Peroxisome Proliferator-activated Receptor γ Agonists Induce Proteasome-dependent Degradation of Cyclin D1 and Estrogen Receptor α in MCF-7 Breast Cancer Cells

INTRODUCTION

Peroxisome proliferators were initially characterized from a large group of synthetic industrial and pharmaceutical chemicals that induced hepatic hypertrophy and hyperplasia in rodents (1, 2, 3) . The effects of these compounds were accompanied by increased numbers and size of liver peroxisomes and induction of enzymes required for oxidative metabolism of fatty acids and members of the cytochrome P4504A (CYP4A) family. The intracellular receptor for peroxisome proliferator-induced hepatic responses was first reported in 1990 as PPARα 3 (4) , and subsequent studies in several laboratories have also characterized PPARβ (or PPARδ), PPARγ, and several isoforms that arise from alternative splicing and promoter use (5, 6, 7, 8, 9) . PPARs are differentially expressed in various tissues and tumors and play a critical role in fatty acid metabolism and energy homeostasis (reviewed in Refs. 1, 2, 3 ). PPARs are ligand-activated transcription factors and members of the nuclear receptor superfamily (10 , 11) . Activation of PPARs is a multistep process that involves ligand binding and heterodimerization with the retinoic X receptor, interaction with sequence-specific gene promoter elements, and recruitment of coactivators and other nuclear coregulatory proteins. PGJ2 is the most potent eicosanoid activator of PPARγ (12 , 13) ; thiazolidinediones such as ciglitazone are synthetic PPARγ agonists used extensively for their antidiabetic properties and treatment of insulin-resistant type II diabetes (14, 15, 16, 17) .

PPARγ is widely expressed in multiple tumors and cell lines, and this receptor has also become a target for developing new anticancer drugs that will take advantage of the antiproliferative effects mediated through PPARγ. For example, a recent study investigated PPARγ expression in 339 clinical tumor samples from colon, breast, lung, prostate, osteosarcomas, glioblastomas, acute myelogenous leukemia, adult T-cell leukemia, B-cell acute lymphoblastic leukemia, B-cell non-Hodgkin’s lymphoma, and myelodisplastic syndrome (18) . Wild-type PPARγ mRNA was expressed in all tumor specimens, and receptor mutants were not detected in any of these samples. The growth-inhibitory effects of endogenous and synthetic PPARγ agonists have been investigated in several tumors and cancer cell lines (19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42) , and a number of these studies show that ligands for this receptor induce apoptosis and/or decrease G₀-G₁→S-phase cell cycle progression, which is accompanied by a decrease in cyclin D1 or modulation of cdk inhibitors and other factors involved in cell growth.

Studies in breast cancer cells show that PPARγ agonists inhibit growth of ER-positive and -negative cell lines. Treatment of ER-positive MCF-7 cells with PPARγ agonists inhibits activation of epidermal growth factor receptors through inhibition of tyrosine phosphorylation (42) and up-regulates PTEN expression in MCF-7 and other cancer cell lines (41) . PGJ2 also repressed cyclin D1 mRNA and protein in MCF-7 cells, and inhibition of transactivation was associated with enhanced recruitment of limiting cellular levels of p300 to PPARγ (39) . This study further investigates the mechanism of PPARγ-induced inhibition of cancer cell growth using MCF-7 human breast cancer cells as a model. The results show that both PGJ2 and ciglitazone (a thiazolidinedione) induce proteasome-dependent degradation of cyclin D1 and ERα, and this represents a novel pathway for PPARγ-mediated growth arrest in breast cancer cells and is consistent with their inhibition of G₀-G₁→S-phase progression.

MATERIALS AND METHODS

Cells, Chemicals, Biochemicals, and Other Materials.

MCF-7 cells were obtained from American Type Culture Collection (Manassas, VA) and maintained in MEM with phenol red and supplemented with 0.22% sodium bicarbonate, 10% FBS, 0.011% sodium pyruvate, 0.1% glucose, 0.24% HEPES, 10⁻⁶% insulin, and 10 ml/liter antibiotic solution. Cells were grown in 150-cm² culture plates in an air:carbon dioxide (95:5) atmosphere at 37°C and passaged every 6 days. Cells were seeded in DMEM:Ham’s F-12 with 5% FBS, and cell proliferation studies were determined using different concentrations of PGJ2 or ciglitazone; cell numbers were determined using a Coulter Z1 cell counter. DMSO, PBS, and 100× antibiotic solution were purchased from Sigma Chemical Co. (St. Louis, MO). PGJ2 (PG-050) and ciglitazone were purchased from Biomol Research Laboratories Inc. (Plymouth Meeting, PA). MG132, PSII, calpeptin, and CII were purchased from CalBiochem-Novabiochem Co. (San Diego, CA). FBS was obtained from Intergen (Purchase, NY). Horseradish peroxidase substrate for Western blot analysis was purchased from New England Nuclear Life Science Products (Boston, MA). Antibodies for cyclin D1 (sc-718 and sc-246), ERα (sc-544 and sc-8005 for Western blot and immunoprecipitation, respectively), PPARγ (sc-7196), Sp1 (sc-59 and sc-420), ubiquitin (sc-8017), cdk4 (sc-260), and preimmune IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Immunoprecipitation antibody for cyclin D1 was obtained from Biosource International (Camarillo, CA). Restriction enzymes, T4-polynucleotide kinase, RNase ONE, and pGL2 luciferase reporter vector were purchased from Promega (Madison, WI) and Boehringer Mannheim (Indianapolis, IN). Reporter Lysis Buffer and Luciferase Reagent for luciferase studies were purchased from Promega. β-Gal reagent was purchased from Tropix (Bedford, MA). SuperFect transfection and plasmid preparation kits were purchased from Qiagen (Santa Clarita, CA). All other chemicals and biochemicals were the highest quality available from commercial sources. Lab-Tek Chamber slides were purchased from Nalge Nunc International (Naperville, IL). InstantImage and Luminometer were purchased from Packard (Meriden, CT).

Cloning and Plasmid Preparation.

The PPRE3Luc luciferase reporter was constructed using the TpGL2 vector, which was constructed by inserting a minimal TATA sequence (GCT-GTA-GGG-TAT-ATA-ATG-GAT-CA) with linkers into the BglII and HindIII sites in the pGL2 basic vector (Promega). The triple consensus PPRE nucleotides were synthesized by Genosys/Sigma (The Woodlands, TX) as a tandem repeat inserted between the SacI and BglII sites. The Gal4 reporter (pGAL4Luc) containing five tandem GAL4 response elements was kindly provided by Dr. Marty Mayo (University of North Carolina, Chapel Hill, NC). Gal4DBD-PPARγ construct (gPPARγ) was a gift of Dr. Jennifer L. Oberfield (Glaxo Wellcome Research and Development, Research Triangle Park, NC), and PPARγ expression plasmid was a gift of Dr. Bruce M. Spiegelman (Harvard University, Boston, MA). All constructs were transformed into TOP10F′ competent cells (Invitrogen, Carlsbad, CA). Plasmids were confirmed by restriction enzyme mapping and DNA sequencing. High quality plasmids for transfection were prepared using Qiagen Plasmid Megaprep Kit.

Transient Transfection and Luciferase Activity Assay.

MCF-7 cells were seeded in 5% FBS DMEM:Ham’s F-12 in 12-well plates 1 day before transfection using the calcium phosphate-DNA coprecipitation method or SuperFect transfection kit. PPRE3Luc reporter plasmid (1.1 μg), PPARγ expression plasmid (0.4 μg), and β-Gal DNA (0.1 μg) were used; alternatively, 1.5 μg of GAL4Luc reporter plasmid, 0.05 μg of gPPARγ, and 0.1 μg of β-Gal DNA were used for transfection. After incubation for 16 h (with calcium phosphate) or 3 h (with SuperFect), cells were washed with PBS and treated with compounds as indicated for 16–20 h in fresh media. Cells were then lysed with 200 μl of 1× Reporter Lysis Buffer; 30 μl of cell extract were used for luciferase and β-Gal assays. LumiCount was used to quantitate luciferase and β-Gal activities, and the luciferase activities were normalized to β-Gal activity.

Western Blot Analysis.

MCF-7 cells were seeded in 5% charcoal-stripped FBS and DMEM:Ham’s F-12 for 24 h and then treated with the indicated compounds. WCLs were obtained using 1× Western sampling buffer. Protein samples were heated at 100°C for 5 min, separated on 10% SDS-PAGE at 160 V for 3 h in 1× running buffer [25 mm Tris-base, 192 mm glycine, and 0.1% SDS (pH 8.3)], and transferred to polyvinylidene difluoride membrane (Amersham) at 100 V for 2 h at 4°C in 1× transfer buffer [48 mm Tris-HCl, 39 mm glycine, and 0.075% SDS]. The polyvinylidene difluoride membrane was blocked in 5% milk-TBS [10 mm Tris-HCl and 150 mm NaCl (pH 8.0)] with gentle shaking for 1 h and incubated in fresh 5% milk-TBS with 1:5000 (for cyclin D1), 1:1000 (for ERα, cdk4, and PPARγ), or 1:8000 (for Sp1) primary antibody (Santa Cruz Biotechnology) for 1 h with gentle shaking. After vigorous washing in 1× TBS for 30 min, secondary antibody (1:5000) in 5% milk-TBS was incubated with shaking for 1 h. The membrane was washed vigorously in TBS buffer for 30 min, incubated in 10 ml of chemiluminescent substrate (ECL; New England Nuclear Life Science Products, Inc.) for 1.0 min, and exposed to Kodak X-OMAT AR autoradiography film immediately. Band intensities were evaluated by scanning laser densitometry (Sharp Electronics Corp.) The same membrane was stripped and then used for Western blot analysis of specific proteins in the same treatment group. The aqueous stripping solution (1 liter) was prepared with Tris-HCl (9.85 g; pH 6.8), SDS (20 g) and β-mercaptoethanol (7.8 ml).

Coimmunoprecipitation/Western Blot.

MCF-7 cells were seeded in 5% DMEM:Ham’s F-12 in 100-mm plates for 24 h and treated as indicated. Cells were rinsed with PBS at room temperature and harvested in 0.6 ml of R1PA buffer (1× PBS, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, and 100 μg/ml phenylmethylsulfonyl fluoride). Cells were then transferred to a fresh tube using a syringe with a 21-gauge needle and incubated with 10 μl of 10 mg/ml phenylmethylsulfonyl fluoride for 30–60 min on ice. Supernatant was collected as WCL after microcentrifuging at 10,000 × g for 10 min at 4°C. WCL containing 1 mg of protein was aliquoted and made to 1 ml with RIPA buffer. Each aliquot of WCL was precleared by incubation with 20 μl of protein A-agarose beads at 4°C for 30 min with shaking. Beads were pelleted by centrifuging at 1,000 × g for 5 min at 4°C, and supernatant (cell lysate) was collected. To 1 mg of the precleared WCL, 2 μg of mouse CD1 or ERα antibody or normal mouse IgG were added. The reaction mixture was incubated at 4°C for 1 h, mixed with 30 μl of resuspended protein A-agarose beads, and incubated for 10 h on a rocker platform at 4°C. The pellet was collected by centrifugation at 1,000 × g for 5 min at 4°C and washed alternately with RIPA (2×) and PBS buffer (2×). After the final wash, the pellet was resuspended in 50 μl of 1× Western sampling buffer, and proteins were separated by 10% SDS-PAGE electrophoresis. Western blot analysis was performed as described above.

Immunocytochemistry.

MCF-7 cells (5 × 10⁴) were seeded in 5% FBS and DMEM:Ham’s F-12 on Lab-Tek Chamber slides (Nalge Nunc International) and treated with PGJ2, ciglitazone, or DMSO for various time intervals. Slides were then washed with PBS, fixed in −20°C methanol for 10 min, air dried, and washed in 0.3% Tween/PBS (use 0.3% Tween in 20 nm PBS for subsequent washing steps) for 5 min. Slides were blocked with 10% serum (the species used for secondary antibody) in 1× antibody dilution buffer (1% BSA in 0.3% Tween/PBS) for 1 h and incubated with 1:500 cyclin D1 or ERα antibody (Santa Cruz Biotechnology) overnight at 4°C in a humid chamber to prevent evaporation of the antibody solution. After washing in 0.3% Tween/PBS for 10 min (3×), slides were reprobed with secondary antibody for 2 h at room temperature in a dark humid chamber and then washed in 0.3% Tween/PBS for 10 min (4×) and rinsed in deionized water. Slides were stained with FITC fluorochrome, mounted with glycerol/phenylenediamine, and visualized. Samples without primary antibody were used as control.

Fluorescence-activated Cell-sorting Analysis.

MCF-7 cells were synchronized in serum-free media for 24 h and then treated with DMSO or different concentrations of the PPARγ agonists for 24 h. Trypsinized cells were then centrifuged and resuspended in staining solution containing 50 μg/ml PI, 4 mm sodium citrate, 30 units/ml RNase, and 0.1% Triton X-100 (pH 7.8). After incubation at 37°C for 10 min, sodium chloride was added to give a final concentration of 0.15 m, and cells were analyzed on a FACSCalibur flow cytometer (Becton Dickinson Immunocytometry Systems, San Jose, CA), using CellQuest (Becton Dickinson) acquisition software. PI fluorescence was collected through a 585/42 nm bandpass filter, and list mode data were acquired on a minimum of 12,000 single cells defined by a dot plot of PI width versus PI area. Data analysis was performed in ModFit LT (Verity Software House, Topsham, ME) using PI width versus PI area to exclude cell aggregates.

Statistical Analysis.

Statistical differences between different groups were determined by ANOVA and Scheffe’s test for significance. The data are presented as means ± SD for at least three separate determinations for each treatment.

RESULTS

Growth Inhibition and Transcription.

The growth-inhibitory effects of PPARγ agonists in MCF-7 cells were determined using PGJ2 and ciglitazone. The results showed that PGJ2 inhibited MCF-7 cell proliferation at a concentration as low as 5 μm (Fig. 1) ⇓ , whereas higher concentrations (30–40 μm) were required for ciglitazone (data not shown). Previous studies showed that PGJ2 decreased immunoreactive cyclin D1 mRNA and protein in MCF-7 cells, and this was accompanied by deceased phosphorylation of retinoblastoma protein and inhibition of G₁ to S-phase cell cycle progression (39) . Our results also show that PGJ2 and ciglitazone inhibit G₀-G₁→S-phase progression in MCF-7 cells, and this was not accompanied by apoptosis (Table 1) ⇓ . The relative potencies of PGJ2 and ciglitazone in MCF-7 cells were also determined in transactivation studies in cells transfected with a consensus PPRE construct (pPPRE₃-luc) or cotransfected with a GAL4-DNA binding domain-PPARγ fusion protein and a reporter construct containing five tandem GAL response elements linked to a luciferase reporter gene. Results of both assays gave comparable results (Fig. 1B) ⇓ showing that PGJ2 was a more potent ligand than ciglitazone in transactivation experiments, and this complements effects of PGJ2 and ciglitazone as inhibitors of MCF-7 cell proliferation.

PPARγ Agonists Induce Down-Regulation of Cyclin D1 and ERα in MCF-7 Cells.

Because PGJ2 and ciglitazone inhibit growth of MCF-7 cells and G₁→S-phase progression, we also investigated the effects of both PPARγ agonists on cyclin D1 protein, which plays a critical role in progression through G₁ to S-phase. The results in Fig. 2A ⇓ demonstrate that 30 or 80 μm PGJ2 or ciglitazone, respectively, induce a time-dependent decrease in cyclin D1 protein within 3–15 h after treatment, and similar results were also observed for ERα protein. The results in Fig. 2B ⇓ also illustrate a concentration-dependent decrease of both cyclin D1 and ERα in MCF-7 cells treated with ciglitazone (40–100 μm) or PGJ2 (10–40 μm) for 12 h. PGJ2 was the more potent PPARγ agonist and induced down-regulation of cyclin D1 and ERα at concentrations as low as 10 μm, and this was further investigated in a separate experiment using lower concentrations of PGJ2 (Fig. 2C) ⇓ . In this experiment, PGJ2 induced down-regulation of cyclin D1 and ERα at concentrations between 5 and 10 μm, whereas levels of PPARγ were unchanged at these concentrations but were slightly decreased after treatment with higher concentrations of PGJ2 (20 μm) for 24 h. In contrast, Sp1 protein was relatively unchanged during these experiments and is used as a loading control for quantitating relative protein levels.

Effects of Proteasome and Protease Inhibitors on Degradation of Cyclin D1 and ERα by PPARγ Agonists.

Previous reports indicate that both antiestrogens and AhR agonists that inhibit hormone-induced growth of MCF-7 cells also induce proteasome-dependent degradation of ERα (43, 44, 45) . Retinoids also inhibit growth of MCF-7 and bronchial epithelial cells (46, 47, 48) and induce down-regulation of cyclin D1 through the proteasome pathway in the latter cells. Therefore, in initial studies, we investigated effects of the proteasome inhibitor MG132 (10 μm) and the protease inhibitor CII (10 μm) on PPARγ agonist-induced down-regulation of cyclin D1 and ERα in MCF-7 cells. In cells treated with solvent control (DMSO), MG132 slightly increased cyclin D1 but not ERα protein levels, whereas CII was inactive. Cotreatment with PPARγ agonists and MG132 or CII showed that MG132 but not CII blocked PGJ2- and ciglitazone-induced down-regulation of cyclin D1 and ERα proteins, suggesting that activation of proteasomes plays an important role in these responses. We further investigated the effects of proteasome and protease inhibitors by quantitatively determining ERα and cyclin D1 protein levels after treating MCF-7 cells with 30 μm PGJ2 for 8 h These experiments were carried out in triplicate, and results are expressed as means ± SD for levels of ERα and cyclin D1 proteins in each lane. The results illustrated in Fig. 3B ⇓ show that PGJ2-induced down-regulation of cyclin D1 and ERα protein levels was inhibited after cotreatment with 1–10 μm MG132 and PSII, another proteasome inhibitor. Levels of Sp1, PPARγ, and cdk4 were essentially unaffected by protease or proteasome inhibitors or PGJ2. Although PGJ2 decreased PPARγ levels after treatment for 24 h (Fig. 2C) ⇓ , levels of the protein were unchanged after 8 h (Fig. 3B) ⇓ , suggesting that the temporal patterns for degradation of PPARγ (8–24 h) are different from those observed for cyclin D1 and ERα. Parallel studies show that two kinase inhibitors, CII and calpeptin, did not affect PGJ2-induced degradation of cyclin D1 or ERα at inhibitor concentrations as high as 50 μm (Fig. 3C) ⇓ . These data (Figs. 2 ⇓ and 3 ⇓ ) demonstrate that PGJ2 activation of PPARγ results in coordinate proteasome-dependent degradations of ERα and cyclin D1 within 3–8 h after treatment, whereas PPARγ was only degraded after prolonged treatment. In contrast, Sp1 and cdk4 proteins are unaffected by PGJ2 or proteasome or protease inhibitors.

Studies illustrated in Figs. 2 ⇓ and 3 ⇓ were obtained by Western blot analyses of whole cell extracts, and we therefore examined the effects of PGJ2 on cyclin D1 (Fig. 4A) ⇓ and ERα (Fig. 4B) ⇓ by immunocytochemical analysis. Treatment with PGJ2 (30 μm) for 10 h decreased ERα and cyclin D1 protein staining, and this was blocked by cotreatment with MG132 but not CII. The inhibitors alone had minimal effects. These data complement results obtained for whole cell extracts and demonstrate PPARγ-induced intracellular degradation of both cyclin D1 and ERα proteins, which are primarily localized in the nucleus of MCF-7 cells.

PGJ2 Enhances Ubiquitination of Cyclin D1.

Proteasome-dependent degradation pathways involve multiple steps including initial ubiquitination of proteins targeted for proteolysis (49, 50, 51, 52) . We therefore investigated the effects of DMSO (solvent control) or 30 μm PGJ2 on formation of ubiquitinated cyclin D1 and ERα (Fig. 5) ⇓ . After treatment of MCF-7 cells for 4 h, whole cells extracts were immunoprecipitated with mouse cyclin D1 IgG2a antibodies or normal mouse IgG, and immunoprecipitated protein and input were analyzed by SDS gel electrophoresis and immunostaining with cyclin D1 or ubiquitin antibodies (Fig. 5A) ⇓ . Cellular protein immunoprecipitated with cyclin D1 antibodies gave a 3–4-fold increase in ubiquitinated bands after treatment with PGJ2, and lower intensity ubiquitinated bands were observed in the solvent (DMSO) control. Western blot analysis of the input band with cyclin D1 antibodies also shows decreased intensity of the cyclin D1 band after treatment with PGJ2. Analysis for ubiquitinated ERα was also determined using WCLs from cells treated with PGJ2 for 2 or 4 h. Western blot analysis of the input shows the ERα band that had not significantly decreased (Fig. 5B) ⇓ ; however, using ubiquitinated antibodies, there was approximately a 2-fold increase in intensity of the ubiquitinated bands in the 4-h treatment group. The increased ubiquitination of cyclin D1 and ERα was observed in replicate studies. Immunoprecipitation of WCLs with mouse IgG did not pull down ERα, cyclin D1, or their corresponding ubiquitinated bands, and in a separate study, we also showed that retinoic acid induced a similar pattern of ubiquitinated cyclin D1 bands (data not shown). A complex ladder of ubiquitinated bands was observed in whole cell extracts (input), and IgG heavy chain and nonspecific bands (obtained with beads alone) were observed as control bands in both immunoprecipitated extracts. These data suggest that PPARγ-induced degradation of cyclin D1 and ERα is accompanied by enhanced ubiquitination, which is consistent with activation of the proteasome pathway.

DISCUSSION

PGJ2 and other PPARγ agonists inhibit growth of breast and other cancer cell lines, and in most cancer cells, these effects are linked to apoptosis or inhibition of G₁→S-phase progression. Our results (Fig. 1) ⇓ confirm the growth-inhibitory effects of both PGJ2 and ciglitazone in MCF-7 cells, and this is accompanied by decreased G₁→S-phase cell cycle progression (Table 1) ⇓ . Wang et al. (39) previously reported that PGJ2 also decreased cyclin D1 protein and mRNA levels in MCF-7 cells, and their studies indicated that transcriptional inhibition was due to competition between PPARγ and c-fos (bound to the cyclin D1 promoter) for limiting cellular levels of p300, an important coregulatory protein. In this study, we also observed that PGJ2 and ciglitazone decreased cyclin D1 protein in MCF-7 cells (Fig. 2) ⇓ , and this response was observed at concentrations of PGJ2 as low as 10 μm. Moreover, we also observed a similar pattern of ERα degradation induced by both PPARγ agonists, and 5–10 μm PGJ2 significantly decreased levels of ERα protein (Fig. 2C) ⇓ , and these responses were observed as early as 3 h after treatment. Thus, pharmacological concentrations (≥5 μm) of PGJ2 were required for these studies. We also observed some degradation of PPARγ at longer time points (Fig. 2C) ⇓ , whereas Sp1 protein is relatively unchanged by PPARγ agonists, as reported previously in other studies using AhR agonists and antiestrogens that also induce degradation of ERα (43 , 44) .

Cyclin D1 plays a critical role in G₀/G₁→S-phase cell cycle progression, and, not surprisingly, there are multiple cell-specific transcriptional and posttranscriptional mechanisms for regulation of cyclin D1 after mitogenic stimuli or after treatment with growth inhibitors (39 , 48 , 53 , 54) . As noted above, PPARγ agonists repress transcription of cyclin D1 by sequestering p300 (39) , and cyclin D1 transcription is also down-regulated in MCF-7 cells by flavopiridol, an inhibitor of cell cycle progression (53) . Serum starvation of NIH3T3 cells results in Ca²⁺-dependent calpain protease-induced degradation of cyclin D1, whereas in the same cells, cyclin B1 is degraded through the proteasome pathway (54) . In other cell lines, cyclin D1 is degraded by proteasomes, and retinoic acid induces proteasome-dependent degradation of cyclin D1 in immortalized human bronchial BEAS-2B epithelial cells (48) . Coordinate down-regulation of both cyclin D1 and ERα by PPARγ or other growth-inhibitory agents has not been reported previously; however, estrogens, some antiestrogens, and AhR agonists induce rapid proteasome-dependent degradation of ERα in breast cancer cells (43, 44, 45) . Therefore, based on the observation that both ciglitazone and PGJ2 induced coordinate degradation of ERα and cyclin D1 in MCF-7 cells (Fig. 2) ⇓ , we further investigated these responses in the presence of proteasome (MG132 and PSII) and protease (CII and calpeptin) inhibitors (Fig. 3) ⇓ . The results clearly demonstrate that over a range of concentrations, proteasome inhibitors block PPARγ-induced down-regulation of both cyclin D1 and ERα, whereas protease inhibitors did not block this response. Moreover, the responses determined by Western blot analysis of whole cell extracts were also observed directly by immunocytochemical analysis of cyclin D1 and ERα proteins after treatment with PGJ2 alone or in combination with MG132 or CII (Fig. 4) ⇓ . The proteasome inhibitor MG132 or the protease inhibitor CII has minimal effects on levels of cyclin D1 protein in MCF-7 cells; however, MG132 (but not CII) inhibited PGJ2-induced degradation of both proteins, which are primarily located in the nucleus of MCF-7 cells.

Because proteasome-dependent degradation of proteins involves prior conjugation of targeted proteins by ubiquitin or ubiquitin-like molecules, we also investigated formation of ERα- and cyclin D1-ubiquitin complexes in MCF-7 cells treated with PGJ2 (Fig. 5) ⇓ . Retinoids also inhibit the growth of breast cancer cells and down-regulate ERα (46, 47, 48) and cyclin D1 protein (55) in MCF-7 cells. Moreover, retinoic acid induced proteasome-dependent degradation of cyclin D1 in immortalized bronchial epithelial cells (48) , and these responses are similar to those induced by PGJ2 in this study. Results illustrated in Fig. 5 ⇓ identify ubiquitinated cyclin D1 and ERα bands that are enhanced after treatment of MCF-7 cells with PGJ2, and this was consistent with PPARγ-mediated proteasome-dependent degradation of cyclin D1 and ERα in MCF-7 cells. Retinoic acid also induced a pattern of cyclin D1-ubiquitinated bands similar to that observed for PGJ2 (data not shown).

In summary, our studies demonstrate that PPARγ agonists that inhibit growth and cell cycle progression (G₀-G₁→S-phase) in MCF-7 cells also induce proteasome-dependent degradation of both cyclin D1 and ERα. This mechanism complements results of previous studies showing that PPARγ also inhibits cyclin D1 transcription (39) , and the combination of both transcriptional and posttranscriptional inhibition of cyclin D1 by PPARγ agonists may contribute to the efficacy of these compounds as inhibitors of breast cancer cell proliferation. Moreover, in ongoing studies, our results indicate that PPARγ agonists inhibit growth and induce proteasome-dependent degradation of cyclin D1 in multiple cancer cell lines, and this may represent an important PPARγ-induced tumor growth-inhibitory pathway.