TGF-β2 inhibits AKT activation and FGF-2-induced corneal endothelial cell proliferation

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / y e x c r

Research Article

TGF-β2 inhibits AKT activation and FGF-2-induced corneal endothelial cell proliferation Jiawei Lua,d , Zhenyu Lub,f , Peter Reinach g , Jingwu Zhanga , Wei Daia,e,f,⁎, Luo Lua,d,f,⁎, Ming Xua,c,f,⁎ a

Institute of Health Science, Shanghai Institutes for Biological Science, Chinese Academy of Sciences, Shanghai JiaoTong University School of Medicine, Shanghai, PR China b Department of Medical Genetics, Shanghai JiaoTong University School of Medicine, Shanghai, PR China c Department of Anesthesia and Critical Care, University of Chicago, Chicago, IL 60637, USA d Division of Molecular Medicine, Harbor-UCLA Medical Center, David Geffen School of Medicine, University of California Los Angeles, Torrance, CA 90502, USA e Division of Molecular Carcinogenesis, New York Medical College, Valhalla, NY 10595, USA f Model Organism Division of E-Institutes of Shanghai Universities, Shanghai JiaoTong University School of Medicine, Shanghai, PR China g Department of Biological Sciences, SUNY College of Optometry, New York, NY 10036, USA

ARTICLE INFORMATION

ABS T R AC T

Article Chronology:

The corneal endothelial cells form a boundary layer between anterior chamber and cornea. This

Received 29 May 2006

single cell layer is important to maintain cornea transparency by eliciting net fluid transport into

Revised version received

the anterior chamber. Injuries of the corneal endothelial layer in humans lead to corneal

7 August 2006

swelling and translucence. This hindrance is thought to be due to limited proliferative capacity of

Accepted 8 August 2006

the endothelial layer. Fibroblast growth factor 2 (FGF-2) and transforming growth factor-beta 2

Available online 10 August 2006

(TGF-β2) are both found in aqueous humor, and these two cytokines promote and inhibit cell growth, respectively. The intracellular signaling mechanisms by which TGF-β2 suppresses the

Keywords:

mitogenic response to FGF-2, however, remain unclear. We have addressed this question by

AKT signaling

investigating potential crosstalk between FGF-2-induced and TGF-β2-regulated intracellular

COX-2 activity

signaling events in cultured bovine corneal endothelial (BCE) cells. We found that TGF-β2 and

PGE-2 production

FGF-2 oppositely affect BCE cell proliferation and TGF-β2 can override the stimulating effects of

Cell proliferation

FGF-2 by increasing COX-2 expression in these cells. Consistent with these findings,

Cell cycle

overexpression of COX-2 significantly reduced FGF-2-induced cell proliferation whereas a COX-2 specific inhibitor NS398 reversed the effect of TGF-β2 on FGF-2-induced cell proliferation. The COX-2 product prostaglandin E2 (PGE-2) blocks FGF-2-induced cell proliferation. Whereas FGF-2 stimulates cell proliferation by activating the AKT pathway, TGF-β2 and PGE-2 both inhibit this pathway. In accordance with the effect of PGE-2, cAMP also inhibits FGF-2-induced AKT activation. These findings suggest that the mitogenic response to FGF-2 in vivo in the corneal endothelial layer may be inhibited by TGF-β2-induced suppression of the PI3-kinase/AKT signaling pathway. © 2006 Elsevier Inc. All rights reserved.

⁎ Corresponding authors. W. Dai is to be contacted at Dept of Medicine, New York Medical College, Basic Science Building A22, Valhalla, NY 10595, USA. L. Lu, Department of Medicine, David Geffen School of Medicine, University of California, 1124 W. Carson St., Torrance, CA 90502, USA. M. Xu, Department of Anesthesia and Critical Care, University of Chicago, Chicago, IL 60637, USA. E-mail addresses: [email protected] (W. Dai), [email protected] (L. Lu), [email protected] (M. Xu). 0014-4827/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2006.08.004

3632

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

Introduction The corneal endothelium is a single layer of cells forming a boundary between the corneal stroma and anterior chamber. Its barrier and pump functions are responsible for maintaining corneal transparency by regulating stromal hydration [1]. This tissue layer is non-replicating, and its capacity for regeneration after injury is severely limited in humans, primates and cats [2]. Consequently, cell loss can result in a permanent decline in cell density and an inability to maintain corneal transparency [3,4]. The fact that human corneal endothelium retains proliferative capacity has led to recent efforts to induce division and increase their cell density. For example, recent studies have demonstrated that adult human corneal endothelial cells can be induced to grow in culture and then be transplanted to recipient corneas ex vivo [5]. FGF-2 is a member of the fibroblast growth factor family and is a multifunctional regulator of cell development, differentiation, regeneration, senescence, proliferation and migration. In normal cornea, FGF-2 is a component of Descemet's membrane that may be necessary for wound repair and proliferation [6,7]. FGF-2 stimulates corneal endothelial proliferation and alters cell shape through activation of the PLC-γ1 and PI3-kinase signaling pathways [8]. The TGF-β family includes three multifunctional proteins, TGF-β1, TGF-β2 and TGF-β3, that are found in aqueous humor [9]. TGF-β is involved in regulating cell differentiation, cell proliferation and other cell functions [10]. This cytokine reportedly inhibits the proliferation of corneal endothelial cells [11,12]. In many cell types, including epithelial, vascular endothelial and blood cells, TGF-β inhibits cell proliferation by blocking cell cycle progression from the G0/G1 phase to the S-phase [13]. The effects of TGF-β are exerted through binding to specific receptors, designated Fig. 1 – Effects of FGF-2 and TGF-β2 on BCE cell proliferation. (A) Dose-dependent effects of FGF-2 on BCE cell proliferation. Different doses of FGF-2 were applied to the BCE cell culture, and cell proliferation was determined by the MTT cell proliferation assay. FGF-2-induced cell proliferation was dose-dependent. (B) Dose-dependent effects of TGF-β2 on BCE cell proliferation. BCE cells were cultured in 10% heat-inactivated fetal bovine serum, and different doses of TGF-β2 were added. TGF-β2 inhibited cell proliferation in a dose-dependent manner. Cell proliferation was determined by the MTT cell proliferation assay. (C) Effects of TGF-β2 on FGF-2-induced BCE cell proliferation. BCE cells were cultured in the presence of 10 ng/ml of FGF-2, and different doses of TGF-β2 were applied to the culture. TGF-β2 inhibited FGF-2-induced BCE cell proliferation, as determined by the MTT cell proliferation assay. Panel C indicates baseline control with no cytokine treatment. (D) Effects of TGF-β2 and FGF-2 on cell cycle regulation of BCE cells. FGF-2 (10 ng/ml) promoted the cells to enter the S phase from the G1 phase whereas TGF-β2 (10 ng/ml) inhibited this pattern. The symbol * represents a significant difference ( p < 0.05) compared with untreated control values. All values were represented as mean ± SEM from triplicate samples (n = 3).

TGF-β type 1 and type 2 receptors. These receptors have serine/threonine kinase activity. Both type 1 and type 2 receptors are necessary for TGF-β signal transduction. After TGF-β binds to type 1 and type 2 receptors, signals are transduced from the receptors to the nucleus by various factors including SMAD family proteins [14–16]. Other signal transducers of TGF-β, including protein kinase C, have been demonstrated in chrondrocytes [17]. Cytokines, such as platelet-derived growth factor, modulate TGF-β mediated control of responses in liver [17]. It was demonstrated that

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

PGE-2 is the primary PG produced by corneal endothelium, which is synthesized and released in substantial quantities by cultured rabbit corneal endothelial cells. Furthermore, TGF-β2 stimulates PGE-2 synthesis in corneal endothelium and inhibits its proliferation in a dose-dependent manner [12]. PGE-2 is a primary arachidonic acid metabolite synthesized by corneal endothelial cells [18] and reaches relatively high levels in aqueous humor following injury (50–60 ng/ml) [19]. Despite these insights, the intracellular signaling mechanisms associated with FGF-2-induced stimulation and TGF-β2suppressed corneal endothelial cell proliferation remain unclear. We have addressed this issue by investigating potential crosstalk between FGF-2-induced and TGF-β2-regulated intracellular signaling pathways. Our results suggest that TGF-β2 blocks FGF-2-induced BCE cell proliferation likely by inhibiting its activation of the PI3-kinase/AKT signaling pathway.

Methods Cell culture A BCE cell line was obtained from American Type Culture Collection (CRL-2048), and this line was established from explants of normal adult bovine corneas. Based on published work, the characters of this line including cellular signaling are very parallel to those of the primary corneal endothelial cells and it has been used as a useful model system that retained many of the key features of the primary cells [20,21]. These cells were grown in Dulbecco's modified Eagle's medium (DMEM) with 10% heat-inactivated fetal bovine serum, 100 μg/ml penicillin and 50 μg/ml streptomycin sulfate. Fig. 2 – Effects of TGF-β2 on COX-2 promoter activity and expression in BCE cells. (A) Effect of TGF-β2 on COX-2 gene promoter activity. BCE cells were transiently co-transfected with a COX-2 reporter COX-2-pGL2 plasmid and an internal control plasmid carrying a β-galactosidase gene. Promoter activities were detected and normalized by taking a ratio of measured luciferase activity and internal control β-galactosidase activity. Luc/Lac represents ratio of luciferase/β-galactosidase activity. 10 ng/ml each of FGF-2 and TGF-β2 was used. (B) Induction COX-2 protein expression by TGF-β2. BCE cells were treated with TGF-β2 at 10 ng/ml for different periods of time, and COX-2 expression was studied by Western blot analysis. The densities were quantitatively scanned and plotted as normalized densities. (C) Effects of COX-2 overexpression on FGF-2-induced BCE cell proliferation. COX-2 protein expression was detected by Western blot analysis, and β-actin levels were detected as loading controls (left panel). FGF-2 (10 ng/ml) was applied to the control and COX-2 transfected cells. On day 3, cell proliferation was determined by the MTT cell proliferation assay (right panel). All values were represented as mean ± SEM from triplicate samples (n = 3). The symbol * represents a significant difference ( p < 0.05) compared with relative control values.

3633

All cell culture reagents were purchased from PAA Laboratories. Cells were cultured in 175-cm2 flasks and maintained in an incubator supplied with 95% air and 5% CO2 at 37°C. They were detached with 0.05% trypsin–EDTA (GIBCO) for experimental use, serial passage or storage.

Antibodies and reagents Polyclonal rabbit antibodies against phosphor-p44/42 MAPK (Thr202/Tyr204), p44/42 MAPK, phosphor-AKT (Ser473) and AKT, and the MEK inhibitor PD98059 were obtained from Cell Signaling Technology, and a polyclonal rabbit antibody against COX-2 was purchased from Chemicon. Horseradish peroxidase (HRP)-conjugated anti-rabbit IgG antibody was purchased from Amersham Bioscience. PGE-2 was obtained from BIOMOL Research Laboratories. Recombinant TGF-β2 was purchased from R&D System. Recombinant FGF-2 was obtained from PeproTech. LY294002 and NS398 were purchased from Sigma.

3634

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

Cell proliferation assay The CellTiter 96 MTT Cell Proliferation Assay was performed according to the instruction manual provided by the manufacturer (Promega). Briefly, after serum starvation, BCE cells were cultured in 96-well plates (BD Biosciences) at a density of 104 cells/100 μl medium per well with each well containing the appropriate treatment as indicated in the text and figure legends. After 72 h of incubation, MTT solution (5 mg/ ml in PBS) was added to each well and the cells were incubated for 4 h at 37°C. A stop solution was then added to dissolve the formazan crystals at 37°C. Cell proliferation was measured based on absorbance at 570 nm using a Sunrise 96-well plate reader (Tecan Group). Each experiment was performed in triplicate and repeated 3 times. The amount of color produced, normalized with the background, is directly proportional to the number of viable cells and is represented as the proliferation index.

Cell cycle analysis Serum-starved BCE cells were collected and treated either with FGF-2, TGF-β2 or both at 10 ng/ml or 1% serum in DMEM for 24 h before they were collected. Ten million cells were spun at 1000×g for 10 min at 4°C and washed once in cold PBS. Cells were resuspended in 0.3 ml of cold PBS and 0.7 ml of cold ethanol was added drop-wise to fix the cells. Cells were placed on ice for 1 h, spun, washed in cold PBS and pelleted. Cell pellets were resuspended in 0.5 ml of PBS staining solution consisting of 50 μg/ml propidium iodide (Sigma) and 0.1 mg/ml RNase A (Sigma). The cell suspension was kept at 4°C for 1 h in the dark. Analysis was performed on a FACSCalibur cytometer (BD Biosciences) using CellQuest and ModFit software, with appropriate gating on the FL2-A and FL2-W channels to exclude cell aggregates.

Transient transfection Approximately 1 × 107 cells were used for each transfection. Transfections were carried out using electroporation with 7 μg Fig. 3 – Effects of COX-2 and PGE-2 on FGF-2-induced BCE cell proliferation. (A) Dose-dependent restoring effect of NS398 on BCE cell proliferation which is inhibited by TGF-β2. BCE cells were pretreated with NS398 at the indicated doses for 30 min followed by TGF-β2 (10 ng/ml) application and cell proliferation was measured. C indicates baseline controls with no TGF-β2 treatment. (B) Up-regulation of PGE-2 secretion by TGF-β2 in BCE cells. 10 ng/ml of TGF-β2 was applied to BCE cells. The culture supernatants were collected at different time points for ELISA. The concentrations of PGE-2 were quantified and plotted. (C) Concentration-dependent inhibition of PGE-2 on FGF-2-induced BCE cell proliferation. 10 ng/ml FGF-2-induced cell proliferation was reversed by PGE-2 dose-dependently. C indicates baseline controls with no treatments. (D) Quantification of the inhibitory effects of PGE-2 (at 50 nM) on FGF-2-induced BCE cell proliferation. All values were represented as mean±SEM from triplicate samples (n =3). The symbol * represents a significant difference (p < 0.05) compared with control groups.

of the reporter and 5 μg of pCMV-β galactosidase as described by the manufacturer's protocol. After overnight culture, BCE cells were washed with ice-cold PBS and starved with serumfree medium. After 24 h, cells were treated with FGF-2, TGF-β2 or both at 10 ng/ml. After another 24 h of incubation, cells were lysed with 400 μl of a reporter lysis buffer and assayed for luciferase and β-galactosidase activities as relative light units using a reporter assay system (Promega). β-galactosidase activity was assessed for the normalization. Lysates (150 μl) were incubated in 150 μl of 2× assay buffer at 37°C for 30 min. Reactions were stopped by the addition of 500 μl of 1 M Na2CO3, and activity was measured with Sunrise 96-well plate reader (Tecan Group) as absorbance at 420 nm. Transfections

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

of a COX-2 overexpression construct or an AKT-dominant negative construct were similarly performed, and the transfections were repeated several times.

Western blot analysis Cells were washed with ice-cold PBS and then lysed with a lysis buffer consisting of 20 mM Tris–HCl (pH 7.5), 137 mM NaCl, 1.5 mM MgCl2, 2 mM EDTA, 10 mM sodium pyrophosphate, 25 mM β-glycerophosphate, 10% glycerol, 1% Triton X-100, 1 mM sodium orthovanadate and freshly supplemented with 1 mM phenylmethylsulfonyl fluoride (Sigma), 10 μM okadaic acid (Calbiochem) and a protease inhibitor cocktail (Sigma), on ice for 40 min. All chemicals were purchased from Shanghai Chemical Reagent Co. Ltd. unless otherwise noted. Cell lysates were cleared by centrifugation at 13,000×g for 25 min and denatured by boiling in 2× Laemmli buffer for 5 min. Equal amounts of protein samples were subjected to 11% SDS-PAGE followed by electrotransfer to nitrocellulose membranes (Bio-Rad). The membranes were blocked with 5% nonfat dry milk in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 1 h and were subsequently incubated overnight at 4°C with primary antibodies diluted in TBST containing 5% nonfat dry milk. After incubation, the membranes were washed in TBST and then incubated for another 1 h at room temperature with horseradish peroxidase (HRP)conjugated donkey anti-rabbit IgG (Amersham Biosciences) diluted to 1:1000 in TBST containing 5% nonfat dry milk. After an additional wash with TBST, signals were developed with a Supersignal West Pico kit (Pierce) and captured on Kodak X-ray films.

3635

the diluted oligomer with the diluted Lipofectamine 2000 and mixed gently and incubated the mix for 20 min at room temperature. We then added the oligomer–Lipofextamine 2000 complexes to overnight seeded BCE cells in 6-well plate and mixed gently. Cells were maintained in an incubator supplied with 95% air and 5% CO2 at 37°C. After overnight incubation, BCE cells were washed with ice-cold PBS and starved with serum-free medium. After 24 h, cells were cultured in 1% serum and were treated with FGF-2 at 10 ng/ ml. After 72 h of treatment, the cells were washed with icecold PBS, then MTT solution (5 mg/ml in PBS) was added to each well and the cells were incubated for 4 h at 37°C. Isopropyl alcohol with 0.04 M hydrochloric acid was added to dissolve the formazan crystals at 37°C after the supernatant was removed. Sequence of the siRNA to knockdown bovine AKT-1 is 5′-CGAGGUGAGUACAUCAAGATT-3′. This oligonucleotide and its complementary RNA strand were synthesized by Shanghai GenePharma, pre-annealed in vitro before transfections. Cell proliferation was measured based on absorbance at 570 nm with Sunrise 96-well plate reader (Tecan Group).

Statistics Data are shown either as original values or in the case of OD absorbance values are indicated as means ± standard errors. Statistical significance was evaluated with the Student's t test by determining if the p value was equal to or less than 0.05.

Results

Quantification of PGE-2

TGF-β2 and FGF-2 oppositely affect cell proliferation

TGF-β2 was applied to BCE cell for indicated time before the supernatants were collected. PGE-2 levels were determined by ELISA according to the instructions from manufacturer (R&D Systems). Briefly, we added 100 μl of pre-acidified samples, 50 μl of the Primary Antibody Solution and 50 μl of PGE2 conjugate to each well and covered with the adhesive strip. We then incubated the mixture for 2 h at room temperature on a horizontal orbital microplate shaker set at 500 ± 50 rpm. After that, we aspirated each well and washed with Wash Buffer (400 μl) for four times and added 200 μl of Substrate Solution to each well and incubated for 30 min at room temperature. Finally, we added 50 μl of Stop Solution to each well and determined the optical density using Sunrise 96-well plate reader (Tecan Group) at 450 nm. We set 570 nm as the wavelength correction. Values were expressed as the mean ± SD of 4 samples.

To investigate the effects of TGF-β2 and FGF-2 on BCE cell proliferation, we determined dose–response curves for these two cytokines in 1% serum following 24 h of exposure to serum-free medium for starvation. FGF-2 promoted BCE cell proliferation in a dose-dependent manner with the maximal effect being reached at 10 ng/ml dose (Fig. 1A). On the other hand, as described, TGF-β2 suppressed BCE cell proliferation (Fig. 1B, [11,20,22]). To evaluate the effect of TGF-β2 on FGF-2induced cell proliferation, we simultaneously treated BCE cells with both cytokines. After 72 h of co-incubation of 10 ng/ml of FGF-2 and different doses of TGF-β2, increasing doses of TGFβ2 suppressed FGF-2-induced cell proliferation (Fig. 1C). To further investigate the effects of FGF-2 and TGF-β2 on cell cycle regulation, we determined the cell cycle distribution in the presence of FGF-2 or TGF-β2. FGF-2 addition significantly decreased the percentage of cells in the G1/G0 phase, from a basal level of 52.2% to 29.6%. By contrast, TGF-β2 significantly increased the percentage of cells in the G1/G0 phase, from a basal level of 52.2% to 61.8% (Fig. 1D). The percentage of cells in the S phase was also increased by the FGF-2 treatment and was decreased by exposure to TGF-β2 (Fig. 1D). These data suggest that FGF-2 promotes BCE cells to pass the G1 to S phase whereas TGF-β2 prevents the cells from entering the S phase. Furthermore, TGF-β2 may block FGF-2induced S phase entry, leaving more cells in the G1 phase. Together, these results suggest that FGF-2 induces and TGF-β2

RNA interference Approximately 1 × 105 cells were used for each lipo-transfection. To prepare oligomer–Lipofectamine 2000 complexes for each well of 6-well plate, 80 pmol of siRNA oligomer was diluted in 200 μl serum-free medium. At the same time, Lipofectamine 2000 (Invitrogen) was diluted 4 ml in 200 ml medium. Both of these two reagents were mixed gently and incubated for 5 min at room temperature. Then, we combined

3636

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

suppresses BCE cell proliferation. Moreover, TGF-β2 can override FGF-2-stimulated BCE cell proliferation.

TGF-β2 increases COX-2 promoter activity and protein expression TGF-β2 may override FGF-2-stimulated cell proliferation by influencing intracellular signaling changes induced by FGF-2. TGF-β1 induces COX-2 expression in hepatic stellate cell [23]

and pulmonary artery smooth muscle cells [24]. To investigate whether COX-2 expression is regulated by TGF-β2, we cotransfected a COX-2 reporter COX-2-pGL2 plasmid with an internal control plasmid carrying a β-galactosidase gene. After overnight culture, BCE cells were starved in a serum-free medium for 24 h. When the BCE cells were treated with 10 ng/ ml of FGF-2 alone, the expression of the reporter gene was not changed (Fig. 2A). By contrast, TGF-β2 (5 ng/ml) significantly up-regulated the activity of the COX-2 reporter gene (Fig. 2A).

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

3637

Moreover, activation of the COX-2 promoter by TGF-β2 was not affected by FGF-2 (Fig. 2A). These results suggest that TGF-β2 up-regulates COX-2 gene expression in BCE cells. To determine whether TGF-β2 also increases COX-2 protein expression, we treated BCE cells with TGF-β2 for different periods of time and used Western blot analysis to measure COX-2 protein expression. Within 4 h of incubation time, TGFβ2 (10 ng/ml) was able to significantly up-regulate COX-2 protein expression (Fig. 2B). These results suggest that TGF-β2 is a positive regulator of COX-2 expression in BCE cells. To further verify the effect of COX-2 in TGF-β2-induced proliferation inhibition, a COX-2 overexpression construct was introduced into BCE cells. The overexpression of COX-2 protein was observed in the transfected cells, and cell proliferation induced by FGF-2 (10 ng/ml) in these cells was significantly decreased compared to the control cells (Fig. 2C). This result suggests that COX-2 has a similar role in overriding FGF-2-stimulated cell proliferation as TGF-β2 and that COX-2 takes part in TGF-β2-induced inhibition of BCE cell proliferation inhibition.

medium. A longer term effect of TGF-β2 on PGE-2 production was also observed, and PGE-2 production went even higher at the 120 min time point. There is evidence that PGE-2 levels could affect corneal endothelial cell proliferation since it is present in the aqueous humor. Its presence there is due to endothelial COX-2 activity, which mediates PGE-2 formation and release [25]. As TGF-β2 stimulates PGE-2 synthesis and inhibits corneal endothelial proliferation, the inhibitory effects of TGF-β2 on BCE cell proliferation could depend on COX-2-mediated production of PGE-2 [12]. To test this possibility, we determined the dose–response relationship of PGE-2 on FGF-2-induced increases in BCE cell proliferation. Similar to that for TGF-β2, increasing doses of PGE-2 dosedependently suppressed FGF-2-induced increase in BCE cell proliferation (Fig. 3C). Quantification of the proliferation index indicates that, like TGF-β2, PGE-2 at a concentration of 50 μM significantly reduced the mitogenic response to FGF2 (Fig. 3D). These results suggest that COX-2 and PGE-2 may be key mediators of the inhibitory effects of TGF-β2 on FGF-2induced stimulation of cell proliferation.

A COX-2-selective inhibitor and PGE-2 affect cell proliferation

FGF-2 stimulates cell proliferation through PI3-kinase/AKT signaling pathway

To investigate the functional significance of TGF-β2-induced COX-2 expression on BCE cell proliferation, we inhibited COX-2 activity with NS398, a non-steroidal anti-inflammatory selective potent COX-2 inhibitor, which reduces PGE-2 production. Increasing doses of NS398 reversed the inhibitory effects of TGF-β2 on BCE cell proliferation. This inhibitor had a maximal effect at 2 μM (Fig. 3A). Then, we measured time course for the secretion of PGE-2 that is one of the major products of COX-2 regulation in BCE cell culture medium after 10 ng/ml of TGF-β2 treatment (Fig. 3B). After 20 min of exposure to TGF-β2, PGE-2 secretion increased in the

To further understand the intracellular signaling pathways related to FGF-2-induced BCE cell proliferation, which can be suppressed by TGF-β2, we determined the time-dependent effects of this mitogen on AKT activation. BCE cells were synchronized in the G1 phase of the cell cycle by serum starvation and then FGF-2 was added at a final concentration of 10 ng/ml to stimulate cell proliferation. FGF-2 treatment activated AKT signaling pathway as indicated by increases in the level of p-AKT over a wide time window (Fig. 4A). AKT was activated 5 min after FGF-2 treatment, and it reached a high

Fig. 4 – Involvement of the PI3 kinase/AKT signaling pathway in FGF-2-stimulated BCE cell proliferation. (A) Activation of the AKT signaling pathway by FGF-2 treatment. BCE cells were treated with FGF-2 at 10 ng/ml, and AKT activation was analyzed by Western blotting using antibodies against phosphor-AKT and total AKT at the indicated time points. (B) Suppression effect of TGF-β2 on FGF-2-induced AKT activation. BCE cells were treated with TGF-β2 and FGF-2 at indicated doses, and AKT activation was analyzed by Western blotting at the indicated time points. Total AKT (t-AKT) levels were detected as loading controls. p-AKT protein intensities in Western blots were quantitatively scanned and plotted as the normalized density (upper panel). (C) Inhibitory effect of PGE-2 on AKT activation induced by FGF-2. 10 ng/ml of FGF-2 and 50 nM of PGE-2 were used. Expression levels of p-AKT were detected by Western blot analysis. t-AKT and β-actin levels were detected as loading controls. p-AKT protein intensities were quantitatively scanned and plotted as the normalized density (upper panel). (D) Effect of PGE-2 on Erk1/2 activation induced by FGF-2. 10 ng/ml of FGF-2 and 50 nM of PGE-2 were used. Expression levels of p-Erk1/2 were detected by Western blot analysis. t-Erk1/2 and β-actin levels were detected as loading controls. p-Erk1/2 protein intensities were quantitatively scanned and plotted as the normalized density (right upper panel). (E) Effect of PI3K and MEK inhibitors on FGF-2-induced BCE cell proliferation. LY294002 and PD98059 at the indicated doses were applied to the cells with FGF-2 (10 ng/ml), respectively. On day 3, BCE cell proliferation was determined by the MTT cell proliferation assay. C in panels C, D and E indicates baseline controls. (F) Effect of knockdown of AKT-1 mRNA on FGF-2-induced BCE cell proliferation. AKT-1-specific siRNA was used to knockdown AKT-1. Expression levels of endogenous AKT protein were detected by Western blot analysis in control and AKT-1 siRNA-transfected BCE cells. β-actin levels were detected as loading controls (left panel). FGF-2 (10 ng/ml) was applied to the control and AKT-1 siRNA transfected BCE cells, and cell proliferation was determined on day 3 by the MTT cell proliferation assay (right panel). (G) Effect of overexpression of a mutant form of AKT on FGF-2-induced BEC cell proliferation. An AKT-1 plasmid was used to overexpress AKT-1. Expression levels of AKT proteins were detected by Western blot analysis in control and AKT-1 transfected BCE cells. β-actin levels were detected as loading controls (left panel). FGF-2 (10 ng/ml) was applied to the control and AKT-1 transfected BCE cells, and cell proliferation was determined on day 3 by the MTT cell proliferation assay (right panel). All values were represented as mean ± SEM from triplicate samples (n = 3). The symbol * represents a significant difference (p < 0.05) compared with control groups.

3638

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

ment, PGE-2 significantly suppressed FGF-2-induced increases in AKT activity at a concentration (Fig. 4C) that also suppressed BCE cell proliferation, and this suppression lasted for at least 30 min. PGE-2 did not have an obvious effect on the activation of Erk1/2 in response to FGF-2 induction (Fig. 4D). To directly investigate the contribution of the AKT and ERK signaling pathways to BCE cell proliferation, we cultured BCE cells in the presence of FGF-2 and different LY294002 concentrations, a PI3-kinase inhibitor, or the MEK1 inhibitor PD98059. LY294002 decreased FGF-2-induced increases in BCE cell proliferation in a dose-dependent manner (Fig. 4E). In contrast, PD98059 had only a slight inhibitory effect on FGF2-stimulated BCE cell proliferation (Fig. 4E). To verify the role of AKT in FGF-2-stimulated signal pathways, the expression of AKT mRNA was knocked down by RNA interference using RNA primers specific to AKT-1. Knockdown of AKT mRNA effectively blocked the effect of FGF-2 on BCE cell proliferation (Fig. 4F). Furthermore, an AKT-dominant negative construct was transfected and overexpressed in BCE cells. Dominant negative expression of AKT significantly suppressed BCE cell proliferation induced by 10 ng/ml FGF-2 (Fig. 4G). Taken together, these results suggest that FGF-2-induced BCE cell proliferation occurs primarily through activating the PI3kinase/AKT signaling pathway, but not the ERK signaling pathway.

cAMP inhibits FGF-2-induced AKT activation

Fig. 5 – Effects of cAMP on FGF-2-induced AKT activation and BCE cell proliferation. (A) Inhibitory effects of cAMP analogs on FGF-2-induced AKT activation. BCE cells were cultured in the presence of FGF-2 (10 ng/ml) and DMSO, forskolin or SP-cAMP at the indicated concentrations, and AKT activation was analyzed by Western blotting. t-AKT proteins were detected as loading controls. p-AKT protein intensities were quantitatively scanned and plotted as the normalized density (upper panel). (B) Effects of cAMP analogs on FGF-2-induced cell proliferation. BCE cells were cultured and treated as above, and cell proliferation was determined with the MTT cell proliferation assay. All values were represented as mean ± SEM from triplicate samples (n = 3). The symbol * represents a significant difference (p < 0.05) compared with control groups.

level after 30 min. Moreover, AKT activity remained high 3–6 h after the treatment and decreased afterwards (Fig. 4A). We next investigated the influence of TGF-β2 and PGE-2 on FGF-2-induced AKT activation. TGF-β2 greatly and dosedependently suppressed FGF-2-induced AKT activation at both 45 min and 6 h time points (Fig. 4B). To assess whether PGE-2 could suppress FGF-2-induced signaling events, we measured the effect of exogenous PGE-2 on FGF-2-induced AKT and Erk1/2 activation. Similar to that after TGF-β2 treat-

PGE-2 increases intracellular cAMP levels. Thus, changes in intracellular levels of cAMP can influence FGF-2-induced AKT activation and BCE cell proliferation. We tested this assumption by determining whether forskolin, its vehicle or SP-cAMP, a cAMP hydrolysis inhibitor, could counter the mitogenic effect of 10 ng/ml FGF-2. Vehicle treatment did not affect FGF2-induced AKT activation, whereas both forskolin and SPcAMP treatment led to a significant suppression in AKT activation (Fig. 5A). When the intracellular levels of cAMP were increased by exposure to forskolin, IMBX, 8-CPT-cAMP or SP-cAMP, FGF-2-induced increases in cell proliferation were all significantly inhibited (Fig. 5B). Therefore, the reduced AKT activation induced by increasing levels of cAMP correlated with an inhibition of FGF-2-induced BCE cell proliferation.

Discussion Under normal physiological conditions, FGF-2 is a component of Descemet's membrane and the corneal endothelial layer is constantly bathed in aqueous humor containing TGF-β2 [6,7,26,27]. Thus, corneal endothelial proliferative activity is determined by the net effect of these cytokines that induce opposing effects on proliferation of this tissue layer. The capacity of human corneal endothelium to proliferate is limited in vivo. This hindrance accounts for why endothelial cells undergo enlargement to compensate for losses in endothelial cell layer integrity following injury. Such a compensatory mechanism can only partially restore corneal endothelium function. In this context, we addressed how TGF-β2 might suppress FGF-2-mediated increases in corneal endothelial cell

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

proliferation. We found that TGF-β2 blocks FGF-2-induced BCE cell proliferation and inhibits the PI3-kinase/AKT signaling pathway. Our studies indicate that FGF-2 alone can increase whereas TGF-β2 alone blocks BCE cell proliferation. These opposing effects are consistent with previous reports that FGF-2 induces a mitogenic response whereas TGF-β2 inhibits proliferation in epithelial, vascular endothelial and blood cells [28–30]. Interestingly, we found that FGF-2-induced mitogenic response in BCE cell is blocked by TGF-β2. Moreover, FGF-2 promotes BCE cells to pass from the G1 to S phase whereas TGF-β2 prevents the cells from entering the S phase. Furthermore, TGF-β2 blocks FGF-2-induced S phase entry, leaving more cells in the G1 phase. To understand how this interaction between FGF-2 and TGF-β2 in BCE cells occurs, we focused on intracellular signaling pathways induced by these two cytokines. It has been shown that FGF-2 uses both PLC-γ1 and PI3-kinase to induce a mitogenic response in corneal endothelial cells [8]. On the other hand, TGF-β2 inhibits this response through increases in PGE-2 expression in the corneal endothelium [12]. We thus hypothesized that TGF-β2 inhibits FGF-2induced cell proliferation through activating COX-2 and the synthesis of PGE-2. Consistent with this assumption, exogenous TGF-β2 induced the expression of a reporter gene that is driven by a COX-2 promoter whereas FGF-2 did not affect the expression of the reporter gene in the presence or absence of TGF-β2. Moreover, TGF-β2 can directly activate COX-2 gene expression. These results indicate that the expression of COX-2 is stimulated by TGF-β2 but not FGF-2 in BCE cells. The synthesis of PGE-2, a downstream product of COX-2, is increased after TGF-β2 treatment. About 20 min after exposure to TGF-β2, the secretion of PGE-2 significantly increased, and this is consistent with the report that long time stimulation of TGF-β2 dramatically activates PGE-2 secretion [12]. Furthermore, we found that a COX-2-specific inhibitor can reverse the inhibitory effect of TGF-β2 on BCE cell proliferation while both overexpression of COX-2 and exogenous PGE-2 significantly reduced FGF-2-induced BCE cell proliferation. These findings support the hypothesis that BCE cells produce COX-2 and PGE-2 after TGF-β2 treatment, which leads to the blockade of FGF-2-stimulated cell proliferation. FGF-2 may increase cell proliferation by activating different cell signaling pathways in different cell types, such as the PI3kinase/AKT pathway [31], and recruit complexes that tether together both PI3-kinase and MAPK/ERK pathways [32]. Consistent with the previous finding that activation of PI3kinase pathway may be involved in FGF-2-promoted cell proliferation, FGF-2 induced the activation of AKT in a timedependent manner. The ERK signaling pathway is also activated by FGF-2 but not suppressed by PGE-2 treatment. In our studies, LY294002, the PI3-kinase inhibitor, dosedependently blocked BCE cell proliferation in the presence of FGF-2. Overexpression of a dominant negative AKT or knockdown of endogenous AKT-1 mRNA by RNA interference reduced BCE cell proliferation stimulated by FGF-2. In contrast, the MEK inhibitor PD98059 had minimal effects on FGF-2induced cell proliferation. These findings suggest that PI3kinase/AKT pathway is a major contributor for FGF-2-stimu-

3639

lated BCE cell proliferation, and the MAPK/ERK is not prominently involved in BCE cell proliferation. Further supporting the notion that TGF-β2 suppresses FGF2-induced BCE cell proliferation by down-regulating the PI3kinase/AKT signaling pathway via PGE-2, we found that exogenous PGE-2 rapidly decreased the activation of AKT induced by FGF-2. PGE-2 treatment increased intracellular cAMP levels, and the elevation of cAMP in BCE cells resulted in both a reduction in FGF-2-induced AKT activation and cell proliferation. Taken together, our findings suggest that TGFβ2 likely down-regulates FGF-2-induced signaling in BCE cells and consequently may suppress BCE cell proliferation. TGF-β2 may do so by up-regulating the expression of COX-2 and the production of PGE-2 and cAMP, and this series of events may collectively inhibit the activation of the PI3-kinase/AKT signaling pathway.

Acknowledgments We thank the members in the joint laboratory for helpful discussions. We are also grateful to Xueyu Chen for assistance. This work was supported in part by a core funding from Chinese Academy of Sciences (CAS), a One Hundred Talent Grant from CAS to MX, the E-Institutes of Shanghai Municipal Education Commission Project Number E03003, the Science and Technology Commission of Shanghai Municipality (04DZ14902), a grant from the Shanghai Education Commission (03BZ03) and a National Key Program grant (973) (NO2002CB512805).

REFERENCES

[1] D.M. Maurice, The location of the fluid pump in the cornea, J. Physiol. 221 (1972) 43–54. [2] D.L. Van Horn, R.A. Hyndiuk, Endothelial wound repair in primate cornea, Exp. Eye Res. 21 (1975) 113–124. [3] K.D. Bell, R.J. Campbell, W.M. Bourne, Pathology of late endothelial failure: late endothelial failure of penetrating keratoplasty: study with light and electron microscopy, Cornea 19 (2000) 40–46. [4] S. Mishima, Clinical investigations on the corneal endothelium—XXXVIII Edward Jackson Memorial Lecture, Am. J. Ophthalmol. 93 (1982) 1–29. [5] Y. Ishino, Y. Sano, T. Nakamura, C.J. Connon, H. Rigby, N.J. Fullwood, S. Kinoshita, Amniotic membrane as a carrier for cultivated human corneal endothelial cell transplantation, Invest. Ophthalmol. Visual Sci. 45 (2004) 800–806. [6] E.P. Kay, S.Y. Park, M.K. Ko, S.C. Lee, Fibroblast growth factor 2 uses PLC-gamma1 for cell proliferation and PI3-kinase for alteration of cell shape and cell proliferation in corneal endothelial cells, Mol. Vis. 4 (1998) 22. [7] E.P. Kay, X. Gu, R.E. Smith, Corneal endothelial modulation: bFGF as direct mediator and corneal endothelium modulation factor as inducer, Invest. Ophthalmol. Visual Sci. 35 (1994) 2427–2435. [8] H.T. Lee, T.Y. Kim, E.P. Kay, Cdk4 and p27Kip1 play a role in PLC-gamma1-mediated mitogenic signaling pathway of 18 kDa FGF-2 in corneal endothelial cells, Mol. Vis. 8 (2002) 17–25. [9] Y. Ochiai, H. Ochiai, Higher concentration of transforming growth factor-beta in aqueous humor of glaucomatous eyes and diabetic eyes, Jpn. J. Ophthalmol. 46 (2002) 249–253.

3640

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 6 3 1 –36 4 0

[10] Y. Shi, J. Massague, Mechanisms of TGF-beta signaling from cell membrane to the nucleus, Cell 113 (2003) 685–700. [11] K.H. Chen, D.L. Harris, N.C. Joyce, TGF-beta2 in aqueous humor suppresses S-phase entry in cultured corneal endothelial cells, Invest. Ophthalmol. Visual Sci. 40 (1999) 2513–2519. [12] K.H. Chen, W.M. Hsu, C.C. Chiang, Y.S. Li, Transforming growth factor-beta2 inhibition of corneal endothelial proliferation mediated by prostaglandin, Curr. Eye Res. 26 (2003) 363–370. [13] K.E. Bachman, B.H. Park, Duel nature of TGF-beta signaling: tumor suppressor vs. tumor promoter, Curr. Opin. Oncol. 17 (2005) 49–54. [14] J. Massague, How cells read TGF-beta signals, Nat. Rev., Mol. Cell Biol. 1 (2000) 169–178. [15] J. Massague, Y.G. Chen, Controlling TGF-beta signaling, Genes Dev. 14 (2000) 627–644. [16] J. Massague, D. Wotton, Transcriptional control by the TGF-beta/Smad signaling system, EMBO J. 19 (2000) 1745–1754. [17] E. Rosado, Z. Schwartz, V.L. Sylvia, D.D. Dean, B.D. Boyan, Transforming growth factor-beta1 regulation of growth zone chondrocytes is mediated by multiple interacting pathways, Biochim. Biophys. Acta 1590 (2002) 1–15. [18] A. Eldor, I. Vlodavsky, E. Hy-Am, R. Atzmon, B.B. Weksler, A. Raz, Z. Fuks, Cultured endothelial cells increase their capacity to synthesize prostacyclin following the formation of a contact inhibited cell monolayer, J. Cell. Physiol. 114 (1983) 179–183. [19] H.E. Bazan, The synthesis and effects of eicosanoids in avascular ocular tissues, Prog. Clin. Biol. Res. 312 (1989) 73–84. [20] H.T. Lee, E.P. Kay, Regulatory role of cAMP on expression of Cdk4 and p27(Kip1) by inhibiting phosphatidylinositol 3-kinase in corneal endothelial cells, Invest. Ophthalmol. Visual Sci. 44 (2003) 3816–3825. [21] T.Y. Kim, W.I. Kim, R.E. Smith, E.D. Kay, Role of p27(Kip1) in cAMP- and TGF-beta2-mediated antiproliferation in rabbit corneal endothelial cells, Invest. Ophthalmol. Visual Sci. 42 (2001) 3142–3149. [22] H.T. Lee, E.P. Kay, Regulatory role of PI 3-kinase on expression of Cdk4 and p27, nuclear localization of Cdk4, and phosphorylation of p27 in corneal endothelial cells, Invest. Ophthalmol. Visual Sci. 44 (2003) 1521–1528.

[23] A.Y. Hui, A.J. Dannenberg, J.J. Sung, K. Subbaramaiah, B. Du, P. Olinga, S.L. Friedman, Prostaglandin E2 inhibits transforming growth factor beta 1-mediated induction of collagen alpha 1(I) in hepatic stellate cells, J. Hepatol. 41 (2004) 251–258. [24] K.K. Sheares, T.K. Jeffery, L. Long, X. Yang, N.W. Morrell, Differential effects of TGF-beta1 and BMP-4 on the hypoxic induction of cyclooxygenase-2 in human pulmonary artery smooth muscle cells, Am J. Physiol.: Lung Cell Mol. Physiol. 287 (2004) L919–L927. [25] N.C. Joyce, S.J. Joyce, S.M. Powell, B. Meklir, EGF and PGE2: effects on corneal endothelial cell migration and monolayer spreading during wound repair in vitro, Curr. Eye Res. 14 (1995) 601–609. [26] N. Landshman, M. Belkin, I. Ben-Hanan, O. Ben-Chaim, E. Assia, N. Savion, Regeneration of cat corneal endothelium induced in vivo by fibroblast growth factor, Exp. Eye Res. 45 (1987) 805–811. [27] I. Vlodavsky, J. Folkman, R. Sullivan, R. Fridman, R. Ishai-Michaeli, J. Sasse, M. Klagsbrun, Endothelial cell-derived basic fibroblast growth factor: synthesis and deposition into subendothelial extracellular matrix, Proc. Natl. Acad. Sci. U. S. A. 84 (1987) 2292–2296. [28] M.B. Datto, Y. Li, J.F. Panus, D.J. Howe, Y. Xiong, X.F. Wang, Transforming growth factor beta induces the cyclin-dependent kinase inhibitor p21 through a p53-independent mechanism, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 5545–5549. [29] G.J. Hannon, D. Beach, p15INK4B is a potential effector of TGF-beta-induced cell cycle arrest, Nature 371 (1994) 257–261. [30] K. Polyak, J.Y. Kato, M.J. Solomon, C.J. Sherr, J. Massague, J.M. Roberts, A. Koff, p27Kip1, a cyclin-Cdk inhibitor, links transforming growth factor-beta and contact inhibition to cell cycle arrest, Genes Dev. 8 (1994) 9–22. [31] J.H. Qi, T. Matsumoto, K. Huang, K. Olausson, R. Christofferson, L. Claesson-Welsh, Phosphoinositide 3 kinase is critical for survival, mitogenesis and migration but not for differentiation of endothelial cells, Angiogenesis 3 (1999) 371–380. [32] S.H. Ong, Y.R. Hadari, N. Gotoh, G.R. Guy, J. Schlessinger, I. Lax, Stimulation of phosphatidylinositol 3-kinase by fibroblast growth factor receptors is mediated by coordinated recruitment of multiple docking proteins, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 6074–6079.