Raloxifene Inhibits Estrogen-induced Up-regulation of Telomerase Activity in a Human Breast Cancer Cell Line*

The mechanism by which raloxifene acts in the chemoprevention of breast cancer remains unclear. Because telomerase activity is involved in estrogen-induced carcinogenesis, we examined the effect of raloxifene on estrogen-induced up-regulation of telomerase activity in MCF-7 human breast cancer cell line. Raloxifene inhibited the induction of cell growth and telomerase activity by 17β-estradiol (E2). Raloxifene inhibited the E2-induced expression of the human telomerase catalytic subunit (hTERT), and transient expression assays using luciferase reporter plasmids containing various fragments of the hTERT promoter showed that the estrogen-responsive element appeared to be partially responsible for the action of raloxifene. E2 induced the phosphorylation of Akt, and pretreatment with a phosphatidylinositol 3-kinase (PI3K) inhibitor, LY294002, attenuated the E2-induced increases of the telomerase activity and hTERT promoter activity. Raloxifene inhibited the E2-induced Akt phosphorylation. In addition, raloxifene also inhibited the E2-induced hTERT expression via the PI3K/Akt/NFκB cascade. Moreover, raloxifene also inhibited the E2-induced phosphorylation of hTERT, association of NFκB with hTERT, and nuclear accumulation of hTERT. These results show that raloxifene inhibited the E2-induced up-regulation of telomerase activity not only by transcriptional regulation of hTERT via an estrogen-responsive element-dependent mechanism and the PI3K/Akt/NFκB cascade but also by post-translational regulation via phosphorylation of hTERT and association with NFκB.

Chemoprevention, defined as the prevention of cancer by the administration of chemical compounds, is a new approach for the management of cancer. Breast cancer remains a significant health problem for women. The large chemoprevention clinical trial with the selective estrogen receptor modulator tamoxifen showed a 38% reduction in breast cancer incidence (1)(2)(3)(4). However, rates of uterine endometrial cancer were increased in all tamoxifen prevention trials (1)(2)(3)(4). Ideal chemopreventive agents are nontoxic. Therefore, tamoxifen cannot yet be recommended as a preventive agent except for women at high risk for breast cancer (5). Raloxifene is a nonsteroidal benzothiophene that has also been classified as a selective estrogen receptor modulator (6) on the basis of studies in which it produced both estrogen-agonistic effects on bone (7) and lipid metabolism (8) and estrogen-antagonistic effects on uterine endometrium (9) and breast tissue (10,11). Because of its ideal tissue selectivity, raloxifene may have fewer side effects than tamoxifen. The MORE (Multiple Outcomes of Raloxifene Evaluation) trial was a randomized study designed to determine whether raloxifene would reduce the risk of fracture in postmenopausal women with osteoporosis (12). The development of breast cancer was a secondary end point of the trial. At a median 48-month followup, raloxifene treatment resulted in a 72% reduction in breast cancer incidence without association with an increased risk of uterine endometrial cancer. However, the mechanism by which raloxifene acts to prevent breast cancer remains unclear.
Telomerase is a cellular reverse transcriptase that catalyzes the synthesis and extension of telomeric DNA (13,14). This enzyme is specifically activated in most malignant tumors but is usually inactive in normal somatic cells, with the result that telomeres are progressively shortened with cell division in normal cells (15,16). Cells require a mechanism to maintain telomere stability to overcome replicative senescence, and telomerase activation may therefore be a rate-limiting or critical step in cellular immortalization and oncogenesis (17). For example, telomerase activity is known to be involved in estrogeninduced carcinogenesis (18). The level of telomerase activity in cells can be regulated by modulating both the expression and phosphorylation of the catalytic subunit (hTERT). 1 The hTERT promoter contains an imperfect palindromic estrogen-responsive element (ERE), and it was reported that estrogen activates telomerase via direct and indirect effects on hTERT in MCF-7 cells (18). However, the mechanism of the indirect effects remains unclear. It was reported that the hTERT promoter contains two putative NFB-binding motifs (19) and that IGF-1 and IL-6 activate the PI3K/Akt/NFB cascade in a human multiple myeloma cell line (20). Thus, it is possible that estrogen enhances the transcription of hTERT via the PI3K/Akt/NFB cascade.
It was reported that the region surrounding Ser-824 in hTERT conforms to a consensus sequence for phosphorylation by Akt and that Akt kinase enhances human telomerase activity through phosphorylation of hTERT (21). In addition, it was reported that an Akt cascade mediates the estrogen-induced S phase entry and cyclin D1 promoter activity in MCF-7 cells (22). Thus, it is possible that estrogen enhances human telomerase activity through an Akt cascade. Another possible mechanism for post-translational modulation of telomerase activity is via the interaction of hTERT with accessory proteins. Recently, it was reported that 14-3-3 proteins (23,24) and NFB (25) are post-translational modifiers of telomerase that function by controlling the intracellular localization of hTERT.
These findings led us to examine whether estrogen induces up-regulation of telomerase activity not only by transcriptional regulation of hTERT via an ERE-dependent mechanism and a PI3K/Akt/NFB cascade but also by post-translational regulation via Akt-dependent phosphorylation of hTERT in MCF-7 cells. In addition, we attempted to clarify the mechanism by which raloxifene inhibits the induction of telomerase activity by estrogen.

EXPERIMENTAL PROCEDURES
Materials-Raloxifene analog LY117018 was a kind gift from Eli Lilly Research Laboratories (Indianapolis, IN). 17␤-Estradiol, TPA, and rabbit IgG were purchased from Sigma. ICI 182780 was obtained from TOCRIS (Ballwin, MO). LY294002 was purchased from Calbiochem. The anti-phospho-Akt, phospho-Akt substrate, and Akt antibodies were obtained from Cell Signaling (Beverly, MA). The anti-HA, IB, phospho-IB, NFB p65, and hTERT antibodies were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). The IB␣ phosphorylation inhibitor BAY-11-7082 was purchased from Alexis Biochemicals (San Diego, CA). The specific NFB nuclear translocation inhibitor SN-50 was purchased from BIOMOL (Plymouth Meeting, PA). Hoechst 33258 was obtained from Molecular Probes (Eugene, OR).
Constructs-pCR3 vector and pCR3-hTERT were kind gifts from Dr.
Takashi Tsuruo (Institute of Molecular and Cellular Biosciences, University of Tokyo, Tokyo, Japan) (24). The pCR-FLAG-p50 and pCR-FLAG-p50⌬NLS constructs were kind gifts from Dr. Gourisankar Ghosh (Department of Chemistry and Biochemistry, University of California, San Diego, CA) (26). Cell Culture-MCF-7 human breast cancer cells were obtained from the American Type Culture Collection. The cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 units/ml penicillin G sodium, and 100 g/ml streptomycin sulfate in the presence of 5% CO 2 at 37°C.
Cell Proliferation Assay-The cells were plated at a density of 10 ϫ 10 3 cells/well in 12-well plates and allowed to attach overnight. The cells were growth-arrested by phenol red-free Dulbecco's modified Eagle's medium with 10% charcoal-stripped serum (CSS) for 48 h and were then treated with vehicle, E2, raloxifene, or E2 ϩ raloxifene by exchanging the culture medium containing these agent(s) with fresh medium every 48 h for 8 days. A Neubauer chamber was used to count the cell number, and a trypan blue exclusion test was carried out to determine the cell viability. All of the experiments were carried out in quadruplicate. The values shown are the means Ϯ S.E. of three independent experiments performed in quadruplicate at three different passages of the cell lines.
Stretch PCR Assay-For quantitative analysis of telomerase activity, stretch PCR assays were performed using the Telochaser system according to the manufacturer's protocol (Toyobo, Tokyo, Japan) as described previously (18). The PCR products were electrophoresed on a 7% polyacrylamide gel and visualized with SYBR green I nucleic acid gel stain (FMC BioProducts, Rockland, ME). To monitor the efficacy of PCR amplification, 10 ng of a internal control consisting of phage DNA (Toyobo) together with 50 pmol of specific primers (Toyobo) were added to the PCR mixture per reaction. Band intensity was measured using NIH Image software.
RT-PCR Analysis-Total cellular RNA was isolated using Tri-Reagent (Molecular Research Center, Inc.). The expression of hTERT mRNA and glyceraldehyde-3-phosphate dehydrogenase mRNA was analyzed by semiquantitative RT-PCR amplification as described previously (18). Briefly, hTERT mRNAs were amplified using the primer pair 5Ј-CGGAAGAGTGTCTGGAGCAA-3Ј and 5Ј-GGATGAAGCGGAGTCT-GGA-3Ј. cDNA was synthesized from 1 g of RNA using an RNA PCR kit version 2 (TaKaRa, Ohtsu, Japan) with random primers. Serially diluted cDNA reverse-transcribed from 1 g of RNA was first amplified by RT-PCR to generate standard curves. The correlation between the band intensity and dose of cDNA template was linear under the conditions described below. Typically, 2-l aliquots of the reverse-transcribed cDNA were amplified by 28 cycles of PCR in 50 l of 1ϫ buffer (10 mM Tris-HCl, pH 8.3, 2.5 mM MgCl 2 , and 50 mM KCl) containing 1 mM each dATP, dCTP, dGTP, and dTTP, 2.5 units of Taq DNA polymerase (TaKaRa), and each specific primer at 0.2 M. Each cycle consisted of denaturation at 94°C for 30 s, annealing at 60°C for 30 s, and extension at 72°C for 45s. PCR products were resolved by electrophoresis in a 1% agarose gel. The efficiency of cDNA synthesis from each sample was estimated by PCR with glyceraldehyde-3-phosphate dehydrogenase-specific primers as described previously (18).
Luciferase Assay-Plasmids pGL3-3328 and pGL3-2000 are hTERT promoter-luciferase reporters in which full-length or 5Ј-deleted promoters including a 77-bp 5Ј-untranslated region are cloned upstream of the luciferase gene in pGL3-Basic at MluI and BglII sites (18). pGL3-EREpromoter was constructed by inserting head-in-tail tetramers of the ERE located at Ϫ2677 in the hTERT promoter, into the enhancer-less vector pGL3-promoter (18). pGL2-HPV31URR-luc vector is an HPV-31 enhancer and promoter-luciferase reporter containing the whole upstream regulatory region (URR) of HPV-31 cloned upstream of the luciferase gene in pGL2-Basic (27). These reporter plasmids were tran-siently transfected into cells for 24 h using LipofectAMINE Plus (Invitrogen) according to the manufacturer's protocol. The cells were harvested and subjected to luciferase assays using a luciferase assay system (Promega) as described previously (18). A plasmid expressing the bacterial ␤-galactosidase gene was also cotransfected in each experiment to serve as an internal control for transfection efficiency.
Western Blot Analysis-The cells were incubated in phenol red-free medium without serum for 16 h and then treated with various agents. They were then washed twice with phosphate-buffered saline and lysed in ice-cold HNTG buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1.5 mM MgCl 2 , 1 mM EDTA, 10 mM sodium pyrophosphate, 100 M sodium orthovanadate, 100 mM NaF, 10 g/ml aprotinin, 10 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride) (28). The lysates were centrifuged at 12,000 ϫ g at 4°C for 15 min, and the protein concentrations of the supernatants were determined using the Bio-Rad protein assay reagent. Equal amounts of proteins were separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. Blocking was done in 10% bovine serum albumin in 1ϫ Tris-buffered saline. Western blot analyses were performed with various specific primary antibodies. For detection of phosphorylated hTERT or association of hTERT with NFB p65, cell lysates were prepared using HNTG buffer. The lysates were incubated with Twenty-four hours later, the cell pellets were collected, and the luciferase assays were performed. The transcriptional activity was normalized relative to ␤-galactosidase activity, and the activity in cells treated with vehicle was set at 1.0. The data are expressed as the mean fold activation Ϯ S.E. of six transfections. D, MCF-7 cells were transfected with 5Ј-deleted hTERT promoter (Ϫ2000) and treated with phenol red-free medium without serum (Serum free, lane 1), or with 10% CSS (lane 2), or 10% CSS ϩ 10 nM raloxifene (lane 3). Twelve hours later, the cell pellets were collected, and the luciferase assays were performed. The transcriptional activity was normalized relative to ␤-galactosidase activity, and the activity in cells treated without serum was set at 1.0. The data are expressed as the mean fold activation Ϯ S.E. of six transfections. E, MCF-7 cells were transfected with HPV31URR-luc and treated with 10 nM raloxifene (lane 2), 100 nM TPA (lane 3), or 100 nM TPA ϩ 10 nM raloxifene (lane 4). Twenty-four hours later, the cell pellets were collected, and the luciferase assays were performed. The transcriptional activity was normalized relative to ␤-galactosidase activity, and the activity in cells treated with vehicle was set at 1.0. The data are expressed as the mean fold activation Ϯ S.E. of six transfections.
anti-hTERT antibody overnight and then immunoprecipitated for 2 h with protein A-Sepharose. Immune complexes were washed with icecold HNTG buffer, electrophoresed, and analyzed by immunoblotting with anti-phospho-Akt substrate antibody or anti-NFB p65 antibody. Immunoreacted bands in the immunoblots were visualized with horseradish peroxidase-coupled goat anti-rabbit or anti-mouse immunoglobulin by using the enhanced chemiluminescence Western blotting system.
Fluorescence Microscopy-MCF-7 cells were grown on glass coverslips in six-well dishes. The cells were transfected with the pCR3-hTERT plasmid for 24 h and then incubated with various reagents. The cells were fixed with 10% formalin for 10 min, permeabilized with 0.5% Triton X-100 for 5 min, and blocked with 3% bovine serum albumin for 1 h. Anti-HA antibody and Alexa Fluor secondary antibody were used at 2 g/l in blocking solution. The cells were counterstained with 10 mmol/liter Hoechst 33258 to visualize the nucleus. The samples were mounted on glass slides with Vectashield (Vector Laboratories), and the cells were examined using fluorescence microscopy. For quantification experiments, 100 cells were scored according to whether hTERT was higher in the nucleus (N), evenly distributed between nucleus and cytoplasm (NϩC), or higher in the cytoplasm (C). The data represent the mean of three independent experiments.
Statistics-Statistical analysis was performed by Student's t test, and p Ͻ 0.01 was considered significant. The data are expressed as the means Ϯ S.E.

RESULTS
Raloxifene Attenuates the E2-induced Up-regulation of Telomerase Activity and hTERT Expression-We first examined whether or not raloxifene regulates the proliferation of MCF-7 human breast cancer cells (Fig. 1A). E2 significantly induced cell growth at 10 nM. Although 10 nM raloxifene had no effect on the basal cell growth, it did significantly inhibit the E2-induced cell growth. To examine the effects of raloxifene on the estrogen-induced telomerase activity, MCF-7 human breast cancer cells were treated with 10 nM E2, 10 nM raloxifene, 10 nM E2 ϩ 10 nM raloxifene, or 10 nM E2 ϩ 1 M ICI 182,780 (a highly selective ER antagonist) for 24 h (Fig. 1B). ICI 182,780 (positive control) attenuated the E2-induced telomerase activity (Fig. 1B, lane 5). Although raloxifene had no effect on basal telomerase activity (Fig. 1B, lane 3), it attenuated the E2induced increase in telomerase activity (Fig. 1B, lane 4). Semiquantitative RT-PCR assays were performed to examine whether the attenuation of estrogen-induced telomerase activity by raloxifene was due to the attenuation by raloxifene of the estrogen-induced up-regulation of the expression of hTERT ( Fig. 2A). As we previously reported (29), treatment of MCF-7 cells with 1 M tamoxifen for 24 h attenuated the up-regulation of hTERT mRNA induced by 10 nM E2 (Fig. 2A, lane 5). Treatment of MCF-7 cells with 10 nM raloxifene for 24 h also attenuated the up-regulation of hTERT mRNA induced by 10 nM E2 (Fig. 2A, lane 6).
To examine the effect of raloxifene on the estrogen-induced transcriptional activity of the hTERT promoter, luciferase assays of cells into which hTERT-promoter reporter plasmids were transfected were performed (Fig. 2B). We have previously reported that an imperfect palindromic ERE is located at Ϫ2677 in the hTERT promoter and is capable of direct association with ER (18). Luciferase reporter plasmids containing the full-length hTERT 5Ј regulatory region (pGL3-3328) or a deletion mutant lacking the imperfect palindromic ERE (pGL3-2000) were transfected into MCF-7 cells. ICI 182,780 attenuated the E2-induced transcriptional activation of pGL3-3328 (Fig. 2B, lane 5). Raloxifene also attenuated the E2-induced transcriptional activation of pGL3-3328 (Fig. 2B, lane 4). To examine whether the ERE at Ϫ2677 is involved in the E2induced up-regulation of the hTERT promoter activity and the inhibitory effect of raloxifene, this putative ERE was cloned upstream of the SV40 promoter in a luciferase reporter plasmid (pGL3-ERE-promoter) and used for transfection. Although E2 had no effect on transcriptional activation of enhancer-less pGL3-promoter containing only SV40 promoter (data not shown), E2 treatment caused transcriptional activation of pGL3-ERE-promoter (Fig. 2C, lane 2), as we reported previously (28). Treatment with raloxifene (Fig. 2C, lane 4) or ICI 182,780 (Fig. 2C, lane 5) attenuated the E2-induced transcriptional activation of the pGL3-ERE-promoter. E2 did activate the transcriptional activity of pGL3-2000 to some extent (Fig.  2B, lane 2), and raloxifene (Fig. 2B, lane 4) or ICI (Fig. 2B, lane 5) attenuated the E2-induced transcriptional activation of pGL3-2000. These results suggest that the ERE at Ϫ2677 is partially responsible for the E2-induced activation of the hTERT promoter and the inhibitory effect of raloxifene.
We examined whether raloxifene has a general inhibitory effect on transcription. Raloxifene did not have an inhibitory effect on 10% CSS-induced hTERT promoter activity using a deletion mutant of the imperfect palindromic ERE (pGL3-2000) (Fig. 2D). In addition, we used HPV31URR-luc, which is a well characterized reporter for examining the AP-1 activity (27). Raloxifene did not have an inhibitory effect on TPAinduced HPV31 promoter activity (Fig. 2E). These results suggest that raloxifene does not have an inhibitory effect on the induction of AP-1 or SRE activity in promoters that lack an ERE.
Raloxifene Inhibits E2-induced Akt Phosphorylation-Because it was previously reported that hTERT expression is induced via an Akt cascade (21), we first examined whether E2 induces Akt phosphorylation in MCF-7 cells. The cells were treated with E2 for various times and then used to prepare lysates that were subjected to Western blotting with anti-phospho-Akt or -Akt antibody. Although E2 did not affect the expression of Akt (Fig. 3A, bottom panel), it induced phosphorylation of Akt (Fig. 3A, middle and top panels). We then examined whether E2 induces telomerase activity and hTERT expression via an Akt cascade. The effects of treatment with PI3K inhibitor LY294002 on E2-induced up-regulation of telomerase activity and hTERT promoter activity were examined. We first confirmed that LY294002 completely blocked the E2induced Akt phosphorylation (Fig. 3B, lane 6). LY294002 clearly attenuated the E2-induced up-regulation of telomerase activity (Fig. 3C, lane 3). LY294002 also decreased the E2induced up-regulation of promoter activity of an hTERT promoter with a deletion of the imperfect palindromic ERE (pGL3-2000) (Fig. 2B, lane 6). These results suggest that an Akt cascade is involved in the E2-induced up-regulation of telomerase activity and hTERT expression.
Next, we examined whether raloxifene attenuates the E2induced Akt phosphorylation. Although raloxifene at 10 nM did not affect the expression of Akt (Fig. 3B, bottom panel, lane 3), it inhibited the E2-induced Akt phosphorylation (Fig. 3B, middle and top panels, lane 4).
Raloxifene Inhibits E2-induced hTERT Expression via PI3K/ Akt/NFB Cascade-It has been reported that the hTERT promoter contains two putative NFB-binding motifs (19) and that IGF-1 and IL-6 activate PI3K/Akt/NFB in a human multiple myeloma cell line (20). Therefore, we examined whether NFB is a nuclear target in the E2-induced up-regulation of hTERT expression via the Akt cascade. NFB is regulated through its association with an inhibitory cofactor, IB, which sequesters NFB in the cytoplasm. Phosphorylation of IB by upstream kinases promotes its degradation, allowing NFB to translocate to the nucleus and induce target genes (30,31). We first examined whether E2 induces the phosphorylation and degradation of IB (Fig. 4A). The cells were treated with E2 for 30 min and used to prepare lysates that were analyzed by Western blotting with anti-phospho-IB, anti-IB, or anti-Akt antibody. Although the each expression of Akt was not changed (Fig. 4A, panel iv), E2-simulated MCF-7 cells showed an increase in phosphorylated IB (Fig. 4A, panel ii, lane 2) and subsequent degradation of IB (Fig. 4A, panel iii, lane 2), and treatment with LY294002 attenuated the E2-induced phosphorylation (Fig. 4A, panel ii, lane 4) and degradation of IB (Fig. 4A, panel  iii, lane 4), suggesting that E2 induces the activation of NFB through phosphorylation and degradation of IB in a PI3K-dependent manner. Moreover, we examined the effect of raloxifene on the E2-induced phosphorylation and degradation of IB. Raloxifene inhibited the E2-induced IB phosphorylation (Fig. 4A, panel ii, lane 3) and degradation (Fig. 4A, panel iii,  lane 3).
We next examined whether NFB is involved in the induction of hTERT promoter activity by E2. Five homologous polypeptides, p50, p65, c-Rel, RelB, and p52, comprise the mammalian Rel/NFB transcription factor family. The subunits associate in a combinatorial fashion to form transcriptionally active homo-and hetero-dimers. The best characterized species of NFB dimer is the p50/p65 hetero-dimer (32). A previous report demonstrated that the nuclear localization signal (NLS) polypeptide of p50 is required for its translocation to the nucleus (33) and that p50⌬NLS lacking the NLS domain inhibits the nucleocytoplasmic shuttling of NFB dimers. Therefore, we examined the effect of p50⌬NLS on the induction of hTERT promoter activity by E2. Cotransfection of p50⌬NLS resulted in a significantly weaker transactivation by E2 of pGL3-2000, a deletion mutant of the imperfect palindromic ERE, compared with the induction in cells expressing wild-type p50 (Fig. 4B). The result suggests that NFB plays a pivotal role in E2-induced transcriptional activation of pGL3-2000.
Raloxifene Inhibits E2-induced Phosphorylation of hTERT at a Putative Akt Phosphorylation Site-Telomerase activity may also be regulated by post-translational modifications of the enzyme. It has been reported that the region surrounding Ser-824 in hTERT conforms to a consensus sequence for phospho- rylation by Akt and that Akt kinase enhances human telomerase activity through phosphorylation of hTERT (21). Therefore, we examined whether E2 induces the phosphorylation of hTERT at a putative Akt phosphorylation site. The cells were treated with E2 for the indicated times and then used to prepare lysates that were immunoprecipitated with anti-hTERT antibody and then subjected to Western blotting with antiphospho-Akt substrate antibody (Fig. 5A, middle panel) or anti-hTERT antibody (Fig. 5A, bottom panel). Although E2 did not affect the expression of hTERT (Fig. 5A, bottom panel), an increase in hTERT phosphorylation at a putative Akt phosphorylation site was induced by E2, reached a peak at 30 min, and declined thereafter (Fig. 5A, middle and top panels, lanes 2 and  3). Although raloxifene did not affect the expression of hTERT (Fig. 5B, bottom panel, lane 3), it inhibited the E2-induced hTERT phosphorylation (Fig. 5B, middle and top panels,  lane 4).
Raloxifene Inhibits E2-induced Association of NFB with hTERT and Nuclear Accumulation of hTERT-One possible mechanism for the post-translational modulation of telomerase activity is via the interaction of hTERT with accessory proteins. Recently, it was reported that NFB is a post-translational modifier of telomerase that functions by controlling the intracellular localization of hTERT (25). Therefore, we examined whether E2 induces the association of NFB p65 with hTERT. The cells were treated with E2 for the indicated times, used to prepare cell lysates that were immunoprecipitated with anti-hTERT antibody or anti-HA antibody (nonrelevant control antibody), and then subjected to Western blotting with anti-NFB p65 (Fig. 6A, middle panel) or anti-hTERT antibody (Fig.  6A, bottom panel). Neither hTERT nor NFB p65 expression was detected in E2 treatment in immunoprecipitates with anti-HA antibody (nonrelevant control antibody) (Fig. 6B, lane 5). E2 did not affect the expression of hTERT (Fig. 6A, bottom  panel), but the association of hTERT with NFB p65 was transiently up-regulated by E2 (Fig. 6A, middle and top panels). Although raloxifene did not affect the expression of hTERT (Fig. 6B, bottom panel, lane 3), it inhibited the E2-induced association of NFB p65 with hTERT (Fig. 6B, middle and top  panels, lane 4).

DISCUSSION
One of the two novel findings in this study was that the induction of telomerase activity by estrogen is due not only to transcriptional regulation of hTERT via an ERE-dependent mechanism and a PI3K/Akt/NFB cascade but also to posttranslational regulation via phosphorylation of hTERT and association with NFB in MCF-7 cells, which is also true of the mechanism by which cytokines modulate telomerase activity (20,25). The other novel finding we made was that raloxifene inhibited the E2-induced up-regulation of telomerase activity via not only transcriptional regulation but also post-translational regulation of hTERT. It thus seems that the inhibition of phosphorylation of Akt by raloxifene is one of the key events in the inhibition of estrogen-induced up-regulation of telomerase activity. This finding about Akt phosphorylation is consistent with a large body of evidence reported in the following publications. Estrogen-induced Akt activation is blocked by the antiestrogens ICI 182,780 and 4-hydroxytamoxifen (35). It has also been reported that constitutive and inducible Akt activity promotes resistance to chemotherapy, trastuzumab, and tamoxifen in breast cancer cells (36), that activation of Akt in breast cancer predicts a worse outcome among endocrinetreated patients (37), and that Akt protects breast cancer cells from tamoxifen-induced apoptosis (38). Thus, Akt may have an important role in the chemoprevention of breast cancer.
How does raloxifene or tamoxifen inhibit the estrogen-induced Akt phosphorylation? Because ICI 182,780 inhibited estrogen-induced Akt phosphorylation (Fig. 3B), nuclear accumulation of hTERT (Fig. 7), and telomerase activation (Fig. 1B), ER may be involved in non-nuclear estrogen-dependent signal- Raloxifene inhibits the estrogen-induced up-regulation of telomerase activity via both binding to the nuclear ER, resulting in the inhibition of hTERT expression, and binding to the non-nuclear ER, resulting in the inhibition of an Aktdependent cascade.
ing in breast cancer cells, because we have reported that estrogen induces the activation of endothelial nitric-oxide synthase via a PI3K/Akt cascade in a nongenomic manner in vascular endothelial cells (39). Moreover, we have reported that raloxifene induces the activation of endothelial nitric-oxide synthase via binding to non-nuclear ER␣ in vascular endothelial cells (40). Thus, raloxifene might also bind to the non-nuclear ER in breast cancer cells and competitively block the action of estrogen. Because raloxifene also inhibited the E2-induced up-regulation of telomerase activity by transcriptional regulation of hTERT via an ERE-dependent mechanism, raloxifene might bind to both the nuclear and non-nuclear ERs, leading to inhibition of the E2-induced up-regulation of telomerase activity.
It was reported that 4-hydroxytamoxifen increases proteintyrosine phosphatase activity in MCF-7 cells (41). Phosphatase and tensin homolog deleted from chromosome 10 (PTEN) functions as a tumor suppressor by negatively regulating the growth/survival signaling of the PI3K/Akt pathway (42,43). Thus, it is possible that phosphatase and tensin homolog deleted from chromosome 10 has a role in the inhibitory effect of raloxifene or tamoxifen in estrogen-induced Akt phosphorylation. We are currently investigating this possibility.
In the present study, we have shown that estrogen induced nuclear accumulation of hTERT by enhancing the association of NFB with hTERT (Fig. 7). The Hsp90 chaperone complex, which includes Hsp90, Hsp70, and p23, is functionally associated with telomerase (44). Expression of this chaperone complex is up-regulated during malignant transformation or in cancer tissues compared with surrounding noncancerous tissues (45), suggesting that up-regulation of the chaperone complex may play a role in the telomerase activation observed in cancer cells. In addition, it was recently reported that 14-3-3 signaling proteins, which work as molecular chaperones and regulate the intracellular localization of their binding partners, are hTERT-binding partners and that 14-3-3 enhances the nuclear localization of hTERT (24). Therefore, it is possible that estrogen induces the association of hTERT with hTERT-binding partners other than NFB, including Hsp and 14-3-3.
hTERT is one of the target genes of estrogen-induced mammary carcinogenesis (18). In this study, we have shown that raloxifene inhibited the estrogen-induced up-regulation of telomerase activity, as has been shown previously for tamoxifen (29). Further investigations will be necessary to examine whether raloxifene also inhibits the up-regulation of other target genes of estrogen-induced mammary carcinogenesis, such as cyclin D1, c-Myc, or vascular endothelial growth factor (46).
According to the large breast cancer chemoprevention clinical trial, raloxifene may have a better profile than tamoxifen, leading to greater efficacy and fewer endometrial cancers (1)(2)(3)(4)(5)12). In the present study, we also showed that the inhibitory effect of 10 nM raloxifene on estrogen-induced hTERT expression was similar to that of 1 M tamoxifen ( Fig. 2A), suggesting greater efficacy of raloxifene. However, there have been no clinical trials directly comparing tamoxifen and raloxifene. The results of the STAR (Study of Tamoxifen and Raloxifene) trials (47), in which postmenopausal women at increased risk for breast cancer are being randomly assigned to receive either tamoxifen or raloxifene for a 5-year period, are awaited with great interest.
Ovarian cancer, like breast cancer, remains a significant health problem for women. It was recently reported that estrogen replacement therapy induces ovarian cancer (48). There is some evidence from observational studies that tamoxifen may produce an antitumor response in a modest proportion of women with relapsed ovarian cancer (49,50). Moreover, it was reported that tamoxifen is a chemoprevention option in BRCA1 and BRCA2 mutation carriers (51). Therefore, it is possible that raloxifene, which shows milder toxicity than tamoxifen, might be effective for chemoprevention of ovarian cancer.
In summary, this study is the first to show that raloxifene inhibits the E2-induced up-regulation of telomerase activity not only by transcriptional regulation of hTERT expression via an ERE-dependent mechanism and a PI3K/Akt/NFB cascade but also by post-translational regulation via phosphorylation of hTERT and association with NFB in MCF-7 cells (Fig. 8).