INTRODUCTION

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)¹ gene is restricted to pulmonary epithelial cells and regulated by developmental and humoral influences (1). SP-C is a hydrophobic protein that enhances the spreading and stability of surfactant phospholipids at the air liquid interface during the respiratory cycle (2). 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 (3, 4). 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. 5 and 6; reviewed in Ref. 7). SP-C gene expression is decreased in acute lung injury and is increased in areas of respiratory epithelial regeneration (8-10). Tumor necrosis factor  (TNF-), inhibits transcription from both the human and mouse SP-C promoters (5). Sequences from the mouse SP-C promoter (318 to +18) are sufficient to confer TNF- regulation on a transfected marker gene (5), 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) (6). 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 (6). TTF-1 activates transcription of several other surfactant-associated genes expressed in overlapping subsets of pulmonary epithelial cells including surfactant protein-A (11), surfactant protein-B, and Clara cell secretory protein (12). 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 (13, 14) and with lower affinity to the half-site TGGCA (15, 16). NFI family members bind to DNA as homo- or heterodimers (15, 17). The NFI proteins are highly homologous in the amino-terminal DNA binding and dimerization domain but diverge in the carboxyl-terminal transactivation domain (18, 19). 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 (20, 21). During mouse embryogenesis, NFI family members have unique, albeit overlapping patterns of expression (22). NFI has been shown to regulate several cell-selective genes including liver-specific (albumin (23, 24), retinol-binding protein (25), vitellogenin (26), -fetoprotein (27)), adipocyte-specific (aP2 (28)), and neuronal (peripherin (29) and myelin basic protein (30)) 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 (31). 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.

MATERIALS AND METHODS

Nuclear extracts were prepared using a mini-extract procedure essentially as described (32). Briefly, nuclear pellets were resuspended in one packed nuclear volume of extraction buffer (20 mM HEPES (pH 7.9), 420 mM NaCl, 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 [-³²P]ATP and T4 polynucleotide kinase.

Double-stranded oligonucleotide probes (see above) were 5 end-labeled on one strand with [-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 (33) as a size marker.

EMSAs were performed as described (12) with slight modifications. Briefly, 5-6 µg of nuclear extract was incubated in EMSA binding buffer (20 mM Tris (pH 7.6), 50 mM KCl, 2 mM MgCl₂, 40 ng/µl poly[d(I-C)] (Boehringer Mannheim), 10% glycerol, 1 mM dithiothreitol, 0.1 mM fresh 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 (34) (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 (12) 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 (6), 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 (6). 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 (35). MLE cells were maintained as described (35), and both MLE-15 and HeLa cells were transfected by the calcium phosphate co-precipitation method (36), with modifications (12).

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-gal (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 -galactosidase assays were performed on 10 µl of the cleared lysates as described (6). Luciferase reporter gene activity was normalized for transfection efficiency based on the -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 -actin promoter (30) 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-gal/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-gal/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.

RESULTS

We previously identified five sites of protein-DNA interaction within the 0.32-kb mouse SP-C promoter region by DNase I footprinting (6). 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 (6) 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).

[View Larger Version of this Image (51K GIF file)]

[View Larger Version of this Image (88K GIF file)]

To determine whether NFI family members bind to these sequences, competitive EMSA were performed using C1 and C3 as probes. Fig. 3A 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) (15). 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. 3B). In contrast, C1m and C3m, containing nucleotide substitutions in the NFI consensus sites in C1 and C3 (see Fig. 3A), did not form complexes with nuclear proteins in EMSA and did not compete for binding (Fig. 3B). As shown in Fig. 3C, 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 (34), 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. 3B), suggesting that only NFI family members interacted with this site.

[View Larger Version of this Image (48K GIF file)]

Previous studies demonstrated that the mouse SP-C promoter is active in transient transfection assays of MLE-15 cells (6). 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 (6), 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.

[View Larger Version of this Image (20K GIF file)]

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. 5A. EMSA analysis using the C4 probe produced two major complexes of different mobilities with MLE-15 nuclear extracts (Fig. 5B, 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. 5A) resulted in loss of this complex (C4m1, Fig. 5B). These results indicate that NFI and another nuclear factor(s) bind to adjacent/overlapping sites in the C4 probe.

[View Larger Version of this Image (59K GIF file)]

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. 6A). 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. 6B). Addition of antiserum to NFI produced supershifted EMSA complexes with C4, C4m1, and C5 (Fig. 6C), 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.

[View Larger Version of this Image (57K GIF file)]

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. 6), 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).

[View Larger Version of this Image (31K GIF file)]

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 (30) containing a mouse NFI-A1.1 cDNA linked to the chicken -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 (Table I). 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 of both enhancer NFI binding sites in pSP-C5mC4m2 had no effect on basal transcription, but reduced the transactivation by NFI-A1.1 approximately 70% (similar to the reductions seen with individual mutation of C1 or C3). Deletion of the entire enhancer region had no effect on basal transcription in HeLa cells, but reduced NFI-A transactivation by 6-fold, suggesting that interactions with upstream sequences are required for proper interaction of NFI-A1.1 with C1 and C3. Mutation of both NFI sites in the promoter region in pSP-C1mC3m had no further effect on basal transcription in HeLa cells, but reduced transactivation by 9-fold. Mutation of all NFI sites in both the promoter and enhancer regions (pSP-C()NFI) reduced the basal transcription in HeLa cells by only 20%, but blocked transactivation with NFI-A1.1 (Table I). These results suggest that in HeLa cells basal transcription from the SP-C promoter is not dependent on NFI binding and NFI-A1.1 interacts with multiple sites in the mouse SP-C promoter to activate transcription.

Table I. NFI-A1 transactivation of SP-C promoter activity in HeLa cells a HeLa cells were co-transfected with 2.6 µg of the SP-C luciferase construct indicated at the left, and 2.6 µg pBETNFI-B1f or the empty pBETvector, and 1.3 µg pRSVgal/well as described under "Materials and Methods." The relative luciferase activity from cells co-transfected with p0.32SP-C and pBETvector was set to 100 for each transfection. b Mean ± S.E. (number of wells). c Fold change in SP-C promoter activity between cells co-transfected with the empty pBETvector and the NFI-A1.1 expression construct (pBETNFI-B1f), based on the 95% confidence interval.

DISCUSSION

In this study, four binding sites for NFI family members were identified in the mouse SP-C promoter, including two sites within the proximal promoter region (C1 and C3) and two sites in the enhancer region (C4 and C5). NFI-A1.1 interacted with both promoter and enhancer sites in the SP-C promoter to activate transcription in HeLa cells. NFI-A1.1 transactivation in HeLa cells was independent of TTF-1 binding since HeLa cells do not have endogenous TTF-1 binding activity (12) and mutation of the TTF-1 binding sites failed to inhibit transactivation by NFI (data not shown). Mutation of all four NFI binding sites in the SP-C promoter produced a non-functional promoter in MLE-15 cells despite the presence of endogenous TTF-1 in this cell line (6, 11). These results are consistent with a critical role for NFI family members in regulation of SP-C gene expression that occurs independently of TTF-1.

Although NFI is thought to be a ubiquitous transcription factor, the complement of NFI isoforms varies between cell types (37-39). Recently, Chaudhry et al. (22) showed that the four NFI genes are expressed in unique but overlapping patterns in the developing mouse embryo consistent with the notion that specific combinations of NFI family members may play a role in tissue-specific gene expression. Some NFI family members have been shown to have different activities when they bind in the promoter position versus the enhancer position. NFI-X1, an isoform well expressed in fibroblasts, does not transactivate NFI sites in the enhancer position, whereas NFI-X2 and NFI-C1 act at both promoter and enhancer positions (31). In this report, NFI-A1.1 acted at both promoter and enhancer positions to transactivate SP-C promoter activity.

Composition of NFI isoforms appears to be an important variable contributing to cell-specific gene expression. Of three human NFI-C splice variants tested, NFI-C1 is most active in transcriptional activation of strong NFI binding sites near the TATA box (19). Recently, NFI-A1.1 was shown to have greater activity on the neurotropic JC viral promoter than NFI-C1 (40). Rat NFI-A1.1 was originally purified and cloned from liver as NFI-L (16), a nuclear protein that binds to the alb and rbp NFI half-sites used as competitors in this study (see Fig. 3). Mouse NFI-A1.1 was isolated from a cerebellum cDNA library and termed NFI-B1f, since it was enriched in brain (30). NFI-A1.1 is 39 nucleotides shorter in the transactivation domain than NFI-A1, an isoform only described in chicken (18). NFI-A family members are expressed at varying levels in many tissues, including lung (22) and freshly isolated pulmonary Type II epithelial cells (data not shown). Whether NFI isoforms other than NFI-A1.1 would induce or repress SP-C promoter activity remains to be determined. Identification of a novel isoform of NFI-B (NFI-B3), which acts as a dominant negative factor in co-transfection experiments (41), suggests that the repertoire of NFI isoforms in a cell may determine whether a NFI dependent gene is activated or silenced. The significance of the differences between most of the NFI family members and splice forms in gene activation and silencing is still poorly understood.

The C1 and C3 NFI sites in the proximal promoter region were critical for SP-C promoter activity in MLE cells, but their absence had minimal effect on basal promoter activity in HeLa cells, suggesting that there are differences in NFI/DNA interactions in these two epithelial cell lines, and possibly in the identity of isoforms expressed in these cells. The precise repertoire of the NFI isoforms expressed in pulmonary Type II epithelial cells in vivo, and their relative contribution to regulation of SP-C promoter activity remain to be studied.

Mutation of the NFI adjacent site in C4m1 had little effect on SP-C promoter activity in MLE cells, or on NFI-A1.1 transactivation in HeLa cells. These results suggest that the factor(s) that bind to this site are not important for SP-C promoter function in transient transfection assays. Since deletion of the entire enhancer region caused a 70% reduction in SP-C promoter activity in MLE-15 cells and inhibited NFI-A1.1 transactivation in HeLa cells, we speculate that other, as yet unidentified, enhancer binding factors may interact with NFI. NFI sites adjacent to or overlapping other transcription factor binding sites occur in several cell-specific promoters, and it has been proposed that the combinatorial interactions of NFI family members with adjacent proteins may ultimately determine the level of transcriptional activation or repression (40, 42). Oct1 binding 2 base pairs from a NFI half-site in the human papillomavirus enhancer causes stronger NFI binding and synergistic promoter activation in epithelial cells (43). In liver cells, NFI binding interferes with transcriptional activation by hepatocyte nuclear factor 3 (HNF-3) binding to an adjacent site in the promoter position, but cooperates with HNF-3 in the context of the multi-component albumin enhancer (24). Identification of the other proteins that bind to the SP-C promoter and interact with NFI family members will be required to understand the interactions that lead to cell-specific expression of the SP-C promoter.

Interaction of NFI with adjacent transcription factors is a hallmark of certain tissue-specific promoters. The combination of NFI and AP1 binding sites occurs in several neurotropic promoters (44), while NFI and HNF-3 are found in liver-specific promoters (24). We have now identified sites for both NFI and TTF-1 (6) in the SP-C promoter. Several potential NFI binding sites are also present in the mouse SP-B enhancer, another lung-specific, TTF-1 responsive gene.² The combinatorial effect of NFI isotype repertoire interacting with multiple sites in the promoter and enhancer region, and the presence of TTF-1, a pulmonary epithelial cell-specific transacting factor, may begin to explain the Type II cell specificity of the SP-C promoter.

Both NFI and TTF-1 undergo post-translational modifications in response to extracellular signals. NFI and TTF-1 are phosphorylated (45, 46), and O-glycosylated (47), and at least some NFI family members are cell cycle responsive (37). Changes in expression of various NFI isoforms, or their secondary modifications may be involved in regulation of SP-C promoter activity in response to extracellular signals.