Troglitazone suppresses transforming growth factor beta-mediated fibrogenesis in retinal pigment epithelial cells

Cell Culture and treatment

Cells of the human RPE line ARPE-19 were obtained from the American Type Culture Collection (ATCC; Manassas, VA). ARPE-19 is an immortalized cell line that spontaneously arose from cultures of human RPE [21]. The cells were cultured in a humidified incubator at 37 °C in 5% CO₂ in 10% fetal bovine serum-defined minimal essential medium (FBS-DMEM)-F12 medium supplemented with 100 U/mL penicillin G and 100 μg/mL streptomycin. When cultures achieved confluence, the spent medium was removed and replaced with fresh FBS-free medium. After 24 h of serum starvation, cells were treated with recombinant human TGFβ2 (R&D Systems, Minneapolis, MN) as previously described [12,13]. For inhibitor studies, cells were preincubated with SB203580 or TGZ (Calbiochem, San Diego, CA) for 1 h followed by TGFβ2 treatment. The inhibitors dissolved in DMSO (DMSO) were added to the cell culture (the final concentration of DMSO was less than 0.05%).

Cell lysis, fractionation, and electrophoresis

ARPE-19 cells were scraped into lysis buffer (150 μL in each 35 mm well of the microtiter plate) containing 20 mM HEPES (pH 7.4), 1% sodium dodecyl sulfate (SDS), 150 mM NaCl, 1 mM EGTA, 5 mM β-glycerophosphate, 10 mM sodium pyrophosphate, 10 mM sodium fluoride, 100 μM sodium orthovanadate, 10 μg/mL leupeptin, and 10 μg/mL aprotinin. The lysate was incubated on ice for 15 min. Cell debris was removed by centrifugation (15,000 rpm, 15 min, 4 °C). For extraction of cytoplasmic and nuclear fractions, the NE-PER™ nuclear and cytoplasmic extraction kit (Pierce, Rockford, IL) was used according to the manufacturer’s instructions. Samples containing 20 μg of protein were analyzed by 12% SDS–PAGE and electrotransferred to polyvinylidene fluoride membranes (Immobilon-P; Millipore, Bedford, MA) and processed for western blotting.

Western blot analysis

Samples were probed with anti-Active p38 pAb (Promega, Madison, WI), anti-phospho-Smad2 (S465/S467, Upstate Biotechnology, Lake Placid, NY), anti-phospho-Smad3 (S423/S425, R&D Systems, Minneapolis, MN), or anti-phospho-HSP27 (Ser78) antibody (Upstate Biotechnology) according to the manufacturer’s instructions, and then washed three times in Tris-buffered saline containing 0.1% Tween 20 (TBST). Antibody directed against p38/SAPK2 (Upstate Biotechnology), Smad2 (Abcam Ltd, Cambridge, UK), Smad3 (Zymed Laboratories, San Francisco, CA), or β-actin (Sigma-Aldrich, St. Louis, MO) were used to verify equal loading of protein. The blots were incubated with horseradish peroxidase-labeled anti-rabbit secondary antibody (Amersham Biosciences, Piscataway, NJ) diluted in TBST for 1 h and washed three times in TBST before visualization using an enhanced chemiluminescent technique. X-ray films were scanned on the Model GS-700 imaging densitometer (Bio-Rad Laboratories, Hercules, CA) and analyzed using Labworks 4.0 software (UVP, Upland, CA). For quantification, blots of at least three independent experiments were used.

Measurement of the C-terminal propeptide of collagen type I

After treatment for 48 h, 20 μL of the conditioned medium was analyzed for collagen type I C-terminal peptide by an enzyme-linked immunosorbant assay, according to the manufacturer’s method (TaKaRa Biochemicals Co., Osaka, Japan).

Semi-quantitative reverse transcriptase-polymerase chain reaction (RT–PCR)

Total RNA was extracted from ARPE-19 cells with TRIzol™ reagent (Invitrogen, Carlsbad, CA), according to the manufacturer’s protocol. Synthesis of cDNA was performed with 1 µg of total RNA at 50 °C for 50 min, using oligo (dT) primers and reverse transcriptase (Superscript III, Invitrogen). The amplification mixture (final volume 20 μl) contained 1 × Taq polymerase buffer, 0.2 mM dNTPs, 1.5 mM MgCl₂, 1 μM primer pair, and 0.5 U of Taq DNA polymerase (Life Technologies, Inc., Gaithersburg, MD). cDNA was equalized in an 18–22 cycle amplification reaction with fibronectin primers 5`-tcgaggaggaaattccaatg-3` (forward) and 5`-ctcttcatgacgcttgtgga −3` (reverse), or collagen type I primers 5`-ggagggaatcactggtgcta −3` (forward) and 5`-agggggaaaaactgctttgt-3` (reverse) yielding a 300-bp product. The number of cycles for the primer set (denaturation for 20s at 94 °C; annealing for 30s and 59 °C; and polymerization for 40s at 72 °C) was chosen to be in the linear range of amplification.

Cell migration assay

Cells (1 × 10⁶) were plated on 6-well culture plates (Corning, Corning, NY) in serum-free DMEM-F12. The cell layer was scratched with a pipette tip. The culture was then incubated with various reagents. Photographs of the wound were taken using a Nikon ECLIPSE TS100 microscope at various time points after monolayer wounding. Cell migration was evaluated by assaying the closure of a linear defect produced in a cell monolayer culture as previously reported [13].

Cell viability

ARPE-19 cells were seeded in 10% FBS-DMEM-F12 medium in 96-well microculture plates (12,000 cells/well, Nunclon, Roskilde, Denmark) for 24 h. The culture medium was then replaced by serum-free DMEM-F12 medium for 16 h. At the time, the cell density before TGFβ2 or TGZ treatment was around 22,000 cells/well. Cells were received serum-free medium containing various concentrations of TGZ or were treated with 4 ng/ml TGFβ2 for 48 h. Cell viability was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, and confirmed by the trypan blue exclusion assay [22]. For performance of trypan blue exclusion assay, at the end of the cytotoxicity test, the cells were incubated with 0.05% trypan blue for 30 min.

Statistical analysis

Data are expressed as mean ± standard deviation (SD) of three independent experiments, each done in triplicate (n=3–4 dishes). The Mann–Whitney U test was used to determine statistically significant differences. P values < 0.05 were considered significant.

TGZ inhibits TGFβ2-induced fibronectin

Since fibronectin and collagen type I are the major ECM components of PVR tissue [13], we investigated whether TGZ could prevent TGFβ2-induced fibronectin protein in ARPE-19 cells. As shown in Figure 1A, western blot analysis revealed that exogenous TGFβ2 significantly increased fibronectin protein levels after treatment for 48 h as compared to 24 h treatment. Cells pretreated with 10 μM TGZ prevent the TGFβ2-mediated induction of fibronectin; at all time periods, the fibronectin level of the pretreated cells were similar as TGFβ2-untreated cells. DMSO pretreatment had no effect. p38 MAPK activation can be induced by TGFβ2 and is essential for TGFβ2-induced ECM production [12,13]. Our results also revealed that a similar inhibition effect was evident in cells pretreated with equal concentrations of a p38 MAPK inhibitor (SB203580) or TGZ.

Exposure of ARPE-19 cells to 1–4 ng/mL TGFβ2 for 48 h increased fibronectin protein levels in a dose-dependent manner (Figure 1B, lanes 2–4). To examine the TGZ dose effect on TGFβ2-induced fibronectin protein, cells were pretreated for 1 h with 1–10 μM TGZ before addition of 4 ng/ml TGFβ2 for a further 48 h. Immunoblotting results revealed that 1 and 5 μM TGZ could still temper suppress TGFβ2-induction of fibronectin protein (approximately 0.5-fold induction, similar to the induction produced by 2 ng/ml TGFβ2) (Figures 1B and 1C).

TGZ inhibits TGFβ2-induced collagen type I

To examine whether TGZ could prevent TGFβ2-induced collagen type I protein, we measured the concentration of the C-terminal propeptide of collagen type I (PICP) in the medium. As shown in Figure 2, exogenous TGFβ2 increased collagen type I protein levels, consistent with previous reports [12,13]. Cells pretreated for 1 h with 1–10 μM TGZ before the 48 h addition of 4 ng/ml TGFβ2 suppressed TGFβ2-induced collagen type I in a dose-dependent fusion. Pretreatment with DMSO had no such effect. In addition, 5 μM TGZ had an inhibitory effect exceeded that obtained following pretreatment with 10 μM SB203580.

TGZ inhibits TGFβ2-induced collagen type I and fibronectin mRNA expression

Next, we investigated whether TGZ was capable of inhibition of TGFβ2-induced collagen type I and fibronectin mRNA expression. As shown in Figure 3, RT–PCR analysis demonstrated that ARPE-19 cells cultured in serum-free medium expressed a basal level of collagen type I and fibronectin genes; treatment with exogenous TGFβ2 for 24 h markedly induced expression of both mRNAs. Cells treated with TGFβ2 for 8 h also displayed an increased expression of both mRNAs but at levels less than following stimulation for 24 h (approximately threefold). The TGFβ2-mediated induction effect was completely blocked by actinomycin D pretreatment, suggesting that the increased mRNA expression was transcription dependent. Cells pretreated for 1 h with 10 μM TGZ before TGFβ2 treatment suppressed TGFβ2-induced collagen type I and fibronectin mRNA to basal levels that were comparable to untreated cells. Pretreatment with DMSO had no such effect. Similarly, 10 μM SB203580 pretreatment partially prevented TGFβ2-induced collagen type I mRNA expression (compare Figure 2 with Figure 3).

Next, we investigated whether TGZ could prevent TGFβ2-induced expression of fibronectin and collagen type I in subconfluent ARPE-19 cells. As shown in Figure 4A, RT–PCR analysis revealed that both mRNAs were markedly upregulated by TGFβ2 treatment for 24 h and the TGFβ2-mediated induction effect was blocked by TGZ pretreatment followed a dose-dependent manner. Pretreatment of subconfluent ARPE-19 cells with TGZ (1–10 μM, 1h) also dose-dependently suppressed TGFβ2-induced fibronectin and collagen type I proteins as assayed by western blot analysis and ELISA (PICP), respectively (Figure 4B and 4C). DMSO pretreatment had no effect. Collectively, our results revealed that TGZ suppresses TGFβ2-mediated fibrogenesis in ARPE-19 cells grown to either confluence or subconfluence.

TGZ inhibits TGFβ2-induced cell migration

TGFβ2-mediated promotion of RPE cell migration has been implicated in the development of PVR [23]. We examined this in more detail using an in vitro wound healing assay in which ARPE-19 cell migration was quantified by the width of remaining defect. As shown in Figure 5B, addition of TGFβ2 significantly enhanced the cell migration (compared untreated cells with TGFβ2-treated cells at 24 h post-wounding; p<0.01). Cells pretreated with 5 and 10 μM TGZ suppressed TGFβ2-promoted cell migration in all the time periods studied (approximately 50 and 72% of the remaining defect as compared to the original defect), but 1 μM TGZ or DMSO had no inhibitory effect. Figure 5A shows representative photographs of cells migrating into scratch wounds. Our results also revealed that a similar inhibition effect was obtained from cells pretreated with 10 μM SB203580 or 5 μM TGZ before TGFβ2 treatment for 48 h (**p<0.05 versus 48 h TGFβ2-treated cells).

To confirm that the effect of TGZ on wound gap was not a result of TGZ cytotoxicity on ARPE-19 cells, cells were exposed to increasing concentrations of TGZ (1–10 μM) in serum-free medium for 48 h, and cell viability was examined using an MTT assay. As shown in Figure 6A, TGZ did not affect cell viability even in cells treated with 10 μM TGZ, although proliferation of ARPE-19 cells was evident; the latter is consistent with another study [13]. However, 10 μM TGZ showed no further inhibitory effect on cell proliferation in the presence of TGFβ2. To further confirm that TGZ at 10 μM does not affect RPE survival at the region of wound healing, cell viability was evaluated using trypan blue exclusion assay. Results revealed that 10 μM TGZ did not affect cell survival while 50 μM TGZ treatment caused massive cell death and trypan blue staining (Figure 6B).

The TGZ inhibition effect is independent to the activation of PPARγ

To investigate whether TGZ acts through the activation of its receptor PPARγ to suppress TGFβ2-mediated reactions, ARPE-19 cells were pretreatment for 2 h with 5–20 μM GW9662, a specific PPARγ antagonist, and then assayed for PPARγ levels, fibronectin levels and cell migration in the presence of 10 μM TGZ combining with TGFβ2. As showed in Figure 7A, western blots showed that there was no change in the PPARγ protein levels with exposure to TGZ in the presence of TGFβ2. In addition, GW9662 did not reverse the inhibition of TGFβ2-induced fibronectin production (Figure 7A). in vitro wound healing assay also revealed that TGZ suppressed TGFβ2-promoted cell migration can not be reversed by GW9662 (Figure 7B, compared TGFβ2+TGZ with GW9662+TGFβ2+TGZ-treated cells at 48 h post-wounding). Therefore, the results led us to suggest that TGZ-mediated inhibitions are through a PPARγ-independent mechanism.

TGZ does not affect the TGFβ2-induced p38 MAPK phosphorylation

We further investigated whether TGZ affected TGFβ2-induced p38 MAPK signaling to suppress TGFβ2-mediated responses in ARPE-19 cells. Cells were treated with TGFβ2 or pretreated for 1 h with 10 μM TGZ before TGFβ2 treatment for various time periods. Subsequently, the phosphorylation status of the p38 MAPK was detected by immunoblotting. As shown in Figure 8A and 8B, TGFβ2 stimulation caused phosphorylation of p38 MAPK, which was transiently upregulated at 6 h as previously reported [13]. However, TGFβ2 still caused a similar p38 MAPK phosphorylation in TGZ-pretreated cells in all the time periods studied, suggesting TGZ can non inhibit the p38 MAPK phosphorylation. In addition, DMSO had no effect on basal and TGFβ2-induced p38 MAPK phosphorylation (data not shown).

To further ensure that TGZ dose not affects p38 MAPK signaling, we examined the levels of phosphorylation of HSP27, a documented substrate of p38 MAPK signaling in ARPE-19 cells [24]. Results showed that TGZ, at the concentration that can attenuate fibronectin expression (Figure 8C), does not change the levels of HSP27 phosphorylation induced by TGFβ2 treatment (Figure 8C and 8D), whereas SB203580 dose-dependently inhibited HSP27 phosphorylation. These suggest that TGZ inhibit TGFβ2-induced fibrogenesis through a p38 MAPK-independent mechanism.

TGZ suppresses TGFβ2-induced Smad phosphorylation

Since Smad signaling has been implicated in TGFβ2-dependent induction of ECM components [9], we examined whether the TGZ could affect C-terminal phosphorylation of Smad 2 and 3 by western blot analysis. As shown in Figure 9A, cell exposure to TGFβ2 induced phosphorylation of both Smad species not only after treatment for 2 h, but also was sustained for 24 h, as in previous studies [9,13]. SB203580 had no significant effect on the phosphorylation of Smad2 in all the time periods studied but partially prevented phosphorylation of Smad3 after TGFβ2 stimulation for 24 h (65 ± 12% lower than TGFβ2-treated cells; Figure 9C). Exposure of cells to 10 μM TGZ partially suppressed TGFβ2-induced Smad2 phosphorylation (36 ± 8% versus TGFβ2 treatment for 2h) but had no effect on the levels of Smad2 phosphorylation stimulated by TGFβ2 for 24 h (Figures 9A and 9B). Notably, TGZ suppressed the levels of Smad3 phosphorylation stimulated by TGFβ2 for either 2 h (41 ± 4%) or 24 h (15 ± 3%) relative to TGFβ2-treated cells (Figures 9A and 9C).

TGFβ binding to its receptor causes phosphorylation of Smad2 and Smad3 to form a hetero-oligomeric complex with Smad4, which then translocates into the nucleus to regulate transcription of target genes [25]. We thus examined if TGZ prevented Smad2 and 3 phosphorylation that could also block cytoplasmic Smad translocation to nucleus. Since TGZ can suppress TGFβ2-induced early (2 h) Smad 2 and 3 phosphorylation, the treated cells were subjected to subcellular fractionation and the TGZ effect was examined by western blotting. As shown in Figure 10, as predicted, TGZ prevented TGFβ2-induced Smad2 and Smad3 nuclear translocation (compare the nuclear fractions of TGFβ2 and TGFβ2+TGZ). DMSO or SB203580 pretreatment had no effect.