Nuclear Factor I Family Members Regulate the Transcription of Surfactant Protein-C
1997; Elsevier BV; Volume: 272; Issue: 52 Linguagem: Inglês
10.1074/jbc.272.52.32759
ISSN1083-351X
AutoresCindy J. Bachurski, Susan Kelly, Stephan W. Glasser, Tracey A. Currier,
Tópico(s)Congenital Diaphragmatic Hernia Studies
ResumoTranscription of the surfactant protein-C (SP-C) gene is restricted to Type II epithelial cells in the adult lung. We have shown previously that the 0.32-kilobase pair (kb) mouse SP-C promoter is functional in transient transfection assays of the lung epithelial cell-derived cell line, MLE-15, and that thyroid transcription factor 1 (TTF-1) transactivates promoter activity. The 0.32-kb SP-C promoter can be separated into a proximal promoter region (−230 to +18) and an enhancer region (−318 to −230). Three DNase I footprints were mapped in the promoter region (C1 through C3) and two in the enhancer region (C4 and C5). We now show that nuclear factor I (NFI) family members bind to both individual NFI half-sites in footprints C1, C3, and C5, and to a composite site in footprint C4 by competition gel retardation and antibody supershift analyses. Mutational analysis of the 0.32-kb mouse SP-C promoter and transient transfection of MLE-15 cells demonstrated that the NFI binding sites are required for promoter activity in this cell type. Site-specific mutation of the proximal or distal NFI sites drastically reduced transactivation by a co-transfected NFI-A expression vector in HeLa cells. These data indicate that NFI family member(s), binding to sites in both the promoter and enhancer regions, regulate SP-C gene expression in a process independent of TTF-1. Transcription of the surfactant protein-C (SP-C) gene is restricted to Type II epithelial cells in the adult lung. We have shown previously that the 0.32-kilobase pair (kb) mouse SP-C promoter is functional in transient transfection assays of the lung epithelial cell-derived cell line, MLE-15, and that thyroid transcription factor 1 (TTF-1) transactivates promoter activity. The 0.32-kb SP-C promoter can be separated into a proximal promoter region (−230 to +18) and an enhancer region (−318 to −230). Three DNase I footprints were mapped in the promoter region (C1 through C3) and two in the enhancer region (C4 and C5). We now show that nuclear factor I (NFI) family members bind to both individual NFI half-sites in footprints C1, C3, and C5, and to a composite site in footprint C4 by competition gel retardation and antibody supershift analyses. Mutational analysis of the 0.32-kb mouse SP-C promoter and transient transfection of MLE-15 cells demonstrated that the NFI binding sites are required for promoter activity in this cell type. Site-specific mutation of the proximal or distal NFI sites drastically reduced transactivation by a co-transfected NFI-A expression vector in HeLa cells. These data indicate that NFI family member(s), binding to sites in both the promoter and enhancer regions, regulate SP-C gene expression in a process independent of TTF-1. Cell-selective transcription of genes during development and modulation of gene expression in response to extracellular signals is mediated by the regulated assembly of specific protein-DNA complexes on promoter and enhancer or silencer sequences. Transcription of the surfactant protein-C (SP-C) 1The abbreviations used are: SP-C, surfactant protein C; TNF-α, tumor necrosis factor-α; TTF-1, thyroid transcription factor 1; NFI, nuclear factor I; EMSA, electrophoretic mobility shift assay. gene is restricted to pulmonary epithelial cells and regulated by developmental and humoral influences (1Beers M.F. Fisher A.B. Am. J. Physiol. 1992; 263: L151-L160PubMed Google Scholar). SP-C is a hydrophobic protein that enhances the spreading and stability of surfactant phospholipids at the air liquid interface during the respiratory cycle (2Weaver T.E. Whitsett J.A. Biochem. J. 1991; 273: 249-264Crossref PubMed Scopus (364) Google Scholar). Expression of the SP-C gene is restricted to the distal respiratory epithelium during branching morphogenesis in the fetal lung, and to alveolar Type II cells postnatally (3Khoor A. Stahlman M.T. Gray M.E. Whitsett J.A. J. Histochem. Cytochem. 1994; 42: 1187-1199Crossref PubMed Scopus (131) Google Scholar, 4Zhou L. Lim L. Costa R.H. Whitsett J.A. J. Histochem. Cytochem. 1996; 44: 1183-1193Crossref PubMed Scopus (248) Google Scholar). The promoter regions of the human and mouse SP-C genes have been cloned and used to direct lung epithelial cell-specific expression in vitro and in transgenic mice in vivo (Refs. 5Bachurski C.J. Pryhuber G.S. Glasser S.W. Kelly S.E. Whitsett J.A. J. Biol. Chem. 1995; 270: 19402-19407Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar and 6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar; reviewed in Ref. 7Glasser S.W. Korfhagen T.R. Wert S.E. Whitsett J.A. Am. J. Physiol. 1994; 267: L489-L497PubMed Google Scholar). SP-C gene expression is decreased in acute lung injury and is increased in areas of respiratory epithelial regeneration (8Coalson J.J. King R.J. Yang F. Winter V. Whitsett J.A. Delemos R.A. Seidner S.R. Am. J. Respir. Crit. Care Med. 1995; 151: 854-866Crossref PubMed Google Scholar, 9Stahlman M.T. Gray M.E. Whitsett J.A. J. Histochem. Cytochem. 1996; 44: 673-678Crossref PubMed Scopus (194) Google Scholar, 10Zsengellèr Z.K. Wert S.E. Bachurski C.J. Kirwin K.L. Trapnell B.C. Whitsett J.A. Hum. Gene Ther. 1997; 8: 1331-1334Crossref PubMed Scopus (28) Google Scholar). Tumor necrosis factor α (TNF-α), inhibits transcription from both the human and mouse SP-C promoters (5Bachurski C.J. Pryhuber G.S. Glasser S.W. Kelly S.E. Whitsett J.A. J. Biol. Chem. 1995; 270: 19402-19407Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar). Sequences from the mouse SP-C promoter (−318 to +18) are sufficient to confer TNF-α regulation on a transfected marker gene (5Bachurski C.J. Pryhuber G.S. Glasser S.W. Kelly S.E. Whitsett J.A. J. Biol. Chem. 1995; 270: 19402-19407Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar), but the mechanism of this regulation is unknown. To dissect the mechanisms regulating SP-C gene transcription, the mouse SP-C promoter was subdivided into a proximal promoter region (−230 to +18 relative to the start of transcription), and an enhancer region (−318 to −230) (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar). DNase I footprint analysis of the 0.32-kb mouse SP-C promoter mapped five areas of protein-DNA interaction (C1 through C5). The homeodomain transcription factor, thyroid transcription factor-1 (TTF-1), activates SP-C transcription through binding sites in the C2 footprint in the proximal promoter region (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar). TTF-1 activates transcription of several other surfactant-associated genes expressed in overlapping subsets of pulmonary epithelial cells including surfactant protein-A (11Bruno M.D. Bohinski R.J. Huelsman K.M. Whitsett J.A. Korfhagen T.R. J. Biol. Chem. 1995; 270: 6531-6536Abstract Full Text Full Text PDF PubMed Scopus (240) Google Scholar), surfactant protein-B, and Clara cell secretory protein (12Bohinski R.J. DiLauro R. Whitsett J.A. Mol. Cell. Biol. 1994; 14: 5671-5681Crossref PubMed Scopus (486) Google Scholar). The identity of the DNA-binding proteins interacting with footprints C1, C3, C4, and C5 of the mouse SP-C promoter has not been reported. Comparison of these regions with a transcription factor binding site data base revealed several potential binding sites for nuclear factor I (NFI). The transcription factor NFI is encoded by a family of four genes (Nfia, Nfib, Nfic, and Nfix). NFI binds with high affinity to the palindromic sequence TGGC(N)5GCCAA (13Borgmeyer U. Nowock J. Sippel A.E. Nucleic Acids Res. 1984; 12: 4295-4311Crossref PubMed Scopus (104) Google Scholar, 14Gronostajski R.M. Nucleic Acids Res. 1987; 15: 5545-5559Crossref PubMed Scopus (75) Google Scholar) and with lower affinity to the half-site TGGCA (15Gounari F. De Francesco R. Schmitt J. van der Vilet P.C. Cortese R. Stunnenberg H. EMBO J. 1990; 9: 559-566Crossref PubMed Scopus (120) Google Scholar, 16Paonessa G. Gounari F. Frank R. Cortese R. EMBO J. 1988; 7: 3115-3123Crossref PubMed Scopus (173) Google Scholar). NFI family members bind to DNA as homo- or heterodimers (15Gounari F. De Francesco R. Schmitt J. van der Vilet P.C. Cortese R. Stunnenberg H. EMBO J. 1990; 9: 559-566Crossref PubMed Scopus (120) Google Scholar, 17Kruse U. Sippel A.E. FEBS Lett. 1994; 348: 46-50Crossref PubMed Scopus (106) Google Scholar). The NFI proteins are highly homologous in the amino-terminal DNA binding and dimerization domain but diverge in the carboxyl-terminal transactivation domain (18Rupp R.A.W. Kruse U. Multhaup G. Göbel U. Beyreuther K. Sippel A.E. Nucleic Acids Res. 1990; 18: 2607-2616Crossref PubMed Scopus (150) Google Scholar, 19Santoro C. Mermod N. Andrews P.C. Tjian R. Nature. 1988; 334: 218-224Crossref PubMed Scopus (492) Google Scholar). Alternative splicing of the NFI genes within the proline-rich transactivation domain adds further diversity to this family, possibly affecting the regulatory properties of individual NFI isoforms (20Mermod N. O'Neill E.A. Kelly T.J. Tjian R. Cell. 1989; 58: 741-753Abstract Full Text PDF PubMed Scopus (542) Google Scholar, 21Kruse U. Sippel A.E. J. Mol. Biol. 1994; 238: 860-865Crossref PubMed Scopus (77) Google Scholar). During mouse embryogenesis, NFI family members have unique, albeit overlapping patterns of expression (22Chaudhry A.Z. Lyons G.E. Gronostajski R.M. Dev. Dyn. 1997; 208: 313-325Crossref PubMed Scopus (179) Google Scholar). NFI has been shown to regulate several cell-selective genes including liver-specific (albumin (23Cereghini S. Raymondjean M. Carranca A.G. Herbomel P. Yaniv M. Cell. 1987; 50: 627-638Abstract Full Text PDF PubMed Scopus (215) Google Scholar, 24Jackson D.A. Rowander K.E. Stevens K. Jiang C. Milos P. Zaret K.S. Mol. Cell. Biol. 1993; 13: 2401-2410Crossref PubMed Scopus (134) Google Scholar), retinol-binding protein (25Eskild W. Simard J. Hansson V. Guérin S.L. Mol. Endocrinol. 1994; 8: 732-745PubMed Google Scholar), vitellogenin (26Cardinaux J.-R. Chapel S. Wahli W. J. Biol. Chem. 1994; 269: 32947-32956Abstract Full Text PDF PubMed Google Scholar), α-fetoprotein (27Bois-Joyeux B. Danan J.-L. Biochem. J. 1994; 301: 49-55Crossref PubMed Scopus (69) Google Scholar)), adipocyte-specific (aP2 (28Graves R.A. Tontonoz P. Ross S.R. Spiegelman B.M. Genes Dev. 1991; 5: 428-437Crossref PubMed Scopus (117) Google Scholar)), and neuronal (peripherin (29Adams A.D. Choate D.M. Thompson M.A. J. Biol. Chem. 1995; 270: 6975-6983Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar) and myelin basic protein (30Inoue T. Tamura T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1990; 265: 19065-19070Abstract Full Text PDF PubMed Google Scholar)) genes. Although NFI is known as a ubiquitous transcription factor, interactions of specific NFI isoforms in distinct cell types may contribute to cell-selective transcriptional activation or silencing of target genes (31Apt D. Liu Y. Bernard H.-U. Nucleic Acids Res. 1994; 22: 3825-3833Crossref PubMed Scopus (76) Google Scholar). We find that NFI binding is required for transcription of the mouse SP-C promoter in lung cells and that the NFI-A1.1 isoform activates transcription of this promoter in HeLa cells. Nuclear extracts were prepared using a mini-extract procedure essentially as described (32Bohinski R.J. Huffman J.A. Whitsett J.A. Lattier D.L. J. Biol. Chem. 1993; 268: 11160-11166Abstract Full Text PDF PubMed Google Scholar). Briefly, nuclear pellets were resuspended in one packed nuclear volume of extraction buffer (20 mm HEPES (pH 7.9), 420 mmNaCl, 1.5 mm MgCl2, 25% glycerol, 1 mm dithiothreitol, 0.5 mm fresh phenylmethylsulfonyl fluoride), the NaCl concentration was adjusted to 400 mm, and the nuclei were incubated on ice with intermittent mixing for 10 min. Nuclei were centrifuged at 14,000 rpm at 4 °C for 15 min, and the supernatants containing nuclear proteins were divided into aliquots and immediately frozen at −80 °C. Nuclear extract protein concentrations were determined by a bicinchoninic acid assay (Sigma) using bovine serum albumin as a standard. Oligonucleotides were synthesized on an Applied Biosystems model 392 DNA/RNA synthesizer using phosphoramidite chemistry and purified using the Applied Biosystems oligonucleotide purification cartridge as described by the manufacturer. Annealed complementary oligonucleotides were diluted and used directly as unlabeled competitor DNA for electrophoretic mobility shift assays (EMSA). For use as EMSA probes, the annealed oligonucleotides were gel-purified using 4% Biogel and the MERmaid kit (Bio101, Vista, CA). Purified probes were end-labeled with [γ-32P]ATP and T4 polynucleotide kinase. Double-stranded oligonucleotide probes (see above) were 5′ end-labeled on one strand with [γ-32P]ATP and partially methylated on guanine residues with dimethyl sulfate. Methylated probes were incubated with 25 μg of MLE-15 cell nuclear extract in EMSA binding buffer for 30 min at room temperature, and probe bound to nuclear proteins was separated from free probe in a nondenaturing 5% polyacrylamide gel. The bound and free fractions were electro-eluted onto DEAE membranes (NA45, Schleicher & Schuell), recovered in 1 m NaCl, 20 mm Tris (pH 8), 0.5 mm EDTA, and ethanol-precipitated. The purified bound and free probes were cleaved at the methylated residues with piperidine, and the fragments were separated on 10% polyacrylamide, 7 m urea sequencing gels using the probe cleaved at guanine nucleotides (33Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989: 13.88-13.89Google Scholar) as a size marker. EMSAs were performed as described (12Bohinski R.J. DiLauro R. Whitsett J.A. Mol. Cell. Biol. 1994; 14: 5671-5681Crossref PubMed Scopus (486) Google Scholar) with slight modifications. Briefly, 5–6 μg of nuclear extract was incubated in EMSA binding buffer (20 mmTris (pH 7.6), 50 mm KCl, 2 mmMgCl2, 40 ng/μl poly[d(I-C)] (Boehringer Mannheim), 10% glycerol, 1 mm dithiothreitol, 0.1 mmfresh phenylmethylsulfonyl fluoride), and when indicated, with unlabeled competitor DNA for 5–10 min at room temperature.32P-End-labeled probe was added (100,000 dpm), and the mixture was incubated for an additional 10 min at room temperature. For antibody supershift-interference assays, 1 μl of rabbit antiserum to full-length Xenopus NFI-B1 recombinant protein (34Puzianowska-Kuznicka M. Shi Y.-B. J. Biol. Chem. 1996; 271: 6273-6282Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar) (a kind gift from M. Puzianowska-Kuznicka, National Institutes of Health, Bethesda, MD) was added, and the incubation was continued for an additional 20 min. Rabbit antiserum to rat TTF-1 (12Bohinski R.J. DiLauro R. Whitsett J.A. Mol. Cell. Biol. 1994; 14: 5671-5681Crossref PubMed Scopus (486) Google Scholar) was used as a control. Bound and free probes were separated by nondenaturing polyacrylamide gel electrophoresis using 5% acrylamide/bisacrylamide (29:1), 2.5% glycerol gels in 0.5 × TBE (1 × TBE = 0.1 m Tris borate (pH 8.3), 2 mm EDTA). The 0.32-kb murine SP-C promoter luciferase reporter construct, p0.32SP-C (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar), and pGL2Basic (Promega, Madison, WI) were used to assay SP-C promoter activity. Polymerase chain reaction-mediated introduction of site-specific mutations was performed as described previously (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar). Each mutant polymerase chain reaction product was double-digested with SmaI and XhoI and ligated to similarly digested pGL2Basic to generate the promoter mutants pSP-C1m, and pSP-C3m, and enhancer mutants pSP-C4m1, pSP-C4m2, and pSP-C5m. Double mutants pSP-C3C1m and pSP-C5C4m2 were generated by the same strategy using single mutants as a template. The plasmids pSP-C(−)NFI and pSP-C5C4C3m were generated by digesting the appropriate double or single mutants with EcoNI, which cuts between the enhancer and promoter region and again within the luciferase gene, and ligating the double mutant enhancer containing fragment to the double or single mutant promoter containing fragment. All mutations were verified by sequencing. Functional assays of the SP-C promoter reporter constructs were performed using transiently transfected MLE-15 cells, a SV40 T-antigen immortalized mouse lung epithelial cell line (35Wikenheiser K.A. Vorbroker D.K. Rice W.R. Clark J.C. Bachurski C.J. Oie H.K. Whitsett J.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 11029-11033Crossref PubMed Scopus (294) Google Scholar). MLE cells were maintained as described (35Wikenheiser K.A. Vorbroker D.K. Rice W.R. Clark J.C. Bachurski C.J. Oie H.K. Whitsett J.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 11029-11033Crossref PubMed Scopus (294) Google Scholar), and both MLE-15 and HeLa cells were transfected by the calcium phosphate co-precipitation method (36Rosenthal N. Methods Enzymol. 1987; 152: 704-720Crossref PubMed Scopus (403) Google Scholar), with modifications (12Bohinski R.J. DiLauro R. Whitsett J.A. Mol. Cell. Biol. 1994; 14: 5671-5681Crossref PubMed Scopus (486) Google Scholar). Six-well plates of MLE-15 cells at 50–60% confluence were transfected with 0.7 pmol of test construct (2.6 μg) and 0.3 pmol of pRSV-βgal (1.3 μg) per well. Transfected cells were allowed to grow to confluence (48 h), and then the plates were washed with phosphate-buffered saline, lysed with 150 μl/well 1 × Reporter Lysis Buffer (Promega), and frozen at −20 °C. Luciferase and β-galactosidase assays were performed on 10 μl of the cleared lysates as described (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar). Luciferase reporter gene activity was normalized for transfection efficiency based on the β-galactosidase activity, and the relative activity of the wild type p0.32SP-C promoter was set to 100. Transfections were performed in triplicate, and the pooled data from at least three independent experiments are shown. Transactivation assays to determine the effect of NFI-A expression on the SP-C promoter were performed in transiently transfected HeLa cells, a cell type that does not express endogenous SP-C mRNA. HeLa cells were maintained in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) with 10% fetal bovine serum. The NFI expression plasmid, pBETNFI-B1f, containing the mouse NFI-A1.1 isoform cDNA linked to the chicken β-actin promoter (30Inoue T. Tamura T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1990; 265: 19065-19070Abstract Full Text PDF PubMed Google Scholar) was a kind gift from Taka-aki Tamura (Chiba University, Chiba City, Japan). A control expression vector with no insert, pBETvector, was generated by double digestion of pBETNFI-B1f with BglII and SalI to liberate the NFI cDNA, filling in with Klenow, and re-circularization of the vector. Varying amounts of pBETNFI-B1f or pBETvector were co-transfected with 2.6 μg of p0.32SP-C, and 1.3 μg of pRSV-βgal/well in six-well plates as described above. To determine the effect of site-specific mutations of the SP-C promoter on NFI-A1.1 transactivation of promoter activity, 2.6 μg of pBETNFI-B1f or pBETvector were co-transfected with 2.6 μg of each SP-C promoter mutant, and 1.3 μg of pRSV-βgal/well. The relative luciferase activity of p0.32SP-C co-transfected with the “empty” pBETvector was set to 100. Relative luciferase activity from wild type and mutant SP-C promoters was analyzed by the two-tailed paired t test statistic using Statview 4.5 software (Abacus Concepts, Berkeley, CA). The -fold change in luciferase activity between cells co-transfected with pBET vector and pBETNFI-B1f was calculated based on the 95% confidence interval of each measurement. We previously identified five sites of protein-DNA interaction within the 0.32-kb mouse SP-C promoter region by DNase I footprinting (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar). The sequence of this region is shown in Fig.1, with the oligonucleotide probes used in this study boxed and labeled relative to the DNase I footprints, C1 through C5. The previously identified sites of interaction with TTF-1 (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar) are also indicated. Comparison of the footprinted regions with a data base of DNA-binding protein consensus sites using the MacVector program revealed several potential NFI half-site recognition sequences (underlined in Fig. 1). Contact point analysis was performed to map sites of protein-DNA interaction in the proximal promoter region using double-stranded oligonucleotide probes that span footprints C1 (−115 to −76) and C3 (−222 to −194). Methylated nucleotides that interfered with protein binding coincided with the predicted NFI binding sites, AGCCAA in footprint C1 and a slightly extended site TGCCAAG in footprint C3 (Fig.2).Figure 2Methylation interference footprinting of C1 and C3 DNA/protein complexes. Double-stranded oligonucleotides (5′-end labeled on the bottom strand for C1 and the top strand for C3) were partially methylated at guanine residues, and incubated with MLE-15 nuclear extract. Bound (B) and free (F) probes were isolated and cleaved at methylated residues, and the fragments were resolved on a 10% acrylamide sequencing gel using probe cleaved at G residues (G) as a marker. The relevant sequence of the analyzed regions is shown on the side of each panel, and the sites of methylated nucleotides that interfere with protein binding are indicated by bold type.View Large Image Figure ViewerDownload Hi-res image Download (PPT) To determine whether NFI family members bind to these sequences, competitive EMSA were performed using C1 and C3 as probes. Fig.3 A shows an alignment of C1 and C3 with a palindromic NFI consensus sequence oligonucleotide and two NFI consensus half-sites, one from the human albumin promoter in an A/T-rich context, termed “alb” (−135 to −110), and one from the human retinol-binding protein promoter in a G/C-rich context, termed “rbp” (−274 to −250) (15Gounari F. De Francesco R. Schmitt J. van der Vilet P.C. Cortese R. Stunnenberg H. EMBO J. 1990; 9: 559-566Crossref PubMed Scopus (120) Google Scholar). Specific complexes were formed between MLE-15 nuclear proteins and probes C1 and C3 that were competed by 100-fold excess unlabeled C1 and C3 and by all competitors containing NFI consensus sequences (Fig. 3 B). In contrast, C1m and C3m, containing nucleotide substitutions in the NFI consensus sites in C1 and C3 (see Fig. 3 A), did not form complexes with nuclear proteins in EMSA and did not compete for binding (Fig. 3 B). As shown in Fig. 3 C, the protein-DNA complexes formed with C1, C3, and the alb control probe were all supershifted by polyclonal antiserum to Xenopus NFI-B (lanes 2,5, 9, 12, and 16), whereas incubation with an irrelevant anti-rat TTF-1 antibody had no effect (lanes 6 and 13). Nuclear extracts from both MLE-15 and MLE-12 cells, which express endogenous SP-C (34Puzianowska-Kuznicka M. Shi Y.-B. J. Biol. Chem. 1996; 271: 6273-6282Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar), showed similar EMSA/supershift patterns with all three probes. Even though an extended methylation interference footprint was detected using the C3 probe, competition with NFI consensus sequences did not reveal any non-NFI binding activity for this probe. Specific mutation of the -GCC- within the NFI consensus site of the C3 probe to -AAA- blocked all binding to the mutant probe (Fig. 3 B), suggesting that only NFI family members interacted with this site. Previous studies demonstrated that the mouse SP-C promoter is active in transient transfection assays of MLE-15 cells (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar). To determine whether the NFI binding sites are required for basal SP-C promoter activity, the nucleotide substitutions that interfered with NFI binding in vitro were introduced into the reporter plasmid, p0.32SP-C (6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar), containing the firefly luciferase gene linked to the mouse SP-C promoter. In transient transfections of MLE-15 cells, mutation of either C1 or C3 resulted in loss of 40–60% of promoter activity. Double mutation of both C1 and C3 reduced expression from the SP-C promoter by 75% (pSP-C3C1m, Fig.4). These results suggest that NFI binding to both C1 and C3 is critical for the SP-C promoter to function properly in lung cells. Inspection of the enhancer region sequence revealed two NFI consensus half-sites, one each in C4 and C5 (see Fig. 1). Alignment of the C4 oligonucleotide (−280 to −301) with the palindromic NFI consensus sequence is shown in Fig.5 A. EMSA analysis using the C4 probe produced two major complexes of different mobilities with MLE-15 nuclear extracts (Fig. 5 B, large and small arrowheads). Specific mutation of the NFI consensus site CC → GT (probe C4m2) interfered with NFI binding (large arrowhead) but not binding of the faster mobility protein-DNA complex (small arrowhead). The faster mobility complex of C4 and C4m2 with MLE-15 nuclear extracts was not affected by 200-fold molar excess of the palindromic NFI competitor oligonucleotide. In contrast, a 4-base pair mutation of sequences adjacent to the NFI binding site (GACA → TTGC, shown in Fig. 5 A) resulted in loss of this complex (C4m1, Fig. 5 B). These results indicate that NFI and another nuclear factor(s) bind to adjacent/overlapping sites in the C4 probe. Alignment of the C5 region probe (−318 to −298) with the palindromic NFI consensus sequence revealed the presence of a potential strong NFI binding site (Fig. 6 A). In a competition EMSA, the NFI consensus sequence competed for binding with the C5 probe, and mutation of the NFI binding site in C5m (GCC → TTT) resulted in loss of binding (Fig. 6 B). Addition of antiserum to NFI produced supershifted EMSA complexes with C4, C4m1, and C5 (Fig.6 C), indicating that NFI family members interact with sites in both C4 and C5. The apparent abundance of the faster migrating C4 complex with MLE-15 nuclear protein(s) was increased in the presence of antibody to NFI, suggesting that binding of this complex and NFI may be mutually exclusive. A binding site for at least one other factor was identified adjacent to the NFI site at C4 by EMSA analysis. To determine the relative contribution of each identified site of protein-DNA interaction to enhancer activity, site-directed mutagenesis was performed on the NFI sites in C4 and C5, and the NFI adjacent site in C4 of the enhancer region of the SP-C promoter. Single site mutations in C4 and C5 did not have significant effects on SP-C promoter activity in transient transfection of MLE-15 cells (Fig. 7, pSP-C4m1, pSP-C4m2, and pSP-C5m). Mutation of both NFI sites in the enhancer region (pSP-C5C4m2) only marginally reduced SP-C transcription, whereas deletion of the entire enhancer region (p0.23SP-C) reduced activity in MLE-15 cells by approximately 70% (Fig. 7, see also Ref. 6Kelly S.E. Bachurski C.J. Burhans M.S. Glasser S.W. J. Biol. Chem. 1996; 271: 6881-6888Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar), suggesting that other factors in addition to NFI are required for SP-C enhancer activity. NFI binding to only C1 showed approximately 30% of wild type promoter activity (Fig. 7, pSP-C5C4C3m). Mutation of all four NFI sites in the SP-C promoter and enhancer resulted in luciferase activity slightly greater than the promoterless vector (p = 0.0028) in MLE-15 cells (Fig. 7). To determine whether co-expression of an NFI family member could transactivate SP-C promoter activity in HeLa cells, the SP-C luciferase constructs were co-transfected with pBETNFI-B1f (30Inoue T. Tamura T. Furuichi T. Mikoshiba K. J. Biol. Chem. 1990; 265: 19065-19070Abstract Full Text PDF PubMed Google Scholar) containing a mouse NFI-A1.1 cDNA linked to the chicken β-actin promoter. A dose-response relationship was observed between SP-C promoter activity and concentration of pBETNFI-B1f, from 0.7 to 5 μg/well in HeLa cells (data not shown). Co-transfection with pBETNFI-B1f (2.6 μg/well) transactivated the wild type 0.32SP-C promoter by 35-fold (TableI). Independent mutation of either the C1- or C3-NFI sites caused only a 25–30% reduction in the low basal transcription of the SP-C promoter in HeLa cells, but reduced NFI-A transactivation by approximately 60% (from 35-fold to 13-fold). In contrast, mutation of the NFI adjacent site in pSP-C4m1 did not have a significant effect on either basal transcription or NFI transactivation (see Table I). Mutation of the NFI site at C5 alone (pSP-C5m) increased basal transcription and decreased transactivation only slightly. Mutation of the NFI site at C4 (pSPC4m2) caused a slight reduction in basal SP-C activity and decreased transactivation approximately 2-fold. Mutation
Referência(s)