Thyroid transcription factor-1 (TTF-1) gene: identification of ZBP-89, Sp1, and TTF-1 sites in the promoter and regulation by TNF-α in lung epithelial cells

Abstract

Thyroid transcription factor-1 (TTF-1/Nkx2.1/TITF1) is a homeodomain-containing transcription factor essential for the morphogenesis and differentiation of the lung. In the lung, TTF-1 controls the expression of surfactant proteins that are essential for lung stability and lung host defense. In this study, we identified functionally important transcription factor binding sites in the TTF-1 proximal promoter and studied tumor necrosis factor-α (TNF-α) regulation of TTF-1 expression. TNF-α, a proinflammatory cytokine, has been implicated in the pathogenesis of acute respiratory distress syndrome (ARDS) and inhibits surfactant protein levels. Deletion analysis of TTF-1 5′-flanking DNA indicated that the TTF-1 proximal promoter retained high-level activity. Electrophoretic mobility shift assay, chromatin immunoprecipitation, and mutational analysis experiments identified functional ZBP-89, Sp1, Sp3, and TTF-1 sites in the TTF-1 proximal promoter. TNF-α inhibited TTF-1 protein levels in H441 and primary alveolar type II cells. TNF-α inhibited TTF-1 gene transcription and promoter activity, indicating that transcriptional mechanisms play important roles in the inhibition of TTF-1 levels. TNF-α inhibited TTF-1 but not Sp1 or hepatocyte nuclear factor-3 DNA binding to TTF-1 promoter. Transactivation experiments in A549 cells indicated that TNF-α inhibited TTF-1 promoter activation by exogenous Sp1 and TTF-1 without altering their levels, suggesting inhibition of transcriptional activities of these proteins. TNF-α inhibition of TTF-1 expression was associated with increased threonine, but not serine, phosphorylation of Sp1. Because TTF-1 serves as a positive regulator for surfactant protein gene expression, TNF-α inhibition of TTF-1 expression could have important implications for the reduction of surfactant protein levels in diseases such as ARDS.

thyroid transcription factor-1 (TTF-1/Nkx2.1/TITF1) is a 38-kDa homeodomain-containing transcription factor that controls the expression of select genes in the thyroid gland and the lung epithelium. TTF-1 serves as an activator for the expression of surfactant protein (SP)A, B, C, and D (9, 12, 16, 32, 41, 46), Clara cell secretory protein (CCSP) (63, 71), lysosomal-associated membrane protein 3, and carcinoembryonic antigen-related cell adhesion molecule (34) genes in the respiratory epithelial cells of the lung. TTF-1 is expressed selectively in the thyroid, lung, and restricted regions of the central nervous system (38). TTF-1 is expressed at the onset of lung morphogenesis and is subject to spatiotemporal pattern of expression during development (72). In the mouse lung, TTF-1 immunoreactivity is first detected in the nuclei of bronchial epithelial cells on gestational day 10. In gestational day 11 lung, high-level TTF-1 expression is detected in the trachea and bronchi. On gestational days 12–16, high-level TTF-1 expression is maintained in the trachea and bronchi with expression in bronchioles, distal acinar buds, and tubules. In gestational day 17 lung, high-level TTF-1 expression is maintained in epithelial cells lining conducting airways and subsets of epithelial cells in the distal lung. In the adult lung, a similar pattern of expression as in day 17 lung is maintained with similar expression levels in type II cells and epithelial cells lining bronchi and bronchioles. The TTF-1-null mice lack lung parenchyma, thyroid gland, and the pituitary and contained extensive defects in the ventral region of the forebrain, underscoring the importance of TTF-1 for the development of the thyroid, lung, forebrain, and the pituitary (33). Partial TTF-1 deficiency in humans attributable to mutations in TTF-1 gene is associated with hypothyroidism, choreoathetosis, respiratory dysfunction, and recurrent pulmonary infections (18, 29, 35). Recent studies have demonstrated that sustained expression of TTF-1 is necessary for the growth and survival of a subset of lung adenocarcinoma, implicating TTF-1 as lineage-specific protooncogene for lung cancer (36, 62, 65).

Rat (19, 43), mouse (50), and human (24, 27, 43) TTF-1 genes have been cloned and characterized. Rat (43) and human (24) TTF-1 genes contain three exons, and multiple transcription start sites and alternative splicing produce mRNAs with heterogeneity at the 5′ end (49). Human TTF-1 gene contains two promoters; one lies within the first intron (proximal promoter), and the other is upstream of the first exon (distal promoter) (24). The proximal promoter does not contain a TATA sequence but instead contains a sequence TAAAA that has some degree of similarity to the TATA element, whereas the distal promoter lacks a TATA-like sequence altogether. The TTF-1 proximal promoter contains two closely located DNA elements that bind hepatocyte nuclear factor (HNF)-3α (FOXA1) and HNF-3β (FOXA2) factors in murine lung epithelial (MLE)-15 cells and are important for promoter activity (28). The TTF-1 distal promoter contains a GC-rich sequence that binds Sp1 and Sp3 factors and is necessary for promoter activity in H441 cells (40). TTF-1 promoter is activated by coexpression of TTF-1 in HepG2 (49) and FRTL-5 thyroid cells (48), indicating that it is subject to positive autoregulation.

Tumor necrosis factor-α (TNF-α), an early response cytokine, is an important mediator of lung inflammation and is present at high levels in the blood and bronchoalveolar lavage fluid of patients with acute respiratory distress syndrome (ARDS) (26) and infants with chronic lung disease (17). The levels of total phospholipids, SP-A, and SP-B in bronchoalveolar lavage are significantly reduced in patients with ARDS and patients at risk for ARDS compared with normal patients (21, 23). The association between high levels of TNF-α and reduced levels of surfactant phospholipids and SPs in patients with ARDS suggests a role for TNF-α in the reduction of surfactant lipids and proteins. TNF-α inhibits the expression of SP-A (67), SP-B, (4, 54), and SP-C (2) genes via transcriptional and posttranscriptional mechanisms. TNF-α inhibition of SP-B promoter activity was associated with reduced TTF-1 DNA binding activity (4).

Considering the important roles that TTF-1 plays in the control of gene expression in the lung and the fact that very little is known about its regulation, we sought to identify additional transcription factors important for TTF-1 promoter regulation and studied TNF-α regulation of TTF-1 expression in H441 lung epithelial cells. We identified functional ZBP-89, Sp1/Sp3, and TTF-1 sites in the TTF-1 proximal promoter region. We found that TNF-α inhibited TTF-1 gene transcription and promoter activity, indicating that transcriptional mechanisms are partly responsible for the inhibition of TTF-1 expression. TNF-α inhibition of TTF-1 promoter activity was associated with reduced levels and binding activity of TTF-1. Although TNF-α did not inhibit Sp1 binding, it inhibited Sp1 activation of TTF-1 promoter, indicating inhibition of Sp1 transcriptional activity.

MATERIALS AND METHODS

Cell culture.

NCI-H441 cells (HTB-174; American Type Culture Collection, Manassas, VA), a human lung adenocarcinoma cell line with characteristics of bronchiolar (Clara) epithelial cells, and A549 cells (CCL-185, American Type Culture Collection), a human lung carcinoma cell line with certain characteristics of type II cells, were grown on plastic tissue culture dishes in RPMI 1640 and F12K medium, respectively, containing 10% fetal bovine serum, penicillin (100 u/ml), streptomycin (100 μg/ml), and amphotericin B (0.25 μg/ml) in a humidified atmosphere of 95% room air-5% CO₂. In some experiments, H441 cells were incubated overnight (16–17 h) in serum-free medium before TNF-α treatment. Alveolar type II cells were isolated from fetal human lung explants and maintained in culture as described previously (1). Type II cells were exposed to TNF-α in culture medium devoid of dibutyryl cAMP. Human fetal lungs collected from second-trimester abortuses were obtained from Advanced Bioscience Resources (Alameda, CA). The Committee for the Protection of Human Subjects of the University of Texas-Houston Health Science Center approved the procurement of fetal human lungs and the protocol for the isolation of type II cells.

Cell viability.

Cell viability was assessed by measuring the release of lactate dehydrogenase (LDH) into medium using CytoTox 96 nonradioactive cytotoxicity assay kit (Promega, Madison, WI) according to the instructions provided by the manufacturer. A standard curve was generated by measuring total LDH released upon lysis of known numbers of cells, and percentage of cell death in test samples was determined from the standard curve. Cell viability was also assessed by the Trypan blue exclusion assay.

Materials.

Recombinant TNF-α was from Upstate (Charlottesville, VA), and mithramycin was from AG Scientific (San Diego, CA). Lipofectamine 2000 was from Invitrogen (Carlsbad, CA).

Plasmid construction.

Human TTF-1 genomic fragment comprising −2,861/+213 bp flanked by Mlu I and Xho I restriction enzyme sites was synthesized by GenScript (Piscataway, NJ), and its sequence was verified. The translation start codon ATG and another ATG codon three nucleotides downstream were mutated (ATG * CTG) to block potential interference with the translation of luciferase reporter gene. Genomic fragments comprising −2,015/+213 bp and −446/+213 bp were obtained by restriction enzyme digestion of the plasmid containing −2,861/+213 bp fragment with Sac I + Xho I and Sma I + Xho I, respectively, and purified by Agarose gel electrophoresis. The cloning of the −401/+213 bp TTF-1 promoter fragment has been described previously (8). Fragments comprising −98/+213 and +44/213 bp were amplified by PCR, and their sequences were verified. TTF-1 promoter fragments were inserted upstream of luciferase gene in the promoterless vector pGL3luc(basic) vector. The sequences of sense primers and anti-sense primer for the amplification of promoter fragments are as follows: 5′-CTCGGGCAAGATGTAGGCTTC-3′ (−98/−78 bp); 5′-CGGTCCACTCCGTTACGTGTA-3′ (+44/+64 bp); 5′-TTGGACTCAGCGACAGGATTC-3′ (+213/+193 bp, antisense).

Site-directed mutagenesis.

Nucleotides within the transcription factor binding sites of TTF-1 promoter were altered by site-directed mutagenesis using QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) according to the manufacturer's instructions. TTF-1 promoter (−446/+213 bp) containing mutation in ZBP-89 site at −219/−214 bp was used as a template to introduce mutation at ZBP-89 site −315/−310 bp. The double ZBP-89 mutant was used as a template to introduce mutations in ZBP-89 sites at −271/−266, −297/−292, and −285/−280 bp in a single reaction using mutagenic primer ZBP-6. Mutated promoter fragments were sequenced to verify the presence of mutations and then cloned into pGL3luc(basic) vector. The sense nucleotide sequences of the mutagenic primers (mutations are underlined) are shown in Table 1.

Table 1.

Sequences of sense oligonucleotide primers used for mutagenesis of transcription factor binding sites in TTF-1 promoter

Oligo Name and Location	Sequence	Description
TTF-1 (+7/+40 bp)	5′- GCCCCCAGCCTCCACACAACCCAATTAAGGAGG -3′	TTF-1 site at +22/+27 bp
TTF-1 (+165/+197 bp)	5′- CCTCTTCCTTCCTCCACCACCCGCCGTCGAATC -3′	TTF-1 site at +179/+184 bp
Sp1 (−65/−31 bp)	5′- GCTTTAGCGCTTACGCCCCGAATCTGGTGGCTGCC -3′	Sp1 site at −50/−42 bp
ZBP-1 (−236/−201 bp)	5′- GGTGGGGGTGTCCAAGTTCTGGGGGAAATGCCTTGG -3′	ZBP-89 site at −219/−214 bp
ZBP-2 (−293/−250 bp)	5′- CCTTCCCCTCCCCTCCACCTCCTAACCGCGCTCCGCTGGTACC -3′	ZBP-89 site at −271/−266 bp
ZBP-3 (−305/−261 bp)	5′- CGGTTCCCTCCTCCTTCCCCTAACCTCCACCTCCTCCCCGCGCTC -3′	ZBP-89 site at −285/−280 bp
ZBP-4 (−318/−275 bp)	5′- CCCTCCCCGCTCTCGGTTCCCTAATCCTTCCCCTCCCCTCCACC -3′	ZBP-89 site at −297/−292 bp
ZBP-5 (−335/−299 bp)	5′- CCCCACCTCCTGCCAGAACCGCTCTCGGTTCC -3′	ZBP-89 site at −315/−310 bp
ZBP-6 (−308/−258 bp)	5′- TCTCGGTTCCCTAATCCTTCCCCTAACCTCCACCTCCTAACCGCGCTCCGC -3′	ZBP-89 sites at −271/−266, −285/−280 and −297/−292 bp

TTF, thyroid transcription factor.

Transient transfection and reporter gene assay.

Plasmid DNAs were amplified in Escherichia coli Top10 strain (Invitrogen) and purified by anion exchange chromatography (Qiagen, Valencia, CA). Plasmids were transiently transfected along with pcDNA3.1 (Invitrogen), a β-galactosidase expression plasmid, into cells by liposome-mediated DNA transfer with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Luciferase and β-galactosidase activities in cell extracts were measured by chemiluminescent assays (Tropix, Bedford, MA and Promega), and luciferase activities were normalized to cotransfected β-galactosidase activity to correct for variations in transfection efficiency. Expression plasmids encoding TTF-1 [pCMV4-thyroid-specific enhancer binding protein (T/EBP1)] and Sp1 (pCMV-Sp1) were kindly provided by Dr. Shioko Kimura, National Cancer Research Institute, Bethesda, MD and Dr. Robert Tjian, University of California, Berkeley, CA, respectively. Expression plasmid encoding ZBP-89 (pcDNA3-Myc-ZBP89-Flag) was kindly provided by Dr. Juanita L. Merchant, University of Michigan, Ann Arbor, MI. Plasmid pSP72 containing Sp3 cDNA was obtained from American Type Culture Collection, and Sp3 cDNA fragment was inserted into pcDNA3 expression vector.

RNA isolation and real-time quantitative RT-PCR.

Total RNA from cells was isolated using RNeasy mini kit (Qiagen) according to the manufacturer's instructions and then treated with RNase-free DNase to remove contaminating genomic DNA. Real-time quantitative RT-PCR was performed using the 7700 Sequence Detector (Applied Biosystems, Foster City, CA). Synthesis of cDNA by reverse transcription and PCR amplification reactions were performed according to standard protocols. Each sample was measured in triplicate plus a control without reverse transcriptase. The names of the genes quantified, the sequences of the primers and probes, and other parameters of the assay are shown in Table 2.

Table 2.

Genes quantified, sequences of the primers and probes, and other parameters of the assay

Gene	Accession No.	Primer and Probe Sequences	Assay Efficiency, %	Amplicon Length	Detection Limit, molecules
Human TITF-1	NM_003317	341(+)- CATGTCGATGAGTCCAAAGC	97	91	160
		420(−)- CTTTCTTGTAGCTTTCCTCCAG
		363(+)FAM- ACGACTCCGTTCTCAGTGTCTGACAT -BHQ1
Human18S rRNA	M10098	1335(+)- CGGCTTAATTTGACTCAACAC	97	68	210
		1401(−)- ATCAATCTGTCAATCCTGTCC
		1359(+) FAM- AAACCTCACCCGGCCCG -BHQ1

Preparation of nuclear extracts and EMSA.

Methods for the preparation of nuclear extracts (57, 60) and double-stranded oligonucleotides (45) have been described previously. The sense strand sequences of human TTF-1 promoter and other oligonucleotides used in mobility shift analysis are shown in Table 3.

Table 3.

Sequences of sense oligonucleotides used in electrophoretic mobility shift assays

Oligo Name and Location	Sequence	Description
TTF-1 (+17/+33 bp)	5′- CTCCACTCAAGCCAATT -3′	TTF-1 site at +22/+27 bp
TTF-1mutant (+17/+33 bp)	5′- CTCCACACAACCCAATT -3′	TTF-1 site at +22/+27 bp
TTF-1 (+174/+190 bp)	5′- TCCTCCTCCAGCCGCCG -3′	TTF-1 site at +179/+184 bp
HNF-3 (−19/+5 bp)	5′- GCCGGGCTAAAACAAACGCGAGGC -3′	HNF-3 site at −11/−6 bp
HNF-3mutant (−19/+5 bp)	5′- GCCGGGCTACCACTAACGCGAGGC -3′	HNF-3 site at −11/−6 bp
HNF-3 (−140/−120 bp)	5′- CTTAAAGGTGTTTACCTTGTC -3′	HNF-3 site at −132/−127 bp
HNF-3mutant (−140/−120 bp)	5′ CTTAAAGGGTAGAACCTTGTC -3′	HNF-3 site at −132/−127 bp
Sp1 (−55/−38 bp)	5′- TTACGCCCCGCCTCTGGT -3′	Sp1 site at −50/−42 bp
Sp1 consensus	5′- ATTCGATCGGGGCGGGGCGAG -3′	consensus Sp1
Sp1mutant	5′- ATTCGATCGGTTCGGGGCGAGC -3′	mutant Sp1
ZBP-1 (−243/−209 bp)	5′- AAATGGGGGTGGGGGTGTCCAAGGGAGGGGGGAA -3′	ZBP-89 site at −219/−214 bp
ZBP-1mutant (−243/−209 bp)	5′- AAATGGGGTATTGGGTGTCCAAGTTCTGGGGGAA -3′	mutant ZBP-89 site at −219/−214 bp
ZBP-2 (−279/−262 bp)	5′- CCACCTCCTCCCCGCGCT -3′	ZBP-89 site at −271/−266 bp
ZBP-2mutant (−279/−262 bp)	5′- CCACCTCCTAACCGCGCT -3′	mutant ZBP-89 site at −271/−266 bp
ZBP-3 (−293/−276 bp)	5′- CCTTCCCCTCCCCTCCAC -3′	ZBP-89 site at −285/−280 bp
ZBP-3mutant (−293/−276 bp)	5′- CCTTCCCCTAACCTCCAC -3′	mutant ZBP-89 site at −285/−280 bp
ZBP-4 (−293/−276 bp)	5′- CCTTCCCCTCCCCTCCAC -3′	ZBP-89 site at −297/−292 bp
ZBP-4mutant (−293/−276 bp)	5′- CCTTCCCCTAACCTCCAC -3′	mutant ZBP-89 site at −297/−292 bp
ZBP-5 (−339/−304 bp)	5′- ACTCCCCCACCCCACCTCCTGCCCTCCCCGCTCTCG -3′	ZBP site at −315/−310 bp
ZBP-5mutant (−339/−304 bp)	5′- ACTCCCAATACAATACTCCTGCCAGAACCGCTCTCG -3′	mutant ZBP site at −315/−310 bp
ZBP-89p21	5′- GAGGGACTGGGGGAGGAGGGAAGTGCCCTC -3′	ZBP-89 site in p21 promoter
TTF-1 SP-B	5′- GCACCTGGAGGGCTCTTCAGAGCAA -3′	TTF-1 site in human SP-B promoter
TTF-1mutant SP-B	5′- GCACCAGGACGGCTCATCACAGCAA -3′	mutant TTF-1 site in human SP-B promoter
HNF-3 SP-B	5′- GCAAAGACAAACACTGAG -3′	HNF-3 site in human SP-B promoter

HNF, hepatocyte nuclear factor; SP, surfactant protein.

Double-stranded oligonucleotides were 5′-end labeled using [γ³²P]-ATP and T4 polynucleotide kinase and electrophoretic mobility shift analysis (EMSA) performed as described previously (45) by incubating 0.5–1.0 ng (100,000 cpm) of labeled oligonucleotide with 5 μg of nuclear protein in 20 μl of binding buffer [13 mM HEPES, pH 7.9, containing 13% glycerol, 80 mM KCl, 5 mM MgCl₂, 1 mM DTT, 1 mM EDTA, and 1 μg of poly(dI-dC) as nonspecific competitor DNA] at 30°C for 20 min. The amount of DTT in nuclear isolation buffers and reaction mixture was 1 mM. For antibody supershift assay, protein-DNA complex was first formed and then incubated with antibody or antiserum for 20 min at room temperature. Polyclonal Sp1, Sp3, and ZBP-89 antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA), and polyclonal rabbit antiserum against the amino-terminal portion of rat T/EBP (TTF-1/Nkx2.1) was kindly provided by Dr. Shioko Kimura. In competition experiments, excess cold oligonucleotide was added along with the labeled oligonucleotide to the incubation mixture. After electrophoresis, the gel was dried and exposed to an X-ray film or Phosphor screen.

ChIP assay.

Procedures for chromatin isolation and immunoprecipitation were according to instructions of chromatin immunoprecipitation (ChiP)-IT express enzymatic kit (Active Motif, Carlsbad, CA). Typically, chromatin was isolated from 2–3 × 10⁷ H441 cells. Normal IgG and TTF-1 (mouse monoclonal antibody; Thermo Scientific, Rockford, IL), Sp1 (Active Motif), and HNF-3α (FoxA1) (Santa Cruz Biotechnology) antibodies were used at 2 μg per reaction in immunoprecipitation. After isolation of DNA, TTF-1 promoter fragment was amplified by 36 cycles of PCR (GoTaq Hot Start Green Master Mix, Promega), and amplified DNA was analyzed by Agarose gel electrophoresis. The sequences of sense and antisense oligonucleotides for the amplification of TTF-1 promoter are as follows: sense: 5′-CGGGCAAGATGTAGGCTTCTATTG-3′ (−96/−73 bp); antisense: 5′-TCTGGTGTTACCTTAACGCCGATC-3′ (+75/+98 bp).

Immunoblotting analysis.

SDS-PAGE separation and transfer of proteins to membrane were carried out with an XCell II Mini-Cell apparatus (Novex, San Diego, CA) according to the manufacturer's instructions. Equal amounts of protein (5–10 μg) were separated by SDS-PAGE on 10% Bis-Tris gels with 3-morpholinopropanesulfonic acid running buffer and electrophoretically transferred to PVDF membranes. Membranes were successively incubated with primary antibodies or antisera at 1:1,000 dilution overnight at 4°C followed by goat anti-rabbit alkaline phosphatase-conjugated secondary antibody (Cell Signaling, Beverly, MA) at 1:2,000 dilution for 1 h at room temperature. Protein bands were visualized by the enhanced chemifluorescence detection method (Amersham Pharmacia Biotech, Piscataway, NJ) and quantified using Quantity One Image Acquisition and Analysis Software (Bio-Rad, Hercules, CA). Polyclonal rabbit antibodies against human actin and Sp1 antibodies were from Santa Cruz Biotechnology. Polyclonal rabbit antisera against the amino-terminal portion of rat T/EBP (TTF-1/Nkx2.1) was kindly provided by Dr. Shioko Kimura, National Cancer Institute, Bethesda, MD, and mouse monoclonal TTF-1 antibody was from Thermo Fisher Scientific. Polyclonal rabbit antibody against ZBP-89 was from Santa Cruz Biotechnology.

Immunofluorescence microscopy.

H441 cells were grown on four-chamber Permanox slides (Nalge Nunc International, Naperville, IL) for 24 h and then treated with or without TNF-α (25 ng/ml) for 24 h. Cells were fixed for 1 h at 4°C in PBS containing 4% paraformaldehyde and 0.1% glutaraldehyde. After fixation, cells were permeabilized with 0.1% Triton X-100 in PBS for 10 min and blocked with 10% goat serum in PBS for 1 h at room temperature. Cells were incubated overnight at 4°C with rabbit polyclonal T/EBP IgG (10 μg/ml) followed by incubation with Rhodamine Red-conjugated goat anti-rabbit IgG for 60 min at room temperature. Nuclei were stained with SYTOX Green for 5 min according to manufacturer's instructions. The immunostained cells were observed under a Nikon Eclipse TE2000–5 inverted fluorescent microscope equipped with UltraVIEW LCI scanning confocal system (Perkin Elmer, Boston, MA). Fluorescence images were captured with a digital coupled-charged device camera (Ultra Pix, Hamamatsu Photonics, Japan) at ×20 or ×60 optical lens (oil) magnification at room temperature using 488- and 568-nm excitation laser lines at a resolution of 1,344 × 1,024 × 12. Imaging SuiteTM version 5.0 acquisition and processing software was used for the acquisition of the images and measuring the colocalization by determining the correlation coefficient of overlap of the green and red fluorescence. The images were imported into Microsoft PowerPoint program for compilation of figures.

Statistical analyses.

Data are shown as means ± SE, unless otherwise indicated. In experiments where TTF-1 mRNA/protein/promoter activity levels were arbitrarily set at 100%, statistical significance was analyzed by a one-sample t-test. For other samples, a paired t-test was used to analyze the statistical significance. One-tailed P values of <0.05 were considered significant.

RESULTS

Deletion analysis of TTF-1 5′ flanking DNA.

There is limited information on the genomic regions controlling TTF-1 promoter activity in bronchiolar (Clara) epithelial cells. H441 cell line is a differentiated cell line that expresses TTF-1, SP-A, SP-B, SP-D, and CCSP and hence is well suited to study regulation of TTF-1 expression in Clara cells. Our deletion analysis experiments (Fig. 1A) demonstrated that deletion of the TTF-1 5′ flanking region from −2,861/+213 bp to −401/+213 bp that resulted in the loss of the distal promoter did not decrease but modestly increased promoter activity. Further deletion of the −401/+213 bp sequence to −98/+213 and −44/+213 bp reduced promoter activities by ∼50% and 80%, respectively, revealing two regions within −401/−98 bp and −98/+44 bp important for promoter activity. Examination of the sequence revealed the presence of several sequence motifs similar to the binding of ZBP-89 transcription factor within the −401/−98 bp region. Sequence motifs similar to binding of Sp1, TTF-1, and NF-Y were found within the −98/+44 bp region. The ZBP-89, Sp1, TTF-1, and NF-Y binding sequences and their locations in the human, rat, and mouse TTF-1 promoters were similar, indicating that they may be important for promoter activity (Fig. 1B).

Fig. 1.

A: deletion analysis of the thyroid transcription factor-1 (TTF-1) 5′ flanking DNA. A schematic diagram of human TTF-1 5′ flanking DNA fragment is shown. The arrows indicate the locations of transcription start sites in the distal and proximal promoters. DNA fragments containing deletions at the 5′ end and possessing a common 3′ end were obtained by restriction enzyme digestion or amplification by PCR. The promoter fragments were inserted into promoterless pGL3luc(basic) luciferase reporter plasmid and transiently transfected along with the β-galactosidase reporter plasmid pcDNA3.0 into H441 cells. After 24 h of incubation, luciferase and β-galactosidase activities of cell extracts were determined and luciferase activity was normalized to β-galactosidase activity to correct for variations in transfection efficiencies. Results are shown as means ± SE (n = 3). *P < 0.05 and ***P < 0.001 vs. −2,861/+213 bp construct, **P < 0.01 for −98/+213 bp vs. −401/+213 bp, *P < 0.05 for +44/+213 bp vs. −401/+213 bp. B: alignment of sequences of human, rat, and mouse TTF-1 promoters. The nucleotide numbers for the human TTF-1 promoter are indicated, and transcription factor binding sites are enclosed in boxes. Alignment indicated that the transcription factor binding sequences and their placement are similar in human, rat, and mouse promoters.

TTF-1 proximal promoter contains functional Sp1, TTF-1, and ZBP-89 DNA elements.

Deletion analysis of TTF-1 5′ flanking DNA demonstrated that the −401/+213 bp TTF-1 fragment retained most of the promoter activity of longer fragments and had slightly higher activity than the −2,861/+213 bp fragment, indicating that it contains DNA elements necessary for promoter activity. Several putative transcription factor binding DNA sequences that have not yet been identified were found in the −401/+213 bp TTF-1 promoter (Fig. 2). The locations of putative DNA binding sequences are as follows: Sp1 at −50/−42 bp, TTF-1 at +22/+27 and +174/+190 bp, ZBP-89 at −219/−214, −271/−266, −285/−280, −297/−292, and −315/−310 bp, and NF-Y at +28/+32 bp.

Fig. 2.

Functional analyses of ZBP-89, Sp1, and TTF-1 binding sites in the TTF-1 promoter. Transcription factor sites were mutated, and the mutated promoter plasmids were transiently transfected along with the β-galactosidase reporter plasmid, pcDNA3.0 into H441 cells. After 24 h of incubation, luciferase and β-galactosidase activities of cell extracts were determined and luciferase activity was normalized to β-galactosidase activity to correct for variations in transfection efficiencies. Data shown are means ± SE (n = 3). A: mutational analysis of ZBP-89 binding sites. Locations of ZBP-89 sites are shown. wt, wild-type; mt, mutant. *P < 0.05, **P < 0.01, ***P < 0.001, and #P < 0.0001 vs. wild-type promoter. B: activation of TTF-1 promoter by exogenous ZBP-89. Plasmid pGL3luc containing −401/+213 bp of TTF-1 promoter DNA was cotransfected with 1.0 μg of empty expression plasmid pcDNA3 or ZBP-89 expression plasmid into H441 cells. After 36 h of incubation, luciferase activities in cell extracts were determined and normalized to total protein content of cell extracts. Data shown are means ± SD (n = 2). **P < 0.01 vs. pcDNA3 transfected cells. C: H441 cells were transiently transfected with no DNA (mock), ZBP-89 expression plasmid (ZBP-89), or empty expression plasmid (pcDNA3.0). After 36 h, ZBP-89 and actin levels were determined by Western immunoblotting. In cells transfected with ZBP-89 plasmid, ZBP-89 levels increased by 3-fold compared with cells transfected with empty expression plasmid. D: mutational analysis of TTF-1 and Sp1 sites. Mutations of TTF-1 (+22/+27 bp) and Sp1 (−50/−42 bp) sites in TTF-1 promoter (−401/+213 bp) were analyzed. **P < 0.01 and ***P < 0.001 vs. wild-type promoter. E: activation of TTF-1 promoter by exogenous Sp1, Sp3, and TTF-1. Plasmid pGL3luc containing −401/+213 bp of TTF-1 promoter DNA was cotransfected with increasing concentrations (0.1, 0.25, 1.0 μg) of pcDNA3, Sp1, Sp3, or TTF-1 expression plasmids into A549 cells. After 36 h of incubation, luciferase activities in cell extracts were determined and normalized to total protein content of cell extracts. Data shown are means ± SE (n = 3–7). *P < 0.05 and **P < 0.01 vs. pcDNA3-transfected cells.

We first determined the importance of the DNA elements in TTF-1 promoter function by mutational analysis and then characterized proteins binding to the DNA elements by EMSA. As the importance of HNF-3 sites has been studied previously (28), we focused on the functions of ZBP-89, Sp1, TTF-1, and NF-Y sites in TTF-1 promoter activity. Transcription factor binding sites were mutated, and the effects of mutations on TTF-1 promoter activity were determined by transient transfection analysis in H441 cells. Mutations of ZBP-89 sites had varying degrees of inhibitory effects on TTF-1 promoter (Fig. 2A). Whereas mutation of ZBP-89 site at −219/−214 bp reduced promoter activity by ∼35%, mutations of ZBP sites at −271/−266, −285/−280, or −315/−318 bp reduced promoter activity by ∼20%. Mutation of the ZBP-89 site at −297/−292 bp had no effect on the promoter activity. Combined mutations of all of the ZBP-89 sites reduced promoter activity by ∼50%, consistent with deletion experiments that showed similar loss of promoter activity upon deletion of ZBP-89 sites. Cotransfection of ZBP-89 expression plasmid activated TTF-1 promoter activity in H441 cells, consistent with the functional roles for ZBP-89 sites (Fig. 2B). TTF-1 promoter activation (∼3-fold compared with cells transfected with empty vector) was associated with similar increases in ZBP-89 protein levels (∼3-fold compared with cells transfected with empty vector) (Fig. 2C). We further assessed the importance of ZBP-89 in TTF-1 promoter function by studying the activation of wild-type and mutant (all ZBP-89 sites) TTF-1 promoters (−446/+213 bp) by exogenous ZBP-89 in H441 cells. Data demonstrated that, whereas ZBP-89 activated wild-type TTF-1 promoter by threefold, it activated mutant TTF-1 promoter by twofold [wild-type, 2.94 ± 0.31; mutant 1.94 ± 0.06 (n = 3), P = 0.034]. These data in agreement with mutational analysis experiments indicated that ZBP-89 sites are important for TTF-1 promoter activation. Mutations of the Sp1 and TTF-1 sites reduced TTF-1 promoter activity by ∼70% and 40%, respectively, indicating the importance of the sites in the promoter function (Fig. 2D). Mutation of the NF-Y site had no effect on TTF-1 promoter activity (data not shown). Cotransfected Sp1, Sp3, and TTF-1 expression plasmids activated TTF-1 promoter activity in a dose-dependent manner in A549 cells, consistent with the functional roles for these sites in the promoter function (Fig. 2E). Sp1 was significantly more potent than Sp3 and TTF-1 in activating TTF-1 promoter. Cotransfection experiments were performed using A549 cells because they express very low or undetectable TTF-1 (10 and unpublished observations, V. Boggaram), enabling easier detection of the effects of exogenous TTF-1 on promoter activity. TTF-1 promoter activation was associated with increased levels of Sp1 and TTF-1 as determined by Western immunoblotting analysis (data not shown). We next characterized proteins binding to ZBP-89, Sp1, TTF-1 sites by EMSA and ChIP experiments. In EMSA experiments, ZBP sites at −219/−214 and −315/−310 bp formed several complexes upon incubation with H441 nuclear proteins. The complexes with slower mobilities (bands I, II, and III) were competed by excess of the same and p21 promoter ZBP-89 oligonucleotide (3), but not mutant oligonucleotide, indicating the interaction of ZBP-89 with the DNA sequence (Fig. 3A). ZBP sites at −271/−266 and −285/−280 bp also formed several complexes upon incubation with H441 nuclear proteins. The complexes with slower mobilities (bands I and II) were competed by excess of the same oligonucleotide and p21 promoter ZBP-89 oligonucleotide but not mutant oligonucleotide, indicating interaction of ZBP-89 with the DNA sequence (Fig. 3B). Oligonucleotide containing ZBP-89 site at −297/−292 bp formed very weak or no band with H441 nuclear proteins, indicating that the site may not be a ZBP-89 binding site (data not shown), consistent with lack of effect of mutation of this site on TTF-1 promoter activity. Interestingly this ZBP-89 site is not conserved in the TTF-1 promoters (Fig. 1B). ZBP-89 bands were also competed by consensus and TTF-1 promoter Sp1, but not mutant Sp1 oligonucleotides, indicating interaction of Sp1 with the ZBP-89 binding sites (Fig. 3, C and D).

Fig. 3.

Electrophoretic mobility shift assay (EMSA) analysis of ZBP-89 binding sites in the TTF-1 promoter. ³²P-labeled double-stranded TTF-1 promoter oligonucleotides were incubated with 5 μg of H441 cell nuclear extract in the absence or presence of 100× molar excess of competitor oligonucleotides. The reactions were analyzed by polyacrylamide gel electrophoresis. Arrows indicate the mobilities of the protein-DNA complexes. A and B: competition with TTF-1 promoter oligonucleotides containing wild-type or mutated ZBP-89 sites and oligonucleotide containing ZBP-89 binding site from p21 promoter. C and D: competition with oligonucleotides containing a consensus (c) or mutant (m) Sp1 site and TTF-1 promoter Sp1 site.

To obtain further evidence for the interaction of ZBP-89 and Sp1 with the ZBP-89 sites in the TTF-1 promoter, we determined the effects of ZBP-89, Sp1, and Sp3 antibodies on complex formation by EMSA (Fig. 4, A and B). We found that, whereas ZBP-89 antibodies modestly reduced the intensities of bands I and II, Sp1 antibodies had no effect on the intensities or the migration of the bands. On the other hand, Sp3 antibodies reduced the intensities of bands I and III for ZBP-89 sites at −219/−214 and −315/−310 bp and bands I and II for ZBP-89 sites at −271/−266 and −285/−280 bp, indicating interaction of Sp3 with ZBP-89 sites. Despite the failure of Sp1 antibodies to interact with the DNA-protein complexes, competition experiments strongly indicated Sp1 interaction with the ZBP-89 sites. The failure of Sp1 antibodies to react with protein-DNA complexes in EMSA experiments suggested that the antibodies may not recognize Sp1. To verify this possibility, we determined the effects of Sp1 and Sp3 antibodies on the mobility of an oligonucleotide containing a Sp1 consensus binding site using H441 nuclear extracts. We found that Sp3 but not Sp1 antibodies blocked protein-DNA complex, indicating that Sp1 antibodies do not recognize Sp1 antigen per se under conditions of EMSA (data not shown).

Fig. 4.

EMSA analysis of nuclear proteins binding to ZBP-89 sites in TTF-1 promoter. A and B: ³²P-labeled double-stranded TTF-1 promoter oligonucleotides were first incubated with 5 μg of H441 cell nuclear protein and then nonimmune IgG (nIgG) or ZBP-89, Sp1, or Sp3 antibodies. The reactions were analyzed by polyacrylamide gel electrophoresis. Arrows indicate the mobilities of protein-DNA complexes.

Sp1 binding site formed two bands upon incubation with H441 nuclear proteins that were competed by excess amounts of unlabeled wild-type and consensus Sp1 oligonucleotides but not by an oligo-carrying mutation in the Sp1 binding site (Fig. 5A). The formation of the Sp1 bands was reduced by Sp1 antibody and nearly prevented by Sp3 antibody (Fig. 5B). Taken together, these data indicated that Sp1 and Sp3 proteins interact with the Sp1 site. TTF-1 binding site at +22/+27 bp formed a single band upon incubation with H441 nuclear proteins, and its formation was reduced by excess amount of wild-type and SP-B TTF-1 but not by mutant oligonucleotide, indicating interaction of TTF-1 (Fig. 5C). The mobility of the DNA-protein band was retarded by TTF-1 antibody, indicating the identity of the bound protein as TTF-1 (Fig. 5D). The putative TTF-1 site at +174/+190 bp did not form complexes with H441 nuclear proteins, indicating that the site may not bind proteins (data not shown). In agreement with the results of EMSA experiments, ChIP assay demonstrated that the TTF-1 proximal promoter bound, Sp1, HNF-3α, and TTF-1 in vivo in H441 cells (Fig. 6). HNF-3α binding to the promoter was weak but was well above normal IgG (nIgG) control. The weak HNF-α binding in ChIP experiments could be due to the weak affinity of the HNF-3α antibody to antigen cross linked to DNA. We performed EMSA to further establish HNF-3 binding to TTF-1 promoter (Fig. 6C). Competition experiments with wild-type and HNF-3 mutant oligonucleotides clearly demonstrated HNF-3 binding to TTF-1 promoter. H441 cells express HNF-3α but not HNF-3β (56).

Fig. 5.

EMSA analysis of nuclear proteins binding to Sp1 and TTF-1 sites in the TTF-1 promoter. ³²P-labeled double-stranded TTF-1 promoter oligonucleotides were incubated with 5 μg of H441 cell nuclear extract in the absence or presence of 100× molar excess of competitor oligonucleotides. For the identification of proteins binding to the sites, ³²P-labeled double-stranded TTF-1 promoter oligonucleotides were first incubated with 5 μg of H441 cell nuclear protein and then nonimmune IgG (nIgG) or Sp1, Sp3, or TTF-1 antibodies. The reactions were analyzed by polyacrylamide gel electrophoresis. Arrows indicate the mobilities of protein-DNA complexes. A: EMSA of Sp1 site in the presence of 100× molar excess of consensus or mutant Sp1 oligonucleotide. B: effects of nonimmune IgG, Sp1, and Sp3 antibodies on the mobilities of oligonucleotide containing Sp1 site. C: EMSA of TTF-1 site in the presence of 100× molar excess of wild-type and mutant TTF-1 and SP-B promoter oligonucleotides. D: effects of nonimmune IgG (nIgG) and TTF-1 antibodies (thyroid-specific enhancer binding protein antisera) on the mobilities of oligonucleotide containing TTF-1 site.

Fig. 6.

Chromatin immunoprecipitation (ChIP) analysis of nuclear proteins binding to TTF-1 promoter. A: H441 cell chromatin was immunoprecipitated with TTF-1, Sp1, or hepatocyte nuclear factor (HNF)-3α antibodies. nIgG and normal serum (nserum) served as controls. After removal of cross links, the immunoprecipitated DNAs were amplified by PCR using TTF-1 primers, and the amplified DNAs were analyzed by agarose gel electrophoresis. Arrow indicates the amplified DNA fragment. B: schematic diagram of the amplified region of TTF-1 promoter with the locations of the transcription factor binding sites is shown. C: EMSA of HNF-3 sites in the presence of 100× molar excess of wild-type and mutant TTF-1 and SP-B promoter oligonucleotides.

TNF-α inhibits TTF-1 mRNA and protein levels.

TNF-α inhibits SP-A, SP-B, and SP-C levels in lung cells and lung tissue (6). TTF-1 is a common transcriptional activator of SP-A, SP-B, and SP-C genes (6). Our previous studies on TNF-α inhibition of SP-B promoter activity indicated that TNF-α reduces TTF-1 DNA binding activity to inhibit SP-B promoter activity in H441 cells (4). TNF-α inhibition of surfactant protein expression could be due to reduced TTF-1 binding activity and/or TTF-1 expression. In the present investigation, we sought to understand molecular mechanisms underlying TNF-α inhibition of TTF-1 levels. We first determined the effects of TNF-α on TTF-1 expression in H441 (Fig. 7, A and B) and human primary type II cells (Fig. 7C) by analyzing its effects on TTF-1 protein and mRNA levels by Western immunoblotting and quantitative RT-PCR, respectively. Results demonstrated that TNF-α reduced TTF-1 levels in a concentration-dependent manner, and TTF-1 inhibition was associated with inhibition of mRNA levels. Analysis of cell viability by assay of lactate dehydrogenase in cell medium showed that TNF-α at concentrations used did not cause cell death [cell death: control cells = 0.69%, TNF-α (25 ng/ml) treated cells at 24 h = 1%, n = 4]. Analysis of the effects of TNF-α on cellular TTF-1 levels in H441 cells by immunofluorescence microscopy showed that TTF-1 was restricted to the nuclear compartment in untreated cells, and TNF-α treatment reduced TTF-1 nuclear content without eliciting changes in cytosolic staining (Fig. 7D).

Fig. 7.

TNF-α inhibits TTF-1 expression in H441 and primary alveolar type II cells. A: H441 cells were treated with or without tumor necrosis factor (TNF)-α for 24 h, and TTF-1 levels in nuclear extracts were determined by Western immunoblotting and normalized to actin. Data shown are means ± SE (n = 5), *P < 0.05 and ***P < 0.001 vs. control cells. B: H441 cells were treated with or without TNF-α (25 ng/ml) for 24 h, and TTF-1 mRNA levels were determined by quantitative RT-PCR and normalized to 18S rRNA levels. Data are means ± SD (n = 2). **P < 0.01 vs. control cells. C: human primary alveolar type II cells were treated with or without TNF-α for 24 h. TTF-1 levels in nuclear extracts were determined by Western immunoblotting and normalized to actin levels. Data shown are means ± SD (n = 2). *P < 0.05 and **P < 0.01 vs. control cells. D: H441 cells were incubated with or without TNF-α (25 ng/ml) for 24 h and then processed for immunofluorescence microscopy with TTF-1 antibodies. Immune-complexes were stained with rhodamine red-conjugated anti-rabbit IgG, and nuclei were visualized by staining with Sytox Green. Similar results were obtained in a second independent experiment.

TNF-α inhibits TTF-1 gene transcription and promoter activity.

To determine whether TNF-α inhibits TTF-1 expression at the transcriptional level, we analyzed the effects of TNF-α on TTF-1 gene transcription and TTF-1 promoter activity by run-on transcription assay in isolated nuclei and transient transfection assay in H441 cells. We found that TNF-α reduced the TTF-1 transcription rate and TTF-1 promoter activity, indicating that transcriptional mechanisms are partly responsible for the inhibition of TTF-1 expression (Fig. 8, A and B). TNF-α inhibited TTF-1 promoter activity in a time-dependent manner with inhibition apparent at 6 h of treatment.

Fig. 8.

TNF-α inhibits TTF-1 gene transcription and TTF-1 promoter activity in H441 cells. A: cells were incubated with or without TNF-α (25 ng/ml) for 24 h and nuclei isolated. Transcription run-on assay was performed on the isolated nuclei using ³²P-UTP, and labeled RNA was hybridized to nitrocellulose filters immobilized with plasmid vector, TTF-1, and actin cDNAs. Similar results were obtained for a second independent experiment. B: cells were transiently transfected with plasmid pGL3luc containing TTF-1 promoter −401/+213 bp along with β-galactosidase reporter pcDNA3.0. Transfected cells were incubated with or without TNF-α for 24 h, and luciferase activities were determined and normalized to β-galactosidase activity. Data shown are means ± SE (n = 3). *P < 0.05, **P < 0.01 vs. control cells. C: cells were transiently transfected with plasmid pGL3luc containing TTF-1 promoter −401/+213 bp along with β-galactosidase reporter pcDNA3.0. Transfected cells were incubated with or without TNF-α (25 ng/ml) for indicated times, and luciferase activities were determined and normalized to β-galactosidase activity. Data shown are means ± SE (n = 3). *P < 0.05 and **P < 0.01 compared with control cells.

TNF-α inhibits TTF-1 DNA binding activity.

Our experiments demonstrated that TNF-α inhibits TTF-1 expression by reducing TTF-1 gene transcription. Our experiments also demonstrated that Sp1 and TTF-1 DNA elements are important for TTF-1 proximal promoter activity. Other studies have shown that HNF-3 elements are important for TTF-1 promoter activity (28). To determine whether TNF-α modulates the DNA binding activities of Sp1, TTF-1, and HNF-3 transcription factors to inhibit TTF-1 promoter activity, we analyzed the effects of TNF-α on Sp1, TTF-1, and HNF-3 DNA binding activities by EMSA. Results indicated that TNF-α reduced TTF-1 DNA binding activity in a concentration-dependent manner, but the DNA binding activities of Sp1 and HNF-3 elements were unaffected (Fig. 9, A and C). Antibody-EMSA demonstrated inhibition of TTF-1 binding in a time-dependent manner in TNF-α-treated cells (Fig. 9B). Interestingly, decreases in TTF-1 DNA binding activities were not accompanied by similar decreases in TTF-1 levels at earlier times of treatment; however, at 24 h both TTF-1 binding activity and levels were decreased. These data indicated that reduced TTF-1 DNA binding activity in addition to reduced TTF-1 levels in TNF-α-treated cells may be involved in the inhibition of TTF-1 expression. Analysis of TTF-1 binding in vivo by ChIP assay also indicated reduced TTF-1 binding to TTF-1 proximal promoter in cells exposed to TNF-α (Fig. 9D). Our preliminary experiments indicated that TNF-α did not alter DNA binding activities of TTF-1 promoter ZBP-89 sites (data not shown).

Fig. 9.

TNF-α inhibits TTF-1 DNA binding activity in H441 cells. A and C: cells were incubated with or without TNF-α for 24 h, and nuclear extracts were prepared. TTF-1, HNF-3, and Sp1 DNA binding activities were analyzed by EMSA. Data are means ± SE (n = 3–5). *P < 0.05 and **P < 0.01 vs. control cells. HNF-3 and Sp1 binding activities in TNF-α-treated cells were not significantly different from those in control cells. B: TTF-1 DNA binding activity and TTF-1 levels in nuclear extracts of cells incubated with or without TNF-α (25 ng/ml) were analyzed by antibody-EMSA and Western immunoblotting, respectively. D: TTF-1 binding in vivo in cells incubated with or without TNF-α (25 ng/ml) for 24 h was analyzed by ChIP assay. Amplified DNA was analyzed by Agarose gel electrophoresis. Noncontiguous lanes are reassembled and are demarcated by white lines. Similar results were obtained for a second independent experiment.

TNF-α inhibits Sp1 and TTF-1 activation of TTF-1 promoter.

Sp1 and TTF-1 are subject to posttranslational modifications by phosphorylation, acetylation, and others that modulate their DNA binding and transcriptional activation functions (42, 70). We investigated whether TNF-α inhibition of TTF-1 promoter activity is associated with changes in the transcriptional activities of Sp1 and TTF-1. We first studied the effects of TNF-α on TTF-1 promoter activation by cotransfected Sp1 and TTF-1 in A549 cells and then correlated the effects of TNF-α on promoter activation with levels of cotransfected Sp1 and TTF-1. A549 cells were chosen for the cotransfection experiments because they do not express detectable levels of TTF-1 (10) (unpublished observations, V. Boggaram). Transfected Sp1 could be easily distinguished from the endogenous Sp1 because of its lower molecular weight. Results indicated that Sp1 and TTF-1 markedly activated TTF-1 promoter activity and that TNF-α treatment significantly inhibited TTF-1 promoter activation (Fig. 10A). Determination of the levels of transfected TTF-1 and Sp1 showed that TNF-α did not reduce their expression, indicating that reduced promoter activity in TNF-α-treated cells is not attributable to reduced Sp1 or TTF-1 expression (Fig. 10B). These data indicated that TNF-α inhibited Sp1 and TTF-1 transcriptional activities to inhibit TTF-1 promoter activation. Because posttranscriptional modifications influence DNA binding and transcriptional activities of Sp1 and TTF-1, we hypothesized that TNF-α-induced changes in the posttranslational modifications of Sp1 and TTF-1 may play roles in TTF-1 promoter inhibition.

Fig. 10.

TNF-α inhibits TTF-1 promoter activation by exogenous Sp1 and TTF-1 without altering Sp1 and TTF-1 levels. TNF-α increases threonine phosphorylation on Sp1. A: A549 lung cells were transiently transfected with plasmid pGL3luc containing −401/+213 bp of TTF-1 promoter along with empty expression plasmid pcDNA3.0 or expression plasmids for Sp1 and TTF-1. Transfected cells were incubated with or without TNF-α (25 ng/ml) for 24 h, and luciferase activities were normalized to total cell protein. Data shown are means ± SE (n = 3). ***P < 0.001 and **P < 0.01 for Sp1 and TTF-1 transfected cells vs. pcDNA3.0 transfected cells. **P < 0.01 for Sp1 or TTF-1 transfected cells treated with TNF-α vs. Sp1 or TTF-1 transfected cells in control medium. B: A549 lung cells were transiently transfected with plasmid pGL3luc containing −401/+213 bp of TTF-1 promoter along with empty expression plasmid pcDNA3.0 or expression plasmids for Sp1 and TTF-1. Transfected cells were incubated with or without TNF-α (25 ng/ml) for 24 h. The levels of exogenous Sp1 and TTF-1 were determined by Western immunoblotting and normalized to actin levels. Data are means ± SE (n = 3). Sp1 and TTF-1 levels in TNF-α-treated cells were not significantly different from levels in control cells. C: H441 cells were treated with or without TNF-α (25 ng/ml) for 6 h, and cell lysates were prepared. Sp1 was immunoprecipitated with Sp1 antibodies and analyzed by Western immunoblotting with phospho-threonine, phospho-serine, and Sp1 antibodies. The same membrane was successively probed with different antibodies. The levels of phosphorylated Sp1 were normalized to total Sp1 levels to correct for loading errors. Quantification showed that TNF-α treatment increased threonine phosphorylation by 2.3 ± 0.51 (mean ± SD, n = 2).

TNF-α stimulates Sp1 phosphorylation.

We focused on Sp1 because the modulation of its DNA binding and transcriptional activities by phosphorylation are well studied (42). The levels of phosphorylated forms of Sp1 were determined by immunoprecipitation using Sp1 antibodies followed by Western immunoblotting with antibodies specific for modifications by phosphorylation. The levels of phosphorylated Sp1 were normalized to total Sp1 levels to correct for loading differences. Results showed that TNF-α significantly increased threonine phosphorylation on Sp1 (2.33 ± 0.5 compared with control, n = 2); however, serine phosphorylation was not altered after 6 h of treatment (Fig. 10C). TNF-α increased threonine phosphorylation on Sp1 by ∼2.5-fold after 6 h of treatment. Increased threonine phosphorylation on Sp1 was also found when Sp1 phosphorylation was analyzed by immunoprecipitation first with phospho-threonine antibodies followed by Western immunoblotting with Sp1 antibodies (data not shown). Our attempts to analyze TTF-1 phosphorylation by immunoprecipitation and Western immunoblotting experiments were unsuccessful. This could be due to the inability of the antibodies to successfully immunoprecipitate TTF-1. These data indicated that TNF-α inhibition of TTF-1 expression is associated with increased threonine phosphorylation of Sp1.

DISCUSSION

The 5′ flanking DNAs of human, rat, and mouse TTF-1 genes display a high degree of sequence similarity and support high-level reporter gene expression in H441 and MLE-15 lung cells, indicating the presence of DNA regulatory elements necessary for lung-specific expression (7). Regulatory elements and interacting transcription factors that control TTF-1 promoter activity particularly in Clara epithelial cells are incompletely defined. Two promoter regions having activity in H441 cells were identified in the human TTF-1 gene; the proximal promoter is located within the first intron, and the distal promoter is located upstream of the first exon (24). Our deletion mapping studies demonstrated that the proximal TTF-1 promoter is sufficient for high-level reporter gene expression in H441 Clara cells and identified two regions, −401/−98 bp and −98/+44 bp, that are necessary for promoter activity. Although the +44/+213 bp fragment lost much of the activity of the larger fragments, it still retained ∼16% of the activity of the −401/+213 bp fragment. This could suggest the presence of a transcription start site within the +44/+213 bp region. TTF-1 gene is known to contain multiple transcription start sites (50). Previous studies found that deletion of human TTF-1 5′ flanking DNA from −2,700/+3 bp to −550/+3 bp that results in the loss of the distal promoter reduced reporter gene activity in H441 cells (27). We found that the deletion of −2,861/+213 bp to −401/+213 bp did not decrease but modestly increased reporter gene activity. The reasons for the differences between our and previous studies (27) are not clear, but a notable distinction between the DNA constructs used is the presence of an additional 119 bp at the 3′ end in the constructs used in our study. It is possible that presently unidentified factors interacting with this region could influence promoter activity, explaining the different results obtained in our studies.

The −401/−98 bp region contained several sequence motifs for the binding of ZBP-89 that were found to be important for promoter activity. Overexpression of ZBP-89 increased TTF-1 promoter activity in a concentration-dependent manner consistent with the stimulatory roles of ZBP-89 sites for promoter activity. ZBP-89 is a member of the Kruppel family of zinc finger proteins that also includes Sp1, Egr1, and GATA1 (11). ZBP-89 is ubiquitously expressed in vertebrates and acts as both a repressor and an activator to modulate gene transcription. Whereas ZBP-89 serves as a positive regulator of promoters of ornithine decarboxylase (37), p21 (3), and GATA1 (52) genes, it represses gastrin (47), vimentin (68), and CD11b (53) genes. ZBP-89 is known to directly interact with Sp1 (68) and also bind the coactivator p300 (3). Our EMSA experiments indicated that the ZBP-89 sites in TTF-1 promoter interact with Sp1, indicating that they also serve as binding sites for Sp1. The occurrence of multiple ZBP-89 binding sites and a Sp1 binding site in the TTF-1 proximal promoter, in the absence of a conventional TATA binding sequence, may facilitate the binding of TFIID, leading to the formation of transcription initiation complexes. Clustering of Sp1 binding sites are found in several TATA-less promoters (5, 25, 39, 44) and are believed to promote the formation of transcription initiation complexes through interactions between the glutamine-rich domain of Sp1 and TAFII-110, a TFIID-associated protein (15, 20). Although TTF-1 proximal promoter lacks multiple Sp1 sites, ZBP-89 sites can bind Sp1 to provide the glutamine-rich domains necessary for interaction with TAFII-110.

The −98/+44 bp region contained functional binding sites for Sp1, Sp3, and TTF-1. Mutational and cotransfection experiments indicated that Sp1 serves as a significantly stronger stimulatory factor than Sp3 and TTF-1. TTF-1 proximal promoter contains DNA elements that bind HNF-3α and HNF-3β factors in MLE-15 lung cells and activate promoter (28). The distal promoter contains a GC-rich sequence that binds Sp1 and Sp3 factors to activate TTF-1 promoter activity in H441 cells (40). A DNA element that binds GATA-6 was identified in the mouse proximal TTF-1 promoter and found to be important for promoter activity in MLE-15 lung cells (59). However, mice deficient in GATA-6 expressed TTF-1 in Clara epithelial cells, suggesting that GATA-6 may not be important for TTF-1 expression in Clara cells (31). The homeobox protein HOX3B binds to a highly conserved DNA element in the TTF-1 proximal promoter and activates promoter activity in cotransfection experiments in HeLa cells (22, 31). The significance of HOX3B regulation of TTF-1 promoter activity is not clear because HOX3B is not expressed in MLE-15 or H441 cells and is detected only in the mesenchyme in the developing lung (58).

The presence of multiple transcription factor binding sites in the TTF-1 promoter indicates that the promoter activity is dependent on combinatorial or cooperative interactions between the transcription factors. Our studies also demonstrated that TTF-1 serves as an activator for its own gene expression by binding to a site in the proximal promoter. Although previous studies demonstrated that TTF-1 coexpression activated its own promoter activity in FRTL-5 thyroid cells (49) and HepG2 cells (48), they did not identify the TTF-1 binding site. The DNA elements involved in the activation of TTF-1 promoter are strikingly similar to those in the promoters of other lung-specific genes such as SP-B (6), CCSP (63), and alveolar type I cell protein T1α (55). The minimal promoters of SP-B, CCSP, and T1α contain functional sites for TTF-1, Sp1/Sp3, and HNF-3 factors similar to TTF-1 proximal promoter, perhaps indicating similarities in the transcriptional networks controlling respiratory epithelial cell functions.

TNF-α plays important roles in the pathogenesis of lung injury and is elevated in patients with ARDS (26) and premature infants with chronic lung disease (17). TNF-α inhibits SP-A, SP-B, and SP-C expression in lung epithelial cells by transcriptional and posttranscriptional mechanisms (6). In H441 cells, TNF-α (4) and ceramide (61) decreased TTF-1 DNA binding activity, suggesting that reduced TTF-1 DNA binding activity is partly responsible for the inhibition of SP expression. We previously concluded that in H441 cells TNF-α reduced TTF-1 DNA binding activity without altering nuclear TTF-1 levels (4). Our present data is at variance with our previous finding. The exact reasons for the discrepancy with our previous study are not clear. It is possible that the low concentration of TNF-α (5 ng/ml) and the prolonged time of incubation (48 h) employed in our previous study could have contributed to the lack of effect on TTF-1 levels. In the present study, we consistently found inhibitory effects of TNF-α at 25 ng/ml on TTF-1 levels by not just one method but by different methods such as, immunofluorescence, Western blotting, qRT-PCR, and promoter and transcription run-on assays. We are, therefore, confident in our present data that clearly show TNF-α inhibition of TTF-1 levels.

The relatively long incubation time (24 h) required for TNF-α inhibition of TTF-1 protein levels perhaps suggests that the inhibition occurs primarily at the transcriptional level and that TTF-1 mRNA and protein are relatively stable with their stabilities unaffected by TNF-α treatment. The relatively shorter incubation time (6 h) required for TNF-α inhibition of TTF-1 promoter activity supports such an interpretation. Other studies have documented relatively long incubation times required for TNF-α inhibition as in the case of TNF-α inhibition SP-A and SP-B mRNA levels in H441 cells that required 16 h for 50% reduction (67). TNF-α binding to its receptor activates signaling pathways that control such diverse processes as apoptosis, cell proliferation, differentiation, and gene expression (3a). Activation of AP-1, NF-κB, p38, and JNK-MAPK signaling pathways commonly occur as a result of TNF-α binding to its receptor (3a). TNF-α has also been suggested to elicit intracellular signaling by altering the cellular redox potential via production of reactive oxygen intermediates (19a). Cross talk between various signaling pathways activated by TNF-α occurs, explaining its multiple biological effects. The importance of cellular signaling pathways for the TNF-α inhibition of TTF-1 expression is not known.

Chronic exposure of rat alveolar type II cells to a combination of TGF-β and TNF-α decreased immunoreactive TTF-1 levels and increased α-smooth muscle actin, indicating induction of epithelial-mesenchymal transition (69). In E10 mouse lung epithelial cells, TGF-β increases TTF-1 levels via TGF-βII receptor and Smad signaling with the involvement of Sp1 and Sp3 (30). In FRTL-5 thyroid cells, TNF-α inhibits TTF-1 DNA binding activity and TTF-1 levels to inhibit thyroglobulin gene expression (51). In the present study, we found that TNF-α decreased TTF-1 mRNA and protein levels. TNF-α reduced TTF-1 promoter activity and TTF-1 gene transcription, indicating that transcriptional mechanisms are partly responsible for the inhibition of TTF-1 expression. Whether posttranscriptional mechanisms such as mRNA stability contribute to TNF-α inhibition of TTF-1 levels is not known. These data indicated that reduced TTF-1 DNA binding activity in TNF-α-treated cells must be partly attributable to reduced TTF-1 levels. Although TNF-α did not decrease Sp1 DNA binding activity in H441 cells, it inhibited Sp1 transactivation of TTF-1 promoter without affecting its levels in A549 cells. Similarly, TNF-α inhibited transactivation of TTF-1 promoter by TTF-1 in A549 cells without affecting the levels of transfected TTF-1. These data suggested that additional mechanisms such as TNF-α-induced changes in the posttranslational modifications of Sp1 and TTF-1, chromatin structure, and others could be responsible for the inhibition of TTF-1 promoter activity. Changes in the posttranslational modifications could inhibit DNA binding and/or transcriptional activities of Sp1 and TTF-1, causing inhibition of TTF-1 promoter activity. Our data showing that TNF-α reduced TTF-1 DNA binding activity independently of TTF-1 levels (Fig. 9B) suggested modulation of binding activity by posttranslational modification(s). Changes in histone acetylation, phosphorylation, and methylation are known to influence chromatin structure, leading to activation or inhibition of promoter activity (13).

Both Sp1 (14, 42) and TTF-1 (7) are subject to a variety of posttranslational modifications such as phosphorylation, oxidation, and acetylation that modulate their activities to alter target gene expression. Whereas, in certain cases, simultaneous DNA binding and promoter activation are accompanied by Sp1 phosphorylation, in others it is not. Oxidation of cysteine residues, under conditions of oxidative stress, in zinc finger-containing transcription factors such as Sp1, result in reduced DNA binding and transcriptional activities (64). Although generally oxidant stress results in reduced Sp1 DNA binding and transcriptional activity, there are instances where oxidant stress increased DNA binding and transcriptional activities of Sp1 and Sp3 to increase promoter activity as in the case of hyperoxia regulation of Na,K-ATPase-β1 (66). Our data indicated that, although TNF-α did not inhibit Sp1 DNA binding activity, it inhibited Sp1 activation of TTF-1 promoter, indicating independent regulation of Sp1 DNA binding activity and promoter activation. Segregation of promoter activation and DNA binding functions of Sp1 are known to be involved in the regulation of target gene expression (14). We found that TNF-α specifically stimulated the phosphorylation of threonine residues in Sp1, suggesting that increased threonine phosphorylation may be involved in inhibition of TTF-1 promoter activity. Further studies are required to directly link increases in threonine phosphorylation of Sp1 with downregulation of TTF-1 promoter activity. It remains to be determined whether TNF-α treatment of H441 cells causes oxidation of Sp1 and Sp3 to reduce TTF-1 promoter activity.

In summary, our studies have indicated that the TTF-1 proximal promoter is sufficient for promoter activity in H441 cells and that ZBP-89, Sp1, Sp3, and TTF-1 serve as positive regulators of TTF-1 promoter. Our studies have also shown that TNF-α reduces TTF-1 expression by inhibiting gene transcription. TNF-α inhibition of TTF-1 promoter activity is due to reduced expression of TTF-1 and reduced transcriptional activities of Sp1 and TTF-1. TNF-α inhibition of TTF-1 expression may be one of the mechanisms for inhibition of SP expression in lung diseases such as ARDS, wherein abnormalities of surfactant composition and function are encountered.

GRANTS

This work was partly supported by National Heart, Lung and Blood Institute Grants HL-R01 48048 (V. Boggaram) and HL-R01 068116 (J. Alcorn).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.