Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Human 12/15-lipoxygenase (EC 1.13.11.33) is a highly regulated, lipid-peroxidating enzyme that is implicated in the pathophysiology of atherosclerosis and airway inflammation (1). The biologic roles of 12/15-lipoxygenase result from both direct enzymatic activity on cellular membranes and the generation of bioactive lipids, including the mono-hydroxyeicosatetraenoic acids (mono-HETEs), di-HETEs, and lipoxins.

Recent studies in human systems have implicated the 12/15-lipoxygenase in the pathogenesis of atherosclerosis through its ability to generate membrane lipid peroxides in cellular membranes that subsequently oxidize low-density lipoprotein into an atherogenic form (1, 2, 5, 6). Although most studies have implicated human 12/15-lipoxygenase in the pathophysiology of atherosclerosis, at least one demonstrated that human 12/15-lipoxygenase may be protective in a rabbit model of atherosclerosis (7). In both rabbit and human reticulocytes, there appears to be homologous 15-lipoxygenase activity that results in reticulocyte maturation through the selective oxygenation of mitochondrial membranes and their respiratory enzymes, which, in turn, results in mitochondrial degradation (8).

Although specific roles of the human 12/15-lipoxygenase in modifying airway inflammation have yet to be established in vivo, the ability of 12/15-lipoxygenase to generate lipid peroxides in the plasma membranes suggests that the expression of the enzyme in ciliated airway cells may augment nonspecific host defense mechanisms (4, 11). Many studies have demonstrated important physiologic functions of human 12/15-lipoxygenase metabolites that are relevant to airway inflammation (3, 12). In certain cases, these metabolites interact with specific cellular receptors. Lipoxin A4 is generated by oxygenation of arachidonic acid by both 5- and 12/15-lipoxygenases and is produced by alveolar macrophages and airway epithelium (13). The induction of an epithelial cell Lipoxin A4 receptor by interleukin (IL)-13 demonstrates that cytokines not only induce specific receptors but may also induce enzyme systems that participate in ligand production for these receptors (16, 17). Similarly, in murine monocytes IL-4 induces expression of the nuclear hormone receptor peroxisome proliferator-activated receptor  (PPAR) and a homologous 12/15-lipoxygenase. In this study, activation of PPAR responses occurred in a lipoxygenase-dependent manner, suggesting an example in which IL-4 induces a specific receptor, PPAR, and an enzyme that generates ligands for that receptor (18).

Human 12/15-lipoxygenase has a specific pattern of expression. Earlier studies demonstrated that IL-4 and IL-13 can induce the enzyme in cultured human peripheral blood monocytes (17, 19). The induction of human 12/15-lipoxygenase by IL-4 in airway epithelial cells was recently confirmed (4, 20, 21). In murine macrophages, a homologous 12/15-lipoxygenase can be induced by IL-4 and this induction is dependent on signal transducer and activator of transcription (Stat)-6 expression (22).

In the present studies, we demonstrate that human 12/ 15-lipoxygenase is induced in nontransformed primary cultures of airway epithelial cells and in the transformed BEAS-2B airway epithelial cell line through an IL-4/Stat6-dependent process. This induction results in increased 12/15-lipoxygenase gene transcription, messenger RNA (mRNA), protein, and enzymatic activity and is dependent upon a Stat6 response element located 952 bases upstream of the translational start codon in the 5'-flanking sequence of the human gene.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture, Reagents, and Vectors

BEAS-2B and HEK-293 cells were obtained from American Type Culture Collection (Rockville, MD) and cultured in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 µg/ml). Cells were maintained at 37°C in 5% CO₂. Normal human bronchial epithelial cells were obtained from Clonetics (San Diego, CA) and maintained in small airway growth medium. IL-4 (2.9 × 10⁷ U/mg) was purchased from R&D Systems (Minneapolis, MN). The expression vector for human Stat6 was a generous gift from William LaRochelle (National Cancer Institute, Bethesda, MD) (23, 24). The -actin/-galactosidase (-gal) vector was obtained from Christopher Glass (USCD, San Diego, CA). The human platelet 12-lipoxygenase expression vector was obtained from Colin Funk (University of Pennsylvania, Philadelphia, PA). All vectors were purified over anion exchange resin columns (Qiagen, Santa Clarita, CA). Purified human 15-lipoxygenase-2 was generously supplied by Alan Brash (Vanderbilt University, Nashville, TN). The -actin antibody was obtained from Sigma (St. Louis, MO).

Immunoblotting

Western blot analysis was performed as previously described, with some modifications. BEAS-2B cells were transfected using standard calcium phosphate precipitation protocols (25). The cells were washed after 6 h and then stimulated with IL-4 (10 ng/ml) for 48 h. Protein extracts (50 µg/lane) were resolved on 10% sodium dodecyl sulfate-polyacrylamide gels. After transferring the proteins to a nitrocellulose filter, human 12/15-lipoxygenase was detected using standard techniques and a rabbit antibody to a human 12/15-lipoxygenase-CheY fusion protein (26). Stat6 was detected using identical protocols and a rabbit polyclonal antibody to human Stat6 (Zymed, South San Francisco, CA).

Enzymatic Activity Assays and Northern Analysis

Enzymatic activity assays were performed as previously described (27). Briefly, cells (5-10 × 10⁶ cells/assay) were incubated in phosphate-buffered saline for 15 min at 37°C in the presence of arachidonic acid (300 µM). Prostaglandin B₂ was added as a control for extraction efficiency. After extracting the free fatty acids, they were separated on a C18 reverse-phase high-performance liquid chromatography (HPLC) column (25 cm) as described. The assay utilizes a Rainin Rabbit HPLC system coupled to a Knauer UV multiple wavelength detector. The limit of detection in this system was approximately 30 to 40 ng of 12- or 15-HETE per assay. The peaks that comigrated with authentic 15S- and 12S-HETE were integrated using Rainin Dynamax software (Walnut Creek, CA). A no-cell control was used to assess nonspecific oxygenation of the arachidonic acid and was subtracted from all conditions. The data are expressed as mean specific activity of three separate cultures for each condition and are reported with a standard error of the mean.

Northern assays were performed as previously described (19). Briefly, 30 µg of total RNA were isolated from Stat6-transfected cells cultured for 3 d in the presence or absence of IL-4 (10 ng/ml). The RNA was separated on a 1% agarose/formaldehyde gel and blotted onto a nylon membrane (Hybond; Amersham, Piscataway, NJ). A human 12/15-lipoxygenase complementary DNA (cDNA) was labeled as described and hybridized to the nylon membrane (3 × 10⁶ counts per min [cpm]/ml of hybridization solution). The membrane was washed as described and standard autoradiography techniques were used to develop the signal. A full-length human -actin cDNA was similarly labeled and hybridized to the filter. The hybridization signal was used as an RNA loading control.

Reporter Gene Vector Construction

A human 12/15-lipoxygenase promoter/luciferase reporter gene was constructed for the transcriptional studies. Using the genomic fragment, 10:11, as a template and primers to the genomic sequence (upstream oligonucleotide 5'-TTAACGCGTGCGTTTAGTTGGAGG-3'; downstream oligonucleotide 5'-GGAACAGGCCTAGAGCGAGGCCCCAGTGGACACGCGGATGCGGTAGAGACAGATCTTGCT-3'), a 459-base pair (bp) fragment was amplified and cloned into pCR-Script (Stratagene, La Jolla, CA) (28). This fragment contains 411 bp of the proximal promoter and 48 bases of the first exon. The amplified fragment also contains a mutation in the native translation initiation codon. The amplified sequence used in this vector was directly sequenced to confirm its identity to the published sequence except for the introduced mutation. All sequencing performed in these studies was completed by the Core Facility of the UCSD Center for AIDS Research (San Diego, CA) using an automated fluorescent DNA sequencer. To this sequence, an additional 4.0 kb of additional 5'-flanking sequence was subcloned upstream of the amplified product. The entire 4.5 kb of 5'-flanking region was then subcloned into the SacI/HindIII sites of pGL3 basic (Promega, Madison, WI) to create the vector LOP4.5 (Figure 1). Vectors LOP3.0, LOP.2, and LOP.05 were created using the XhoI, EcoRI, and OxaI restriction sites indicated in Figure 1.

View larger version (7K): [in this window] [in a new window]   Figure 1.   Construction of the human 12/15-lipoxygenase promoter/luciferase reporter genes. A 12/15-lipoxygenase promoter/ luciferase reporter gene (LOP4.5) was constructed that included approximately 4.5 kb of 5'-flanking sequence and the initial 48 bases of the first exon. The SacI (S), AflII (A), KpnI (K), XhoI (X), EcoRI (E), and OxaI (O) sites are indicated. The vertical arrow marks the position of the putative Stat6 response element investigated in these studies. The horizontal arrow is located at the transcriptional start site. The thick line indicates the position of published sequence.

A variant of vector LOP3.0 was generated (LOP3.0m) that was identical to the parent vector except that the Stat6 response element 952 bases upstream of the translational stop codon was mutated from 5'-TTCCTGAGAA-3' to 5'-GTCCTGAGGT-3' by site-directed mutagenesis (Quickchange; Stratagene). The Stat6 site was first mutated in a 12/15-lipoxygenase genomic fragment clones into Bluescript. Direct sequencing revealed an unintentional G-to-C exchange 128 bases downstream from the Stat6 response element and a CC insertion in a GC-rich region 145 bases downstream of the Stat6 site. These mutations did not involve the proximal promoter or any potential Stat6 response elements. Using this vector and standard subcloning techniques, LOP3.0m was generated. This 12/15- lipoxygenase promoter/luciferase reporter (parent vector pGL3 basic) gene contains the mutated Stat6 response element in the context of the native promoter.

The Stat6 site was catemerized and subcloned into vector OTLO, a pSP65-based vector that contains the luciferase coding sequence driven by 109 bases of the herpes simplex virus (HSV) thymidine kinase promoter. Briefly, sense and antisense oligonucleotides with flanking 5' XhoI and 3' SalI restriction sites were annealed, ligated, and subcloned into the XhoI site of Bluescript (Stratagene, San Diego, CA). One clone with a 70-bp insert was directly sequenced and found to have four direct repeats of the Stat6 response element. This insert was subsequently cloned into the SalI site of OTLO.

Transient Transfections

BEAS-2B cells were transfected using standard calcium phosphate precipitation protocols (25). Briefly, 5 × 10⁵ cells were plated in 60-mm dishes 1 d before transfection. Calcium phosphate was coprecipitated with the luciferase reporter genes and a -actin promoter/-gal reporter gene. The calcium phosphate precipitates were washed from the cells after 6 h and then IL-4 (10 ng/ml) was added to the media. After 36 h of IL-4 treatment, protein extracts were made and assessed for luciferase (Promega) and -gal (Tropix, Bedford, MA) activity using commercial chemiluminescent kits. The luciferase data are expressed as the means of triplicates per 100,000 -gal units ± standard error. Each study is representative of at least three separate transfections.

Electrophoretic Mobility Shift Assays

Electrophoretic mobility shift assays (EMSAs) were performed as described elsewhere (23, 24). Nuclear extracts were prepared from Stat6-transfected BEAS-2B cells after treatment with IL-4 (250 ng/ml) for 60 min. IL-4-treated and untreated cells (5 × 10⁶) were resuspended in 2 ml lysis buffer (0.1% Nonidet P-40, 20 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid [pH 7.0], 10 mM KCl, 1 mM MgCl₂, 1 mM dithiothreitol, 0.25 mM phenylmethylsulfonyl fluoride, 1 mM orthovanadate, and 20% glycerol). Nuclei were harvested after Dounce homogenization of the cells and resuspended in 500 µl of the lysis buffer containing 500 mM NaCl. The nuclear extracts were then centrifuged at 14,000 rpm for 15 min. Nuclear extract protein was quantitated and stored at 70°C until needed.

Probes for the EMSAs were generated by annealing the sequence 5'-CGGGTAAGACTTTCCTGAGAAACCGGAGGTGAA-3' with an antisense oligonucleotide containing the 12/15-lipoxygenase Stat6 response element with 10 bp of flanking sequence on either side. The double-stranded oligonucleotide was labeled using T4 polynucleotide kinase and [³²P]adenosine triphosphate. The probe was gel-purified on a 5% polyacrylamide gel. After eluting the probe from the gel, it was ethanol precipitated and resuspended in TE buffer. Approximately 100,000 cpm of purified probe were mixed with the thawed nuclear extracts in binding buffer for 15 min at room temperature. The protein complexes were resolved on a 0.5 × Tris- borate, ethylenediaminetetraacetic acid (TBE) nondenaturing 5% polyacrylamide gel. The gels were dried and exposed using standard autoradiography techniques.

Binding specificity was assessed with competition and Stat6 supershift assays. The competition assays were performed as indicated earlier except that the unlabeled native and mutant probes were added to the binding reaction at concentrations that were 100-fold greater than the radioactive probe. The sense mutant probe sequence was 5'-CGGGTAAGACTGTCCTGAGGTACCGGAGGTGAA-3'. Supershift assays were performed by adding 1 and 3 µg of an antihuman Stat6 rabbit polyclonal antibody (Zymed) to the binding reaction.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Stat6 Expression in Human Airway Epithelium

Although Stat6 was previously noted to be expressed in lung tissue, the specific cellular localization has not been reported. Figure 2 demonstrates strong expression of Stat6 in primary cultures of nontransformed human airway epithelial cells and no expression in the transformed airway epithelial cell line, BEAS-2B. BEAS-2B cells are artificially transformed and were generated from normal human airway epithelial cells. They were selected for further studies because of their transfection efficiency and functional IL-4 receptors. Because BEAS-2B cells lack expression of endogenous Stat6, this cell line allowed us to establish clearly the role of this transcription factor in regulating human 12/15-lipoxygenase promoter activity, mRNA, protein, and enzymatic activity. When BEAS-2B cells were transfected with a human Stat6 expression vector (100 ng), the level of expression was comparable to that seen in nontransformed cells (Figure 2). Immunoblot analysis of nuclear extracts of Stat6 expressing BEAS-2B cells showed the predicted IL-4-dependent translocation of Stat6 to the nucleus (D. Conrad and M. Lu, unpublished observations).

View larger version (28K): [in this window] [in a new window]   Figure 2.   Stat6 expression in human airway epithelial cells. Immunoblots were performed using whole-cell extracts (50 µg/lane) from nontransformed human airway epithelial cells (lane 2), BEAS-2B (lane 1) cells, and BEAS-2B cells transfected with a human Stat6 expression vector (lane 3). The nitrocellulose filters were probed with a rabbit antibody to human Stat6 (1:2,000 titer) and chemiluminescent techniques were used to develop the signal. The arrow indicates the position of a 105-kD band that is specific for human Stat6.

Endogenous 12/15-Lipoxygenase Gene Expression in Transformed and Nontransformed Airway Epithelial Cells

12/15-Lipoxygenase expression in BEAS-2B cells occurs in an IL-4-dependent and Stat6-dependent manner (Figure 3). BEAS-2B cells cultured in the presence of IL-4 (10 ng/ml) did not express endogenous 12/15-lipoxygenase unless Stat6 was expressed in the cells. Conversely, BEAS-2B cells expressing Stat6 did not express 12/15-lipoxygenase unless they were treated with IL-4. Finally, immunoblot analysis of nontransformed human airway epithelial cells grown in submersion cultures demonstrated a similar IL-4- dependent induction of the 12/15-lipoxygenase.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Regulation of endogenous airway 12/15-lipoxygenase by IL-4 and Stat6. BEAS-2B cells were transfected with a Stat6 expression vector (250 ng) and treated with IL-4 (10 ng/ml) for 48 h as indicated. Normal human bronchial epithelial (NHBE) cells were cultured for 5 d in the absence or presence of IL-4 (10 ng/ml). Immunoblot analysis of cellular extracts (50 µg/lane) was performed using a primary rabbit antibody to a CheY-human 15-lipoxygenase fusion protein (26). The arrow indicates the position of a specific band that comigrates with a 72-kD human 12/15-lipoxygenase standard (15LO).

In Figure 4, we confirm that the protein induced in BEAS-2B cells by IL-4 and Stat6 was 12/15-lipoxygenase and not human platelet 12-lipoxygenase, human 12(R)-lipoxygenase, or the recently cloned 15-lipoxygenase-2 protein. Figure 4a shows an immunoblot of the IL-4-induced protein in Stat6-transfected BEAS-2B cells (Figure 4a, lanes 6 and 7). This protein comigrated with human 12/15-lipoxygenase that was expressed from a mammalian expression vector in HEK-293 cells (Figure 4a, lane 2). The antibody used in these studies demonstrated no significant immunocrossreactivity with the human platelet 12-lipoxygenase expressed in HEK-293 cells (Figure 4a, lane 1) or with purified human 15-lipoxygenase-2 protein (Figure 4a, lanes 3-5).

View larger version (50K): [in this window] [in a new window]   Figure 4.   Characterization of the IL-4-dependent lipoxygenase in BEAS-2B cells. (a) To determine the specificity of the antibody used in these studies, the rabbit antibody to the human CheY/15-lipoxygenase fusion protein was used in immunoblots with extracts from HEK cells expressing human platelet 12-lipoxygenase with enzymatic activity of 151 ng 12-HETE/106 cells (lane 1); human 12/15-lipoxygenase with enzymatic activity of 184 ng 15-HETE/106 cells (lane 2); 50, 100, and 500 ng of purified 15-lipoxygenase-2 (lanes 3, 4, and 5, respectively); Stat6-transfected BEAS-2B cells cultured without IL-4 (lane 6); and Stat6-transfected BEAS-2B cells cultured with IL-4 (10 ng/ml) for 2 d (lane 7). A total of 10 µg of whole-cell extracts were analyzed in lanes 1 and 2, while 50 µg were used in lanes 6 and 7. The migration of the molecular weight standards are indicated. (b) Lipoxygenase enzymatic activity was determined in Stat6-transfected BEAS-2B cells cultured in the absence (open bars) or presence (crosshatched bars) of IL-4 (10 ng/ml) for 3 d. The data were normalized for cell number and corrected for extraction efficiency as described in MATERIALS AND METHODS. (c) Northern analysis of Stat6-transfected BEAS-2B cells treated with IL-4 as indicated for 3 d. Total RNA was isolated and separated on 1% agarose/ formaldehyde gel and immobilized on nylon hybridization filter as described in MATERIALS AND METHODS. Full-length cDNAs of the human 12/15-lipoxygenase and -actin were labeled and hybridized to the filters.

In addition, enzymatic activity assays demonstrated a marked induction of 15- and 12-HETE in Stat6-transfected BEAS-2B cells cultured in the presence of IL-4. The amount of 15- or 12-HETE generated from Stat6-transfected BEAS-2B cells cultured in the absence of IL-4 was just above the detection limit of this assay (i.e., 30 ng/ 10⁷ cells). If Stat6 was not expressed in the BEAS-2B cells, 15- and 12-HETE were not detected in this assay even when the cells were cultured in the presence of IL-4 (D. Conrad and M. Lu, unpublished observations). The generation of 15- and 12-HETE at approximately an 8:1 ratio is consistent with the enzymatic properties of the human 12/15-lipoxygenase (26). Further, this pattern is not consistent with human platelet 12-lipoxygenase, human 12(R)-lipoxygenase, or the keratinocyte 15-lipoxygenase (29). Finally, in Figure 4c, we used Northern analysis to demonstrate the double-banded hybridization signal (2.7 and 3.9 kb) that is characteristic of the human gene and results from two polyadenylation and cleavage signals in the 3' untranslated region (28).

IL-4-Dependent 12/15-Lipoxygenase Promoter Activity

12/15-Lipoxygenase promoter activity is also dependent upon Stat6 expression in BEAS-2B cells. BEAS-2B cells were cotransfected with the 12/15-lipoxygenase promoter/ luciferase reporter gene LOP4.5 and a -actin promoter/-gal reporter gene to normalize for transfection efficiency. In addition, the cells were transfected with an expression vector for human Stat6 and treated with IL-4 (10 ng/ml) as indicated (Figure 5). These studies, which parallel those in Figure 3, demonstrate that the IL-4-dependent 12/15-lipoxygenase promoter activity requires Stat6 and shows a 20-fold increase in promoter activity with IL-4 treatment. There was no IL-4-dependent 12/15-lipoxygenase promoter activity without the cotransfected Stat6 protein. These assays demonstrate that 12/15-lipoxygenase promoter activity is dependent on both IL-4 and Stat6.

View larger version (13K): [in this window] [in a new window]   Figure 5.   Regulation of 15-lipoxygenase promoter activity by IL-4 and Stat6. BEAS-2B cells were cotransfected with LOP4.5 (2 µg), a -actin/-gal reporter gene (50 ng), and a Stat6 expression vector (100 ng) and treated with IL-4 (10 ng/ml) for 36 h as indicated. Whole-cell extracts were analyzed for luciferase and -gal activity. The luciferase data are normalized for 100,000 -gal units and are expressed as means of triplicates ± standard error. The data are representative of three similar experiments.

Mapping of the 12/15-Lipoxygenase 5'-Flanking Sequence for Functional Stat6 Response Elements

LOP4.5 contains approximately 4.5 kb of 5'-flanking sequence, of which 1.8 kb has been sequenced. In the region of known sequence, there are several putative Stat6 response elements (28). A series of lipoxygenase promoter/ luciferase reporter gene vectors was constructed and used to map the functional regions responsible for IL-4-dependent regulation of 12/15-lipoxygenase. These studies show that the IL-4-dependent transcriptional activity is mediated by sequences between the XhoI and EcoRI sites located at 3.0 kb and 181 bp upstream of the translational start codon (Figure 6). The 12/15-lipoxygenase promoter/ luciferase vectors were generated from the pGL3 basic vector, which expresses, at most, 50 percent of the normalized luciferase activity seen from the shortest lipoxygenase promoter reporter gene, LOP.05 (D. Conrad and M. Lu, unpublished observations). Both pGL3 control and pGL3 basic lack IL-4-dependent luciferase activity, indicating a lack of functional Stat6 response elements in these vectors.

View larger version (10K): [in this window] [in a new window]   Figure 6.   Mapping the 12/15-lipoxygenase 5'-flanking region for sequences responsive to the IL-4/ Stat6 signal transduction pathway. A series of 12/15-lipoxygenase promoter/luciferase reporter genes based on LOP4.5 were constructed. pGL3 control is a reporter gene that uses an SV40 promoter. BEAS-2B cells were cotransfected with the 12/15-lipoxygenase promoter/reporter genes (2 µg each), a -actin/-gal reporter gene (50 ng), and the Stat6 expression vector (100 ng), and cultured for 36 h in the presence (open bars) or absence (solid bars) of IL-4 (10 ng/ml) as indicated. The protein extracts were analyzed for luciferase and -gal activity. The luciferase data are normalized for 100,000 -gal units and are expressed as means of triplicates ± standard error. The data are representative of at least three separate experiments.

The putative Stat6 response element located 952 bp upstream of the translational start site was the initial focus of these studies. A vector was constructed that contained four direct repeats of this element cloned upstream of the HSV thymidine kinase promoter/luciferase reporter protein (ST6X4TLO). This vector demonstrated IL-4/Stat6- dependent transcriptional activity that was not evident in the parental vector, OTLO (Figure 7).

View larger version (13K): [in this window] [in a new window]   Figure 7.   Functional studies of the putative Stat6 response element. The Stat6 response element, located 952 bases upstream of the translational start codon, was catermerized and ligated into the SalI site of vector OTLO as described in MATERIALS AND METHODS. The vector, ST6X4TLO, contains four copies of the Stat6 sites as direct repeats upstream of the HSV thymidine kinase promoter (109 bases of proximal promoter). LOP3.0m expresses the mutated Stat6 response element in the context of the native 12/15-lipoxygenase 5'-flanking sequence. BEAS-2B cells were cotransfected with a Stat6 expression vector (100 ng), a -actin/-gal reporter gene (50 ng), and 2 µg of either OTLO, ST6X4TLO, LOP3.0, or LOP3.0m. After transfection, the cells were cultured for 24 h in the presence (open bars) or absence (solid bars) of IL-4 (10 ng/ml). The luciferase data are normalized for 100,000 -gal units and are expressed as means of triplicates with a standard error. The data are representative of at least three separate experiments.

To establish that this was the only functional Stat6- dependent site in this region, site-directed mutagenesis was used to obliterate this Stat6 response element. This mutant vector, LOP3.0m, had no IL-4/Stat6-dependent activity when cotransfected into BEAS-2B cells (Figure 7) although it contained the basal activity noted in the parent vector LOP3.0. Thus, these studies identified no other functional Stat6-dependent response elements in this 3-kb region.

IL-4-Dependent Binding of Stat6 to the 12/15-Lipoxygenase Promoter

EMSAs were used to confirm the Stat6 interaction with the 12/15-lipoxygenase promoter. The nuclear extracts used in Figure 8 were prepared from BEAS-2B cells transfected with a Stat6 expression vector. The probe (30 bases) was generated by end-labeling annealed oligonucleotides that contain the functional Stat6 response element with 10 bases of flanking sequence on either side. The EMSA showed a single IL-4-dependent probe:protein complex and demonstrated the prompt nuclear translocation of Stat6 (Figure 8a). A second, similar probe in which the native Stat6 response element contained the same bp changes as those used in LOP3.0m showed no binding activity in EMSAs using the same nuclear extracts (D. Conrad and M. Lu, unpublished observations).

View larger version (23K): [in this window] [in a new window]   Figure 8.   Specificity of Stat6 binding to the functional response element in the 12/15-lipoxygenase 5'-flanking sequence. Nuclear extracts were made from BEAS-2B cells that were transfected with the Stat6 expression vector and incubated with IL-4 (250 ng/ml) for 60 min. The end-labeled probe consisted of sense and antisense oligonucleotides that spanned the functional Stat6 site and 10 bases of flanking sequence. Nuclear extracts and probe (100,000 cpm) were used in binding reactions that were resolved on 5% 0.5× TBE gels as described in MATERIALS AND METHODS. (a) Escalating amounts of nuclear extract prepared from untreated and IL-4-treated cells were used in EMSAs as indicated. (b) The specificity of binding to the Stat6 response element was demonstrated by adding unlabeled probe and mutant probe to the binding reaction at 100-fold molar excess concentration of the labeled probe. (c) Supershift assays with a rabbit polyclonal antihuman Stat6 antibody were performed by adding the antibody (1 to 3 µg) to the binding reaction. In a-c, the IL-4-dependent binding complexes are indicated by the solid arrows. In c, the open arrow indicates the position of the supershifted complex. In b and c, the migration of the free probe is indicated at the bottom of each gel.

Competition assays were used to assess whether the binding activity noted in the IL-4-treated BEAS-2B nuclear extracts was specific to the Stat6 response element sequence. In the presence of 100-fold molar excess of unlabeled probe in the binding reaction, the IL-4-dependent binding complex was not detected but was seen when the same concentration of unlabeled mutant probe was used in the reactions (Figure 8b).

Finally, antibody supershift assays were used to determine the presence of Stat6 in these binding complexes. With 3 µg of a rabbit polyclonal antibody to human Stat6, we detected a quantitative supershift of the IL-4-dependent binding complex (Figure 8c). No shift was detected using the antibody alone. These studies confirm the presence of Stat6 in these probe:nuclear protein binding complexes.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Human 12/15-lipoxygenase belongs to a group of homologous enzymes including the murine monocyte 12-lipoxygenase and a rabbit reticulocyte 15-lipoxygenase that appear related because of their similarity of primary amino acid sequence, immunocrossreactivity, and tissue pattern of expression (9, 22, 32). These homologous enzymes also display similar catalytic properties in terms of their substrate specificity and regiospecificity of arachidonic acid oxygenation. These enzymes catalyze the oxygenation of several polyenoic fatty acids, including those esterified into complex lipids (8, 19, 36). In addition, the murine, human, and rabbit enzymes generate both 15S- and 12S-HETE from arachidonic acid, which distinguishes their lipoxygenation reaction from those seen with the platelet 12-lipoxygenases, human 12(R)-lipoxygenase, and human 15-lipoxygenase-2 (29).

The rabbit reticulocyte 15-lipoxygenase was the first of these enzymes to be purified in a search for cellular mechanisms that resulted in mitochondrial membrane oxidation and subsequent degradation (8, 10). The presence of a 15-lipoxygenase in human reticulocytes suggests that a similar mechanism is relevant in human erythrocyte maturation (9).

The generation of lipid peroxides in macrophage plasma membranes has implicated 12/15-lipoxygenase in the oxidative modification of low-density lipoprotein into its atherogenic form (1, 2, 5, 25, 37, 38). Although most of these studies support a proatherogenic effect of 12/15-lipoxygenase in humans and models of atherosclerosis in rabbits and mice, one study demonstrated a protective effect of human 12/15-lipoxygenase (7). In this study, human 15-lipoxygenase was overexpressed from a lysozyme promoter in the monocytes of rabbits and showed not only good expression of the enzyme monocytes but also decreased levels of the fatty streak formation when compared with control animals.

Specific biologic roles for the airway 12/15-lipoxygenases have yet to be established in vivo. 12/15-Lipoxygenase enzymatic activity was demonstrated earlier in humans and dogs (39). Later, the human 12/15-lipoxygenase was purified from eosinophils and cloned from human reticulocytes and airway epithelial cells (27, 40, 41). These studies allowed other investigations that linked the 12/15-lipoxygenase of reticulocytes, eosinophils, macrophages, and airway epithelial cells and eventually led to the identification of a single human 12/15-lipoxygenase gene (28, 34).

The biology established in macrophage and reticulocyte systems suggest that the airway 12/15-lipoxygenase could have similar effects in airway cells. This led to the hypothesis that the generation of lipid peroxides in cell membranes of airway epithelial cells expressing the 12/ 15-lipoxygenase could augment airway host defenses (4, 11). Alternatively, the airway 12/15-lipoxygenase could affect airway epithelial cell function through the generation of bioactive lipids such as the mono-HETEs, di-HETEs, hydroperoxyeicosatetraenoic acids (HPETEs), and lipoxins (3, 12, 42). Although some of these metabolites exhibit anti-inflammatory responses that antagonize the 5-lipoxygenase pathway, other studies indicate that the 12/15-lipoxygenase metabolites could modify inflammation through other specific receptors such as PPAR (43). Recent studies in monocytes demonstrate activation of PPAR in a 12/15-lipoxygenase-dependent manner (18). Activation of PPAR in monocytes is associated with downregulation of the inducible nitric oxide synthase and gelatinase B. If similar mechanisms are true in airway epithelial cells, this mechanism may partly explain the anti-inflammatory properties of 12/15-lipoxygenase and its metabolites.

Given the roles for 12/15-lipoxygenase in atherosclerosis and its potential role in modifying airway inflammation, the study of the mechanisms responsible for regulating 12/ 15-lipoxygenase expression become very important. Earlier studies have shown that IL-4 and IL-13 upregulate this enzyme in cultured monocytes (17, 19). More recently, IL-4 was shown to upregulate 12/15-lipoxygenase expression in airway epithelial cells (11, 20, 21). Investigations of the homologous gene in murine systems have shown that the IL-4- dependent induction requires Stat6; however, a direct interaction with the murine 12/15-lipoxygenase promoter was not demonstrated (22). Further, one study demonstrated that murine 12/15-lipoxygenase protein expression does not require Stat6 activation and that other signal transduction pathways may also be involved (46).

The data presented here show that IL-4 induces endogenous human 12/15-lipoxygenase in transformed and nontransformed airway epithelial cells. Our assays establish the specificity of the immunoblot analysis and demonstrate the IL-4-dependent generation of 15- and 12-HETE at the characteristic 8:1 ratio. The fact that BEAS-2B cells do not express 12/15-lipoxygenase in response to IL-4 but regain the ability to synthesize the protein with Stat6 confirms the importance of this transcription factor in mediating IL4-dependent responses in general and the expression of 12/15-lipoxygenase in particular.

Human 12/15-lipoxygenase gene expression results from transcriptional mechanisms mediated through a simple Stat6 response element in the 5'-flanking sequence located 952 bases upstream from the translational start codon. The Stat6 response element characterized in this paper conforms to the characteristics of a consensus Stat6 response element (47). It contains the TTC...GAA consensus sequence present in most Stat response elements. More importantly, it contains the 4-bp spacing between these elements that is a necessary determinant for specific Stat6 dimer binding and transcriptional transactivation. The 12/15-lipoxygenase Stat6 response element appears to be a simple response element. Primary sequence analysis of the region flanking the Stat6 response element did not identify other response elements that could potentially compete or interact with Stat6 binding at this site. Our EMSA data did not suggest the presence of other proteins binding in the immediate region of the Stat6 site. The EMSA data are consistent with the IL-4-dependent phosphorylation and translocation of Stat6 to the nucleus and is consistent with the known events of this signal transduction pathway (48).

In these studies, we demonstrate the immunologic and functional presence of Stat6 in transformed and nontransformed human airway epithelial cells. By expressing Stat6 in BEAS-2B cells, the IL-4/Stat6 signal transduction pathway normally present in airway epithelial cells is reconstituted and restores Stat6-dependent 12/15-lipoxygenase gene regulation. The specific reason for the loss of Stat6 expression in this transformed airway epithelial cell line is not known. Inasmuch as there is no immunologic evidence of Stat6 protein in BEAS-2B cells, it is unlikely that its lack of expression can be explained by the presence of Stat6c, a Stat6 isoform associated with dominant negative activity (24).

Histologic studies demonstrate human 12/15-lipoxygenase expression in ciliated cells of the airway epithelium under normal conditions; however, its expression is not limited to cells with this mature phenotype. In fact, we show that nontransformed cells grown in submersion cultures express 12/15-lipoxygenase even though this culture system results in cells that lack a mature phenotype. In addition, several studies have demonstrated expression in two naturally transformed cell lines (20, 21).

There is a clear role for Stat6 in humoral responses and specifically those involved in mediating allergic inflammation. Recent studies have also confirmed an important role for Stat6 in mediating allergic inflammation in murine models of asthma, including bronchial hyperreactivity and mucus hypersecretion (49). In these models of asthma, activation of Stat6 via IL-13 is particularly important (50, 51). Although many of these effects likely result from Stat6 expression in mononuclear cells, they do not exclude important roles for the Stat6 expressed in other airway cells, including the airway epithelium (51). The expression of Stat6 in airway epithelial cells confirms the important role of these cells in pulmonary host defenses and supports a downstream role for the 12/15-lipoxygenase in the IL-4/ Stat6 signal transduction pathway.