Mechanism of 17-β-Estradiol-induced Erk1/2 Activation in Breast Cancer Cells A ROLE FOR HER2 AND PKC-δ

Activation of mitogen-activated protein kinase (Erk/MAPK) is a critical signal transduction event for estrogen (E2)-mediated cell proliferation. Recent studies from our group and others have shown that persistent activation of Erk plays a major role in cell migration and tumor progression. The signaling mechanism(s) responsible for persistent Erk activation are not fully characterized, however. In this study, we have shown that E2 induces a slow but persistent activation of Erk in MCF-7 breast carcinoma cells. The E2-induced Erk activation is dependent on new protein synthesis, suggesting that E2-induced growth factors play a major role in Erk activation. When MCF-7 cells were treated with E2 in the presence of an anti-HER-2 monoclonal antibody (herceptin), 60–70% of E2-induced Erk activation is blocked. In addition, when untreated MCF-7 cells were exposed to conditioned medium from E2-treated cells, Erk activity was significantly enhanced. Furthermore Erk activity was blocked by an antibody against HER-2 or by heregulin (HRG) depletion from the conditioned medium through immunoprecipitation. In contrast, epidermal growth factor receptor (Ab528) antibody only blocked 10–20% of E2-induced Erk activation, suggesting that E2-induced Erk activation is predominantly mediated through the secretion of HRG and activation of HER-2 by an autoctine/paracrine mechanism. Inhibition of PKC-δ-mediated signaling by a dominant negative mutant or the relatively specific PKC-δ inhibitor rottlerin blocked most of the E2-induced Erk activation but had no effect on TGFα-induced Erk activation. By contrast inhibition of Ras, by inhibition of farnesyl transferase (Ftase-1) or dominant negative (N17)-Ras, significantly inhibited both E2- and TGFα-induced Erk activation. This evaluation of downstream signaling revealed that E2-induced Erk activation is mediated by a HRG/HER-2/PKC-δ/Ras pathway that could be crucial for E2-dependent growth-promoting effects in early stages of tumor progression.

Normal mammary development and breast cancer growth are under the influence of steroid hormones, particularly E 2 . 1 The effects of estrogens are mediated primarily through interaction with the estrogen receptor ␣ leading to cell proliferation (1,2). Estrogen receptor ␣ has an NH 2 -terminal domain with a hormone-independent transcriptional activation function (AF-1) (3), a central DNA binding domain, and a COOHterminal ligand binding domain with a hormone-dependent transcriptional activation function (AF-2) (4,5). In addition to steroid hormones, a host of polypeptide growth factors may also play an important role in the growth regulation of breast cancer by autocrine, paracrine, or endocrine mechanism(s) (6 -10). Several groups, including ours, have confirmed that estrogendependent breast cancer cells synthesize and secrete growth factors in response to estrogen stimulation and that estrogenindependent breast cells secrete these growth factors constitutively (11)(12)(13). Because breast cancer cells have membrane receptors for several of the growth factors they secrete, it has been proposed that secreted growth factors produced by the cells can bind to receptors on the surface and activate Erk/ MAPK by autocrine and/or paracrine mechanisms (6,8,10,13).
Cell transformation often results from activation of components in signaling pathways that control cell proliferation and differentiation. These pathways are initiated from various cell surface receptors, and may converge on the MAPK cascade, a module consisting of MAP kinase kinase (MEK) and MAPK (14,15). Action of the MAPK cascade appears to be necessary for cell growth. Growth factor-regulated gene transcription and cell proliferation are blocked in mammalian cells by inhibiting MAPK activity (16) or microinjection of dominantly interfering mutants of MAPK or of antisense RNA complementary to MAPK transcripts (17,18). These studies indicate that MEK and MAPK are necessary components of cell growth.
Many polypeptide growth factors exert their function by binding to cell surface receptors that have intrinsic protein tyrosine kinase activity. A large number of receptor tyrosine kinase subclasses have been described, among which the type I/ErbB family of receptor tyrosine kinases is of particular interest due to their frequent involvement in human cancer. Four members of this family are currently known: epidermal growth factor receptor (EGFR)/ErbB-1, ErbB-2/HER-2, ErbB-3, and ErbB-4 (19 -21). Co-expression of estrogen receptor ␣ and ErbB receptors is commonly detected in normal and breast cancer cells (22). Aberrant expression of EGFR has been observed in various human tumors (23). Overexpression of ErbB-2 in the presence or absence of gene amplification is frequently found in tumors arising at many sites, especially of the breast and ovary, where it correlates with poor patient prognosis (24). Multiple lines of experimental evidence suggest that overexpression of HER-2 confers antiestrogen resistance to breast tumor cells. MCF-7 human breast cancer cells transfected with either a full-length HER-2 cDNA or with ectopic heregulin-␤1 (HRG), the HER3/4 ligand that also activates HER-2, lose sensitivity to tamoxifen (25)(26)(27)(28).
Erk can also be efficiently activated by protein kinase C (PKC) (29,30). The PKC family is comprised of at least 12 serine/threonine kinases that participate in signal transduction events in response to specific hormonal and growth factor stimuli (31). Differences in their structure and substrate requirements have permitted classification of the isoforms into three groups: 1) conventional PKCs (␣, ␤I, ␤II, and ␥), which are Ca 2ϩ -dependent and activated by both phosphatidylserine and the second messenger diacylglycerol; 2) novel PKCs (␦, ⑀, , and ), which are Ca 2ϩ -independent and regulated by diacylglycerol and phosphatidylserine; and 3) atypical PKCs ( and /), which are Ca 2ϩ -independent and do not require diacylglycerol for activation although phosphatidylserine regulates activity (32,33). The activity of PKC is several times higher in ER-negative breast cancer cell lines than ER-positive cell lines (34). Breast cancer biopsies exhibit higher levels of total PKC activity compared with the surrounding normal tissue (35). PKC-␦ plays a major role in 12-O-tetradecanoylphorbol-13acetate (TPA)-induced Raf-MEK-Erk activation (36,37).
In the present study, we have shown that HER-2 and PKC-␦ play a major role in E 2 -induced Erk activation. The results indicate that E 2 activates Erk by an autocrine/paracine mechanism through activation of a HER-2/PKC-␦/Ras signaling pathway. In contrast, TGF␣ activates Erk by stimulating an EGFR/Ras pathway. These findings indicate that HER-2 and PKC-␦ play a major role in estrogen-mediated signaling leading to Erk activation and cell proliferation.

EXPERIMENTAL PROCEDURES
Cell Lines and Reagents-The human breast carcinoma cell line MCF-7 was obtained from the American Type Culture Collection and maintained in Dulbecco's modified Eagle's medium, supplemented with 5% fetal calf serum, insulin, and antibiotics penicillin and streptomycin (Invitrogen). MCF-7 cells that have been stably transfected with fulllength HER2 cDNA (MCF-7/HER2-18) or control vector (MCF-7/neo) were generously provided by C. K. Osborne (Baylor College of Medicine, Houston, TX) and have been described previously (38). TGF-␣ was purchased from R&D Systems (Minneapolis, MN). 17␤-Estradiol, cycloheximide, and TPA were purchased from Sigma. ICI 182,780 was purchased from Tocris (Ballwin, MO). UO126 was purchased from Promega (Madison, WI). EGFR-specific monoclonal antibody clone-528, bisindolylmaleimide, rottlerin, GO6976, and Ftase-1 were purchased from Calbiochem (San Diego, CA). Herceptin (humanized anti-HER2 antibody) was generously provided by Genentech. A mouse monoclonal antibody to phosphorylated MAPK was purchased from New England Biolabs (Beverly, MA), and a rabbit polyclonal antibody to total MAPK was purchased from Zymed Laboratories Inc.(San Francisco, CA). An anti-HA antibody (clone Y-11) was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA).
Isolation of Stable Transfectants-MCF-7 cells were stably transfected with full-length EGFR cDNA (pRCMV3 vector obtained from Dr. Gordan, University of California, San Diego, CA) (39) or with HA-tagged dominant negative PKCϪ␦ cDNA with Lys-Arg point mutation in the ATP binding site (Dr. Weinstein, Columbia University) (40) by electroporation using a Bio-Rad gene pulsar at 950 microfarads and 0.22 kV/cm (t ϭ 12-14 ms). Stable transfectants were selected in the presence of 250 g/ml G418 (Invitrogen) for two to three weeks. Individual antibiotic-resistant colonies were isolated and screened for the expression of the corresponding protein by Western analysis using anti-EGFR antibody (clone-528) or anti-HA antibody (clone Y-11). All cell lines were routinely tested for mycoplasma contamination and found to be negative.
Immunoprecipitation-Cells were lysed as described above, and HER2 and EGFR were immunoprecipitated from 1 mg of total protein with 2 g of herceptin (anti-HER-2) or clone528 (anti-EGFR) by incubating overnight at 4°C. Immunocomplexes were pulled down by protein G-or protein A-coupled-agarose beads (Calbiochem) for 3-4 h at 4°C. The immunoprecipitates were separated on SDS-PAGE, transferred to nitrocellulose and probed with an anti-phosphotyrosine antibody (Signal Transduction labs) as described in Western immunoblot analysis.
Assay for Ras-GTP Levels-The GTP-bound form of Ras was isolated using a minimal Ras binding domain of Raf (41) coupled to agarose beads following the procedure previously described (42). In brief, proteins were extracted from 1 ϫ 10 6 cells in 500 l of lysis buffer (50 mM Hepes-sodium (pH 7.4), 150 mM NaCl, 0.5% Triton X-100, 20 mM MgCl 2 ) and supplemented with protease inhibitors (0.5% leupeptin, 0.5% aprotinin, and 0.02% phenylmethylsulfonyl fluoride) and 0.25% sodium deoxycholate. Affinity purification of Ras-GTP was performed from 300 l of cell lysate at 4°C for 30 min using 50 l of Raf-Ras binding domain beads (50% suspension). Raf-Ras binding domain beads were pelleted, washed twice and resuspended in 20 l of 2ϫ SDS-PAGE sample buffer. An additional 150 l of cell lysate was precipitated with cold trichloroacetic acid (10% final concentration). Precipitates were dissolved in 15 l of 1 M NaHCO 3 and 30 l of 2ϫ SDS-PAGE sample buffer. The samples were then separated by SDS-PAGE and transferred to nitrocellulose membrane and probed with anti-Ras monoclonal antibody (Transduction Laboratories, Lexington, KY). Band intensities were quantified as described in Western analysis and active Ras (Ras-GTP) was normalized to corresponding total Ras levels.
Transient Transfections-Dominant negative Ha-Ras (N17Ras) (43) or wild type Ha-Ras cDNA was transiently transfected into MCF-7 cells by LipofectAMINE (Invitrogen) according to the manufacturer's protocol. After 24 h the cells were treated with E 2 , TGF␣, or inhibitors. Subsequent procedures for cell lysis and western immunoblots were as described above.
Cell Proliferation Assay-Cell proliferation was measured using Promega's aqueous one solution cell proliferation assay according to the manufacturer's protocol (Promega). In brief, 5000 cells were plated in each well of a 96-well plate in Dulbecco's modified Eagle's medium with serum. After 24 h the medium was exchanged with serum-free and phenol red-free medium, and the cells were incubated for another 24 h. The cells were then pretreated with the inhibitors or antibody 1 h prior to 24 h of stimulation with E 2 or growth factors. The reagent was added and incubated for 2-3 h at 37°C, 5% CO 2 . The intensity of the color was measured at 490 nm using a 96-well plate reader.

E 2 and TGF␣ Induce Erk Activation by Different Signaling
Pathways in MCF-7 Cells-Erk activity is tightly controlled by dual phosphorylation of specific threonine and tyrosine residues (Thr-183 and Tyr-185) by MEK, an upstream activator of Erk (44,45). Using a specific antibody that recognizes the active, phosphorylated forms of p44/p42 MAPK (Erk1/Erk2), Western immunoblot analysis revealed that activation of Erk in MCF-7 cells occurs after treatment with E 2 and TGF␣. TGF␣ induced a rapid and transient activation of Erk that peaked within 10 min and disappeared by 2 h (Fig. 1B). In contrast, E 2 -induced a slow activation of Erk starting at 2 h, which peaked by 4 h, and was sustained for at least 24 h (Fig. 1A). To rule out a role for non-genomic E 2 -mediated signaling we treated MCF-7 cells with membrane-impermeable BSA-conjugated E 2 (BSA-E 2 ). MCF-7 cells treated with BSA-E 2 did not exhibit Erk activation at 15-min, 1-, 2-, 4-, and 24-h time points (data not shown). By using antibodies that recognize total Erk we showed that both TGF␣ and E 2 modulate Erk phosphorylation and activity, but total protein levels were unaltered. In all subsequent experiments we used a 10-min time point for TGF␣ stimulation studies and a 4-h time point for E 2 -induced Erk activation studies unless otherwise specifically stated. When MCF-7 cells were pretreated with the antiestrogen ICI 182,780 for 1 h and then stimulated with E 2 , most of the E 2 -induced Erk activation was inhibited, whereas TGF␣-induced Erk activation was maintained (Fig. 1C). E 2 -induced Erk Activation Requires Synthesis and Secretion of New Proteins-Different groups including ours have previously shown that E 2 induces synthesis and secretion of peptide growth factors such as TGF␣, EGF, HRG, etc. Because breast cancer cells have membrane receptors for several of the growth factors they secrete, it has been suggested that secreted growth factors produced by the cells can bind to receptors on the cell surface and regulate growth by autocrine and/or paracrine stimulation (7,8,12,13). To investigate the possibility of such a mechanism in E 2 -induced Erk activation we used cycloheximide (CHX), a potent inhibitor of protein synthesis. When cells were treated with CHX 1 h prior to addition of E 2 a complete inhibition of E 2 -induced Erk activation was observed ( Fig. 2A). However, when cells were treated with CHX at different time points subsequent to E 2 -stimulation, E 2 -induced Erk activation gradually escaped CHX-mediated inhibition by 3 h (Fig. 2A). To further confirm that E 2 -induced proteins such as growth factors mediate Erk activation by an autocrine/paracrine mechanism, we collected the conditioned medium after treating the cells with E 2 for 0, 1, 2, and 4 h. Phenol red-free and serumstarved MCF-7 cells were exposed to the above conditioned medium for 10 min (to avoid E 2 -induced Erk activation). We observed highest Erk activation (ϳ3-fold higher when compared with control cells) in the cells exposed to conditioned medium from the 4-h E 2 treatment (Fig. 2B). However, conditioned medium from cells exposed to E 2 and CHX at the same time did not stimulate Erk activity in MCF-7 cells. CHX did not inhibit growth factor-induced Erk activity ( Fig. 2A). Based on the above data we conclude that E 2 -induced Erk activation is mediated by an autocrine or paracrine mechanism. E 2 -induced Erk Activation Is Predominantly Mediated through HER-2 Activation-Previous studies have shown that MCF-7 cells express all four EGF growth factor receptors, EGFR, ErbB2/HER-2, ErbB-3, and ErbB4 (46). To identify the receptors involved in the E 2 -induced Erk activation, we used antibodies against EGFR and HER-2 receptors. It was previously shown that EGFR antibody Ab528 can block the activation of receptor tyrosine kinases by EGF and TGF␣ and that anti-HER-2 antibody (Ab herceptin) can block HER-2 signaling (47)(48)(49)(50). Our data showed that E 2 -induced Erk activation was only partially blocked (10 -20%) by the EGFR antibody, but the antibody against HER-2/neu significantly blocked E 2 -induced Erk activity (60 -70%) (Fig. 3A). In addition we have shown that the EGFR antibody blocked Erk activation by exogenously added TGF␣ and that herceptin antibody blocked Erk activation by exogenously added HRG (Fig. 3A).
To determine the role of HRG in E 2 -induced Erk activation, we collected the conditioned medium from cells treated with E 2 for 4 h and removed HRG from the conditioned medium with a neutralizing antibody. When serum-starved MCF-7 cells were exposed to the conditioned medium that had been depleted of HRG, Erk activity was significantly reduced (Fig. 3B). Similar experiments with a TGF␣ neutralizing antibody did not inhibit Erk activity, however. These results provide additional evidence to show that E 2 -induced Erk activation is predominantly mediated by the HER-2 receptor, possibly through interaction with HER-3 or HER-4 (51).
To further confirm that E 2 -induced growth factors bind and phosphorylate the HER-2/neu receptor, we used HER-2/MCF-7 and EGFR/MCF-7 cells. MCF-7 cells express low levels of EGF and HER-2/neu receptors, and it is very difficult to identify receptor phosphorylation by Western immunoblotting. To overcome this problem we used MCF-7 cells that are stably transfected with EGFR (EGFR/MCF-7 cells) or HER-2 (HER-2/ MCF-7 cells). Overexpression of either EGFR or HER-2/neu did not significantly alter estrogen receptor levels or E 2 -induced signaling in these cells (data not shown). After 4 h of E 2 treatment, cell lysates were prepared and immunoprecipitated with either EGFR or HER-2/neu antibody and probed with antiphosphotyrosine antibody. E 2 -induced an ϳ2-3-fold increase in HER-2/neu receptor phosphorylation but had little effect on EGFR phosphorylation (Fig. 3C). These data further confirm that E 2 -induced growth factors bind to cell surface receptors and activate signaling by autocrine and paracrine mechanisms.
PKC-␦ Mediates E 2 -and HRG-induced Erk Activation-Activation of Erk can be induced by a variety of extracellular stimuli mediated through receptor tyrosine kinases and cytokine receptors (52,53). To further define the signaling molecules involved in E 2 -induced Erk activation, we investigated whether the E 2 -dependent activation of Erk in MCF-7 cells requires activation of PKC or Ras, which are downstream of growth factor-induced, type I receptor-mediated signaling (54). When MCF-7 cells were treated with E 2 or TGF␣ in the presence or absence of the nonspecific PKC inhibitor bisindolylmaleimide (Bis), a significant inhibition of E 2 -induced Erk activation but not of TGF␣-induced Erk activation was found (Fig.  4A). To investigate the specific PKC isoform(s) involved in E 2 -indued Erk activation, we used an inhibitor of Ca 2ϩ -dependent PKC isoforms (G06976) (55) and a PKC-␦ selective inhibitor (rottlerin) (56, 57). Rottlerin strongly inhibited E 2 -induced Erk activation, indicating a PKC-␦ requirement for E 2 -induced Erk activation (Fig. 4B). However, the Ca 2ϩ -dependent PKC inhib-itor had no effect on E 2 -induced Erk activation. Neither of the PKC inhibitors was able to block TGF␣-induced Erk activation (Fig. 4B), indicating that E 2 and TGF␣ induce Erk activation by distinct signaling pathways.
To further confirm that PKC-␦-mediated signaling plays a major role in E 2 -induced Erk activation, we developed MCF-7 cells that stably express a dominant negative mutant of PKC-␦. Expression of dominant negative PKC-␦ mutants was verified by taking advantage of the observation that PKC down-regulation in response to TPA is dependent upon an active kinase (56) and an HA tag by Western blotting (Fig. 5A). In MCF-7 cells transfected with a control vector, PKC-␦ was almost totally degraded by a 24-h treatment with TPA. In contrast, PKC-␦ levels in three independent cell lines expressing dominant negative PKC-␦ were only marginally reduced in the presence of TPA, indicating that kinase-dead dominant negative PKC mutants were expressed (data not shown). We examined the ability of E 2 and TGF␣ to activate Erk in MCF-7 cells that express the dominant negative PKC-␦. As expected, the dominant negative PKC-␦ inhibited E 2 -induced Erk activation in all three clones (Fig. 5B). Dominant negative PKC-␦ expression did not, however, inhibit TGF␣-induced Erk activation

FIG. 2. Effect of CHX on E 2 -induced Erk activation and the ability of the conditioned media from E 2 -treated cells to induce Erk activation. A, cells
were serum-starved for 48 h and treated with 25 ng/ml CHX for 1 h prior to 10 Ϫ8 M E 2 stimulation for 4 h or 50 ng/ml TGF␣ stimulation for 10 min. In some cases CHX was added subsequent to E 2 stimulation as shown. B, CM was collected from MCF-7 cells treated with 10 Ϫ8 M E 2 at different times as indicated or from cells treated with 25 ng/ml CHX prior to 4 h of E 2 treatment. Freshly serum-starved MCF-7 cells were treated with the above CM for 10 min. Cell lysates from the above samples were analyzed for phospho-Erk and total Erk as described in Fig. 1.

FIG. 3. Effect of antibodies against EGFR, HER-2, HRG, and TGF␣ on E 2 -induced Erk activation and activation of HER-2 and EGFR in response to E 2 treatment.
A, after 48 h of serum starvation cells were pretreated with 10 g/ml monoclonal antibody against EGFR (Ab528) or 10 g/ml monoclonal antibody against HER-2 (herceptin) 1 h prior to 4h of 10 Ϫ8 M E 2 or 10 min of 50 ng/ml TGF␣ or HRG stimulation. B, CM collected from 4 h of E 2 -treated cells was neutralized by immunoprecipitation for 2 h with antibodies against HRG or TGF␣ and precleared with protein A-agarose beads. Fresh serum-starved MCF-7 cells were then treated with the above neutralized CM or precleared CM for 10 min. Cell lysates were analyzed for phospho-Erk and total Erk as described in Fig 1. C, cell lysates obtained from the cells treated with 4 h of E 2 or 10 min of HRG, TGF␣, or ICIϩE 2 were immunoprecipitated overnight with 2 g/ml antibodies against HER-2 (herceptin) and EGFR (Ab528). Immunocomplexes were pull down by protein A (herceptin)-and protein G (Ab528)-agarose beads, separated by SDS-PAGE, and analyzed for phospho-Tyr by Western immunoblotting.

FIG. 4. Effect of various PKC inhibitors on E 2 -and TGF␣-induced Erk activation.
After 48 h of serum starvation cells were treated with 5 M Bis, a pan-PKC inhibitor, (A) or 5 M rottlerin (Rot), a PKC-␦ specific inhibitor or an inhibitor of calcium-dependent PKC isoforms, 10 nM GO6979 (G0), 1 h prior to E 2 or TGF␣ stimulation (B). Cell lysates were analyzed for phospho-Erk and total Erk as described in Fig. 1.  (Fig. 5C), indicating that PKC-␦ plays a selective role in E 2induced Erk activation in MCF-7 cells. HRG-dependent activation of HER-2/neu signaling by heterodimerization with ErbB-3 is well established in MCF-7 cells (51). We investigated whether inhibiting PKC-␦ can block HRG-induced Erk activation. All three PKC-␦ dominant negative-expressing clones exhibited greatly reduced HRG-induced Erk activation (Fig. 5D). These data confirm that HRG-induced Erk activation is mediated through PKC-␦-mediated signaling. E 2 Activates Ras through the PKC-␦ Signaling Pathway-Previous reports have shown that PKC can regulate the Ras/ Raf/MEK/ERK pathway either through direct activation of Raf independent of Ras (58) or through phosphorylation of Shc upstream of the Grb2-SOS complex (59), enabling Ras activation. We investigated whether the activation of Erk in MCF-7 cells in response to E 2 and TGF␣ requires activation of Ras. When MCF-7 cells were treated with E 2 or TGF␣ in the presence or absence of an inhibitor of farnesyl transferase (Ftase-1), which should prevent the maturation of Ras, both E 2 -and TGF␣-induced Erk activation was significantly inhibited (Fig.  6A). This indicates that both E 2 and TGF␣ induce Erk activation in a Ras-dependent manner. To further establish the role of Ras, we transiently expressed HA-tagged N17-Ras and wild type-Ras in MCF-7 cells and then treated these cells with E 2 for 4 h and TGF␣ for 10 min. As expected expression of a dominant negative Ras in which amino acid 17 is changed to Asn (N17-Ras) significantly inhibited both E 2 -and TGF␣-induced Erk activation (Fig. 6B), confirming that E 2 -induced Erk activation pathway operates in a Ras-dependent manner. The efficiency of the transient transfection was monitored by Western blotting for the HA tag (Fig. 6B).
We also examined the effect of PKC inhibitors on Ras activation by E 2 to establish whether Ras is upstream or downstream of PKC-␦ in the Erk activation cascade. Ras activation was measured by affinity isolation of Ras-GTP using a Ras binding domain of Raf coupled to agarose beads (41,60). When MCF-7 cells were treated with bisindolylmaleimide (a general PKC inhibitor) or rottlerin (a PKC-␦-specific inhibitor) prior to E 2 treatment, there was a significant inhibition of E 2 -induced Ras activation (Fig. 7A). Similar experiments with GO6976, which inhibits all calcium-dependent PKC isoforms, had no effect on the E 2 -induced Ras activation. These data suggest that Ras is down stream of PKC-␦ in E 2 -induced signaling leading to Erk activation. TPA was previously shown to activate Ras in a PKC-dependent manner; therefore, we used TPAtreated cells as positive controls in these experiments. It is significant that the general PKC inhibitor, Bis, was unable to block TGF␣-induced Ras activation (Fig. 7B), which further supports a distinct pathway for TGF␣ as compared with E 2 signaling. Pretreatment of cells with Ftase-1 significantly inhibited E 2 -, TGF␣-, and TPA-stimulated Ras activation (Fig. 7,  A and B). Together, these results provide strong evidence that E 2 -induced Erk activation is predominantly mediated through HRG, HER-2/neu, PKC-␦, and Ras in MCF-7 cells.
Cell Proliferation-To investigate the effects of herceptin and PKC inhibitors on E 2 -and growth factor-induced cell proliferation, we examined its effects on MCF-7 cells. E 2 -, TGF␣and HRG-induced a 2-3-fold increase in cell proliferation. Anti-estrogens such as ICI-182,780 inhibited E 2 -induced cell proliferation, but it had no effect on growth factor-induced cell proliferation. An antibody against HER-2, herceptin, blocked both E 2 -and HRG-induced cell proliferation, whereas an EGFR antibody predominantly blocked only TGF␣-induced cell proliferation. Bisindolylmaleimide (a general PKC inhibitor) or rottlerin (a PKC-␦ specific inhibitor) were able to inhibit E 2 -and HRG-induced cell proliferation (Fig. 8A). However, GO6976, which inhibits all the calcium-dependent PKC isoforms, was unable to inhibit either E 2 -or growth factor-induced cell proliferation. The MEK inhibitor UO126 almost completely blocked all E 2 -and growth factor-induced cell proliferation, confirming that Erk activation is downstream of PKC and Ras signaling. Together, these results provide strong evidence that  Despite accumulating evidence showing that the ER plays a central role in cell proliferation (25,61,62), the signaling mechanisms responsible for Erk activation by E 2 are not fully characterized. The current model is that E 2 diffuses freely into the cytoplasm and enters the nucleus by passive diffusion (63). Upon activation by hormone binding, the ER interacts specifically with a cis-acting DNA sequence called the estrogen response element, which further facilitates the direct interaction of receptor dimers with promoter regions of target genes to enhance the synthesis and production of growth factors and other proteins (6,64). Consistent with this classical dogma of steroid hormone action, we have shown that when MCF-7 cells were exposed to E 2 , it leads to slow (after 2 h of E 2 exposure) but sustained activation of Erk up to 24 h. In addition to receptor-mediated (genomic) estrogenic effects, which take place starting ϳ1-6 h following hormone treatment, some effects of E 2 occur within seconds to minutes after E 2 administration to cultured cells and are thought to be mediated, at least in part, by non-genomic ER (65,66). A non-genomic estrogenic effect in our study was discounted because of the delayed Erk activation in response to E 2 -stimuation of MCF-7 cells. In addition BSA-conjugated estrogen, which prevents E 2 from entering the cell, was unable to activate estrogen response element-Luc reporter activity (data not shown). Furthermore, we have shown that E 2 -induced Erk activation can be blocked by antiestrogen (ICI 182,780) and an inhibitor of new protein synthesis, CHX.
Erk activation in response to growth factors such as EGF and platelet-derived growth factor classically proceeds via a Rasmediated cascade that involves Raf-1, MEK, and MAPK/Erk activation (16,54). In this study we have shown that blockade of HER-2-mediated signaling inhibits E 2 -induced Erk activation by 60 -70%, suggesting that HER-2 mediates most of the E 2 -induced Erk activation. Removal of HRG from the E 2treated conditioned medium (CM) significantly reduces the ability of CM to induce Erk activation. In addition, we have also shown a significant increase in HER-2 receptor phosphorylation when compared with EGFR phosphorylation in response to E 2 stimulation. Although ErbB2/HER-2 has no known ligand, it participates in type I receptor signaling by heterodimerization with ligand-bound members of the ErbB receptor family (67). Herceptin, a monoclonal antibody-based therapy targeted to HER-2, is presently approved for breast cancer patients and was shown to increase median survival time in metastatic breast cancer patients whose tumors overexpress ErbB-2 (68). The current study shows that normal or low levels of HER-2 play a major role in E 2 -mediated signaling and Erk activation. The overexpression of HER-2 or HRG in MCF-7 cells leads to a hormone-independent phenotype, however (25,69). It was previously shown that while E 2 increased TGF␣ expression in cultured breast cells, blockade of EGFR with a panel of EGFR antibodies failed to inhibit estrogen-induced growth of MCF-7, T47D, or ZR75 cells (7). One of the reasons that EGFR antibody to block E 2 induced cell proliferation may be due to the fact that E 2 -induced Erk activation and cell proliferation are predominantly mediated through HER-2/neu-dependent signaling.
There are some data to show that E 2 -induces PKC activation, and there are several studies that focus on PKC isoforms and their potential role in Erk activation (30,58,70,71). There are apparently inconsistent observations about the role of Ras in phorbol ester-induced activation of Erk. For example in PC12 cells and NIH3T3 cells, TPA-induced activation of Erk involves Ras (70,72,73). On the other hand, N17-Ras fails to inhibit TPA-induced Erk activation in Rat 1 cells, COS cells, and 293 cells (74 -76). It should be noted that these are all observations using TPA or other phorbol esters and not E 2 -stimulation or direct examination of activated PKC isoforms. In the present study, we have shown that inhibition of PKC-␦ can block E 2induced activation of Erk. To determine whether PKC activates Erk through Ras or by directly activating Raf and independently of Ras, we blocked Ras activation by either a dominant negative mutant Ras (N17-Ras) or a farnesyl transferase inhibitor. Both the Ras inhibitors blocked E 2 -and TGF␣-induced Erk activation suggesting that the Ras/Raf/MEK pathway is involved in Erk activation.
In this study we have identified PKC-␦ as a critical and specific element of E 2 -dependent activation of Erk. Inhibition of PKC-␦ significantly blocked E 2 -induced Ras and Erk activation, but not TGF␣-induced Ras and Erk activation. These data were further confirmed using stably transfected, dominant negative PKC-␦ expressing clones of MCF-7 cells (DN-PKC-␦/ MCF-7). Our analysis of Ras activation in the presence and absence of PKC inhibitors shows that E 2 -induced Ras activation is dependent on PKC-␦ in MCF-7 cells, suggesting that PKC-␦ is upstream of Ras in the E 2 -induced Erk activation signaling pathway. Interestingly, the DN-PKC-␦/MCF-7 clones also showed a significantly reduced Erk activation in response to HRG stimulation but maintain their TGF␣-induced Erk activation. These observations support our hypothesis that E 2induced Erk activation is predominantly mediated by HRG/ HER-2/PKC-␦/Ras/Raf/MEK. In contrast TGF␣ activates Erk more directly through activation of Ras/Raf/MEK signaling (Fig. 9). Ueda et al. (30) have demonstrated, using constitutively active mutants of PKC isoforms, that PKC-␦ but not PKC-␣ or PKC-⑀ activates MEK1 and Erk. These results indicate that activation of PKC-␦ is sufficient to activate the MEK-Erk pathway. However, at this time it is not clear if activated PKC-␦ will be able to support cell proliferation in the absence of E 2 or alter antiestrogenic effects in MCF-7 cells.
The interaction between growth factors and estrogen signaling is complex and occurs at multiple levels. Previous work has implicated the estrogen receptor itself as a target for growth factor signaling pathways involving Erk (3). Conversely, activation of Erk by estrogen and ER also occurs in various tissues and cell types, although the exact mechanisms remain to be determined (77,78). Even though previous studies have shown a strong association of high levels of HER-2 and PKC-␦ with hormone-independent aggressive forms of tumor cells, our results show for the first time that HER-2-and PKC-␦-mediated signaling also plays a major role in E 2 -mediated signaling in hormone-dependent, ER-positive tumor cells.