Cyclooxygenase-2 is overexpressed in HER-2/neu-positive breast cancer: evidence for involvement of AP-1 and PEA3.

Markedly increased levels of cyclooxygenase-2 (COX-2) mRNA, protein, and prostaglandin E(2) synthesis were detected in HER-2/neu-transformed human mammary epithelial cells (184B5/HER) compared with its nontransformed partner cell line (184B5). HER-2/neu stimulated COX-2 transcription via the Ras --> Raf --> MAPK pathway. The inductive effects of HER-2/neu were mediated, in part, by enhanced binding of AP-1 (c-Jun, c-Fos, and ATF-2) to the cyclic AMP-response element (-59/-53) of the COX-2 promoter. The potential contribution of the transcription factor PEA3 was also investigated. Elevated levels of PEA3 were detected in 184B5/HER cells. A PEA3 site (-75/-72) was identified juxtaposed to the cyclic AMP-response element. HER-2/neu-mediated activation of the COX-2 promoter was blocked by mutagenizing the PEA3 site or overexpressing antisense to PEA3. To determine whether HER-2/neu status was also a determinant of COX-2 expression in vivo, we compared levels of COX-2 protein in HER-2/neu-positive and -negative human breast cancers. Increased amounts of COX-2 were detected in HER-2/neu-positive tumors. Taken together, these results suggest that closely spaced PEA3 and cyclic AMP-response elements are required for HER-2/neu-mediated induction of COX-2 transcription. The clear relationship between HER-2/neu status and COX-2 expression in human breast tumors suggests that this mechanism is likely to be operative in vivo.

The HER-2/neu (erbB-2) gene encodes a 185-kDa transmembrane receptor with tyrosine kinase activity that belongs to the family of receptors for epidermal growth factor (1). Amplification and/or overexpression of HER-2/neu occurs in 20 -30% of human breast cancers, and increased expression has been associated with a poor prognosis for the patient (2)(3)(4). Overexpression of HER-2/neu causes non-neoplastic mammary epithelial cells to undergo malignant transformation (5). Transgenic mice that express HER-2/neu develop mammary tumors (6,7). Antibodies directed at HER-2/neu inhibit the growth of tumors that express high levels of this receptor (8,9). Overexpression of HER-2/neu activates the Ras pathway and increases mito-genic signaling (10). Although the precise mechanism by which HER-2/neu regulates oncogenesis is incompletely understood, the PEA3 subfamily of ets genes appears to be important (11,12). PEA3 is overexpressed in 93% of HER-2/neu-positive human breast tumors (13). Moreover, expression of dominant negative PEA3 in the mammary gland of MMTV-neu transgenic mice dramatically delayed the onset of mammary tumors and reduced the number and size of tumors in individual mice (12). It is logical to postulate, therefore, that the identification of PEA3 target genes should provide new insights into the mechanism by which overexpression of HER-2/neu causes breast cancer.
Multiple lines of evidence suggest that cyclooxygenase-2 (COX-2), 1 an enzyme that catalyzes the formation of prostaglandins (PGs), is also important in carcinogenesis. COX-2 is overexpressed in transformed cells (14,15) and in malignant tissues (16 -23). Recently, overexpression of COX-2 was found to be sufficient to induce breast cancer in multiparous transgenic mice (24). Mice engineered to be null for COX-2 were protected against the development of both intestinal and skin tumors (25,26). In addition to the genetic evidence implicating COX-2 in carcinogenesis, selective inhibitors of COX-2 reduce the formation and growth of tumors in experimental animals (27)(28)(29)(30)(31) and decrease the number of intestinal tumors in familial adenomatous polyposis patients (32). Several different mechanisms can potentially explain the link between COX-2 and cancer. Enhanced synthesis of COX-2-derived PGs favors tumor growth by stimulating cell proliferation (33), promoting angiogenesis (34,35), increasing invasiveness (36,37), and inhibiting apoptosis (38,39).
A link between HER-2/neu signaling and COX-2 expression has been established (40,41). Overexpression of HER-2/neu in the biliary epithelium of transgenic mice led to increased levels of COX-2 (40). Additionally, activation of the HER2/HER3 pathway induced COX-2 in colorectal cancer cells (41). Although these studies established a relationship between HER-2/neu signaling and COX-2 expression, the underlying mechanism was not evaluated. The main purpose of the current study was to elucidate the signaling mechanism and cis-acting elements in the COX-2 promoter that mediate the inductive effects of HER-2/neu. We show that HER-2/neu stimulates COX-2 transcription via the Ras pathway in cultured human mammary epithelial cells. Notably, closely spaced PEA3 and AP-1 sites are necessary for HER-2/neu-mediated induction of COX-2. This mechanism is likely to be operative in vivo because COX-2 was also overexpressed in HER-2/neu-positive human breast cancers. Tissue Culture-The 184B5 and 184B5/HER cell lines have been described previously (5,42). The 184B5 cell line is an immortalized but FIG. 1. HER-2/neu-mediated transformation of human mammary epithelial cells is associated with increased rates of COX-2 transcription. A, 184B5 and 184B5/HER cells were grown in culture medium for 6 h. This medium was replaced with fresh medium containing 10 M sodium arachidonate. 30 min later, the medium was collected to determine amounts of PGE 2 . Production of PGE 2 was determined by enzyme immunoassay. Columns, means; bars, S.D.; n ϭ 6. *, p Ͻ 0.001 compared with 184B5 cells. B, cellular lysate protein (25 g/lane) was loaded onto a 10% SDS-polyacrylamide gel, electrophoresed, and subsequently transferred onto nitrocellulose. The immunoblot was probed with antibody specific for COX-2. Cell lysates were prepared from 184B5 (lane 2) and 184B5/HER (lane 3) cells. Lane 1 represents an ovine Cox-2 standard. C, total cellular RNA was isolated from 184B5 (lane 1) and 184B5/HER (lane 2) cells. 10 g of RNA was added to each lane. The blot was hybridized with probes that recognized COX-2 mRNA and 18 S rRNA. D, nuclei were isolated from 184B5 (lane 1) and 184B5/HER (lane 2) cells. The COX-2 and 18 S rRNA cDNAs were immobilized onto nitrocellulose membranes and hybridized with labeled nascent RNA transcripts. E, total cellular RNA was isolated from HEK293 cells transfected with empty vector (lane 1) or HER-2/neu (lane 2). 10 g of RNA was added to each lane. The blot was hybridized with probes that recognized COX-2 mRNA and 18 S rRNA.

Materials-Minimum
FIG. 2. HER-2/neu stimulates COX-2 promoter activity via the Ras pathway. 184B5 cells were transfected with 0.9 g of a human COX-2 promoter construct (Ϫ327/ϩ59) (Control) and 0.2 g of pSV␤gal. A, cells were co-transfected with 0.4 g of expression vectors for HER-2/neu, Ras, and dominant negative Ras. B, cells were co-transfected with 0.4 g of expression vectors for HER-2/neu, Raf-1, and dominant negative Raf-1. The total amount of DNA in each reaction was kept constant at 2 g by using corresponding empty expression vectors. Luciferase activity represents data that have been normalized to ␤-galactosidase activity. Columns, means; bars, S.D.; n ϭ 6.
non-tumorigenic human breast epithelial cell line that was established from a reduction mammoplasty (42). The 184B5/HER cell line was derived by stably transfecting 184B5 cells with a mutationally activated human HER-2/neu oncogene (5). These cells form tumors when injected into athymic nude mice (5). Cells were maintained in minimum Eagle's medium/KBM mixed in a ratio of 1:1 (basal medium) containing epidermal growth factor (10 ng/ml), hydrocortisone (0.5 g/ml), transferrin (10 g/ml), gentamicin (5 g/ml), and insulin (10 g/ml) (growth medium). Cells were grown to 60% confluence, trypsinized with 0.05% trypsin, 2 mM EDTA, and plated for experimental use. HEK293 cells were obtained from American Type Culture Collection (Manassas, VA) and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. PGE 2 Production-5 ϫ 10 4 cells/well were plated in 6-well dishes and grown to 60% confluence in growth medium. The levels of PGE 2 released by cells were measured by enzyme immunoassay. Production of PGE 2 was normalized to protein concentrations.
Analysis of Human Breast Cancers-Immunohistochemistry for HER-2/neu was performed on formalin-fixed paraffin-embedded tissue sections using avidin/biotin/peroxidase as described (43), with counterstain using hematoxylin. Pretreatment consisted of microwave heating for 5 min in 0.01 M citrate buffer. Anti-HER-2/neu antibody was used at a dilution of 1:1000; normal rabbit serum was used as the primary antibody for negative control sections. Immunoreactivity was scored as positive if greater than 20% of the tumor cells were reactive and intensity of signal was subjectively scored as 0 to 3 ϩ by visual examination. For the studies described here, cases scored as 2 or 3 ϩ were considered HER-2/neu-positive; cases scored as 0 were considered HER-2/neunegative. Protocols for procurement and use of tissues for research have been approved by the Committee on Human Rights in Research.
Purification of frozen tumor tissue was performed under direct microscopic guidance using cryostat histologic sections as a guide. Nonneoplastic tissue was trimmed away, and only non-necrotic carcinoma was used for immunoblot analysis. Samples were composed of greater than 80% tumor.
Ten mg of frozen microdissected human breast cancer tissue was thawed. The tissue was sonicated in 1 ml of lysis buffer and then centrifuged at 10,000 ϫ g for 10 min at 4°C. After discarding the pellet, the supernatant was preabsorbed with 20 l of normal goat IgG and 20 l of rabbit IgG at 4°C; 20 l of protein G PLUS-agarose was then added. The mixture was then centrifuged at 3,000 ϫ g for 5 min at 4°C. The pellet was discarded. 20 l of rabbit anti-human COX-2 antiserum was then added to the supernatant; the mixture was then incubated at 4°C on a rocker platform for 1 h. 20 l of protein A-agarose was then added, and the mixture was incubated on a rocker platform for 16 h at 4°C; the mixture was then centrifuged at 3,000 ϫ g for 5 min at 4°C. The supernatant was discarded. After washing the pellet four times with RIPA buffer, the pellet was resuspended. SDS-PAGE and immunoblotting were then performed as described below.
Western Blotting-Cell lysates were prepared by treating cells with lysis buffer (150 mM NaCl, 100 mM Tris (pH 8.0), 1% Tween 20, 50 mM diethyldithiocarbamate, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml trypsin inhibitor, and 10 g/ml leupeptin). Lysates were sonicated for 20 s on ice and centrifuged at 10,000 ϫ g for 10 min to sediment the particulate material. The protein concentration of the supernatant was measured by the method of Lowry et al. (44). SDS-PAGE was performed under reducing conditions on 10% polyacrylamide gels as described by Laemmli (45). The resolved proteins were transferred onto nitrocellulose sheets as detailed by Towbin et al. (46). The nitrocellulose membrane was then incubated with primary antibodies. Secondary antibody to IgG conjugated to horseradish FIG. 4. HER-2/neu induces COX-2 promoter activity via ERK1, JNK, and p38 MAP kinases. Cells were transfected with 0.9 g of a human COX-2 promoter construct ligated to luciferase (Ϫ327/ϩ59) and 0.2 g of pSV␤gal. A, cells received COX-2 promoter alone (Control) or COX-2 promoter and 0.4 g of expression vectors for HER-2/neu, ERK1, HER-2/neu plus ERK1 or HER-2/neu plus ERK1 dominant negative. B, cells received COX-2 promoter alone (Control) or COX-2 promoter and 0.4 g of expression vectors for HER-2/neu, JNK, HER-2/neu plus JNK or HER-2/neu plus JNK dominant negative. C, cells received COX-2 promoter alone (Control) or COX-2 promoter and 0.4 g of expression vectors for HER-2/neu, p38, HER-2/neu plus p38 or HER-2/neu plus p38 dominant negative. The total amount of DNA in each reaction was kept constant at 2 g by using corresponding empty expression vectors. Luciferase activity represents data that have been normalized to ␤-galactosidase activity. Columns, means; bars, S.D.; n ϭ 6.
peroxidase was used. The blots were probed with the ECL Western blot detection system according to the manufacturer's instructions.
Northern Blotting-Total cellular RNA was isolated from cell monolayers using an RNA isolation kit from Qiagen Inc. 10 g of total cellular RNA per lane were electrophoresed in a formaldehyde-containing 1.2% agarose gel and transferred to nylon-supported membranes. After baking, membranes were prehybridized overnight in a solution containing 50% formamide, 5ϫ sodium chloride/sodium phosphate/ EDTA buffer (SSPE), 5ϫ Denhardt's solution, 0.1% SDS, and 100 g/ml single-stranded salmon sperm DNA and then hybridized for 12 h at 42°C with radiolabeled cDNA probes for human COX-2 and 18 S rRNA.
COX-2 and 18 S rRNA probes were labeled with [ 32 P]CTP by random priming. After hybridization, membranes were washed twice for 20 min at room temperature in 2ϫ SSPE, 0.1% SDS, twice for 20 min in the same solution at 55°C, and twice for 20 min in 0.1ϫ SSPE, 0.1% SDS at 55°C. Washed membranes were then subjected to autoradiography.
Nuclear Run-off Assay-2.5 ϫ 10 5 cells were plated in four T150 dishes for each condition. Cells were grown in growth medium until ϳ60% confluent. Nuclei were isolated and stored in liquid nitrogen. For the transcription assay, nuclei (1.0 ϫ 10 7 ) were thawed and incubated in reaction buffer (10 mM Tris (pH 8), 5 mM MgCl 2 , and 0.3 M KCl) containing 100 Ci of uridine 5Ј-[␣-32 P]triphosphate and 1 mM unlabeled nucleotides. After 30 min, labeled nascent RNA transcripts were isolated. The human COX-2 and 18 S rRNA cDNAs were immobilized onto nitrocellulose and prehybridized overnight in hybridization buffer. Hybridization was carried out at 42°C for 24 h using equal cpm/ml of labeled nascent RNA transcripts for each treatment group. The membranes were washed twice with 2ϫ SSC buffer for 1 h at 55°C and then FIG. 5. HER-2/neu-mediated induction of COX-2 promoter activity is mediated via the cyclic AMP-response element. A, shown is a schematic of the human COX-2 promoter. B, 184B5 cells were transfected with 0.9 g of a series of human COX-2 promoter deletion constructs ligated to luciferase (Ϫ1432/ϩ59, Ϫ327/ϩ59, Ϫ220/ϩ59, Ϫ124/ϩ59, Ϫ52/ϩ59) alone (empty bars) or 0.9 g of the indicated COX-2 promoter deletion construct plus 0.9 g of expression vector for HER-2/neu (black bars). C, 184B5 cells were transfected with 0.9 g of a series of human COX-2 promoter-luciferase constructs (Ϫ327/ϩ59; KBM; ILM; CRM). The bars labeled HER-2/neu also received 0.9 g of expression vector for HER-2/neu. KBM represents the Ϫ327/ϩ59 COX-2 promoter construct in which the NF-B site was mutagenized; ILM represents the Ϫ327/ϩ59 COX-2 promoter construct in which the NF-IL6 site was mutagenized; CRM refers to the Ϫ327/ϩ59 COX-2 promoter construct in which the CRE was mutagenized. B and C, all cells received 0.2 g of pSV␤gal. The total amount of DNA in each reaction was kept constant at 2 g by using empty vector. Reporter activities were measured in cellular extract 24 h following transfection. Luciferase activity represents data that have been normalized to ␤-galactosidase activity. Columns, means; bars, S.D.; n ϭ 6.

Cyclooxygenase-2 Is Induced by the HER-2/neu Oncogene
Transient Transfection Assays-184B5 and 184B5/HER cells were seeded at a density of 5 ϫ 10 4 cells/well in 6-well dishes and grown to 50 -60% confluence. For each well, 2 g of plasmid DNA were introduced into cells using 8 g of LipofectAMINE as per the manufacturer's instructions. After 7 h of incubation, the medium was replaced with basal medium. The activities of luciferase and ␤-galactosidase were measured in cellular extract as described previously (48).
Electrophoretic Mobility Shift Assay-Cells were harvested and nuclear extracts were prepared. For binding studies, oligonucleotides containing the CRE or PEA3 sites of the COX-2 promoter were used. The complementary oligonucleotides were annealed in 20 mM Tris (pH 7.6), 50 mM NaCl, 10 mM MgCl 2 , and 1 mM dithiothreitol. The annealed oligonucleotide was phosphorylated at the 5Ј-end with [␥-32 P]ATP and T4 polynucleotide kinase. The binding reaction was performed by incubating 5 g of nuclear protein in 20 mM HEPES (pH 7.9), 10% glycerol, 300 g of bovine serum albumin and 1 g of poly(dI⅐dC) in a final volume of 10 l for 10 min at 25°C. The labeled oligonucleotide was added to the reaction mixture and allowed to incubate for an additional 20 min at 25°C. The samples were electrophoresed on a 4% nondenaturing polyacrylamide gel. The gel was then dried and subjected to autoradiography at Ϫ80°C.
Statistics-Comparisons between groups were made by the Student's t test or the 2 test of proportions. A difference between groups of p Ͻ 0.05 was considered significant.

COX-2 Is Overexpressed in HER-2/neu-transformed Human
Mammary Epithelial Cells-Cell culture was used to investigate whether HER-2/neu regulated the expression of COX-2. PGE 2 synthesis was increased by more than 10-fold in 184B5/ HER cells compared with the 184B5 counterpart (Fig. 1A). Western blotting was carried out to determine whether the

FIG. 7. c-Jun is important for HER-2/neu-mediated activation of the COX-2 promoter.
A, cells were transfected with 0.9 g of a human COX-2 promoter construct ligated to luciferase (Ϫ327/ϩ59) and 0.2 g of pSV␤gal. Cells received COX-2 promoter alone (Control) or COX-2 promoter construct and 0.4 g of expression vectors for HER-2/ neu, c-Jun, HER-2/neu plus c-Jun, or HER-2/neu plus c-Jun dominant negative. The total amount of DNA in each reaction was kept constant at 2 g by using empty vector. Reporter activities were measured 24 h after transfection. Luciferase activity represents data that have been normalized with ␤-galactosidase activity. Columns, means; bars, S.D.; n ϭ 6. B, cellular lysate protein was prepared from 184B5 (lane 1) and 184B5/HER cells (lane 2) and loaded (50 g/lane) onto a 10% SDSpolyacrylamide gel. The immunoblot was probed with an antibody to c-Jun.

FIG. 8. PEA3 is important for HER-2/neu-mediated induction of COX-2.
A, cellular lysate protein (25 g/lane) was loaded onto a 10% SDS-polyacrylamide gel, electrophoresed, and subsequently transferred onto nitrocellulose. The immunoblot was probed with antibody specific for PEA3. Cell lysates were prepared from 184B5 (lane 1) and 184B5/ HER (lane 2) cells. Lane 3 represents a PEA3 standard. B, 184B5 cells were transfected with 0.9 g of a human COX-2 promoter construct ligated to luciferase (Ϫ327/ϩ59) (Control) or COX-2 promoter plus expression vector for HER-2/neu (0.4 g) or COX-2 promoter, HER-2/ neu plus antisense to PEA3 (0.4 g). The total amount of DNA in each reaction was kept constant at 2 g by using empty vector. Reporter activities were measured 24 h after transfection. Luciferase activity represents data that have been normalized with ␤-galactosidase activity. Columns, means; bars, S.D.; n ϭ 6.
Cyclooxygenase-2 Is Induced by the HER-2/neu Oncogene differences in PGE 2 production were related to differences in amounts of COX-2. Fig. 1B shows that levels of COX-2 protein were much higher in 184B5/HER cells than in 184B5 cells. COX-1 was not detectable by immunoblotting in these cell lines. To elucidate further the mechanism responsible for the changes in amounts of COX-2 protein, we determined steadystate levels of COX-2 mRNA by Northern blotting. As shown in Fig. 1C, higher levels of COX-2 mRNA were also detected in 184B5/HER cells than in 184B5 cells. Differences in levels of mRNA could reflect altered rates of transcription. To investigate this possibility, nuclear run-offs were performed. Higher rates of synthesis of nascent COX-2 mRNA were observed in 184B5/HER cells than in 184B5 cells (Fig. 1D). The link between HER-2/neu and COX-2 was also established in HEK293 cells. Overexpressing HER-2/neu in these cells induced COX-2 (Fig. 1E).
Defining the Signaling Mechanism by Which HER-2/neu Induces COX-2-Transient transfections were performed to determine whether the inductive effects of HER-2/neu on COX-2 were mediated via the Ras pathway. Overexpression of either HER-2/neu or ras stimulated COX-2 promoter activity in 184B5 cells ( Fig. 2A). Moreover, HER-2/neu-mediated stimulation of COX-2 promoter activity was suppressed by dominant negative ras. A downstream target of Ras is Raf-1. Hence, we determined whether Raf-1 mediated the inductive effects of HER-2/neu on COX-2. As shown in Fig. 2B, overexpressing Raf-1 led to about a 3-fold increase in COX-2 promoter activity. The stimulation of COX-2 promoter activity by HER-2/neu was blocked by kinase-deficient Raf-1 (Fig. 2B).
Ras signaling can alter gene expression by three distinct MAPK cascades (49). It was important, therefore, to investigate whether increased MAPK activity contributed to the induction of COX-2 in HER-2/neu transformed cells. The activities of extracellular signal-regulated kinase 1/2 (ERK1/2), p38 MAPK, and c-Jun N-terminal kinase (JNK) were higher in 184B5/HER cells than in 184B5 cells (Figs. 3, A-C). Subsequently, experiments were done to show that increased MAPK activity was linked to elevated levels of COX-2 in HER-2/neu transformed cells. In the first experiment, we utilized PD 98059, a specific inhibitor of MAPK kinase, which prevents activation of ERK1 and ERK2 (50). Treatment with PD 98059 caused a decrease in amounts of COX-2 in 184B5/HER cells (Fig. 3D). Similarly, SB 202190, a selective inhibitor of p38 MAPK (51), down-regulated amounts of COX-2 in these cells (Fig. 3E). To investigate further the importance of MAPK in mediating the induction of COX-2 in HER-2/neu transformed cells, a series of transient transfections was performed (Fig. 4). Transiently overexpressing ERK1, JNK, or p38 MAPKs led to severalfold increases in COX-2 promoter activity. The induction of COX-2 promoter activity by HER-2/neu was inhibited by dominant negatives for ERK1, JNK, and p38 (Fig. 4).
The Cyclic AMP-response Element and AP-1 Are Necessary for the Induction of COX-2 by HER-2/neu in Mammary Epithelial Cells-We next were interested in identifying the region of the COX-2 promoter that was important for mediating the inductive effects of HER-2/neu. Transient transfections were performed with a series of human COX-2 promoter-deletion constructs (Fig. 5A) in 184B5 cells. As shown in Fig. 5B, overexpression of HER-2/neu led to nearly a 3-fold increase in COX-2 promoter activity when a Ϫ1432/ϩ59 COX-2 promoter construct was utilized. A stepwise decrease in basal COX-2 promoter activity was observed when shorter constructs were used. However, the magnitude of induction by HER-2/neu remained nearly 3-fold with all promoter deletion constructs except the Ϫ52/ϩ59 construct (Fig. 5B). The Ϫ52/ϩ59 COX-2 promoter construct was not stimulated by HER-2/neu. This result implies that one or more promoter elements lying between Ϫ53 and Ϫ123 is necessary for HER-2/neu-mediated induction of COX-2. A CRE is present between nucleotides Ϫ59 and Ϫ53 raising the possibility that this element could be involved in mediating the inductive effects of HER-2/neu. To test this notion, transient transfections were performed utilizing COX-2 promoter constructs in which specific known enhancer elements including the CRE were mutagenized. As shown in Fig. 5C, HER-2/neu-mediated stimulation of COX-2 promoter activity was abrogated by mutagenizing the CRE site. By contrast, mutagenizing the NF-B and NF-IL6 sites had no effect on COX-2 promoter function. To confirm the importance of the CRE for mediating the induction of COX-2 by HER-2/ neu, a separate series of transient transfections was performed. We examined the effects of a CRE-decoy oligonucleotide on HER-2/neu-mediated stimulation of COX-2 promoter activity. The CRE-decoy oligonucleotide effectively inhibited HER-2/ neu-mediated activation of the COX-2 promoter (Fig. 6A). In contrast, neither a CRE mismatch oligonucleotide nor a nonsense-sequence palindrome blocked HER-2/neu-mediated induction of COX-2 promoter activity.
Electrophoretic mobility shift assays were performed to identify the transcription factor that contributed to the induction of COX-2 in HER-2/neu transformed cells. Increased binding of nuclear proteins to the CRE site of the COX-2 promoter was detected (Fig. 6B). By contrast, binding to the NF-B and NF-IL6 sites was similar in these two cell lines (data not shown). Supershift analysis identified c-Jun, c-Fos, and ATF-2 in the binding complex (Fig. 6C). Consistent with this finding, binding was also prevented by incubating nuclear extract from 184B5/HER cells with an excess of AP-1 cold probe (Fig. 6B). Transient transfections were performed to confirm the importance of AP-1 for mediating the induction of COX-2 by HER-2/ neu. A dominant negative form of c-Jun inhibited the induction of COX-2 promoter activity by HER-2/neu (Fig. 7A). Changes in either the amount or phosphorylation state of c-Jun can alter AP-1-mediated gene expression. Hence, we compared the amounts of c-Jun and phosphorylated c-Jun in 184B5/HER and 184B5 cells. Levels of c-Jun protein (Fig. 7B) and phosphorylated c-Jun protein (data not shown) were higher in 184B5/HER cells compared with its nontransformed 184B5 partner cell line. AP-1 activity was also increased in the transformed 184B5/HER cell line compared with its nontransformed 184B5 partner cell line (data not shown). Thus, in response to overexpression of HER-2/neu, increased MAPK signaling activates AP-1, which, in turn, contributes to enhanced COX-2 gene expression via the CRE in the COX-2 promoter.
PEA3 Is Also Necessary for HER-2/neu-mediated Induction of COX-2-Increased levels of PEA3 have been detected in more than 90% of HER-2/neu-overexpressing breast cancers (13). A variety of genes are regulated by closely spaced PEA3/ets and AP-1 sequences (52). Experiments were therefore carried out to investigate the potential role of PEA3 in HER-2/neu-mediated induction of COX-2. Higher levels of PEA3 were detected in 184B5/HER cells than in 184B5 cells (Fig. 8A). To investigate whether PEA3 is important for the induction of COX-2 by HER-2/neu, transient transfections were performed. As shown in Fig. 8B, overexpressing antisense to PEA3 blocked the stimulation of COX-2 promoter activity by HER-2/neu.
There are several possible PEA3 sites (GGAA) in the COX-2 promoter (53) (Fig. 9A). Site-directed mutagenesis was used to create COX-2 promoter constructs in which each of these candidate PEA3 sites (mut1-mut3) was altered. HER-2/neu stimulated COX-2 promoter activity except when PEA3 site 1 (Ϫ72/ Ϫ75) was mutagenized (Fig. 9B). To evaluate further the importance of this site, electrophoretic mobility shift assays Cyclooxygenase-2 Is Induced by the HER-2/neu Oncogene were performed. Nuclear protein from both 184B5 and 184B5/ HER cells was incubated with a labeled oligonucleotide containing PEA3 site 1 of the COX-2 promoter. Extracts from HER-2/neu transformed cells led to increased binding to PEA3 site 1 (Fig. 9C). This binding was abolished when an excess of unlabeled consensus PEA3 oligonucleotide was added. Supershift analysis identified PEA3 in the binding complex (Fig. 9D). Taken together, these results suggest that both PEA3 site 1 and the CRE are necessary for HER-2/neu-mediated induction of COX-2.
COX-2 Is Overexpressed in HER-2/neu-positive Human Breast Cancers-Based on the above in vitro findings, it was important to determine whether COX-2 was overexpressed in HER-2/neu-positive human breast cancers. By immunoblot analysis, COX-2 was readily detected in 14 of 15 cases of HER-2/neu-overexpressing human breast cancer. Fig. 10 is a representative blot. In contrast, COX-2 was only detected in 4 of 14 cases of HER-2/neu-negative human breast cancer (p ϭ 0.0005); moreover, the levels of COX-2 were much lower in these tumors than in any of the tumors in which HER-2/neu was overexpressed (Fig. 10). DISCUSSION In the current study, we found that levels of COX-2 were increased in HER-2/neu-overexpressing human mammary epithelial cells and breast cancers. The induction of COX-2 by FIG. 9. Localization of PEA3 site that is responsible for HER-2/neu-mediated activation of COX-2 promoter. A, shown is a schematic of the human COX-2 promoter containing three potential PEA3-binding sites (see numbers  1-3). B, 184B5 cells were transfected with 0.9 g of a human COX-2 promoter construct ligated to luciferase (Ϫ327/ϩ59) or COX-2 promoter plus expression vector for HER-2/neu (0.4 g) or COX-2 promoter containing mutagenized PEA3 site 1 (mut1) plus HER-2/neu or COX-2 promoter containing mutagenized PEA3 site 2 (mut2) plus HER-2/neu or COX-2 promoter containing mutagenized PEA3 site 3 (mut3) plus HER-2/neu. All cells received 0.2 g of pSV␤gal. The total amount of DNA in each reaction was kept constant at 2 g by using empty vector. Reporter activities were measured 24 h after transfection. Columns, means; bars, S.D.; n ϭ 6. C, in lanes 1-3, 5 g of nuclear protein was incubated with a 32 Plabeled oligonucleotide containing the PEA3 site (1) of COX-2. Lane 1 represents nuclear protein from 184B5 cells; lane 2 represents nuclear protein from 184B5/ HER cells; lane 3 represents nuclear protein from 184B5/HER cells incubated with a 100-fold excess of unlabeled oligonucleotide containing a PEA3 consensus site. D, 5 g of nuclear protein was incubated with a 32 P-labeled oligonucleotide containing the PEA3 site (1) of COX-2. Lane 1 represents nuclear protein from 184B5/HER cells; lane 2 represents nuclear extract incubated with antibody to PEA3. C and D, the protein DNA complex that formed was separated on a 4% polyacrylamide gel.
HER-2/neu was mediated by the Ras pathway (Fig. 11). Ras can regulate gene expression by stimulating MAPK activities (49). Several lines of evidence suggest that HER-2/neu induced COX-2 via activation of ERK, JNK, and p38 MAPKs. First, the activities of ERK1/2, JNK, and p38 were increased in HER-2/ neu transformed cells. Second, inhibitors of MAPK kinase and p38 decreased amounts of COX-2 in HER-2/neu transformed cells. Third, overexpression of dominant negatives for ERK1, JNK, and p38 suppressed the induction of COX-2 promoter activity by HER-2/neu.
We also report that the induction of COX-2 promoter activity by HER-2/neu is mediated through closely spaced PEA3/ets and AP-1 sites located 72 and 53 nucleotides upstream of the transcriptional start site, respectively. Several observations support a role for AP-1 in mediating the induction of COX-2 by HER-2/neu. Increased binding of AP-1 to the CRE of the COX-2 promoter was detected in HER-2/neu transformed cells (Fig.  6B); c-Jun, c-Fos, and ATF-2 were identified in the DNA binding complex (Fig. 6C). The functional importance of AP-1 was established because HER-2/neu-mediated activation of the COX-2 promoter was suppressed by mutagenizing the CRE or overexpressing dominant negative c-Jun (Fig. 7A). Our results are consistent with the findings of Xie and Herschman (54,55). These investigators showed that, in response to expression of v-Src or treatment with platelet-derived growth factor, c-Jun induced murine Cox-2 via the CRE site. Tumor-promoting phorbol esters also stimulate AP-1-mediated activation of COX-2 transcription via the CRE site (56). Several findings also suggest the involvement of the PEA3/ets-binding site in mediating the induction of COX-2 by HER-2/neu. First, levels of PEA3 were elevated in HER-2/neu transformed cells. Second, increased binding of PEA3 to the PEA3 site (Ϫ75/Ϫ72) was detected when nuclear extracts were prepared from 184B5/ HER versus 184B5 cell lines (Fig. 9, C and D). Third, the ability of HER-2/neu to stimulate the COX-2 promoter was abrogated by overexpressing antisense to PEA3 or by mutagenizing the PEA3 site. Although PEA3 motifs have been identified in the COX-2 promoter (53), this is the first time a functional PEA3 site has been described. In this context, it is noteworthy that MAPKs can regulate the activities of both AP-1 and PEA3 (49,57). ERK1/2 stimulates AP-1 activity by inducing c-Fos which heterodimerizes with c-Jun (58). JNK can induce AP-1 activity by increasing the expression and phosphorylation of c-Jun (58). p38 MAPK induces AP-1 by phosphorylating ATF-2. A heterodimer composed of phospho-ATF-2 and c-Jun can induce c-Jun expression (59). Either JNK or ERK MAPK can activate PEA3 at least, in part, by phosphorylation (57). It is not surprising, therefore, that ERK1/2, JNK, and p38 MAPKs are important for HER-2/neu-mediated activation of COX-2 transcription.
Both PEA3 and AP-1 sites are necessary for HER-2/neumediated activation of the COX-2 promoter. Thus, mutating either of these closely spaced sites abolished the stimulation of COX-2 promoter activity by HER-2/neu. Although this mechanism of regulation is unprecedented for COX-2, the involvement of juxtaposed PEA3/AP-1 sites has been reported for other genes (52, 60 -63). Indeed, the involvement of closely spaced PEA3/AP-1 elements has been reported for several inducible genes including urokinase, type I collagenase, keratin 18, and tumor necrosis factor-␣ (60 -63).
Importantly, we were able to extend the above mechanistic studies to human breast cancers. As predicted from the cell culture findings, HER-2/neu status was a determinant of COX-2 expression in human breast tumors. Both the frequency and magnitude of COX-2 overexpression were markedly enhanced in HER-2/neu-positive breast cancers. The fact that PEA3 levels are elevated in 93% of HER-2/neu-positive breast cancers (13) suggests that the mechanism of regulation discussed above is likely to be operative in vivo. Recently, PEA3 subfamily Ets proteins were found to play an essential role in Neu-mediated mammary oncogenesis (12). Moreover, overexpression of COX-2 was sufficient to induce mammary cancer in multiparous transgenic mice (24). Our results suggest, therefore, that the interaction between PEA3 and COX-2 could be important for understanding Neu-induced tumor formation.
The results of this study provide other potentially significant insights. HER-2/neu induces the expression of vascular endothelial growth factor (64). COX-2-derived PGs enhance the production of vascular endothelial growth factor (65). It is reasonable to postulate, therefore, that the increased levels of vascular endothelial growth factor and angiogenesis in HER-2/neu-positive breast cancers are a consequence, in part, of HER-2/neu-mediated induction of COX-2 and PG biosynthesis. Another important issue concerns the role of nonsteroidal antiinflammatory drugs, prototypic inhibitors of COX, in preventing cancer. The finding that COX-2 is undetectable in most cases of HER-2/neu-negative breast cancer may help to explain why nonsteroidal anti-inflammatory drugs have not been shown consistently to protect against breast cancer (66,67). Our findings also imply that selective COX-2 inhibitors may be useful in preventing or treating the subset of cancers in which HER-2/neu is overexpressed. In support of this idea, treatment with a selective COX-2 inhibitor reduced the growth rate of a HER-2/neu-postive colon cancer cell line in vitro and in vivo (68). Future studies will be needed to determine whether selective inhibitors of COX-2 have a role in preventing or treating HER-2/neu-positive human breast cancers.