Thiazolidinediones, such as pioglitazone, seem to exert direct antiatherosclerotic and antirestenotic effects on type 2 diabetes, in part due to an induction of vascular smooth muscle cell (VSMC) apoptosis. We aimed to study the role of transforming growth factor (TGF)-β in rat aortic VSMC. Pioglitazone at 100 μmol/l increased apoptosis without affecting DNA synthesis, and this effect was reversed by an anti-TGF-β1 antibody. Extracellular TGF-β1 levels were rapidly increased after treatment with pioglitazone in a peroxisome proliferator-activated receptor (PPAR)-γ-dependent mechanism because this secretion was blocked by the PPAR-γ inhibitor GW9662. Pioglitazone subsequently increased the nuclear recruitment of phospho-Smad2, without any effect on protein expression. According to our results, we propose that the apoptotic effect of pioglitazone on VSMC depends on the following sequence: PPAR-γ activation, TGF-β1 release, and selective phospho-Smad2 nuclear recruitment. Management of Smad signaling on VSMC might provide future clinical benefits in vascular diseases.
Type 2 diabetes is a major risk factor for atherosclerosis, the leading cause of morbidity and mortality in developed countries (1). In addition, diabetic patients show an increased risk for suffering postangioplasty restenosis (2). Vascular wall size depends on a relative balance between proliferation and apoptosis (3). When the ratio of proliferation to apoptosis increases, an early atherosclerotic plaque or a postangioplasty restenosis is faced.
Thiazolidinediones (TZDs) are an emerging class of antidiabetic drugs that enhance insulin sensitivity in a peroxisome proliferator-activated receptor (PPAR)-γ-dependent fashion (4). However, they have also been demonstrated to exert some direct effects in decreasing neointimal growth after balloon injury (5) and angioplasty (6), two situations with an increased ratio for vascular smooth muscle cell (VSMC) proliferation to apoptosis. Many of the studies concerning the cellular effects of TZD have focused on cell cycle and DNA synthesis regulation. Surprisingly, these effects seem to be PPAR-γ independent (7). On the other hand, at higher doses, TZDs have been proven to induce apoptosis in VSMCs by involving PPAR-γ (8).
Transforming growth factor (TGF)-β1 is an essential cytokine involved in the control of the balance between proliferation and apoptosis in VSMCs. TZDs and TGF-β1 share the induction of apoptosis by similar mechanisms, such as GADD45 (9,10). In addition, some crosstalk between TGF-β1 and PPAR-γ has been described; thus, in the long term, TGF-β1 inhibits PPAR-γ expression (11), whereas PPAR-γ agonists and PPAR-γ itself interact with the TGF-β1 signaling pathway by inhibiting Smad3 (12). Therefore, Smad2 appears as a likely target for TGF-β in TZD-treated VSMCs.
In the present study, we show that the apoptotic effect of the TZD pioglitazone in VSMC depends on the following sequence: PPAR-γ agonism, rapid TGF-β1 autocrine secretion, and selective phospho-Smad2 nuclear recruitment.
RESEARCH DESIGN AND METHODS
Primary cultures of rat VSMCs were obtained from enzymatically dissociated rat thoracic aorta, as described (13). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Madrid, Spain) containing 10% FCS (Gibco), supplemented with 100 IU/ml penicillin G (sodium salt) and 100 μg/ml streptomycin (Gibco).
To rule out the unspecific effects of the serum, cells were washed twice with sterile PBS and medium was changed to a serum-free medium composed by DMEM plus 0.2% BSA and 10 ng/ml platelet-derived growth factor (PDGF) isoform BB before every set of experiments. Because maintenance culture medium has 10% FCS, which contains ∼16 ng/ml TGF-β1 (14), we considered control time 0 to be the time in which cells were incubated in DMEM plus 0.2% BSA. Experiments were performed using subconfluent cells (4,000 cells/cm2) at passages three through five.
Male Wistar-Kyoto rats 8 weeks of age weighing 250 g were used. All animal experimental procedures had ethical approval and followed the guidelines for animal care issued by the Universidad Complutense.
Cells were seeded onto 24-well plates at a density of 4,000 cells/well and allowed to attach for 24 h. The cells were then treated with pioglitazone at 50 and 100 μmol/l or control. Then they were trypsinized and counted using a Neubauer′s chamber.
Analysis of BrdU incorporation.
To determine cell proliferation, VSMCs were plated onto 96-well plates and allowed to attach for 24 h. The cells were serum-starved for 24 h in 0.4% FCS-containing medium in order to synchronize the cells in the G0/G1 phase of the cell cycle. They were then treated with PDGF-containing medium in the presence or absence of pioglitazone at 100 μmol/l for 24 h. The cells were loaded with BrdU (10 μmol/l) for the last 3 h of treatment. BrdU incorporation was measured by a solid-phase enzyme-linked immunosorbent assay (ELISA) kit (Amersham Life Science, Barcelona, Spain) following the manufacturer’s instructions.
Measurement of cellular DNA fragmentation.
The cells were plated on 96-well plates and allowed to attach for 24 h. Pioglitazone (50 and 100 μmol/l) was added to the culture medium for 24, 48, and 72 h. Cellular DNA fragmentation was measured with a commercially available cellular DNA fragmentation ELISA kit (Roche-Boehringer). Proliferating cells in 96-well microtiter plates were labeled with 10 μmol/l BrdU overnight, washed with PBS, and treated in the presence or absence of pioglitazone for 24, 48, and 72 h. In another set of experiments, exogenous recombinant TGF-β1 (R&D Systems, Madrid, Spain) was added at a concentration of 400 pg/ml. After 1 h of incubation, plates were washed with sterile PBS and medium changed to a TGF-β1-free medium. In both sets of experiments, we also performed an inhibition by adding anti-TGF-β1 at 50 μmol/l (R&D Systems).
After treatment, the cells were washed with PBS and incubated with the kit lysis buffer (BSA, EDTA, and Tween 20) for 30 min at room temperature, and soluble BrdU-labeled DNA fragments present in the buffer were quantified using the ELISA kit following the manufacturer’s instructions. DNA fragmentation was expressed as the fold increase of the control values.
Caspase-3 cellular activity.
Caspase-3 activity was assessed as another measurement of apoptosis. A commercially available kit was used for this purpose, following manufacturer’s instructions (Calbiochem, Scwalbach, Germany). Fluorescence (405 nm) was assessed at 24 h.
Cellular DNA content was measured by fluorescence-activated cell sorting (FACS). Cells were plated, allowed to attach for 24 h, and then incubated in the presence or absence of pioglitazone for 24 h. In some experiments, VSMCs were incubated with pioglitazone at 100 μmol/l plus anti-TGF-β1 at 50 μg/ml. The cells were then harvested by trypsinization, washed with PBS, pelleted, and resuspended in PBS containing 0.6% Nonidet P-40 and 100 μg/ml propidium iodide, to which RNAse was added to a final concentration of 100 μg/ml. Flow cytometric analysis was carried out with a FACScan (Becton Dickinson) flow cytometer equipped with a 15-mW Argon laser emitting at 488 nm. Propidium iodide fluorescence was determined through a 575/24-BP filter; 10,000 cells were acquired per sample, and a double discriminator module was used to ensure detection of single cells.
Measurements of TGF-β1 levels.
To determine TGF-β1 levels in the culture medium of control and treated samples, we used a solid-phase TGF-β1-specific ELISA, following manufacturer instructions (R&D Systems). Cells were seeded onto 24-well plates and allowed to attach for 24 h. The cells were then treated with pioglitazone at 100 μmol/l or control for a defined time period (30 min to 24 h). In addition, in some experiments, GW9662, a PPAR-γ inhibitor (2 μmol/l), was added 30 min before pioglitazone. In some experiments, we used the physiological PPAR-γ ligand 15d-PGJ2 (5 μmol/l) in order to establish the effect of a nonthiazolidinedionic agent on this parameter.
The subcellular location of phospho-Smad proteins was also analyzed by confocal images of immunofluorescence-stained samples. Cells were plated onto cover slips and allowed to attach for 24 h. Then, the cells were cultured in the presence or absence of pioglitazone at 100 μmol/l from 30 min to 3 h. In some experiments, GW9662 (2 μmol/l) was added 30 min before treatment. Cells were washed with PBS and fixed for 20 min in 4% paraformaldehyde in PBS and permeabilized with 0.4% Triton-x100 for 30 min at room temperature. After blocking with 3% BSA in PBS, cells were then incubated with goat polyclonal anti-phospho-Smad2 or rabbit polyclonal anti-phospho-Smad2 (1:100) for 1 h. Excess primary antibody was removed by washing with blocking solution, followed by incubation with donkey anti-goat Alexa 568 (1:100; Molecular Probes) or goat anti-rabbit Alexa 568 (1:100; Molecular, Probes) for 1h. The cells were washed four times with blocking buffer every 5 min. Images were captured using a Leica TCS SP2 inverted microscope. Intensity of staining was analyzed by an Image J software.
In some experiments, we added an anti-TGF-β1 (50 μg/ml) antibody to rule out the unspecific nuclear/cytoplasmic shuttle of phospho-Smad2 and thus ensure this pioglitazone-mediated nuclear recruitment is TGF-β1 dependent.
Cells were plated onto 6-well plates and allowed to attach for 24 h. Cells were treated with pioglitazone at 100 μmol/l in DMEM containing PDGF-BB at 10 ng/ml and BSA at 0.2% for 24 h. Then they were washed with ice-cold PBS and lysed on ice with 200 μl lysis buffer (10% glycerol, 2.3% SDS, 62.5 mmol/l Tris, pH 6.8, 150 mmol/l NaCl, 10 mmol/l EDTA, 1 μg/ml eupeptic, 1 μg/ml pepstatin, 5 μg/ml chymostatin, 1 μg/ml aprotinin, and 1 mmol/l phenylmethylsulphonyl fluoride) and boiled for 5 min. Equal amounts of protein were run on 10% SDS-polyacrylamide gel electrophoresis. Proteins were then transferred to polyvinylidene difluoride membranes (Immobilon-P; Amersham, Madrid, Spain) and blocked overnight at 4°C in blocking solution (5% skimmed milk in Tris-buffered saline with Tween [TBS-T]: 25 mmol/l Trizma base, 75 mmol/l NaCl, pH 7.4, and 0.1% vol/vol Tween 20). Blots were incubated with agitation at 4°C overnight in the presence of a specific anti-Smad2 (Santa Cruz Biotechnology) or anti-Smad3 (Calbiochem, Schwalbach, Germany) at 1:1,000 in TBS-T. After washing in TBS-T solution, the blots were further incubated for 1 h at room temperature with horseradish peroxidase-conjugated anti-rabbit secondary antibody diluted 1:10,000 (Santa Cruz Biotechnology) in the blocking solution. The blots were then washed five times in TBS-T, and antibody-bound protein was visualized with an enhanced chemiluminescence kit (Amersham Biosciences, Barcelona, Spain). Smooth muscle α-actin was used as a housekeeping protein and was analyzed following the same procedure, using a specific anti-α-actin mouse monoclonal antibody (Sigma-Aldrich, Madrid, Spain) at 1:1,000 in TBS-T.
Pioglitazone was a generous gift from Takeda Chemical Industries (Osaka, Japan) and was directly diluted in cell culture media. GW9662 was from Tocris (Bristol, U.K.), and 15d-PGJ2 was obtained from Sigma. These two molecules were diluted in DMSO, at a final concentration of 1:1,000.
The results are expressed as the mean ± SEM and accompanied by the number of observations. A statistical analysis of the data were carried out using a Student’s t test or by a one-way ANOVA when necessary. Differences with a P value <0.05 were considered statistically significant.
Proliferation versus apoptotic effect of pioglitazone: role of TGF-β1.
As shown in Fig. 1A, PDGF induced an increase in cell number in a time-dependent manner. Treatment with pioglitazone decreased the proliferation induced by PDGF in a concentration-dependent fashion. At 24 h only the highest concentration (pioglitazone at 100 μmol/l) decreased cell counting, whereas at 72 h both concentrations reduced cell number.
Despite this effect of pioglitazone in cell counting, we did not find any effect of the drug on DNA synthesis measured by BrdU incorporation (Fig. 1B). To elucidate this discrepancy, we analyzed cell death by apoptosis by analysis of DNA fragmentation, FACS, and caspase-3 assay. Previously, we tested all of these concentrations of pioglitazone for cytotoxicity in a set of lactate dehydrogenase (LDH) assays and found them insignificant (data not shown).
Only the highest concentration of pioglitazone (100 μmol/l) induced a significant increase in DNA fragmentation (Fig. 2A). This effect was reversed by anti-TGF-β1, revealing that TGF-β1 is essential in the apoptotic effect of pioglitazone (Fig. 2B and D). Addition of exogenous recombinant TGF-β1 to the cell culture medium for only 1 h also induced apoptosis at 48 h and therefore mimicked the effect of pioglitazone (from a 1.09 ± 0.1- to a 9.7 ± 3.9-fold increase in DNA fragmentation). As expected, the effect of TGF-β1 was reversed by anti-TGF-β1 (from 9.7 ± 3.9 to 1.38 ± 0.12).
To confirm that TGF-β1 induces apoptosis in our model, in a further set of experiments we chose the maximum peak of TGF-β1 that we had previously measured in an ELISA (400 pg/ml) as the concentration of exogenously added TGF-β1 for DNA fragmentation experiments. As expected, 1 h of incubation exerted an apoptotic effect that mimicked the one observed with pioglitazone at 100 μmol/l. Anti-TGF-β1 was able to reverse it.
Flow cytometry experiments showed a similar result: pioglitazone at 100 μmol/l enhanced apoptosis, and anti-TGF-β1 was able to reverse it in a significant manner (Fig. 2C and D). In addition, measurement of caspase-3 absorbance at 24 h is also similar (absorbancies from 0.070 ± 0.0060 in control to 0.1120 ± 0.020 with pioglitazone at 100 μmol/l and back again to 0.060 ± 0.0090 in pioglitazone at 100 μmol/l plus 50 μg/ml anti-TGF-β1).
Pioglitazone induces the release of TGF-β1.
Because TGF-β1 seems to be involved in the apoptotic effect of pioglitazone, we studied whether pioglitazone was able to increase the release of TGF-β1 in rat VSMCs. Figure 3A shows a rapid and transient increase in the release of total TGF-β1 induced by pioglitazone 100 μmol/l as early as 30 min after treatment. Pretreatment with the PPAR-γ inhibitor GW9662 completely blocked this effect of pioglitazone on the TGF-β1 release at the 30 min time-point (Fig. 3B). These effects are similar to those observed with 15d-PGJ2, which increased from TGF-β1 666 ± 51.0 pg/mg protein (control) to 914.0 ± 50.0 (P < 0.05 for 15d-PGJ2 vs. control, n = 2). Yet again, GW9662 at 2 μmol/l inhibited the increase in TGF-β1 levels induced by 15dPGJ2 (914.0 ± 50.0 vs. 706.0 ± 83.0 for 15dPGJ2 and 15dPGJ2 + GW9662, respectively; P < 0.05, n = 2).
We also measured active TGF-β in control and treated samples. As expected because of its short half-life, no significant traces were found (data not shown).
Pioglitazone enhanced the nuclear recruitment of phospho-Smad2.
Pioglitazone at 100 μmol/l rapidly increased the nuclear recruitment of phospho-Smad2, as postulated (Fig. 4A). The addition of GW9662 (Fig. 4B, right) or anti-TGF-β1 (Fig. 5B) completely prevented this effect, showing a PPAR-γ–and a TGF-β1-dependent mechanism in Smad2 activation. Analysis of the images obtained (Fig. 4B, left) showed that the nuclear recruitment of phospho-Smad2 (measured as the ratio of nucleus to cytoplasm) increased after 30 min of stimulation with pioglitazone at 100 μmol/l compared with control. Staining of pSmad 2/3 remained unchanged (Fig. 6).
To rule out that the increase observed in the nuclear recruitment of phospho-Smad2 induced by pioglitazone could be due to an increase in the expression of Smad2, we analyzed the expression of this protein by Western blotting following pioglitazone (100 μmol/l) treatment. Figure 7 shows no differences in the expression of Smad2 in VSMC treated with pioglitazone for 24 h.
The present study demonstrates that the apoptotic effect of pioglitazone in VSMCs could be mediated by a mechanism that includes the activation of PPAR-γ, the release of TGF-β1, and the nuclear recruitment of phospho-Smad2.
In addition to their insulin-sensitizing effect, TZDs have been reported to exert hypoglycemic-independent benefits in atherosclerosis (1) and postangioplasty restenosis (6). Formerly considered as mere PPAR-γ agonists, PPAR-γ-independent effects of TZD have also been described, specially when DNA synthesis and cell cycle regulation are studied (7). However, since insulin-sensitizing effects depend on PPAR-γ agonism (4), the study of PPAR-γ-mediated effects is still essential for understanding the effects of TZD in therapeutics.
The antiproliferative and proapoptotic effect of TZD on vascular cells is nowadays a well-known fact, although the molecular mechanisms involved are not completely elucidated. It has been described by De Dios et al. (15) that TZDs inhibit serum-stimulated VSMC proliferation by a decrease in phosphorylated retinoblastoma (a marker of cell cycle activation). In contrast, our experiments do not show a TZD-induced inhibition of DNA synthesis measured as BrdU uptake, although we found that pioglitazone reduced the increase of cell number induced by PDGF. This discrepancy might be due to several issues. To begin with, in that previous study, serum at 5% was used as mitogen, whereas we used PDGF-BB at 10 ng/ml. Second, our cells were plated at lower confluence, and current research from our laboratory is remarking this feature as an essential parameter in DNA synthesis induced by several substances (unpublished data). And third, there seems to be a dose dependency from cell cycle arrest to apoptosis. In the above-mentioned study, the maximum effect of pioglitazone on cell counting took place at 100 μmol/l, whereas the effect on cell cycle regulatory proteins was just measured when concentrations of pioglitazone up to 30 μmol/l were reached.
Due to a higher selectivity in PPAR-γ agonism, pioglitazone is a less potent molecule than troglitazone (clinically withdrawn); therefore, despite using high concentrations of pioglitazone, these are within the range of concentrations used by others to inhibit cell counting in in vitro studies (15), and these are the concentrations that have been described to induce VSMC apoptosis by a PPAR-γ-dependent mechanism (8).
On the other hand, TGF-β has been regarded as an important therapeutic target in cardiovascular pharmacology (16). Our results show that pioglitazone rapidly increased the release of TGF-β1; this is in accordance with some results observed in neuroendocrine cells, where TGF-β1 is synthesized and stored in secretion granules and released upon stimulation (17). Pioglitazone and 15d-PGJ2-induced TGF-β1 release is a PPAR-γ-dependent effect, as it is inhibited by the PPAR-γ antagonist GW9662. Nuclear receptors such as PPAR-γ are being described to have rapid or “nongenomic” effects, which happen within minutes after stimulation and could be mediated by a membrane-bound receptor population (18).
Another important issue would be to demonstrate the link between the apoptotic effect of pioglitazone and the involvement of phospho-Smad2. Our results in terms of time-course in the release of TGF-β1 and nuclear recruitment of phospho-Smad2 seem to indicate that this mechanism is involved in the apoptotic effect of pioglitazone. The inhibition of pioglitazone-mediated pSmad2 nuclear recruitment by anti-TGF-β1 rules out any unspecific TGF-β-independent nuclear/cytoplasmic Smad shuttling in regard to this context.
At elapsed time-points such as 72 h, data suggest there might be several other mechanisms, since the anti-TGF-β1 is unable to totally protect VSMC from pioglitazone-induced apoptosis. In relation to this, there are several reports that link PPAR-γ agonism with other elapsed apoptotic pathways in VSMCs, such as transcriptional activation of the PPAR-γ→interferon regulatory factor-1→p21cip1 pathway (19). On the other hand, TGF-β is able to activate other signaling cascades that interact with Smad2, such as the mitogen-activate protein kinase extracelluar signal-related kinase (11).
The recent report (20) of the first death associated with pioglitazone highlights the importance of a comprehensive approach in order to ensure more rational therapeutics. The outcome of this study could help in understanding the mechanism of the apoptotic actions of pioglitazone. Presumably, future in vivo and clinical studies will broaden our knowledge on this class of drugs.
This work was funded by Fondo de Investigaciones Sanitarias (FISS 01/0815, a Health Research Fund from the Spanish Ministry of Health).