Angiotensin II-induced Transcriptional Activation of the Cyclin D1 Gene Is Mediated by Egr-1 in CHO-AT1A Cells*

Cyclin D1 protein expression is regulated by mitogenic stimuli and is a critical component in the regulation of G1 to S phase progression of the cell cycle. Angiotensin II (Ang II) binds to specific G protein-coupled receptors and is mitogenic in Chinese hamster ovary cells stably expressing the rat vascular Ang II type 1A receptor (CHO-AT1A). We recently reported that in these cells, Ang II induced cyclin D1 promoter activation and protein expression in a phosphatidylinositol 3-kinase (PI3K)-, SHP-2-, and mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK)-dependent manner (Guillemot, L., Levy, A., Zhao, Z. J., Béréziat, G., and Rothhut, B. (2000)J. Biol. Chem. 275, 26349–26358). In this report, transfection studies using a series of deleted cyclin D1 promoters revealed that two regions between base pairs (bp) −136 and −96 and between bp −29 and +139 of the human cyclin D1 promoter contained regulatory elements required for Ang II-mediated induction. Mutational analysis in the −136 to −96 bp region provided evidence that a Sp1/early growth response protein (Egr) motif was responsible for cyclin D1 promoter activation by Ang II. Gel shift and supershift studies showed that Ang II-induced Egr-1 binding involved de novo protein synthesis and correlated well with Egr-1 promoter activation. Both U0126 (an inhibitor of the MAPK/ERK kinase MEK) and wortmannin (an inhibitor of PI3K) abrogated Egr-1 endogenous expression and Egr-1 promoter activity induced by Ang II. Moreover, using a co-transfection approach, we found that Ang II induction of Egr-1 promoter activity was blocked by dominant-negative p21 ras , Raf-1, and tyrosine phosphatase SHP-2 mutants. Identical effects were obtained when inhibitors and dominant negative mutants were tested on the −29 to +139 bp region of the cyclin D1 promoter. Taken together, these findings demonstrate that Ang II-induced cyclin D1 up-regulation is mediated by the activation and specific interaction of Egr-1 with the −136 to −96 bp region of the cyclin D1 promoter and by activation of the −29 to +139 bp region, both in a p21 ras /Raf-1/MEK/ERK-dependent manner, and also involves PI3K and SHP-2.

The control of mammalian cell proliferation by extracellular signals in G 1 to S phase progression of the cell cycle is largely mediated by serine/threonine cyclin-dependent kinases CDK4 1 and CDK6, which interact with specific D-type cyclins. The CDK-D-type cyclin complexes induce phosphorylation of the retinoblastoma protein (pRb), thereby releasing the transcription factor E2F, which is required for the transcription of S phase-specific genes (1)(2)(3)(4). Activation by mitogenic stimuli of D-type cyclins during the G 1 phase appears to be an essential and rate-limiting step in G 1 to S phase progression of the cell cycle (5)(6)(7). The cyclin D1 gene expression seems to be essentially regulated at the transcription level. The promoter region of the cyclin D1 gene contains multiple potential cis-regulatory elements including binding sites for AP1, E2F, Oct, Egr-1, Sp1, ATF/CREB, NF-B, and signal transducers and activators of transcription (7)(8)(9)(10)(11)(12). Transcriptional activation of the cyclin D1 gene occurs in a cell type-and mitogen-specific manner. For instance, it has been shown that the AP1 binding site was implicated in Ang II-induced cyclin D1 activation in the human adrenal cell line H295R (13), whereas signal transducers and activators of transcription binding sites were involved in cytokine-dependent growth of hematopoietic cells (12) and Sp1 and a cAMP-responsive element in serum-stimulated vascular endothelial cells (9). Numerous studies have shown that cyclin D1 gene activation by growth factors and by Ang II is dependent upon the Ras/extracellular signal-regulated kinase pathway (13)(14)(15)(16)(17)(18).
Ang II, the major effector molecule of the renin-angiotensin system, has long been implicated in the pathobiology of hypertension. This octapeptide hormone exerts diverse biological effects including induction of cell hypertrophy and/or hyperplasia and stimulation of hormone synthesis and ion transport in the heart, kidney, and adrenal (19). Ang II functions as a growth factor in vascular smooth muscle cells, cardiac fibroblasts, and Chinese hamster ovary cells stably expressing the rat vascular Ang II type 1A receptor (CHO-AT 1A ) (20 -22). Many of the known biological actions of Ang II are mediated by stimulation of the AT 1 receptor subtype, a member of the G protein-coupled seven-transmembrane-spanning receptor family (23). Signal transduction through the AT 1 receptor involves phospholipase C, phospholipase A2, phospholipase D, adenylate cyclase, and the release of intracellular calcium (24 -27). Moreover, Ang II mediates many intracellular signaling pathways similar to those induced by classical growth factors and cytokines. Ang II induces rapid tyrosine phosphorylation and activation of phospholipase C-␥1 (28), the Janus kinase/signal transducers and activators of transcription pathway (29), protein-tyrosine phosphatase SHP-2 (18,30), MAPK/ERK (31,32), and phosphatidylinositol 3-kinase (PI3K) (33). Ang II also activates c-Jun N-terminal kinase (34) and p38/MAPK (35). The array of genes activated by Ang II includes those encoding growth factors and its receptors (36 -38), genes encoding extracellular matrix proteins (39 -41), and several immediately early growth response genes such as c-fos, c-jun, c-myc, and the early growth response gene-1 (egr-1) (42)(43)(44)(45)(46). egr-1 (also known as NGF1-A, Krox24, Tis8, and zif268) (47) is a member of the immediate early gene family that encodes an 80 -82-kDa nuclear protein. Egr-1 is a DNA-binding protein containing three zinc finger motifs that regulates gene transcription by interacting with a consensus G ϩ C-rich sequence 5Ј-GCG(T/G)GGGCG-3Ј (48). The expression of the Egr-1 protein is rapidly and transiently induced by growth factors and other extracellular signals and is a critical upstream mediator of cell proliferation (49), differentiation (50), and apoptosis (51). In turn, Egr-1 regulates expression of many genes such as those for growth factors, cytokines, and adhesion molecules (52)(53)(54).
In a recent study (18), we established that Ang II-induced cyclin D1 protein expression and promoter activation required p21 ras , Raf-1, MEK, PI3K, and also the catalytic activity of SHP-2 and its Src homology 2 (SH2) domains through the regulation of MAPK/ERK activity. However, the regulatory mechanisms underlying Ang II induction of the cyclin D1 promoter are not known in CHO-AT 1A cells. Here we show that the Egr-1 transcription factor is largely implicated in Ang II-dependent activation of the cyclin D1 promoter and that induction of DNA binding activity and transcriptional activity via the Egr-1 site is mediated by the Ras/MEK/ERK-dependent pathway. We also demonstrate that Ang II induction of Egr-1 and cyclin D1 is modulated by PI3K and SHP-2 in CHO-AT 1A cells. In addition, we found that the second region stimulated by Ang II and located between nucleotides Ϫ29 and ϩ139 of the 5Ј-untranslated region of the promoter is also regulated by Ras/MEK/ERK, PI3K, and SHP-2.
Transient Transfection and Luciferase Activity Assay-Firefly luciferase reporter gene constructs (D1⌬, pGLE, or pT81 constructs) were transiently transfected in CHO-AT 1A cells (640 ng/3 ϫ 10 6 cells in 100-mm dishes) with or without different amounts of relevant expression vectors (i.e. 2.5 g of Ras N17 mutant, 2.5 g of Raf-1 C4 mutant, 5 g of SHP-2 CS mutant, 5 g of the SH2 domains of SHP-2, 5 g of SHP-1 CS mutant, 5 g of Egr-1 WT, 5 g of Egr-1 RW mutant, or the corresponding empty vector) using LipofectAMINE-PLUS reagent following the protocol provided by the supplier. To control transfection efficiency and to normalize firefly luciferase values in experiments using pT81 constructs, cells were co-transfected with 6 ng of the internal control vector pRL-TK (herpes simplex virus thymidine kinase promoter driving Renilla luciferase expression). Twenty-four hours following transfection, cells were trypsinized and aliquoted into 12-well cell culture dishes (250,000 cells/well). Six hours later, cells were serum-starved for 24 h in Ham's F-12 medium supplemented with 0.5 mg/ml bovine serum albumin before preincubation or not with inhibitors for 1 h and then stimulated for the indicated times with 10 Ϫ7 M Ang II. Cells were lysed with 200 l of passive lysis buffer (1ϫ), and a 20-l aliquot was assayed for luciferase activity in a luminometer (Lumat LB9507, Berthold) using the kit from Promega.
Preparation of Nuclear Extracts-CHO-AT 1A cells pretreated or not with inhibitors for 1 h and exposed to Ang II 10 Ϫ7 M for the indicated times were washed once with ice-cold phosphate-buffered saline containing 1 mM Na 3 VO 4 at 4°C and then scraped into 5 ml of cold phosphate-buffered saline. Cells were pelleted by centrifugation at 1300 rpm for 5 min at 4°C. The pellet was resuspended in 500 l of ice-cold hypotonic solution (buffer A) consisting of 5 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.2% Nonidet P-40, 50 mM NaF, 1 mM Na 3 VO 4 , 5 mM dithiothreitol, 0.1 mg/ml 4-(2-aminoethyl)-benzenesulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml antipain, 1 g/ml pepstatin and repelleted by centrifugation at 5000 rpm for 4 min at 4°C. The pellet was lysed for 15 min at 4°C by the addition of 500 l of ice-cold buffer A. The suspension was centrifuged at 3000 rpm for 10 min at 4°C, and then nuclei were lysed for 30 min at 4°C in 30 l of ice-cold solution (buffer C) consisting of 20 mM HEPES, pH 7.9, 25% glycerol, 0.5 M NaCl, 1.5 mM MgCl 2 , 0.5 mM EDTA, 50 mM NaF, 1 mM Na 3 VO 4 , 5 mM dithiothreitol, 0.1 mg/ml 4-(2-aminoethyl)-benzenesulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml antipain, 1 g/ml pepstatin. The nuclear extract was clarified by centrifugation at 15,000 rpm for 30 min at 4°C. The supernatant fraction was immediately frozen on dry ice and stored at Ϫ80°C prior to use. Protein concentrations were measured with the bicinchoninic acid assay from Pierce.
Electrophoretic Mobility Shift Assay (EMSA)-Binding reactions for gel shift assays were performed in 20 l of 10 mM HEPES, pH 8, 0.1 mM EDTA, 50 mM NaCl, 50 mM KCl, 5 mM MgCl 2 , 4 mM spermidine, 2 mM dithiothreitol, 0.1 mg/ml bovine serum albumin, 4% Ficoll, 4 g of poly(dI-dC)-poly(dI-dC), 32 P-labeled oligonucleotide probe (200,000 cpm) and 7-8 g of nuclear extract. The reaction was allowed to continue for 10 min at 4°C. In supershift studies, 2 g of the appropriate affinity-purified antibody was preincubated with the crude nuclear extract for 10 min at 4°C before addition of the labeled probe. In competition experiments, the nuclear extract was incubated with a 100-fold molar excess of the appropriate unlabeled competitor oligonucleotides. Electrophoresis of the different samples was carried out on 6% nondenaturating polyacrylamide gels with 0.5ϫ TBE (45 mM tris(hydroxymethyl)aminomethane, 45 mM boric acid, 1 mM EDTA, pH 8) and run at 15 V/cm for ϳ2 h. The gel was dried under vacuum and exposed to Fuji Super RX film overnight at room temperature.
Western Blotting-Nuclear extracts prepared as described previously (18) were resuspended in SDS-sample buffer and boiled for 5 min at 95°C. Reaction mixtures were size-fractionated by 10% denaturing SDS-polyacrylamide gel electrophoresis (10% gels), and transferred to nitrocellulose membranes. Blots were stained with Ponceau S to confirm equal loading of nuclear proteins and then probed with the indicated antibodies: monoclonal anti-Sp1 (1C6) or polyclonal anti-Egr-1 (C-19). Immunoblots were developed using appropriate secondary horseradish peroxidase-coupled antibodies and an enhanced chemiluminescence (ECL) kit (Pierce).

Identification of Regions in the Human Cyclin D1 Promoter Required for Ang II-induced Gene
Expression-To map ciselements that were required for Ang II-induced transcriptional activity of the cyclin D1 promoter, we used a panel of luciferase promoter/reporter gene constructs in transient transfection experiments. Quiescent CHO-AT 1A cells were transfected with the 5Ј-deleted constructs and were treated with 10 Ϫ7 M Ang II for 24 h prior to lysis for luciferase assays. As shown in Fig. 1, Ang II stimulated cyclin D1 promoter/reporter gene activity ϳ6-fold for D1⌬-973 to D1⌬-136 promoter constructs. When the sequence between nucleotides Ϫ136 and Ϫ96 was deleted, the stimulatory effect was greatly diminished and further deletion to Ϫ29 bp did not reduce this activity. These results indicate that the region between nucleotides Ϫ136 to Ϫ96 contains critical sequences required for transcriptional activation of cyclin D1. A 2.4-fold stimulation was retained for D1⌬-96 and D1⌬-29 promoter constructs, suggesting that another Ang II sensitive region is located between nucleotides Ϫ29 and ϩ139 of the 5Ј-untranslated region within the cyclin D1 promoter.
The sequence between nucleotides Ϫ136 and Ϫ96 contains two putative early growth response protein (Egr) motifs and two potential Sp1 motifs that overlap with the 3Ј putative Egr motif ( Fig. 2A). To examine more precisely the function of each putative binding site, we tested various constructs encompassing the region extending from nucleotide Ϫ137 to Ϫ99 subcloned into the heterologous herpes simplex virus thymidine kinase minimal promoter fused to the luciferase reporter gene in the pT81 plasmid (7). In addition to the wild type region, we also tested three mutations affecting the putative Egr binding sites ( Fig. 2A). Quiescent CHO-AT 1A cells transiently transfected with these constructs were incubated with 10 Ϫ7 M Ang II for 24 h prior to lysis for luciferase activity assay. As shown in Fig. 2B, the wild type (WT) construct was clearly induced by Ang II, and stimulation was maintained as long as the 3Ј putative Egr/Sp1 motifs were preserved. The response to Ang II stimulation displayed by mutant 2 was fully conserved (2-fold). However, the basal level was clearly greater than with the wild type. Therefore, we hypothesize that the 5Ј putative Egr motif may bind repressor protein(s) to flanking nucleotides. With mutant 1 and mutant 3, the basal activity was severely diminished, and Ang II was unable to stimulate the promoter. Taken together, these results suggest that only the 3Ј putative Egr/ Sp1 motif is responsible for cyclin D1 activation in the Ϫ136 to Ϫ96 bp promoter region.
Ang II Induces the Specific Interaction of Nuclear Proteins with Cis-regulatory Elements between Nucleotides Ϫ136 and Ϫ96 of the Cyclin D1 Promoter-To determine whether Ang II induces an interaction between nuclear proteins from CHO-AT 1A cells and the 3Ј putative Egr/Sp1 motif, a 32 P-labeled double-stranded oligonucleotide ( 32 P-Oligo A) bearing the Ϫ117/Ϫ99 bp cyclin D1 promoter sequence ( Fig. 2A) was incubated with nuclear extracts from CHO-AT 1A cells exposed or not to 10 Ϫ7 M Ang II for different times. Electrophoretic mobility shift assays revealed the formation of two major distinct nucleoprotein complexes (A and B) (Fig. 3A). However, only complex B was induced in response to Ang II. Time course experiments defined the transient nature with which this nucleoprotein complex formed. Complex B was totally absent in nuclear extracts from unstimulated cells but appeared within 1 h after induction, reaching a maximum at 4 h, and then declined progressively until 24 h. Interestingly, induction of complex B seemed to be correlated with the earliest activation of the cyclin D1 promoter/reporter gene in CHO-AT 1A exposed to Ang II (18).
The same experiment was performed with a double-stranded 32 P-labeled oligonucleotide ( 32 P-Oligo A-TA-) bearing the mutated sequence of the 3Ј putative Egr/Sp1 motif that corre- sponds to the sequence found in mutant 1 and mutant 3 as described in the legend to Fig. 2A. Under these conditions, complex A normally present (Fig. 3B, lanes 1 and 2) in nuclear extracts from unstimulated cells or cells exposed to Ang II for 4 h was severely reduced when the mutated oligonucleotide was tested (Fig. 3B, lanes 3 and 4). Accordingly, Ang II was unable to induce the formation of complex B between this oligonucleotide and nuclear extracts of cells exposed to Ang II for 4 h (Fig. 3B, lane 4). These data are correlated with the previous observation using mutant 1 and mutant 3 in transient transfection experiments (Fig. 2B).
Next we performed competition studies to examine the binding specificity of the factors constituting complexes A and B. A 100-fold molar excess of unlabeled Oligo A abrogated the formation of both complexes formed with 32 P-Oligo A in presence of nuclear extracts from unstimulated and Ang II-stimulated cells for 4 h (Fig. 4A, lanes 2 and 6). The same molar excess of an unrelated NF1 consensus oligonucleotide (NF1 cons ) had no effect (Fig. 4A, lanes 4 and 8). However, the same molar excess of an unlabeled Sp1 consensus oligonucleotide (Sp1 cons ) abrogated the formation of complex A (Fig. 4A, lanes 3 and 7) but not that of complex B induced by Ang II (Fig. 4A, lane 7). These results suggest that, in contrast to complex A, nuclear factor from complex B is not related to Sp1. Finally, supershift analysis was performed in an attempt to identify all these complexes (Fig. 4B). Complex A was partially supershifted when nuclear extracts from unstimulated cells or cells exposed to Ang II for 4 h were preincubated with an anti-Sp1 antibody ( Fig. 4B, lanes 2 and 4). Incomplete supershifts with Sp1 antibodies have been reported elsewhere (52). The inducible complex B was completely supershifted when nuclear extracts from Ang II-stimulated cells were preincubated with an anti-Egr-1 antibody (Fig. 4B, lane 5). As a control, we used an antibody directed against AP-2, another zinc-finger transcription factor that interacts with the G ϩ C-rich sequence, 5Ј-(C/T)(C/G)(C/ G)CC(C/A)N(G/C)(G/C)(G/C)-3Ј (57,58). Under these conditions, this antibody failed to supershift any complex (Fig. 4B,  lane 6), although it was able to supershift AP-2 protein (data not shown). Taken together, these results show that Ang II induces binding of Egr-1 transcription factor to the cyclin D1 promoter in a specific and transient manner.
Overexpression of a Mutant Form of Egr-1 Inhibits Ang IIinduced Cyclin D1 Promoter Activity-To further investigate the functional role of Egr-1, we co-transfected an expression vector encoding either a WT or a mutant form of Egr-1 (RW) together with the cyclin D1 promoter constructs pT81 WT or pT81 mutant 2 (see Fig. 2). Overexpression of Egr-1 RW almost completely inhibited Ang II-induced WT and mutant 2 promoter activities (Fig. 5, lanes 3 and 6). In addition, the induction of both cyclin D1 promoter constructs by Ang II was increased by Egr-1 WT overexpression (Fig. 5, lanes 2 and 5). Altogether, these results show that Ang II-induced cyclin D1 promoter activity is mediated by Egr-1.
Ang II-induced Egr-1 Binding Activity Requires de Novo Protein Synthesis-To determine whether the specific interaction of the Egr-1 transcription factor with the cyclin D1 pro-FIG. 2. Functional role of Egr and Egr/Sp1 sites in the cyclin D1 promoter. Analysis by mutagenesis. A, design of the Ϫ137 to Ϫ99 bp region of the cyclin D1 promoter containing two putative Egr motifs and two potential Sp1 motifs that overlap with the 3Ј Egr motif. The "wild type" and three mutant oligonucleotides (mutants 1-3) were fused to the minimal promoter HSV-TK luciferase reporter gene in the pT81 plasmid as described previously (7). B, quiescent CHO-AT 1A cells transiently transfected with these constructs (empty pT81 plasmid, WT, or mutants 1-3) were treated with 10 Ϫ7 M Ang II for 24 h prior to lysis for luciferase activity assays. Firefly luciferase activity was normalized to Renilla luciferase activity. The data are presented as means Ϯ S.E. (bars) from three independent experiments, each performed in triplicate. *, significantly different compared with basal luciferase activity (p Ͻ 0.05; Student's t test). moter is dependent on de novo protein synthesis, we performed experiments in the presence of cycloheximide (Fig. 6). Quiescent CHO-AT 1A cells were pretreated (CHX (ϩ) ) or not (CHX (-) ) with cycloheximide at 10 g/ml for 1 h. Cells were then maintained either in the absence (Ϫ) or the presence (ϩ) of the same concentration of cycloheximide and exposed or not to 10 Ϫ7 M Ang II for 4 h. The inducible Egr-1 complex appeared upon Ang II stimulation in the absence of cycloheximide (Fig. 6, lane 2) and even after pretreatment with cycloheximide for 1 h (Fig. 6,  lane 4). However, the Egr-1 complex was absent after Ang II stimulation in the continued presence of cycloheximide (Fig. 6, lane 5), although the noninducible Sp1 complex was not affected by this treatment. These results probably indicate that Ang II-mediated induction of the Egr-1 complex is dependent on de novo protein synthesis.

Ang II-induced Endogenous Egr-1 Protein Expression Is Dependent on PI3K and MEK Activities in CHO-AT 1A Cells-To
further analyze the transduction pathway following Ang II stimulation, we first determined the kinetics of Egr-1 protein expression. Quiescent CHO-AT 1A cells were treated with 10 Ϫ7 M Ang II and harvested at different times. The level of the Egr-1 protein was analyzed by immunoblotting nuclear extracts with a polyclonal anti-Egr-1 antibody (C-19). As shown in Fig. 7A, in unstimulated quiescent cells, Egr-1 protein was undetectable. Ang II treatment led to an increase in the Egr-1 level detectable at 1 h poststimulation, reaching a maximum at 4 h before declining by 24 h. Interestingly, this time course of Egr-1 protein induction nicely correlated with binding to the 32 P-Oligo A probe (Fig. 3A).
We previously demonstrated the important role that MAPK/ ERK and PI3K proteins play in Ang II-induced cyclin D1 protein expression (18). To investigate whether these proteins are also involved in Ang II-induced Egr-1 protein expression, we tested the effects of U0126, an inhibitor of MEK (the dual specific kinase that activates p44/p42 MAPK/ERK by phosphorylation), and wortmannin, an inhibitor of PI3K. Quiescent CHO-AT 1A cells were pretreated or not for 1 h with U0126 at 50 M or wortmannin at 200 nM. Then 10 Ϫ7 M Ang II was added for an additional 4 h incubation period. Both drugs severely inhibited Ang II-induced Egr-1 protein expression (Fig. 7B, lanes 3  and 4). Similar results were obtained with the second PI3K inhibitor, LY294002 at 50 M (data not shown). These results indicate that MAPK/ERK and PI3K are both required for upregulation of Egr-1 protein expression in response to Ang II in CHO-AT 1A cells, as previously observed for cyclin D1 protein expression. In contrast, the Sp1 protein was detectable in unstimulated quiescent cells (Fig. 7C, lane 1) but was not induced by Ang II (Fig. 7C, lane 2). Moreover, U0126 and wortmannin had no effect on Sp1 protein expression (Fig. 7C, lanes 3 and 4), and similar results were obtained with LY294002 (data not shown).
Altogether, these results indicate that Ang II regulates Egr-1 protein expression and demonstrate that this transcription factor can be considered as the modulator of cyclin D1 expression in CHO-AT 1A cells.
Ang II-induced Egr-1 Promoter Activity Is Dependent upon Ras/Raf-1/MEK/ERK, PI3K, and the Protein-tyrosine Phosphatase SHP-2-To determine whether Ang II was able to induce Egr-1 promoter activity, we used the Ϫ697 bp human Egr-1 promoter fragment linked to the firefly luciferase reporter gene (pGLE) (55). Quiescent CHO-AT 1A cells transiently transfected with pGLE were treated with 10 Ϫ7 M Ang II for different times before luciferase assay measurements. As shown in Fig. 8A, Ang II induced Egr-1 promoter/reporter activity in a time-dependent manner, reaching 6.5-fold at 4 h, and this time was chosen for subsequent experiments. This

FIG. 3. EMSA reveals an Ang II-inducible interaction of nuclear proteins with the 3 putative Egr motif of the cyclin D1 promoter.
A, nuclear extracts from CHO-AT 1A cells treated or not (control (C)) with 10 Ϫ7 M Ang II for the indicated times were incubated with the double-stranded oligonucleotide probe ( 32 P-Oligo A) bearing the 3Ј putative Egr site, which overlaps two Sp1 consensus motifs (see Fig. 2). The arrows indicate the position of two major nucleoprotein complexes (A and B). B, nuclear extracts from CHO-AT 1A cells treated (lanes 2 and 4) or not (C, lanes 1 and 3) with 10 Ϫ7 M Ang II for 4 h were incubated with either 32 P-Oligo A (wild type) (lanes 1 and 2) or 32 P-Oligo A-TA-(mutant) probes (lanes 3 and 4). Electrophoretic mobility shift assays were performed in the same way as in A. Nucleotide sequences of Oligo A and Oligo A-TA-probes are 5Ј-GCGCCCGCCCCCGCCCCCC-3Ј and 5Ј-GCGCCCGCCCTAGCCCCCC-3Ј, respectively. kinetics correlated exactly with the Ang II-induced increase in the Egr-1 protein expression shown in Fig. 7A.
We have previously demonstrated that p21 ras , Raf-1, PI3K, and also the catalytic activity of SHP-2 and its SH2 domains were required for cyclin D1 promoter activation by Ang II through the regulation of the MAPK/ERK activity (18). To confirm the possible relationship between Ang II-induced Egr-1 protein expression and cyclin D1 promoter activity, we first tested the effects of the dominant negative Ras N17 mutant on Ang II-induced Egr-1 promoter/reporter gene activity. Quiescent CHO-AT 1A cells transiently co-transfected with the Egr-1 promoter/luciferase construct and Ras N17 were treated with 10 Ϫ7 M Ang II for 4 h before luciferase activity measurements. As shown in Fig. 8B, overexpression of Ras N17 reduced Ang II-induced Egr-1 gene promoter activity by ϳ50% (lane 2). Likewise, pretreatment of quiescent CHO-AT 1A cells with 200 nM wortmannin partially inhibited the activity of the Egr-1 promoter by ϳ50% (Fig. 8B, lane 3). When the effect of Ras N17 overexpression in the presence of 200 nM wortmannin was tested, the inhibition was complete (Fig. 8B, lane 4). Similar results were obtained with 50 M LY294002 (data not shown).
To further investigate the possible role of SHP-2, the dominant negative SHP-2 CS mutant and its SH2 domains were transiently co-transfected. Under these conditions, both forms strongly prevented the stimulation of the Egr-1 promoter/luciferase construct by Ang II (Fig. 8C, lanes 2 and 3). The catalytically inactive form of the closely related tyrosine phosphatase SHP-1 (SHP-1 CS) had no effect (Fig. 8C, lane 4), and cotransfection with the wild type SHP-2 did not affect the activity of the construct (data not shown).
Taken together, these results reveal that the Ras/Raf-1/ MEK/ERK pathway, PI3K, and also the catalytic activity of SHP-2 and its SH2 domains are required to connect Ang II to cyclin D1 promoter activation by inducing Egr-1 transcription factor in CHO-AT 1A cells.
Ang II-induced Cyclin D1 Proximal Promoter Region Activity Is Dependent upon Ras/Raf-1/MEK/ERK, PI3K, and SHP-2-We have shown in Fig. 1 that another Ang II-sensitive region seemed to be located between nucleotides Ϫ29 and ϩ139 of the cyclin D1 promoter. To gain insight into the transcriptional control of this region, we tested the expression of the dominant negative Ras N17 mutant on Ang II-induced D1⌬-29 promoter/reporter gene activity. Quiescent CHO-AT 1A cells transiently co-transfected with D1⌬-29 and Ras N17 constructs were treated with 10 Ϫ7 M Ang II for 24 h followed by luciferase activity assays. As shown in Fig. 9A, overexpression of Ras N17 inhibited the activity of the D1⌬-29 promoter/reporter gene activity construct by ϳ50% (lane 2). In the case of cells pretreated with 50 M LY294002, the activity of this promoter construct was reduced by ϳ50% (Fig. 9A, lane 3). When the effect of Ras N17 overexpression was tested in the presence of 50 M LY294002, the inhibition was complete (Fig. 9A, lane 4). Similar results were obtained with 200 nM wortmannin (data not shown). Complete inhibition was also observed when CHO-AT 1A cells were transiently co-transfected with D1⌬-29 and the dominant negative Raf-1 C4 mutant (Fig. 9A, lane 5) or when CHO-AT 1A cells were pretreated with 50 M U0126 (Fig. 9A,  lane 6).
The dominant negative SHP-2 CS mutant and the SH2 domains of SHP-2 were also transiently co-transfected with D1⌬-29 promoter/reporter gene construct. Under these conditions, Ang II induction was greatly inhibited (Fig. 9B, lanes 2  and 3). No effect was observed when SHP-1 CS (Fig. 9B, lane 4) or SHP-2 WT (data not shown) were used instead.
Taken together, these data indicate that Ang II-induced cyclin D1 proximal promoter region activity is dependent on the Ras/Raf-1/MEK/ERK pathway and PI3K activity but also on the catalytic activity of SHP-2 and its SH2 domains. DISCUSSION Cyclin D1 protein expression is regulated by mitogenic stimuli, and its assembly with cyclin-dependent kinases CDK4 or CDK6 is a rate-limiting step in the G 1 /S phase progression of the cell cycle. The aim of our study was to examine the transcriptional regulation of the cyclin D1 gene by Ang II in CHO-AT 1A cells employing serial cyclin D1 promoter deletion constructs fused to the luciferase reporter gene. It was previously FIG. 6. Effect of cycloheximide on the Ang II-induced Egr-1 binding activity. To inhibit de novo protein synthesis, CHO-AT 1A cells were pretreated (CHX (ϩ) , lanes [3][4][5] or not (CHX (Ϫ) , lanes 1 and 2) with cycloheximide at 10 g/ml for 1 h. Then cells were maintained in the absence ((Ϫ), lanes [1][2][3][4] or presence ((ϩ), lane 5) of cycloheximide and exposed (lanes 2 and 4 -5) or not (lanes 1 and 3) to 10 Ϫ7 M Ang II for 4 h. Nuclear extracts were prepared, and EMSA was performed with 32 P-Oligo A probe. reported that Ang II-induced cyclin D1 promoter activity relies on the binding of c-Fos and c-Jun only to the AP-1-responsive element located at Ϫ954 bp in the human adrenal cell line H295R (13). However, in CHO-AT 1A cells, deletion of this putative AP-1 site did not reduce Ang II-induced promoter activity (Fig. 1). By using a similar approach, we also found an AP-1 independent regulation of cyclin D1 promoter in cultured rat aortic smooth muscle cells (RASMC). 2 This effect has also been reported by others in transforming growth factor-␣-treated esophageal squamous carcinoma cells (7) in estrogen-treated (10) and pp60 v-src -expressing (59) human mammary carcinoma cells MCF-7, in cytokine-activated hematopoietic cells (12), and in serum-stimulated vascular endothelial cells (9). These observations argue in favor of a cell type-and agonist-specific effect. In CHO-AT 1A cells, we have mapped a key region localized between nucleotides Ϫ136 and Ϫ96 that contributes to the activation of the cyclin D1 promoter by Ang II.
It was also reported that the ATF/cAMP-responsive elementbinding protein-binding site in the cyclin D1 promoter was implicated in transcriptional activation of the gene (9,10,59). However, results from Fig. 1 reveal that further deletion of this site located between nucleotides Ϫ96 and Ϫ29 did not affect the 2 L. Guillemot, A. Levy, and B. Rothhut, unpublished data. 2.4-fold stimulation obtained with the D1⌬-96 construct and therefore identify another Ang II-responsive region located between nucleotides Ϫ29 and ϩ139 of the 5Ј-untranslated region of the promoter. A possible candidate element within the Ϫ29 bp proximal region of the cyclin D1 promoter has been described by others (14) and shown to be a target for cEts-2.
Here we report that the transcriptional activation of the cyclin D1 gene by Ang II is mainly mediated through the induction of the early growth response (Egr-1) transcriptional factor binding activity with a cis-regulatory element in the cyclin D1 promoter within nucleotides Ϫ136 to Ϫ96. EMSAs demonstrated that Egr-1 binds only one of the two putative Egr motifs within this region of the cyclin D1 promoter (i.e. the 3Ј Egr motif located at Ϫ104 bp, which overlaps two consecutive Sp1 sites). A dual interplay between Egr-1 and Sp1 response elements has been previously described in the proximal platelet-derived growth factor A-chain promoter (52), the human interleukin 2 gene promoter (60), the rat luteinizing hormone-␤ gene promoter (61), and the rat phenylethanolamine N-methyltransferase gene promoter (62) but not in the human cyclin D1 promoter (7). The model involves the displacement of Sp1 by Egr-1 at overlapping motifs in the proximal promoter (52,63). Other regulatory mechanisms have been described for the upor down-regulation of promoters: the inducible presence of a heat-labile and protease-sensitive Sp1-negative regulator (64) or the reduced O-glycosylation of Sp1 leading to an increased susceptibility to proteasome degradation (65).
Using a functional approach, we show that cells transfected with plasmids containing a mutation in the 3Ј Egr motif (mutant 1 and mutant 3) abrogated Ang II-induced promoter/ reporter gene activity, while mutation in the 5Ј Egr-like site (mutant 2) did not abolish this effect (Fig. 2B). Although several reports have suggested a role for Sp1 sites in the transcriptional activation of the cyclin D1 gene (9,66), in CHO-AT 1A cells, Sp1 does not seem to play a central role in Ang IImediated cyclin D1 promoter activity but rather seems to be responsible for the basal expression of cyclin D1. Occupancy of the site may prevent the formation of a repressive chromatin structure (67). This could then explain the previously observed data with mutant 1 and mutant 3 in luciferase assays (Fig. 2B). Basal activities were severely diminished due to the possible inability of the mutated promoter sequences to bind the Sp1 protein, as also suggested in EMSAs (Fig. 3B, lanes 3 and 4). In addition, using RASMC, we found that the 3Ј Egr motif on the cyclin D1 promoter was also regulated by Ang II. Indeed, with mutant 1 and mutant 3, Ang II-induced promoter activity was completely abolished. 2 These results argue in favor of CHO-AT 1A cells to represent a valid model with which to characterize Ang II-dependent gene activation.
egr-1 belongs to a family of genes termed immediate early genes, including c-fos, c-myc, and c-jun that are rapidly induced by signals stimulating mitogenesis and differentiation (68). The MAPK/ERK activation pathway has been implicated in the induction of egr-1 expression by urea (69), endothelin-3 (70), growth hormone (71), cellular stress (55,72,73), and Ang II (74). We have previously reported that in CHO-AT 1A cells, Ang II-induced cyclin D1 promoter activity was dependent upon Ras/Raf-1/MEK/ERK, PI3K, and SHP-2 activities (18). Here we demonstrate that this induction is partly mediated by Egr-1 transcription factor synthesis and binding to an Egr response element of the cyclin D1 promoter in a specific and time-dependent manner. We show that Egr-1 is rapidly and transiently induced by Ang II and that this activation required de novo protein synthesis. In addition, U0126 completely inhibited Ang II-induced Egr-1-promoter activity and endogenous protein expression, while an inhibitor of PI3K and a dominant negative mutant (SHP-2 CS) or SH2 domains of SHP-2, which partially inhibited Ang II-induced ERK phosphorylation (18), also attenuated this effect. Taken together, these results suggest that Ang II-dependent activation of the Ras/MEK/ERK pathway contributes to Ang II-stimulated cyclin D1 expression through induction of Egr-1 and raise the possibility that PI3K and SHP-2 may regulate transcriptional activation of Egr-1. In addition, we found that the Ϫ29 bp proximal region of the cyclin D1 promoter is also dependent upon Ras/MEK/ERK, PI3K, and SHP-2 activities.
In CHO-AT 1A cells, Ang II-responsive sequences of the Egr-1 promoter and interacting proteins have not yet been identified. The human Egr-1 promoter contains five serum response elements (SREs) organized into a downstream and an upstream cluster. It has been shown that transcriptional activation of this gene in response to human granulocyte-macrophage colony-stimulating factor, mouse interleukin-3, urea, and antigen cross-linking (69,75,76) is invariably mediated through SREs that are regulated by complexes composed of serum response factor and Ets protein family members. However, other functional cis-acting elements have been reported (i.e. EBS (Egr-1 binding site) (48) and cAMP response elements (77)). SRE/Ets binding sites on the Egr-1 promoter are occupied by multiprotein complexes that are similar to the ternary complex described for the c-fos SREs. Possible candidates of MAPK/ERK activation could be the Ets domain-containing proteins Elk1, Sap-1a, and Fli-1, members of the ternary complex factor family of transcription factors. Indeed, a ternary complex of serum response factor and Elk-1 or Sap-1a bound to SRE has been shown to mediate growth hormone-induced transcription of Egr-1 (78). Both Elk-1 and Sap-1a can be activated by ERKs through phosphorylation of serine residues (79 -81). Elk-1 but not c-Jun activity is involved in Egr-1 activation by fluid shear stress (55), and the Ets protein Fli-1 and serum response factor form a complex on the SRE of the Egr-1 promoter, which is required for egr-1 gene transcription (82).
SHP-2 has been recently implicated as a positive regulator of the Egr-1 promoter by leptin receptor stimulation via activation of the MAPK/ERK pathway (83). Here we demonstrate for the first time that G-protein-coupled receptor induction of cyclin D1 is mainly mediated by Egr-1 and that the pathways connecting Ang II to this activation are dependent upon Ras/ MEK/ERK, PI3K and SHP-2 activities.