Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Secretory leukoprotease inhibitor (SLPI) is a nonglycosylated serine protease inhibitor consisting of 107 amino acids (M_r 11,726). Its major physiologic function is considered to be antineutrophil elastase protection at inflammatory sites (1). This serine protease inhibitor has been demonstrated in several normal tissues, and produced and released into mucus by secretory cells in respiratory, genital, and lacrimal glands, but not in the liver, endocrine glands, or hematologic system (2). In the respiratory tract, SLPI was observed in alveolar type II epithelial cells, serous cells of submucosal glands, and nonciliated epithelial cells (2).

It was reported that SLPI gene expression in several lung epithelial cell lines is upregulated by phorbol ester (3), neutrophil elastase (4), corticosteroids (5), interleukin (IL)-1, and tumor necrosis factor- (TNF-) (6). The stimulatory effect of phorbol ester is mediated by changes in gene transcription and mRNA stability (3). Moreover, it was found that up to 1.3 kilobase pair (kbp) of the 5' flanking sequence was transcriptionally active in human lung carcinoma cell lines, supporting the concept that this promoter region contains cis-active elements necessary for SLPI gene expression (7). However, little is known about the tissue-specific transcriptional regulation of the SLPI gene, nor has the interaction between cis-acting sequences and nuclear factors required for the regulatory pathway been demonstrated.

In the present study, using the transiently expressed reporter gene assay as well as DNase-I hypersensitive mapping, we define the cis-acting region that activates SLPI gene transcription only in human lung epithelial cell lines. Within this lung cell-specific regulatory region, we further reveal an 11-bp recognition sequence, 102 to 92 bp upstream of the transcription start site, for two nuclear proteins. Because one of the binding proteins is abundant in lung cell lines and another in nonlung cell lines, the ratio of these two nuclear factors is likely to be critical for cell type-specific transcription of the SLPI gene.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Cell Culture

Human lung adenocarcinoma A549 cells, human utero-cervical carcinoma HeLa cells, and human hepatocellular carcinoma Hep G2 cells were distributed by the Cancer Cell Repository (Tohoku University, Sendai, Japan). Human lung squamous cell carcinoma HS-24 cells were provided by Dr. W. Ebert (Thoraxklinik der LVA Baden, Heidelberg-Rohrbach, Germany). Cells were cultured as a monolayer in Dulbecco's modified Eagle's medium (DMEM; GIBCO-BRL, Gaithersburg, MD) supplemented with 10% fetal calf serum (FCS) at 37°C in a humidified 5% CO₂ atmosphere.

RNA Isolation and Northern Blot Hybridization

Total cellular RNA was isolated from cultured cells by acid guanidinium thiocyanate-phenol chloroform extraction (8), and poly(A)⁺ RNA was prepared from total cellular RNA by oligo(dT)-cellulose column chromatography. Two micrograms of poly(A)⁺ RNA were electrophoresed on a 1% agarose gel containing 2.2 M formaldehyde, and transferred to a nitrocellulose membrane. A cDNA probe was radiolabeled with [-³²P]dCTP (~ 111 TBq/mmol) (Du Pont, Wilmington, DE) using the random primers DNA labeling system (GIBCO-BRL). Filter hybridization with the probe proceeded for 16 h at 68°C in hybridization buffer containing 6× SSC, 5× Denhardt's solution, 0.5% sodium dodecyl sulfate (SDS), and 100 µg/ml denatured fragmented salmon sperm DNA. After hybridization, the membrane was washed successively with 2× SSC, 0.5% SDS at room temperature for 5 min with 2× SSC, 0.1% SDS at room temperature for 15 min with 0.1× SSC, 0.5% SDS at 37°C for 1 h, and with 0.1× SSC, 0.5% SDS at 68°C for 1 h. The membrane was dried and exposed to a Fuji imaging plate (Fuji Photo Film Co., Minamiashigara, Japan). Scanning of the signals was performed using a bio-imaging analyzer system (Fuji Photo Film Co.).

DNase-I Hypersensitive Site Mapping

Nuclei were prepared by homogenizing 108 cells in a glass Dounce homogenizer (Wheaton, Millville, NJ) with 15 strokes of a B pestle. The nuclei were resuspended in 1 ml of buffer A (10 mM Tris-HCl [pH 7.4], 10 mM KCl, 3 mM MgCl₂), and 118 µl of the suspension was treated with DNase I for each digestion time as follows. Briefly, after adjusting the volume of the nuclear suspension to 1 ml with buffer A, DNase I (Boehringer Mannheim Biochemicals, Indianapolis, IN) was added to a final concentration of 7.8 µg/ml. After incubation for 0 to 4 min at 30°C, digestion was stopped by the addition of 100 µl of 5% SDS, and 125 mM EDTA, and 25 µl of proteinase K (10 mg/ml). Following overnight incubation at 37°C, the DNA was purified by phenol-chloroform extraction and ethanol precipitation. DNA (48 µg) from each sample was completely restricted with BamH I, electrophoresed on 1.2% agarose, and subjected to Southern hybridization analysis using a ³²P-labeled DNA probe (pAK3) encompassing almost all of the intron I and exon II ([Pst I]-[BamH I] fragment; Figure 2a). The membrane was dried and analyzed by the bio-imaging analyzer system as used for Northern analysis.

View larger version (32K): [in this window] [in a new window]   Figure 2.   DNase-I hypersensitive site mapping in the SLPI gene. (a) Partial map of the SLPI gene and 5' flanking region with the sites for relevant restriction enzymes. Two BamH I sites are located in this area, one in exon II and the other 6 kbp upstream of exon II. DNase-I hypersensitive sites are evaluated within this 6-kbp BamH I genomic DNA fragment, using a 0.8-kbp (Pst I) (BamH I) probe (thick black line) corresponding to +140 ~ +961 bp downstream of the transcription start site. Four DNase-I hypersensitive sites (DH1, DH2, DH3, and DH4) detected in both A549 and HeLa cells are indicated with arrows. (b) Southern analysis demonstrating DNase-I hypersensitive sites in the chromatin of HeLa cells. Nuclei isolated from HeLa cells were treated with DNase I for the time periods indicated (top). The DNA was extracted, digested with BamH I, and analyzed by Southern blot hybridization with a 0.8-kbp (Pst I) (BamH I) probe (a). The parental fragment and four distinct fragments liberated by DNase-I digestion (DH1, DH2, DH3, and DH4) are indicated by arrows, with the sizes in kilobase pairs within parentheses.

Construction of Luciferase Expression Vector

The serially deleted fragments of the SLPI promoter were synthesized by polymerase chain reaction (PCR) using AmpliTaq (Perkin-Elmer Cetus, Norwalk, CT), with appropriate forward primers, a common reverse primer, and plasmid pAK2 as template, pAK2 is a pGEM-4Z (Promega, Madison, WI) with an inserted 1.2 kbp of the 5' flanking region of the SLPI gene (1,228 to +145 bp relative to the transcription start site) (9). A Kpn-I site was created near the 5' end of each forward primer, and an Xho-I site was created near that of the reverse primer (10). The PCR product was digested with both Kpn I and Xho I and inserted into the polylinker region of the pGL2-Basic (Promega), keeping the direction of the promoter upstream of the firefly luciferase gene. The sequence of each construct was confirmed by the dideoxy chain-termination method (11) with Sequenase version 2.0 (United States Biochemical Corp., Cleveland, OH).

Transfection, Luciferase Assay, and Chloramphenicol Acetyltransferase Assay

Cells were transfected by the calcium phosphate technique (12) as follows. Twenty-four hours prior to transfection, A549 and HeLa cells were seeded onto plastic dishes (9 cm in diameter) at a density of 1 × 10⁶/plate, and Hep G2 cells at a density of 4 × 10⁶/plate. Twenty micrograms of the luciferase construct and 5 µg of pRSVcat, in which the coding region of the chloramphenicol acetyltransferase (CAT) gene is ligated to the long terminal repeat of the Rous sarcoma virus (13), were coprecipitated and left on the cells for 4 h. The cells were then exposed to 15% glycerol for 30 s (HeLa cells) or 2 min (A549 and Hep G2 cells), and refed with growth medium. Cells were harvested 48 h after transfection, and divided into equal aliquots for luciferase and CAT assays. For the luciferase assay (14), pelleted cells were lysed in 250 µl of 25 mM Tris-phosphate (pH 7.8), 2 mM dithiothreitol (DTT), 2 mM 1,2-diaminocyclohexane-N,N,N',N'-tetraacetic acid, 10% glycerol, and 1% Triton X-100, and incubated for 15 min at room temperature. The cell-free extract prepared by centrifugation was analyzed for luciferase activity by a luminometer Lumat LB 9501 (Berthold, Berlin, Germany). The reaction was started by mixing 20 µl of the cell extract with 100 µl of assay buffer (Toyo Ink Mfg. Co., Tokyo, Japan), which was composed of 20 mM Tricine, 1.07 mM (MgCO₃)₄ Mg(OH)₂ · 5 H₂O, 2.67 mM MgSO₄, 0.1 mM EDTA, 33.3 mM DTT, 270 µM coenzyme A, 470 µM luciferin, and 530 µM ATP. Light emission was measured for 10 s after starting the reaction, and integrated over time by the luminometer. The value was corrected after protein quantification of the extract according to the method of Bradford (15). Luciferase activity was assayed within the linear range in terms of the reaction time and the amount of the enzyme. To monitor the transfection efficiency for each dish, the CAT activity was assayed as described by Gorman and colleagues (16). Following standardization with the acetylation ratio of [¹⁴C]chloramphenicol, the level of luciferase activity in each sample was normalized relative to the activity of p-1228LUC plasmid in A549 cells. The result was expressed as the mean ± SEM of three independent transfections for each construct. Statistical comparison between a construct and the subsequently truncated construct was made by a two-tailed Student's t test, and a value of P < 0.01 was accepted as indicating statistical significance.

Preparation of Nuclear Extracts

Nuclear extract was prepared as described by Schreiber and colleagues (17). Cells (3.5 to 4.5 × 10⁶) were washed with phosphate-buffered saline (PBS), and resuspended in 400 µl of buffer A (10 mM Hepes [pH 7.8], 10 mM KCl, 0.1 mM EDTA, 2 mM DTT, 1 mM phenylmethylsulfonyl fluoride [PMSF]). The cells were incubated on ice for 15 min, then 25 µl of 10% Nonidet P-40 (NP-40) was added, and the tube was vigorously vortexed for 10 s. The homogenate was centrifuged for 30 s in a microcentrifuge, and the nuclear pellet was resuspended in 50 µl of buffer C (20 mM Hepes [pH 7.8], 420 mM NaCl, 5 mM EDTA, 5 mM DTT, 1 mM PMSF, 10% glycerol). The tube was vigorously rocked at 4°C for 30 min on a shaking platform, and centrifuged in a microfuge for 10 min. The supernatant was frozen in aliquots at 70°C. Protein concentration was determined by the method of Bradford (15), using the protein assay dye reagent provided by Bio-Rad Laboratories (Richmond, VA) with bovine serum albumin as a standard.

Electrophoretic Mobility Shift Assay

A duplex oligonucleotide probe was labeled with [-³²P] dATP (~ 222 TBq/mmol)(Du Pont) by filling in a 5' overhanging end with Klenow DNA polymerase. The end-labeled probe (~ 0.5 ng of DNA) was incubated in the binding reaction buffer (10% glycerol, 5 mM Tris-HCl [pH 7.5], 25 mM NaCl, 0.25 mM DTT, 0.05% NP-40) with 0.3 µg of poly(dI-dC) · poly(dI-dC) and 3 µg of a crude nuclear extract for 20 min at room temperature. Following the binding reaction, the mixture was electrophoresed through a low ionic strength gel (5% polyacrylamide, 7 mM Tris-HCl [pH 7.5], 3 mM sodium acetate, 1 mM EDTA). Electrophoresis was carried out at 10 V/cm for 2.5 h. The gel was dried and analyzed by the bio-imaging analyzer system as used for Northern analysis. For a competition experiment, 20 ng of specific or nonspecific competitor DNA was incubated in the mixture prior to addition of the labeled probe.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Expression of the SLPI Gene in A549 and HeLa Cells

Northern analysis revealed that the poly(A)⁺ RNA prepared from A549 and HeLa cells contained 0.7-kb RNA transcripts that hybridized with the SLPI cDNA, but no band was detected when Hep G2 cells were used (Figure 1). The amount of SLPI mRNA transcripts is less abundant in type II pneumocyte cell line A549 as compared with the utero-cervical cell line HeLa. The entirety of the mRNA was confirmed using a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA probe, even with hepatocyte cell line Hep G2, which showed undetectable SLPI mRNA transcripts.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Expression of the secretory leukoprotease inhibitor (SLPI) gene in human epithelial cell lines. Shown is the result of Northern analysis of the poly(A)+ RNA isolated from Hep G2 cells, A549 cells, and HeLa cells. Two micrograms of poly(A)+ RNA was loaded on each lane, electrophoresed on a 1.0% agarose gel containing 2.2 M formaldehyde, and blotted onto nitrocellulose membrane. The membrane was hybridized with a labeled probe of SLPI cDNA or GAPDH cDNA followed by autoradiography. Although poly(A)+ RNA from each cell line contains a similar amount of GAPDH transcripts with a size of 1.4 kb, SLPI transcripts with a size of 0.7 kb were detected in poly(A)+ RNA from A549 cells and HeLa cells but not in that from Hep G2 cells.

Mapping of DNase-I Hypersensitive Sites

In chromatin analysis, four DNase-I hypersensitive sites were identified in a 6-kbp BamH I restriction endonuclease area containing the first exon of the SLPI gene (Figure 2a). Using a 0.8-kbp (Pst I)-(BamH I) probe (+140 to +961 bp downstream of the transcription start site), four DNA fragments (1.5, 1.2, 1.0, and 0.7 kbp), were detected in Southern analysis of DNase-I digested HeLa nuclei (Figure 2b). In the context of the SLPI gene structure, corresponding DNase-I hypersensitive sites, designated DH1 to DH4, were located 0.6, 0.3, and 0.1 kbp upstream from the beginning of exon I, and within the first intron. These four DNase-I hypersensitive sites were also observed with nuclei prepared from A549 cells (data not shown). This result suggests that candidates as cis-regulatory elements are likely to lie within 1 kbp upstream of the SLPI gene.

Functional Analysis of the SLPI Promoter

Evaluation of the SLPI promoter by reporter gene assay encompassing DNase-I hypersensitive sites. When the luciferase reporter genes with various lengths of the SLPI promoter region were constructed and evaluated, the 5' flanking region of the SLPI gene from 1,228 to +22 bp is sufficient to direct transcription in a cell type-specific manner. The longest SLPI promoter region from 1,228 to +22 bp (p-1228LUC) yielded 125-fold and 185-fold greater luciferase activities than pGL2-Basic in SLPI-expressing cell lines, A549 and HeLa cells, respectively (Figure 3). In contrast, transfection of this construct (p-1228LUC) into Hep G2 cells supported only slight luciferase activity above that of the promoterless vector (pGL2-Basic). These results are consistent with the transcriptional activity or inactivity of the SLPI gene assessed by Northern blot analysis.

View larger version (20K): [in this window] [in a new window]   Figure 3.   Transient transfection analysis of the SLPI promoter in A549, HeLa, and Hep G2 cells. On the left is shown a 5' deletion series. Various lengths (up to 1,228 bp relative to the transcription start site) of the SLPI promoter and the first exon comprising 22 bp of the 5' untranslated region were linked upstream of the luciferase gene (LUC). The numbers to the left of each construct indicate the 5' end point. These constructs and a plasmid lacking a promoter (pGL2-Basic) were transiently transfected into A549, HeLa, and Hep G2 cells, and the luciferase activities were assayed. On the right is listed the relative luciferase activity of each construct expressed as a percentage of the activity of p-1228LUC in A549 cells (see MATERIALS AND METHODS for details). The values represent the mean ± SEM of three independent transfections. Asterisks denote significant differences (P < 0.01) compared with the subsequently truncated construct.

Deletion analysis of the sequence flanking the SLPI promoter. Transfection of fusion genes composed of sequentially deleted SLPI promoter and a luciferase reporter gene into A549 cells proved that three distinct regions are concerned in the promoter function (Figure 3). Luciferase activity varied significantly in A549 cells with truncations from p-1059LUC (2-fold activation), p-849LUC (3-fold reduction), and p-132LUC (12-fold reduction), while activity remained constant in both HeLa and Hep G2 cells. These data indicate that three distinct cis-acting elements that function only in A549 cells lie in the 5' flanking region of the SLPI gene, and that no specific segment from 1228 to 92 bp affects SLPI transcription in either HeLa or Hep G2 cells.

Similar deletion analysis previously demonstrated that the SLPI promoter spanning 115 to 97 bp is essential for constitutive and phorbul myristate acetate (PMA)- induced SLPI expression in human squamous lung carcinoma HS-24 cells (3). Among three A549-restricted regulatory regions revealed in our deletion study, the most proximal region from 132 to 92 bp not only corresponds to the cis-acting region detected in HS-24 cells, but also is responsible for the striking magnitude (12-fold) of transcriptional activation in A549 cells. These observations suggest that this 41-bp promoter sequence (132 to 92 bp) shares cis-acting elements associated with lung cell-specific expression of the SLPI gene.

Binding of Nuclear Proteins to the SLPI Promoter

A DNA-protein interaction assay using A549 nuclear extracts revealed that two nuclear proteins interact with the lung cell-specific promoter (132 to 92 bp) in a sequence-specific manner (Figures 4 and 5). In the analysis, two DNA-protein complexes, termed SLPI-B1 and SLPI- B2, were detected (Figure 5, lane 2). The formation of SLPI-B1 and SLPI-B2 complexes was diminished by competition with excess unlabeled specific competitor S (Figure 4, lane 3), but was not diminished by competition with excess unlabeled nonspecific competitor (Figure 4, lane 4), demonstrating the sequence specificity of these two DNA- protein interactions.

View larger version (14K): [in this window] [in a new window]   Figure 4.   The nucleotide sequence of the 5' flanking region and first exon of the SLPI gene. The numbering indicates the nucleotide position relative to the transcription start site, which is designated as +1. Coding sequence for amino acids (boldface) begins +23 bp downstream. The CAAT (88 to 84 bp) and TATA (27 to 22 bp) boxes are also indicated. Schematic representation of the duplex oligonucleotides used in the electrophoretic mobility shift assay (EMSA) is inserted. The positions of the 41-bp specific competitor (S) and three short duplex oligonucleotides (DO1, DO2, and DO3) are indicated.

View larger version (49K): [in this window] [in a new window]   Figure 5.   EMSA analysis of the 41-bp lung cell-specific SLPI promoter sequence corresponding to competitor S. Nuclear extract from A549 cells (A549 NE) was prepared as described in MATERIALS AND METHODS. Three micrograms of nuclear protein was incubated with the 32P-labeled 41-bp DNA probe (132 to 92 bp). Protein-DNA complexes were resolved by electrophoresis on a 5% polyacrylamide gel. Two specific DNA-protein complexes, SLPI-B1 and SLPI-B2, are indicated by arrows on the right. Lane 1, without nuclear extract; lane 2, with nuclear extract in the absence () of competitor (Comp.); lane 3, with nuclear extract in the presence of the 41 bp-specific competitor S (132 to 92 bp); lane 4, with nuclear extract in the presence of nonspecific competitor NS. As a nonspecific competitor, the 41-bp duplex oligonucleotide without homology to the specific competitor S was used.

SLPI-B1 and SLPI-B2 Complexes among Epithelial Cell Lines

Although SLPI-B1 and SLPI-B2 complexes were seen in the cell lines used, the signal intensity varied according to the SLPI promoter function of the cell type (Figure 6). In this context, when using nuclear extracts from lung cell lines (HS-24 or A549 cells) in which the 41-bp SLPI promoter region (132 to 92 bp) is transcriptionally active, the SLPI-B1 complex is much more abundant than the SLPI-B2 complex (Figure 6, lanes 3 or 4). In contrast, when using nuclear extract from nonlung cell lines (Hep G2 or HeLa cells) in which the 41-bp promoter region is transcriptionally inactive, the SLPI-B2 complex is dominant over the SLPI-B1 complex (Hep G2 cells; Figure 6, lane 1) or as abundant as the SLPI-B1 complex (HeLa cells, Figure 6, lane 2). This experimental evidence suggests that the ratio of SLPI-B1 and SLPI-B2 binding proteins has important implications for cell type-specific function of this 41-bp SLPI promoter sequence.

View larger version (42K): [in this window] [in a new window]   Figure 6.   Analysis of nuclear protein-DNA complexes of SLPI- B1 and SLPI-B2 in several epithelial cell lines. The 32P-labeled 41-bp DNA probe (132 to 92 bp) was incubated with 3 µg of nuclear extracts (NE) from Hep G2 (lane 1), HeLa (lane 2), HS-24 (lane 3), and A549 cells (lane 4). The reaction mixture was resolved on a 5% polyacrylamide gel. Relative binding of SLPI-B1 and SLPI-B2 are indicated at the bottom. ++, Strong; +, weak.

Sequence Requirements of SLPI-B1 and SLPI-B2 Complexes

Another DNA competition assay revealed that the 11-bp sequence within the 41-bp lung cell-specific promoter region is indispensable for the formation of SLPI-B1 and SLPI-B2 complexes (Figure 7). To evaluate the sequence requirements of SLPI-B1 and SLPI-B2, HeLa cells expressing abundant SLPI-B1 and SLPI-B2 binding proteins (Figure 6, lane 2) were used in this competition assay. Nuclear extract was preincubated with an excess amount of a cold short competitor (DO1, DO2, and DO3; Figure 4), and an electrophoretic mobility shift assay (EMSA) was performed with the ³²P-labeled 41-bp probe (132 to 92 bp). As a result, using oligonucleotide DO1 (132 to 113 bp) and DO2 (122 to 103 bp) as the cold competitor (Figure 7, lanes 4 and 5), a pattern similar to that of the nonspecific competitor NS (Figure 7, lane 3) in the formation of SLPI-B1 and SLPI-B2 complexes was observed, indicating that the DNA sequence between 132 and 103 bp is not responsible for the DNA-protein binding. However, when oligonucleotide DO3 (112 to 92 bp) was used as the competitor, the formation of both SLPI-B1 and SLPI-B2 complexes was strongly suppressed (Figure 7, lane 6). These data demonstrate that both SLPI- B1 and SLPI-B2 binding proteins interact with the identical sequence in DO3 but not in DO1 and DO2. A short 11-bp promoter sequence, from 102 to 92 bp, is likely to be a target sequence for these nuclear proteins in a mutual interactive manner.

View larger version (68K): [in this window] [in a new window]   Figure 7.   Comparison of complex formation with partial specific competitors. EMSA competition experiment was performed as described in MATERIALS AND METHODS. The 32P-labeled 41-bp DNA probe (132 to 92 bp) and HeLa cell nuclear extract were incubated with or without competitor (Comp.). Lane 1, in the absence of competitor; lane 2, in the presence of the 41 bp-specific competitor S (132 to 92 bp); lane 3, in the presence of nonspecific competitor NS; lanes 4-6, in the presence of short specific competitors DO1, DO2, and DO3. Relative binding of SLPI-B1 and SLPI-B2 are indicated at the bottom. ++, Strong; ±, barely detectable; , undetectable.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

To provide a basic framework for understanding the shared mechanisms of gene regulation in respiratory epithelial cells, transcriptional elements in the surfactant protein (SP) (A, B, and C) and Clara cell secretory protein (CCSP) have been analyzed. Concerning the SP-B gene, 218 bp from the transcription start site is sufficient to confer lung epithelial cell gene expression in vitro, and thyroid transcription factor 1 (TTF-1) and hepatocyte nuclear factor 3/forkhead (HNF-3/fkh) family members bind to and activate the SPB-f1 site (113 to 90 bp) and SP-B-f2 site (91 to 70 bp), respectively (18, 19).

The essential process that regulates the lung cell-specific expression of the SLPI gene was shown, using cultured cell lines (9), to occur during transcription in vitro. Moreover, Maruyama and colleagues have demonstrated, using transient transfection of lung cancer-derived HS-24 cells, that a positive cis-active element is located between 115 and 97 bp relative to the transcription start site (3). However, little information is available for identification of transcriptional motifs controlling the SLPI gene expression, and nuclear proteins from lung cells have not been shown to bind the regulatory cis elements.

In this study, we have examined the mechanisms for the lung epithelium-specific transcription of the SLPI gene, focusing on cis-acting regions in the promoter and nuclear proteins interacting with the sequences. By transient luciferase expression assay, we identified three cis-acting regulatory regions functioning specifically in lung cell line A549. Among these three regions, the most proximal sequence (from 132 to 92 bp) is responsible for the most striking magnitude of transcriptional activation in A549 cells. This region contains the promoter sequence from 115 to 97 bp, which was reported to share the transcriptional regulatory elements in HS-24 cells (3). Thus, we consider the 41-bp promoter region from 132 to 92 bp to be associated with lung cell-specific transcription of the SLPI gene. Analysis by EMSA of this 41-bp promoter sequence indicates that two nuclear factors, SLPI-B1 and SLPI-B2 binding proteins, interact with the short sequence from 102 to 92 bp in a sequence-specific manner. Furthermore, when using nuclear extract of A549 or HS-24 cells, in which the 41-bp promoter sequence is shown to be requisite for the transcriptional activation by the transient transfection assay, the nuclear protein-DNA complex of SLPI-B1 is much more abundant than that of SLPI-B2 (Figure 6, lanes 3 and 4). Conversely, when using nuclear extract of Hep G2 or HeLa cells, in which deletion of the 41-bp promoter sequence made no change in terms of transcriptional activity, the nuclear protein-DNA complex of SLPI-B2 is much more abundant than that of SLPI-B1 (Hep G2 cells) or is as abundant as that of SLPI-B1 (HeLa cells) (Figure 6, lanes 1 and 2). A hypothetical interpretation of these results is as follows: in pulmonary epithelial cells such as A549 and HS-24 cells, abundant SLPI-B1 binding protein interacts with the 11-bp promoter sequence, and this binding provides crucial signals required for the activation of the transcriptional unit of the SLPI gene. In contrast, in nonlung cell lines such as Hep G2 and HeLa cells abundant SLPI-B2 binding protein dominates the 11-bp promoter sequence, and inactivates the cis-acting sequence by interacting with the binding of the positively acting SLPI-B1 nuclear protein.

It is interesting to note that this 11-bp promoter sequence contains the sequence 5' CGTTTCC 3' (102 to 96 bp; Figure 4), which shows homology (five of seven nucleotides match) to the core sequence T(G/A)TTTA(C/ T) found in the binding site for a number of HNF-3/fkh families (20). The consensus HNF-3 binding site (22), centered around the HNF-3/fkh core sequence, has been identified in the SP-A (23), SP-B (24), and CCSP (25, 26) promoters, and it has been suggested that HNF-3 proteins are likely to play a critical role in regulating the expression of these genes in the lung. Despite an agreement of the SLPI promoter sequence (105 to 94 bp) with the consensus HNF-3 binding site sequence (8 of 12 nucleotides match), the binding motif of any known HNF-3/fkh family members so far does not completely correspond to this SLPI promoter sequence. Sequence variations in the core and flanking sequence suggest that as yet unidentified HNF-3/fkh family members may also be implicated in SLPI promoter function. Future experiments will be required to identify these nuclear factors.

In summary, we have identified the lung cell-specific regulatory region in the SLPI promoter, and have shown that the amount of two nuclear proteins interacting with this regulatory sequence is different between lung and nonlung cell lines. The identification and characterization of these proteins should further our understanding of the molecular mechanisms of SLPI gene regulation.