Peroxisome proliferator-activated receptor gamma and ligands inhibit surfactant protein B gene expression in the lung.

Pulmonary nonciliated bronchiolar epithelial cells (Clara cells) and alveolar type II (AT II) epithelial cells are responsible for surfactant synthesis and secretion. These cells are highly lipogenic with a high lipid turnover rate. Although only 10% of surfactant lipids are neutral lipids, they play very important roles in maintaining pulmonary surfactant homeostasis. Many metabolic intermediate products of neutral lipids serve as ligands for various nuclear receptors that bind to target genes to influence gene transcription. In this report, the functional role of the neutral lipid metabolites, 15-deoxy-Delta12,14-prostaglandin J2 and 9-hydroxyoctadecanoic acids, and peroxisome proliferator-activated receptor gamma was evaluated in surfactant protein B gene regulation. These reagents down-regulated surfactant protein B gene expression in respiratory epithelial cells at the transcriptional level in both cell line and whole lung explant systems. The studies support the concept that surfactant protein B homeostasis is influenced by neutral lipid metabolites in the lung.

During the respiratory cycles, the pulmonary surfactant protects the lung from collapse by lowering the air-liquid interface tension. Pulmonary surfactant is composed of 90 -95% lipids and 5-10% surfactant proteins that are synthesized, stored, and secreted by alveolar type II (AT II) 1 epithelial cells and nonciliated bronchiolar epithelial cells (Clara cells). Similar to the liver, the lung is an organ with a high lipid turnover rate to fulfill the need for surfactant synthesis. The majority of surfactant lipids are phospholipids (ϳ80%). Disaturated phos-phatidylcholine (PC), principally dipalmitoyl-PC, is the major phospholipid component in surfactant. In addition, there are about ϳ10% neutral lipids in pulmonary surfactant. Although extensive characterization of phospholipids has been performed in maintaining surfactant function and homeostasis, the functional roles of neutral lipids in the lung are less clear. It has been documented that many neutral lipid metabolites are the ligands for nuclear receptors that are potent transcription factors controlling gene regulation.
In neutral lipids, cholesteryl ester and triglycerides can be hydrolyzed by lysosomal acid lipase in the lysosome of cells to generate free cholesterol and free fatty acids (1). Free cholesterol and free fatty acid derivative compounds, hydroxyeicosatetraenoic acids, hydroxyoctadecanoic acids (HODEs), and 15deoxy-⌬ 12,14 -prostaglandin J 2 (15d-PGJ 2 ), are ligands for peroxisome proliferator-activated receptor ␥ (PPAR␥). PPAR␥ belongs to the nuclear receptor superfamily. PPAP␥ is involved in a broad range of physiological functions, including macrophage inflammatory responses (2,3), adipocyte differentiation (4 -6), glucose homeostasis (7), and apoptosis (8). Upon binding to ligands, PPAR␥ interacts with the retinoid X receptor (RXR) to form the PPAR␥⅐RXR dimer that subsequently binds to specific PPAR-responsive elements (PPRE) on target genes and recruits nuclear receptor coactivators. Nuclear receptor coactivators possess intrinsic histone acetyltransferase activities to activate gene expression. PPAR␥ functions as both a positive and a negative regulator for target genes. Since AT II epithelial cells are responsible for synthesizing surfactant and highly lipogenic, it is important to know how gene expression of surfactant proteins in these cells is regulated by neutral lipid metabolic pathways.
Among the surfactant proteins, SP-B is a 79-amino acid amphipathic peptide that is synthesized and produced in Clara cells and AT II epithelial cells. SP-B is secreted from AT II epithelial cells along with phospholipids into the alveolar lumen to form the surfactant. SP-B facilitates lamellar body formation in AT II epithelial cells and phospholipid spreading during the respiratory cycles (9). Null mutations in the SP-B gene cause lethal respiratory distress in newborn infants and in SP-B-deficient mice produced by gene targeting (10,11). Therefore, SP-B is essential for alveolar maturation and postnatal respiratory adaptation in newborns. Regulation of SP-B homeostasis is important to maintain the surfactant membrane structure and normal lung functions. Transcriptional regulation of gene expression is an important aspect of SP-B homeostasis. Elucidation of mechanisms by which SP-B gene transcription is controlled by positive and negative transcription factors and signaling molecules is critical to understand SP-B synthesis and physiological functions in the lung.
In this report, we assessed the effect of fatty acid metabolite 15d-PGJ 2 on SP-B gene expression in respiratory epithelial cells and whole lung explants. The role of the downstream mediator of this metabolite, PPAR␥⅐RXR␣, in regulating SP-B gene transcription was also studied. The studies support the concept that PPAR␥⅐RXR␣ and its ligands serve as negative regulators for SP-B gene expression in respiratory epithelial cells. These studies add a new insight into the mechanism whereby SP-B gene expression and homeostasis are regulated by neutral lipid metabolites and nuclear receptors in the lung.

MATERIALS AND METHODS
Plasmids and Transgenic Mice-The mammalian expression vectors for wild type PPAR␥ and ACTR were kind gifts from Dr. Ron Evans (the Salk Institute). The mammalian expression vector for dominant negative PPAR␥ was a kind gift from Dr. Krishna K Chatterjee (Cambridge, UK). The SRC-1 vector was kindly provided by Dr. B. O'Malley. The RXR␥ and TIF2 expression vectors were kindly provided by Dr. P. Chambon. The hSP-B 500 and hSP-B 218 luciferase reporter constructs were made previously (12). The hSP-B 150 luciferase report construct was made by following a procedure previously described (12). The hSP-B 1.5-kb lacZ gene transgenic mouse line was generated previously (13).
H441 Cell Maintenance-Human pulmonary adenocarcinoma H441 cells were originally obtained from American Type Culture Collection (ATCC) and cultured in RPMI supplemented with 10% fetal calf serum, glutamine, and penicillin/streptomycin. Cells were maintained and passaged weekly at 37°C in 5% CO 2 /air.
RT-PCR-H441 cells were seeded and treated with 15d-PGJ 2 (Cayman Chemical Co., Ann Arbor, MI) at a final concentration of 10 M overnight. Total RNAs were isolated from cells using the RNA purification kit (Qiagen Co.), and RNA concentrations were determined. One g of total RNAs from each sample was subject to RT-PCR assay using a pair of primers corresponding to the SP-B coding region and the SuperScript One-Step RT-PCR kit (Invitrogen) as described previously (14). A pair of primers corresponding to the glyceraldehyde-3-phosphate dehydrogenase coding region was used as a control. PCR signal intensities were determined by gel electrophoresis.
H441 Cell Transient Transfection and Luciferase Assay-H441 cells were seeded at densities of 2 ϫ 10 5 cells/well in six-well plates. For ligand study, various hSP-B luciferase reporter gene constructs (0.25 g) and pCMV-␤gal plasmid (0.5 g) were cotransfected into H441 cells by FuGene6 (Roche Applied Science). H441 cells were treated with 10 M 15d-PGJ 2 (Cayman Chemical Co.) on the following day. After 2 days of incubation to allow reporter protein expression, cells were lysed, and luciferase activities were determined using the luciferase assay system (Promega). For PPAR␥⅐RXR␣ transfection study, various concentrations of PPAR␥⅐RXR␣ expression vectors were co-transfected with the hSP-B 150-bp luciferase reporter gene construct (0.5 g) and 0.5 g of pCMV-␤gal construct into H441 cells. On the next day, cells were treated with 10 M 15d-PGJ 2 . After a 48-h incubation, cells were lysed, and luciferase activities were determined. For the nuclear receptor coactivator transfection study, various concentrations of SRC-1, ACTR, and TIF2 expression vectors were co-transfected with the hSP-B 150-bp luciferase reporter gene construct (0.5 g) and 0.5 g of pCMV-␤gal construct into H441 cells. On the next day, cells were treated with 10 M 15d-PGJ 2 . After a 48-h incubation, cells were lysed, and luciferase activities were determined. In each transfection study above, ␤-galactosidase activities were determined for normalization of transfection efficiency.
AT II Epithelial Cell Culturing and ␤-Galactosidase Activity Measurement-As previously described (15), AT II epithelial cells were isolated from 6-week-old hSP-B 1.5-kb lacZ transgenic FVB/N mice (13) and cultured on a Matrigel matrix/rat tail collagen (70:30, v/v) in bronchial epithelial cell growth medium minus hydrocortisone plus 5% charcoal-stripped fetal bovine serum and 10 ng/ml keratinocyte growth factor. Cells were treated with or without 10 M 15d-PGJ 2 or (ϩ)-9-HODE (Cayman Chemical Co.) for 2 days to allow reporter protein expression. Cells were lysed, and ϳ1 ng of protein was used for the ␤-galactosidase assay.
Lung Explant Culture and ␤-Galactosidase Staining-Lung buds were dissected out from embryonic day 11 embryos of nontransgenic or hSP-B 1.5-kb lacZ gene transgenic mice following a procedure described previously (13). The lung buds were cultured on top of 1% low melting point agarose gel in Dulbecco's modified Eagle's medium/F-12 medium with supplementation of 10% fetal calf serum in a 30-mm dish. The next day, a final concentration of 0.2 mg/ml X-gal with or without 10 M 15d-PGJ 2 was added to the culture medium and incubated for an additional 2 days. The medium containing X-gal and 15d-PGJ 2 for cultured lung explants was changed each day.
Immunohistochemistry-The lungs from wild type FVB/N adult mice were infused with a fixative solution (4% paraformaldehyde, 1ϫ phosphate-buffered saline, or PBS), removed from the chest, and stored in fixative at 4°C for ϳ24 h. After fixation, dehydration, and embedding in paraffin, lung tissue sections were cut to 5 m thick. The adult lung slides were baked at 60°C for 2 h and immersed in xylene and ethanol to remove paraffin from the tissues. After rehydration, an antigen retrieval procedure was performed, and endogenous peroxidase activity was removed from the adult mouse lung tissues by incubating the tissue slides in methanol and hydrogen peroxide for 15 min. Tissue slides were incubated overnight at 4°C with primary PPAR␥ antibody (1:500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), TTF1 antibody (1:10,000; from Dr. R. Di Lauro, Naples, Italy), and FLAG antibody (1:4,000; Sigma). The tissues were washed and treated with biotinylated secondary antibodies. The interactions were detected with a Vectastain Elite ABC kit to visualize the signals following a procedure recommended by the manufacturer.
EMSA-The fragment of the hSP-B Ϫ150 to ϩ41 bp promoter region was amplified by PCR using the hSP-B 218 luciferase reporter gene construct as a template. The PPRE response element (GTCGACAGGG-GACCAGGACAAAGGTCACGTTCGGGAGTCGAC) from the acyl-CoA oxidase gene (16) was synthesized, annealed, and purified as a positive control. To make the PPAR␥-GST fusion protein, the full-length PPAR␥ was subcloned into the pGEX4T-1 GST vector (Amersham Biosciences) at the SalI and NotI restriction enzyme sites by PCR. The plasmids were transformed into BL21 bacterial strains for protein expression. After 3 h of induction at 37°C by 1 mM isopropyl ␤-D-thiogalactoside, the bacteria were harvested and resuspended in 1ϫ PBS, followed by sonication and treatment with 1% Triton X-100. The proteins were purified by incubation with 50% slurry of glutathione-Sepharose 4B beads (Amersham Biosciences) for 30 min at room temperature and then eluted from beads using glutathione (GS) elution buffer followed by dialysis. Protein expression was confirmed by Coomassie Brilliant Blue staining and Western blot using PPAR␥ antibody (Amersham Biosciences) after gel electrophoresis. Protein concentrations were determined by the bicinchoninic acid protein assay kit (Sigma). The RXR␣ protein was made by the TNT in vitro transcription/translation kit (Promega) as recommended by the manufacturer. The oligonucleotides were radiolabeled by [␥-32 P]ATP and kinase and incubated with 1 g of the purified PPAR␥-GST fusion protein and 1 l of in vitro transcribed and translated RXR␣. EMSA was performed by following the procedure described previously (17).

Inhibition of SP-B mRNA Expression by 15d-PGJ 2 in H441
Cells-To assess how endogenous SP-B gene expression is affected by neutral lipid metabolites, monolayers of H441 cells were treated with or without 10 M 15d-PGJ 2 overnight. Total RNAs were isolated from these cells. A pair of hSP-B cDNA-specific primers was used for RT-PCR assessment of hSP-B mRNA expression. As shown in Fig. 1, hSP-B mRNA expression was reduced by 15d-PGJ 2 treatment in H441 cells. As a control, a pair of glyceraldehyde-3-phosphate dehydrogenase cDNA-specific primers was also used for RT-PCR. No significant change was observed in glyceraldehyde-3-phosphate Total RNAs were isolated from H441 cells treated or untreated with 10 M 15d-PGJ 2 overnight. One g of total RNAs from each sample was subject to an RT-PCR assay using a pair of primers corresponding to the SP-B coding region. A pair of primers corresponding to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) coding region was used as a control. PCR signal intensities were determined by gel electrophoresis. The experiment was done in triplicate. dehydrogenase mRNA expression. Since the SP-B mRNA expression level can be affected at the transcriptional level or at the posttranscriptional level (mRNA stability and nuclear transportation), further studies are necessary to determine whether this inhibitory effect is a result of hSP-B transcriptional repression.
Inhibition of the hSP-B 1.5-kb lacZ Gene by 15d-PGJ 2 and 9-HODE in Primary AT II Epithelial Cells-Previously, a transgenic mouse system carrying an hSP-B 1.5-kb 5Ј-flanking regulatory region and the lacZ gene was created. In this in vivo system, the hSP-B 1.5-kb lacZ gene recapitulates the endogenous SP-B gene expression pattern in the lung in a highly tissue-and cell type-specific manner (13). Since the hSP-B 1.5-kb 5Ј-flanking regulatory sequence resides in AT II cells and Clara cells in this transgenic mouse line, it is an ideal system to study the hSP-B transcriptional activity in vivo. To assess whether neutral lipid metabolites affect the hSP-B 1.5-kb transcriptional activity in primary AT II epithelial cells, adult AT II epithelial cells were isolated from hSP-B 1.5-kb lacZ gene transgenic mice and cultured in vitro. The next day, primary AT II epithelial cell monolayers were treated with or without a 10 M concentration of either 15d-PGJ 2 or 9-HODE. The nontransgenic AT II epithelial cells were also isolated and cultured in vitro as a control. After 4 days of incubation with 9-HODE, AT II epithelial cells were lysed and assayed for ␤-galactosidase activities. There was no significant difference between 15d-PGJ 2 -or 9-HODE-treated and nontreated AT II epithelial cells from nontransgenic mice. In contrast, a significant decrease in ␤-galactosidase activity was observed in 15d-PGJ 2 -or 9-HODE-treated AT II epithelial cells that were isolated from hSP-B 1.5-kb lacZ gene transgenic mice (Fig. 2, A  and B). This result indicates that PPAR␥ ligands inhibit hSP-B gene expression at the promoter transcriptional level in AT II epithelial cells.
Inhibition of the hSP-B 1.5-kb lacZ Gene during Lung Branching Morphogenesis by 15d-PGJ 2 -Since expression of the hSP-B 1.5-kb lacZ gene is developmentally controlled as previously reported (13), the effect of neutral lipid metabolite 15d-PGJ 2 on hSP-B 1.5-kb lacZ gene transcription was assessed in lung branching morphogenesis. Intact fetal lungs were isolated from hSP-B 1.5-kb lacZ transgenic mice at 11 days of gestation and cultured in vitro in the presence of X-gal. Nontransgenic fetal lungs were also isolated and cultured as a control. On the next day, lung explants were treated with or without 10 M 15d-PGJ 2 for 2 days. In nontransgenic control, no lacZ gene expression was detected in lung explants (Fig. 3). In hSP-B 1.5-kb lacZ gene transgenic mice, the in vitro cultured lung explant showed robust expression of the hSP-B 1.5-kb lacZ gene in newly formed epithelial tubules (Fig. 3). In contrast, hSP-B gene expression in branching epithelial tubules was significantly inhibited in the presence of 15d-PGJ 2 . This study indicates that 15d-PGJ 2 can inhibit the hSP-B transcriptional activity during the developmental branching process. Suppression of SP-B gene expression seems not to affect the whole lung branching structure in this study. This is in agreement with the previous finding that ablation of the SP-B gene in mice does not affect embryonic lung development (10).
Inhibition of the hSP-B Gene by 15d-PGJ 2 Is Mediated within the Ϫ218 to ϩ41 Region-To define the cis-acting region that mediates 15d-PGJ 2 inhibition on the hSP-B promoter, shorter hSP-B 500 and hSP-B 218 luciferase reporter constructs were tested in H441 cells by transient transfection and luciferase assay. The H441 cell line retains many characteristics of Clara cells. After 48 h of cell incubation with various concentrations of 15d-PGJ 2, a significant decrease of both hSP-B 500 and hSP-B 218 luciferase activities was observed in H441 cells in a dose-dependent fashion (Fig. 4). In addition, a  3. Inhibition of the hSP-B 1.5-kb lacZ gene by 15d-PGJ 2 in lung explants. Fetal lungs were isolated from nontransgenic (Non-Tg) and hSP-B 1.5-kb lacZ transgenic (Tg) embryos (embryonic day 11; E11) and cultured in vitro on top of 1% low melting point agarose gel in Dulbecco's modified Eagle's medium/F-12 medium with supplementation of 10% fetal calf serum. Next day, a final concentration of 0.2 mg/ml X-gal with or without 10 M 15d-PGJ 2 was added to the culture medium and incubated for an additional 2 days. Lung explants isolated from hSP-B 1.5-kb lacZ transgenic mice without ligand treatment (Ϫ) served as a positive control. Lung explants isolated from nontransgenic mice served as a negative control. Shown is a representative demonstration from multiple lung bud studies. chimeric construct that contains the hSP-B-(Ϫ500/Ϫ331) enhancer region and the SV40 basic promoter was also studied. No 15d-PGJ 2 inhibitory effect was observed on this construct. Instead, a moderate increase of luciferase activity was observed at lower 15d-PGJ 2 concentrations. This hSP-B-(Ϫ500/Ϫ331) enhancer region mediates RAR and nuclear receptor coactivator stimulation on hSP-B gene expression (18,19). As a control, the SV40 basic promoter construct (PGL2P) showed no change in response to 15d-PGJ 2 . It is unlikely that the hSP-B-(Ϫ500/ Ϫ331) enhancer region mediates 15d-PGJ 2 inhibition of the hSP-B gene. Previously, the hSP-B-(Ϫ500/Ϫ331) enhancer region has been identified to mediate RA/RAR signaling stimulation of the hSP-B gene (18). Taken together, 15d-PGJ 2 inhibition of the hSP-B gene is mediated through the Ϫ218 to ϩ41 region.
PPAR␥⅐RXR␣ Inhibits the hSP-B Gene Transcription-The compound 15d-PGJ 2 exhibits both PPAR␥-dependent and independent mechanisms. To test whether PPAR␥ mediates the 15d-PGJ 2 inhibitory effect on the hSP-B gene expression, a PPAR␥ and hSP-B luciferase reporter gene cotransfection study was performed in H441 cells. Monolayers of H441 cells were transiently transfected with the hSP-B 218 or hSP-B 150 luciferase reporter gene constructs. After subsequent 48-h cell incubation with 5 M 15d-PGJ 2 , a significant decrease of both hSP-B 218 and hSP-B 150 luciferase activities was observed in H441 cells (Fig 5A). When hSP-B 218 or hSP-B 150 luciferase reporter gene constructs were co-transfected with various concentrations of PPAR␥⅐RXR␣, further inhibition was observed. This is a strong indication that PPAR␥⅐RXR␣ mediates 15d-PGJ 2 inhibition. To confirm this observation, a mutant PPAR␥ was chosen for the cotransfection assay. In the mutant, Leu 468 and Glu 471 in helix 12 of the ligand binding domain of PPAR␥ are mutated to alanine. The mutant retains ligand and DNA binding activities but exhibits no co-activator recruitment activity (20). When the hSP-B 150 luciferase reporter construct was cotransfected with increasing amounts of mutant PPAR␥ into H441 cells, no inhibitory effect was observed, Fig. 5B. Instead, a modest increase of the hSP-B 150 luciferase reporter gene activity was observed. This reversal of repression is probably due to the interference of endogenous PPAR␥ inhibitory effects by the mutant PPAR␥ protein.

PPAR␥ Expression in Mouse AT II Epithelial Cells and Clara
Cells-Since PPAR␥ can functionally inhibit SP-B gene expression, immunohistochemical staining was performed to test whether PPAR␥ is localized in AT II epithelial cells and Clara cells where SP-B is expressed and processed. Sections from adult mouse lungs were prepared and stained using an antibody against PPAR␥. Sections were also stained with an antibody against thyroid transcription factor 1 (TTF1) as a positive control. TTF1 is a tissue-and cell type-specific marker whose expression is specifically restricted to AT II and Clara cells in the lung. Staining for TTF1 and PPAR␥ antibodies in mouse lungs was detected in bronchiolar and AT II epithelial cells where SP-B is synthesized (Fig. 6) (13). In a negative control, no specific staining was detected by a FLAG antibody. Previously, RXR␣ expression was detected in AT II epithelial cells and Clara cells by immunohistochemistry (19). Therefore, PPAR␥⅐RXR␣ not only functionally influences SP-B gene expression but also is colocalized physically with SP-B in the lung.
PPAR␥⅐RXR␣ Does Not Directly Bind to the hSP-B Ϫ150 to ϩ41 Region-One mechanism for nuclear receptors to exert negative effect on gene transcription is to bind to cis-acting sites within the promoter region of the target genes and recruit corepressor complexes (21). To test whether the PPAR␥⅐RXR␣ inhibitory effect is mediated by direct DNA binding to the hSP-B-(Ϫ150 to ϩ41) promoter region, the PPAR␥-GST fusion protein was expressed in bacteria and purified by chromatography. The RXR␣ protein was prepared by in vitro transcription and translation. Two proteins were incubated with the 32 P-radiolabled hSP-B Ϫ150 to ϩ41 bp fragment for EMSA in the presence of 15d-PGJ 2 . As shown in Fig. 7, the PPAR␥-GST⅐RXR␣ complex did not form a protein-DNA complex with the hSP-B Ϫ150 to ϩ41 promoter region. A homology search did not reveal the consensus sequence for PPRE in this region. In a positive control study, PPAR␥-GST⅐RXR␣ formed a complex with a PPRE from the acyl-CoA oxidase gene, known to bind to PPAR␥⅐RXR␣. Therefore, the inhibitory effect of PPAR␥ and its ligands on the hSP-B gene is not mediated through direct DNA binding to the promoter region.
Reversal of the 15d-PGJ 2 Inhibition by Nuclear Receptor Coactivator TIF2-It has also been reported that PPAR␥ inhib- its gene transcription through the trans-repression mechanism by competing with a limited amount of the nuclear receptor coactivator pool in the cell (22). If PPAR␥ and its ligands inhibit the hSP-B transcription through the trans-repression mechanism, increasing the amount of nuclear receptor co-activators should reverse the inhibitory effect. To test this possibility, p160 family nuclear receptor co-activators, including SRC-1, ACTR, and TIF2, were co-transfected with the hSP-B 150 luciferase reporter construct into H441 cells to increase nuclear receptor co-activator concentrations. Without nuclear receptor co-activator cotransfection, 5 M 15d-PGJ 2 significantly decreased the hSP-B 150 luciferase activity in H441 cells (Fig. 8). When increasing amounts of TIF2 were cotransfected into H441 cells, the inhibitory effect of the hSP-B 150 luciferase reporter activity was reversed. The reversal effect was dose-de-  7. EMSA of PPAR␥⅐RXR␣ to hSP-B ؊150 to ؉41 DNA region. The synthetic PPRE from the acyl-CoA oxidase gene or the PCR-amplified hSP-B Ϫ150 to ϩ41 fragment was radiolabeled and incubated with 1 g of purified PPAR␥-GST and 1 l of TNT in vitro transcribed and translated RXR␣ proteins. Free probes and DNA-protein complexes were separated on 4% nondenaturing polyacrylamide gels followed by autoradiography. pendent (Fig. 8). On the other hand, SRC-1 and ACTR did not show the reversal of inhibitory effect on the hSP-B 150 luciferase reporter gene. This indicates that among p160 nuclear receptor co-activators, only TIF2 plays a role in the hSP-B Ϫ150 to ϩ41 promoter region to activate gene transcription, which can be competed off by overexpression of PPAR␥.

DISCUSSION
Although there are only about 10% neutral lipids in pulmonary surfactant, their metabolites can serve as ligands to bind to various nuclear receptors to regulate surfactant protein gene expression. Prostaglandins are derived from fatty acids (primarily from arachidonate) through a series of biochemical conversions in vivo. In this report, we investigated the functional role of 15d-PGJ 2 that serves as a ligand for PPAR␥ in SP-B gene expression in respiratory epithelial cells. Ligand 15d-PGJ 2 inhibited endogenous SP-B mRNA synthesis in H441 cells (Fig. 1), indicating that the inhibitory effect can happen as early as 18 h. This apparently occurs at the gene transcriptional level, because expression of the lacZ reporter gene under the control of the hSP-B 1.5-kb 5Ј-flanking regulatory region was also inhibited by 15d-PGJ 2 in primary AT II epithelial cells (Fig. 2). Using hSP-B promoter deletion constructs, the inhibitory effect was narrowed down to the hSP-B Ϫ150 to ϩ41 bp promoter region (Fig. 5A). This region is highly conserved in both human and murine species (23). PPAR␥⅐RXR␣ that mediates the 15d-PGJ 2 action also inhibited hSP-B promoter transcription (Fig. 5, A and B). Endogenous expression of PPAR␥ was detected in both AT II and Clara cells in the mouse by immunohistochemical staining (Fig. 6). These results collectively suggest that PPAR␥ is colocalized with SP-B in AT II epithelial cells and Clara cells and functionally influences SP-B gene expression.
Nuclear receptors exert negative regulation on gene regulation through various mechanisms. Some nuclear receptors re-press transcription by binding to cis-acting sites within the promoter region of the target genes by recruiting corepressor complexes (21). In an EMSA study, the PPAR␥⅐RXR␣ complex failed to bind to the hSP-B Ϫ150 to ϩ41 promoter region that mediates the 15d-PGJ 2 /PPAR␥ inhibitory effect. Therefore, it is unlikely that this is the mechanism for PPAR␥⅐RXR␣ and its ligands to inhibit hSP-B promoter transcription. A second mechanism involves nuclear receptor (e.g. PPAR␥) transrepression through coactivator competition (22). Since there are limiting amounts of coactivators (CREB-binding protein/p300 and p160 coactivators including SRC-1, ACTR, and TIF2) in a given cell, nuclear receptors mutually antagonize each other by competing with these coactivators. In this mechanism, inhibition usually can be reversed by increasing concentrations of nuclear receptor coactivators in the cell. Indeed, cotransfection of nuclear receptor coactivator TIF2 caused the reversal of hSP-B gene inhibition by 15d-PGJ 2 (Fig. 8). This was confirmed by the observation that a PPAR␥ mutant lacking the nuclear receptor coactivator recruiting ability reversed PPAR␥ inhibitory activity in H441 cells (Fig. 5B). The study suggests that there may be another unidentified transcription factor (e.g. nuclear receptor) in hSP-B Ϫ150 to ϩ41 promoter region, which can recruit TIF2. In a similar situation, the sequestering of CREB-binding protein and SRC-1 by activated PPAR␥ accounts for transrepression of the inducible nitric-oxide synthase promoter in response to ligand binding (22). In this system, transrepression requires PPAR␥ interactions with LXXLL-containing coactivators through functional domains. It has been shown that a deletion of the ligand-dependent activation domain AF2 of PPAR␥ and point mutations of the critical charge clamp residues (PPAR␥ EA469 and KG301) weakened interaction with SRC-1 and CREB-binding protein. This deletion and point mutations led to complete loss of transrepression function (22). PPAR␥ transrepression has also been reported in cyclooxygenase-2 gene regulation (24).
SP-B gene expression is also subject to positive regulation by certain members of nuclear receptors. Previously, we have shown that RA and RAR transactivate SP-B gene transcription by recruiting nuclear receptor coactivators through an enhancer located in the hSP-B Ϫ500 to Ϫ331 region upon retinoic binding (18,19). The formation of the transcriptional complex is dependent on RAR⅐RXR, TTF1, and nuclear receptor coactivators (18). Deletion of this highly conserved enhancer region abolished hSP-B gene expression in Clara cells and reduced its expression in AT II epithelial cells in transgenic mice (13). When a dominant negative RAR␣ was overexpressed in H441 The hSP-B Ϫ500 to Ϫ331 enhancer region binds to enhanceosome containing TTF-1, RAR⅐RXR, and nuclear receptor coactivators. The hSP-B Ϫ150 to ϩ41 region binds to TTF1 and HNF3. PPAR␥⅐RXR␣ and ligands exert their inhibitory effect on the hSP-B Ϫ150 to ϩ41 region but not on the hSP-B Ϫ500 to Ϫ331 region. The question mark represents an unidentified transcription factor that recruits TIF2. cells, it strongly suppressed hSP-B transcription (25). In a doxycycline-inducible transgenic mouse system, overexpression of dominant negative RAR␣ in respiratory epithelial cells caused abnormal alveolar formation in neonatal lungs, probably due to suppression of SP-B and other functional genes (25). To determine whether PPAR␥ interferes with RA/RAR transactivation in the hSP-B Ϫ500 to Ϫ331 enhancer region by coactivator sequestration, a chimeric hSP-B-(Ϫ500/Ϫ331)/ SV40 promoter was treated with 15d-PGJ 2 in H441 cells. Interestingly, there was no transrepression observed (Fig. 4). Although the detailed mechanism is not clear at this moment, it seems that PPAR␥ and its ligands selectively compete with the nuclear receptor coactivator recruiting process in the hSP-B Ϫ150 to ϩ41 promoter region but not with that in the hSP-B Ϫ500 to Ϫ331 enhancer region (Fig. 9). This selectivity must be determined by surrounding transcription factors and DNA cis-acting elements in the hSP-B 5Ј-flanking regulatory region. Whereas all p160 nuclear receptor coactivators stimulate hSP-B (Ϫ500 to Ϫ331)/SV40 luciferase reporter gene transcription (18), only TIF2 stimulated hSP-B 150 luciferase reporter gene transcription (data not shown). In agreement with this observation, only an increase of TIF2 concentration in H441 cells reversed the 15d-PGJ 2 inhibitory effect of the hSP-B 150 luciferase reporter gene (Fig. 8).
PPAR␥ ligands also affect hSP-B temporal/spatial expression during lung branching morphogenesis. In lung explant in vitro culturing study (Fig. 3), hSP-B 1.5-kb lacZ gene expression was dramatically inhibited by PPAR␥ ligand treatment, implicating that the unidentified TIF2 recruiting factor in the hSP-B Ϫ150 to ϩ41 promoter region is required for SP-B gene expression during lung development in newly formed epithelial tubules. This suggests that during lung development, both positive and negative regulation by nuclear receptors and coactivators are essential for proper SP-B gene expression. Previously, the glucocorticoid receptor has also been reported to inhibit hSP-B transcription (14), although the mechanism is unknown.
PPAR␥ has pleiotrophic effects in various systems. There are studies suggesting that PPAR␥ plays a role in antagonizing proinflammatory response (3, 26 -28). SP-B gene expression is stimulated by proinflammatory cytokines. We previously reported that SP-B gene expression is stimulated by proinflammatory cytokine interleukin-6 family members in the lung. Interleukin-6 family cytokines stimulate downstream target genes by activation of signal transducers and activators of transcription 3 through tyrosine phosphorylation at Tyr 705 . How PPAR␥ and its ligands affect this process of proinflammatory stimulation to keep the balance of SP-B gene expression during pro-and anti-inflammatory responses in the respiratory system is an important issue for future investigation and has potential clinical application.
In summary, neutral lipid metabolites in lipogenic AT II epithelial cells influence surfactant protein B gene expression and homeostasis in the lung. The inhibitory effect of PPAR␥ and ligands on SP-B gene expression represents a novel mech-anism whereby SP-B homeostasis is regulated in respiratory epithelial cells. In pulmonary surfactant, phospholipids are the major structural component, whereas neutral lipid metabolites can serve as sensor signals to balance pulmonary surfactant homeostasis in responding to changes of various physiologic conditions and inflammatory responses in host defenses.