Blocking airway mucous cell metaplasia by inhibiting EGFR antiapoptosis and IL-13 transdifferentiation signals

Abstract

Epithelial hyperplasia and metaplasia are common features of inflammatory and neoplastic disease, but the basis for the altered epithelial phenotype is often uncertain. Here we show that long-term ciliated cell hyperplasia coincides with mucous (goblet) cell metaplasia after respiratory viral clearance in mouse airways. This chronic switch in epithelial behavior exhibits genetic susceptibility and depends on persistent activation of EGFR signaling to PI3K that prevents apoptosis of ciliated cells and on IL-13 signaling that promotes transdifferentiation of ciliated to goblet cells. Thus, EGFR blockade (using an irreversible EGFR kinase inhibitor designated EKB-569) prevents virus-induced increases in ciliated and goblet cells whereas IL-13 blockade (using s-IL-13Ralpha2-Fc) exacerbates ciliated cell hyperplasia but still inhibits goblet cell metaplasia. The distinct effects of EGFR and IL-13 inhibitors after viral reprogramming suggest that these combined therapeutic strategies may also correct epithelial architecture in the setting of airway inflammatory disorders characterized by a similar pattern of chronic EGFR activation, IL-13 expression, and ciliated-to-goblet cell metaplasia.

Figure 1

Persistent EGFR activation on ciliated epithelial cells after viral infection. (A) Representative photomicrographs of airway sections from C57BL/6J mice obtained at 21 days after inoculation with SeV or an equivalent amount of SeV-UV and then immunostained for EGFR and p-EGFR as well as competition by 50-fold antigen (Ag) excess. Scale bar: 20 μm. (B) Representative photomicrographs of airway sections obtained from mice at 21 days after inoculation with SeV and then subjected to immunofluorescent staining for EGFR, β-tubulin, CCSP, and MUC5AC alone and in combination. Primary anti-EGFR Ab binding was detected by anti-CY3 Ab (red fluorescence) while others were detected by anti-FITC Ab (green fluorescence). Scale bar: 20 μm. Similar results were obtained for mice treated with SeV-UV.

Figure 2

Effect of EGFR blockade on airway epithelial remodeling after viral infection. (A) Photomicrographs of airway sections obtained at 21 days after inoculation with SeV or SeV-UV and then subjected to immunofluorescent staining for β-tubulin (green fluorescence) and CCSP (red fluorescence) and immunostaining for MUC5AC. Immunostaining with nonimmune IgG gave no signal above background (data not shown). Scale bars: 20 μm. (B) Corresponding quantitative data for conditions in A plus postinoculation day 12 (PI day 12) without treatment and day 21 after treatment with EKB-569 for days 10–21 after inoculation. Values represent mean ± SEM. *Significant difference from SeV-UV control.

Figure 3

Airway epithelial remodeling without cellular proliferation in genetically susceptible mice. (A) Representative photomicrographs of airway sections obtained from C57BL/6J and Balb/cJ mice at indicated days after inoculation with SeV and subjection to immunostaining for BrdU. (B) Corresponding quantitative data for conditions in A. (C) Representative photomicrographs of airway sections from indicated conditions, immunostained for p-EGFR. (D) Representative photomicrographs of airway sections from indicated conditions, immunostained for MUC5AC. (E) Quantitative morphometry for airway sections that were obtained from Balb/cJ mice at 21 days after inoculation with SeV or SeV-UV and subjection to immunostaining for β-tubulin, CCSP, and MUC5AC. For B and E, values represent mean ± SEM. *Significant difference from day 0 or SeV-UV control. Scale bars: 20 μm.

Figure 4

Effect of EGFR signaling pathways on ciliated epithelial cell death in culture. (A) Representative photomicrographs of airway epithelial cell cultures placed under air-liquid interface conditions for 10 days followed by immunostaining for EGFR (top panel) or double immunofluorescence and confocal microscopy for β-tubulin and either EGFR (middle panel) or p-EGFR (bottom panel). Scale bars: 20 μm. (B) Western blot analysis of mTEC cultures that were placed in basic medium for 1 day and then treated with EGF (1 or 10 ng/ml) for 10 minutes with or without concomitant inhibitor. Each inhibitor was added at maximal effective concentrations to the lower chamber for 6 hours and the upper chamber for 2.5 hours before addition of EGF to both chambers. For each condition, cell lysates were blotted against anti-EGFR, p-EGFR, p-Akt, or p-ERK1/2 Ab, and Ab binding was detected by enhanced chemiluminescence. (C) Representative photomicrographs of mTEC cultures that were treated with vehicle or PD153035 (0.3 μM) for 7 days at 37°C and subjected to immunofluorescent staining for β-tubulin IV and Hoescht 33432. Scale bar: 20 μm. (D) Quantitative analysis of β-tubulin staining cells (expressed as a percentage of total Hoechst staining cells) without and with treatment with PD153035, LY294002, and PD98059, given at the indicated doses for 7 days. *Significant difference versus 0 μM.

Figure 5

Effect of EGFR signaling pathways on apoptosis in airway epithelial cell cultures. (A) Representative photomicrogaphs of mTEC cultures that were treated with vehicle, PD153035 (0.3 μM), LY294002 (50 μM), and PD98059 (50 μM) for 3 days at 37°C and then subjected to immunofluorescent staining for cleaved fragment of active caspase-3 (act caspase-3) or TUNEL reaction. Scale bar: 20 μm. (B) Quantitative analysis of information in A for active caspase-3 staining cells (expressed as percentages of total Hoechst staining cells), using treatment conditions from A as well as PD15305 plus z-VAD-fmk (100 μM). (C) Immunoblot analysis of active caspase-3 and caspase-9 in cell lysates from mTEC cultures using treatment conditions from A. Anti–caspase-9 antibody recognizes precursor (caspase-9) and the cleaved fragment of active caspase-9. (D) Flow cytometric analysis of JC-1 staining of mTEC cultures using treatment conditions from A. Values represent percentages of cells with decreased mitochondrial membrane potential (ΔΨm) detected by shift from FL2 to FL1. For B and D, values represent mean ± SEM. *Significant difference from vehicle alone.

Figure 6

Effect of EGFR inhibition on ciliated but not goblet cell death in vitro. (A) Representative photomicrographs of mTEC cultures treated with or without IL-13 (100 ng/ml for 5 days) and with or without subsequent PD153035 (0.3 μM for 3 days) and subjected to immunofluorescent staining for MUC5AC (red) and active caspase-3 (green) as well as counterstaining with Hoescht dye (blue). Scale bar: 20 μm. (B) Corresponding quantitative data for A. Values represent mean ± SEM for percentage of active caspase-3⁺ goblet cells (number of MUC5AC⁺ active caspase-3⁺ cells × 100 / number of MUC5AC⁺ cells) and active caspase-3⁺ nongoblet cells (total number of active caspase-3⁺ cells × 100 / total Hoescht staining cells). *Significant difference from vehicle control.

Figure 7

Identification of IL-13–dependent ciliated-to-goblet cell transdifferentiation in vitro. Representative transmission electron micrographs are shown for cultured mTECs before treatment (upper left panel) and then after treatment with IL-13 (100 ng/ml for 2 days at 37°C; all other panels). Early cilia-goblet cells are identified with cilia that are visible on the surface of cells that also contain a few mucous granules, late cilia-goblet cells exhibit greater numbers of mucous granules in the cytoplasm, and mature goblet cells contain characteristic mucous granules with no cilia.

Figure 8

Identification and blockade of cilia-to-goblet cell transdifferentiation in vitro and in vivo. (A) Representative photomicrographs of airway sections obtained from mice at 21 days after SeV inoculation and subjected to confocal immunofluorescence microscopy for β-tubulin (green) and MUC5AC (red). Arrows indicate ciliated cells staining for β-tubulin (ci), goblet cells staining for MUC5AC (g), and cells staining for both β-tubulin and MUC5AC (cig). (B) Representative photomicrographs of airway sections obtained as in A but immunostained for p-EGFR (red) and MUC5AC (green). Arrows indicate ciliated cells staining for p-EGFR (ci), goblet cells staining for MUC5AC (g), and cells staining for both p-EGFR and MUC5AC (cig). (C) Representative photomicrographs of airway sections obtained as in A but immunostained for CCSP (green) and MUC5AC (red). Arrows indicate cells staining for CCSP (cc) or CCSP and MUC5AC (ccg). Scale bars: 20 μm. (D) Quantitative analysis of MUC5AC-expressing cells that also immunostained for CCSP or β-tubulin. (E) Real-time PCR analysis of lung IL-13, mCLCA3, and MUC5AC mRNA levels corrected for GAPDH control level at indicated times after SeV inoculation. (F) Quantitative analysis of β-tubulin, CCSP, and Muc5AC immunostaining in mice inoculated with SeV and treated with sIL-13Rα2-Fc or control IgG on days 12, 14, 17, and 20 after inoculation. Values represent mean ± SEM *Significant difference from corresponding SeV-UV control for D and E or IgG treatment for F.

Figure 9

Evidence of cilia-to-goblet cell transdifferentiation in human epithelial cells in vivo and in vitro. (A) Representative photomicrographs from lung sections obtained from COPD patients and immunostained for β-tubulin and MUC5AC, counterstained with DAPI, and viewed with immunofluorescence microscopy (left panel) or immunostained for β-tubulin and MUC5AC or CCSP and MUC5AC and viewed with laser confocal scanning microscopy (middle and right panels). Arrows and outlines indicate goblet cells that express MUC5AC (g), Clara cells that express CCSP (cc), cilia-goblet cells that coexpress β-tubulin and MUC5AC (cig), and goblet cells that coexpress CCSP (ccg). (B) Representative photomicrographs of human large airway epithelial cells (hLAECs) cultured from COPD patients, incubated with IL-13 (100 ng/ml) for 5 days, and then immunostained for β-tubulin (red) and MUC5AC (green). (C) Representative photomicrographs of hLAECs cultured from control (non-COPD) subjects, incubated with IL-13 for 1 day, immunostained as in B, and then viewed with laser confocal scanning microscopy in x-y axis and z axis views. In B and C, arrows indicate cells that immunostained for both γ-tubulin and MUC5AC. Scale bar: 20 μm.

Figure 10

Scheme for virus-inducible EGFR- and IL-13–dependent pathways controlling epithelial host response and remodeling. EGFR activation with receptor dimerization and receptor tyrosine kinase phosphorylation leads to Gab1 recruitment followed by PI3K activation that causes generation of phosphatidylinositol-3,4,5-phosphate (PI-3,4,5-P₃) and activation of PDK1/2 and then Akt that inactivates proapoptotic factors at the level of the mitochondria and promote cell survival. IL-13 signaling activates IRS1/2-dependent cascade to ERK1/2 and Stat6, which each contribute to upregulation of genes (CLCA and MUC) that promote cilia to goblet cell transdifferentiation. Under physiologic conditions, these pathways may (in conjunction with IFN-dependent activation of Stat1) lead to protection from viral infection, but if there is persistent activation in a susceptible genetic background, the same pathways may lead to ciliated cell hyperplasia and goblet cell metaplasia. Rational use of specific inhibitors, e.g., EGFR and IL-13R blockers, may fully restore normal epithelial architecture. Grb2, growth factor receptor–bound 2; PTEN, phosphatase and tensin homologue deleted on chromosome 10; Sos, son-of-sevenless.