MCF-7 (American Type Culture Collection, Rockville, MD), MCF-7CLU and SUM-159PT (University of Michigan Human Breast Cell/Tissue Bank, Ann Arbor, MI) were cultured in alpha-MEM medium (Life Technologies Inc, Gaithersburg, MD) containing 25 mM HEPES and 5% fetal bovine serum (ATLAS Biologicals, Fort Collins, CO). Cells were routinely passaged every 4-5 days. For cell growth and morphological assays, MCF-7 and MCF-7CLU cells were plated at either 10,000 or 20,000 cells/well in 24-well plates and were treated with 1, 5 or 10 micromolar tamoxifen (Sigma, St. Louis, MO), or 1, 5 or 10 ng/ml TNFalpha(Sigma) or ethanol vehicle for the indicated times, starting two days after plating.
Over-expression of clusterin
MCF-7 cells were engineered to over-express the glycosylated clusterin isoform using Gateway Technology (Invitrogen Corporation, Carlsbad, California) according to the manufacturer's protocol. Briefly, Gateway's attB1 and attB2 recombination sites were added to the 5' and 3' ends respectively of human clusterin cDNA  by PCR amplification using primers containing the attB sites. Using Gateway Technology, which allows for the transfer of DNA among different vectors based on lambda-phage recombination, the clusterin cDNA was inserted into the pDONOR 201 plasmid to generate the entry clone. This vector was subsequently transferred into the pcDNA-DEST47 destination vector to create the expression clone pEXP-CLU which allows for constitutive gene expression under the control of a CMV promoter. A second construct was also engineered in which the stop codon was altered by a single base pair mutation of TGA to GGA yielding a glycine residue substitution. This alteration was generated to allow for expression of a C-terminal GFP-tagged clusterin (pEXP-CLUGFP). The construct mutations and correct transfer into pcDNA-DEST47 were confirmed in all cases by automated DNA sequencing.
The expression constructs (pEXP-CLU and pEXP-CLUGFP) were independently transfected into MCF-7 cells using Lipofectamine 2000 (GibcoBRL) according to the manufacturer's direction and MCF-7CLU cells were selected for by growth in appropriate media containing neomycin (BRL, Gaithersberg, MD).
Cell growth was determined by crystal violet assays. For quantitation of total adherent cell number, cells were fixed with 1% gluteraldehyde (Fisher Scientific, Pittsburgh, PA) for 15 min, incubated with 0.1% crystal violet (Fisher Scientific) for 30 min, destained with H2O, and solubilized with 0.2% Triton X-100 (Sigma). Absorbance at 590 nm, which is proportional to total adherent cell number, was determined on a microplate reader.
DNA was isolated from MCF-7 and MCF-7CLUGFP cells using a commercially available assay according to the manufacturer's protocol (Qiagen, Valencia, CA). Primers were designed that spanned the CLUGFP gene sequence and the pcDNA-DEST vector. DNA was amplified directly using primers specific for the CLU-pcDNA-DEST vector to generate a product of 380 bp. For MCF-7CLU-pcDNA-DEST47 the primers were as follows:
The amplification conditions were as follows: A denaturation step for 5 min at 95°C, followed by 45 sec at 95°C, 45 sec at 64°C and 45 sec at 72°C for 30 cycles, followed by 15 min at 72°C. 20 microliters of the PCR reaction was analyzed on a 1.5% agarose gel.
MCF-7 and MCF-7CLU cells were plated at 10,000 cells/well onto poly-L-lysine coated Lab-Tek II chamber slides (Fisher Scientific) and fixed in 4% formaldehyde after 72 h growth. Cells were analyzed by phase and fluorescence microscopy and photographed using an Olympus AX70 microscope equipped with a Spot RT digital camera.
MCF-7 and MCF-7CLU cells were plated at a density of 8 × 105 cells/150-mm flask. 48 h following plating, cells were treated with 1, 2.5 or 5 micromolar tamoxifen or 1, 2.5 or 5 ng/ml TNFalpha or ethanol vehicle for the indicated times. For analysis of cell cycle kinetics, MCF-7 and MCF-7CLU cells were harvested by trypsinization, pooled with floating cells, pelleted by centrifugation (500 × g, 5 min, 4°C), fixed and permeabilized with 90% ethanol at -20°C. Cells were stained with 5 micrograms/ml propidium iodide (Sigma) containing RNase A (Roche Molecular Biochemicals, Mannheim, Germany). For analysis of DNA fragmentation, MCF-7 and MCF-7CLU cells were harvested by trypsinization at the indicated times, collected by centrifugation, fixed in 2% formaldehyde in PBS, and permeabilized in 70% ethanol at -20°C. DNA strand breaks in cells undergoing apoptosis were indirectly labeled with bromodeoxyuridine by terminal transferase (Roche Molecular Biochemicals) and detected by FITC-conjugated monoclonal antibody to bromodeoxyuridine using the APO-BrdU kit according to manufacturer's protocol (Phoenix Flow Systems, San Diego, CA). Cells were counterstained with 5 microgram/ml propidium iodide containing RNase A for detection of total DNA, and two-color analysis of DNA strand breaks and cell cycle was achieved by flow cytometry. All flow cytometric analyses were performed on an Epics XL Flow Cytometer (Coulter Corp., Miami, FL) equipped with an argon laser. Data was modeled with the Multiplus AV software (Phoenix Flow Systems).
Preparation of cell lysates, conditioned media and subcellular fractions
For cell lysates, monolayers were washed with PBS, scraped into 2 × Laemlli buffer (containing protease and phosphatase inhibitors) and sonicated prior to protein determination using the BCA protein assay (Pierce, Rockland, IL).
For conditioned media fractions, serum free media from cells was centrifuged (500 × g, 5 min, 4°C) to pellet any floating cells and concentrated using ultra-centrifugation filter units (5 KDa cut-off) according to the manufacturer's protocol (Fisher Scientific).
For subcellular fractions, adherent and floating cells were pelleted by centrifugation (500 × g, 5 min, 4°C), resuspended in wash buffer (25 mM Tris pH 7.5, 250 mM sucrose, 2.5 mM MgCl2, protease and phosphatase inhibitors), pelleted (500 × g, 5 min, 4°C), resuspended in 3 volumes of buffer A (20 mM HEPES-KOH, pH 7.5, 10 mM KCL, 1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA, 250 mM sucrose, protease and phosphatase inhibitors), lysed with a Dounce homogenizer, and fractionated by differential centrifugation. Briefly, homogenates were centrifuged twice (500 × g, 5 min, 4°C) and the supernatants transferred to an ultracentrifuge tube. The pellets (containing nuclei) were resuspended in Buffer A, sonicated and stored at -80°C. The combined supernatants were ultracentrifuged (100,000 × g, 1 h, 4°C) to generate the cytosolic fraction (supernatant) and the non-nuclear membrane fraction (NNMF) (pellet). The pellets were resuspended in 100 ml Buffer A, sonicated and stored at -80°C. All protein samples were analyzed for total protein using the Micro BCA protein assay (Pierce, Rockford, IL).
Western blot analysis
Subcellular fractions, conditioned media, or cell lysates isolated as described above were solubilized in Laemmli sample buffer, separated under reducing conditions by SDS-PAGE and transferred to nitrocellulose. Proteins derived from the NNMF and cytosolic extracts were immunoblotted with a rabbit polyclonal antibody directed against Bax (Clone 13666E; Pharmingen, San Diego, CA) and normalized for protein loading with mouse monoclonal antibodies directed against GAPDH (Clone 6G5; Biogenesis, Brentwood, NH) and ATP synthase (Molecular Probes, Inc., Eugene, OR) diluted 1:500, or 1:1000 respectively. Proteins derived from nuclear fractions were immunoblotted with a mouse monoclonal antibody directed against p21WAF1 (Clone EA10; Calbiochem, San Diego, CA) diluted 1:50. Protein samples were normalized by comparison to a mouse monoclonal antibody antibodies directed against lamin A/C (Clone sc-7293; Santa Cruz, Santa Cruz, CA) diluted 1:100. Proteins derived from cell lysates were immunoblotted with mouse monoclonal directed against ER-alpha (Clone 6F11; Nova Castra, New Castle upon Tyne, UK) or clusterin (Clone 7D1; RDI, Flanders, NJ) diluted 1:50 and 1:1000 respectively and normalized for protein loading by immunoblotting with GAPDH antibody. All antibodies were diluted in PBS plus 5% skim milk. Proteins derived from conditioned media were immunoblotted with anti-clusterin antibody (Clone 7D1; RDI, Flanders, NJ). Specific antibody binding was detected by the appropriate horseradish peroxidase-conjugated secondary antibodies (BioRad, Hercules, CA) diluted 1:5000 in PBS plus 5% skimmed milk and detected using enhanced chemiluminescence. (Pierce, Rockford, IL). The band intensities were measured and analyzed with the Kodak 1D imaging software. Changes in protein levels were normalized relative to the appropriate loading control (lamin A/C for the nuclear fraction; GAPDH for the cytosolic fraction, and ATP-synthase-alpha for the mitochondrial fraction) and then plotted relative to the appropriate control (which was normalized to a value of 1).
Matrigel invasion assay
Invasion was assessed using 8 micron invasion chambers coated with growth factor reduced Matrigel™ (Becton Dickinson, Bedford, MA) according to the manufacturer's protocol with the exception that fibroblast conditioned media (from NIH 3T3 cells) was used as a chemo-attractant. The number of invasive MCF-7, MCF-7CLU and SUM-159PT cells was determined 24 h after 1 × 104 cells were seeded into the Matrigel™ chambers. At that time the Matrigel™ was removed from the inserts and the cells on the underside of the membranes were fixed using 1% gluteraldehyde for 15 min, stained with 0.1% crystal violet for 30 min and rinsed with water to remove unbound dye. Membranes were dried, and numbers of invasive cells on each insert were counted by viewing under an Olympus AX70 microscope.
Orthotopic xenograft of MCF-7CLU cells
Six week old ovariectomized female NCr-nu mice (Taconic Farms, Germantown, NY) supplemented with 17-beta-estradiol-sustained release pellets (1.7 mg/90 days) (Innovative Research, Sarasota, FL), were used as hosts for MCF-7 and MCF-7CLU xenografts. Mice were inoculated orthotopically with 2 × 106 cells/0.1 ml injection directly into the inguinal mammary fat pad under anesthesia. Body weight and tumor size was monitored weekly by caliper measurements of the length, width and height, and volume was calculated using the formula for a semi-ellipsoid (4/3Br3/2). When tumor volume reached approximately 200 mm3 (4-5 weeks), mice were randomized to control or tamoxifen treatment groups. Anesthetized mice were implanted subcutaneously with placebo or tamoxifen (15 mg/60 day) continuous release pellets. After 6 weeks of treatment, mice were euthanized, blood was collected by cardiac puncture, tumors and organs were removed, examined and fixed in 4% formalin (Fisher Scientific) for histological analysis. All animal studies were carried out in accordance with an IACUC-approved protocol at the University of Notre Dame.
Histological analysis of tumors and organs
For routine morphological assessment, tumors and organs were embedded in paraffin, sectioned at 5 microns, and stained with hematoxylin (Gill's formulation 3, Fisher Scientific) and eosin Y (Sigma, St. Louis, MO). Mitotic and apoptotic cells were identified by immunohistochemistry for proliferating cell nuclear antigen (PCNA) and by TUNEL assay. For PCNA, sections were incubated with mouse monoclonal anti-PCNA (Nova Castra Laboratories) at a 1:50 dilution in 1% BSA-PBS. The secondary biotin conjugated antibody (Vector Laboratories, Burlingame, CA) was applied at a 1:200 dilution in 1% BSA-PBS. The ABC technique followed by diaminobenzidine was used to localize peroxidase, and sections were counterstained with Harris' modified hematoxylin (Fisher). TUNEL was performed with a commercially available assay according to manufacturer's directions (Boehringer Mannheim, Indianapolis, IN).
For quantitation of proliferation and apoptosis PCNA and TUNEL positive cells were counted as a percentage of total cells. At least 4 mice from each group were chosen and 2 tumor sections (approximately 500 cells/section) from each mouse were quantitated for PCNA and TUNEL positivity.
Statistical comparisons were performed using Student's unpaired t tests (for two groups) or one-way non-parametric ANOVA for more than two groups. Data are expressed as the mean ± SD or mean ± SEM, depending on the context as indicated in the figure legends, and differences between means were considered significant at P < 0.05.