Neutrophil transmigration triggers repair of the lung epithelium via β-catenin signaling

Abstract

Injury to the epithelium is integral to the pathogenesis of many inflammatory lung diseases, and epithelial repair is a critical determinant of clinical outcome. However, the signaling pathways regulating such repair are incompletely understood. We used in vitro and in vivo models to define these pathways. Human neutrophils were induced to transmigrate across monolayers of human lung epithelial cells in the physiological basolateral-to-apical direction. This allowed study of the neutrophil contribution not only to the initial epithelial injury, but also to its repair, as manifested by restoration of transepithelial resistance and reepithelialization of the denuded epithelium. Microarray analysis of epithelial gene expression revealed that neutrophil transmigration activated β-catenin signaling, and this was verified by real-time PCR, nuclear translocation of β-catenin, and TOPFlash reporter activity. Leukocyte elastase, likely via cleavage of E-cadherin, was required for activation of β-catenin signaling in response to neutrophil transmigration. Knockdown of β-catenin using shRNA delayed epithelial repair. In mice treated with intratracheal LPS or keratinocyte chemokine, neutrophil emigration resulted in activation of β-catenin signaling in alveolar type II epithelial cells, as demonstrated by cyclin D1 expression and/or reporter activity in TOPGAL mice. Attenuation of β-catenin signaling by IQ-1 inhibited alveolar type II epithelial cell proliferation in response to neutrophil migration induced by intratracheal keratinocyte chemokine. We conclude that β-catenin signaling is activated in lung epithelial cells during neutrophil transmigration, likely via elastase-mediated cleavage of E-cadherin, and regulates epithelial repair. This pathway represents a potential therapeutic target to accelerate physiological recovery in inflammatory lung diseases.

Many inflammatory lung diseases of the airways and parenchyma are characterized by migration of neutrophils from the pulmonary vasculature across the epithelium into the airspaces (1–3). Although the role of neutrophil migration in epithelial injury is a matter of debate (4), under pathological conditions neutrophil migration results in severe anatomical and physiological injury to the epithelium, including disruption of the intercellular junctions, apoptosis, and denudation, resulting in epithelial permeability (5, 6). This injury phase is followed by a repair phase in which an anatomically and functionally intact epithelium is restored in part through the proliferation of epithelial cells (reepithelialization) (7). Importantly, restitution of epithelial integrity is an important determinant of recovery and survival in acute lung injury (8). Furthermore, dysfunctional epithelial repair has been implicated in the pathogenesis of other inflammatory lung diseases, including chronic obstructive pulmonary disease and asthma (9). In addition to the well-known broad degradative effects of leukocyte proteinases, neutrophil transmigration can activate specific intracellular signaling pathways in epithelial cells by nondegradative mechanisms, including proteolytic processing (10). An intriguing possibility is that such signaling pathways contribute to the proliferation of epithelial cells that is critical for reepithelialization.

β-catenin exists in multiple cellular pools and serves diverse functions. In the basal state, β-catenin is localized mainly to the intercellular (adherens) junctions, where it is bound to the cytoplasmic domain of E-cadherin and plays a role in regulation of cell–cell adhesion (11). In the presence of canonical WNT signaling or as a consequence of destabilization of the cadherin–catenin complex, activated β-catenin translocates to the nucleus and binds to the TCF/LEF family of transcription factors, resulting in expression of specific target genes (11, 12). Because many β-catenin target genes, such as c-Myc, WISP1, and cyclin D1, are involved in cell proliferation, β-catenin signaling plays an important role in development and neoplasia (11). The WNT/β-catenin pathway also has been implicated in injury and repair (13), although not in all circumstances (14). Notably, β-catenin signaling is involved in the proliferation of epithelial cells that characterizes the dysfunctional epithelial repair of pulmonary fibrosis (15–17) and emphysema (18). To date there is limited evidence implicating β-catenin signaling in inflammatory lung injury (19, 20).

In the present study, we used transcriptional profiling to identify the activation of β-catenin signaling in lung epithelial cells in response to neutrophil transmigration. We hypothesized that β-catenin–dependent transcription is activated by neutrophil transmigration via specific signaling pathways and is critical to repair of the lung epithelium. Our data are relevant to the respiratory diseases in which epithelial repair after inflammatory injury is important for pathogenesis or recovery.

Results

Neutrophil Transmigration Induces Injury Followed by Repair of Lung Epithelial Cells.

Neutrophils were induced to transmigrate across epithelial monolayers in the physiological basolateral-to-apical direction. As illustrated in Fig. 1A, neutrophil transmigration induced a transient decrease in transepithelial resistance (TER). During neutrophil transmigration, individual scout neutrophils migrated across the epithelium at specific sites, and trailing neutrophils followed in the tracks of leading cells, crossing and ultimately disrupting the monolayer at these sites (Fig. 1 B–D). This pattern of neutrophil migration resulted in the formation of circular areas of epithelial denudation termed “wounds” in both Calu-3 cells (Fig. 1C) and primary human alveolar type II (ATII) epithelial cells (Fig. 1D). Over the subsequent 48–72 h, permeability resolved (Fig. 1A), and defects in the monolayer were reepithelialized (Fig. 1C).

Fig. 1.

Neutrophil transmigration induces injury, followed by repair of the lung epithelium. Polymorphonuclear neutrophils (PMN) were induced to migrate across Calu-3 (A and B) or Calu-3–GFP (C) or primary human ATII cell monolayers (D) for 90–120 min by a gradient of N-FORMYL-MET-LEU-PHE (fMLP) (1 μM). (A) TER was measured. (B) Cells were fixed, stained with H&E, and captured by digital images. Arrows indicate individual scout neutrophils migrating, closed arrowheads indicate trailing neutrophils, and the open arrowhead indicates the site of epithelial disruption. (C) Fluorescent images (10×) were acquired at the specified time points, and the total cross-sectional area of epithelial defects in 10 randomly selected HPF was measured. (D) Primary human ATII cells were imaged (10×) at the indicated time points. Arrows indicate neutrophils. *P < 0.05; n = 3–4. Error bars represent SEM.

Neutrophil Transmigration Induces Increased Expression of β-Catenin Target Genes in Lung Epithelial Cells in Vitro.

To identify the signaling pathways activated in epithelial cells in response to neutrophil transmigration, we examined cDNA isolated from purified lung epithelial cells after migration using a 44K (G4112F) whole-genome expression microarray. Ingenuity Pathway Analysis revealed significant modulation of the WNT/β-catenin pathway (Fig. 2A), as well as other pathways (SI Appendix, Fig. 1). Heatmap analysis revealed increased expression of multiple β-catenin target genes, including WISP1, MMP3, Axin2, DKK1, Fzd7, cyclin D1, and c-Myc (Fig. 2B).

Fig. 2.

Neutrophil transmigration induces β-catenin signaling in lung epithelial cells. (A and B) Neutrophils were induced to migrate across Calu-3 cells for 90 min. Epithelial cells were purified from contaminating neutrophils by magnetic-activated cell sorting separation. (A) cDNA from four separate experiments was pooled and analyzed using a whole-genome expression microarray (Agilent). Gene expression at the 24-h time point was analyzed using Ingenuity Pathway Analysis. A ratio of the number of molecules from the dataset that map to the specified pathway divided by the total number of molecules that map to that pathway is displayed. (B) Heatmap analysis of WNT/β-catenin pathway gene expression at 4 h after migration was constructed using hierarchical clustering. (C and D) Neutrophils were induced to migrate across Calu-3 or primary human ATII cells for 90–120 min. (C) Real-time qPCR analysis of indicated genes was performed on epithelial cDNA isolated at 2 h after the end of migration. (D) Epithelial monolayers were immunostained for c-Myc and WISP1 at 28 h after migration and imaged (10–20×). (Inset) Staining with isotype control antibody. *P < 0.05; n = 3–4. Error bars represent SEM.

Real-time quantitative PCR (qPCR) confirmed increased expression of the β-catenin–dependent genes c-Myc, WISP1, Fzd7, MMP3, and Axin2 (Fig. 2C), as well as others (SI Appendix, Fig. 2), in both Calu-3 and primary human ATII cells after neutrophil transmigration. Furthermore, neutrophil transmigration resulted in increased expression of c-Myc and WISP1 on the protein level, particularly at the sites of neutrophil migration, as determined by immunostaining (Fig. 2D). Knockdown of β-catenin using lentiviral-mediated shRNA (SI Appendix, Fig. 3), attenuated the increased expression of c-Myc and WISP1 in response to neutrophil transmigration (SI Appendix, Fig. 4).

Neutrophil Transmigration Induces Nuclear Translocation of β-Catenin and Activation of the TCF/LEF Reporter TOPFlash.

β-catenin was localized to the interepithelial junctions in quiescent Calu-3 and primary ATII cells. After neutrophil transmigration, there was decreased junctional β-catenin with accumulation in the nucleus (Fig. 3 A and B and SI Appendix, Fig. 5). As shown in Fig. 3C, neutrophil transepithelial migration resulted in activation of the TCF/LEF reporter TOPFlash but did not induce activity of FOPFlash, a reporter with mutated TCF/LEF-binding sites.

Fig. 3.

Neutrophil transmigration induces nuclear translocation of β-catenin and β-catenin–dependent transcription. (A and B) Neutrophils were induced to migrate across Calu-3 cells for 90 min. Epithelial cells were fixed, stained with the indicated antibody, and imaged. (Original magnification, 40×.) Arrowheads indicate nuclear staining, and arrows indicate cytoplasmic sparing of the stain. (Inset) Staining with isotype control antibody. (C) Neutrophils were induced to migrate for 90 min across Calu-3 cells cotransfected with Super8× TOPFlash (TOP) or Super8×FOPFlash (FOP) and CMV–β-galactosidase vectors. Then, 2 h later, luciferase and β-galactosidase assays were performed on cell lysates. *P < 0.05; n = 4. Error bars represent SEM.

β-Catenin Signaling Mitigates Epithelial Injury and Accelerates Epithelial Repair After Neutrophil Transmigration.

Knockdown of β-catenin by shRNA (SI Appendix, Fig. 3) resulted in intensified epithelial injury and delayed repair after neutrophil transmigration, as measured by TER, the total area of epithelial denudation, and the number of epithelial defects (Fig. 4 and SI Appendix, Fig. 6). Notably, epithelial cells were able to generate comparable baseline TER after knockdown of β-catenin.

Fig. 4.

Knockdown of β-catenin exacerbates injury and delays repair of the lung epithelium after neutrophil transmigration. Neutrophils were induced to migrate across Calu-3 cells transduced with the pGIPZ lentivirus expressing either shRNA to β-catenin or a nonsilencing shRNA. (A) TER was measured with an Evometer. (B) Images were captured at the indicated time points. (Original magnification, 10×.) The total cross-sectional area of epithelial defects was measured in 10 HPFs per experiment using ImageJ software. *P < 0.05; ^‡P < 0.01; n = 5. Error bars represent SEM.

Neutrophil Transmigration Induces Activation of β-Catenin–Dependent Transcription in ATII Cells in Murine Models.

To address the relevance of the findings in our in vitro model, we assessed whether β-catenin signaling was activated in the alveolar epithelium by neutrophil transmigration in murine models of lung injury. Intratracheal (i.t.) LPS treatment induced β-catenin–dependent transcription in ATII cells, as evidenced by β-galactosidase activity that colocalized with pro-surfactant protein C (SPC) expression in the TCF/LEF reporter strain TOPGAL (Fig. 5A) (21). LPS treatment also resulted in enhanced expression of cyclin D1 in C57BL/6 mice (Fig. 5B and SI Appendix, Fig. 7).

Fig. 5.

β-catenin signaling is induced by neutrophil transmigration and is critical for repair of the alveolar epithelium in murine models. (A) TOPGAL mice were treated with saline or LPS 20 μg i.t. Mice were euthanized at 24 h. (B and C) C57BL/6 mice were treated with saline or KC 1 μg i.t. or LPS 20 μg i.t. and euthanized at 6 or 24 h, respectively. (C) Mice were treated with anti-Ly6G antibody 125 μg i.p. 24 h before KC 1 μg i.t. (D) C57BL/6 mice were treated with IQ-1 1 mg s.c. at 2 h after KC 1 μg i.t. and euthanized at 24–48 h. (A–C) β-galactosidase (LacZ) and immunohistochemical staining for pro-SPC and/or cyclin D1 was performed on serial sections (B, Right). (D) Immunofluorescence staining with BrdU and pro-SPC or isotype control antibodies (Inset) was performed. (B–D) Images were captured at 20–40× magnification, and positive cells were counted in a blinded manner on 25 random HPFs. Red arrows indicate cells that stain for LacZ and pro-SPC. Black arrows indicate cells that stain for cyclin D1. Open black arrowheads indicate cells that stain for both cyclin D1 and pro-SPC. Closed black arrowheads indicate cells that stain for pro-SPC, but not for cyclin D1. White arrowheads indicate cells that stain for both BrdU and pro-SPC *P ≤ 0.05; n = 3–6 mice/group. Error bars represent SEM.

Because LPS is a broadly acting stimulus that affects many cell types, we sought to confirm that β-catenin activation in lung epithelial cells was attributable to neutrophil transmigration. Mice were treated with i.t. keratinocyte chemokine (KC), a neutrophil chemoattractant that induces robust neutrophil migration into the lung and epithelial permeability, as measured by concentration of IgM in bronchoalveolar lavage (BAL) fluid (SI Appendix, Fig. 8), but has little if any direct effect on epithelial cells. KC treatment increased expression of cyclin D1 in ATII cells, as determined by immunostaining for cyclin D1 and pro-SPC (Fig. 5B). Pretreatment with anti-Ly6G antibody effectively depleted neutrophils, as demonstrated by FACS analysis of peripheral blood (SI Appendix, Fig. 9), and abrogated the increased expression of cyclin D1 in response to i.t. KC (Fig. 5C).

β-Catenin Signaling Is Critical for Repair of the Injured Alveolar Epithelium in Vivo.

Mice were treated with the β-catenin/p300 inhibitor IQ-1 (22), which attenuated the activation of β-catenin as measured by cyclin D1 up-regulation in ATII cells after i.t. KC (SI Appendix, Fig. 10A), a trend that did not achieve statistical significance (P = 0.09). Importantly, attenuation of β-catenin activation inhibited ATII cell proliferation in response to neutrophil transmigration, as assessed by BrdU (Fig. 5D) or Ki-67 (SI Appendix, Fig. 10B) staining. Notably, β-catenin inhibition did not affect neutrophil influx or the magnitude of epithelial injury, as measured by IgM in BAL fluid in response to i.t. KC (SI Appendix, Fig. 11).

Neutrophil Transmigration Induces Elastase-Mediated Cleavage of E-Cadherin with Up-Regulation of β-Catenin Signaling.

Neutrophil transmigration results in loss of β-catenin (Fig. 3A and SI Appendix, Fig. 5) and E-cadherin (Fig. 6A) from the adherens junctions. To assess for E-cadherin cleavage during transmigration, we concentrated epithelial cell supernatants and analyzed them by immunoblot using an antibody (DECMA-1) that recognizes the extracellular domain of E-cadherin. Neutrophil transmigration resulted in cleavage of a ∼50-kDA extracellular fragment of E-cadherin from the ∼115-kDA full-length protein on the surface of Calu-3 or primary human ATII cells (Fig. 6B and SI Appendix, Fig. 12). Preincubation of neutrophils with the cell permeant compound human neutrophil elastase inhibitor IV attenuated E-cadherin cleavage during neutrophil migration (Fig. 6B). In addition, elastase inhibition prevented nuclear translocation of β-catenin (SI Appendix, Fig. 13A), activation of TOPFlash (Fig. 6C), and up-regulation of c-Myc and WISP1 (Fig. 6D and SI Appendix, Fig. 13B) in epithelial cells in response to neutrophil transmigration. Notably, inhibition of elastase abrogated (rather than delayed) the up-regulation of β-catenin target genes after neutrophil migration (SI Appendix, Fig. 14). Finally, direct stimulation of epithelial cells with leukocyte elastase increased epithelial expression of Axin2 (Fig. 6E). These observations are consistent with the notion that during neutrophil transmigration, elastase-mediated cleavage of epithelial E-cadherin results in the translocation of β-catenin from the junctional pool to the nucleus, where it triggers the transcription of specific target genes.

Fig. 6.

Neutrophil transmigration induces elastase-mediated cleavage of E-cadherin and activation of β-catenin signaling. Neutrophils were left untreated (A) or were pretreated with 0.4% DMSO or 100 μM human neutrophil elastase inhibitor IV for 45 min (B–D) and induced to migrate across Calu-3 cells (A–D) or primary human ATII cells (B) for 90–120 min. (A) Cells were fixed, stained with anti–E-cadherin or isotype control (Inset) antibodies, and imaged. (Original magnification, 40×.) (B) Epithelial cell supernatants were harvested, concentrated by ultracentrifugation through a 10-kDA MW filter, separated by SDS/PAGE, and immunoblotted with an antibody that recognizes the extracellular domain of E-cadherin (DECMA-1). (C) Calu-3 cells were cotransfected with Super8× TOPFlash and renilla luciferase vectors before transmigration. Firefly and renilla luciferase assays were performed on cell lysates. (E) Calu-3 cells were treated with 0.1–0.25 U/mL purified elastase for 1 h, then incubated in media for 2 h. (D and E) Epithelial cDNA was subjected to real-time qPCR for c-Myc or Axin2. *P < 0.05. A, B, and D are representative of at least three experiments. Error bars represent SEM.

Discussion

Epithelial injury is central to the pathogenesis of diverse inflammatory lung diseases and may be attributable in part to neutrophil transmigration across the epithelium (23). In this study, we investigated the novel hypothesis that as neutrophils migrate across the lung epithelium, they activate specific pathways in surviving epithelial cells that promote repair, a concept supported by a nascent literature (24, 25). We demonstrated that neutrophil transmigration activates β-catenin signaling in lung epithelial cells in vitro and in animal models. Because repair of the denuded epithelium in lung injury requires proliferation of ATII cells (7, 9) and many β-catenin target genes are involved in cell proliferation, we hypothesized that enhanced β-catenin signaling in alveolar epithelial cells in response to neutrophil transmigration contributes to epithelial repair after injury. We demonstrated in both in vitro and in vivo models that inhibition of β-catenin delays epithelial repair after neutrophil-mediated injury.

In our in vitro studies, inhibition of β-catenin delayed, but did not completely prevent, reepithelialization of the denuded epithelium (Fig. 4B). Potential explanations for this finding include incomplete knockdown of β-catenin (SI Appendix, Fig. 3), substitution of transcriptionally active γ-catenin in the setting of β-catenin deficiency (26), and activation of other repair or growth factor pathways. In our in vivo model, inhibition of β-catenin/p300 blocked ATII cell proliferation, possibly by impairing the asymmetric differentiation and thereby activation of quiescent progenitor cells into transient amplifying cells. The relative importance of ATII cell proliferation versus terminal differentiation to ATI cells in repair of the alveolar epithelium after neutrophil transmigration, and the role of β-catenin signaling in these distinct processes, are unknown and merit further investigation, as does the role of individual β-catenin target genes in epithelial repair.

Our observations are consistent with previous reports of the role of β-catenin signaling in tissue repair (13, 27). In the lung, β-catenin signaling has been implicated in pulmonary fibrosis, the pathogenesis of which is thought to involve repetitive cycles of injury and repair of ATII cells (15–17). In emphysema, dysfunctional repair is associated with decreased β-catenin signaling (18). Up-regulation of β-catenin target gene transcription has been demonstrated in animal models (19, 20) and patients (28) with acute lung injury, and inhibition of β-catenin signaling impairs epithelial repair after injury in some (19), but not all (14), circumstances. Further studies are indicated to determine whether activation of β-catenin–dependent signaling pathways in response to neutrophil transmigration in acute inflammatory injury leads to physiological repair or pathological remodeling.

Although β-catenin signaling has been implicated in tissue repair in other contexts, our data provide important mechanistic information suggesting that these responses may be mediated by specific, well-regulated signaling events initiated by neutrophil proteinases. Cleavage of cadherins by proteinases, with subsequent nuclear translocation of β-catenin and activation of β-catenin–dependent transcriptional events, has been documented in endothelial cells (29, 30), in keratinocytes (27), and during pancreatitis (31). Shedding of an extracellular fragment of E-cadherin in animal models of lung injury has been reported (32, 33). The results reported here support the novel concept that β-catenin signaling is activated in lung epithelial cells via elastase-mediated cleavage of E-cadherin during neutrophil transmigration. The release of β-catenin from E-cadherin may be a result of a conformational change in E-cadherin in response to cleavage of its extracellular domain. It is relevant to note that when neutrophils are in close proximity to epithelial cells during transmigration, the adjacent membranes form a protected space in which neutrophil proteinases can attain high concentrations and from which larger molecules, such as anti-proteinases, are excluded (34). The 50-kDA E-cadherin fragment detected in the present study is likely attributable to elastase cleavage (Fig. 6B). However, cleavage of E-cadherin by other proteinases, particularly matrix metalloproteinases, has been reported and can stimulate diverse repair processes (27, 33), including the retention of protective CD103⁺ dendritic cells in the lung (35). Further studies are indicated to determine whether similar phenomena may be occurring in our model. In addition to the release of cadherin-bound β-catenin, canonical WNT signaling might be activated by neutrophil transmigration, given the up-regulation of several epithelial WNTs (Fig. 2B and SI Appendix, Fig. 2).

Although the primary focus of the present study was on repair, our findings also suggest that inhibition of β-catenin may exacerbate epithelial injury (Fig. 4). The interpretation of these studies is complex, because injury and repair processes are likely to occur concurrently, such that interference with repair by β-catenin inhibition might appear to exacerbate injury. In addition, β-catenin signaling may attenuate injury by preventing the disruption of the intercellular junctions and the consequent paracellular permeability that results from neutrophil migration. In addition, because β-catenin target genes are involved in cell survival (11), and, as we previously demonstrated, apoptosis is one mechanism of epithelial injury due to neutrophil transmigration (5), it is possible that inhibition of β-catenin signaling exacerbates injury via down-regulation of survival genes.

A limitation of the present study is the use of Calu-3 cells, neoplastic cells likely of airway rather than alveolar epithelial origin. Despite this limitation, Calu-3 cells are a robust cell line for the study of injury and repair because of their ability to form an impermeable monolayer with electrical resistance (Fig. 1). We have confirmed key findings in both primary human ATII cells and mouse models, suggesting that the biological behavior of the Calu-3 cells in this setting is reflective of both airway and alveolar epithelial cells. Nonetheless, a monolayer of primary human ATII cells does not fully recapitulate the in vivo situation, in which the alveolar epithelium is lined with both ATII cells and ATI cells. In fact, neutrophil migration likely occurs at the tricellular junctions between two ATI cells and one ATII cell, rather than between two ATII cells (36). We chose to establish our in vitro model of neutrophil transepithelial migration using an ATII rather than an ATI cell monolayer, because the ATII cells proliferate (and differentiate into ATI cells) to reconstitute the denuded alveolar epithelium after injury (7).

In conclusion, we have provided strong evidence that neutrophil transmigration across the lung epithelium induces β-catenin signaling, likely via elastase-mediated cleavage of E-cadherin, and that this pathway plays a critical role in repair of the injured epithelium. Importantly, because epithelial repair after injury is critical to the clinical outcome in inflammatory lung diseases, including acute lung injury (1, 8, 9), pharmacologic manipulation of the β-catenin pathway has promising therapeutic potential.

Methods

Cell Culture and Neutrophil Transmigration.

Calu-3 cells (0.75 × 10⁶/well) were seeded on the undersurface of polycarbonate membranes in Transwell tissue culture inserts (Corning) and grown to confluence (TER ≥ 1,000 Ω·cm²). Human ATII cells (1 × 10⁶/well) obtained from explanted human lungs not suitable for transplantation were cultured in Transwell tissue culture inserts (Millicell). Six million freshly isolated human neutrophils, pretreated with 100 μM human neutrophil elastase inhibitor IV where indicated, were added to the upper chamber, and 1 μM fMLP was added to the lower chamber. Transmigration was allowed to proceed for 90–120 min. TER was measured, and images were obtained.

Magnetic-Activated Cell Sorting Separation.

After migration, epithelial cells were purified from neutrophils by magnetic-activated cell sorting separation using anti–epithelial cell adhesion molecule antibodies.

Expression Microarray Analysis.

cDNA isolated from purified epithelial cells was hybridized on Agilent 44K human whole-genome arrays. Data from the array were analyzed using GeneSpring GX9 (Agilent Technologies) and Ingenuity Pathway Analysis (Ingenuity Systems).

Real-Time qPCR.

Epithelial cell cDNA was analyzed by the WNT signaling pathway RT² Profiler PCR Array (SABiosciences) or real-time qPCR using specific primers for Axin2, c-Myc, Fzd7, MMP3, WISP1, GAPDH, and HHPRT.

Immunoblotting.

Epithelial cell lysates or supernatants were analyzed by SDS/PAGE and immunoblotting for β-catenin, α-tubulin, or E-cadherin (DECMA-1).

Immunofluorescence.

Epithelial monolayers were fixed and stained for active β-catenin, β-catenin, E-cadherin, c-Myc, and WISP1.

Transfection.

Calu-3 cells were transfected with Super8× TOPFlash or Super8× FOPFlash and CMV–β-galactosidase or renilla luciferase vectors. Neutrophil transmigration was performed, and firefly and renilla luciferase and β-galactosidase activity was measured.

Lentiviral Transduction.

Calu-3-GFP cells were generated by transduction of the HIV-1 GFP lentiviral vector into Calu-3 cells. pGIPZ lentivirus containing shRNA to β-catenin or nonsilencing shRNA was transduced into Calu-3 cells.

Preparation of Epithelial Cell Supernatants.

After transmigration, supernatants on the apical surface of the epithelial monolayer were concentrated by centrifugation and boiled in Laemmli buffer.

Elastase Treatment.

Calu-3 cells were treated with 0.1–0.25 U/mL of human leukocyte elastase at 37 °C for 1 h and then incubated in media for 2 h.

Animal Models.

Female C57BL/6 or TOPGAL(B6) mice were treated with 20 μg of LPS or 1 μg of recombinant murine KC i.t. In selected experiments, mice were treated with 125 μg of anti-Ly6G antibody i.p. at 24 h before i.t. KC or with 1 mg of IQ-1 s.c. at 2 h after i.t. KC. Mice were euthanized at selected time points, BAL was performed, and lungs were inflation-fixed. IgM concentrations in BAL fluid were measured by ELISA.

Immunohistochemistry and LacZ Staining.

Immunohistochemistry for cyclin D1, BrdU, Ki-67, and pro-SPC and LacZ staining was performed on lung sections.

Statistical Analysis.

Data are expressed as mean ± SEM. Unless indicated otherwise, data were analyzed from three or more independent experiments done in duplicate or triplicate. Multiple comparisons were performed by one-way ANOVA with the Tukey or Bonferroni (post hoc) test for determination of differences between groups. Statistical analysis was performed using the Student paired or unpaired t test or the Wilcoxon signed-rank test as indicated. For analysis of the area of microscopic epithelial defects, the t test was performed on log₁₀ of the total cross-sectional area. P < 0.05 was considered significant. GraphPad PRISM software was used for all statistical calculations.