cell injury and in DRCB, epithelial thinning and squamous cell metaplasia of the small airways was identified. Underlying these epithelial abnormalities, the study detected regions of airway wall thickening and collagen deposition similar to those reported by King et al⁷ and in a more recent study by Krefft et al.¹⁹ Although these publications did not specifically mention the presence or absence of small airway metaplasia, this finding was described in a case report of DRCB identified in a veteran with Gulf War Veterans’ Illnesses.⁹ The chronic epithelial abnormalities observed in murine CB were preceded by a transient increase in CCSP levels in both BALF and serum samples during and (or) immediately after sustained club cell injury. As CCSP levels were much higher in BALF versus serum samples, this increase likely reflects leakage of CCSP from dying club cells into the small airways' lumen and then to the systemic circulation. Increased CCSP levels in murine CB returned to baseline by day 20, suggesting that CCSP could be a biomarker of acute or ongoing club cell injury but may not be useful at the chronic phase of this disorder. Despite challenges inherent to evaluating CCSP levels in actively deployed soldiers, the current findings provide strong rationale for future studies designed to assess whether alterations in CCSP levels are associated with respiratory exposures, symptoms, and (or) histopathologic features of DRCB.

To enhance our understanding of the pathogenesis of CB, temporal, molecular, and cellular relationships between sustained club cell injury and small airway fibrosis were investigated. The findings identify an initial acute injury response followed by a chronic fibrotic response. The acute injury phase was characterized by an increase in tumor necrosis factor-α, IL-6, and BAFF in BALF as well as accumulation of multiple myeloid subsets in the lung. Whether this phase represents a non-specific response to acute cellular death or is integral to the development of airway fibrosis warrants further investigation. It will be especially important to determine whether the increased production of tumor necrosis factor-α in the acute club cell injury phase promotes the subsequent increase in activated TGF-β1 (observed in the chronic phase of club cell injury), as this has been implicated as a profibrotic mechanism in mice exposed to asbestos.^49,50 Interestingly, levels of BAFF were increased in both the BALF and serum at day 10 and day 20. Increased BAFF levels have been reported in the lungs of mice with bleomycin-induced fibrosis, and BAFF blockade was found to ameliorate lung fibrosis.⁵¹ Thus, in addition to serving as a potential biomarker in soldiers with acute and (or) chronic airborne hazard exposures, future studies could also explore the role of BAFF in small airway fibrosis and assess whether it might be advantageous to target it therapeutically in this model.

The transition from the acute injury phase to the chronic fibrotic phase of the response to sustained club cell injury was marked by increased expression of genes associated with oxidative stress. This finding is intriguing as oxidative stress has been implicated in the development of Gulf War Veterans’ Illnesses⁴⁶ and in the response to exposure to sulfur mustard⁴⁷ and particulate matter.³⁷ To our knowledge, no studies have specifically evaluated whether soldiers or veterans with DRCB exhibit findings indicative of increased oxidative stress. Yet, increased oxidative stress has been shown to be associated with accumulations of large, foamy, alternatively activated lung macrophages, which were detected in the lumen of small airways in our CC-DTA mice during the chronic fibrotic phase of the response to sustained club cell injury and in lung sections obtained from a veteran with DRCB. Although the study verified sustained macrophage accumulation in our murine model of CB, the ability to perform a quantitative or semiquantitative analysis of macrophage accumulation in DRCB was constrained by the paucity of tissue samples available from patients with DRCB that can be used for research purposes. However, the representative findings are consistent with the observations described by King et al,⁷ who identified respiratory bronchiolitis (defined as intraluminal accumulations of large airway macrophages) in 27 of 38 soldiers diagnosed with DRCB.

The increased level of free active TGF-β1 and the persistent accumulation of alternatively activated macrophages in response to sustained club cell injury parallels the prior findings,^44,52 where alternatively activated macrophages accumulated in response to sustained targeted injury to type II alveolar epithelial cells. As both club cells and type II alveolar epithelial cells are regional progenitor cells within the airway and alveolar compartments, respectively, these findings suggest that sustained injury to lung progenitor cells might trigger comparable pathologic pathways in their corresponding microanatomic locations. Multiple studies44, 46, 47 showing that impairing accumulation of alternatively activated lung macrophages during alveolar epithelial injury can limit lung fibrogenesis, were seminal in testing whether depletion of lung macrophages would protect against CB in mice subjected to sustained club cell injury. As alveolar macrophages constituted the predominant leukocyte subpopulation during the chronic fibrotic phase (100-fold more that exudate macrophages), the current study focused on their depletion via oropharyngeal aspiration of CLs. Alveolar macrophage depletion ameliorated CB and was associated with reduced levels of free active TGF-β1. These findings suggest that alveolar macrophages promote peribronchiolar fibrosis in this animal model. In response to long-term alveolar macrophage depletion, increased accumulation of exudate macrophages in the lungs were identified. The possibility that these newly recruited exudate macrophages also play an antifibrotic role cannot be ruled out. Additional studies are needed to further delineate the protective effects of macrophage depletion in this model and to test whether strategies to inhibit macrophage accumulation or alternative macrophage activation could be used to treat CB.

The current findings demonstrate that sustained club cell injury is sufficient to induce CB and that external antigenic stimuli accompanied by host lymphocyte activation is not required. This finding is of interest as Liu et al⁵³ reported that, after allogeneic, but not syngeneic, lung transplant, transient club cell injury resulted in CD8⁺ T-cell–dependent obliterative bronchiolitis. Despite the substantial differences between the two models, their intriguing findings may suggest that antigenic stimuli during exposure to inhalational toxins could exacerbate club cell injury and the severity of the ensuing CB.

Altogether, the current data identify several persistent cellular and molecular pathways triggered by sustained club cell injury that are likely detrimental and can constitute attractive therapeutic targets. Furthermore, key histopathologic features of murine CB induced by sustained club cell injury are also present in the lungs of soldiers and veterans with DRCB, thereby implicating sustained club cell injury in the pathogenesis of this disorder. Additional investigations using this murine model may identify much-needed biomarkers of DRCB and lead to the design of future treatments.

Author Contributions

J.J.O. and S.T.-T. conceived the study; J.J.O., S.T.-T., A.-K.T.P., T.H.S., and V.N.L. developed study protocols; S.T.-T., S.P.V., A.J., Q.P., H.A., K.N.S., S.L., N.W., and N.S. collected data; J.J.O., S.T.-T., and T.H.S. analyzed and interpreted data; J.J.O., S.T.-T., A.-K.T.P., and T.H.S. prepared and edited the manuscript.

Supplemental Data

Supplemental Figure S1.

Identification of lung myeloid cells in murine constrictive bronchiolitis using flow cytometric analysis. Gating strategy used in flow cytometric analysis to identify CD45⁺ leukocytes (CD45⁺), neutrophils, eosinophils (Eos), CD103⁺ conventional dendritic cells (CD103⁺ cDCs), CD11b⁺ conventional dendritic cells (CD11b⁺ cDCs), monocyte-derived dendritic cells (moDCs), Ly6C⁺ monocytes (Ly6C⁺ monos), alveolar macrophages (AMs), and exudate macrophages (ExMs) within lung-derived single-cell suspensions obtained from CC-DTA and control mice at the indicated time points (representative plots from a CC-DTA mouse at protocol day 20 are shown; initial gates set to eliminate doublets, debris, and nonviable cells are not shown). FSC, forward scatter; MHC, major histocompatibility complex; SSC, side scatter.

Supplemental Figure S2.

Monocyte and macrophage accumulation in murine constrictive bronchiolitis (CB). Representative images of lung sections obtained from a CC-DTA mouse at protocol day 20. Murine frozen sections were stained with isotype control antibody (A and B), anti-F4/80 (C and D), and anti-CD11c (E and F) using an aminoethyl carbazole peroxidase reaction (red staining). Both specimens were counterstained with hematoxylin. Note the presence of smaller F4/80⁺ mononuclear cells localized to the subepithelial infiltrates (red arrows; C and D) and the larger CD11c⁺ cells with macrophage morphology residing within airway lumens (blue arrows; F). Scale bars: 100 μm (A); 40 μm (B–F). AW, airway.

Supplemental Figure S3.

Alveolar macrophage depletion using clodronate liposomes. A: Protocol day 5 enumeration of alveolar and exudate macrophages (by flow cytometric analysis) in the lungs of CC-DTA mice exposed to doxycycline and administered clodronate liposomes (CLs) or empty liposomes (ELs) on days 0 and 3. Data are from one experiment and presented as data from three mice. B: Protocol day 20 enumeration of alveolar and exudate macrophages in the lungs of CC-DTA and control mice exposed to doxycycline and administered clodronate or empty liposomes on protocol days 0, 3, 7, 10, 13, and 17. Cumulative data from two independent experiments are shown and presented as data from 10 mice. CC-DTA mice, open bars; control mice, closed bars. Data are presented as means ± SEM (A and B). ∗P < 0.05, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001 [unpaired t-test (A) or one-way analysis of variance, followed by the Tukey multiple comparisons test (B)].