Osmotic Response Element Enhancer Activity

Hypertonicity induces a group of genes that are responsible for the intracellular accumulation of protective organic osmolytes such as sorbitol and betaine. Two representative genes are the aldose reductase enzyme (AR, EC 1.1.1.21), which is responsible for the conversion of glucose to sorbitol, and the betaine transporter (BGT1), which mediates Na+-coupled betaine uptake in response to osmotic stress. We recently reported that the induction of BGT1 mRNA in the renal epithelial Madin-Darby canine kidney cell line is inhibited by SB203580, a specific p38 kinase inhibitor. In these studies we report that the hypertonic induction of aldose reductase mRNA in HepG2 cells as well as the osmotic response element (ORE)-driven reporter gene expression in transfected HepG2 cells are both inhibited by SB203580, suggesting that p38 kinase mediates the activation and/or binding of the transcription factor(s) to the ORE. Electrophoretic gel mobility shift assays with cell extracts prepared from SB203580-treated, hypertonically stressed HepG2 cells further show that the binding of trans-acting factors to the ORE is prevented and is thus also dependent on the activity of p38 kinase. Similarly, treatment of hypertonically stressed cells with PD098059, a mitogen-activated extracellular regulated kinase kinase (MEK1) inhibitor, results in inhibition of the hypertonic induction of aldose reductase mRNA, ORE-driven reporter gene expression, and the binding of trans-acting factors to the ORE. ORE-driven reporter gene expression was not affected by p38 kinase inhibition or MEK1 inhibition in cells incubated in iso-osmotic media. These data indicate that p38 kinase and MEK1 are involved in the regulation of the hyperosmotic stress response.

Hypertonicity induces a group of genes that are responsible for the intracellular accumulation of protective organic osmolytes such as sorbitol and betaine. Two representative genes are the aldose reductase enzyme (AR, EC 1.1.1.21), which is responsible for the conversion of glucose to sorbitol, and the betaine transporter (BGT1), which mediates Na ؉ -coupled betaine uptake in response to osmotic stress. We recently reported that the induction of BGT1 mRNA in the renal epithelial Madin-Darby canine kidney cell line is inhibited by SB203580, a specific p38 kinase inhibitor. In these studies we report that the hypertonic induction of aldose reductase mRNA in HepG2 cells as well as the osmotic response element (ORE)-driven reporter gene expression in transfected HepG2 cells are both inhibited by SB203580, suggesting that p38 kinase mediates the activation and/or binding of the transcription factor(s) to the ORE. Electrophoretic gel mobility shift assays with cell extracts prepared from SB203580-treated, hypertonically stressed HepG2 cells further show that the binding of trans-acting factors to the ORE is prevented and is thus also dependent on the activity of p38 kinase. Similarly, treatment of hypertonically stressed cells with PD098059, a mitogen-activated extracellular regulated kinase kinase (MEK1) inhibitor, results in inhibition of the hypertonic induction of aldose reductase mRNA, ORE-driven reporter gene expression, and the binding of trans-acting factors to the ORE. ORE-driven reporter gene expression was not affected by p38 kinase inhibition or MEK1 inhibition in cells incubated in isoosmotic media. These data indicate that p38 kinase and MEK1 are involved in the regulation of the hyperosmotic stress response.
Many organisms, including bacteria, yeast, plants, and animals, adapt to sustained hyperosmotic stress by the preferential accumulation of compatible organic osmolytes (1). The induction of osmoprotective genes has been shown to be part of the protective response of the kidney medulla tubular cells to exposure to hypertonicity during urinary concentration (2). The accumulation of these osmolytes is facilitated by the induction of specific proteins as follows: betaine/␥-amino-n-butyric acid transporter (BGT1) 1 for betaine (3); Na ϩ -dependent myo-inositol transporter (SMIT) for inositol (4); taurine transporter for the amino acid taurine (5), as well as the aldose reductase enzyme (AR) that catalyzes the NADPH-mediated reduction of aldehydes such as D-glucose to sorbitol (6,7).
AR is a high K m (30 -80 mM) enzyme for glucose and other sugars (6), and it was shown to play a major role in the pathogenesis of a variety of diabetic complications via an increased flux of glucose through the sorbitol pathway during hyperglycemia, i.e. the pathway is driven by increased intracellular glucose availability leading to sorbitol accumulation (8). However, it is also possible to accumulate large amounts of intracellular sorbitol at normal blood glucose levels by increasing the transcription of AR upon exposure of cells to hypertonicity. The result is an enzyme-driven intracellular accumulation of sorbitol at normal blood glucose levels, which is one of several redundant mechanisms involved in osmoregulation in the renal medulla (2).
Hypertonicity increases synthesis of AR mRNA 15-fold in 24 h without a detectable change in the rate of degradation of the AR protein (9). AR transcription is regulated by a promoter that contains diverse regulatory elements including an osmotic response element (ORE) present at 1235 bp upstream of the transcription start site, which mediates its induction during hypertonic stress (10 -12). Studies in the MDCK renal epithelial cell line showed that a specific p38 kinase inhibitor, SB203580, blocked the induction of BGT1 mRNA in response to hyperosmotic media (13) suggesting that the p38 kinase pathway is essential for the hypertonicity response. Additional studies in monocytes showed that SB203580 also blocked the hypertonic induction of mRNA of both SMIT and BGT1 (14). However, the role of p38 kinase in the hypertonic induction of these genes remains unknown.

EXPERIMENTAL PROCEDURES
Preparation of Constructs-A 132-bp fragment (GenBank TM accession number L14440, nucleotides 2032-2163) (10) containing the osmotic response element (ORE) and its surrounding sequence was excised from the AR promoter construct pARP4.2 and inserted into the pCAT SV40 promoter vector (Promega) using BamHI/PstI cloning sites as described previously (11). Introduction of SacI and NheI sites by polymerase chain reaction using primers 5Ј-TGGTTTGTCCGAGCT-CATCAATGTATCTTATC and 5Ј-CAGGAAACAGCTAGCACCATGAT-TACGCCA, respectively, allowed the ligation of this fragment into the pGL3-promoter vector that carries the SV40 promoter upstream of the luciferase gene (Promega). A vector construct containing the minimal 12-bp sequence necessary for ORE enhancer activity (GenBank TM accession number L14440, nucleotides 2110 -2121) was prepared by annealing a synthetic oligonucleotide fragment containing the ORE, 5Ј-gagctcTGGAAAATCACCgctagc and its reverse complement, flanked by NheI and SacI sites, and inserted into the pGL3 promoter vector. Similarly, an insert spanning 1.5 kb of the AR gene (GenBank TM accession number L14440, nucleotides 1801-3300) which includes the ORE enhancer and other identified elements of the AR promoter (10 -12) was amplified by polymerase chain reaction using primers 5Ј-TTTAGGCTCGAGTTCAAATTCTATTACTTGG and 5Ј-CCATGGAA-GCTTCGCTCCCCAGACCCC. The resulting fragment was inserted into the pGL3 basic vector using XhoI and HindIII restriction sites. All constructs were sequenced to verify the sequence and orientation (sequencing kit from U. S. Biochemical Corp.).
Tissue Culture and Transfections-HepG2 cells were grown at 37°C and 5% CO 2 in DMEM low glucose medium supplemented with 10% fetal calf serum, 1% penicillin/streptomycin, and 2 mM glutamine (Life Technologies, Inc.). Cells (0.25 ϫ 10 6 , passages 30 -40) were seeded into 24-well plates (16-mm well diameter) and grown for 24 h. The cells were transfected with 2 g of DNA/well of pCAT (chloramphenicol acetyltransferase) or pGL3 (luciferase) vectors, driven by either the 12-or 132-bp ORE constructs. The 1.5-kb promoter-driven pGL3 basic vector construct was transfected at 4 g of DNA/well. A control firefly luciferase vector pRSVL (Promega, 0.2 g of DNA/well) was cotransfected with ORE-pCAT plasmids to correct for transfection efficiency. Similarly, a control plasmid, pRLTK (Promega, 0.2 g of DNA/well) which expresses the Renilla luciferase, was used as a control for transfection efficiency of the ORE-pGL3 plasmids, which express the firefly luciferase. The two luciferases exhibit luminescence under different conditions in the dual luciferase assay, thus allowing independent quantitation. Transfections were done using the calcium phosphate precipitation method. The cells were shocked after 6 h with 15% glycerol and allowed to recover for 24 h in regular media. Hypertonic stress was induced by the addition of 5 M NaCl to regular DMEM to make a final concentration of 220 mM NaCl (515 mosmol/kg). Regular DMEM was added to cells exposed to iso-smotic conditions. The kinase inhibitors SB203580 or PD098059 (Calbiochem) were added in 10 l of Me 2 SO (Sigma) to achieve the desired concentrations. Me 2 SO was also added to the controls to ensure that all wells had the same concentration of Me 2 SO. The cells were harvested after 16 h incubation.
Reporter Gene Assays-In the CAT reporter gene assays, the cells were harvested by scraping in 200 l of phosphate buffer (0.1 M potassium phosphate, pH 7.9) and lysed by 3 freeze-thaw cycles at Ϫ80 and 37°C. The lysate was centrifuged, and 10 l of the supernatant was analyzed for luciferase activity (15) using a Monolight 2010 luminometer (Analytical Luminescence Laboratory, Sparks, MD). Extracts (2-50 l) containing equal luciferase activity were then incubated with [ 14 C]chloramphenicol (Amersham Pharmacia Biotech) and 1 mM acetyl-CoA (Sigma) in phosphate buffer. The acetylated chloramphenicol products were separated by thin layer chromatography and visualized by autoradiography. The bands visualized by autoradiography were then excised from the chromatography plate and quantitated by liquid scintillation counting.
In the dual luciferase transfections, passive lysis buffer (200 l) supplied by the manufacturer (Promega) was added to each well, and the plate was shaken on an orbital shaker for 30 min at room temperature. The resulting cell lysate was collected in amber tubes and centrifuged, and 2-5 l of the supernatant was analyzed for Renilla and firefly luciferase activities in a Monolight 2010 luminometer using the Dual Luciferase Assay kit (Promega). Kinase Assays-MAPK activity was determined as described previously (16,17) with slight modifications. The cells were exposed to hyperosmotic conditions in the presence or absence of inhibitors as described above, scraped into the experimental medium, and harvested by centrifugation. The cell pellet was lysed in Triton Lysis Buffer (TLB) consisting of 20 mM Tris, pH 7.4, 137 mM NaCl, 2 mM EDTA, 1% Triton X-100, 25 mM ␤-glycerophosphate, 1 mM sodium orthovanadate, 2 mM sodium pyrophosphate, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, and 1 g/ml leupeptin. The supernatant was collected by centrifugation at 15,000 ϫ g for 10 min. Anti-ERK antibody (Upstate Biotechnology Inc. NY; the antibody cross-reacts with ERK1 and ERK2) was bound to protein A-and G-agarose. Cell lysate (100 g of protein) was then added to the agarose beads, and ERK was precipitated by the antibody-agarose complex. The beads were washed 3 times in TLB, followed by 3 additional washes in kinase buffer (25 mM Hepes, pH 7.4, 25 mM ␤-glycerophosphate, 25 mM MgCl 2 , 2 mM dithiothreitol, 0.1 mM sodium vanadate). Myelin basic protein (Sigma) in kinase buffer containing 25 M [␣-32 P]dATP was added to the beads and incubated for 20 min at 30°C. The kinase reaction was stopped by centrifugation at 12,000 ϫ g for 2 min. The supernatant was resolved on a 15% SDS-PAGE, and the gel was dried and autoradiographed. ERK1 and -2 activity was determined by the extent of incorporation of 32 P in the myelin basic protein substrate. For JNK1, p38, and p38␤ assays, the above procedure is used except for the following: anti-p38, anti-p38␤, or anti-JNK1 antibodies (Santa Cruz, CA) are used for immunoprecipitation, and Pansorbin (Calbiochem, CA) is used to immobilize the antibodies; activated transcription factor 2 is used as substrate for the p38 kinases, whereas the NH 2 -terminal peptide-(1-79) derived from c-Jun (Santa Cruz Biotechnology) is used as substrate for JNK1. The reaction supernatant is resolved on 10% SDS-PAGE.
Gel Shift Assays-The oligonucleotides 5Ј-GTGAAGCACCAAATG-GAAAATCACCGGCATGGAGT and 5Ј-GTGACTCCATGCCGGTGA-TTTTCCATTTGGTGCTT containing the minimal ORE (in bold) were annealed, and the 5Ј-overhangs were labeled using Klenow polymerase (Roche Molecular Biochemicals), [␣-32 P]dCTP, and [␣-32 P]dATP (ICN Radiochemicals). The labeled probe was purified using a Select G-25 spin column (5 Prime 3 3 Prime, Inc., Boulder, CO). Specific competitor fragments were generated in a similar fashion using non-radioactive ␣-dCTP and ␣-dATP. For the nonspecific competitor, a single base (underlined and bold) was changed in the ORE element to yield oligonucleotides 5Ј-AAGCACCAAATAGAAAATCACCGGCATGGAGT and 5Ј-ACTCCATGCCGGTGATTTTCTATTTGGTGCTT which were annealed and used directly. This single nucleotide mutation was previously shown in transfection studies to abolish the hypertonicity response of the ORE (11).
Whole cell extracts were prepared from HepG2 cells seeded into 10-cm dishes and grown in regular DMEM until they reached 90% confluency. The cells were then exposed to isotonic or hypertonic media containing the appropriate concentrations of kinase inhibitors (10 M SB203580 or 10 M PD098059) for 16 h. The plates were washed with phosphate-buffered saline, and the cells were scraped and collected in extraction buffer (1% Nonidet P-40, 5 g/ml soybean trypsin inhibitor, 1 mM phenylmethylsulfonyl fluoride, 100 M leupeptin, and 100 M benzamidine in phosphate-buffered saline) and incubated on ice for 30 min. The lysate was centrifuged, and the supernatant was stored in aliquots at Ϫ70°C. Aliquots containing 5 g of total protein were incubated for 30 min in gel shift buffer (10 mM Tris, 50 mM NaCl, 10% glycerol) containing radiolabeled probe (50,000 cpm), 100 ng of poly(dI/ dC) and cold competitor (0, 10, or 50ϫ). The reaction products were then separated on a 4% acrylamide gel at 4°C and visualized by autoradiography.
Northern Blot Analysis-HepG2 cells at 90% confluency were exposed to iso-osmotic or hyperosmotic medium (with and without 10 M SB203580 or PD098059) for 16 h as described above. The cells were scraped and pelleted by centrifugation at 5000 ϫ g for 5 min at 4°C. Total RNA was isolated from the cell pellet using RNAzol (Tel-Test, Friendswood, TX). Equal amounts of RNA per lane were loaded onto 1% agarose, 2.2 M formaldehyde gel. The gel was electrophoresed and transferred to GeneScreen membrane (NEN Life Science Products). Human AR cDNA and introns 6 -8 from the human AR gene (6) and human full-length ␤-actin cDNA (CLONTECH) were labeled with [␣-32 P]dCTP (Random Primed DNA Labeling Kit, Roche Molecular Bio-chemicals) for use as probes. Probes were hybridized to the blots overnight at 42°C in a solution containing 40% formamide, 5ϫ SSC (0.75 M NaCl, 75 mM trisodium citrate, pH 7), 5ϫ Denhardt's solution (0.5% (w/v) polyvinylpyrrolidone, 0.5% (w/v) Ficoll 400), 0.5% SDS, 250 g/ml salmon sperm DNA, 10 mM Tris, pH 7.5, and 10% dextran sulfate. The blots were washed under high stringency at 65°C as follows: twice in 3ϫ SSC, 0.5% SDS for 30 min, and twice in 0.3ϫ SSC, 0.5% SDS for 30 min. The blots were autoradiographed, and relative band intensities were quantitated using Image Tool (University of Texas Health Science Center, San Antonio) software. Relative band intensities were normalized to ␤-actin.

Hypertonic Induction of AR mRNA in HepG2 Cells Is p38
Kinase-and MEK1-dependent-We examined the effects of SB203580, a specific p38 kinase inhibitor, on the hypertonic induction of AR mRNA in HepG2 cells. Fig. 1 shows that 16 h of exposure to hypertonicity induces a 15-fold increase in AR mRNA levels in comparison with cells incubated in isotonic media. The increase in mRNA is attenuated in the presence of a 5 M concentration of the p38 kinase inhibitor. These osmotically stressed cells express two specific AR mRNA species in equal abundance (1.5 and 2.5 kb) as compared with cells incubated in isotonic media, where only the smaller mRNA is detected. Northern blots were hybridized with genomic probes derived from introns 6 -8 of the human AR gene (6). Only the larger 2.5-kb mRNA species hybridized to the intron 7 probe but not to introns 6 or 8 probes, suggesting that this band represents incompletely processed mRNA (6) containing intron 7 sequences (data not shown). The hypertonic induction of AR mRNA in HepG2 cells was also attenuated by PD098059, a MEK1 inhibitor, at a 10 M concentration (Fig. 1). These data suggest that p38 kinase and MEK1 regulate the hypertonically induced native promoter expression and the level of native AR mRNA in these cells.
Regulation of Hypertonic Induction by p38 Kinase and MEK1 Is ORE-mediated-To determine the specific sites in the AR gene promoter that are under the regulatory control of p38 kinase and MEK1, we examined the expression of ORE-driven CAT or luciferase reporter constructs in hypertonically stressed HepG2 cells in the presence of the p38 kinase inhibitor (SB203580) and MEK1 inhibitor (PD098059). Transfection experiments with a CAT reporter construct driven by a 132-bp sequence that contains the basic ORE and surrounding se-quence 2 paralleled the mRNA findings (Fig. 2), with OREdriven CAT reporter gene expression rising 5-fold in hypertonically stressed cells. This response was attenuated in the presence of the p38 kinase inhibitor in a dose-dependent manner with complete inhibition at a 25 M concentration (Fig. 2). Transfection experiments using a luciferase reporter construct driven by the same 132-bp enhancer sequence showed similar results (Fig. 3A), albeit complete inhibition of the hypertonic induction of reporter gene product expression was achieved at a lower concentration of p38 kinase inhibitor (10 M). Since the results obtained with the luciferase assay confirmed those obtained with the CAT assay, the luciferase assay was used for the remainder of the studies. The effects of MEK1 inhibition on ORE-driven luciferase expression in osmotically stressed cells were also similar (Fig. 3B). A 5 M concentration of PD098059 was sufficient to inhibit completely the hypertonic induction response.
To define better the cis-elements regulated by p38 kinase and MEK1, we examined the expression of luciferase reporter constructs, driven by one of the following enhancer/promoter constructs: the 12-bp basic ORE, TGGAAAATCACC; 132-bp sequence, which contains the ORE complex as well as the full 1.5-kb human AR promoter (10 -12). Hypertonicity increases ORE-driven luciferase reporter expression 5.9-, 4.7-, and 1.6fold with the 1.5-kb, 132-, and 12-bp ORE sequences, respectively. This induction is attenuated in the presence of either MEK1 or p38 kinase inhibitor, in a dose-dependent manner, with complete inhibition at 10 M SB203580 or 5 M PD098059 (Table I). However, the expression of the reporter gene was not affected by either inhibitor under isotonic conditions. Although the magnitude of hypertonic induction of the reporter gene product differed among the various constructs (1.5 kb Ͼ 132 bp Ͼ 12 bp), the effects of the inhibitors were similar ( Table I). The data suggest that the hypertonicity-induced regulatory effects of p38 kinase and MEK1 on AR expression are ORE-specific.
p38 Kinase and MEK1 Regulate the Binding of trans-Acting Elements to the ORE-We examined the effects of p38 kinase and MEK1 inhibitors on the interaction between the ORE and transcription factors. The basic 12-bp ORE fragment was radiolabeled and incubated with whole cell extracts derived from isotonic or hypertonically stressed HepG2 cells. Two prominent DNA-protein complexes were observed, designated as complex 1 and 2 (Fig. 4). Complex 2 was seen most prominently in extracts from osmotically stressed cells, whereas complex 1 was observed in all extracts and is intensified 15-fold under hyper-  2. Effect of p38 kinase inhibitor, SB203580, on ORE-driven CAT reporter construct in response to hypertonicity. Autoradiograph of the CAT reporter assay from HepG2 cells transfected with 132-bp ORE-driven CAT reporter construct. Top band represents acetylated chloramphenicol products, and the bottom band represents nonacetylated chloramphenicol. Lanes 1 and 2, cells exposed to isotonic media; lanes 3 and 4, cells exposed to hypertonic media containing 5 M SB203580; lanes 5 and 6, cells exposed to hypertonic media containing 10 M SB203580; lanes 7 and 8, cells exposed to hypertonic media containing 25 M SB203580; lanes 9 and 10, cells exposed to hypertonic media containing 50 M SB203580; lanes 11 and 12, cells exposed to hypertonic media containing 100 M SB203580; lanes 13 and 14, cells exposed to hypertonic media. tonic conditions (Fig. 4). These complexes can be competitively prevented from forming by incubation with excess cold ORE. The formation of both complexes was not affected by incubation with excess cold mutant ORE (TAGAAAATCACC), indicating binding specificity. Under hypertonic conditions, treatment with either p38 kinase or MEK1 inhibitor abolishes the formation of complex 2 and diminishes the formation of complex 1. The data suggest that the binding of transcription factors to the ORE upon exposure of the cells to hypertonic stress is dependent on the activities of both MEK1 and p38 kinase.
Regulatory Effects of MEK1 on ORE Are ERK-independent-PD098059 is a specific inhibitor of MEK1 (an upstream activator of ERK1 and ERK2) and is commonly used for down-regulation of ERK pathways. To determine whether the effects of PD098059 are indeed ERK-mediated, we examined ERK activity in PD098059-treated, osmotically stressed HepG2 cells, by measuring the in vitro activities of immunoprecipitated ERK1 and ERK2. As shown in Fig. 5, the sum of ERK1 and ERK2 activities in lysates of PD098059-treated cells is higher than that observed in untreated cells, irrespective of medium tonicity, indicating that ERK1 and ERK2 are not inhibited by PD098059. The data suggest that the PD098059-mediated effects are MEK1 but not ERK1/2-dependent. Since the effects of MEK1 were not mediated through the ERK pathway, we examined the activities of p38 kinase, p38␤ kinase, and JNK1. As shown in Fig. 5, treatment with PD098059 does not affect the activities of these kinases, suggesting that the MEK1 effects in hypertonically stressed HepG2 cells are in addition p38 kinase-, p38␤ kinase-, and JNK1-independent. DISCUSSION Exposure of cells to hypertonicity induces a sequence of events that results in the expression of a variety of genes, some protective, whereas others may have harmful effects. The induction of these genes results from a complex of signal cascades that ultimately affect the promoter of the genes involved. The promoters of AR, BGT1, and other genes that have been shown to be induced by the hypertonic response have a 12-bp minimal enhancer element that has been shown to respond to hypertonicity (11,12,18,19). The minimal osmotic response element (ORE or TonE) appears to be widely distributed in the genome and can be found in the promoters of other genes that respond to stimuli other than hypertonicity. For example, the nuclear factor of activated T lymphocytes (NFAT) binds to a DNA region that contains a consensus 12-bp ORE motif, TG-GAAAATTTGT, that responds to hypertonicity in transfection studies. 3 In activated T lymphocytes, the NFAT transcription factor binds the consensus ORE in conjunction with a Jun-Fos AP-1 heterodimer to form a large NFAT-AP-1-DNA complex (20). It thus appears that discrete response and cellular specificities may be attained through multiple levels of interactions.
In yeast, the induction of GPD1 (glycerol-3-phosphate dehydrogenase 1), which is responsible for the synthesis of glycerol in response to osmotic stress, is dependent on the HOG1 mitogen-activated protein kinase (MAPK), a mammalian p38 homologue, suggesting that the p38 kinase pathway may be important in the regulation of osmoprotective genes by hypertonicity in mammalian cells (21). Although the activation of ERK, JNK, and p38 kinases by osmotic stress has been demonstrated in mammalian cells (22)(23)(24), ERK activity was not found to be essential for transcriptional regulation of betaine and inositol transporters (25). In addition, inhibition of MKK3 (one of the upstream activators of p38 kinase) or SEK1 (an upstream activator of JNK) did not affect the induction of osmoprotective genes in mammalian cells (26). However, using a specific p38 kinase inhibitor, SB203580, it was shown that the p38 kinase pathway is essential for the hypertonic induction of BGT1 mRNA in the canine renal epithelial cells, MDCK (13), and the hypertonic induction of both SMIT and BGT1 mRNAs in monocytes (14).
In these experiments (Fig. 1), we show that the hypertonicity-induced AR mRNA increase in HepG2 cells is attenuated by the p38 kinase and MEK1 inhibitors, SB203580 and PD098059, respectively. In addition, we show in transfection experiments that the specific locus of action of these inhibitors is the activation of the ORE in response to hypertonicity (Figs. 2 and 3). The data in Table I, describing the effect of the inhibitors on the minimal 12-bp ORE, the more responsive 132-bp ORE, and the full 1.5-kb AR promoter, confirm that the ORE motif is the main effector, since the inhibitory effects are comparable among the three constructs despite quantitative gradations in the response to hypertonicity.
Our data also demonstrate the involvement of p38 kinase in the binding of trans-acting elements to the ORE. Two specific major complexes were observed in the electrophoretic mobility shift assays (Fig. 4). The faster mobility complex 1, which is also seen with extracts derived from HepG2 cells maintained in isotonic media, increases 15-20-fold with hypertonicity, whereas the slower mobility complex 2 was detected only upon exposure of the cells to hypertonic media. Inhibition of p38 kinase during exposure of cells to hyperosmotic media abolishes the induction of complex 2 and markedly diminishes the formation of complex 1. A recent report by Miyakawa et al. (27) described the cloning of a human cDNA coding for a TonEbinding protein (TonEBP) that binds to TonE (ORE) sequences via a Rel-like DNA binding domain (28). The latter domain,

FIG. 3. Effect of p38 kinase inhibitor (A) or MEK1 inhibitor (B) on ORE-driven luciferase reporter construct expression.
Firefly luciferase activity of extracts from HepG2 cells transfected with 132-bp ORE-driven luciferase reporter construct were normalized to control Renilla luciferase activity. Boxes indicate extracts from cells exposed to isotonic media, and triangles indicate extracts from cells exposed to hypertonic media. Values represent the mean of six independent transfections in three separate experiments. Error bars indicate Ϯ S.D.
which is present at the NH 2 -terminal end of TonEBP, has a 45% amino acid identity with the corresponding region of the NH 2 -terminal DNA binding domains of the NFAT transcription factors (27). The x-ray crystallographic structure of the NFAT-AP-1-DNA complex (20) demonstrates that this region forms part of a large groove, which along with the grooves generated by the DNA-facing surfaces of Fos and Jun bind the DNA and form the large multimer. All of the above data clearly establish that p38 kinase is involved in the activation and/or synthesis of the transcription factor binding to the ORE.
PD098059 is a selective inhibitor of MEK1, without significant inhibitory activity against ERK itself. It has no inhibitory activity against 18 different protein Ser/Thr kinases, including two other MEK1 homologues that participate in stress and interleukin-1-stimulated kinase cascades, suggesting high specificity for MEK1 (29). Inhibition of MEK by PD098059 was shown to prevent the activation of ERK and subsequent phosphorylation of ERK substrates both in vitro and in intact cells (30). Our data with PD098059 indicate that MEK1 is involved in the regulation of hypertonic induction of ORE-mediated gene expression. Inhibition of MEK1 in hypertonically stressed cells results in attenuation of AR mRNA abundance, inhibition of ORE-driven reporter construct expression, and the binding of trans-acting elements to labeled ORE. However, the MEK1 effect does not appear to be ERK-mediated, since the ERK1/ ERK2 activity is in fact increased in response to inhibitor treatment. Similarly, the induction of GLUT-1 gene by hypertonicity in L6 muscle cells has been shown to be MEK1-dependent, yet ERK-independent, suggesting that in the context of osmotic stress, the MEK1 effects are ERK-independent (31). In addition, the activities of p38 kinase, p38␤ kinase, and JNK1 in PD098059-treated and hypertonically stressed cells are not affected, indicating that MEK1 effects are not mediated through these kinases. The basis for MEK1 effect on OREmediated gene expression remains to be determined. The pyridinyl imidazole group of compounds to which SB203580 belongs demonstrate a highly specific and potent inhibitory activity against p38 kinase. In earlier studies these compounds showed no inhibitory activity against ERK1 and ERK2, JNK, MAPK-activated protein kinase-2 (MAPKAPK-2), MAPKK, proteinkinaseC,calmodulin-dependentproteinkinase,orcAMPdependent protein kinase (32,33). In fact, the binding specificity of these compounds was utilized to identify and affinity purify p38 kinase (33). It should be noted that in Fig. 5 SB203580-treated cells do not show any change in the activities of immunoprecipitated p38 and p38␤ kinases, since the drug is removed after repeated washes of the precipitates. As was TABLE I The effect of p38 kinase inhibitor (SB203580, abbreviated as SB) or MEK1 inhibitor (PD098059, abbreviated as PD) on the expression of various ORE-driven luciferase reporter constructs is shown. Values represent the ratio of ORE-driven firefly luciferase expression relative to control Renilla luciferase (see "Experimental Procedures") and are a mean of six independent transfections (ϮS.E.) in three separate experiments. I, isotonic medium; H, hypertonic medium.
FIG. 4. Effect of p38 kinase or MEK1 inhibitor on the electrophoretic mobility of ORE and HepG2 cell-derived trans-acting elements. Whole cell extracts were incubated with 32 P-labeled ORE, and the reactions were electrophoresed on 4% polyacrylamide gel. Lane 1, extract from cells in isotonic media; lanes 2 and 3, extract from cells in isotonic media plus 10ϫ or 50ϫ of unlabeled ORE as competitor, respectively; lanes 4 and 5, extract from cells in isotonic media plus 10 or 50ϫ of unlabeled mutated ORE, respectively; lane 6, extract from cells in hypertonic media; lanes 7 and 8, extract from cells in hypertonic media plus 10 or 50ϫ of unlabeled ORE as competitor, respectively; lanes 9 and 10, extract from cells in hypertonic media plus 10 or 50ϫ of unlabeled mutated ORE as competitor, respectively; lane 11, extract from cells in hypertonic media plus SB203580; lane 12, extract from cells in hypertonic media plus PD098059; lane 13, labeled probe alone.
FIG. 5. In vitro activity of native ERK1/2, p38 kinase, p38␤ kinase, and JNK1 from SB203580-or PD098059-treated and hypertonically stressed HepG2 cells. Whole cell extracts from HepG2 cells exposed to various stimuli were analyzed for native ERK1/2, p38, p38␤, and JNK1 activities after immunoprecipitation. The labeled substrates were resolved and autoradiographed. Panel 1, ERK1 and ERK2; panel 2, p38 kinase; panel 3, p38␤ kinase; and panel 4, JNK1. Lanes 1 and 2, extracts from cells in isotonic media; lanes 3 and 4, extracts from cells in isotonic media containing SB203580; lanes 5 and 6, extracts from cells in isotonic media containing PD098059; lanes 7 and 8, extracts from cells in hypertonic media; lanes 9 and 10, extracts from cells in hypertonic media containing SB203580; lanes 11 and 12, extracts from cells in hypertonic media containing PD098059. previously shown in MDCK cells (13), JNK1 activity is upregulated in HepG2 cells when p38 kinase is inhibited by SB203580. The significance of this finding remains to be determined.
Our data indicate that the hypertonic induction of AR mRNA in HepG2 cells is regulated by p38 kinase and MEK1 and is mediated by the ORE. Inhibition of either kinase independently results in loss of the hypertonic effects suggesting that both signal pathways are necessary for the hyperosmotic response. It is not known at this time whether these kinases affect only the binding of the transcription factors to the ORE, their nuclear translocation and activation, or both.