Role of Transforming Growth Factor β in Rat Bladder Smooth Muscle Cell Proliferation

Urinary bladder outlet obstruction (BOO) is a frequent consequence of benign prostatic enlargement or urethral strictures. It leads to an increased bladder size and micturition disorders (Andersson, 2003), the former involving both hypertrophy and hyperplasia at the cellular level. However, the mediators and molecular mechanisms leading to bladder hypertrophy have remained elusive. Studies in rats (Chul Kim et al., 2001) and patients with BOO (MacRae Dell et al., 2000; Monga et al., 2001) have found an increased urinary excretion of transforming growth factor (TGF) β. Hence, TGFβ has been proposed to play a role in the pathophysiology of BOO and bladder hypertrophy.

TGFβ is a pluripotent growth factor that has been implicated in a variety of physiological processes such as proliferation, apoptosis, phenotypic switching, differentiation, and specification of developmental fate (Massagué, 2000). Interestingly, TGFβ can have differential or even opposite effects depending on the tissue or cell type under investigation (Roberts and Sporn, 1993). For example, TGFβ can increase the proliferation of airway smooth muscle cells (Chen and Khalil, 2006), but it potently inhibits proliferation in vascular smooth muscle cells due to a G₀/G₁ arrest via the p38 pathway (Seay et al., 2005). In the heart, TGFβ can induce cardiomyocyte hypertrophy, whereas cardiac fibroblasts respond with differentiation into myofibroblasts and synthesis of collagen (Liu et al., 2006; Watkins et al., 2006). Therefore, it remains largely unclear whether increased TGFβ excretion in experimental and clinical BOO reflects a role as mediator of cell growth, as part of a negative feedback loop limiting cell growth, or as a by-product that at best can be used as a biomarker.

The heterogeneity of TGFβ responses may at least partly be explained by the existence of multiple TGFβ isoforms as well as a complex signal transduction cascade. Thus, three mammalian isoforms of TGFβ exist, i.e., TGFβ1, TGFβ2, and TGFβ3, which exhibit partly overlapping as well as isoform-specific functions (Azhar et al., 2003; Lebrin et al., 2005; Akutsu et al., 2006). Unfortunately, many reports in the field have not specified which isoform has been used. All three isoforms bind to heteromeric receptor complexes consisting of type I and II transmembrane proteins with intrinsic serine-threonine kinase activity. These complexes activate two principal signaling pathways. Originally, it was assumed that the main signaling occurs via SMAD proteins, which translocate into the nucleus to regulate gene transcription (Feng and Derynck, 2005). More recently, a parallel signaling pathway of TGFβ became evident that operates independently of SMAD proteins and rather involves one or more members of the family of mitogen-activated protein kinases (MAPKs) (Roberts, 1998; Derynck and Zhang, 2003). This includes the extracellular signal-regulated kinase (ERK), c-Jun NH₂-terminal kinase (JNK), and p38 subfamilies, all of which have been implicated in the regulation of cell growth and apoptosis.

Our lack of understanding of the role of TGFβ in the urinary bladder may partly relate to species differences. Thus, in rabbits acute BOO reduces TGFβ expression (Chen et al., 1994), whereas relief of a longer lasting BOO increases TGFβ expression (Levin et al., 1994). Conversely, the urinary excretion of TGFβ is increased in rats (Chul Kim et al., 2001) and patients with BOO (MacRae Dell et al., 2000; Monga et al., 2001). Species differences have also been proposed at the functional level. Thus, on the one hand, a hypertrophic and fibrotic response was found in cultured human bladder smooth muscle cells (BSMCs) upon addition of exogenous TGFβ (Howard et al., 2005). On the other hand, a similar response, i.e., a hypertrophied lamina propria and muscularis externa with myofibroblast differentiation and proliferation, has been observed by removal of TGFβ tone, i.e., in TGFβ receptor type II gene knockout mice (Sharif-Afshar et al., 2005). In rats, TGFβ may act as a mediator of mesenchymal-epithelial interactions necessary for the differentiation of BSMCs during embryonic development (Baskin et al., 1996).

Against this background, we have investigated the role of TGFβ isoforms in the modulation of cell growth and death in cultured rat BSMCs. In this regard, we have specifically focused on the role of MAPKs because they are considered as universal regulators of cell growth and apoptosis (Widmann et al., 1999) and because cell-permeable, low-molecular-weight inhibitors of various MAPKs are available that allow to determine their role in functional effects.

Materials and Methods

Chemicals. RPMI 1640 medium, fetal calf serum (FCS), penicillin, streptomycin, 0.4% trypan blue stain (5×), and calcium-free phosphate-buffered saline (PBS) were from Invitrogen (Breda, The Netherlands). Trypsin, collagenase (type IV; Worthington Biochemicals, Freehold, NJ), TGFβ type I receptor antagonist SB 431542 and specific inhibitors of ERK activation (PD 98059), JNK (SP 600125), and p38 (SB 203580) were purchased from Sigma-Aldrich Chemie GmbH (Schnelldorf, Germany). Human recombinant TGFβ1, TGFβ2, and TGFβ3 were from R&D Systems Europe, Ltd. (Abingdon, UK). p38, phospho-p38, JNK, phospho-JNK, ERK, and phospho-ERK antibodies were purchased from Cell Signaling Technology Inc. (Danvers, MA).

Cell Culture. Rat BSMCs were harvested using an enzymatic dispersion method (Ma et al., 2002) with minor modifications. Bladders from male Wistar rats (240–260 g) were excised. After discarding the bladder base and removing the outer connective and adipose tissue and the inner urothelial layers with a cotton swab, the bladder body was incubated at 37°C for 30 min in PBS containing 0.2% trypsin. The tissue was cut into 1- to 3-mm pieces and incubated with 0.1% collagenase in RPMI 1640 medium for 30 min at 37°C. The resulting cell suspension was centrifuged at 250g at 4°C for 5 min. The supernatant was discarded, and the cells were resuspended in RPMI 1640 medium containing 10% FCS by repeated agitation with a pipette, and then the cells were centrifuged again at 250g for 2 min at 4°C. The supernatant was placed in a culture dish, and the cells from the pellet were resuspended in RPMI 1640 medium and centrifuged at 125g at 4°C for 2 min. The supernatant was added to that of the previous centrifugation, and cells were cultured in plastic flasks in RPMI 1640 medium containing 10% FCS in a humidified atmosphere containing 5% CO₂ at 37°C. Based upon α-actin staining, >95% of the cultured cells were smooth muscle cells. Cells of passage 2 and 3 with approximately 90% confluence at time of harvesting were used for all experiments; under confluent conditions, these cells did not differ morphologically from cells of passage 0 (data not shown). In some cases, cells were serum-starved for 24 h before addition of TGFβ. Cells were harvested by washing with PBS and trypsin solution, and in some cases centrifugation at 200g for 5 min.

Flow Cytometric Analysis. To assess cell size and the presence of apoptosis, flow cytometry was performed on a FACScan using an annexin V staining protocol (BD Biosciences, Sunnyvale, CA). Serum-starved cells were treated with either 10% FCS or FCS plus 10^–9 g/ml of the TGFβ isoforms for 24, 48, 72, or 96 h. The cells were incubated for 15 min in PBS containing annexin V-FLUOS and propidium iodide labeling solution according to the manufacturer's protocol. Forward scatter parameter analysis of 10,000 cells per sample was performed to estimate the size of the viable cells. Simultaneous dual parameter analysis of annexin V and propidium iodide was used to estimate the percentage of apoptotic cells defined as annexin V-positive and propidium iodide-negative; viable cells were defined as those that were negative for annexin V and propidium iodide staining, whereas necrotic cells were defined as propidium iodide- and annexin V-positive (van der Poel, 2005).

Mitochondrial Activity. Mitochondrial activity as an integrated measure of cell number and size was measured using the CellTiter 96 AQueous One proliferation assay (Promega, Madison, WI) according to the manufacturer's protocol. Serum-starved BSMCs were treated with either 10% FCS or the TGFβ isoforms in the absence or presence of 10% FCS for 48 h.

5-Bromo-2′-deoxyuridine Incorporation. DNA synthesis, as a second measure of cellular proliferation, was assessed as BrdU incorporation in a commercially available enzyme-linked immunosorbent assay (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's protocol. Serum-starved cells were treated with the indicated agents in the absence or presence of 10% FCS for 48 h, with 10 μM BrdU being present during the final 24 h. To validate our method, we have reproduced the previous observation by others (Seay et al., 2005) that TGFβ inhibits DNA synthesis in aortic smooth muscle cells (data not shown).

Manual Counting. Serum-starved cells were treated with the TGFβ isoforms (10^–9 g/ml) in the presence of 10% FCS for 48 h. After harvesting, they were collected by centrifugation and resuspended in RPMI 1640 medium. Viable cells were identified as those excluding trypan blue by light microscopy in a Burker-Turk hemocytometer (Emergo, Landsmeer, The Netherlands).

Western Blot Analyses. Serum-starved cells were treated for 10 min in the absence or presence of 10% FCS or FCS plus the indicated TGFβ isoforms (10^–9 g/ml). Incubations were stopped by washing with ice-cold PBS and incubating with 500 μl of extraction buffer [20 mM Tris, 1 mM EDTA, 1 mM EGTA, 1% (v/v) Triton X-100, 150 mM NaCl, 1 mM Na₃VO₄, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, and 1× protease inhibitor cocktail, pH 7.5 (Pierce Chemical, Rockford, IL)] for 10 min at 4°C. Cells were scraped and spun down for 10 min at 15,000g. Protein concentrations of the supernatant were determined using the BCA protein assay kit (Pierce Chemical), according to the manufacturer's instructions. SDS-polyacrylamide gel electrophoresis and Western blotting were performed using the NuPAGE electrophoresis system (4–12% bis-Tris polyacrylamide gel) from Invitrogen (Carlsbad, CA). The blots were washed with TBS/T [Tris-buffered saline + 0.1% (v/v) Tween 20] and incubated with blocking solution [TBS/T + 5% (w/v) chicken serum] for 1 h at room temperature. Afterward, the blots were incubated overnight at 4°C with primary antibodies against total or phosphorylated forms of ERK, JNK, and p38 in TBS/T containing 1% (w/v) chicken serum. Blots were washed 3 × 5 min with TBS/T at room temperature, and the donkey anti-rabbit IgG horseradish peroxidase-conjugated antibody (1:10,000; GE Healthcare, Little Chalfont, Buckinghamshire, UK) was incubated for 1 h. After washing for 3 × 5 min, chemiluminescent detection by enhanced chemiluminescence (Roche Diagnostics, Basel, Switzerland) was performed, according to the manufacturer's protocol. Band intensity on the blots was analyzed by two-dimensional densitometry (ImageJ, version 1.37; Wayne Rasband, National Institutes of Health, Bethesda, MD; http://rsb.info.nih.gov/ij/).

Terminal Deoxynucleotidyl Transferase dUTP Nick-End Labeling Measurements. Nuclear DNA fragmentation as a measure of apoptosis was measured by the DeadEnd Fluorometric TUNEL system (Promega) according to the manufacturer's protocol. Cells were grown on eight-well Lab-Tek chamber slides (Nalge Nunc International, Rochester, NY) in RPMI 1640 medium containing 10% FCS in a humidified 5% CO₂ atmosphere at 37°C. After 24 h, cells were serum-starved for 24 h. Thereafter, they were incubated for 48 h with the TGFβ isoforms (10^–9 g/ml) in the absence or presence of 10% FCS. After washing, cell nuclei were stained with VECTASHIELD 4,6-diamidino-2-phenylindole nuclear stain in mounting medium (Vector Laboratories, Burlingame, CA), and fluorescence microscopy was used to detect green fluorescence against a blue background to indicate apoptosis.

Data Analysis. Data are presented as the mean ± S.E.M. of n experiments or as a representative of at least three separate experiments with similar results. Statistical analysis (one-sample t test and ANOVA with post hoc Dunnett's multiple comparison test) was performed using GraphPad Prism 4.0 (GraphPad Software Inc., San Diego, CA).

Results

MAP Kinase Phosphorylation. Neither the presence of FCS nor the additional presence of any of the three TGFβ isoforms affected the abundance of total immunoreactive ERK, JNK, and p38 during 10-min incubation (Fig. 1). Although 10% FCS did not affect the phosphorylation of p38 to a major extent, it markedly increased the phosphorylated forms of ERK and JNK (Fig. 1). In the presence of FCS, all three TGFβ isoforms markedly reduced the phosphorylation of ERK, JNK, and p38 to levels even below those seen in the absence of FCS, although this did not reach statistical significance for the effect of TGFβ1 on p38 (Fig. 1).

Apoptosis. Based upon a TUNEL assay, none of the three TGFβ isoforms (10^–9 g/ml) induced apoptosis in the absence or presence of 10% FCS during a 48-h incubation, whereas DNase treatment induced clear fluorescent signals in almost all treated cells (three experiments each; data not shown). In quantitative experiments with an annexin V-based FACS method 0.39 to 6.80% of cells were identified as being apoptotic after 24 to 96 h of incubation in the presence of 10% FCS; the three TGFβ isoforms did not significantly enhance the percentage of apoptotic cells except for TGFβ2 at the 72-h time point (Fig. 2). At the 96-h time point, TGFβ1 caused a significant decrease of apoptotic cells (Fig. 2).

DNA Synthesis. To determine possible effects on DNA synthesis, we have measured TGFβ effects on BrdU incorporation (Fig. 3; Table 1). All three TGFβ isoforms concentration-dependently inhibited BrdU incorporation with similar potency, and they similarly caused a maximal reduction of approximately 50%. Although basal BrdU incorporation was markedly higher (15.1-fold; n = 9) in the presence than in the absence of 10% FCS, the effects of all three TGFβ isoforms were quantitatively similar under both experimental conditions (Table 1).

The TGFβ receptor type I antagonist SB 431542 (10 μM) enhanced basal BrdU incorporation in the absence and presence of 10% FCS by approximately 4.6- and 3.5-fold (n = 9 each), respectively. SB 431542 abolished the inhibitory effects of all three TGFβ isoforms in the presence of FCS (Fig. 3) and also in its absence (data not shown), demonstrating that the TGFβ effects were receptor-mediated.

In the presence of 10% FCS, the p38 inhibitor SB 203580 (10 μM) enhanced basal DNA synthesis by 69 ± 9% (n = 9), whereas PD 98059 (10 μM), an inhibitor of ERK activation, and the JNK inhibitor SP 600125 (10 μM) reduced basal BrdU incorporation by 44 ± 2 and 56 ± 5%, respectively (all p < 0.0001). Alternatively, none of the three inhibitors significantly altered the potency or efficacy of any of the three TGFβ isoforms to inhibit DNA synthesis (Table 2), demonstrating that inhibition of DNA synthesis by TGFβ did not involve any of the three MAPKs.

Cell Growth. To assess consequences of the inhibition of DNA synthesis by TGFβ isoforms, their effect on the total number of viable cells was determined during 48-h incubation. The number of cells counted in the presence of TGFβ1, TGFβ2, or TGFβ3 (10^–9 g/ml each) was 100, 103, and 103%, respectively, of those found in the absence of TGFβ in paired experiments (n = 3; not significant).

To quantify possible TGFβ isoform effects on cell size, forward scatter was determined by FACS analysis (Fig. 4). During incubations of 24 to 96 h, none of the TGFβ isoforms (10^–9 g/ml each) affected cell size. Similarly, none of the TGFβ isoforms (10^–9 g/ml each) affected mitochondrial activity as measured by the AQueous One assay during a 48-h incubation in the absence or presence of 10% FCS (data not shown), whereas the presence of FCS increased basal activity in this assay from 1.40 ± 0.03 to 1.93 ± 0.03 arbitrary units (n = 3). Finally, the three TGFβ isoforms (10^–9 g/ml each) also failed to cause major morphological alterations of the BSMCs during a 48-h incubation (Fig. 5).

Discussion

An enhanced urinary excretion of TGFβ has repeatedly been shown in rats and patients with BOO (MacRae Dell et al., 2000; Chul Kim et al., 2001; Monga et al., 2001). As the functional role of TGFβ in bladder hypertrophy remains largely unclear, we have used primary cultures of rat BSMCs to investigate the functional effects of TGFβ isoforms on growth properties of these cells. In line with results from several other cell types (Sterle et al., 1997; Seay et al., 2005; Akutsu et al., 2006), we have not detected any effects of TGFβ on BSMC apoptosis using two independent methods. Therefore, we have subsequently investigated whether TGFβ affects parameters of cell growth.

TGFβ concentration-dependently inhibited BSMC DNA synthesis. In agreement with reports on human BSMCs (Howard et al., 2005), this inhibition of DNA synthesis in rat BSMC was not reflected by reductions in cell number. Although a discrepancy between reduced DNA synthesis and unaltered cell number may be surprising, this is most likely related to the very slow doubling of BSMCs (Malkowicz et al., 1995; Kropp et al., 1999; Boselli et al., 2002), which makes it technically difficult to detect reduced cell numbers unless cells are observed for very long time courses. The inhibition of DNA synthesis may reflect a TGFβ-induced cell cycle arrest, which has been observed in many other cell types (Ewen et al., 1993; Reddy and Howe, 1993; Reynisdóttir et al., 1995; Wiesmann et al., 2000; Seay et al., 2005).

Although some studies have shown quantitatively or even qualitatively different effects of the TGFβ isoforms (Lebrin et al., 2005), all three isoforms behaved qualitatively and quantitatively similarly with regard to DNA synthesis in rat BSMCs. This shared effect of the three isoforms did not reflect unspecific action because their potencies were quite similar to those in other cell types (Reddy and Howe, 1993; Seay et al., 2005; Kapoun et al., 2006). More importantly, all isoforms were similarly inhibited by the specific TGFβ receptor antagonist SB 431542 (Herrmann et al., 2006; Ninomiya et al., 2006). It is also noteworthy that the inhibitory effects of TGFβ were similarly seen regardless of the basal rate of DNA synthesis, i.e., TGFβ was similarly potent and efficacious in the absence or presence of FCS, a finding that is in line with the proposed cell cycle arrest.

The signal transduction of TGFβ receptors involves two parallel pathways, i.e., SMAD proteins and MAPK (Derynck and Zhang, 2003). We have focused on MAPK because they are well established to play a role in cellular growth control (Widmann et al., 1999) and because selective inhibitors are available to functionally study their role in growth responses. All three TGFβ isoforms inhibited phosphorylation and hence activity of all three MAPKs. In many cell types, ERK, JNK, and p38 exert differential effects on cellular growth (Widmann et al., 1999). This was confirmed for rat BSMCs in our study, because inhibition of p38 increased, whereas inhibition of ERK and JNK decreased basal DNA synthesis, respectively. Nevertheless, inhibition of any of the MAPK did not alter the potency or efficacy of any of the three TGFβ isoforms to inhibit DNA synthesis. Taken together, these findings suggest that pathways distinct from MAPKs, possibly a SMAD pathway, are involved in the inhibition of DNA synthesis by TGFβ isoforms in rat BSMCs. Although activation of a SMAD pathway is a ubiquitous signaling response to TGFβ, a functional involvement of this pathway is difficult to test experimentally since interference with SMAD pathways requires transfection and/or cellular permeabilization techniques, which are notoriously difficult with primary BSMC cultures.

Studies with human BSMCs have demonstrated that TGFβ1 can promote hypertrophy at the cellular level and can also enhance extracellular matrix deposition (Howard et al., 2005). In contrast, we did not observe TGFβ-induced increases in cell size as measured by FACS or microscopic inspection or in cell function as measured by mitochondrial activity in our rat BSMCs, possibly reflecting the previously reported species differences in TGFβ expression and function in the bladder (Chen et al., 1994; Levin et al., 1994; Baskin et al., 1996; Macrae Dell et al., 2000; Chul Kim et al., 2001; Monga et al., 2001; Howard et al., 2005; Sharif-Afshar et al., 2005).

In conclusion, our data demonstrate that all three TGFβ isoforms inhibit DNA synthesis in rat BSMCs with similar potency and efficacy by a receptor-mediated mechanism, apparently not involving MAPKs. At least in rats, this is not associated with a switch from a proliferative to a hypertrophic phenotype. These data do not support the hypothesis that TGFβ is a mediator of BSMC hypertrophy and hyperplasia and hence may not be a mediator of such processes under conditions of BOO. Because these data are partly different from those obtained in human BSMCs, our findings urge caution in the extrapolation of data in rat bladder to the human situation.