Breast cancer still represents one of the most relevant mortality factors in women throughout the world, despite the significant advancements in its early detection and therapeutic approaches. Some evidence has established dietary essential fats as potential regulators of breast tumour cell growth(Reference Apantaku1, Reference Hilakivi-Clarke, Olivo, Shajahan, Khan, Zhu, Zwart, Cho and Clarke2). It is known that n-3 and n-6 fatty acids, the two major families of essential fatty acids, play an important role in tumour proliferation, though their bio-clinical mechanisms on tumour cells are still not properly understood(Reference Hilakivi-Clarke, Olivo, Shajahan, Khan, Zhu, Zwart, Cho and Clarke2–Reference Hammamieh, Chakraborty, Miller, Waddy, Barmada, Das, Peel, Day and Jett4). n-6 Fatty acids were found to stimulate the growth and metastasis of human breast cancer cells(Reference Cohen5, Reference Lanson, Bougnoux, Besson, Lansac, Hubert, Couet and Le Floch6), while n-3 fatty acids appeared to exert a protective effect(Reference Wu, Harvey, Ruzmetov, Welch, Sech, Jackson, Stillwell, Zaloga and Siddiqui7). More recently, it has been reported that both n-3 and n-6 PUFA are able to reduce the growth of different human cancer cells, although to different extents(Reference Monjazeb, High, Connoy, Hart, Koumenis and Chilton8).
Literature data suggest that the intake of PUFA may modulate cell behaviour and growth by a variety of mechanisms, including modification of tumour cell membranes which, in turn, can affect cell signalling pathways, lipid peroxidation and oxidative stress(Reference Schonberg, Rudra, Noding, Skorpen, Bjerve and Krokan9), eicosanoid production, fatty acid metabolism(Reference Sampath and Ntambi10) and regulation of gene expression(Reference Narayanan, Narayanan and Reddy11).
Among n-6 PUFA, 20 : 4n-6 arachidonic acid (AA) is a relatively minor PUFA found in cell membrane glycerolipids(Reference Lands12), which has been reported to inhibit the growth of some breast cancer cell lines(Reference Hammamieh, Chakraborty, Miller, Waddy, Barmada, Das, Peel, Day and Jett4). In contrast to other more abundant unsaturated fatty acids (linoleic or linolenic acid), levels of unesterified AA are stringently controlled within mammalian cells, and the pathways of AA uptake, incorporation and remodelling are well documented(Reference Chilton, Fonteh, Surette, Triggiani and Winkler13). The intracellular levels of AA are regulated through distinct and non-overlapping mechanisms. In resting cells, low concentrations of AA are tightly maintained by basal levels of rapid catabolism, by membrane glycerolipid recycling(Reference Brash14) and by esterification(Reference Cao, Traer, Zimmerman, McIntyre and Prescott15). On the other hand, a stimulus-induced release of AA by cytosolic phospholipase 2α results in its rapid metabolism induced by cyclooxygenase-2, thereby limiting the intracellular AA pools(Reference Brash14).
Once released from the membrane, AA and its metabolites are important signals that regulate a wide variety of cellular functions(Reference Monjazeb, Clay, High and Chilton16). The range of biological processes in which AA and its metabolites participate is vast; they regulate the transcription of several families of genes including heat-shock-protein genes(Reference Jurivich, Sistonen, Sarge and Morimoto17), genes involved in cell-cycle control(Reference Razanamahefa, Prouff and Bardon18), inflammation(Reference Bordin, Priante, Musacchio, Giunco, Tibaldi, Clari and Baggio19), steroid biosynthesis(Reference Wang, Dyson, Mondillo, Patrignani, Pignataro and Stocco20) and proto-oncogenes(Reference Hughes-Fulford, Chen and Tjandrawinata21, Reference Bernard-Gallon, Vissac-Sabatier, Antoine-Vincent, Rio, Maurizis, Fustier and Bignon22).
There are several other mechanisms probably implied in AA effects, not to be considered mutually exclusive, such as mechanisms implied in AA-induced apoptosis. In this regard, it has been reported that AA-induced apoptosis is transcriptionally dependent and involves the regulation of the key families of genes involved in cell survival and apoptosis.
For example, an alternate downstream target of unesterified intracellular AA is ceramide signalling of apoptosis initiated via the TNF-α pathway(Reference Jayadev, Linardic and Hannun23), since cancer cells treated with exogenous AA at concentrations that induce apoptosis accumulate ceramide(Reference Pettus, Bielawska and Subramanian24). It has been reported that exogenous AA and inhibitors of AA metabolism that lead to the accumulation of unesterified AA are cytotoxic to the colon cancer cell line(Reference Monjazeb, High, Connoy, Hart, Koumenis and Chilton8). Of interest is the observation that AA induced the suppression of the growth of different cancer cells through a mechanism that involves lipid peroxidation and PPAR activation(Reference Trombetta, Maggiora, Martinasso, Cotogni, Canuto and Muzio25).
PPAR are transcription factors with a pivotal role in lipid metabolism and homeostasis(Reference Chinetti, Fruchart and Staels26); they function by forming heterodimers with the retinoid X receptor(Reference Kostadinova, Wahli and Michalik27, Reference Kota, Huang and Roufogalis28). There are three PPAR subtypes (α, β and γ) which bind to different ligands, among them fatty acids and their metabolites, so regulating the expression of genes involved in lipid transport and metabolism within the cell. In mammals, PPARα is most highly expressed in brown adipose tissue, followed by liver, kidney, and heart. Activation of rat liver PPARα provides an anti-apoptotic mechanism(Reference Roberts, Chevalier, Hasmall, James, Cosulich and Macdonald29). PPARβ is expressed in all tissues studied to date(Reference Burdick, Kim, Peraza, Gonzalez and Peters30), whereas PPARγ is highly expressed in adipose tissue, as well as in muscle, colon, and liver. The importance of PPAR activation in preventing epithelial carcinogenesis is evident in different tumour systems. Activation of PPARγ through troglitazone and other PPARγ activators causes the inhibition of proliferation and the induction of apoptosis, both in vitro and in vivo (Reference Li, Lee, Mok, Warner, Yim and Chen31). PPAR can mediate inflammation, and this effect is potentiated by elements from AA metabolic pathways. These include leukotriene B4, derived from the lipoxygenase pathway, and 15-deoxy-Δ12,14-prostaglandin J2, related to the cyclooxygenase pathway. 15-Deoxy-Δ12,14-prostaglandin J2 is a PPARγ activator that has been shown to counteract the effects of the pro-inflammatory cytokines TNFα, IL-1, and IL-6(Reference Bianchi, Moulin, Sebillaud, Koufany, Galteau, Netter, Terlain and Jouzeau32).
Eicosanoids and a range of fatty acids, such as the PUFA linoleic acid and AA, bind to PPAR at physiological concentrations and regulate gene transcription(Reference Hihi, Michalik and Wahli33). Furthermore, all three PPAR subtypes have been implicated in carcinogenesis(Reference Kersten, Desvergne and Wahli34) and it is known that PPAR signalling can influence cell survival and apoptosis(Reference Clay, Namen, Atsumi, Trimboli, Fonteh, High and Chilton35). The objective of the current study was to understand the potential mechanisms of action of AA modulation of breast cancer cell growth; for this purpose, we compared the influence of AA on proliferation, signal transduction and apoptosis in two breast cancer cell lines, the well differentiated oestrogen receptor (ER)α(+) MCF-7 cells and the poorly differentiated ERα( − ) MDA-MB-231 cells. In particular we focused our attention on the interaction between the AA and PPAR pathways, an interaction that could provide an opportunity to develop drug combinations that maximize growth arrest and apoptosis in breast cancer cells.
Materials and methods
Antibodies and chemicals
MDA-MB-231 and MCF-7 breast cancer cell lines were a kind gift from Professor Sebastiano Andò, Department of Pharmaco-Biology, Faculty of Pharmacy, University of Calabria, Cosenza, Italy. AA, PPARα antagonist (MK886), PPARγ antagonist (GW9662), culture media, mouse monoclonal antibody specific to β-actin and chemicals were purchased from Sigma Chemical (MO, USA). Cell culture plasticware was from TPP (Trasadingen, Switzerland). Rabbit polyclonal antibody specific for PPARα (sc-9000), PPARβ (sc-7197), PPARγ (sc-7196), Bak (sc-832), Bcl-2, caspase-3 (sc-7148), caspase-9p10 (sc-7885), caspase-8p20 (sc-7890) and for extracellular signal-regulated protein kinase (ERK)1 (sc-94), mouse monoclonal antibody specific for phospho-ERK1/2 (sc-7383), goat anti-rabbit (sc-2004) and anti-mouse (sc-2005) secondary antibodies were obtained from Santa Cruz Biotechnology (CA, USA). The enhanced chemiluminescence detection system was from Amersham Pharmacia Biotech (Uppsala, Sweden). The protein assay kit and iQ™ SYBR® Green SuperMix Bio-Rad were from Hercules (CA, USA). The RNeasy Mini Kit® was from QIAGEN (GmbH, Germany); the cDNA Archive kit was from Applied Biosystems (Foster City, CA, USA).
Cell culture
MDA-MB-231 and MCF-7 breast cancer cells were grown in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10 % foetal bovine serum, 100 U penicillin/ml, 100 μg streptomycin/ml and 25 μg amphotericin B/ml. Cells were cultured at 37°C in a humidified incubator with 5 % CO2 and 95 % air and regularly examined using an inverted microscope. For treatments, cells were seeded at a density of 3 × 104 cells/cm2 and cultured for 24 h to allow them to adhere to the substratum. The medium was then replaced with serum-free DMEM supplemented with 100 U penicillin/ml and 100 μg streptomycin/ml, 25 μg amphotericin B/ml, 2 mm glutamine, 1 % ITS (insulin–transferrin–sodium selenite), 1 % vitamin solution, 0·4 % serum bovine albumin (fatty acid free) and AA. AA was dissolved in foetal bovine serum and the concentration of foetal bovine serum was adjusted so it was the same in all experiments and the final concentration was no stronger than 0·01 % (v/v). Control groups received the same amount of foetal bovine serum. When GW9662 (PPARγ antagonist) and MK886 (PPARα antagonist) were used, they were added to the medium 1 h before AA treatment. GW9662 is an irreversible PPARγ antagonist, identified in a competition-binding assay against the human ligand-binding domain; it binds PPARγ and covalently modifies a cystein residue in the ligand binding site of PPARγ(Reference Leesnitzer, Parks and Bledsoe36). MK886 inhibits PPARα by a non-competitive mechanism; it prevents the conformational change necessary for PPARα ligand–receptor interaction(Reference Kehrer, Biswal, La, Thuillier, Datta, Fischer and Vanden Heuvel37).
Viability and growth rate determination
The exponentially growing cells were harvested with 0·25 % trypsin–0·02 % EDTA treatment and seeded in twelve-well culture plates. After overnight incubation to allow cell attachment, the medium was removed and replaced with fresh serum-free DMEM containing AA at a series of concentrations with or without 5 μm GW9662 or MK886. Viability and cell number were determined using the trypan blue (0·5 % in NaCl) exclusion assay. Briefly, treated cells were washed with PBS and trypsinized. Aliquots of cell suspension (100 μl) were incubated with the same volume of trypan blue for 5 min. Finally, cells were transferred to the Bürker chamber and counted by a light microscope. Dead cells were defined as those stained with the dye. Samples were measured in three replicates and each experiment was repeated at least three times.
Lactate dehydrogenase assay
Cells were seeded in twelve-well-culture plates and appropriately treated. After treatment, the cell supernatant was collected for the measurement of lactate dehydrogenase release. The lactate dehydrogenase activity was determined spectrophotometrically by an assay based on the oxidation of NADH and the rate of decrease in absorbance at 340 nm. The activity of lactate dehydrogenase was calculated as nanomoles of NADH consumed per ml per min. Samples were measured in three replicates and each experiment was repeated at least three times.
Analysis of nuclear morphology
Cells were plated on glass slides in twelve-well plates and appropriately treated. Changes in nuclear morphology were labelled by 4′,6-diamidino-2-phenylindole and examined by fluorescent microscopy. The cells were fixed with 95 % ice-cold ethanol for 5 min and incubated with 4′,6-diamidino-2-phenylindole (1 mg/ml in methanol) for 30 min at 37°C in the dark, then nuclear morphology was observed under a fluorescence microscope equipped with a UV light filter. Cells which exhibited reduced nuclear size, intense fluorescence, chromatin condensation and nuclear fragmentation were considered as apoptotic. Each experiment was repeated at least three times.
Protein extraction and Western blotting
Cells were seeded in 75 cm2 plates and then appropriately treated. Collected cells were suspended in lysis buffer containing 20 mm Tris(hydroxymethyl) amino methane (Tris)–HCl (pH 7·4), 150 mm NaCl, 5 mm EDTA, 0·1 mm phenylmethyl-sulfonyl fluoride, 0·05 % aprotinin, 0·1 % Igepal and then incubated for 30 min at 4°C. The suspension was centrifuged for 25 min at 12 000 rpm, and the supernatant from this centrifugation was saved as total extracts. Protein contents in supernatants were measured using a commercially available assay (Biorad) with bovine serum albumin as a standard.
Equal amounts of proteins (60 μg/well) were mixed with a solubilization buffer containing 250 mm Tris, pH 8·8, 4 % SDS, 16 % glycerol, 8 % 2-mercaptoethanol and 0·1 % bromophenol blue, and then fractionated by electrophoresis on SDS–PAGE. Proteins were transferred onto nitrocellulose for 2 h in a Biorad electroblotting device. Nitrocellulose matrices were blocked with 5 % milk in TBST (1 m Tris buffer, pH 7·4, 5 m NaCl, 0·1 % Tween-20) for 1 h at room temperature. For immunodetection, matrices were incubated overnight at 4°C with a primary antibody. The matrices were then detected by incubation for 1 h at room temperature with the corresponding horseradish peroxidase-conjugated secondary antibody. The immunoreactive bands were visualized using the enhanced chemiluminescence system. Band intensities were quantified by densitometry and the expression of proteins was reported as a proportion of β-actin or ERK1 protein expression to control for any discrepancies in gel loading. Fold change v. control values were calculated by normalizing densitometric values obtained from the various proteins with those obtained for β-actin or ERK1 (VersaDoc Imaging System 3000, Biorad). Each experiment was repeated at least three times.
Evaluation of PPARα expression by real-time PCR
Total RNA was extracted using RNeasy Mini Kit®. Real-time PCR was performed using single-stranded cDNA prepared from total RNA (1 μg) with a High Capacity cDNA Archive kit. Forward and reverse primers were designed using Beacon Designer software (Bio-Rad, Hercules, CA) (see Table 1).
Table 1 Primer design (FW, forward primer and RV, reverse primer) for evaluation of PPARα expression by real-time PCR
A PCR mixture (25 μl), containing cDNA template equivalent to 80 ng total RNA, 5 pmole each of the forward and reverse primers and 2 × iQ™ SYBR® Green SuperMix, were amplified using an iCycler PCR (Bio-Rad, Hercules, CA). Each sample was tested six times and the threshold cycle (Ct) values were the corresponding mean. The fold change was defined as the relative expression compared to that at time 0 (time of seeding cells), calculated as 2− ΔΔCt, where ΔCt = Ctsample − CtGAPDH and ΔΔCt = ΔCtsample − ΔCttime 0.
Statistical analysis
Differences between the means were analysed for significance using the one-way ANOVA test with the Bonferroni post hoc multiple comparisons used to assess the differences between independent groups. All values are expressed as means with standard deviation, and differences were considered significant at P < 0·05.
Results
Cell viability
To assess the effect exerted on human breast cancer cell proliferation, MCF-7 and MDA-MB-231 cancer cells were treated with increasing concentrations (1–50 μm) of AA. Cell viability was examined by the trypan blue assay, which provides a direct count of cells that exclude the trypan blue dye, including apoptotic cells. Twenty-four-hour AA treatment remarkably decreased the survival of both cell lines in a dose-dependent manner (Fig. 1), with a strong inhibitory effect against MDA-MB-231 cells (50 % inhibition with 50 μm AA). Thus, compared with the well-differentiated MCF-7 cells (Fig. 1 (A)), the poorly differentiated MDA-MB-231 cells (Fig. 1 (B)) appeared to be more susceptible to the AA growth inhibitory effect.
Fig. 1 Arachidonic acid (AA)-related growth inhibition of MCF-7 (A) and MDA-MB-231 (B) breast cancer cells. Cells were incubated with different concentrations of AA (1–50 μm) for 24 h and viability was determined by the trypan blue exclusion test. Values are presented as the percentage of control, with standard deviation shown by vertical bars, from three independent experiments conducted in triplicate. Mean values were significantly different from controls: *P < 0·05, **P < 0·01, ***P < 0·001.
The growth-inhibitory effect of AA is accompanied by a cytotoxic side effect detected as the cell release of lactate dehydrogenase in the culture media (Fig. 2). In MCF-7 cells, the release is significantly increased at all the concentrations of AA used (Fig. 2 (A)), while in MDA-MB-231 cells the release is significantly augmented only with high concentrations of AA, with a maximal effect at 50 μm dose (Fig. 2 (B)).
Fig. 2 Cytotoxic activity of arachidonic acid (AA). MCF-7 (A) and MDA-MB-231 (B) breast cancer cells were treated with different concentrations of AA (1–50 μM) for 24 h, then supernatants were collected for measurement of lactate dehydrogenase (LDH) amounts. LDH release is expressed as the percentages of nmoles of NADH consumed per ml per min. Values are means with standard deviation shown by vertical bars, from three independent experiments conducted in triplicate. Mean values were significantly different from controls: *P < 0·05, **P < 0·01.
The most effective condition in inhibiting cell survival in both cell lines was obtained after 24 h treatment with 50 μm AA, so we used this experimental protocol for further experiments.
Apoptosis induction
To assess whether the decrease of cell growth by AA treatment was due to the induction of apoptosis, we further evaluated the effects of AA on the expression of apoptosis-regulating proteins such as Bak, Bcl-2 (Fig. 3) and procaspase-3, -8 and -9 (Fig. 4).
Fig. 3 Effect of arachidonic acid (AA) on pro- and anti-apoptotic proteins. MCF-7 (A) and MDA-MB-231 (B) breast cancer cells were incubated with 1, 25 or 50 μm AA for 24 h, and total cell lysates were separated on 10 % SDS–PAGE gel, transferred to nitrocellulose membrane and probed with anti-bak (
) or anti-bcl-2 (
) antibody. Protein contents were normalized by probing the same membrane with anti-β-actin antibody. The densitometric values of the bands are shown as means with standard deviation indicated by vertical bars, from three independent experiments. Mean values were significantly different from controls (c): *P < 0·05, **P < 0·01.
Fig. 4 Effect of arachidonic acid (AA) on procaspase-8, -9 and -3. MCF-7 (A) and MDA-MB-231 (B) breast cancer cells were incubated with 1, 25 or 50 μm AA for 24 h, and total cell lysates were separated on 10 % SDS–PAGE gel, transferred to nitrocellulose membrane and probed with an anti-caspase-8p20 (), anti-caspase-9p10 () and anti-caspase-3 antibody (□). Protein contents were normalized by probing the same membrane with an anti-β-actin antibody. The densitometric values of the bands are shown as means with standard deviation indicated by vertical bars, from three independent experiments. Mean values were significantly different from controls (c): **P < 0·01, ***P < 0·001.
In MCF-7 cells the inhibition of cell growth did not correlate with apoptosis induction, as AA caused a decrease in the content of the pro-apoptotic protein Bak concomitantly to an increase of the anti-apoptotic protein Bcl-2 (Fig. 3 (A)). On the contrary, a pro-apoptotic effect was detected in MDA-MB-231 (Fig. 3 (B)), in which AA induced a significant increase in the expression of Bak without any effect on the anti-apoptotic protein Bcl-2, which was expressed at basal condition. In addition, in MCF-7 cells AA did not modify the levels of procaspase-8 and -9 (Fig. 4 (A)); in this cell line procaspase-3 was undetectable, since the caspase-3 gene is not expressed(Reference Yang, Stennicke, Wang, Green, Janicke, Srinivasan, Seth, Salvesen and Froelich38). In MDA-MB-231 (Fig. 4 (B)), AA induced a strong decrease of procaspase-3 and -9 levels, suggesting that the cleavage of the proteins to the active form (caspase-3 and -9) also occurred. The fact that in this cell line expression of procaspase-8 was not modified seems to exclude the activation of extrinsic pathways of apoptosis.
In MDA-MB-231 cells evidence of apoptosis was confirmed by the analysis of nuclear morphology by staining with specific DNA fluorochrome 4′,6-diamidino-2-phenylindole, which showed nuclear fragmentation and condensation above all after 24 h treatment with AA at 1 μm concentration (Fig. 5 (B)).
Fig. 5 Effect of arachidonic acid (AA) on nuclear morphology. MCF-7 (A) and MDA-MB-231 (B) breast cancer cells were treated with 1, 25 or 50 μm AA for 24 h. To detect nuclei the cells were stained with 4′,6-diamidino-2-phenylindole (1 mg/ml in methanol) and viewed under a fluorescence microscope equipped with a UV light filter.
Involvement of extracellular signal-regulated protein kinase–mitogen activated protein kinase pathway
In an attempt to explore the nature of the antiproliferative response exhibited by MCF-7 and MDA-MB-231 cells, we analyzed the AA effect on the ERK–mitogen-activated protein kinase signal transduction pathway, a large network of signalling molecules regulating cell growth and differentiation (Figs. 6 and 7).
Fig. 6 Effect of arachidonic acid (AA) on extracellular signal-regulated protein kinase (ERK)1/2 phosphorylation state in MCF-7 breast cancer cells. MCF-7 cells were incubated with 1, 25 or 50 μm AA for 24 h (A) or with 50 μm AA for 1, 4, 8 or 16 h (B). After treatment, total cell lysates were obtained and probed with an anti-phosphoERK1/2 antibody. Protein contents were normalized by probing the same membrane with an anti-ERK1 antibody. The densitometric values of the bands are shown as means with standard deviation indicated by vertical bars, from three independent experiments. Mean values were significantly different from controls (c): *P < 0·05, **P < 0·01.
Fig. 7 Effect of arachidonic acid (AA) on the extracellular signal-regulated protein kinase (ERK)1/2 phosphorylation state in MDA-MB-231 breast cancer cells. MDA-MB-231 cells were incubated with 1, 25 or 50 μm AA for 24 h (A) or with 25 μm AA for 1, 4, 8 or 16 h (B). After treatment, total cell lysates were obtained and probed with an anti-phosphoERK1/2 antibody. Protein contents were normalized by probing the same membrane with an anti-ERK1 antibody. The densitometric values of the bands are shown as means with standard deviation indicated by vertical bars, from three independent experiments. Mean values were significantly different from controls (c): *P < 0·05, **P < 0·01.
In MCF-7 cells 24 h treatment with AA caused a dose-dependent decrease of phospho-ERK expression (active form), with the highest magnitude of reduction at 50 μm concentration (Fig. 6 (A)).
The analysis of the short-time effect (1, 4, 8 and 16 h) of 50 μm AA on phospho-ERK level (Fig. 6 (B)) indicated a significant reduction of the ERK phosphorylation state starting from 8 h treatment. Also in MDA-MB-231 cells 24 h treatment with AA induced a strong decrease of phospho-ERK level, in this case the maximal downregulatory effect was observed at 25 μm concentration (Fig. 7 (A)). The short-time analysis of the AA effect in this case indicated a reduction of phospho-ERK from 8 h from the start of treatment (Fig. 7 (B)).
PPAR expression
Effects on PPAR expression were further examined to study the AA influence on mechanisms involved in cell growth regulation. We examined changes in the protein contents of different PPAR isoforms by western blot analysis. As shown in Fig. 8, 24 h treatment with AA dose-dependently enhanced PPARα expression in both cell lines. Also the PPARγ level was up-regulated by AA, which was more effective in MCF-7 cells. On the contrary, after AA treatment the PPARβ level did not display substantial changes in MCF-7 cells and increased only slightly in MDA-MB-231 cells.
Fig. 8 Effect of arachidonic acid (AA) on PPAR expression. MCF-7 (A) and MDA-MB-231 (B) breast cancer cells were treated with 1, 25 or 50 μm AA for 24 h and total cell lysates were probed with an anti-PPARα (), anti-PPARβ () and anti-PPARγ (□) antibody. Protein contents were normalized by probing the same membrane with an anti-β-actin antibody. The densitometric values of the bands are shown as means with standard deviation indicated by vertical bars, from three independent experiments. Mean values were significantly different from controls (c): *P < 0·05, **P < 0·01.
To evaluate whether AA could be directly responsible for the induction of PPARα and γ, we also analysed its effects in the presence of a specific PPARα (MK886) and PPARγ (GW9662) antagonist. Both antagonists were administered to MCF-7 and MDA-MB-231 cells at 5 μm concentration, chosen on the basis of the existing literature(Reference Kehrer, Biswal, La, Thuillier, Datta, Fischer and Vanden Heuvel37, Reference Cimini, Cristiano, Colafarina, Benedetti, Di Loreto, Festuccia, Amicarelli, Canuto and Ceru39). As shown in Fig. 9 ((A) and (B)), in both cell lines pre-treatment (1 h) with MK886 prevented an AA stimulatory effect on PPARα expression; on the contrary, GW9662 was not able to influence AA effects on PPARγ expression (data not shown).
Fig. 9 Effect of arachidonic acid (AA) on cell growth and PPARα expression in the presence of MK886 (MK). MCF-7 (A, C) and MDA-MB-231 (B, D) breast cancer cells were treated for 24 h with 50 μm AA, 5 μm antagonist alone or 5 μm antagonist 1 h before 50 μm AA. Total cell lysates were probed with an anti-PPARγ antibody and normalized by probing the same membrane with an anti-β-actin antibody. In (A) and (B) the densitometric values of the bands are shown as means with standard deviation indicated by vertical bars, from three independent experiments. Mean values were significantly different from controls (c): *P < 0·05, **P < 0·01. For MCF-7 (C) and MDA-MB-231 (D) cells were treated as described, and cell viability was determined by the trypan blue exclusion test. Values are presented as the percentage of control, as mean values with standard deviation shown by vertical bars, from three independent experiments conducted in triplicate. Mean values were significantly different from controls: **P < 0·01, ***P < 0·001.
MCF-7 (Fig. 10 (A)) and MDA-MB-231 (Fig. 10 (B)) cells were cultured with AA, with or without 1 h pre-incubation with MK886, for 24 h and analysed by real-time PCR for subsequent changes in the PPARα mRNA expression levels relative to control cells. In both cell lines PPARα mRNA level increased after short time incubation with AA (data not shown) and declined below baseline level at 24 h, when PPARα protein amount was higher than control cells. The pre-treatment of the cells with the antagonist, alone or in the presence of AA, caused in both cases the increase of PPARα mRNA level.
Fig. 10 Effect of arachidonic acid (AA) on PPARα mRNA expression in the presence of MK886. MCF-7 and MDA-MB-231 breast cancer cells were treated for 24 h with 50 μm AA, with or without 5 μm MK886. PPARα mRNA levels in MCF-7 (A) and MDA-MB-231 (B) cells were determined by real-time PCR. Each sample was tested six times and data are reported as variation calculated taking the values of control cells (C) as 1. Mean values were significantly different from controls: *P < 0·05, **P < 0·01.
The effect of the antagonist can probably be related to the fact that it binds to PPARα protein, thus preventing the conformational change necessary for PPARα ligand–receptor interaction(Reference Kehrer, Biswal, La, Thuillier, Datta, Fischer and Vanden Heuvel37).
Involvement of PPARα in cell proliferation, extracellular signal-regulated protein kinase–mitogen activated protein kinase pathway and apoptosis
To thoroughly evaluate whether AA-induced cell growth inhibition may be ascribed to PPARα activation, the growth inhibitory effects of AA in MCF-7 and MDA-MB-231 cells were compared to those obtained in the presence of the specific antagonist MK886. As shown in Fig. 9 ((C) and (D)), in both cell lines the reduction of cell growth induced by AA was significantly prevented by MK886 administration, thus confirming the involvement of PPARα in AA growth inhibitory activity. The pre-treatment of both MCF-7 (Fig. 11 (A)) and MDA-MB-231 (Fig. 11 (B)) cells with the antagonist was able to reduce the inhibitory effect of AA on the phosphorylation state of ERK1/2 (active form). In addition, in MDA-MB-231 cells the antagonist also reduced the pro-apoptotic effects of AA, as it strongly down regulated Bak expression and augmented procaspase-3 levels (Fig. 11 (B)).
Fig. 11 Effect of arachidonic acid (AA) on phospho-ERK1/2, Bak and procaspase-3 levels in the presence of MK886. MCF-7 (A) and MDA-MB-231 (B) breast cancer cells were incubated for 24 h with 50 μm AA, 5 μm antagonist alone or 5 μm antagonist 1 h before 50 μm AA and the total cell lysates were probed with an anti-phosphoERK1/2 (for MCF-7 and MDA-MB-231 cells ()), anti-Bak () or anti-caspase-3 antibody (■) (for MDA-MB-231 cells). Protein contents were normalized by probing the same membrane with anti-β-actin antibody. The densitometric values of the bands are shown as means with standard deviation indicated by vertical bars, from three independent experiments. Mean values were significantly different from controls (c): *P < 0·05, **P < 0·01.
Taken together, these results indicate that PPARα may be involved, at least in part, in the growth inhibitory activity of AA on MCF-7 and MDA-MB-231 cells; furthermore, in MDA-MB-231 cells, cross-talk between PPARα induction and signalling pathways involved in the induction of apoptosis may occur.
Discussion
Breast cancer is one of the most frequently diagnosed cancers and the second most common cause of cancer death in women. Epidemiological and experimental studies conducted over the past few decades suggest a protective role for n-3 PUFA against the development of breast and colon cancer(Reference Hilakivi-Clarke, Olivo, Shajahan, Khan, Zhu, Zwart, Cho and Clarke2, Reference Schley, Jijon, Robinson and Field3). It has been recently reported that also some n-6 PUFA, often believed to exert stimulatory effects on cancer cell growth and metastasis, can reduce the growth of different human breast and colon cancer cells(Reference Hammamieh, Chakraborty, Miller, Waddy, Barmada, Das, Peel, Day and Jett4).
In this study, we investigated the effects of long-chain polyunsaturated 20 : 4n-6 AA on ERα(+) MCF-7 and ERα( − ) MDA-MB-231 human breast cancer cell lines with the aim to elucidate the mechanisms by which AA regulates cell growth.
AA, like other fatty acids, represents a ligand of PPAR, transcription factors belonging to the nuclear hormone receptor super-family that includes receptors for steroids, thyroid hormones, retinoic acid and vitamin D(Reference Kota, Huang and Roufogalis28). Increasingly studies have linked PPAR with mammary tumourigenesis(Reference Saez, Rosenfeld, Livolsi, Olson, Lombardo, Nelson, Banayo, Cardiff, Izpisua-Belmonte and Evans40) and, in particular, PPARγ activators have been trialled as a therapy for breast cancer(Reference Burstein, Demetri, Mueller, Sarraf, Spiegelman and Winer41).
PPARγ has an important role in cell growth regulation, differentiation and fat metabolism(Reference Kota, Huang and Roufogalis28). Recent data have shown that activation of PPARγ causes inhibition of cell proliferation and extensive lipid accumulation(Reference Burstein, Demetri, Mueller, Sarraf, Spiegelman and Winer41) in human breast cancer cells that are characterized by high levels of PPARγ.
In contrast to the growth-inhibitory effect of PPARγ on breast cancer cells(Reference Mueller, Sarraf, Tontonoz, Evans, Martin, Zhang, Fletcher, Singer and Spiegelman42), the anti-tumour effect of PPARα appears to be less commonly reported. One reason for this may be the perception that PPARα agonists are carcinogenic in rodents(Reference Roberts-Thomson and Snyderwine43) and increase proliferation in some breast cancer cells(Reference Suchanek, May, Robinson, Lee, Holman, Monteith and Roberts-Thomson44). This tumourigenic effect, however, is controversial as PPARα agonists have demonstrated antiproliferative efficacy in vitro against melanoma(Reference Grabacka, Plonka, Urbanska and Reiss45), endometrial and breast cancer cell lines(Reference Saidi, Holland, Charnock-Jones and Smith46).
In our experimental condition AA induces a strong increase in both PPARα and PPARγ expression, while it does not interfere significantly with PPARβ expression. The order of magnitude of this effect differs according to the cell line considered: we detected higher levels of PPARα in the ERα ( − ) MDA-MB-231 than in the ERα (+) MCF-7, while PPARγ was more strongly expressed above all in MCF-7 cells.
These results indicate the involvement of distinct PPAR pathways and are in accordance with studies showing a relationship between the relative levels of PPARα and ERα in breast cancer cells, where high levels of ERα mRNA expression are associated with reduced levels of PPARα(Reference Faddy, Robinson, Lee, Holman, Monteith and Roberts-Thomson47). It has been reported that induction of ERα expression in MDA-MB-231 cells reduces PPARα levels whereas inhibition of ERα activity in ERα (+) MCF-7 cells increases PPARα levels(Reference Suchanek, May, Robinson, Lee, Holman, Monteith and Roberts-Thomson44). These observations suggest that the relationship between ERα and PPARα is more than correlative and reflects the ability of ERα to dynamically regulate PPARα activity and expression(Reference Faddy, Robinson, Lee, Holman, Monteith and Roberts-Thomson47). The precise relationship between PPARα and ERα is not completely elucidated, but studies on PPARα promoter indicate the presence of at least two nuclear receptor response elements(Reference Pineda Torra, Jamshidi, Flavell, Fruchart and Staels48) that could potentially be regulated by activated ERα.
Although we found that AA induces the expression of both PPAR isotypes in the two cell lines tested, the fact that the PPARα antagonist completely prevents AA stimulatory effect on PPARα expression, while the PPARγ antagonist is ineffective, most likely indicates a specific involvement of PPARα in AA-induced growth inhibition.
PPARα exerts a dynamic regulation in the mammary gland during pregnancy and lactation and potentially it is hormonally regulated. It is expressed in the mouse mammary gland with the maximum levels at 2 weeks of age and with declining levels during pregnancy and lactation(Reference Master, Hartman, D'Cruz, Moody, Keiper, Ha, Cox, Belka and Chodosh49), but this potential role of PPARα may extend beyond the normal physiological regulation of the mammary gland to tumourigenesis(Reference Roberts-Thomson and Snyderwine43). Literature data outline various mechanisms to explain the effect of PPARα in cancer cells; in particular, PPARα regulates networks involved in the control of cell growth, cell cycle and apoptosis and its agonists probably act through the involvement of the mitogen-activated protein kinase pathway and eicosanoid production(Reference Roberts-Thomson and Snyderwine43). In line with these reports, our study showed that in both cell lines treatment with AA produces a reduction of the ERK phosphorylation state, a reduction that was prevented by pre-treatment with PPARα antagonist MK886. In addition, MK886 was also capable of attenuating AA-mediated cell growth inhibition, while it was notable that, in both cases, pre-treatment with PPARγ antagonist was ineffective.
We showed that AA inhibits the growth of both cell lines in a dose-dependent manner, MDA-MB-231 cells being more sensitive than MCF-7 cells. The growth inhibitory effect of AA is not characterized by alterations in cell cycle progression (data not shown) but is accompanied by different contributions from apoptosis and necrosis. In MDA-MB-231 cells AA growth inhibitory activity is likely related to apoptosis induction. Here again, the PPARα antagonist was able to prevent AA effects, thus suggesting that the modulation of PPARα expression contributes to some degree to cell growth inhibition by AA.
Emerging evidence has shown that elevated intracellular AA can induce cell death via the mitochondrial-mediated apoptosis pathway(Reference Pompeia, Lima and Curi50). In addition, AA displays pro-apoptotic characteristics such as the ability to be converted to pro-apoptotic eicosanoids(Reference Chen, Shen and Tsai51, Reference Kwon, Jung, Lee, Moon and Baik52) or to regulate the expression of genes involved in susceptibility and resistance to apoptosis(Reference Monjazeb, Clay, High and Chilton16).
Collectively, AA shows a different behaviour with regard to cell growth inhibition or cell death induction in correlation with variations in PPAR isotype expression levels. Our results most likely indicate that a PPARγ-mediated pathway is not apparently involved in the growth inhibitory activity of AA, whereas a connection can be established with the PPARα pathway, both in MCF-7 and MDA-MB-231 breast cancer cell lines. The distinct response to AA that we detected could be related to the differences between MCF-7 and MDA-MB-231 cells in ERα levels. Of interest is the observation that the greatest increase of PPARα protein occurred in ERα ( − ) MDA-MB-231 cells which underwent induction of apoptosis, in accordance with reports indicating that PPARα may be correlated to the apoptotic programme execution(Reference Roberts, Chevalier, Hasmall, James, Cosulich and Macdonald29).
These features demonstrate that the 20 : 4n-6 PUFA, AA, can exert its growth inhibitory activity on breast cancer cells in a similar manner to n-3 PUFA, and support the hypothesis that it can be considered also as an anti-tumourigenic compound.
Acknowledgements
This work was supported by grants from University of Torino, Italy. No conflict of interest exists.
