Background
Breast cancer is the most common malignancy and the second leading cause of cancer related death among American women [
1]. Despite of the fact that recent research efforts have significantly improved the outcome of breast cancer, the complexity and heterogeneity of this disease still urges the necessity to explore new and more specific drug targets. Peroxisome proliferator-activated receptor gamma (PPARγ), a member of the nuclear-hormone receptor family, has shown potential as a therapeutic target for prevention and treatment of breast cancer. PPARγ is a ligand-activated transcription factor. There are two isoforms of PPARγ protein, PPARγ1 and PPARγ2, the latter of which has the addition of 30 N'-terminal amino acids as a result of the usage of a different promoter and alternative splicing [
2]. PPARγ plays an important role in adipocyte differentiation, insulin sensitivity, energy metabolism, immune response, and the development of the nervous system [
3‐
5]. It is predominantly expressed in adipose tissues; although, it is also detected in various tissues such as cardiac and skeletal muscle, intestine, vascular smooth muscle, lung, breast, colon, and prostate [
6,
7]. Some polyunsaturated fatty acids [
8‐
10] and arachidonic acid metabolites [
11] are considered to be the natural ligands of PPARγ. Synthetic ligands of PPARγ include the thiazolidinedione class of anti-diabetic drugs (TZDs) such as rosiglitazone, pioglitazone, troglitazone [
12,
13], some non-steroidal anti-inflammatory drugs (NSAID) [
14], and non-thiazolidinedione tyrosine [
15]. In addition, a ligand-independent mechanism of PPARγ activation has also been observed due to altered phosphorylation status of the receptor [
16].
Recently, PPARγ has emerged as a promising target for cancer therapy based on the fact that its activation by synthetic ligands such as TZDs have been shown to induce cell cycle arrest, apoptosis and differentiation in many human malignancies [
17,
18]. Several studies have demonstrated that PPARγ activation by agonists can promote growth inhibition and apoptosis in both primary and metastatic breast malignancies [
19‐
22]. In addition to the anti-proliferative and pro-apoptotic effects, PPARγ ligands have also been reported to inhibit invasion and metastasis of human breast cancer cells [
23,
24]. However, these results were questioned by several studies that demonstrated the ability of PPARγ ligands to elicit anti-tumor effects via PPARγ-independent pathways and in the absence of PPARγ receptors [
25,
26]. Moreover, there is a debate that the concentrations of PPARγ ligands used in many studies are above the saturation level of the receptor. In fact, Roziglitazone, a widely studied PPARγ agonist, has shown to induce opposing effects when used in low versus high doses [
27]. Furthermore, PPARγ antagonists have also shown anticancer effects in a wide range of epithelial cancer cell lines, usually with greater potency than agonists [
28].
Existing data from
in vivo studies is also controversial. Recent animal studies have demonstrated that PPARγ agonists can prevent mammary carcinogenesis and reduce the development of tumors in nude mice [
29]. In contrast, another study has demonstrated an increase in the number of tumors when PPARγ ligand was administered [
30]. To clarify the controversy arising from the use of pharmacological approaches, several animal studies utilized techniques that allowed evaluation of the consequences of PPARγ transactivation in breast cancer independent of exogenous stimulation. Studies which employed a genetic approach to explore the intrinsic role of PPARγ signaling have demonstrated that an increase in PPARγ signaling accelerates mammary gland tumor development and constitutive over-expression of PPARγ increases incidence of breast cancer in mice already susceptible to the disease [
31]. This group has also shown that mice heterozygous for a null PPARγ mutation develop tumors with the same kinetics as those that carry two functional copies [
31]. Furthermore, the ablation of PPARγ expression in the mouse mammary gland using a Cre- Lox recombination system has demonstrated that no tumors developed in mammary glands lacking PPARγ suggesting that PPARγ is not a tumor suppressor [
32]. In summary, these observations suggest that reduced PPARγ expression does not contribute to the initiation of breast cancer; however, acceleration of PPARγ signaling after tumor initiation markedly promotes breast cancer development.
In this study, we have begun to elucidate the functional significance of endogenous PPARγ activation in breast cancer using an
in vitro model. We have previously reported that PPARγ1, not PPARγ2, is expressed in normal mammary epithelial cells and breast cancer cell lines [
33]. Our lab and others have also demonstrated that the level of PPARγ1 expression is significantly higher in breast cancer cell lines as compared to normal epithelial cells [
33‐
36]. In addition, we have shown that a distinct promoter regulates PPARγ1 expression in MCF-7 cells and that
promoter switching mediates differential PPARγ1 expression levels between normal and cancer cells [
33]. The Myc-associated zinc finger protein (MAZ) has been identified as a transcriptional mediator of PPARγ1 in MCF-7 cells [
37]. MAZ is a transcriptional factor that controls the expression of various genes through interactions between GC-rich DNA binding sites within the promoter sequence of target genes and the carboxyl-terminal zinc finger motifs of MAZ [
38]. Here, we demonstrated that an increase in expression and endogenous transactivation of PPARγ1 in MCF-7 breast cancer cells enhances cell proliferation by accelerating cell transition from G
1 to the S phase. This data was confirmed using a dominant-negative PPARγ1 mutant as an alternative approach to inhibit endogenous activity of PPARγ1 in two different cell lines, MCF-7 and T47D. We also found that in the absence of exogenous stimulation high expression of PPARγ1 significantly inhibits apoptosis in MCF-7 cells.
Discussion
In this report, we confirmed that PPARγ1 is highly expressed in cultured breast cancer cell lines as compared to HMEC [
37,
42]. High expression of PPARγ1 has also been reported in human breast cancer tissues [
43]. However, questions about the mechanism and role of endogenous transactivation of PPARγ1 during development of breast cancer still remain unanswered. We have previously shown that the increase in expression of PPARγ1 from normal human mammary epithelia to breast cancer is due to the recruitment of a distal, tumor-specific promoter [
34]. We identified MAZ as a transcription factor that directly binds to this promoter and drives expression of PPARγ1 in MCF-7 breast cancer cells [
37]. Our results also indicated that MAZ is highly expressed in MCF-7 cells as compared to HMEC [
36]. In this study, statistical analysis of three different Western Blots demonstrated an increase in PPARγ1 expression in a panel of different breast cancer lines and confirmed that it is a feature attributed not only to MCF-7 cells but also to other tested breast cancer cell lines. This observation suggests that the proposed model of endogenous PPARγ1 transactivation may apply not only to a particular cell line, but also to breast cancer in general. Currently this hypothesis is being tested in the lab using pathological sections from normal and breast cancer specimens.
In efforts to explore the role of PPARγ activation in cancer, most of the recent studies employed pharmacological approaches. The anti-cancer activity of PPARγ ligands, such as TZDs, demonstrated in multiple
in vitro studies, has raised discussion about the possibility of using PPARγ receptors as a target for breast cancer therapy. However, the "off target" effects of PPARγ agonists [
44,
45], the dual role that some ligands play when they are applied to the cells at different concentrations [
46], and the paradoxical anti-cancer effect of PPARγ antagonists [
47] necessitated the use of other approaches to evaluate the consequences of PPARγ transactivation in cancer. For the first time, using an
in vitro model, we have addressed questions about the role that endogenous transactivation of PPARγ1 plays in the pathogenesis of breast cancer. Using RNAi techniques to inhibit PPARγ1 expression we demonstrated that an increase in PPARγ1 signaling can significantly affect proliferation and apoptosis in breast cancer cells. It is widely accepted that the dysfunctional balance between cellular proliferation and apoptosis can contribute to the initiation and progression of cancer. Here, we demonstrated that down-regulation of PPARγ1, directly or indirectly via knock-down of its transcriptional regulator MAZ, leads to a decrease in cellular proliferation in MCF-7 breast cancer cells. Interestingly, changes in cellular proliferation caused by direct PPARγ1 inhibition by PPARγ shRNA were analogous to changes in PPARγ1 expression when inhibited via down-regulation of MAZ. This suggests that in addition to its role as a mediator of tumor-specific expression of PPARγ1, MAZ may also be involved in the regulation of other growth control genes in MCF-7 breast cancer cells. The ongoing project in our lab is to further investigate the role of MAZ in breast cancer development.
The observed pro-survival effect of PPARγ1 signaling in MCF-7 cancer cells was also confirmed by using a different approach to inhibit endogenous activity of PPARγ1. We took advantage of a PPARγ1 mutant, Δ462, which lacks helix12, critical for ligand binding and co-activators recruitment. Thus, Δ462 functions in a dominant-negative manner. Data from BrdU proliferation assays demonstrated that inhibition of PPARγ1 activity using Δ462 decreases cell proliferation not only in MCF-7 cells but in another widely studied breast cancer cell line, T47D. In our previous study, we have shown that the T47D cell line also has a functional peroxisomal response [
42]. Here, using Western blot analysis, we demonstrated that T47D cancer cells as well as MCF-7cells have high level of PPARγ1 expression as compared to HMEC. However, the direct comparison of PPARγ1 expression in MCF-7 and T47D cells showed the lower level of PPARγ1 in a latter cell line. The differential expression of PPARγ1 in these cell lines can explain the more prominent changes in cellular proliferation in MCF-7 cells compare to T47D cells when Δ462 are applied to the cells. The specificity of Δ462 in the inhibition of endogenous PPARγ1 activity was confirmed using Luciferase assay. We measured PPRE-mediated reporter activity when either MCF-7 or T47D cells were transfected with Δ462 or control plasmids and then treated with10 μM Rosi. Data revealed that PPRE reporter activity is significantly lower in cells transfected with the Δ462 expression plasmid compared to control, thus, providing the evidence that this mutant acts in dominant-negative manner, decreasing activity of PPARγ1 in MCF-7 and T47D cancer cells. Moreover, a similar effect was observed with Rosi, confirming the specificity of Δ462 action in these cell lines. In summary, these results suggest that PPARγ1 transactivation enhances cell growth in breast cancer cells and that this phenomenon is not specific to MCF-7 cells.
To further investigate the mechanism by which PPARγ1 regulates cell growth, we performed fluorescence-activated cell sorting (FACS). Cell cycle distribution analysis confirmed results from the BrdU proliferation assay and demonstrated that an increase in PPARγ1 signaling accelerates the transition of cells from G1-phase to S-phase and, thus, increases cellular proliferation.
Blockage of apoptosis is a likely requirement for cancer maintenance [
48]. In fact, FACS analysis revealed that the number of cells which undergo apoptosis is much higher in MCF-7 cells with decreased PPARγ1 expression. This observation was tested and confirmed by measuring DNA fragmentation in control and PPARγ or MAZ shRNA transfected cells. The data demonstrated that down-regulation of PPARγ1 expression in MCF-7 cells leads to a significant increase in apoptosis. The induction of apoptosis in cells with PPARγ1 or MAZ knockdown was also confirmed by analyzing the expression of Bcl2, a protein that is known to block cell death [
40], and by evaluation of PARP-1 cleavage, a widely accepted marker for apoptosis. The results showed that inhibition of PPARγ1 leads to down-regulation of Bcl2 which may in turn favor re-activation of signaling pathways to induce apoptosis. The increase in PARP-1 cleavage in MCF-7 cells, which have a decreased level of PPARγ1 expression, verified the induction of apoptosis as well. Together these results suggest that accelerated PPARγ1 signaling can interfere with apoptotic pathways and promote cancer cell survival during breast tumor development. However, the molecular mechanisms that drive these events are not known and will be the subject of future investigation in the lab.
In summary, this study demonstrates that the increase in PPARγ1 expression observed in breast cancer results in an increase in PPARγ1 signaling that in turn promotes proliferation and inhibits apoptosis and thus, may significantly contribute to the progression of disease to a more malignant stage. Our findings are consistent with results from a study that evaluated the consequences of intrinsic PPARγ1 activation using transgenic mice. This study demonstrated that constitutive over-expression of PPARγ1 in mice, which were predisposed to breast cancer, leads to a greater number of tumors and higher mortality in both male and female animals, thus suggesting that increased PPARγ1 signaling serves as a tumor promoter in the mammary gland [
31]. Since constitutive PPARγ1 signaling did not affect mammary gland differentiation or function when introduced in wild-type mice, the authors emphasized that consequences of PPARγ1 transactivation are different in normal and transformed cells. This observation is consistent with our previous data, which demonstrated different mechanisms of transcriptional regulation of PPARγ1 in breast cancer cells as compared to HMEC [
34,
37].
Methods
Cell culture
MCF-7 and T47D breast cancer cells were obtained from the American Type Culture Collection (Rockville, MD). Cells were cultured in modified DMEM (Gibco BRL, Gaithersburg, MD) supplemented with 10% fetal bovine serum (Hyclone). Normal human mammary epithelial cells (HMEC) (Cambrex) were cultured in MEGM® with SingleQuot® supplements. All cell lines were grown in media lacking phenol red at 37°C in a 5% CO2 atmosphere.
Western blot analysis
The whole cell lysates were prepared using passive lysis buffer. Concentrations were determined using a Bradford Assay (BioRad). 30 μg of total cell lysate per sample was run on a 10% or 12% SDS polyacrylamide gel. The proteins were transferred to a nitrocellulose membrane, blocked in 5% TBST, and incubated at 4°C overnight with primary antibody. The membrane was then washed and incubated for 4 hours with secondary IgG-HRP antibody. After incubation the membrane was washed, incubated with Chemiluminescence substrate (Pierce) for 5 min, and expose to film. The following primary antibodies were used: PPARγ mouse monoclonal IgG antibody 1:200 dilution (Santa Cruz Biotechnology, sc-7273), Bcl2 mouse monoclonal antibody 1:1000 dilution (Santa Cruz Biotechnology, sc-509), PARP-1 rabbit polyclonal antibody 1:1000 dilution (Santa Cruz Biotechnology, sc-2578). Appropriate secondary goat anti-mouse (Santa Cruz Biotechnology, sc-2055), or bovine anti-goat (Santa Cruz Biotechnology, sc-2378), or goat anti-rabbit secondary antibody 1:1000 dilution (Santa Cruz Biotechnology, sc-2004) antibodies 1:1000 dilution were applied. Anti-actin raised in rabbit 1:2000 dilution (Sigma, A 5060) and goat anti-rabbit secondary antibody 1:1000 dilution (Santa Cruz Biotechnology, sc-2004) were used to visualize actin. Western Blot Stripping Buffer (Pierce, # 21059) was used to restore membranes.
shRNAs constructs
The set of five shRNAs for PPARγ1 (TRCN 0000001670-74) and MAZ (TRCN 0000015343-47) genes as well as scrambled shRNA and non-hairpin TRC controls were purchased from The RNAi consortium (TRC) Human shRNA Library (Open Biosystems). The shRNA construct includes a hairpin of 21 base pairs, a sense and antisense stem, and a 6 base-pair loop. Each hairpin sequence is cloned into a lentoviral vector (pLKo1). Based on structural evaluation and Western blot analysis the most efficient shRNAs for PPARγ1 (TRCN 0000001672) and MAZ (TRCN 0000015345) were chosen for transient transfections.
Dominant-negative PPARγ1 construct
The dominant-negative PPARγ1 mutant was a kind gift of Dr. Stephen O'Rahilly and Dr. V. Krishna K Chatterjee, Department of Medicine, Addenbrook's Hospital, Cambridge, U.K. Sequence analysis revealed a single base deletion introducing a premature stop codon (5'-1380 GACAGACTGA1390-3') leading to translation of protein truncated just before the AF-2 domain.
Transfection assays
Cells were transiently transfected with 3.6 μg of pGL3 plasmid containing 3XPPRE-mTK-Luc and Renilla (Allred, 2005) per 24-well plate and then co-transfected with scrambled, PPARγ or MAZ shRNAs, Δ462, or control plasmids using FuGENE transfection reagent (Roche). 4 hours after transfection cells were subsequently treated with 10 μM Rosi for 20 hours. Cells were lysed in 80 μl of passive lysis buffer and treated according to manufacture's instructions (Promega dual luciferase assay kit). Luminometry was performed on a Berthold Technologies Lumat 9507 (Wildbad, Germany). Results were calculated as raw Luciferase units divided by raw Renilla units (RLU's). Data is presented as mean fold changes in treated cells as compared to control cells.
Real-time PCR
Total mRNA was isolated using an RNeasy Mini Kit (Qiagen, CA) according to manufactures instructions. Real-time PCR was performed on total RNA using the TaqMan One-Step RT-PCR Master Mix Reagents Kit (Applied Biosystems). The pre-optimized primers and probes for MAZ, PPARγ1, and 18S were purchased from Applied Biosystems.
BrdU proliferation assay
MCF-7 cells were seeded at 0.1 × 104 cells/well in 96-well tissue culture plates. Cells were transiently transfected on the second and third day using 0.05 μg of plasmid and 0.3 μl of FuGENE 6 transfection reagent (Roche) per well. Control MCF-7 cells were treated with FuGENE6 only. 16 wells per each shRNA and control were used. The same experimental set-up was used when cells were transfected with a dominant-negative form of PPARγ1, Δ462. The media was changed before the second transfection. Cell Proliferation ELISA, BrdU (colometric) (Roche) was performed on the fifth day according to the manufacture instructions.
Apoptosis assay
The same protocol as for the proliferation assay was used to plate and transfect MCF-7 cells. A Cell Death Detection ELISA (Roche) was performed on the fifth day. The assay is based on quantitative sandwich-enzyme-immunoassay-principle and uses mouse monoclonal antibodies directed against DNA and histones. Cells were lysed in a 96-well plate, centrifuged, and 20 μl of supernatant was transferred into streptavidin-coated wells. A mixture of antibodies was added and the plate was then incubated for 2 hours. The unbound components were removed by washing. ABTS substrate was added and the amount of mono- and oligonucleosomes were measured photometrically using the ELISA-plate reader according manufacture instructions (Roche).
Cell cycle analysis
The DNA content of control and shRNAs transfected MCF-7 cells was analyzed using a detergent-trypsin method (Vindelov, 1983). MCF-7 cells were seeded at 1 × 106 cells in 100 mm culture plates. Cells were transiently transfected on the second and third day with 6 μg of plasmid using 18 μl of FuGENE6 transfection reagent (Roche). On the fifth day the propidium iodide labeling procedure and fluorescence-activated cell sorting (FACS) using Mod FitLT V.3.1 software was performed (University of Kentucky Flow Cytometry Facility).
Statistics
Data was analyzed by a two-way analysis of variance (ANOVA) using the StatServer 6.1(Insightful, Seattle, WA) from the server maintained by the University of Kentucky's Department of Statistics. In every 2-way ANOVA, Tukey's pair-wise comparison test was used post-hoc. P-values of less than 0.05 were considered to be significant. One-way ANOVA with Fisher's LSD or Tukey's pair-wise comparison post-hoc test were also used where appropriate. When appropriate, Student's t-test was also used for data analysis on Microsoft Excel.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
All authors significantly contributed in the design of the study, data interpretation, and manuscript drafts. YYZ carried out all experiments. RCS performed all statistical analyses and figure design. NKW and XW assisted with Δ462 and PPRE-Luciferase experiments. MWK coordinated this study.