Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The pulmonary inflammatory response entails a beneficial influx of neutrophils as part of the natural defense of the lung. Part of this defense mechanism involves the release of proteolytic enzymes into the epithelial lining fluid. However, release of unregulated proteolytic enzymes by inflammatory cells within the lung may be critical to the pathogenesis of inflammatory lung disease (1). Protection against these proteinases, of which neutrophil elastase (NE) is a prototype, is mediated by antiproteinases that include 1-proteinase inhibitor (1-PI), elafin, and secretory leukocyte protease inhibitor (SLPI) (1, 2). Whereas 1-PI is produced mainly by the liver and reaches the lung via passive diffusion, elafin and SLPI are locally produced within the pulmonary system (3). Elafin, also known as skin-derived antileukoprotease and trappin-2, is synthesized as a 117-amino acid preprotein that is structurally related to SLPI by virtue of a shared C-terminal, four-disulfide core whey acidic protein (WAP) domain (6). Elafin is the 57- amino acid C-terminal proteolytic fragment of pre-elafin, which contains only the WAP domain. Elafin was originally isolated from psoriatic scales (7, 8) and was subsequently identified in the trachea (3) and breast (9). Elafin is expressed in multiple epithelia (10), is found in bronchoalveolar lavage fluid from normal subjects, and is increased in patients with inflammatory lung disease (11). Although elafin is present in the trachea (3), the exact cellular source in the lower respiratory tract has not yet been identified. SLPI, also known as antileukoprotease, is a 12-kD, nonglycosylated, cationic protein that is produced by serous cells of the submucosal bronchial glands, by nonciliated cells of the bronchial epithelium, and by neutrophils (2, 6). Its major physiologic function is considered to be the inhibition of NE, but it is also a potent inhibitor of a variety of other proteinases, including cathepsin G and tryptase. Similar to SLPI, elafin also inhibits NE but has also been shown to inhibit an additional neutrophil-derived proteinase, proteinase-3 (6). Besides these shared antiproteinase activities, elafin and SLPI have antibacterial activities (6).

Little is known about the regulation of SLPI and elafin expression within the lung. Studies with both airway epithelial cell lines and primary cultures have indicated that these cells express elafin and SLPI. This expression can be increased by proinflammatory stimuli such as tumor necrosis factor (TNF)- and interleukin (IL)-1 and by 12-O-tetradecanoylphorbol acetate (PMA) (4, 5, 12). Levels of elafin messenger RNA (mRNA) and protein are generally lower in lung-derived cell lines compared with SLPI. We have previously reported that both elafin and SLPI mRNA are constitutively expressed by primary human type II cells in culture (13). These findings suggest that the pulmonary epithelium may respond to inflammatory stimuli by increasing its antiproteinase shield. The regulation of elafin and SLPI may also be directly influenced by inflammatory neutrophils. During inflammation, neutrophils release serine proteinases into the extracellular lining fluid. NE has been shown to increase elafin and SLPI mRNA expression in lung epithelial cells in vitro (5, 14). These findings suggest that components of the inflammatory lung milieu may also play roles in regulating the expression and secretion of elafin and SLPI in the lung in vivo.

Elafin and SLPI are colocalized (within 75 kB) on chromosome 20q12-13.12, where they reside in close proximity to two rapidly evolving seminal vesicle-transcribed (REST) proteins Semenogelin I and II (6). REST proteins are transglutaminase substrates, as is pre-elafin. REST genes and pre-elafin have recently been proposed to be members of the trappin gene family, where trappin stands for "transglutaminase substrate and WAP domain containing protein," and refers to the property of these proteins of becoming "trapped" in a tissue and functioning as an anchored protein (6). It is considered that elafin arose by exon shuffling where exons are derived from both a REST gene and an ancestral WAP gene. SLPI contains two tandemly repeated WAP domains.

Limited studies have previously been undertaken to study the regulatory mechanisms that govern the expression of the elafin gene. Most of what is known derives from studies in breast cell lines. Elafin is expressed in normal mammary epithelial cells, but not in most breast carcinoma-derived cell lines (9). Transient transfection analysis identified a positive regulatory element between positions 575 and 434 of the proximal promoter (9), containing an activator protein (AP)-1 site that conferred high transcription rates in normal mammary epithelial cells and PMA-inducible transcription in carcinoma cell lines (15). In keratinocytes, induction of elafin has been shown to be linked to differentiation, and it has recently been shown that TNF- and serum-mediated induction of elafin expression in these same cells is mediated by a p38 mitogen-activated protein kinase-dependent pathway (16). NE has recently been shown to induce expression of the elafin proximal promoter in a pulmonary cell line (14). In the present study we analyzed the molecular basis on which the elafin gene proximal promoter is regulated by the proinflammatory cytokines TNF- and IL-1 in airway epithelial cells, and identified a role for nuclear factor (NF)-B, acting through a response element in the proximal promoter, in this process.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of Chimeric Reporter Genes and Expression Plasmids

Specific regions of the human elafin gene promoter were amplified by polymerase chain reaction (PCR) using oligonucleotide primers containing 5' SacI sites and 3' BamH1 sites. The 5' extent of the oligonucleotides corresponded to 749, 505, 448, 248, and 105 relative to the published transcription start site (9). All constructs terminated at the same 3' position, +30 relative to the transcription start site. All PCR products were subcloned into the SacI and BglII site of pGL-3 Luc basic (Promega, Southampton, UK). The human IL-8 minimal promoter construct in pGL-3 Luc and a human NF-B-inducing kinase (NIK) complementary DNA (cDNA) in pCDNA3 were provided by Dr. Endre Kiss-Toth (University of Sheffield, Sheffield, UK) (17). The human RelA expression plasmid in pCDNA3 was a gift of Dr. Franco Carlotti (University of Sheffield) (18). A human SLPI promoter construct containing 1,228 base pairs (bp) of the proximal promoter in pGL-2 Luc basic was provided by Tatsuya Abe (Tohoku Hospital, Sendai, Japan) (19). The irrelevant expression plasmid pCMV-TTF-1, containing the coding region of the lung enriched homeodomain protein TTF-1, was used as a negative control. Site-directed mutagenesis of reporter constructs was performed using overlapping PCR primers and introduced the nucleotides CTC in place of GGG within the putative NF-B site (see Figure 4A). All reporter constructs were sequenced by an ABI377 sequencer to determine orientation and to confirm the fidelity of the amplification.

View larger version (50K): [in this window] [in a new window]   Figure 4.   (A) The sequence of the elafin gene proximal promoter region used in the gel mobility shift analysis. The putative NF-B binding site is aligned with the NF-B consensus site used as a competitor. The mutation used in the gel shift studies and in the generation of the mutated reporter construct is shown in bold above the sequence. (B) Gel mobility shift analysis of nuclear extracts from control A549 cells (lanes 1-3) or A549 cells treated with IL-1 (lanes 4-6) or TNF- (lanes 7-9) using a labeled probe corresponding to 105 to 78 of the elafin gene proximal promoter. Reactions were performed in the absence (lanes 1, 4, and 7) or presence (lanes 2, 3, 5, 6, 8, and 9) of a 100-fold excess of unlabeled specific (lanes 2, 5, and 8) or nonspecific (lanes 3, 6, and 9) competitor oligonucleotide. Nonspecific complexes are indicated by the asterisk. (C) Gel mobility shift analysis of Oct-1 binding in control or IL-1- or TNF--stimulated A549 cell nuclear extracts. Reactions were performed in the absence (lanes 1, 3, and 5) or presence (lanes 2, 4, and 6) of a 100-fold excess of wild-type binding site oligonucleotides. (D) Gel mobility shift analysis was performed with TNF--stimulated A549 cell nuclear extracts using the NF-B site as a probe in the absence (lane 1) or presence of a 100-fold excess of either unlabeled wild-type (lane 2) or mutant (lane 3) binding site, and unrelated NF-B binding site (lane 4) or a nonspecific competitor (lane 5). Further reactions were incubated with 1 µl of antibodies to p65 (lane 6) or p50 (lane 7) or a nonspecific antibody (anti-Oct-1) (lane 8) before resolution of the gel. The supershifted complexes are indicated by the A and B. (E) Gel mobility shift analysis was performed with in vitro-translated RelA in the absence (lane 1) or presence of a 100-fold excess of unlabeled wild-type (lane 2), mutant (lane 3), or nonspecific (lane 4) competitor oligonucleotide.

Isolation and Culture of Primary Human Alveolar Epithelial Type II Cells

Human lung tissue was obtained after lobectomy for carcinoma of the lung. Normal regions distal to the site of the tumor were used. The cells were isolated as described (20). Briefly, tissue was lavaged with 0.15 M NaCl using a 20-ml syringe and 21-gauge needle to remove alveolar macrophages and blood cells, until the tissue was gray in appearance. The tissue was then inflated with 0.25% trypsin in Hanks' balanced salt solution (HBSS) and incubated in a petri dish for 45 min at 37°C. Trypsin was replaced as necessary. The tissue was then chopped finely in the presence of normal calf serum. Deoxyribonuclease (250 µg/ml HBSS) was added to the minced tissue suspension and shaken vigorously for 5 min at room temperature. The suspension was filtered through coarse-, then fine-grade mesh to yield single cells. The type II (TII) cell-enriched filtrate was centrifuged at 300 × g for 10 min. Contaminating macrophages and fibroblasts were removed by adherence for 1 h in serum-free medium and then 10% serum, respectively, at 37°C. The TII cell-enriched supernatant was centrifuged as before, and the cell pellet was suspended in 10% normal calf serum in low-protein hybridoma media with antibiotics and plated out at 0.5 million cells/well of a Vitrogen-coated 24-well tissue culture plate. The cells were allowed to reach confluence (approximately 3 d) before study. The cells were characterized as TII on the basis of their morphology, alkaline phosphatase activity, and expression of surfactant protein (SP)-A and SP-C (13).

Tissue Culture, Transient Transfections, and Reporter Gene Assays

A549 and NCI-H441 cells were obtained from American Tissue Culture Collection (Manassas, VA) and maintained as monolayers in RPMI medium containing 10% (vol/vol) fetal calf serum, 100 U/ml penicillin, and 150 µg/ml streptomycin. NCI-H322, NCI-H358, and NCI-H226 cells were a gift of Professor J. Carmichael (University of Nottingham, Nottingham, UK) and maintained as for the other cell lines.

For RNA expression studies and nuclear extract preparation, subconfluent monolayers of cells were used either untreated or treated overnight with IL-1 (20 ng/ml), TNF- (20 ng/ml), or lipopolysaccharide (LPS) (100 ng/ml) before RNA extraction and/ or nuclear extract production (see later text). Stimulations for RNA analysis were performed on multiple occasions.

Cells were transfected at about 60% confluency in 12-well plates, using 200 ng of luciferase reporter constructs and 50 ng of pTK renilla (Promega) as an internal control. Total DNA was made up to 500 ng with the appropriate carrier DNA and 1 µl of lipofectamine (Life Technologies, Paisley, UK) in Optimem serum-free medium (Life Technologies). Transfection mixtures were placed on the cells for 6 h, after which time the transfection mixture was removed and replaced with complete growth medium. In cotransfection experiments, 20 ng of pCDNA3-RelA, pCDNA3-NIK, pCMV-TTF-1, or empty pCDNA3 were added to the reaction mixtures in place of a corresponding amount of carrier DNA. By the use of cotransfected enhanced green fluorescent protein in representative assays (17) we were able to determine that transfection efficiency was > 20%. In the stimulation experiments, wells were treated for the last 16 h with IL-1 (20 ng/ml) or TNF- (20 ng/ml) before the assay. Control cells were mock-treated with a change of medium. Luciferase assays were performed 40 h after transfection according to the manufacturer's instructions (Promega) and the results corrected for internal renilla luciferase activity. All assays were performed on multiple occasions using at least triplicate wells in each assay.

RNA Isolation and Expression Studies

Total RNA was isolated from human lung-derived cell lines using the RNAeasy system (Qiagen, Crawley, UK). Samples were resolved on denaturing agarose gels, Northern-blotted, and hybridized with random-primed cDNA probes as previously described (21). The human elafin cDNA probe was a gift of Dr. Joost Schalkwijk (University Hospital, Nijmegen, The Netherlands) (8).

Nuclear Extract Production, In Vitro Translation, and Gel Mobility Shift Assays

Subconfluent monolayers of cells, either control or cytokine-treated (see earlier text), were used for the generation of nuclear extracts as previously described (22). Aliquots of extracts were frozen at 80°C and used after only a single thaw cycle to avoid degradation of proteins. In vitro transcription and translation of the human RelA expression construct was performed using the TnT coupled system (Promega) primed with T7 RNA polymerase (for sense translation) and SP6 RNA polymerase (for antisense translation). For gel mobility shift experiments, double-stranded oligonucleotides were annealed and labeled by filling in with [³²P]deoxycytidine triphosphate and Klenow polymerase. The probe used in the assay corresponded to nucleotides 101 to 78 of the human elafin promoter, 5'-GCCTGAGGGAAAGCCCCCAGGTCCC-3'. The italicized G was added to improve probe labeling. The following probes were also used as double-stranded competitors (bold type indicates the mutant residues): 101Mut1, 5'-GCCTGACTCAAAGCCCCCAGGTCCC-3'; 101Mut2, 5'-GCCTGAGGGAAAGTATCCAGGTCCC-3'; NF-B consensus, 5'-TTTGACAGAGGGGACTTTCCGAGAGGAAA-3'. The nonspecific competitor was a hepatocyte leukemia factor binding site (23). For standardization of the protein extracts we used a wild-type octamer binding site from the human TTF-1 gene (24). Approximately 1 ng of purified probe (specific activity > 10⁷ counts per min/µg) was used per binding reaction. Binding reactions were performed at room temperature as described (22) in a Ficoll-based binding buffer, using 1 µl of nuclear extracts or 1 µl of in vitro-translated protein. Unlabeled competitor oligonucleotides were included at 100-fold excess to the probe. In some reactions the competitor oligonucleotides were replaced with specific antisera to p65 or p50 proteins (Santa Cruz Biotechnologies, Santa Cruz, CA) or an unrelated (Oct-1) antibody. At the end of the binding reaction, samples were resolved on 4% polyacrylamide minigels, dried, and exposed to film at 80°C.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Constitutive and Cytokine-Induced Expression of the Human Elafin Gene Is Limited to a Subset of Pulmonary-Derived Epithelial Cell Lines

By the use of Northern blot analysis we could show that elafin mRNA was constitutively expressed in primary TII cells (Figure 1, lane 1). Levels were found to be upregulated by IL-1 and TNF treatment (Figure 1, lanes 3 and 4), whereas LPS treatment had no effect (Figure 1, lane 2). Due to the difficulty in obtaining primary TII cells in sufficient numbers and the difficulties associated with transfection of primary pulmonary cells, we searched for cell lines in which we could analyze cytokine-mediated transcription of the human elafin gene. For this we screened a number of human lung-derived cell lines for constitutive and IL-1- and TNF--induced elafin gene expression. Low levels of elafin mRNA were dectable in H322 (Figure 1, lane 5) and A549 (Figure 1, lane 8) cells. After IL-1 and TNF- stimulation of the cells there was a marked induction of expression of elafin in the H322 and A549 cells (Figure 1, lanes 6, 7, 9, and 10). Constitutive elafin expression was not seen in H441, H226, and H358 cells (results not shown). These results extend previously published observations that cytokines induce expression of elafin in a limited subset of lung-derived cell lines and suggest that these lines would be useful for further analysis of the regulatory mechanisms that govern its expression.

View larger version (69K): [in this window] [in a new window]   Figure 1.   Northern blot analysis of elafin mRNA expression was performed as described in MATERIALS AND METHODS using total RNA isolated from primary human TII cells (lanes 1-4) and pulmonary cell lines (H322 cells [lanes 5-7] and A549 [lanes 8-10]), either unstimulated or cytokine-stimulated. Control, lanes 1, 5, and 8; LPS, lanes 2, IL-1, lanes 3, 6, and 9; TNF- lanes 4, 7, and 10. Blots were hybridized with 32P-labeled cDNA elafin probe as described (upper), and loading was confirmed using ribosomal RNA visualization after ethidium bromide staining of the gel (lower).

The Proximal Promoter of the Human Elafin Gene Is Transcriptionally Active in Pulmonary Epithelial Cells

To search for the transcriptional mechanism that governs the cytokine-mediated induction of the elafin gene, we analyzed the proximal promoter (Figure 2A) in a series of transient transfection experiments. A deletion series of reporter constructs was generated and analyzed by transfection into the elafin-expressing cell line (A549). This line was chosen on the basis of the expression studies presented in Figure 1 and the ease with which they could be transiently transfected. Constructs containing 749, 505, and 448 of the proximal promoter were essentially as active (Figure 2B), suggesting that the AP-1 and putative octamer sites located within these regions (Figure 2A) play little functional role in the basal regulation of the elafin gene in pulmonary epithelial cells. Removal of the region from 448 to 248 led to a reduction in reporter gene activation, suggesting that this region may contain a positive-acting element (Figure 2B). Removal of the region between 248 and 105 caused a marked drop in reporter gene expression (Figure 2B), suggesting that a strong positive-acting element is located within this region. Similar results were found in studies performed in the non-elafin expressing cell line H441 (results not shown). These results suggest that although the elements that regulate the cell type-specific expression of the elafin gene reside outside of the proximal promoter, important regulatory elements are present within this region. We therefore performed experiments to see whether cytokine-mediated induction of elafin expression involves any of these elements.

View larger version (24K): [in this window] [in a new window]   Figure 2.   (A) Schematic representation of the deletion constructs used in the study corresponding to 749, 505, 448, 248, and 105 to +30 relative to the transcription start site. The locations of the previously identified putative transcription factor binding sites are identified. (B) Transient transfection analysis of elafin gene reporter constructs was performed in A549 cells as described in MATERIALS AND METHODS. At 40 h after transfection, luciferase assays were performed and the results, relative to the activity of the 505 Luc construct, are presented as means +/ standard error of the mean (SEM) (n = > 8).

IL-1 and TNF- Induce Expression of the Elafin Gene Proximal Promoter

As an initial experiment we transfected the longest elafin promoter construct that contains the AP-1 site, shown to be important in PMA induction of elafin gene expression in mammary cells (15), into A549 cells. We then stimulated the cells for 16 h with IL-1 or TNF- before assaying for reporter gene activity. As a positive control we used the cytokine-responsive IL-8 minimal promoter. In view of the observation that the SLPI gene is also induced by cytokine treatment in this cell line (5), we also transfected cells with a human SLPI gene reporter construct corresponding to 1,228 bp 5' of the transcription start site. Cytokine stimulation of cells transfected with the IL-8 reporter construct gave rise to 9.5-fold (IL-1) and 6-fold (TNF-) inductions of activity, confirming that these cells could respond to these mediators in the expected manner (Figure 3). When cells transfected with the 749 elafin reporter gene construct were stimulated with IL-1 or TNF- we could show that activity was increased by 10-fold (IL-1) and 5.5-fold (TNF-). In contrast, we could not show an activation of expression of the SLPI gene construct in the same cells (Figure 3). Although the regulation of SLPI expression was not the prime focus of this study, these results suggest that the cytokine-mediated induction of SLPI expression (which is less marked than that seen with elafin [5, 12, 13]) is mediated by elements outside of the proximal promoter. Empty pGL3-luc expression was not induced by stimulation with either cytokine (results not shown). These results suggest that the cytokine-mediated induction of elafin expression is mediated, at least in part, by elements present within the proximal promoter. Additional transfections performed with the deletions of the promoter in the presence of IL-1 or TNF- clearly show that all the constructs tested from 749 to 105 are responsive to both cytokines (Figure 3), although the induction seen in the 105 construct was less marked than that seen with the longer constructs. We conclude from this that the region from 105 to +30 contains a functional cytokine-responsive element.

View larger version (24K): [in this window] [in a new window]   Figure 3.   Transient transfection analysis of elafin, SLPI, and IL-8 gene reporter constructs were performed in A549 cells as described in MATERIALS AND METHODS. At 24 h after transfection, the cells were stimulated with either IL-1 or TNF-, or with fresh medium as a control. After a further 16-h incubation, luciferase assays were performed as described in MATERIALS AND METHODS. Results are expressed as relative induction compared with each individual construct, as means ± SEM n = 18-19).

NF-B Binds to an Element within the Human Elafin Gene Proximal Promoter

Previous sequence analysis (9) suggested that an NF-B binding site is located at 90 to 78 in the elafin gene proximal promoter, and we chose to study whether this region of the gene was responsive to cytokines. To directly test whether the putative NF-B binding site (shown in Figure 4A) was able to bind to nuclear proteins it was used as a labeled probe in gel mobility shift studies with nuclear extracts isolated from control and cytokine-stimulated A549 cells. When nuclear extracts prepared from control A549 cells were interacted with the labeled probe, little specific binding was noted (Figure 4B, lanes 1-3). However, extracts from cells stimulated overnight with IL-1 or TNF- contained more proteins than were capable of binding to this probe (Figure 4B, lanes 4 and 7). The specificity of this interaction was shown by the observation that the complex was competed with an excess of unlabeled binding site (Figure 4B, lanes 5 and 8) but not by an excess of an unlabeled irrelevant competitor site (Figure 4B, lanes 6 and 9). The specificity of this cytokine induction of binding was shown by the fact that extracts from all three samples bound to an Oct-1 binding site probe equally (Figure 4C).

To identify proteins present within the induced extracts, we performed a further gel shift assay with the TNF--stimulated extracts. In this assay the two major bands in the retarded complex were competed with an excess of unlabeled probe (Figure 4D, lane 2) but not by an excess of an oligonucleotide in which the putative NF-B site had been mutated (Figure 4D, lane 3). A further unrelated NF-B site also served as an effective competitor (Figure 4D, lane 4), whereas a second mutant of the elafin probe did not (results not shown). An additional specific band of lesser mobility was also formed with the same probe (Figure 4D, indicated by the C). This band was competed equally well for the wild-type and mutant elafin oligonucleotides as well as by the NF-B consensus, but not by the nonspecific oligonucleotide, suggesting that this complex was due to the interaction of an additional protein with the probe. To confirm that NF-B-related proteins were present within the retarded complex, we used specific antiserum in supershift studies. Addition of antiserum to the p65 and p50 components of the NF-B complex produced supershifts (indicated by the A and B in Figure 4D, lanes 6 and 7) that were not generated by the addition of an unrelated antibody (anti-Oct-1) to the reaction (Figure 4D, lane 8). As with the oligonucleotide competition studies, the slower migrating band was not influenced by the addition of any of the antibodies tested. As further confirmation that NF-B proteins could bind to this probe we could also show that in vitro-translated RelA was able to form a complex with the same specificity on the elafin promoter probe (Figure 4E). These results clearly show that NF-B was able to bind to the elafin probe. Next, we wished to test whether the cytokine-mediated induction of transcriptional activity of the elafin proximal promoter was mediated through this interaction.

Cotransfected RelA and Cytokine Stimulation Induce Expression of the Elafin Promoter through the Proximally Located NF-B Binding Site

To test whether NF-B was a direct inducer of elafin gene expression, we performed a further series of transfection studies. When the 749 elafin reporter construct was cotransfected with an expression plasmid for RelA, the transcriptionally active component of the NF-B complex, activity was significantly induced compared with activity when the reporter was transfected with either empty pCR3.1 (Figure 5) or with the irrelevant expression construct pCR3.1 TTF-1 (results not shown). As seen in the cytokine-stimulation experiments, the magnitude of induction was more marked than that seen with the IL-8 minimal promoter construct. Cotransfections performed with NIK, which induces degradation of I-B and results in increased activation via NF-B, gave similar results (results not shown). These results clearly show that the elafin proximal promoter is a direct transcriptional target of NF-B. Cotransfection of the 105 construct with RelA confirmed that the minimal promoter region, containing the NF-B binding element, was responsible for this induction of reporter gene activation (Figure 5). As a final confirmation that this site was functionally significant we generated a site-directed mutation of the sequence, identical to that used in the gel shift studies, within the context of the 248 construct (Figure 6A). Direct comparison of the basal activity of the wild-type and mutant constructs showed that this mutation led to an increase in reporter gene activity (Figure 6B). This mutation was, however, able to block both RelA- and cytokine-mediated induction of the promoter activity (Figure 6B). These results clearly suggest that the cytokine-mediated increase in elafin expression is regulated by the proximal NF-B binding site.

View larger version (28K): [in this window] [in a new window]   Figure 5.   Transient cotransfection analysis of elafin and IL-8 gene reporter constructs were performed in A549 cells as described in MATERIALS AND METHODS in the presence or absence of 20 ng of pCDNA RelA or appropriate controls. Luciferase assays were performed after 40 h as described in MATERIALS AND METHODS. Results are expressed as relative induction compared with each individual construct as means ± SEM (n = > 8).

View larger version (20K): [in this window] [in a new window]   Figure 6.   (A) Schematic representation of the wild-type and mutant elafin gene reporter constructs showing the mutated region. The relative position of the Sp1/Sp3 binding site is indicated. Constructs were generated in the 248-bp backbone. (B) Transient transfection and cotransfection analysis of wild-type and mutant elafin gene proximal promoter constructs were performed in A549 cells as described in MATERIALS AND METHODS. In the stimulation experiments, 24 h after transfection the cells were stimulated with either IL-1 or TNF- or with fresh medium as a control. After a further 16-h incubation, luciferase assays were performed as described in MATERIALS AND METHODS. The cotransfection experiments were performed in the presence or absence of 20 ng of pCDNA RelA or appropriate controls. After a 40-h incubation, luciferase assays were performed as described in MATERIALS AND METHODS. Results are expressed as relative induction compared with each individual construct as means ± SEM (n = 9).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

A variety of proteinase inhibitors are found within the lung lining fluid and these provide a defense mechanism against the activity of serine proteinases produced by inflammatory cells, principally neutrophils, present within the lung. Elafin, along with the related inhibitor SLPI, and 1-PI are the major serine proteinase inhibitors found within the lung lining fluid (1, 2). In inflammatory situations, where the proteinase burden is increased due to the increased numbers of infiltrating cells, the pulmonary system counters this threat by increasing production of inhibitors. In certain situations local inhibitor production is insufficient to counter the proteinases, resulting in destruction of the lung parenchyma and a loss of lung function (1). Elafin has previously been shown to be expressed in multiple epithelia (10) and to be induced by a variety of inflammatory mediators (5, 6), but the molecular mechanisms that govern its expression are unresolved. In the present study we have shown that the elafin gene is constitutively expressed in primary human TII cells in culture and is induced by IL-1 and TNF. We further undertook to study the mechanisms regulating the constitutive and inducible expression of the human elafin gene in the TII cell-like airway epithelial cell line A549. Our studies have shown that constitutive expression of elafin is limited to a subset of airway epithelial cells. This is in contrast to the constitutive expression of the related antiproteinase SLPI that is found in a wider set of cells, including those that express elafin. This observation suggests that the molecular regulation of the elafin gene may involve transcriptional regulatory mechanisms that overlap those which regulate the SLPI gene.

Our initial transfection experiments performed in elafin-expressing and -nonexpressing cell lines suggest that the two lines were able to support similar levels of reporter gene activity. This suggests that regions of the elafin gene responsible for the cell type specificity of expression reside outside of the proximal promoter region that we studied. The identification of the exact regions that mediate this cell type-specific expression requires further study. However, such observations have frequently been made with other genes expressed within the pulmonary epithelium. For example, the promoters of the Clara cell secretory protein and SP genes are active in cell lines which do not express the endogenous gene products (22, 25). But notwithstanding this observation, such lines have proved useful for the molecular analysis of such genes. By analyzing a deletion series of promoter constructs in the elafin-expressing cell line A549, we were able to identify two positive-acting regions within the proximal promoter. One was located between 448 and 248 bp and the second, stronger site was located between 248 and 105 bp. These results are significantly different from a previously published analysis of the same promoter performed in mammary epithelial cells (9). In these earlier studies a strong regulatory region was mapped to between 505 and 450 bp. Removal of this region reduced reporter activity to background (9). The authors subsequently showed that this region contained an AP-1 site which was critically required for PMA-mediated induction of the elafin gene in mammary cells (15). Our results suggest that this AP-1 site does not play a significant role in the regulation of constitutive and induced elafin expression in pulmonary epithelial cells. This observation suggests that different molecular mechanisms may govern the regulation of elafin gene expression in these two cell types. Consistent with this, we have been unable to show induction of elafin mRNA in pulmonary epithelial cells by PMA treatment (results not shown).

It has previously been shown that elafin expression in pulmonary epithelial cells is regulated by a variety of inflammatory mediators, including IL-1, TNF-, and NE (5, 14). Our results clearly demonstrate that the AP-1 site in the elafin proximal promoter is not responsible for the cytokine-mediated induction of elafin expression in airway epithelial cells but rather that this effect is mediated through a functional NF-B site located within 100 bp of the transcription start site. We have shown that this site binds NF-B and is transactivated by both RelA and NIK cotransfection. Mutation of this site completely abolishes the inducibility of the promoter. Several other cytokine- inducible gene products have been shown to be regulated by NF-B in an analogous manner in airway epithelial cells (26). Although we have not directly tested the effect of NE in this system we would expect that NE-induced activation of the elafin promoter would be regulated through this site. A previous study of the elafin proximal promoter in airway epithelial cells has shown that activation occurs through the region 505 to +30 (14), which contains both the AP-1 and NF-B sites. NE treatment of airway epithelial cells has been shown to cause structural changes in cells (30, 31) and to induce expression of a variety of proinflammatory mediators, including IL-6 and IL-8 (32). It is therefore likely that the effect of NE will be mediated through the NF-B site. The functional consequence of induction of elafin gene expression by inflammatory mediators is that during pulmonary inflammation, levels of elafin will be increased to serve a protective role within the lung. Levels of elafin in lung lavage have been shown to be increased in patients with a variety of hypersensitivity pneumonitis (11), and levels of the related inhibitor SLPI have been shown to be increased in experimental bacterial pneumonia (33, 34).

One interesting finding from our analysis of the NF-B site in the elafin proximal promoter is that when we abolished the ability of the region to bind to NF-B by mutation, the resultant promoter construct had a higher constitutive activity (Figure 6B). Our interpretation of this observation is that mutation of the site caused an alteration in the affinity of additional binding sites surrounding the NF-B site. Our gel shift studies identified an additional non-NFB-related band formed on the probe (Figure 4D, indicated by the C) and this protein may be responsible. In further gel mobility shift experiments we show that a functional Sp1/Sp3 binding site is located immediately 3' of this region and partially overlaps, and that other, as yet uncharacterized, DNA binding proteins interact with the region 5' of the site (results not shown). Further studies will be required to show whether these additional sites are able to influence the cytokine-mediated induction of the elafin gene.

In summary, we have shown that the TNF-- and IL-1-mediated induction of elafin gene expression is regulated by a functional NF-B site located within the first 100 bp of the proximal promoter. Our results suggest that it may prove possible to selectively upregulate expression of the endogenous elafin gene as a potential therapeutic tool in the treatment of inflammatory lung disease.