SPDEF regulates goblet cell hyperplasia in the airway epithelium (original) (raw)
SPDEF is expressed in pulmonary epithelial cells. Mouse Spdef mRNA was detected in adult mouse lungs and trachea, but not in the MLE-12 cell line, an SV40 immortalized mouse lung epithelial cell line with characteristics of type II alveolar cells (Figure 1A). By in situ hybridization, Spdef mRNA was found to be present in subsets of respiratory epithelial cells in conducting airways of the mouse lung from E17.5 to adulthood (Figure 1 and Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI29176DS1). In lungs of adult mice, Spdef mRNA was readily detected in subsets of epithelial cells in trachea, bronchi, and epithelial cells of tracheal glands (Figure 1, B–E). Spdef mRNA was also present in H441, a human pulmonary adenocarcinoma cell line, and HTEpC human tracheal epithelial cells, but not in HeLa cells (Figure 1A). In situ hybridization demonstrated the presence of Spdef mRNA in epithelial cells of the stomach, small intestine, caecum, colon, oviduct, dorsal and ventral prostate, coagulating gland, and seminal vesicles (Supplemental Figure 2), consistent with the previously reported distribution of Spdef mRNA in the adult mouse (25). An Spdef sense probe did not hybridize (Supplemental Figure 3).
Spdef mRNA in mouse respiratory epithelial cells, trachea, and tracheal glands. (A) SPDEF and GAPDH expression was identified by RT-PCR using RNA extracts from human lung adenocarcinoma H441 cells, cervical adenocarcinoma HeLa cells, normal human tracheal epithelial HTEpC cells, SV40 large T immortalized mouse lung epithelial MLE-12 cells, mouse lung (mLu), and mouse trachea (mTra). Spdef mRNA was detected in H441 cells, HTEpC cells, and mouse lung and trachea, but not in HeLa or MLE-12 cells. PCR without RT (–) showed no product. (B–E) In situ hybridization for Spdef mRNA was performed on sections of trachea and tracheal glands (B and D) and lung (C and E) in adult mice. Spdef mRNA was detected in the epithelium lining trachea, bronchi (B), and tracheal glands (arrows), but not in bronchioles (Br) or blood vessels (V). Inset shows phase microscopy of the hybridized tracheal glands; original magnification, ×4. Scale bars: 200 μm. C, cartilage.
Polyclonal antisera were produced against recombinant SPDEF that detected a single protein of approximately 37 kDa by immunoblot. SPDEF antiserum–immunostained HeLa cells were transfected with a full-length mouse Spdef cDNA (Supplemental Figure 4). Consistent with our mRNA data, SPDEF was detected in epithelial cells lining the adult mouse trachea (Figure 2). Neither SPDEF staining nor mRNA was detected in bronchioles or alveoli (data not shown). Nuclear staining for SPDEF was observed in epithelial cells in trachea, bronchi, and tracheal glands, consistent with the distribution of Spdef mRNA detected by in situ hybridization (Figure 2). Sites of SPDEF expression in conducting airways overlapped with expression of TTF-1, SRY-box containing gene 17 (SOX17), FOXJ1, and SCGB1a1 (Figure 2). Spdef mRNA and protein were observed in prostate, oviduct, colon, and seminal vesicles (Supplemental Figures 2 and 5). In the adult lung, levels of staining intensity for SPDEF varied in nuclei of respiratory epithelial cells lining conducting airways (Figure 2, A and B). Relatively high levels of Spdef mRNA and immunostaining were observed in epithelial cells of tracheal glands (Figures 1 and 2). During development, Spdef mRNA was first detected in conducting airway epithelial cells of the fetal mouse lung at E17.5 (Supplemental Figure 1). Thereafter, Spdef was present in epithelial cells lining trachea and bronchi, but was not detected by either in situ hybridization or immunohistochemistry in peripheral bronchiolar and alveolar epithelial cells (Figures 1 and 2). Timing and sites of expression of SPDEF support its potential role in cell differentiation or gene expression in tracheal glands and proximal conducting airways, but not in peripheral bronchioles or alveolar epithelial cells in the mouse.
SPDEF immunohistochemistry in mouse trachea and tracheal glands. Immunohistochemistry was performed on sections of adult lungs. (A and B) SPDEF staining was detected in nuclei in epithelial cells lining the trachea (A) and tracheal glands (B). (C–F) Immunohistochemistry for SPDEF (A and B), TTF-1 (C), SOX17 (D), FOXJ1 (E), and SCGB1a1 (F) in tracheal epithelium. Scale bars: 50 μm.
SPDEF interacts with TTF-1 and regulates gene expression in respiratory epithelial cells. Because SPDEF was selectively expressed in epithelial cells lining tracheal glands and proximal conducting airways, the ability of SPDEF to regulate potential transcriptional targets expressed at these cellular sites was assessed by transfection assays in vitro. Surfactant protein A is a host defense protein that is selectively expressed in epithelial cells of tracheal glands as well as bronchiolar and alveolar type II cells (26). SPDEF enhanced the activity of the Sftpa promoter in vitro (Figure 3A). Cotransfection of SPDEF with TTF-1 further activated the Sftpa promoter (Figure 3A). Potential SPDEF binding motifs GGAA/T (27) were identified in the promoter; however, repeated attempts to bind recombinant SPDEF or the DNA-binding domain of SPDEF to consensus SPDEF elements in the Sftpa promoter (aa 247–335) by EMSA were unsuccessful. Deletion of these potential SPDEF binding sites in the Sftpa gene promoter did not inhibit its activation in the transfection assays (data not shown), suggesting that the effects of SPDEF on transcription may be mediated indirectly.
SPDEF and TTF-1 activate gene transcription in vitro. Reporter assays were performed using plasmids expressing SPDEF and TTF-1 and reporter plasmids in which firefly luciferase gene (luc) is controlled by Sftpa (A), Foxj1 (B), Scgb1a1 (C), and Sox17 (D) gene promoters as described in Methods. SPDEF activated Sftpa, Foxj1, Scgb1a1, and Sox17 promoters in the presence and absence of TTF-1. Controls (Con) were cells transfected with the reported and empty expression plasmids. Assays were repeated in triplicate in at least 3 experiments with similar results. Values are mean ± SD (n = 3). P values shown were obtained by ANOVA.
Because SPDEF was expressed in conducting airway epithelial cells, we tested whether SPDEF regulates the promoters of other genes expressed selectively in the proximal airways, including Foxj1, Sox17, Scgb1a1, and mucin 5, subtypes A and C, (Muc5A/C). SPDEF acted additively or synergistically with TTF-1 on the promoters of Foxj1, Scgb1a1, and Sox17 (Figure 3, B–D), but did not directly activate the MUC5A/C gene promoter (data not shown). Again, we were unable to demonstrate direct binding of the recombinant SPDEF homeodomain or full-length SPDEF recombinant protein to the Foxj1 elements or to a previously reported consensus SPDEF binding site identified in the PSA gene by EMSA (24). Deletion and mutation of several potential SPDEF binding sites identified in the Foxj1 promoter did not block its activation by SPDEF (Supplemental Figure 6).
SPDEF interacts with the C-terminal domain of TTF-1. Because of the synergistic response to TTF-1 and SPDEF that we observed in several promoters, potential interactions between TTF-1 and SPDEF were assessed by mammalian 2-hybrid assays in HeLa cells and by glutathione-S-transferase (GST) pulldown assays in vitro. TTF-1 and SPDEF directly interacted in both assays (Figure 4). Interactions of TTF-1 with SPDEF were mediated by the C-terminal domain of TTF-1 and did not require the TTF-1 homeodomain, as assessed by both mammalian 2-hybrid assays and coimmunoprecipitation assays with GST-SPDEF (Figure 4). Thus, activation of target genes by SPDEF may be mediated, at least in part, by its interactions with TTF-1, known to bind and activate the promoters of a number of genes selectively expressed in the respiratory epithelium, including Sftpb, Sftpa, Sftpc, and Scgb1a1 (13–16).
SPDEF interacts with TTF-1 via the C-terminal domain of TTF-1. (A) Mammalian 2-hybrid assay was performed using the luciferase reporter pG5-luc and pACT and pBIND as described in Methods. Full-length TTF-1 and a series of TTF-1 deletion mutants (see Supplemental Table 1) were inserted to pBIND vector. Recombinant plasmids were cotransfected with the pG5-luc plasmids, and their activity was compared with that of cells transfected with pACT-SPDEF plus pBIND–TTF-1. Values are mean ± SD (n = 3). Assays were repeated 3 times with similar results. HD, homeodomain. (B) GST pulldown assays were performed with GST-SPDEF that was immobilized on glutathione-sepharose beads. Protein extracts were prepared from HeLa cells transiently transfected with the expression plasmids encoding for 3XFLAG–TTF-1, 3XFLAGΔ14, and 3XFLAGΔ3 as described in Methods. The extracts were incubated with GST or GST-SPDEF. Both GST and GST-SPDEF beads were washed several times before boiling, run on 10% SDS-polyacrylamide gels, and analyzed by immunoblot using a monoclonal antibody that recognizes the FLAG sequences.
Distinct effects of SPDEF and ETS variant gene 5. SPDEF enhanced expression of the promoters from the Sftpa, Foxj1, Sox17, and Scgb1a1 genes (Figure 3), but also activated the Sftpc gene, which is not normally expressed in conducting airway cells (data not shown). ETS variant gene 5 (ERM), a related member of the ETS family, is expressed in the distal epithelium during early lung development and is abundantly expressed in the peripheral lung saccules and in alveolar type II cells in the fetal and postnatal lung (28). ERM binds TTF-1 and activates Sftpc gene transcription in vitro (29). While SPDEF activated the Sftpa, Foxj1, Scgb1a1, and Sox17 promoters, ERM was inactive or less active (Figure 5). The distinct sites of expression of the ETS family members ERM and SPDEF and their selective influence on target genes may influence gene expression in a regionally specific manner.
Comparison of SPDEF and ERM on target gene transcription. Reporter assays were performed using plasmids expressing SPDEF and ERM, an ETS family transcription factor also expressed in the lung. The plasmids were cotransfected with the firefly luciferase reporter plasmids under control of Sftpa (A), Foxj1 (B), Scgb1a1 (C), and Sox17 (D) gene promoters. While SPDEF activated the Sftpa, Foxj1, Scgb1a1, and Sox17 promoters, ERM was less active. Each assay was repeated in at least 3 experiments with similar results. Values are mean ± SD (n = 3). P values shown were obtained by ANOVA.
SPDEF causes goblet cell hyperplasia in vivo. Because SPDEF was expressed in a subset of epithelial cells in the trachea, bronchi, and tracheal glands and stimulated transcriptional activity of genes normally expressed in proximal airway epithelial cells in vitro, its role was assessed in vivo. SPDEF was conditionally expressed under control of Clara cell secretory protein–reverse tetracycline transactivator (CCSP-rtTA; Figure 6A). Spdef transgene mRNA was not detected in lungs unless the mice were treated with doxycycline (Figure 6B and Figure 7). In situ hybridization and immunohistochemistry demonstrated the induction of Spdef mRNA and protein in subsets of cells in the respiratory epithelium lining conducting airways (Figure 6, D and H) and alveoli (data not shown), consistent with the activity of the CCSP (Scgb1a1) promoter in this mouse line, which is known to selectively direct expression to Clara cells in the conducting airways (30, 31).
Conditional expression of SPDEF in vivo. (A) Construct and strategy used to express SPDEF in Clara cells in the conducting airway. (B) RT-PCR using primers that selectively detect transgenic Spdef mRNA revealed that transgene-specific Spdef mRNA was increased in whole lung in the presence, but not the absence, of doxycycline (Dox). GAPDH expression was detected as an internal control. (C–H) In situ hybridization (C and D) and immunostaining (G and H) were performed to detect the expression of Spdef mRNA and protein in conducting airways and lung parenchyma in the absence (C, E, and G) and the presence (D, F, and H) of doxycycline. (E and F) Serial sections from C and D were stained with hematoxylin and eosin. Spdef mRNA and protein were detected at the sites of goblet cell morphology in the conducting airways of CCSP-rtTA/TRE2-Spdef mice treated with doxycycline (arrows), but were not detected in the bronchiolar epithelium of the transgenic mice without doxycycline. Scale bars: 200 μm (C–F); 50 μm (G and H).
Expression of SPDEF caused goblet cell hyperplasia in the conducting airways. CCSP-rtTA/TRE2-Spdef mice were maintained with (B, D, and F) or without (A, C, and E) doxycycline from E0 to postnatal day 14. Lung sections were stained with Alcian blue (A and B) or by immunohistochemistry for MUC5A/C (C and D) and CCSP (E and F). Increased Alcian blue and MUC5A/C staining was readily detected in the conducting airways of mice in the presence (B and D), but not in the absence (A and C), of doxycycline. Expression of SPDEF caused decreased CCSP staining (F) compared with controls without doxycycline (E). Scale bars: 100 μm.
SPDEF caused goblet cell differentiation in conducting airways of the transgenic mice (Figures 6–8). We observed similar results in 2 independent tetracycline response element 2–Spdef) (TRE2-Spdef) mouse lines, which were dependent upon doxycycline. Alcian blue staining (Figure 7, A and B) and immunostaining for MUC5A/C (Figure 7, C and D) were increased at the sites of SPDEF expression in the trachea and bronchi and in peripheral airways, including smaller bronchioles that normally lack goblet cells. Morphometric analysis demonstrated that SPDEF induced goblet cell hyperplasia in large (cartilagenous), medium-sized (central), and lateral (distal) airways, but not in terminal bronchioles or in the alveoli (Figure 8). There was a significant increase in the amount of Alcian blue staining found in the doxycycline-treated group compared with the untreated group (P = 0.004; Kruskal-Wallis 1-way ANOVA on ranks). Pairwise comparisons of each airway category revealed that the extent of Alcian blue staining in large, proximal, hilar airways increased significantly in the doxycycline-treated group compared with the untreated group (P = 0.026; Figure 8). The extent of Alcian blue staining along the airway correlated with staining for SPDEF in larger conducting airways, but was not significant in the more distal bronchioles, acinar ducts, or alveoli (Figure 8). Goblet cell hyperplasia occurred in the absence of inflammation, leukocytic infiltration, or altered expression of _TGF-_α, heparin-binding EGF (HB-EGF), IL-4, IL-6, and IL-13 mRNAs (Supplemental Figure 7). CCSP staining was decreased in regions lined by goblet cells (Figure 7, E and F), whereas the staining pattern for Foxj1, a ciliated cell marker, was not altered (Figure 9, A and B). Because the CCSP-rtTA driven Spdef transgene is expressed selectively in Clara cells, the paucity of CCSP staining and the presence of goblet cell hyperplasia that we observed in vivo are consistent with a cell-autonomous effect of SPDEF on the differentiation of Clara cells into goblet cells. Since loss of FOXA2 was previously shown to cause goblet cell differentiation (9), we assessed the effect of SPDEF on FOXA2 expression. FOXA2 staining was absent at sites of goblet cell hyperplasia induced by SPDEF (Figure 9, C and D). Phosphorylated histone H3 (pH3) staining was used to identify proliferating cells. The ectopic goblet cells did not stain for pH3, supporting the concept that expression of SPDEF in the airway epithelium influenced cell differentiation rather than proliferation (Supplemental Figure 8).
Morphometric analysis of goblet cell hyperplasia. (A) Alcian blue staining (μm2/mm) was increased significantly in CCSP-rtTA/TRE2-Spdef mice exposed to doxycycline compared with unexposed mice (P = 0.004; Kruskal-Wallis 1-way ANOVA on ranks). Staining was increased in all 4 categories of airways: cartilagenous, proximal (noncartilagenous), central, and distal. Pairwise comparisons of the data for the 4 different airway categories demonstrated that the increase in Alcian blue staining was most significant in the proximal airways of the doxycycline-exposed double-transgenic mice. *P = 0.026; Mann-Whitney rank-sum test. (B) Alcian blue staining was observed only in airways associated with cartilage, not in the epithelia of noncartilagenous airways, in double-transgenic mice without doxycycline. (C) Alcian blue staining was observed in epithelial cells lining these regions conducting airways in the presence of doxycycline. (D) Minimal SPDEF staining was observed in control mice at this antibody dilution. (E) Increased staining for SPDEF was observed in conducting airways of doxycycline-exposed double-transgenic mice. Scale bars: 500 μm.
FOXJ1 and loss of FOXA2 staining in lungs of CCSP-rtTA/TRE2-Spdef transgenic mice. Immunohistochemistry for FOXJ1 (A and B) and FOXA2 (C and D) was performed on the lung sections of control mice (A and C) and transgenic mice expressing SPDEF (B and D). The normal staining pattern of FOXJ1, a ciliated cell marker, was unaltered by expression of SPDEF (A and B). FOXA2 staining was observed throughout the epithelium in control mice (C). FOXA2 was not detected in the goblet cells lining the conducting airways of the transgenic mice, but persisted in ciliated cells (D). Scale bars: 50 μm.
Induction of SPDEF expression in mouse models of goblet cell hyperplasia. Increased expression of either IL-4 or IL-13 (32–35) and allergen challenge (36) cause goblet cell hyperplasia in vivo. We tested whether increased SPDEF expression is associated with goblet cell hyperplasia in mice expressing IL-13 in Clara cells under conditional control of doxycycline. Increased SPDEF staining and mRNA were associated with goblet cell hyperplasia in conducting airways of adult mice as assessed by RT-PCR, in situ hybridization, and immunostaining (Figure 10). SPDEF staining was observed in both the cytoplasm and the nuclei of epithelial cells in conducting airways. Likewise, intratracheal IL-13 treatment caused goblet cell hyperplasia in association with increased SPDEF staining in control but not in _Stat-6_–/– mice (Figure 11, A and B). Goblet cell hyperplasia and increased SPDEF staining were observed following repeated intratracheal administration of dust mite allergen to wild-type mice but was not observed in treated _IL-13_–/– mice (Figure 11, C and D). Thus, IL-13 and allergen exposure increased Spdef mRNA and staining, extending expression in smaller conducting airways.
IL-13 induces expression of SPDEF. At 5 weeks of age, CCSP-rtTA/tetO–CMV–IL-13 mice were maintained with or without doxycycline for 1 week. (A) RT-PCR for SPDEF was performed using total RNA from the transgenic mice. Spdef mRNA was increased in the transgenic mice treated with doxycycline, while GAPDH was unchanged. (B and C) In situ hybridization (B) and SPDEF immunostaining (C) were performed on lung sections from the transgenic mice. Spdef mRNA was induced in the conducting airways of the transgenic mice treated with doxycycline but was not detected in the untreated mice (B). SPDEF staining was detected in the conducting airways of the transgenic mice treated with doxycycline (C), consistent with the sites of Spdef mRNA expression. SPDEF was not detected in the absence of doxycycline under these conditions. Insets represent higher magnification views of these airways; original magnification, ×20. Scale bars: 200 μm.
IL-13 and dust mite allergen induce SPDEF and cause goblet cell hyperplasia. (A and B) Immunohistochemistry for SPDEF was performed on lung sections of control (A) and Stat-6–/– (B) mice that were treated intratracheally with IL-13. SPDEF staining was increased at sites of goblet cell hyperplasia, but absent in conducting airways of Stat-6–/– mice. (C and D) SPDEF was increased in association with goblet cell hyperplasia caused by intratracheal exposure to house dust mite allergens in wild-type mice (C), but not in the conducting airways of exposed IL-13–/– mice (D). Insets for A–D represent higher magnification views of these airways; original magnification, ×20. Scale bars: 200 μm.