Identification of a Novel NF-κB-binding Site with Regulation of the Murine α2(I) Collagen Promoter

Hepatic fibrosis is due to the increased synthesis and deposition of type I collagen. Acetaldehyde activates type I collagen promoters. Nuclear factor κB (NF-κB) was previously shown to inhibit expression of murine α1(I) and human α2(I) collagen promoters. The present study identifies binding of NF-κB, present in nuclear extracts of stellate cells, to a region between -553 and -537 of the murine α2(I) collagen promoter. The NF-κB (p65) expression vector inhibited promoter activity. Mutation of the promoter at the NF-κB-binding site increased basal promoter activity and abrogated the activating and inhibitory effects of transforming growth factor β and tumor necrosis factor α, respectively, on promoter activity. Acetaldehyde increased IκB-α kinase activity and phosphorylated IκB-α, NF-κB nuclear protein, and its binding to the promoter. However, the activating effect of acetaldehyde was not affected by the mutation of the promoter. In conclusion, although acetaldehyde increases the binding of NF-κB to the murine α2(I) collagen promoter, this binding does not mediate the activating effect of acetaldehyde on promoter activity. The effects of acetaldehyde in increasing the translocation of NF-κB to the nucleus with increased DNA binding activity may be important in mediating the effects of acetaldehyde on other genes.

␣ 1 (I) and ␣ 2 (I) collagen gene transcription and messages as well as type I collagen production by cultured stellate cells (10 -13). In previous studies, we showed that acetaldehyde enhances the activity of the murine ␣ 2 (I) collagen promoter in stellate cells and that this effect is mediated by increased binding of nuclear proteins including nuclear factor I to region Ϫ315 to Ϫ295 of the promoter (12). Although Sp1 is important in maintaining basal activity of the ␣ 2 (I) collagen promoter, it plays no role in mediating the effects of acetaldehyde (14).
Acetaldehyde was shown to enhance binding of NF-B of nuclear extract from HepG2 cells to the NF-B consensus sequence in one study (15), whereas in another study, acetaldehyde inhibited lipopolysaccharide-stimulated DNA binding of NF-B by nuclear extracts from Kuppfer cells (16). Upon analysis of the nucleotide sequence of the murine ␣ 2 (I) collagen promoter, we found the presence of a NF-B-binding motif (5Ј-ACTGGGGAAATTAGGGG-3Ј) at Ϫ553 to Ϫ537 from the start of transcription. The purpose of this study was to determine the regulatory role of NF-B on the ␣ 2 (I) collagen promoter and the effect of acetaldehyde on the regulation.

EXPERIMENTAL PROCEDURES
Animals and Materials-Adult male Sprague-Dawley rats were obtained from Charles River Laboratories (Wilmington, MA). All of the animals received humane care in compliance with the guidelines of the Animal Care and Use Committee of the Johns Hopkins University. TNF-␣ and TGF-␤1 were obtained from R & D Systems (Minneapolis, MN). Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum, agarose, and the restriction enzymes XbaI and HindIII were purchased from Invitrogen. Plastic tissue culture flasks were purchased from BD Biosciences. [␣-32 P]dATP, [␣-32 P]dCTP, and [␥-32 P]ATP were purchased from ICN (Costa Mesa, CA). Acetaldehyde was from Fisher.
Plasmids-The plasmid pAZ1009, which contains a 2.0-kb region of the mouse ␣ 2 (I) collagen promoter, was provided by Dr. Benoit de Crombrugghe from the M. D. Anderson Cancer Center (Houston, TX). The luciferase construct of the ␣ 2 (I) collagen promoter (pGL3-1009) was made by inserting the 2.0-kb promoter into the HindIII site of the pGL3 enhancer vector as described previously (13). The NF-B expression vectors RSV-NF-B (p50) and RSV-Rel (p65) and the empty vector k1 were obtained from Prof. Guidalberto Manfioletti (University of Trieste, Trieste, Italy).
Site-directed Mutagenesis-Oligonucleotides containing 2-bp substitutions were designed for site-directed mutagenesis. The method employed for the mutagenesis of the pGL3-1009 promoter was based on the strategy of overlap extension using PCR as described by Ho et al. (17). The DNA sequencing facility of our Department of Biological Chemistry confirmed successful mutations.
Isolation and Culture of Hepatic Stellate Cells-Rats weighing ϳ400 g were used in the procedure of hepatic stellate cell isolation, as described previously (13). The procedure follows the method of Friedman and Roll (18) except for the use of a Nycodenz gradient. In brief, the stellate cells were isolated in situ by the perfusion of the portal vein under sterile conditions sequentially with 0.2% Pronase E and 0.015% collagenase in DMEM. The liver cell suspension obtained was centrifuged at 1400 ϫ g in a two-step discontinuous Nycodenz gradient. The isolated cells were suspended in DMEM and seeded in 25-cm 2 tissue culture flasks maintained in DMEM containing 10% fetal bovine serum, fungizone (2.5 g/ml), penicillin (100 units/ml), and streptomycin (100 g/ml) at 37°C with a humidified atmosphere of 5% CO 2 and 95% air. The medium was changed every 48 h, while the cells transformed into activated cells after 10 -14 days in culture.
Nuclear Protein Extraction-Nuclear extracts from cultured stellate cells were prepared as described previously (19). The nuclear protein extracts were aliquoted and stored under nitrogen at Ϫ80°C. The protein content of the nuclear extracts was determined by the method of Lowry et al. (20).
Electrophoretic Mobility Shift Assays (EMSA)-The sequence of the wild-type oligonucleotide, corresponding to region Ϫ553 to Ϫ537 of the ␣ 2 (I) collagen promoter, used initially for EMSA was 5Ј-ACTGGG-GAAATTAGGGG-3Ј. Mutated oligonucleotides, which contained 2-bp nucleotide substitutions, are shown in the legend to Fig. 2. Complimentary strands of each oligonucleotide were annealed, and the doublestranded oligonucleotides were labeled with [␣-32 P]dATP and [␣-32 P]dCTP using Klenow enzyme according to the method of Feinberg and Vogelstein (21). DNA-protein binding reactions were performed following the previously described EMSA procedure (22). Nuclear extracts from stellate cells were incubated with the labeled oligonucleotide probes (2.5-25 fmol) at room temperature for 30 min in 25 l of reaction buffer containing 25 mM HEPES, pH 7.8, 50 mM KCl, 0.1 mM ZnCl 2 , 1 mM dithiothreitol, 2 g of poly(dI-dC), and 10% glycerol. The competition assays were performed by incubating molar excess of unlabeled oligonucleotides for 30 min with nuclear proteins prior to the addition of labeled oligonucleotide probes. For "supershift" EMSA experiments, rabbit polyclonal antibodies to NF-B (p65) and to NF-B (p50) obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) were used. These antibodies were added separately to the reaction at the completion of DNA-protein binding and incubated for an additional 30 min at room temperature. The resultant samples were resolved on a 5% nondenaturing polyacrylamide gel and visualized on x-ray film.
Transient Transfections and Luciferase Assay-Activated stellate cells, after no more than three passages, cultured in DMEM containing 20% fetal bovine serum, penicillin (100 units/ml), streptomycin (100 g/ml), and fungizone (2.5 g/ml) were grown until they were 60% confluent. Each cell culture flask was transfected with 5 g of pGL3-1009 using calcium phosphate precipitation (23). Where expression vectors were used, the controls were transfected with the empty k1 vector, and the total DNA quantities added to each flask were balanced by the addition of sheared salmon sperm DNA. Transfection efficiency was determined by co-transfection of 0.4 g of the Renilla luciferase vector pRL-CMV (Promega). Four h after transfecting, the cells were washed twice with DMEM and then shocked with 10% (IH 3 ) 2 SO for 3 min. The cells were then returned to DMEM containing 10% fetal bovine serum. For the experiments with acetaldehyde, the medium was changed after 1 h to serum-free DMEM containing the following six supplemental growth factors: epidermal growth factor (10 g/liter), transferrin (0.5 mg/liter), selenous acid (5 g/liter), linoleic acid (0.5 mg/liter), bovine serum albumin (0.5 mg/liter), and fetuin (0.5 mg/liter). The acetaldehyde was added to a final concentration of 200 M, and the flasks were tightly capped. The cells were spiked with acetaldehyde at 12-h intervals for a total treatment time of 40 h. The cells were harvested 40 -44 h after transfection and exposed to one freeze-thaw cycle in reporter lysis buffer (Promega). Firefly luciferase activity was determined and normalized to Renilla luciferase activity using the dual luciferase assay system of Promega.
Ultraviolet Cross-linking of Nuclear Proteins to Oligonucleotides and Immunoblot Analysis-The binding of nuclear proteins to the oligonucleotide probe was performed as for EMSA. Reactions using 8 g of nuclear protein and 50 fmol of the radioactively labeled oligonucleotide were used. Following the binding reaction, UV cross-linking was performed as described previously (24). The membranes were then incu- IB Kinase Assay-IB-␣ kinase activity was determined by the method of Mercurio et al. (25). The cytosolic protein was incubated with antibody to IKK-␣ overnight. Twenty l of protein A-Sepharose (Rockland, Gilbertsville, PA) was then added. Following a 2-h incubation, with gentle rocking, the antigen-antibody-protein A-Sepharose complex was precipitated, washed, and then assayed for kinase activity. The reaction mixture for the kinase activity consisted of 20 mM HEPES buffer, pH 7.4, 20 mM MgCl 2 , 2 mM MnCl 2, 10 mM ␤-glycerophosphate, 10 mM NaF, 10 mM p-nitrophenyl phosphate, 300 M Na 3 VO 4 , 1 mM benzamidine, 1 mM dithiothreitol, 2 M phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml pepstatin, 10 g/ml aprotinin, 20 mCi of [␥-32 P]ATP, 10 M ATP, and 2 g of GST-IB-␣ (Santa Cruz Biotechnology). The total volume of reaction mixture was 25 l. After incubation at 30°C for 30 min, the reaction was terminated by boiling with 5 l of 6ϫ SDS sample buffer for 5 min. The protein was resolved on a 10% polyacrylamide gel under reducing conditions. The gel was dried, and the radioactive bands were visualized by phosphorimaging.
Statistical Analysis-The data were analyzed with the Student's t test when appropriate or by two-way analysis of variance when comparing the means of more than two groups.

RESULTS
Nuclear Factor B Binds to the Murine ␣ 2 (I) Collagen Promoter, and This Binding Activity Is Increased by TNF-␣-Increased nuclear protein binding to the oligonucleotide specifying the wild-type Ϫ553 to Ϫ537 region of the ␣ 2 (I) collagen promoter is shown in Fig. 1A. The nuclear protein binding is competed away by a 100-fold molar excess of the cold oligonucleotide. Antibodies to NF-B (p50) and to NF-B (p65) resulted in the appearance of supershifted bands (Fig. 1B). Only the antibody of NF-B (p50) resulted in the complete disappearance of nuclear protein binding.
A mutational analysis of the NF-B-binding site of the ␣ 2 (I) collagen promoter was performed by EMSA of protein-DNA complex formed with labeled oligonucleotides that were sequentially mutated by 2 base pairs between Ϫ550 and Ϫ537

FIG. 4. Immunoblot blot of the effects of TNF-␣ (T) on nuclear NF-B (p50) and NF-B (p65) proteins in NE of rat stellate cells.
The cells in culture were exposed to TNF-␣ (0.6 nM) for 30 min. C, not exposed. (Fig. 2). This analysis revealed that the 2-bp mutation of the oligonucleotide M2 or of the oligonucleotide M3 eliminates the upper protein-DNA complex. The 2-bp mutation of oligonucleotide M6 results in a decrease in the upper protein-DNA complex. Thereafter the ␣ 2 (I) collagen promoter (pGL3-1009) was mutated by the 2-bp mutation of the M2 oligonucleotide and named pGL3-1009mut.
Exposure of the stellate cells to TNF-␣ (0.6 nM) for 30 min resulted in an increase in the upper protein-DNA complexes formed by nuclear extracts with the wild-type NF-B oligonucleotide (Fig. 3, lane 3). Antibody to NF-B (p65) results in a decrease in the upper complex formed in the presence of 7 g of NE (lane 5) and in the appearance of a supershifted complexes when the NE concentration used in the EMSA was increased to 12 g (lanes 8 and 9). TNF-␣ increased NF-B (p65) but not NF-B (p50) protein in nuclear extracts of the stellate cells (Fig. 4).
Nuclear Factor B Inhibits the Activity Murine ␣ 2 (I) Collagen Promoter-The effects of NF-B (p65) and NF-B (p50) expression vectors were determined on the activity of the wild-type (pGL3-1009) and mutated (pGL3-1009mut) ␣ 2 (I) collagen promoters in co-transfection experiments. The NF-B (p65) expression vector resulted in the inhibition of the activity of pGL3-1009 (p Ͻ 0.05) (Table I), whereas the NF-B (p50) expression vector had no effect. The combination of the NF-B (p65) and NF-B (p50) expression vectors caused inhibition of pGL3-1009 activity (p Ͻ 0.05) similar to that observed with NF-B (p65) expression vector alone. The basal activity of the transfected pGL3-1009mut was greater than the activity of the wild-type pGL3-1009 (p Ͻ 0.01). NF-B (p65) alone or NF-B (p50) alone had no significant effect on pGL3-1009mut, but the combination of NF-B (p65) and NF-B (p50) resulted in a paradoxical increase in the activity of pGL3-1009mut (p Ͻ 0.05).
Acetaldehyde Increases Nuclear Factor B Binding to the Murine ␣ 2 (I) Collagen Promoter-Exposure of stellate cells to acetaldehyde (200 M) for 4 and 24 h increased the binding of NF-B to the wild-type NF-B oligonucleotide (Fig. 5). Acetaldehyde increased nuclear NF-B (p65) protein after 4 h of exposure but had no effect on nuclear NF-B (p50) protein ( Fig   6). The relative densitometries of cross-linked NF-B (p65) shown in Fig 6B were 100 Ϯ 9 for control and 139 Ϯ 7 after 4 h of acetaldehyde exposure (p Ͻ 0.05).
Acetaldehyde Activates NF-B by Enhancing IB Kinase and IB Phosphorylation-Acetaldehyde resulted in a small decrease in IB-␣ kinase activity at 0.5 h, followed by an increase at 4 h. (Fig. 7A). Acetaldehyde exposure for 4 h resulted in a decrease in phosphorylated IB-␣ at 0.5 h followed by an increase at 4 h (Fig. 7B). Acetaldehyde increased IB-␣ at 0.5 and 4 h and increased IB-␤ at 4 h (Fig. 7B). Acetaldehyde increased IKK-␣ protein but did not affect IKK-␤ protein (Fig.  7C). The effects of acetaldehyde in enhancing IB kinase and IB phosphorylation at 4 h were associated with increased cytosolic NF-B (p65) protein, but there was no change in cytosolic NF-B (p50) protein (Fig. 7D).
The Activating Effect of Acetaldehyde on Enhancing the Activity of the Murine ␣ 2 (I) Collagen Promoter Is Not Mediated by NF-B Binding-Acetaldehyde (200 M) increased the activity of both pGL3-1009 and pGL3-1009mut (Table III). DISCUSSION This study demonstrates that NF-B binds to a novel site in the murine ␣ 2 (I) collagen promoter and that binding to this site results in inhibition of promoter activity. Prior studies have shown that NF-B binds and inhibits the murine ␣ 1 (I) (8) and human ␣ 2 (I) collagen promoters (9). TNF-␣ is well known to activate latent NF-B by degradation of IB inhibitory cytoplasmic retention proteins, leading to it nuclear translocation and gene activation (26). In this study, TNF-␣ increased NF-B (p65) protein and binding to the newly described NF-B-binding site in the murine ␣ 2 (I) collagen promoter and inhibited the activity of the transfected promoter in stellate cells. Previously, TNF-␣ had been shown to increase NF-B binding activity to the murine ␣ 1 (I) (8) and to the human ␣ 2 (I) (9) collagen promoters, resulting in depressed collagen gene expression (8,9). TGF-␤1 is a strong stimulator of the production of extracellular matrix proteins such as collagen and fibronectin (27). The activation by TGF-␤ of the ␣ 2 (I) collagen promoter in this study required NF-B binding to the newly described binding site, because activation was abrogated by mutation of this binding site. This finding was surprising because TGF-␤ was shown to regulate the human ␣ 2 (I) collagen promoter principally through the cellular Smad signal transduction pathway (28). However, it has become apparent that the receptor-activated Smads regulate transcription in the context of other transcription factors such as Sp1 (29) and that in addition other signaling pathways including the RAS pathway may play a role as mediators of the effect of TGF-␤ (30). Furthermore, in one recent study, TGF-␤ was shown to increase NF-B binding in nuclear extracts from cultured fibroblast-like synoviocytes to a consensus NF-B oligonucleotide (31).
Acetaldehyde increased nuclear NF-B (p65) protein in stellate cells and its binding to the murine ␣ 2 (I) collagen. NF-B is usually present as a heterodimer of two proteins and p65 and p50 subunits. In the cytoplasm, NF-B is bound to IB-␣ and IB-␤, which when phosphorylated by kinases are degraded, causing a release of NF-B, which then translocates to the nucleus where it binds to target genes (32). The effects of acetaldehyde in increasing IB-␣ kinase activity and phosphorylated IB-␣ indicate that acetaldehyde enhances the translocation of NF-B to the nucleus by increasing the degradation of IB-␣. Previous studies of the effects of acetaldehyde on NF-B binding to a consensus sequence NF-B oligonucleotide gave contradictory results. In one study using lipopolysaccharidestimulated Kupffer cells, acetaldehyde increased cytosolic IB-␣ protein and nuclear protein NF-B DNA binding (16), whereas in another study with HepG2 cells, acetaldehyde decreased IB-␣ protein and increased nuclear protein NF-B DNA binding (15). Although our study demonstrates that acetaldehyde increases the binding of NF-B to a specific site in the ␣ 2 (I) collagen promoter, this binding site does not mediate the activating effect of acetaldehyde on promoter activity, because the activation was not affected by mutation of the NF-B-binding site. The effects of acetaldehyde in increasing the translocation of NF-B to the nucleus with changes in its DNA binding activity may be important in mediating the effects of acetaldehyde on other genes.
The effects of acetaldehyde on type I collagen regulation may be partially mediated by TGF-␤. Acetaldehyde was shown increase secretion of TGF-␤ by stellate cells and to increase expression of the type II TGF-␤ receptor (33). Neutralizing antibody to TGF-␤ decreased acetaldehyde-induced augmentation of the mouse ␣ 2 (I) collagen mRNA in stellate cells (13). Also, neutralizing antibodies to TGF-␤ or to the type II TGF-␤ receptor as well as antisense type II TGF-␤ receptor resulted in partial blockage of the activating effect of acetaldehyde on the rat ␣ 1 (I) collagen promoter (27). The regulation of the type I collagens involves various transcription factors binding to proximal TGF-␤-responsive elements. In previous studies, we showed that acetaldehyde enhances the activity of the ␣ 2 (I) collagen promoter in stellate cells and that this effect is mediated by increased binding of nuclear proteins, including nuclear factor I to a region, which is located in the TGF-␤-responsive element (12). The binding of Sp1, by contrast, was not affected by acetaldehyde (14). In studies with the mouse ␣ 1 (I) collagen promoter, it was found that the activating effect of acetaldehyde was mediated by increased binding of CCAAT/ enhancer-binding protein ␤ to a region that overlaps with the TGF-␤-responsive element (34,35), but in this case, the activating effect of acetaldehyde was not modified by neutralizing antibody to TGF-␤ (34).
In conclusion, this study shows that NF-B binds to a novel site in the murine ␣ 2 (I) collagen promoter and that binding to this site results in inhibition of promoter activity. Acetaldehyde increases the translocation to the nucleus and the binding of NF-B to the promoter by increasing the phosphorylation-dependent degradation of IB-␣. This increased NF-B binding, however, plays no role in the activating effect of acetaldehyde on the activity of this collagen promoter. The effects of acetaldehyde in increasing the binding activity of NF-B to DNA may be important in the regulation of other genes.