Thyroid Transcription Factor 1 Interactions

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 24, pp. 16716 –16726, June 16, 2006 © 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

Erm/Thyroid Transcription Factor 1 Interactions Modulate Surfactant Protein C Transcription* Received for publication, March 9, 2006 Published, JBC Papers in Press, April 13, 2006, DOI 10.1074/jbc.M602221200

Sui Lin, Anne-Karina T. Perl, and John M. Shannon1 From the Division of Pulmonary Biology, Cincinnati Children’s Hospital Medical Center and the University of Cincinnati College of Medicine, Cincinnati, Ohio 45229-3039

Mouse lung development begins on embryonic day 9.5 (E9.5)2 as two evaginations from the floor of the foregut endoderm into mesenchyme derived from splanchnic mesoderm. After a series of dichotomous and lateral branchings that form the pulmonary tree, the most distal epithelium continues to expand and mature during late gestation and early postnatal life to give rise to mature alveoli. Two cell types constitute the alveolar epithelium, alveolar type I and type II cells. Type I cells have a highly attenuated morphology that is ideally suited for gas exchange (1),

* This work was supported by a Franklin Delano Roosevelt Fellowship from the March of Dimes and by NHLBI, National Institutes of Health Grants HL-56387 and HL-071898. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondence should be addressed: Division of Pulmonary Biology, Children’s Hospital Medical Center, 3333 Burnet Ave., Cincinnati, OH 45229-3039. Tel.: 513-636-2938; Fax: 513-636-7868; E-mail: [email protected]. 2 The abbreviations used are: E, embryonic day; FGF, fibroblast growth factor; FGFR, FGF receptor; SP-A, SP-B, SP-C, surfactant protein A, B, C; TTF-1, thyroid transcription factor 1; Erm, Ets-related molecule; EMSA, electrophoretic mobility shift assay; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Spry-4, Sprouty 4; siRNA, small interfering RNA; kb, kilobase.

16716 JOURNAL OF BIOLOGICAL CHEMISTRY

whereas the more cuboidal type II cells function in a number of capacities. These include serving as progenitor cells for type I cells, directionally transporting sodium from apical to basolateral cell surfaces to minimize alveolar fluid, producing molecules involved in innate host defense, and synthesizing and secreting pulmonary surfactant (2). Proximate tissue interactions between the epithelium and mesenchyme play a critical role in lung morphogenesis and differentiation. Tissue recombination studies have shown that distal lung epithelial differentiation is induced as a specific response to signals produced by lung mesenchyme (3, 4). Furthermore, tissue recombination experiments in which epithelial and mesenchymal rudiments are physically separated by an interposed filter have shown that these inductive cues are diffusible (5, 6). Studies using mesenchyme-free culture of embryonic lung (7–9) or tracheal (10, 11) endoderm have demonstrated a primary role for members of the FGF family, particularly FGF7, in the induction and maintenance of type II cell differentiation, including the expression of surfactant protein C (SP-C; mouse genomic designation Sftpc). SP-C, which is found only in the lung, is expressed in the distal epithelium during early lung development (12), then is restricted to alveolar type II cells late in gestation and postnatally. Restriction of SP-C expression to the lung epithelium is conferred by regulatory elements located within ⫺215 bp of the human SP-C promoter (13). This region, which is highly homologous between humans and mice, contains conserved binding sites for thyroid transcription factor 1, a phosphorylated, homeodomain-containing member of the Nkx nuclear transcription factor family (TTF-1; also called Nkx2.1 and T-EBP; mouse genomic designation Titf1) (14). TTF-1 is expressed in the epithelium of the developing lung, thyroid, and in the central nervous system. TTF-1 is critical for normal lung formation, since Titf1 null mutant mice die at birth from a lack of peripheral lung tissues (15). The phenotype of mice in which the serine phosphorylation sites of TTF-1 are mutated (Titf1PM/PM) is also neonatal lethal, and the expression of known TTF-1 target proteins is decreased (16). The effects on lung morphogenesis are not as severe in the phosphorylation mutant, since lobulation and early branching morphogenesis are maintained, but formation of peripheral structures late in development is still compromised. Experiments done in our laboratory using tissue recombinants (3) and mesenchyme-free culture (10) have shown that the entire embryonic respiratory tract epithelium exhibits a substantial plasticity in its eventual phenotype. These observations led us to speculate that the differences in embryonic trachea and lung are defined by a limited subset of genes and that a molecular comparison of these two tissues might provide new information on genes that are important for both lung and trachea development. We, therefore, used microarray analysis to identify genes differentially regulated in the embryonic lung and trachea (17). One of the differentially expressed genes we identified was Erm (Ets-related molecule), a member of the Ets transcription factor family. In the present study we demonstrate that Erm is increasingly restricted to the distal epithelium as lung development progresses and is

VOLUME 281 • NUMBER 24 • JUNE 16, 2006

Downloaded from www.jbc.org at Childrens Hospital Medical Center on December 4, 2006

Expression of surfactant protein C (SP-C), which is restricted to alveolar type II epithelial cells of the adult lung, is critically dependent on thyroid transcription factor 1 (TTF-1). In the present study we have demonstrated that Erm, a member of the Ets family of transcription factors, is expressed in the distal lung epithelium during development and is also restricted to alveolar type II cells in the adult. Erm was up-regulated by fibroblast growth factors (FGFs) in culture, and blocking FGF signaling inhibited Erm expression both in vivo and in vitro. The SP-C minimal promoter was found to contain two potential Ets binding sites, and electrophoretic mobility shift assays showed that two 20-bp wild-type oligonucleotides containing the 5ⴕ-GGA(A/T)-3ⴕ Ets consensus binding motif were shifted by nuclear extracts from MLE15 cells. Co-transfection assays showed that Erm by itself had little effect on SP-C promoter activity but that Erm significantly enhanced TTF-1-mediated SP-C transcription. Mutation of one of the Ets binding sites reduced SP-C transcription to background levels, whereas mutation of the other site resulted in increased SP-C transcription. Protein-protein interactions between Erm and TTF-1 were demonstrated by mammalian two-hybrid assays and by co-immunoprecipitation assays. Mapping studies showed that the Ets domain of Erm and the combined N terminus and homeodomain of TTF-1 were critical for this interaction. Treatment of primary cultures of adult alveolar type II cells with siRNA targeting Erm diminished expression of both Erm and SP-C but had no effect on ␤-actin or GAPDH (glyceraldehyde-3phosphate dehydrogenase). Taken together, these results demonstrate that Erm is involved in SP-C regulation, which results from an interaction with TTF-1.

Erm Is Involved in SP-C Transcriptional Regulation found only in type II cells in the adult lung. Embryonic tracheal epithelium does not contain Erm or SP-C but can be induced by lung mesenchyme to express both genes in the same spatial domain. Erm, like SP-C, is regulated by FGF signaling both in vivo and in vitro. We show that, although Erm by itself does not affect SP-C expression, it interacts with TTF-1 to increase SP-C transcription. Finally, we demonstrate that siRNA-mediated suppression of Erm in primary cultures of type II cells concomitantly decreases expression of SP-C.

EXPERIMENTAL PROCEDURES

JUNE 16, 2006 • VOLUME 281 • NUMBER 24

JOURNAL OF BIOLOGICAL CHEMISTRY

16717


Animals—All protocols involving the use of mice were reviewed and approved by the Institutional Animal Care and Use Committee at Cincinnati Children’s Medical Center. Lung and tracheal tissues that were used for the isolation of Erm cDNA, the ontogenetic expression of Erm, tissue recombination studies, and mesenchyme-free culture were obtained from FVB/N mice. Adult type II cells were obtained from C57Bl/6 mice. Titf1⫺/⫺ mice were generated on a C57Bl/6 background (15). The generation of transgenic FVB/N mice that conditionally express a soluble dominant-negative FGF receptor (FGFR-Hc) in the lung has been described elsewhere (18). Generation of transgenic mice conditionally expressing the FGF antagonist Sprouty-4 (Spry-4) in the developing lung epithelium has been described (19). Timed-pregnant females of both transgenic strains were given doxycycline in their food (625 mg/kg; Harlan-Teklad, Madison, WI) beginning on day E0.5, and embryonic lungs were harvested on day E13.5. Generation of Erm Construct—All procedures involving oligonucleotide and cDNA manipulations were performed essentially as described in Sambrook et al. (20). We generated the Erm construct by reverse transcription-PCR using total lung RNA from E11.5 embryos. A cDNA fragment comprising the Erm coding sequence to which a FLAG tag was added to the 3⬘ end was generated using the primers 5⬘-primer (5⬘-GCGGATCCACCATGGATGGGTTTTGT3⬘) specific to Erm and adding a BamHI restriction site (underlined) to the 5⬘ end and 3⬘-primer (5⬘-GCTCTAGATTActtgtcatcgtcgtccttgtagtcGTAAGCGAAGCCTTC-3⬘) specific to the last 15 bp of Erm coding sequence (uppercase letters) and 24 bp of the FLAG epitope tag (lowercase letters) followed by a stop codon and adding a XbaI site (underlined) to the 3⬘ end. The PCR product was subsequently digested with BamHI and XbaI restriction enzymes (Stratagene, La Jolla, CA) and subcloned into pcDNA3 (Invitrogen) that had been digested with BamHI and XbaI. The resulting expression construct, hereafter pErmFLAG, was sequenced in both directions, and no mutations were detected. Plasmids, Cell Culture, Transfection, and Reporter Gene Assays—The reporter plasmid pGL3/SPCluc (a gift of Dr. C. Bachurski, Cincinnati Children’s Hospital) was created by cloning the 0.32-kb minimal mouse SP-C promoter sequence (14) into the firefly luciferase reporter vector pGL3 (Promega, Madison, WI). The two putative Ets binding sites in this promoter were mutated by PCR to XhoI (CTCGAG) sites to generate the single mutations M2 (⫺200 to ⫺195) and M1 (⫺125 to 120) as well as a M1M2 double mutation (Fig. 5A). The three mutated fragments were subcloned into pGL3 to generate SPC/M1luc, SPC/M2luc, and SPC/M1M2luc, and their sequences were confirmed. Genomatrix software was used to verify that no known enhancer/repressor sites were generated. The TTF-1 expression plasmid pCMV-TTF-1 was a gift of Dr. R. DiLauro (Stazione Zoologica, Naples, Italy). The pRL-TK Renilla luciferase plasmid (Promega) was used to monitor and normalize transfection efficiency. Human JEG-3 choriocarcinoma cells were grown in Eagle’s minimal essential medium (Invitrogen) supplemented with 10% fetal bovine

serum (Sigma), 2 mM L-glutamine, 100 units/ml penicillin, and 100 ␮g/ml streptomycin. MLE-15 cells were grown in HITES medium (21). 2 ⫻ 105 cells were seeded into 6-well dishes and grown to 50 –70% confluence. The amount of plasmid DNA transfected into the cells was kept constant at 800 ng/well; any differences among groups were compensated using empty pcDNA3 vector. All plasmids were transiently transfected into JEG-3 or MLE-15 cells using Effectene transfection reagent (Qiagen, Valencia, CA). Expression constructs were transfected at the following concentrations: pGL3/SPCluc, 270 ng/well; pErmFLAG, 170 ng/well; pCMV-TTF-1, 170 ng/well; pRL-TK, 20 ng/well. Constructs for the Ets site mutation analysis were transfected into JEG-3 cells at the following concentrations: SPCluc, SPC/M1luc, SPC/M2luc, and SPC/ M1M2luc, 380 ng/well; pErmFLAG, 200 ng/well; pCMV-TTF-1, 200 ng/well; pRL-TK, 20 ng/well. Cells were lysed 48 h after transfection, and luciferase activities were assessed using the dual-luciferase reporter assay system (Promega), with light units quantified by luminometry (Monolight 3010, Analytical Luminescence Laboratory, San Diego, CA). Tissue Recombinations, Mesenchyme-free Epithelial Culture, and Explant Culture—Tissue recombinants composed of lung mesenchyme plus either lung epithelium or tracheal epithelium were prepared as previously described (3). Tracheal epithelial rudiments for mesenchyme-free culture were isolated from E11.5 embryos and enrobed in growth factor-reduced Matrigel (BD Biomedical) as previously described (10). Rudiments were cultured in Dulbecco’s modified Eagle’s medium/F-12 containing 3% charcoal-stripped fetal bovine serum, 10 ␮g/ml insulin, 5 ␮g/ml human transferrin (both from BD Biomedical), 1 ␮g/ml cholera toxin (ICN Pharmaceuticals, Irvine, CA), 25 ng/ml mouse recombinant epidermal growth factor, 10 ng/ml human recombinant hepatocyte growth factor, 200 ng/ml bovine FGF1 (all from R&D Systems, Minneapolis, MN), and 25 ng/ml human recombinant FGF7 (PreproTech Inc., Rocky Hill, NJ). Lungs from E12.5 embryos were cultured for 24 h on 8-␮m pore nucleopore track-etch membranes (Whatman International, Clifton, NJ) in Dulbecco’s modified Eagle’s medium/F-12 plus 5% fetal bovine serum. Explants were treated with the FGF receptor antagonist SU5402 (EMD Biosciences, La Jolla, CA) at 10 ␮M; controls were treated with 0.1% Me2SO. Real-time PCR—Gene expression was assessed by real-time PCR using a Smart Cycler (Cepheid, Sunnyvale, CA) and DNA master SYBR Green I (Roche Applied Science). Template cDNA was obtained by reverse transcription of total RNA as described above, and an equal amount of input cDNA was used in each set of PCR reactions. Template cDNA was mixed with 1⫻ SYBR mix, 2.5 mM MgCl2, and 0.25 ␮M concentrations of each primer. Reaction conditions were 95 °C for 15 s followed by 40 cycles of amplification step (95 °C for 10 s, 60 °C for 15 s, and 72 °C for 20 s). ␤-actin and L32 were used as normalization standards. Changes in gene expression are given as percent changes versus controls. For the gels shown in Figs. 2– 4, parallel real-time PCR reactions were performed, with the modification that all reactions were terminated when the sample(s) with the highest level of expression reached the plateau. The sequences of the primers used were: Erm (151 bp), forward primer 5⬘-GCAGGGGGAAGTAGTTCAAATCC-3⬘, reverse primer 5⬘-TGTGCCAGTCAAGGGACACG-3⬘; Pea3 (150 bp), forward primer 5⬘GCCCCGCCACCTTAGTTGTG-3⬘, reverse primer 5⬘-GCACTCGCTGGTCCAAGTATCC-3⬘; TTF-1 (187 bp), forward primer 5⬘-GTTCCAGAACCACCGCTACA-3⬘, reverse primer 5⬘-GGGTTTGCCGTCTTTGACTA-3⬘, ␤-actin (350 bp), forward primer 5⬘-TGGAATCCTGTGGCATCCATGAAAC-3⬘, reverse primer 5⬘-TAAAACGCAGCTCAGTAACAGTCCG-3⬘, L32 (257 bp), forward primer 5⬘-GTG-

Erm Is Involved in SP-C Transcriptional Regulation

AAGCCCAAGATCGTC-3⬘, reverse primer 5⬘-AGCAATCTCAGCACAGTAAG-3⬘. In Situ Hybridization—The 2.2-kb mouse cDNA for the 3⬘-untranslated region of Erm contained in IMAGE clone IM2645600 (Invitrogen) and the 758-bp cDNA for mouse SP-C (12) were used as templates for generating riboprobes. Digoxigenin-labeled RNA probes were used for whole mount in situ hybridization (22), and riboprobes labeled with [33P]UTP were used for tissue section in situ hybridization (7). Electrophoretic Mobility Shift Assay (EMSA)—Double-stranded wildtype oligonucleotide 1 (⫺132 to ⫺113) and oligonucleotide 2 (⫺207 to ⫺188) (Fig. 5A), each of which contains one of the two consensus Ets binding sites present in the 320-bp SP-C promoter, were end-labeled with ␥-[32P]ATP and T4 polynucleotide kinase. EMSA were performed as described elsewhere (23). For antibody supershift assays, 2 ␮l of goat anti-Erm antibody (Santa Cruz Biotechnology, Santa Cruz, CA) or control IgG were added, and the incubation was continued for an additional 20 min. The DNA-protein complex was separated from free probes by nondenaturing electrophoresis on a 5% polyacrylamide gel. Erm and TTF-1 Co-precipitation Assays—Nuclear extracts were prepared from JEG-3 cells 48 h after co-transfection with the expression plasmids pErmFLAG and/or pCMV-TTF-1; in one set of controls pErmFLAG was omitted from the transfection. The nuclear extracts were incubated with 10 ␮l of a rabbit polyclonal Erm antibody (Santa Cruz) or normal rabbit serum (as a control) and protein G-Sepharose beads in immunoprecipitation buffer (20 mM Tris (pH 7.6), 150 mM NaCl, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 10 ␮g/ml pepstatin, 10 ␮g/ml leupeptin, 30 ␮g/␮l Bacitracin, 1 mM o-phenanthroline) overnight at 4 °C. After extensive washing with immunopre-


cipitation buffer minus Triton X-100, the beads were boiled in Laemmli buffer (62.5 mM Tris (pH 6.8), 2% SDS, 25% glycerol, 0.01% bromphenol blue). The precipitated proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. Western blots were incubated with a 1:5000 dilution of a monoclonal TTF-1 antibody (24). Mammalian Two-hybrid System Assays—The plasmid constructs pVP16/Erm-AD and pM/TTF-1-BD (hereafter AD/Erm and BD/TTF-1) were made by subcloning the coding sequences of Erm and TTF-1 into the vectors pVP16 and pM (Clontech, Palo Alto, CA), respectively. The reporter construct pG5Luc, which contains the luciferase gene downstream of five GAL4 DNA binding sites and the adenovirus E1b minimal promoter (25), was provided by Dr. C. Yan (Cincinnati Children’s Hospital). Erm deletion constructs, which included the N-terminal (Em-N), acidic (Em-AD), Ets (Em-Ets), and C-terminal (Em-C) domains as well as the Ets plus C-terminal (Em-⌬N), were generated by PCR and subcloned into pVP16 vector between EcoRI and XbaI sites. All constructs were sequenced, and no mutations were detected. TTF-1 deletions were excised from previously generated TTF-1 domain constructs in pVP16 (26) and subcloned into pMBD vector between EcoRI and XbaI sites. JEG-3 cells were transfected with 300 ng/well of the paired AD/Erm and BD/TTF-1 chimeric plasmids along with 180 ng/well of pG5Luc and 20 ng/well of pRL-TK. Cells were lysed 48 h after transfection, and luciferase activity was assayed by luminometry as described above. Adult Type II Cell Culture and Transfection with siRNAs—Adult mouse type II cells in Dulbecco’s modified Eagle’s medium plus 10% fetal bovine serum were prepared as previously described (27) and seeded onto Matrigel in 12-well dishes. Non-adherent cells were removed after 24 h, and the cells were switched to BEGM medium



FIGURE 1. Erm expression is associated with distal lung epithelial differentiation. In situ hybridization was performed on embryonic lung/trachea complexes over the course of gestation as well as in the adult lung. On day E10.5 (A ) Erm was detected in both lung (arrowheads) and tracheal epithelium (arrow). By E11.5 (B) expression was restricted to the lung epithelium, which continued through day E13.5 (C ). From day E16.5 (D) onward Erm expression was increasingly restricted to distal acini (arrowheads); note the lack of Erm signal in the proximal epithelium (asterisks). Proximal bronchiolar (br) areas in the adult lung did not express Erm, which was seen only in alveolar type II cells (E and F ).


(Clonetics, Walkersville, MD) minus hydrocortisone but additionally containing 5% rat serum and 12.5 ng/ml FGF7 (PreproTech). Cells were then transfected with 400 ng of siRNA targeting bases 194 –214 (5⬘AAGUUCCUGAUGAUGAGCAGU-3⬘) of Erm using Gene Silencer reagent (Gene Therapy Systems, San Diego, CA). Sham-transfected cells and cells transfected with an equal amount of a Cy3-labeled siRNA targeting luciferase (Cy3luc; Dharmacon Research, Lafayette, CO) served as controls. Cells were harvested 72 h post-transfection, and expression of Erm, SP-C, ␤-actin, and GAPDH was determined by realtime PCR. Statistical Analysis—Data were analyzed and graphed using GraphPad Prism version 4.0 (GraphPad Software, San Diego, CA). Comparisons among groups were made by one-way analysis of variance using the Tukey-



FIGURE 2. Erm is induced by embryonic distal lung mesenchyme. When purified E11.5 tracheal epithelium, which is negative for Erm expression (Fig. 1B), was recombined with lung mesenchyme, it was induced to branch in a lung-like pattern (A and B). Branching of the reprogrammed epithelium was accompanied by induction of Erm (A), which like SP-C (B) was expressed in the distal epithelium (arrowheads). E18.5 fetuses with a targeted deletion of the Titf1 gene had poorly branched, hypoplastic lungs (lu; C) that did not contain detectable Erm by in situ hybridization (D). Normal littermates showed Erm expression in the distal epithelium, whereas proximal epithelial tubules (asterisks) were negative (E and F). Equivalent amounts of input cDNA from normal and Titf1⫺/⫺ fetuses were analyzed in parallel realtime PCR reactions. In one set all reactions were terminated when the highest expressing sample(s) reached a plateau and run on a gel to give the qualitative image shown here (G) that accurately reflects the real-time PCR data. Normal E18.5 lungs contained both Erm and SP-C. SP-C was not detected in Titf1⫺/⫺ lungs, but a faint signal for Erm was seen. L32 was used as a positive control for normal and Titf1⫺/⫺ lungs. Control reactions that contained no template (H2O) gave no signal.

Kramer multiple comparisons test. Data are presented as the means ⫾ S.E., with p ⬍ 0.05 considered significant.

RESULTS Erm Is Expressed in Lung Epithelium—Using cDNA microarray analysis, we identified Erm, a member of the Ets transcription factor family, as differentially expressed in the E11.5 lung versus the trachea (17). In situ hybridization showed that Erm was expressed in both lung and tracheal epithelium on day E10.5 (Fig. 1A) but was restricted to the lung epithelium by E11.5 (Fig. 1B), an observation that was confirmed by reverse transcription-PCR (data not shown). Erm became progressively more restricted to the distal lung epithelium on days E13.5 (Fig. 1C) and E16.5 (Fig. 1D). In the adult lung we observed that Erm was expressed


16719



FIGURE 3. Erm expression is enhanced by FGFs in vitro. E11.5 tracheal epithelium cultured in Dulbecco’s modified Eagle’s medium/F-12 medium containing insulin, transferrin, cholera toxin, epidermal growth factor, hepatocyte growth factor, FGF1, and FGF7 showed widespread hybridization with an antisense Erm probe (A), whereas a sense probe gave no signal (B). Induction of SP-C also occurred in these cultures (C), and the pattern of expression was identical to that seen for Erm. Equivalent amounts of input cDNA from E11.5 tracheal epithelial rudiments cultured in the presence or absence of FGF1 and FGF7 were analyzed in parallel real-time PCR reactions. In one set, all reactions were terminated when the highest expressing sample(s) reached a plateau and run on a gel to give the qualitative image shown here (D) that accurately reflects the real-time PCR data. Elimination of FGFs from the medium (⫺FGFs lane) ablated expression of SP-C and reduced Erm expression by 67%. ␤-actin, Pea3, and TTF-1 expression were unaffected by the absence of FGFs. Controls containing no template cDNA (H2O) gave no signal.

pared with control littermates, the lungs from mice expressing FGFR-HFc or Spry-4 had reduced expression of Erm (⬃65% decrease) and SP-C (⬃80% decrease) (n ⫽ 1; Fig. 4G). Erm Binds to the SP-C Promoter and Enhances TTF-1 Activation of SP-C in MLE-15 and JEG-3 Cells—The fact that the expression of Erm and SP-C was spatially and temporally restricted in an identical pattern suggested the possibility that Erm may be involved in the regulation of SP-C transcription itself. We, therefore, examined the 0.32-kb minimal mouse SP-C promoter, which contains several cis-acting elements essential for its transcription, for Ets binding sites. As shown in Fig. 5A, two consensus Ets binding motifs (5⬘-GGA(A/T)-3⬘) were identified at positions ⫺124 to ⫺121 and ⫺199 to ⫺196, the latter of which was located inside the 70-bp TTF-1-responsive fragment that includes the TTF-1 consensus elements T2 to T6 (14). Two oligonucleotide probes based on these Ets sites bound to protein in MLE-15 nuclear extracts in EMSA (Fig. 5B). Competition with a 100-fold excess of either unlabeled probe eliminated binding, whereas competition with an unlabeled nonspecific oligonucleotide that included the proximal TTF-1 binding site of the SP-A promoter (30) had no effect on binding. The protein-DNA complexes formed by MLE-15 nuclear extracts with the Erm probes were disrupted by incubation with anti-Erm antibody. Transfection of MLE-15 cells, which contain both Erm and TTF-1 protein, with a pErmFLAG and pGL3/SPCluc had no significant effect on SP-C transcription (Fig. 5C). In agreement with previous results (14, 31), transfection with pCMV-TTF-1 resulted in significant ( p ⬍ 0.001) stimulation of SP-C transcription. However, transfection with both Erm and TTF-1 caused a significant ( p ⬍ 0.001) further enhancement in SP-C transactivation compared with cells transfected with TTF-1 alone, suggesting that Erm interacted with TTF-1 to regulate SP-C transcription. To further assess the role of Erm in SP-C transcription, we mutated the putative Ets binding sites of pGL3/SPCluc singly and in combination (Fig. 5D). These experiments were done with JEG-3 cells, which do not contain endogenous Erm or TTF-1. Mutating the Ets site between bp



exclusively in alveolar type II cells (Fig. 1, E and F ) but not in more proximal epithelial cell types, an expression pattern identical to that seen for SP-C (12, 28). The co-localization of SP-C and Erm in the adult lung was underscored by the fact that expression of both genes was similarly enriched in freshly isolated alveolar type II cells when compared with whole lung by real-time PCR (6.1- and 7.7-fold increases for Erm and SP-C, respectively; n ⫽ 4). We did not detect Pea3 in adult type II cells by reverse transcription-PCR (data not shown). Erm Expression Is Associated with the Distal Lung Epithelial Cell Phenotype—Like Erm, SP-C is not expressed in the tracheal epithelium at stage E11.5 or later. We have previously shown, however, that embryonic tracheal epithelium can be reprogrammed by lung mesenchyme to express morphological, biochemical, and molecular markers of differentiated distal lung epithelium, including SP-C (3). Because the ontogeny data suggested an association of Erm with distal lung epithelial differentiation, we examined tissue recombinations of E11.5 tracheal epithelium with same-stage lung mesenchyme for Erm expression. We found that tracheal epithelium induced to branch by lung mesenchyme expressed Erm in the most distal epithelial cells (Fig. 2A) and that the spatial expression of Erm coincided with that of SP-C (Fig. 2B), suggesting that Erm expression was associated with distal lung epithelial phenotype. We, therefore, examined Erm expression in the lungs of Titf1(⫺/⫺) fetuses, which lack markers of distal epithelial differentiation, including SP-C (29). The severely hypoplastic lungs from E18.5 Titf1⫺/⫺ fetuses showed no detectable signal for Erm by in situ hybridization (Fig. 2, C and D), whereas wild type and heterozygous littermates showed strong expression in the distal epithelium (Fig. 2, E and F ). Although real-time PCR confirmed that Titf1⫺/⫺ lungs contained no SP-C, we did detect Erm but at a much lower level (reduced by 90%) than controls (n ⫽ 2; Fig. 2G). Erm Expression Is Regulated by FGFs—We have previously shown that the reprogramming of tracheal epithelium to a lung phenotype by lung mesenchyme can be mimicked in mesenchyme-free culture by a complex growth medium and that members of the FGF family are critical for the induction of SP-C (10, 11). Day E11.5 tracheal epithelium cultured in the presence of FGF1 and FGF7 showed extensive growth and budding, and Erm (Fig. 3A) and SP-C (Fig. 3C) expression were detectable by both in situ hybridization and real-time PCR (Fig. 3D). Whereas deletion of FGF1 and FGF7 from the medium resulted in the complete elimination of SP-C (Fig. 3D), Erm was still expressed but was significantly decreased by 67 ⫾ 2% ( p ⬍ 0.02, n ⫽ 3; Fig. 3D) when normalized to ␤-actin. Cultured tracheal epithelium also contained Pea3, but expression levels were unaffected by the presence or absence of FGFs (Fig. 3D). TTF-1 expression was not influenced by the presence or absence of FGFs (Fig. 3D). To assess how FGFs affect Erm expression in a more physiological context, we tested the effects of blocking FGF signaling in intact lungs both in vitro and in vivo. E12.5 lungs (Fig. 4A) cultured for 1 day continued to branch at the distal tips and to form lateral branch points (Fig. 4B). In contrast, lungs cultured with the FGFR inhibitor SU5402 ceased formation of clefts in distal tips and lateral budding (Fig. 4C), although individual buds continued to elongate (arrow, Fig. 4C). When normalized to ␤-actin levels, real-time PCR analysis showed that SU5402 treatment decreased the expression of Erm by 83 ⫾ 3% ( p ⬍ 0.002, n ⫽ 3) and SP-C by 92 ⫾ 3.1% ( p ⬍ 0.006, n ⫽ 3; Fig. 4G). Expression of TTF-1 was unaffected by SU5402 (data not shown). Disrupting FGF signaling in the lung in vivo either by expressing a soluble dominant-negative FGFR (FGFR-HFc) (18) or by expressing the FGF antagonist Spry-4 (19) resulted in severe lung hypoplasia (Fig. 4, D–F). Real-time PCR analysis showed that, com-


⫺200 to ⫺195 (SPC/M2luc) not only abolished the synergism of Erm with TTF-1 but also decreased TTF-1-induced SP-C transcription to background levels. Mutating the Ets site between bp ⫺125 to ⫺120 (SPC/M1luc) did not affect the significant ( p ⬍ 0.001) increase in SP-C transcription in response to TTF-1. On the contrary, this mutation resulted in a response to TTF-1 that was significantly greater ( p ⬍ 0.001) than that seen with the wild-type SP-C promoter. The addition of both Erm and TTF-1 resulted in a significant ( p ⬍ 0.001) increase in transactivation above that seen with TTF-1 alone; again, the response of the SPC/M1luc mutant was greater than that seen with wild-type pGL3/ SPCluc. Mutation of both Ets sites (SPC/M1M2luc) resulted in no significant change in SP-C transcription in response to Erm, TTF-1, or Erm plus TTF-1. Erm and TTF-I Interact to Regulate SP-C Transcription—Erm, like other members of the Ets family, influences transcriptional activation by complexing with other nuclear proteins through proteinprotein interaction (32, 33). We examined whether Erm enhances the TTF-1-mediated SP-C transcription as a result of protein-protein interaction in two ways. First, a mammalian two-hybrid assay was used to assess the physical relationship between Erm and TTF-1. Erm was subcloned into vector pVP16 (AD/Erm), whereas TTF-1 was subcloned into vector pM (BD/TTF-1) (Fig. 6A). Transfection of AD/Erm and BD/TTF-1 together increased luciferase activity 23-fold ( p ⬍ 0.001) over either construct by itself (Fig. 6B). We observed similar results ( p ⬍ 0.05) when we transfected HeLa cells (Fig. 6C), which contain endogenous Erm. We performed co-precip-


itation assays as a second means of demonstrating the interaction of Erm and TTF-1. Consistent with these results, TTF-1 was co-immunoprecipitated when nuclear extracts from JEG-3 cells transfected with Erm and TTF-1 expression plasmids were incubated with antiErm antibody (Fig. 6D). Importantly, TTF-1 was not detected in the absence of transfected Erm or when normal rabbit serum was substituted for anti-Erm antibody. To map the Erm domains required for its interaction with TTF-1, we generated constructs in pVP16 that encoded the N-terminal (Em-N), acidic (Em-AD), Ets (Em-Ets), and C-terminal (Em-C) domains as well as the combined Ets plus C-terminal domains (Em-⌬N). We assessed the interactions between the mutant fusion proteins and TTF-1 by mammalian two-hybrid assay using JEG-3 cells. We found that both the Em-Ets and Em-⌬N constructs supported interaction with TTF-1, albeit at a reduced level, whereas the other deletion constructs did not (Fig. 6E). The fact that the interaction of Em-Ets or Em-⌬N with TTF-1 resulted in reduced activity compared with full-length Erm suggests that both the N- and C-terminal domains may be required for the correct conformation of the Ets domain. We next used deletion constructs of TTF-1 to determine which domains were required for the interaction with Erm in JEG-3 cells (Fig. 6F ). The construct with a deletion of the C-terminal domain (T1-⌬C) functioned as well as intact TTF-1. However, neither the N-terminal domain (T1-N) nor the homeodomain (T1-HD) alone were capable of interacting with Erm, indicating that the combination of these two domains is critical for the interaction of TTF-1 with Erm. The C-termi-


16721


FIGURE 4. Inhibition of FGF signaling decreases Erm expression in vitro and in vivo. E12.5 lungs (A) cultured in medium containing Me2SO (B) continued to branch; note the increase in buds in the right middle lobe (arrow, B). Whereas treatment with the FGFR antagonist SU5402 completely blocked further branching (C ), individual lobes appeared to continue elongating (arrow, C ). Lungs from E13.5 wild-type fetuses (D) showed extensive branching, with many distal tips; only the left lung (LL) is shown here. In contrast, lungs from mice expressing FGFR-HFc (E ), a soluble dominantnegative FGFR, were severely hypoplastic, and the left lung (LL) had few distal tips. Similarly, mice expressing the FGF antagonist Spry-4 (F ) in the epithelium were hypoplastic and showed little branching. Note that the lack of lateral branching in the epithelium in lungs from mice expressing FGFR-HFc or Spry-4 results in an elongated appearance, similar to that seen in SU5402-treated explants (C). Equivalent amounts of input cDNA from the lungs of normal and transgenic fetuses and from lung explant cultures were analyzed in parallel real-time PCR reactions. In one set, all reactions were terminated when the highest expressing sample(s) reached a plateau and run on a gel to give the qualitative image shown here (G) that accurately reflects the real-time PCR data. Lungs from wild-type E13.5 fetuses expressed both Erm and SP-C. The lungs of fetuses expressing FGFR-HFc or Spry-4 expressed both Erm and SP-C but at lower levels (⬃65 and ⬃80% decreases, respectively). E12.5 lungs cultured for 24 h in medium containing 0.1% Me2SO expressed Erm and SP-C at levels similar to those seen in E13.5 lungs. Treatment with the FGFR antagonist SU5402 decreased expression of Erm by 83% and SP-C by 92%. Control PCR reactions that contained no template (H2O) were negative.


Downloaded from www.jbc.org at Childrens Hospital Medical Center on December 4, 2006 FIGURE 5. Erm enhances TTF-1 activity in regulating SP-C transcription. The sequence of the 0.32-kb mouse SP-C promoter is shown (A). The core sequence of Ets binding sites (GGA(A/T)) is shown in boldface. Mutated sequences M1 and M2 are shown in lowercase and boldface. TTF-1 binding sites are shown in boldface italics. The oligonucleotide probes used in electrophoretic mobility shift assays are overlined and labeled. Oligo 1 and Oligo 2 correspond to (⫺132 to ⫺113) and (⫺207 to ⫺188) of the murine SP-C sequence, respectively. The TATA promoter is boxed, and the transcription start site is indicated by a bent arrow and labeled ⫹1. EMSA (B) using MLE-15 cell nuclear extract showed that both Oligo 1 and Oligo 2 formed protein-DNA complexes that were competed by a 100-fold excess unlabeled self oligonucleotide but not by the same amount of an unrelated oligonucleotide. The addition of anti-Erm antibody disrupted binding of the oligonucleotides. The open arrow indicates the position of the free probe; the gray arrow indicates specific protein-DNA complexes. The reporter pGL3/SPCluc was transfected into MLE-15 (C) cells by itself, paired with pErmFLAG (Erm), paired with pCMV-TTF-1 (TTF-1), or with both pErmFLAG and pCMV-TTF-1. Cells were lysed 48 h after transfection, and relative luciferase activity was determined. Whereas Erm by itself did not affect SP-C promoter activation, TTF-1 significantly (*, p ⬍ 0.001) increased promoter activity. Co-transfection of both Erm and TTF-1 significantly enhanced SP-C promoter activity above that seen with TTF-1 alone (#, p ⬍ 0.001). Data are plotted as the means ⫾ S.E. (n ⫽ 7). Transfection of JEG-3 cells (D) with the pGL3/SPCluc plus Erm had no effect on SP-C transcription. Like MLE-15 cells, JEG-3 cells transfected with the reporter and TTF-1 showed significantly (*, p ⬍ 0.01) increased SP-C transcription, and the further addition of Erm resulted in a significant (#, p ⬍ 0.05) increase in activity. Mutation of the putative Ets binding site between bp ⫺200 to ⫺195 (M2) resulted in no significant SP-C transcription in the presence of Erm, TTF-1, or Erm plus TTF-1. Mutation of the putative Ets site between bp ⫺125 to ⫺120 (M1) resulted in no promoter activity in the presence of Erm. The M1 mutation, however, resulted in increased (*, p ⬍ 0.001) TTF-1-mediated stimulation of SP-C transcription that was also significantly (†, p ⬍ 0.001) greater than that seen with wild-type pGL3/SPCluc. Transfection of Erm plus TTF-1 significantly (#, p ⬍ 0.001) increased SP-C transcription above that seen with TTF-1 alone, and this stimulation was greater (†, p ⬍ 0.001) than that seen with the wild-type promoter. Mutation of both putative Ets binding sites resulted in no significant promoter activity with the addition of Erm, TTF-1, or Erm plus TTF-1. Data are plotted as means ⫾ S.E. (n ⫽ 5).

nal domain by itself (T1-C) interacted with full-length Erm, which appeared to be at odds with our observations that its deletion (T1-⌬C) had no effect on activity and that it had no activity when combined with the TTF-1 homeodomain (T1-⌬N). In an attempt to resolve this dispar-


ity, we assessed the interactions of the T1-C and T1-⌬C mutants with the active mutants Em-Ets and Em-⌬N. We found that T1-⌬C functionally interacted with both Erm mutants, whereas T1-C showed no interaction with either (data not shown).



Suppression of Erm with siRNA Inhibits Both Erm and SP-C Expression—To determine whether SP-C expression is regulated by Erm in a biologically relevant context, we transfected primary cultures of adult mouse alveolar type II cells with a siRNA that targeted bases 194 –214 of the Erm coding region (Erm194). To assess the efficacy of



FIGURE 6. Erm interacts with TTF-1. A, a model of the mammalian two-hybrid assay system is shown. The reporter plasmid pG5Luc was co-transfected into JEG-3 (B) and HeLa (C) cells with paired pVP16/ Erm-AD (AD/Erm) and pM/TTF-1-BD (BD/TTF-1). Control cultures were co-transfected with pG5Luc ⫹ BD/TTF-1 and empty pVP16, pG5Luc ⫹ AD/Erm and empty pMBD, or pG5Luc ⫹ empty pVP16 and empty pMBD. JEG-3 cells transfected with pG5Luc ⫹ BD/TTF-1 and empty pVP16 or pG5Luc ⫹ AD/Erm and empty pMBD showed no significant increase in luciferase activity when compared with pG5Luc ⫹ empty pVP16 and pMBD. JEG-3 cells transfected with pG5Luc ⫹ BD/TTF-1 and AD/Erm showed a 2300% increase in promoter activity ( p ⬍ 0.001, n ⫽ 5). An increase of 870% ( p ⬍ 0.05, n ⫽ 3) was seen when pG5Luc ⫹ BD/TTF-1 and AD/Erm were cotransfected into HeLa cells, which contain endogenous Erm. Data are plotted as the means ⫾ S.E. D, nuclear extracts were prepared from JEG-3 cells 48 h after transfection with pErmFLAG and/or pCMV-TTF-1. An aliquot of nuclear extract was incubated with anti-Erm antibody, and the immunoprecipitated proteins were subjected to SDSPAGE followed by Western blotting using antiTTF-1 (lane 3). Omitting pErmFLAG from the transfection mixture resulted in no reaction with anti-TTF-1 (lane 2). Substitution of normal rabbit serum (NRS) for anti-Erm resulted in no signal for TTF-1 (lane 4). An equivalent aliquot of nuclear extract that was not immunoprecipitated is shown in lane 1 to identify the TTF-1 band. E, identification of Erm domains required for interaction with TTF-1. Mutant Erm cDNAs consisting of the N-terminal (Em-N), Ets (Em-Ets), acidic (Em-AD), C-terminal (Em-C), and the Ets plus C-terminal (Em-⌬N) domains were generated by PCR, cloned into pVP16, and co-transfected with BD/TTF-1 ⫹ the reporter pG5Luc in the JEG-3 cells. Compared with BD/TTF-1 plus empty pVP16 (top lane), both Em-Ets and Em-⌬N significantly (*, p ⬍ 0.001, n ⫽ 7) interacted with TTF-1, but the activity was less than that seen with full-length Erm. F, identification of TTF-1 domains required for interaction with Erm. TTF-1 fragments, including the homeodomain (T1-HD), the N terminus (T1-N), the C terminus (T1-C), the N terminus plus homeodomain (T1-⌬C), and the C terminus plus homeodomain (T1-⌬N) were cloned into pMBD. These constructs were co-transfected with AD/Erm ⫹ pG5Luc in the JEG-3 cells. Compared with AD/Erm plus empty pMBD (top lane), both T1-C and T1-⌬C showed significant (*, p ⬍ 0.001, n ⫽ 7) activity in interacting with Erm. Only T1-⌬C, however, was able to interact with Em-Ets or Em-⌬N (not shown). Control cultures in which full-length TTF-1 or the deletion constructs were transfected with empty pVP16 vector gave no activity (not shown).

transfecting primary type II cell cultures with siRNAs, we transfected cells with a Cy3-labeled siRNA that targets luciferase (Cy3luc), which also served as a specificity control siRNA. As shown in Fig. 7A, adult type II cells cultured on Matrigel formed multicellular aggregates (34) after 4 days. Most of the cells in the type II cell aggregates transfected


16723


with Cy3luc contained labeled siRNA (Fig. 7B) on culture day 4 (72 h after transfection), demonstrating the efficacy of transfection. Quantitative real-time PCR showed that Erm194 inhibited Erm expression by 40% 72 h after transfection ( p ⬍ 0.01; Fig. 7C). This decrease was accompanied by a concomitant 51% reduction in SP-C expression ( p ⬍ 0.01; Fig. 7D). Expression of ␤-actin (Fig. 7E) and GAPDH (Fig. 7F) were unaffected by transfection with Erm194. Two other siRNAs, targeting bases 920 –940 and 1240 –1260 of Erm, had no effect on the expression of either Erm or SP-C (data not shown). Control cultures transfected with Cy3luc showed no diminution of any of the genes examined (Fig. 7, C–F).

DISCUSSION The Ets family of transcription factors in humans and mice currently comprises 27 genes that have been divided into 12 subfamilies (35). Ets proteins act as nuclear targets of signal transduction pathways and have been shown to play important roles in diverse biological processes, including cell proliferation, differentiation, embryogenesis, apoptosis, immune response, and oncogenic transformation (36 –38). All Ets proteins share a structurally conserved DNA binding domain of ⬃85 amino acids. The DNA binding site of all Ets proteins contains a central 5⬘-GGA(A/T)-3⬘ binding motif, and the sequences flanking this motif


confer the specificity and affinity of binding for individual Ets proteins (36, 38). Although most Ets proteins are transcriptional activators, some act as repressors (36). Protein-protein interactions have been shown to regulate the DNA binding, target gene selection, subcellular localization, and transcriptional activity of Ets proteins (33). A characteristic of Ets family members is that they interact with other transcription factors, such as AP-1 (39), AML1 (40), c-Myb (41), Pax-5 (42), NF␬B (43), Stat-5 (44), and others (for review, see Ref. 33) to exert combinatorial control over the expression of cell-specific genes. For example, Erm has been shown to synergize with c-Jun to activate a synthetic promoter element (32) as well as to partner with ␤-catenin-Tcf in transactivating matrilysin (45). Erm is a member of the Pea3 subfamily of Ets transcription factors that includes Pea3/Etv4, Erm/Etv5, and Er81/Etv1. Pea3 subfamily members are expressed at sites of active branching morphogenesis in the developing lung, salivary gland, kidney, and mammary gland (46, 47), suggesting that they may be downstream effectors of epithelialmesenchymal interactions. Our observation that Erm was induced when lung mesenchyme reprogrammed tracheal epithelium to a lung phenotype supports the concept that Erm is a gene expressed in the distal lung epithelium in response to factors produced in the mesenchyme. A role for Erm in lung development is further supported by the



FIGURE 7. siRNA-mediated suppression of Erm concomitantly decreases SP-C expression. Adult mouse alveolar type II cells were seeded onto Matrigel, transfected with siRNAs after 24 h, then harvested after an additional 72 h. At the end of the culture period the type II cells had formed multicellular aggregates on the surface of the gel (A). Cultures transfected with a Cy3-labeled siRNA targeting luciferase (Cy3luc) show that most cells contain siRNA (B), indicating a high efficiency of transfection. Real-time PCR demonstrated that sham-transfected cells or cells transfected with Cy3luc had no significant change in the content of Erm (C), SP-C (D), ␤-actin (E), or GAPDH (F). Cells transfected with siRNA targeting Erm, however, showed a 40% decrease in Erm (C; p ⬍ 0.01, n ⫽ 6) and a 51% decrease in SP-C (D; p ⬍ 0.01, n ⫽ 4).



found, however, that Erm by itself did not activate the minimal SP-C promoter but instead interacted with TTF-1 to enhance SP-C transcription. TTF-1 has been shown to be a key regulator of lung epithelial differentiation genes including CCSP (Clara cell secretory protein) (54), SP-A (30), SP-B (55), and SP-C (14). TTF-1 has been shown to function cooperatively with GATA-6 (26), NFI (31), and TAZ (56) in regulating SP-C transcription, suggesting that TTF-1 forms complexes with several transcriptional co-activators on the SP-C promoter. Our results indicate that Erm should be added to the list of TTF-1 partners. The activity of Erm in regulating SP-C transcription was further demonstrated by mutating the Ets binding sites. Mutation of the Ets site between bp ⫺200 to ⫺195 (M2) negated the synergistic effect of Erm on TTF-1-mediated SP-C transcription, suggesting that this is a critical site of Erm activity. This mutation also inhibited the activity of TTF-1 transfected without Erm. We do not think that this was due to the mutation interfering with the ability of TTF-1 to interact with the TTF-1 site at ⫺205 to ⫺202, because it has been previously demonstrated that this site has minimal activity in SP-C transcription (14). Mutation of the Ets site between bp ⫺125 to ⫺120 (M1) gave the contrasting but equally interesting result of significantly enhancing the effect of TTF-1 or TTF-1 plus Erm on SP-C transcription. We do not yet know the mechanism underlying this observation. One possibility is that Erm or perhaps another Ets factor represses SP-C transcription by binding to this site. The fact that the increases in SP-C transcription seen with the M1 mutation were ablated when both sites were mutated emphasizes the importance of the M2 site. TTF-1 is a critical regulator of SP-C transcription in lung epithelial cells. Our data demonstrate that Erm interacts with TTF-1 to synergistically activate SP-C transcription. The observation that Erm is regulated by FGFs may explain in part the fact that SP-C expression is FGFdependent. Our demonstration that siRNA-mediated suppression of Erm resulted in concomitant suppression of SP-C in primary cultures of adult type II cells gives biological relevance to the transfection data. Our observations, considered together with a recent study (48) demonstrating that Erm is required for normal lung morphogenesis, suggest that Erm plays multiple roles in lung development. Acknowledgments—We thank Dr. Jeffrey A. Whitsett for the anti-TTF-1 antibody, Dr. Cong Yan for the pG5Luc plasmid, and Dr. Cindy J. Bachurski for the pGL3/SPCluc plasmid. We are grateful to Dr. Stephan W. Glasser for helpful comments and suggestions. We also thank Xiaofei Shangguan, Kalpana Srivastava, and Kathleen McCormick-Shannon for expert technical assistance.

REFERENCES 1. Williams, M. C. (2003) Annu. Rev. Physiol. 65, 669 – 695 2. Mason, R. J., and Shannon, J. M. (1997) in The Lung: Scientific Foundations (R.G. Crystal, J. B. W., et al., eds) 2nd Ed., pp. 543–555, Lippincott-Raven, Philadelphia, PA 3. Shannon, J. M., Nielsen, L. D., Gebb, S. A., and Randell, S. H. (1998) Dev. Dyn. 212, 482– 494 4. Shannon, J. M. (1994) Dev. Biol. 166, 600 – 614 5. Taderera, J. T. (1967) Dev. Biol. 16, 489 –512 6. Shannon, J. M. (1996) J. Jpn. Med. Soc. Biol. Interface 26, (suppl.) 47S– 65S 7. Deterding, R. R., and Shannon, J. M. (1995) J. Clin. Investig. 95, 2963–2972 8. Deterding, R. R., Jacoby, C. R., and Shannon, J. M. (1996) Am. J. Physiol. 271, L495–L505 9. Cardoso, W. V., Itoh, A., Nogawa, H., Mason, I., and Brody, J. S. (1997) Dev. Dyn. 208, 398 – 405 10. Shannon, J. M., Gebb, S. A., and Nielsen, L. D. (1999) Development 126, 1675–1688 11. Hyatt, B. A., Shangguan, X., and Shannon, J. M. (2002) Dev. Dyn. 225, 153–165 12. Wert, S. E., Glasser, S. W., Korfhagen, T. R., and Whitsett, J. A. (1993) Dev. Biol. 156, 426 – 443 13. Glasser, S. W., Burhans, M. S., Eszterhas, S. K., Bruno, M. D., and Korfhagen, T. R. (2000) Am. J. Physiol. Lung Cell. Mol. Physiol. 278, L933–L945 14. Kelly, S. E., Bachurski, C. J., Burhans, M. S., and Glasser, S. W. (1996) J. Biol. Chem.


16725


observation that branching morphogenesis was disrupted when the 3.7-kb human SP-C promoter was used to drive expression of a truncated form of Erm fused to the Drosophila Engrailed repressor domain (EngR-Erm) in the distal lung epithelium (48). The mechanism by which Erm disrupts lung epithelial branching is not known. In addition to its involvement in lung branching morphogenesis, Erm also plays a role in distal lung epithelial differentiation. The distal epithelium in the lungs of fetuses expressing EngR-Erm primarily contained immature type II cells, and morphologically identifiable type I cells expressing Aquaporin 5 (Aqp5) did not differentiate (48). Our observation that Erm enhanced TTF-1-mediated transcriptional regulation of SP-C in transfected cells is the first demonstration of a role for Erm in a specific aspect of distal lung epithelial differentiation. Importantly, our observation that siRNA mediated suppression of Erm in primary cultures of type II cells concomitantly inhibits SP-C expression gives biological relevance to the transfection data. Because Er81 is expressed only in the lung mesenchyme and Pea3 is not detectable in the adult lung by Northern blot (46) or by reverse transcription-PCR (this paper), the ability to influence SP-C expression in adult type II cells is restricted to Erm among the Pea3 subfamily members. Because Erm regulates SP-C transcription, our data support the suggestion by Liu et al. (48) that the EngR-Erm transgene, which was expressed under control of the SP-C promoter, may have negatively regulated its own expression and thereby reduced the severity of its effects on the lung. Erm has been shown to be a nuclear target of the mitogen-activated protein kinase signaling cascade (49). Of the many upstream activators of mitogen-activated protein kinase signaling, the FGF family has been shown to be particularly important in activating Erm expression (50 – 52). The hypothesis that FGFs regulate Erm in the lung is supported by the observation that implanting FGF-soaked beads into the proximal region of embryonic lung explants resulted in the ectopic expression of both Erm and Pea3 (48) and by the fact that distal lung endoderm cultured in the presence of FGF7 maintained expression of Erm. Our observations on SU5402-treated lung explants confirm these results and extend them by demonstrating that inhibiting endogenous FGF signaling decreased Erm expression. Furthermore, our data showing that Erm is decreased in transgenic mice overexpressing a dominant-negative FGF receptor (FGFR-HFc) or an FGF antagonist (Spry-4) suggest that FGFs are also important regulators of Erm in vivo. Although FGFs clearly play an important role in Erm activation in the lung, they are not its sole regulators. This is indicated by our observation that SP-C expression, which is dependent on FGF signaling in the embryonic lung, was effectively ablated by SU5402, but Erm expression was still readily detectable although at a reduced level. Similarly, when we cultured embryonic tracheal epithelium, which expresses neither Erm nor SP-C, in a complex medium containing FGF1 and FGF7, we found that both Erm and SP-C were induced. However, when FGFs were omitted from the medium, the tracheal rudiments contained no detectable SP-C, but Erm was still induced. Erm expression in these cultures was likely due to globally elevated cAMP levels induced by cholera toxin, since Erm is also a nuclear target of protein kinase A (49, 53). The fact that SP-C was not induced in these cells even though Erm was present is consistent with our transient transfection studies showing that Erm by itself cannot transactivate the minimal SP-C promoter. These data also suggest that FGFs regulate factors that are additionally required for SP-C expression. Our observation that Erm and SP-C were expressed in the same cells in the developing and adult lung along with the result that the ectopic induction of Erm in tracheal epithelium was accompanied by co-expression of SP-C suggested that Erm might regulate SP-C transcription. We



143–156 35. Kurpios, N. A., Sabolic, N. A., Shepherd, T. G., Fidalgo, G. M., and Hassell, J. A. (2003) J. Mammary Gland Biol. Neoplasia 8, 177–190 36. Sharrocks, A. D. (2001) Nat. Rev. Mol. Cell Biol. 2, 827– 837 37. Sementchenko, V. I., and Watson, D. K. (2000) Oncogene 19, 6533– 6548 38. Oikawa, T., and Yamada, T. (2003) Gene (Amst.) 303, 11–34 39. Bassuk, A. G., and Leiden, J. M. (1995) Immunity 3, 223–237 40. Mao, S., Frank, R. C., Zhang, J., Miyazaki, Y., and Nimer, S. D. (1999) Mol. Cell. Biol. 19, 3635–3644 41. Shapiro, L. H. (1995) J. Biol. Chem. 270, 8763– 8771 42. Fitzsimmons, D., Hodsdon, W., Wheat, W., Maira, S. M., Wasylyk, B., and Hagman, J. (1996) Genes Dev. 10, 2198 –2211 43. Gitlin, S. D., Dittmer, J., Shin, R. C., and Brady, J. N. (1993) J. Virol. 67, 7307–7316 44. Rameil, P., Lecine, P., Ghysdael, J., Gouilleux, F., Kahn-Perles, B., and Imbert, J. (2000) Oncogene 19, 2086 –2097 45. Crawford, H. C., Fingleton, B., Gustavson, M. D., Kurpios, N., Wagenaar, R. A., Hassell, J. A., and Matrisian, L. M. (2001) Mol. Cell. Biol. 21, 1370 –1383 46. Chotteau-Lelievre, A., Desbiens, X., Pelczar, H., Defossez, P. A., and de Launoit, Y. (1997) Oncogene 15, 937–952 47. Chotteau-Lelievre, A., Montesano, R., Soriano, J., Soulie, P., Desbiens, X., and de Launoit, Y. (2003) Dev. Biol. 259, 241–257 48. Liu, Y., Jiang, H., Crawford, H. C., and Hogan, B. L. (2003) Dev. Biol. 261, 10 –24 49. Janknecht, R., Monte, D., Baert, J. L., and de Launoit, Y. (1996) Oncogene 13, 1745–1754 50. Raible, F., and Brand, M. (2001) Mech. Dev. 107, 105–117 51. Roehl, H., and Nusslein-Volhard, C. (2001) Curr. Biol. 11, 503–507 52. Firnberg, N., and Neubuser, A. (2002) Dev. Biol. 247, 237–250 53. Baert, J. L., Beaudoin, C., Coutte, L., and de Launoit, Y. (2002) J. Biol. Chem. 277, 1002–1012 54. Zhang, L., Whitsett, J. A., and Stripp, B. R. (1997) Biochim. Biophys. Acta 1350, 359 –367 55. Bohinski, R. J., Di Lauro, R., and Whitsett, J. A. (1994) Mol. Cell. Biol. 14, 5671–5681 56. Park, K. S., Whitsett, J. A., Di Palma, T., Hong, J. H., Yaffe, M. B., and Zannini, M. (2004) J. Biol. Chem. 279, 17384 –17390



271, 6881– 6888 15. Kimura, S., Hara, Y., Pineau, T., Fernandez-Salguero, P., Fox, C. H., Ward, J. M., and Gonzalez, F. J. (1996) Genes Dev. 10, 60 – 69 16. DeFelice, M., Silberschmidt, D., DiLauro, R., Xu, Y., Wert, S. E., Weaver, T. E., Bachurski, C. J., Clark, J. C., and Whitsett, J. A. (2003) J. Biol. Chem. 278, 35574 –35583 17. Lin, S., and Shannon, J. M. (2002) Chest 121, (suppl.) 80S– 81S 18. Hokuto, I., Perl, A. K., and Whitsett, J. A. (2003) J. Biol. Chem. 278, 415– 421 19. Perl, A. K., Hokuto, I., Impagnatiello, M. A., Christofori, G., and Whitsett, J. A. (2003) Dev. Biol. 258, 154 –168 20. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Cold Spring Harbor Laboratory, pp. 153–185, Cold Spring Harbor, New York 21. Wikenheiser, K. A., Vorbroker, D. K., Rice, W. R., Clark, J. C., Bachurski, C. J., Oie, H. K., and Whitsett, J. A. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 11029 –11033 22. Wilkinson, D. G. (1992) in In Situ Hybridization, A Practical Approach (Wilkinson, D. G., ed) pp. 75– 83, IRL Press, New York 23. Bachurski, C. J., Kelly, S. E., Glasser, S. W., and Currier, T. A. (1997) J. Biol. Chem. 272, 32759 –32766 24. Holzinger, A., Dingle, S., Bejarano, P. A., Miller, M. A., Weaver, T. E., DiLauro, R., and Whitsett, J. A. (1996) Hybridoma 15, 49 –53 25. Naltner, A., Ghaffari, M., Whitsett, J. A., and Yan, C. (2000) J. Biol. Chem. 275, 56 – 62 26. Liu, C., Glasser, S. W., Wan, H., and Whitsett, J. A. (2002) J. Biol. Chem. 277, 4519 – 4525 27. Rice, W. R., Conkright, J. J., Na, C. L., Ikegami, M., Shannon, J. M., and Weaver, T. E. (2002) Am. J. Physiol. Lung Cell. Mol. Physiol. 283, L256 –L264 28. Kalina, M., Mason, R. J., and Shannon, J. M. (1992) Am. J. Respir. Cell Mol. Biol. 6, 595– 600 29. Minoo, P., Su, G., Drum, H., Bringas, P., and Kimura, S. (1999) Dev. Biol. 209, 60 –71 30. Bruno, M. D., Bohinski, R. J., Huelsman, K. M., Whitsett, J. A., and Korfhagen, T. R. (1995) J. Biol. Chem. 270, 6531– 6536 31. Bachurski, C. J., Yang, G. H., Currier, T. A., Gronostajski, R. M., and Hong, D. (2003) Mol. Cell. Biol. 23, 9014 –9024 32. Nakae, K., Nakajima, K., Inazawa, J., Kitaoka, T., and Hirano, T. (1995) J. Biol. Chem. 270, 23795–23800 33. Li, R., Pei, H., and Watson, D. K. (2000) Oncogene 19, 6514 – 6523 34. Shannon, J. M., Mason, R. J., and Jennings, S. D. (1987) Biochim. Biophys. Acta 931,