Tamoxifen but Not 4-Hydroxytamoxifen Initiates Apoptosis in p53(−) Normal Human Mammary Epithelial Cells by Inducing Mitochondrial Depolarization*

Despite the widespread clinical use of tamoxifen as a breast cancer prevention agent, the molecular mechanism of tamoxifen chemoprevention is poorly understood. Abnormal expression of p53 is felt to be an early event in mammary carcinogenesis. We developed an in vitro model of early breast cancer prevention to investigate how tamoxifen and 4-hydroxytamoxifen may act in normal human mammary epithelial cells (HMECs) that have acutely lost p53 function. p53 function was suppressed by retrovirally mediated expression of the human papillomavirus type 16 E6 protein. Tamoxifen, but not 4-hydroxytamoxifen, rapidly induced apoptosis in p53(−) HMEC-E6 cells as evidenced by characteristic morphologic changes, annexin V binding, and DNA fragmentation. We observed that a decrease in mitochondrial membrane potential, mitochondrial condensation, and caspase activation preceded the morphologic appearance of apoptosis in tamoxifen-treated early passage p53(−) HMEC-E6 cells. p53(−) HMEC-E6 cells rapidly developed resistance to tamoxifen-mediated apoptosis within 10 passages in vitro. Resistance to tamoxifen in late passage p53(−) HMEC-E6 cells correlated with an increase in mitochondrial mass and a lack of mitochondrial depolarization and caspase activation following tamoxifen treatment. We hypothesize that an early event in the induction of apoptosis by tamoxifen involves mitochondrial depolarization and caspase activation, and this may be important for effective chemoprevention.

Despite the widespread clinical use of tamoxifen as a breast cancer prevention agent, the molecular mechanism of tamoxifen chemoprevention is poorly understood. Abnormal expression of p53 is felt to be an early event in mammary carcinogenesis. We developed an in vitro model of early breast cancer prevention to investigate how tamoxifen and 4-hydroxytamoxifen may act in normal human mammary epithelial cells (HMECs) that have acutely lost p53 function. p53 function was suppressed by retrovirally mediated expression of the human papillomavirus type 16 E6 protein. Tamoxifen, but not 4-hydroxytamoxifen, rapidly induced apoptosis in p53(؊) HMEC-E6 cells as evidenced by characteristic morphologic changes, annexin V binding, and DNA fragmentation. We observed that a decrease in mitochondrial membrane potential, mitochondrial condensation, and caspase activation preceded the morphologic appearance of apoptosis in tamoxifen-treated early passage p53(؊) HMEC-E6 cells. p53(؊) HMEC-E6 cells rapidly developed resistance to tamoxifen-mediated apoptosis within 10 passages in vitro. Resistance to tamoxifen in late passage p53(؊) HMEC-E6 cells correlated with an increase in mitochondrial mass and a lack of mitochondrial depolarization and caspase activation following tamoxifen treatment. We hypothesize that an early event in the induction of apoptosis by tamoxifen involves mitochondrial depolarization and caspase activation, and this may be important for effective chemoprevention.
Apoptosis, or programmed cell death, is critical for embryogenesis and for normal tissue homeostasis (1). Deregulated apoptotic signaling is felt to contribute to human cancer and autoimmune disorders (2,3). Chemotherapeutic agents are also felt to exert many of their cytotoxic effects by induction of apoptosis, and chemotherapy resistance frequently correlates with resistance to apoptotic signaling (4).
Apoptosis is morphologically characterized by specific structural changes including margination of chromatin, nuclear condensation, cell shrinkage, and formation of apoptotic bodies (5). There is much evidence that apoptotic signaling activates highly regulated and specific proteolysis mediated by caspases (6). Caspases are a highly conserved family of aspartic acidspecific proteases that are synthesized as zymogens and are converted to active heterodimers by proteolytic cleavage (7,8). Activated caspases are thought to be responsible, in part, for cellular changes that occur during the execution phase of apoptosis such as DNA fragmentation, chromatin condensation, and formation of apoptotic bodies (9). A large body of evidence supports a cascade model for effector caspase activation; a proapoptotic signal culminates in release of mitochondrial cytochrome c, resulting in activation of initiator caspases that, in turn, activate effector caspases, resulting in cellular disassembly (9).
Recent evidence suggests that mitochondria play a central role in apoptosis as integrators of cellular apoptotic signal transduction and in amplification of the apoptotic response (10). Disruption of mitochondrial electron transport and energy metabolism is recognized as an early event in apoptosis and precedes the appearance of morphologic changes characteristic of apoptosis (10). Mitochondrial dysfunction is characterized by an increase in mitochondrial membrane permeability and loss of membrane potential (⌬⌿ m ). Associated with this decrease in ⌬⌿ m , cytochrome c is translocated from the intermembrane compartment of the mitochondria to the cytosol. Cytosolic cytochrome c forms an essential part of the "apoptosome," composed of cytochrome c, Apaf-1, and procaspase-9 (11). This results in activation of procaspase-9 and that, in turn, activates other caspases, such as caspase-3, to orchestrate the execution phase of apoptosis (10).
The estrogen agonist/antagonist tamoxifen is a triphenylethylene that has been shown to act as both a chemotherapeutic agent for the treatment of breast cancer and, more recently, as a breast cancer chemoprevention agent. The Breast Cancer Prevention Trial demonstrated a 45% reduction in breast cancer incidence among the participants who took tamoxifen, 6 years after its inception (12). This was the first study to demonstrate that a chemopreventive agent could reduce the incidence of breast cancer. However, many questions surround the results from the Breast Cancer Prevention Trial. 1) Did "true" chemoprevention occur or were the benefits of tamoxifen due to ablation of preclinical breast cancer? 2) Tamoxifen has been shown to induce both growth arrest and apoptosis (13)(14)(15). Did tamoxifen act as a cytostatic or cytotoxic agent in the Breast Cancer Prevention Trial? p53 is a critical regulator of cell cycle control, and the high frequency with which p53 is functionally inactivated in early human breast cancer attests to its key role in preventing mammary carcinogenesis (16 -18). Approximately 50% of all primary node-negative breast cancers have deleted or mutated p53, and individuals with germ line heterozygous mutations in p53, or Li-Fraumeni's syndrome, demonstrate an increased risk of breast cancer (19 -21). Furthermore, aberrant expression of p53 in mammary epithelial cells is a predictor of risk for the subsequent development of breast cancer. 1) The accumulation of p53 protein (but not c-ErbB-2) in benign breast lesions is a significant predictor for the subsequent development of breast cancer (22,23). 2) p53 protein is frequently overexpressed (36%) in benign mammary epithelial cells obtained from high risk women but not observed (0%) in low risk women (22). 3) Aberrant expression of p53 in the setting of mammary hyperplasia is a significant predictor of the subsequent development of breast cancer in high risk women (23). These observations suggest that loss of p53 function may be an early event in breast carcinogenesis.
Although abnormal expression of p53 predicts a poor response to tamoxifen chemotherapy, little is known about the fate of normal human mammary epithelial cells (HMECs) 1 that acutely lose p53 function during tamoxifen chemoprevention. We sought to model tamoxifen chemoprevention in normal HMECs that have acutely lost p53 function. p53 function was suppressed utilizing retrovirally mediated introduction of human papilloma type 16 (HPV-16) E6 protein (24). The E6 protein of the cancer-associated HPV-16 binds to p53 and targets it for degradation through the ubiquitin pathway (25,26) and provides a model for the isolated loss of p53 function.
Tamoxifen is extensively metabolized in vivo, producing demethylated and hydroxylated derivatives that can be detected in patients receiving tamoxifen treatment (27)(28)(29). The major metabolites of tamoxifen are felt to be 4-hydroxytamoxifen and N-demethyltamoxifen (27). Whereas certain metabolites are biologically inactive, 4-hydroxytamoxifen has a higher affinity for the estrogen receptor than tamoxifen in vitro (18,30). The ability of tamoxifen to inhibit proliferation has been extensively studied. However, the molecular mechanism by which tamoxifen initiates apoptosis is poorly understood. We sought to investigate whether tamoxifen and its related metabolite, 4-hydroxytamoxifen, may induce growth arrest and/or apoptosis in HMECs that have acutely lost p53 function as a model of anti-estrogen chemoprevention. Surprisingly, we observed that tamoxifen but not 4-hydroxytamoxifen induced apoptosis in p53(Ϫ) HMECs. Tamoxifen-induced apoptosis was associated with a fall in mitochondrial potential, mitochondrial condensation, and caspase activation suggesting a critical role for mitochondrial targeting in mediating sensitivity to tamoxifen-induced apoptosis.

EXPERIMENTAL PROCEDURES
Materials-A 1.0 mM stock solution of tamoxifen and 4-hydroxytamoxifen (Sigma) was prepared in 100% ethanol and stored in opaque tubes at Ϫ70°C. Control cultures received equivalent volumes of the ethanol solvent. Stocks were used under reduced light. Cell culture plasticware was from Corning Glass (Corning, NY).
Cell Culture and Media-Normal human mammary epithelial cell (HMEC) strain AG11132 (M. Stampfer, 172R/AA7) was purchased from the National Institute of Aging, Cell Culture Repository (Coriell Institute) (32). HMEC strain AG11132 was established from normal tissue obtained at reduction mammoplasty, has a limited life span in culture, and fails to divide after ϳ20 -25 passages. AG11132 cells exhibit a low level of estrogen receptor staining, characteristic of normal mammary cells. AG11132 was at passage 8 at the time of receipt. Cells were grown in mammary epithelial cell basal medium (Clonetics, San Diego, CA) supplemented with 4 l/ml bovine pituitary extract (Clonetics), 5 g/ml insulin (Upstate Biotechnology Inc., Lake Placid, NY), 10 ng/ml epidermal growth factor (Upstate Biotechnology Inc.), 0.5 g/ml hydrocortisone (Sigma), 10 Ϫ5 M isoproterenol (Sigma), 10 mM HEPES buffer (Sigma) (Standard Medium). G418 (Life Technologies, Inc.) containing medium was prepared by the addition of 300 g/ml of G418 to Standard Medium. Cells were cultured at 37°C in a humidified incubator with 5% CO 2 , 95% air. Mycoplasma testing was performed as reported previously (33).
Cell Synchronization-Approximately 2 ϫ 10 6 p53(ϩ) HMEC-LXSN or p53(Ϫ) HMEC-E6 cells were plated in a T-75 flask on Day Ϫ5 in Standard Medium and grown for 4 days (Day Ϫ1). We previously observed that on Day Ϫ1 greater than 85% of the cells that had become growth factor-depleted are in G 1/0 phase, trypsinize without difficulty, and rapidly resume proliferation in the presence of fresh Standard Medium. 2 Cells were synchronized by this method prior to each experiment.
Retroviral Transduction-The LXSN16E6 retroviral vector containing the HPV-16 E6 coding sequence (provided by D. Galloway) has been described previously (24). AG11132 normal human mammary epithelial cells (passage 9) were plated in four T-75 tissue culture flasks in Standard Medium and grown to 50% confluency. Transducing virions from either the PA317-LXSN16E6 or the control PA317-LXSN (without insert) retroviral producer line were added at a multiplicity of infection at 1:1 in the presence of 4 g/ml Polybrene (Sigma) to log phase cells grown in T-75 flasks. The two remaining T-75 flasks were not infected with virus. After 48 h the 2 flasks containing transduced cells and 1 flask with untransduced cells were selected with Standard Medium containing 300 g/ml G418. Cells were continued in G418 containing medium for 1 week, until 100% of control untransduced cells were dead. The 4th flask of unselected, untransduced parental control cells was passaged in parallel with the selected, transduced experimental and vector control cells. Parental AG11132 cells are designated HMEC-P, and transduced cells expressing the HPV-16E6 construct are designated p53(Ϫ) HMEC-E6, and vector control clones are designated p53(ϩ) HMEC-LXSN.
Western Blotting-Preparation of cellular lysates and immunoblotting were performed as described previously (34,35). Equal amounts of protein lysates (ϳ100 g of total protein) were loaded on 10% polyacrylamide gels, and the gels were run and then electroblotted (Hoeffer) at 80 mA for 45 min onto Hybond-ECL membrane (Amersham Pharmacia Biotech). The membrane was blocked with 20% bovine serum albumin (Sigma) in PBS overnight at RT and then incubated with a 1:100 dilution of mouse anti-human p53 (Oncogene Science Ab-2). The membrane was washed three to five times at RT with 250 ml of PBS containing 0.1% Tween and then incubated with either a horseradish peroxidase-conjugated goat anti-mouse IgG (Jackson ImmunoResearch) at a 1:35,000 dilution, or a 1:2000 dilution of horseradish peroxidase-conjugated protein A (Sigma) for 1 h at RT. The blot was washed again, and complexes were detected by ECL Western blotting Detection Reagents (Amersham Pharmacia Biotech) as described by the manufacturer.
Growth Curves-p53(ϩ) HMEC-LXSN vector controls and p53(Ϫ) HMEC-E6 cells were plated in duplicate at 1 ϫ 10 4 cells per 12-well tissue culture plates on Day Ϫ1 and allowed to adhere. On Day 0 the medium was replaced with Standard Medium with or without 1.0 M tamoxifen or 4-hydroxytamoxifen. Untreated controls received an equivalent volume of ethanol solvent (0.1% final concentration). Cells were trypsinized at 24-h time intervals and counted in triplicate.
Detection of Apoptosis-annexin V Staining-Annexin V-FITC/a (Boehringer Ingelheim, Heidelberg, Germany) was used as per manufacturer's recommendation with some modification. Approximately 5 ϫ 10 5 p53(ϩ) HMEC-LXSN or p53(Ϫ) HMEC-E6 cells were plated in T-75 flasks on Day Ϫ1 and allowed to adhere. On Day 0 the medium was replaced with fresh Standard Medium, and tamoxifen or 4-hydroxytamoxifen was added for a final concentration of 1.0 M. Untreated controls received an equivalent volume of ethanol solvent (0.1%). Cells were harvested after 24 h (Day 1) and did not exceed 25% confluency. For the tamoxifen-apoptosis time course cells were treated on Day 0 with 1.0 M tamoxifen in Standard Medium and harvested 0, 1, 3, 6, 12, or 24 h after treatment. Cells were trypsinized, washed in PBS, resuspended in binding buffer (10 mM HEPES/NaOH, pH 7.4, 140 mM NaCl, 2.5 mM CaCl 2 ; filtered through a 0.2-m pore filter). Cell density was adjusted to 2-5 ϫ 10 5 cells/ml. Five l of recombinant human annexin V-FITC/a (BMS306F/a) was added to 195 l of cell suspension, and the mixture was briefly mixed and incubated for 10 min at room temperature in the dark. Cells were washed once and resuspended in 190 l binding buffer. Cells were analyzed by FACScan as described below.
Diphenylamine Assay-2.5-5.0 ϫ 10 5 cells were plated per T-25 flask and were grown in Standard Medium. Tamoxifen or 4-hydroxytamoxifen stock was added directly to the media on Day 0 to bring the final concentration to 1.0 M. Confluency did not exceed 70%. Cells were trypsinized 0, 6, 12, 18, or 24 h after treatment, washed in cold PBS, pelleted, and then resuspended in 100 l of lysis buffer (5 mM Tris, pH 8.0, 20 mM EDTA, 0.5% Triton X-100) on ice for 15 min. The lysate was spun at 12,000 rpm for 30 min in a refrigerated microcentrifuge. The supernatant was removed and transferred to a second microcentrifuge tube. Both tubes were placed on ice, and 1.0 ml of 0.5 N perchloric acid was added to the nuclear pellet in the first tube and vortexed. Five hundred l of 1 N perchloric acid was added to the cytoplasmic fraction in the second tube and vortexed. Both tubes were spun at 12,000 rpm for 15 min. The supernatants were then discarded, and 1.0 ml of 0.5 N perchloric acid was added to the pellets. The tubes were heated to 70°C for 20 min to hydrolyze the DNA, then cooled to RT, and 1.0 ml of diphenylamine solution (1.5 g of diphenylamine (Aldrich) in 100 ml of glacial acetic acid, to which was added 1.5 ml of sulfuric acid and 0.1 ml of 1.6% acetaldehyde on the day of use) was added. The tubes were incubated 16 -20 h at 30°C. Absorbance was read at 600 nm (36).
Transmission Electron Microscopy-p53(Ϫ) HMEC-E6 cells and p53(ϩ) HMEC-LXSN vector control cells were plated on Day Ϫ1 in 6-well tissue culture plates. On Day 0 cells were treated with 1.0 M tamoxifen or 4-hydroxytamoxifen for 0, 1, 3, 6, 12, 18, and 24 h. Cells were then fixed in half-strength Karnovsky's fixative (37) for 6 h, rinsed in 0.1 M sodium cacodylate buffer, and post-fixed in 1% collidine-buffered osmium tetroxide. Dehydration in graded ethanol and propylene oxide was followed by infiltration and embedding in Epon 812. Approximately 70 -90-nm sections were stained using saturated aqueous uranyl acetate and lead tartrate. Photographs were taken using a JEOL 100 SX transmission electron microscope operating at 80 kV. Approximately 200 cells were surveyed per data point following treatment for the presence or absence of apoptosis.
Caspase Assays-Activated caspase-3 and -9 were detected utilizing the ApoAlert TM Caspase Fluorescent Assay Kit (CLONTECH). p53(Ϫ) HMEC-E6 cells and p53(ϩ) HMEC-LXSN vector control cells were plated on Day Ϫ1 and treated with 1.0 M tamoxifen or 4-hydroxytamoxifen in duplicate. Cells were trypsinized and counted, and 1 ϫ 10 6 cells were pelleted and frozen at Ϫ80°C. On the day of assay, the pellet was thawed, resuspended in 50 ml of chilled cell Lysis Buffer (CLON-TECH), and incubated on ice for 10 min. Cell lysates were centrifuged for 3 min at 4°C to remove debris, and lysates were then transferred to a new microcentrifuge tube. Fifty l of 2ϫ Reaction Buffer (CLON-TECH), with 10 l of 1 M dithiothreitol, and 5.0 l of either 1.0 mM caspase-3 substrate (DEVD-AFC; 50 M final concentration) or 5 mM caspase-9 substrate (LEHD-AMC; 250 M final concentration) was added to each tube and incubated for 1 h at 37°C. To confirm the correlation between protease activity and product formation, either 1.0 l of caspase-3 inhibitor (DEVD-CHO) or 2.0 l of caspase-9 inhibitor (LEHD-CHO) was added to the reaction mixture of an induced sample and incubated for 1 h at 37°C before adding the caspase-3 or caspase-9 substrate, respectively. Samples were read in a Shimzdzu RF-1501 spectrofluorophotometer with at a 400 nm excitation filter and a 505 nm emission filter (caspase-3) or 380 nm excitation filter and a 460 nm emission filter.
Assessment of Mitochondrial Changes-Mitochondrial transmembrane potential was measured by rhodamine 123 (Molecular Probes) (38) and JC-1 red fluorescence (CLONTECH) (39). JC-1 mitochondrial aggregate formation was measured by JC-1 green fluorescence (CLON-TECH) (39). Relative mitochondrial mass was measured by flow cytometry using 1 n-nonyl acridine orange (NAO; Molecular Probes) (40). For rhodamine 123 staining, 1 ϫ 10 6 cells/ml were incubated at 37°C in 0.5 mg/ml rhodamine 123. For JC-1 staining, 1 ϫ 10 6 cells incubated with 10 g/ml JC-1 for 10 min at 37°C and were analyzed for red and green fluorescence. For NAO staining, 1 ϫ 10 6 cells were resuspended in 1.0 ml of 1.0 M NAO in PBS. Fluorescence of individual nuclei and whole cells was performed using a FACScan flow cytometer equipped with an argon-ion laser at 488 nm and 250 milliwatts light output and Lysis II software (Becton Dickinson Immunocytometry Systems). Forward and side scatter were used to establish size gates and exclude cellular debris. The excitation wavelength was 488 nm. The observation wavelengths were 530 nm for green fluorescence and 585 nm for red fluorescence. The red and green JC-1 fluorescence emissions from each cell were separated and measured using the standard optics of the FACScan. Ten thousand events were collected in list mode fashion, stored, and analyzed on Multicycle AV software (Phoenix Flow Systems).

RESULTS
p53 Protein Suppression in HMECs-Retrovirally mediated expression of the HPV-16 E6 protein was utilized to suppress normal intracellular p53 protein levels in HMECs. Western blots were performed on p53(ϩ) HMEC-LXSN vector controls (passages 10 and 18) and p53(Ϫ) HMEC-E6 transduced cells (passages 10 and 18) to determine the relative levels of p53 protein expression. Expression of p53 protein was observed in p53(ϩ) HMEC-LXSN vector controls but was not detectable by Western analysis in early or late passage p53(Ϫ) HMEC-E6 transduced cells (Fig. 1). Tamoxifen radiolabeled tamoxifen and analyzed by HPLC at 24 h. There was no difference in tamoxifen metabolism in p53(ϩ) or p53(Ϫ) HMECs (data not shown). No tamoxifen was metabolized to 4-hydroxytamoxifen, and all radioactivity was recovered in the tamoxifen peak.
Tamoxifen but Not 4-Hydroxytamoxifen Induces Apoptosis in Early Passage p53(Ϫ) HMEC-E6 Cells-We investigated the mechanism by which tamoxifen might potentiate increased cytotoxicity in early passage p53(Ϫ) HMEC-E6 cells. We observed that early passage p53(Ϫ) HMEC-E6 cells treated with tamoxifen underwent apoptosis as evidenced by annexin V binding, characteristic morphologic changes, and by the presence of fragmented cytoplasmic DNA. In contrast, early passage p53(Ϫ) HMEC-E6 cells did not demonstrate evidence of apoptosis when treated with 4-hydroxytamoxifen.
Annexin V exhibits anti-phospholipase activity and binds to phosphatidylserine (41). Cells undergoing apoptosis acquire annexin V-binding sites during apoptosis and provide a convenient method for detection of cells undergoing apoptosis. We utilized FITC-conjugated annexin V followed by FACS analysis to detect the presence or absence of apoptosis in cells treated  (Fig. 3D). In contrast, early passage p53(Ϫ) HMEC-E6 cells did not undergo apoptosis by this measure when treated with 1.0 M 4-hydroxytamoxifen (Fig. 3E). In addition p53(ϩ) HMEC-LXSN vector controls (passage 10) and late passage p53(Ϫ) HMEC-E6 cells (passage 18) were resistant to tamoxifen-mediated apoptosis in vitro as evidenced by lack of annexin V binding (Fig. 3, B and G).
Electron microscopy of tamoxifen-treated early passage p53(Ϫ) HMEC-E6 cells (passage 10) revealed morphologic changes characteristic of the effector phase of apoptosis including margination of chromatin, cell shrinkage, and formation of apoptotic bodies (5). Margination of chromatin was the first morphologic change detected and was observed 12 h after treatment with 1.0 M tamoxifen (Fig. 4E). After 24 h, 99% of cells exhibited cell shrinkage, condensed chromatin, and formation of apoptotic bodies (Fig. 3D). In contrast, early passage p53(Ϫ) HMEC-E6 cells treated with 1.0 M 4-hydroxytamoxifen did not exhibit evidence of apoptosis by morphologic criteria (data not shown). In addition, neither p53(ϩ) HMEC-LXSN vector controls (passage 10) nor late passage p53(Ϫ) HMEC-E6 cells (passage 21) demonstrated morphologic evidence of apoptosis after treatment with 1.0 M tamoxifen for 24 h (Fig. 4, B and H, and data not shown).
Sequential changes in mitochondrial ultrastructure were identified in tamoxifen-treated early passage p53(Ϫ) HMEC-E6 cells undergoing apoptosis but not in tamoxifentreated p53(ϩ) HMEC-LXSN cells nor in late passage p53(Ϫ) HMEC-E6 cells. One hour after treatment with 1.0 M tamoxifen the mitochondrial matrix in early passage p53(Ϫ) HMEC-E6 cells (passage 10) is condensed, and the cristae are well formed, transversing the mitochondrion (Fig. 6D). Six hours following treatment with tamoxifen, mitochondria in early passage p53(Ϫ) HMEC-E6 cells exhibit further morphologic changes. Mitochondria are small, and the outer mitochondrial membrane appears indistinct, and internal structures are obscured by highly electron-dense material (Fig. 6E). Overall, these changes are consistent with mitochondrial matrix condensation and volume loss. After 6 h of treatment, nuclear chromatin is normal, and there is no morphologic evidence of apoptosis (data not shown). At 12 h, a majority of cells exhibited margination of chromatin and mitochondrial condensation ( Fig. 4E and data not shown). In contrast, tamoxifen-treated p53(ϩ) HMEC-LXSN vector controls (passage 10) and late passage p53(Ϫ) HMEC-E6 cells (passage 21) do not exhibit changes in mitochondrial morphology after 24 h of treatment (Fig. 6, B and H). These observations indicate that mitochondrial matrix condensation, starting at 1 h, precedes execution phase morphologic changes in early passage p53(Ϫ) HMEC-E6 cells treated with tamoxifen but not 4-hydroxytamoxifen.
Caspase Activation in Tamoxifen-sensitive and -resistant HMECs-Caspases are thought to be responsible, in part, for cellular changes that occur during apoptosis such as DNA fragmentation, chromatin condensation, and formation of apoptotic bodies. Effector caspases are constitutively expressed in their inactive form and are activated through intracellular caspase cascades. We investigated the relationship between caspase-9 and caspase-3 activation in tamoxifen-treated, apoptosis-sensitive and -resistant HMECs.
Recent evidence suggests that apoptosis can be triggered by inducing mitochondrial release of cytochrome c. The initiator caspase, caspase-9, is activated by mitochondrial release cytochrome c and that in turn activates caspase-3 (9). Activation of caspase-9 was observed in apoptosis-sensitive, early passage p53(Ϫ) HMEC-E6 cells (passage 11) treated with 1.0 M tamoxifen but not 1.0 M 4-hydroxytamoxifen for 6 h (Fig. 7A). In contrast, p53(ϩ) HMEC-LXSN vector controls (passage 12) and apoptosis-resistant, late passage p53(Ϫ) HMEC-E6 cells (pas- sage 19) did not exhibit caspase-9 activation when treated with tamoxifen (Fig. 7A). Caspase-9 was maximally activated 1 h after tamoxifen treatment in early passage p53(Ϫ) HMEC-E6 cells and correlated with appearance of mitochondrial condensation first detected by electron microscopy at 1 h. Caspase-9 activation significantly preceded the detection of apoptosis by annexin V binding at 12 h (data not shown) and the detection of chromatin condensation, cell shrinkage, and formation of apoptotic bodies at 24 h (Fig. 7B). These observations indicate that casapse-9 activation may be an early event in tamoxifeninduced apoptosis.

FIG. 4. Morphologic evidence of apoptosis in tamoxifen-treated but not in 4-hydroxytamoxifen-treated early passage p53(؊) HMEC-E6 cells or in tamoxifen-treated p53(؉) HMEC-LXSN vector controls or late passage p53(؊) HMEC-E6 cells. A and
Caspase-3 is an active cell death protease involved in the execution phase of apoptosis. A 5-fold activation of caspase-3 was observed in apoptosis-sensitive, early passage p53(Ϫ) HMEC-E6 cells (passage 11) treated with 1.0 M tamoxifen but not 1.0 M 4-hydroxytamoxifen for 24 h (Fig. 8A). In contrast, apoptosis-resistant, late passage p53(Ϫ) HMEC-E6 cells (passage 19) did not exhibit activation of caspase-3 when treated with tamoxifen (Fig. 8A). We next tested whether the temporal activation of caspase-3 correlated with morphologic changes observed in early passage p53(Ϫ) HMEC-E6 cells undergoing tamoxifen-mediated apoptosis. Caspase activation was first detected at 12 h after tamoxifen treatment and preceded the detection of apoptosis by electron microscopy at 24 h (Fig. 8B).
Relationship between Mitochondrial Transmembrane (⌬⌿ m ) Potential and Sensitivity to Tamoxifen-induced Apoptosis-Disruption of mitochondrial electron transport is an early feature of apoptosis. The mitochondrial abnormalities observed by electron microscopy during tamoxifen-induced apoptosis were further evaluated using fluorescent measures of mitochondrial mass and membrane potential. Mitochondrial mass was measured by staining with NAO, a fluorescent dye that specifically binds to the mitochondrial inner membrane independent of the transmembrane potential (40). To measure mitochondria potential (⌬⌿ m ), cells were stained with rhodamine 123 (38) or the J-aggregate forming cationic dye JC-1 (39). JC-1 is a dye that normally exists as a monomer emitting green fluorescence. JC-1 is taken up by mitochondria and in response to the mitochondrial membrane potential forms multimers, which then emit red fluorescence (39). Decreasing mitochondrial transmembrane potential results in a decrease in JC-1 red fluorescence and an increase in JC-1 green fluorescence.
The relative mitochondrial mass was similar in p53(ϩ) HMEC-LXSN cells (passage 12) and early passage p53(Ϫ) HMEC-E6 cells (passage 11) (1.0 Ϯ 0.05 and 0.92 Ϯ 0.07, respectively) ( Table I). These data are consistent with the observation made by electron microscopy that there is a qualitative increase in the number of mitochondria in apoptosis-resistant p53(Ϫ) HMEC-E6 cells relative to apoptosis-sensitive p53(Ϫ) HMEC-E6 cells and p53(ϩ) vector controls (Fig. 4). These observations suggest that an increase in mitochondrial mass may be associated with the development of resistance to tamoxifen-induced apoptosis.
To Compare the Base-line Values of ⌬⌿ m in apoptosis-sensitive and -resistant cells, ⌬⌿ m was normalized to mitochondrial mass (⌬⌿ m /NAO fluorescence). The normalized values of ⌬⌿ m were decreased in both untreated apoptosis-sensitive (passage 11) and untreated apoptosis-resistant (passage 19) p53(Ϫ) HMEC-E6 cells relative to p53(ϩ) HMEC-LXSN vector control cells (passage 12) ( Table I). The decreased base-line ⌬⌿ m and normal mitochondrial content in apoptosis-sensitive p53(Ϫ) HMEC-E6 cells correlates with sensitivity to tamoxifeninduced apoptosis. DISCUSSION Apoptosis is a dynamic process that involves initiation by a pharmacologic or DNA-damaging agent, activation of proteolytic enzymes, and execution of characteristic morphologic changes. Mitochondria serve as sensors and amplifiers of the apoptotic process (43). Diverse apoptotic stimuli converge at the mitochondria resulting in activation of the caspase proteolytic cascade that ultimately leads to cellular disassembly (9). In this study, we present evidence that the acute suppression of p53(Ϫ) function in HMECs, mediated by HPV-16 E6, results in increased sensitivity to tamoxifen-induced apoptosis via a signaling pathway that involves mitochondrial depolarization and caspase activation. In contrast, the related metabolite, 4-hydroxytamoxifen fails to induce either mitochondrial depolarization or caspase activation and subsequently does not induce apoptosis.
We observe that early passage p53(Ϫ) HMEC-E6 cells treated with tamoxifen for 12-24 h exhibit morphologic and biochemical changes characteristic of the execution phase of apoptosis. These apoptotic changes are preceded by mitochondrial matrix condensation, mitochondrial depolarization (⌬⌿ m ), and caspase-9 activation, all first observed 1 h after tamoxifen treatment. p53(Ϫ) HMEC-E6 cells rapidly developed resistance to tamoxifen-induced apoptosis after about 10 passages in vitro. Resistance to tamoxifen was associated with an increase in mitochondrial mass, the appearance of a branching mitochondrial morphology, and lack of membrane depolarization and caspase activation following tamoxifen treatment. Taken together, these observations suggest a critical role for mitochondrial signaling in mediating sensitivity to tamoxifeninduced apoptosis.
To date, most studies have investigated tamoxifen action using human breast cancer cell lines that contain complex chromosomal rearrangements and are a poor model for chemoprevention. Although the importance of p53 as a tumor suppressor is well documented, little is known about the fate of normal human cells that acutely lose p53 function in the con-text of tamoxifen chemoprevention. A majority of cellular studies investigating the role of p53 in tamoxifen sensitivity have been made in experimentally transformed cells lines or in cancer cell lines. Loss of p53 function confers genetic instability, and studies of p53 function in these model systems may be complicated by mutations acquired subsequent to p53 inactivation. We observe that p53(Ϫ) HMEC-E6 cells are genetically unstable and acquire major chromosomal rearrangements and deletions within 10 passages of transduction in vitro. 3 It has been observed recently that the acute loss of p53 function results in enhanced sensitivity to apoptosis in normal human fibroblasts expressing HPV-16 E6, human placental cells expressing SV40 T antigen, and mouse embryonic fibroblasts isolated from p53 Ϫ/Ϫ transgenic mice (44 -46). We compared base-line mitochondrial membrane potential (⌬⌿ m ) standardized to mitochondrial mass (M m ) in p53(ϩ) and p53(Ϫ) HMECs to investigate whether a decrease in base-line ⌬⌿ m /M m might correlate with sensitivity to tamoxifen-induced apoptosis. We observed that ⌬⌿ m /M m in apoptosis-insensitive p53(ϩ) HMEC-LXSN controls was significantly increased relative to apoptosis-sensitive early passage p53(Ϫ) HMEC-E6 cells. This observed base-line mitochondrial membrane depolarization may provide a mechanism for the increased sensitivity of early passage p53(Ϫ) HMECs to apoptotic stimuli. Interestingly, the development of tamoxifen resistance in late passage p53(Ϫ) HMEC-E6 cells was not the result of increased 3  ⌬⌿ m /M m . This indicates that although sensitivity to tamoxifen apoptosis correlates with a decrease in baseline ⌬⌿ m /M m , resistance to tamoxifen is not associated with a return to baseline mitochondrial membrane potential.
Tamoxifen has been extensively studied as a chemotherapeutic agent; however, the mechanism of tamoxifen chemoprevention in normal mammary tissue is poorly understood. Tamoxifen chemotherapy is felt to involve the following mechanism: ligand-bound steroid receptors bind to specific promoter elements and thereby activate or inhibit the expression of target genes (47). This "genomic" mechanism of tamoxifen action requires the presence of the estrogen receptor and both transcription and translation. However, normal proliferating luminal mammary epithelial cells exhibit both low levels of estrogen receptor expression and estrogen binding (48). We observe that tamoxifen, but not an equimolar concentration of 4-hydroxytamoxifen, induces apoptosis in early passage p53(Ϫ) HMEC-E6 cells. Since 4-hydroxytamoxifen has a higher affinity for the estrogen receptor than does tamoxifen, we expected that if apoptosis was initiated via an estrogen receptor-derived signal, 4-hydroxytamoxifen would exhibit equal or increased ability to induce apoptosis relative to tamoxifen (18,30). We observe, however, that whereas tamoxifen is able to induce apoptosis in p53(Ϫ) HMEC-E6 cells, 4-hydroxytamoxifen induces growth arrest alone.
It is hypothesized that patients treated with tamoxifen may exhibit de novo or acquired resistance through changes in drug metabolism. Recently it has been observed that there is an accumulation of 4-hydroxytamoxifen in primary breast cancers that have acquired resistance to tamoxifen chemotherapy (49). In addition, significantly higher levels of 4-hydroxytamoxifen, relative to tamoxifen, were observed in the plasma samples taken from the acquired resistance group (49). These observations lead the investigators to hypothesize that perhaps the appearance of increased levels of 4-hydroxytamoxifen could account for tamoxifen resistance and are consistent with observations in our in vitro system.
Although the tamoxifen metabolite, 4-hydroxytamoxifen, is detected in the plasma and tissue of women treated with tamoxifen, we do not detect 4-hydroxytamoxifen in tamoxifentreated HMECs by HPLC. This observed lack of tamoxifen metabolism allows us to assess independently the ability of tamoxifen and 4-hydroxytamoxifen to induce apoptosis. Since 1) normal mammary epithelial tissue exhibits low levels of estrogen receptor expression and 2) treatment of p53(Ϫ) HMEC-E6 cells with an agent that has increased in vitro affinity for the estrogen receptor fails to induce apoptosis, this raises the possibility that tamoxifen chemoprevention may be, in part, mediated by an estrogen receptor-independent pathway.
Recent evidence suggests that estrogens and perhaps antiestrogens may also act through nongenomic, calcium-mediated signaling pathways (47). Estrogen has been shown to induce extremely rapid increases in intracellular calcium and cAMP (50 -52) and thereby activate mitogen-activated protein kinase (52). We observe that induction of apoptosis in early passage p53(Ϫ) HMEC-E6 cells occurs within 1 h of tamoxifen treatment. This rapid induction of apoptosis is consistent with a nongenomic mechanism of anti-estrogen signaling. Tamoxifen and 4-hydroxytamoxifen are structurally and functionally very similar but differ in their affinity for calmodulin (53). Calmodulin has been shown recently to be a mediator of apoptosis in several cell systems (54,55), and calcium/calmodulin-dependent protein kinase has been shown to regulate apoptosis through the death-associated protein kinase-2 (31, 57). We speculate that apoptosis mediated by anti-estrogens in HMECs with low levels of estrogen receptor expression may involve a calcium-mediated signaling pathway, perhaps modulated by calmodulin binding.
Cumulatively, the data presented in this study support the hypothesis that tamoxifen-mediated apoptosis in early passage p53(Ϫ) HMEC-E6 cells rapidly occurs via a mitochondrial signaling pathway, requiring mitochondrial membrane depolarization and the activation of caspase-3 and caspase-9. Whereas tamoxifen readily induces apoptosis in early passage p53(Ϫ), the tamoxifen metabolite 4-hydroxytamoxifen does not induce mitochondrial changes and hence does not induce apoptosis. These data suggest a need to investigate the relative contributions of genomic and nongenomic mechanisms in the induction of apoptosis by tamoxifen, an activity that is likely to support the development of novel pharmacologic agents for breast cancer chemoprevention.  (⌬⌿ m ) is decreased in early passage p53(Ϫ) HMEC-E6 cells relative to p53(ϩ) HMEC-LXSN vector controls Mitochondrial potential (⌬⌿ m ) is measured by rhodamine 123 staining and by JC-1 (red) fluorescence and normalized to mitochondrial mass, measured by NAO staining. Mitochondrial mass in apoptosis-resistant, late passage p53(Ϫ) HMEC-E6 cells (passage 21) is relative to p53(ϩ) HMEC-LXSN vector controls (passage 11) and apoptosis-sensitive, early passage p53(Ϫ) HMEC-E6 cells (passage 11). Base-line mitochondrial potential normalized to mitochondrial mass (JC-1 red (⌬⌿ m )/NAO or rhodamine (⌬⌿ m )/NAO fluorescence) in p53(Ϫ) HMEC-E6 cells is relative to p53(ϩ) HMEC-LXSN vector controls. Fluorescent values are reported relative to ethanol-treated controls. Reported values represent the average of three separate experiments. Tam, tamoxifen.