 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cell Biology |
Concentration Specifies Differential Signaling Profiles of Growth Arrest/Differentiation and Apoptosis in Podocytes
Departments of * Molecular Genetics, Medicine, Albert Einstein College of Medicine, Bronx, New York, and Department of Medicine, Mount Sinai School of Medicine, New York, New York
Address correspondence to: Dr. Erwin P. Böttinger, Mount Sinai School of Medicine, One Gustave L. Levy Place, Box 1118, New York, NY 10029-6574. Phone: 212-659-8242; Fax: 212-849-2643; E-mail: erwin.bottinger{at}mssm.edu
Received for publication December 7, 2004. Accepted for publication August 24, 2005.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, which is typically upregulated in injured glomeruli. The TGF- are multifunctional cytokines that regulate growth, differentiation, and apoptosis in most cells. Determinants of functional specificity of TGF- signaling in cell-cycle control and apoptosis remain poorly understood. Using a unique system of conditionally immortalized podocytes, it is demonstrated that autocrine TGF-2 induces G0/G1 arrest and differentiation under nonpermissive culture through Smad3-dependent induction of the cyclin-dependent kinase inhibitor p15Ink4b (Cdkn2b). When exposed to recombinant TGF-1 (or TGF-2), nonpermissive culture podocytes switch to G2/M arrest and apoptosis, selectively at advanced TGF- concentrations and specifically in association with suppression of Cdkn2b and activation of proapoptotic p38 mitogen-activated protein kinase. Thus, distinct signaling profiles activated in a concentration-dependent manner by TGF- were identified. Autocrine TGF-2/Smad3/Cdkn2b signaling in podocytes specifies G0/G1 arrest associated with podocyte differentiation, whereas increasing TGF- concentrations beyond a critical threshold induces G2/M block and apoptosis associated with selective p38 mitogen-activated protein kinase activation and with suppression of Cdkn2b. In summary, the results suggest a new functional requirement of TGF-2 in growth arrest and differentiation of murine podocytes in vitro and demonstrate that a critical TGF- concentration threshold may specify a molecular switch to proapoptotic signaling profiles and apoptosis. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
superfamily consists of secreted peptides, of which the three TGF- isoforms (TGF-1 through 3), activins, and bone morphogenetic proteins (BMP) are best known in mammalian development, homeostasis, and pathobiology. The TGF- isoforms are widely expressed and act on virtually every cell type in mammals by engaging in an intracellular signaling cascade of Smad family proteins through ligand-induced activation of heteromeric transmembrane TGF- receptor kinases. Receptor-activated Smad protein complexes accumulate in the nucleus, where they participate directly in transcriptional activation of target genes. In addition, TGF- receptors can activate Smad-independent signaling mechanisms such as mitogen-activated protein kinases (MAPK) (1,2). The multifunctional TGF- may alter cell behavior by controlling growth, differentiation, death, and function of cells in development and disease, including cancer, immune, inflammatory, and fibrotic conditions (2). However, the underlying molecular mechanisms by which TGF- controls and switches cellular processes are thought to depend on specific (patho)physiologic context and remain poorly understood.
Wilms tumor antigen (WT1)-positive progenitors of glomerular visceral epithelial cells (podocytes) form mesenchyme-derived epithelial structures of glomeruli and proliferate actively during kidney development and glomerulogenesis. Starting with the S-shaped stage of glomerular development, podocyte progenitors exit from the cell cycle to acquire a quiescent, mature phenotype that is characterized by interdigitated foot processes with filtration slit diaphragms (3,4). Cell-cycle exit of podocyte progenitors in S-shaped stage has been associated with de novo expression of cyclin-dependent kinase inhibitors Cdkn1a (p21Waf1/Cip1), Cdkn1c (p57Kip2) (5), and Cdkn1b (p27Kip1) (6). Conditionally immortalized murine podocytes in culture have been shown to recapitulate the phenotypic transitions of podocytes observed during glomerular development (7). These cells express a thermolabile SV40 large T antigen oncogene when maintained under permissive culture conditions (33°C with IFN-
). In the presence of oncogene, these cells proliferate vigorously, forming an epithelial monolayer that resembles proliferative columnar epithelial podocyte progenitors characteristic of the S-shaped stage of glomerular development. When shifted to nonpermissive culture (NPC) at 37°C without IFN-, rapid loss of expression and degradation of the large T antigen is associated with growth arrest and differentiation, characterized by cytoskeletal reorganization and formation of foot process–like cell protrusions that mimic the differentiated podocyte phenotype observed in mature glomeruli (8). A recent report underscores the utility of conditionally immortalized podocyte cultures in dissecting autocrine and paracrine signaling loops by demonstrating that TGF- signaling stimulates podocyte secretion of vascular endothelial growth factor and its actions, such as increasing 
3(IV) collagen production of podocytes (9).
Here we report a new functional requirement of TGF-2 in the differentiation of murine podocytes in vitro and demonstrate TGF- concentration-dependent switching of molecular signaling profiles and biologic specificity in a physiologically relevant model system. Autocrine TGF-2/Smad3/Cdkn2b signaling in podocytes specifies G0/G1 arrest associated with podocyte differentiation, whereas increasing TGF- concentrations beyond a critical threshold induces G2/M block and apoptosis associated with selective activation of p38 MAPK and with suppression of Cdkn2b.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(GibcoBRL Life Technologies, Grand Island, NY) or under NPC at 37°C without IFN-. Conditionally immortalized Smad3–/– podocyte lines were generated from the offspring of Smad3dex8/dex8 mice (11) and ImmortoMouse (Charles River Laboratories, Wilmington, MA) for wild-type (7) and knockout cells (12), following our protocols.
Cytokine and Antibody Treatment
Cells that were maintained in complete medium under NPC were treated with recombinant human TGF-1 or TGF-2 (R&D Systems, Minneapolis, MN), anti–TGF- neutralizing antibody 2G7 (gift from Dr. John Letterio, National Cancer Institute, Bethesda, MD) (13), or IgG2b isotype-matched negative control antibody (Sigma, St. Louis, MO). Early response assays (<24 h of treatment) were performed under serum-free conditions.
Flow Cytometry
For quantification of synaptopodin expression and cell-cycle distribution, podocytes were detached from culture surface with 0.5 mM EDTA and acetone fixed, then incubated at 4°C in 10% FBS in phosphate-buffered solution with primary anti-synaptopodin (8,14) or IgG1 isotype-matched control antibody (Sigma). After washes and incubation at 4°C with secondary FITC-conjugated goat anti-mouse antibody (Jackson ImmunoResearch, West Grove, PA), the cells were resuspended in PBS that contained RNase (Roche, Mannheim, Germany) and propidium iodide (Sigma), and data were collected for 10,000 propidium iodide–positive events using a SCAN flow cytometer (Becton Dickinson). Expression was analyzed using Cellquest (Becton-Dickinson) and WinMDI 2.7 (The Scripps Research Institute, La Jolla, CA) software. Cell-cycle and apoptosis analyses were determined with MODFIT v.3.0 software (Verity Software House Inc., Topsham, ME).
Immunodetection of Podocyte Markers In Vitro and In Situ
Indirect immunofluorescence labeling was performed as described previously (10,15). Cell preparations were incubated with anti-synaptopodin (mouse) and polyclonal anti–zona occludens-1 (anti–ZO-1; goat) antibodies (Zymed, San Francisco, CA), followed by incubation with FITC- or Cy3-conjugated secondary antibodies (Jackson ImmunoResearch). Images were captured using a cooled CCD camera system (Diagnostic Instruments Inc., Sterling Heights, MI) connected to a Nikon Eclipse TE 300 fluorescence microscope (Nikon Inc., Melville, NY) and were digitally processed using Adobe Photoshop 5.0.2 (Adobe Systems Inc., San Jose, CA).
Quantitative Real-Time PCR
Total RNA was prepared from cell lysates using Qiagen RNeasy mini columns (Qiagen, Valencia, CA) and then reverse transcribed with Superscript II reverse transcriptase (Invitrogen Life Technologies, Carlsbad, CA) following commercial protocols. Quantitative amplification of the cDNA using gene-specific primer pairs (Supplemental Table 1 available online) designed with Primer3 software (16) and 2x SYBR-Green PCR Master Mix (Applied Biosystems, Foster City, CA) was performed in an ABI-Prism 7900HYT Sequence Detection System (Applied Biosystems). Results were evaluated using SDS version 2.0 software (Applied Biosystems). Normalization across samples was performed using the average expression level of the constitutive murine gene hypoxanthine guanine phosphoribosyl transferase.
TGF- ELISA
For quantification of active TGF-1 and TGF-2 in podocyte culture supernatants, sandwich ELISA were performed using isoform-specific capture and detection antibodies (R&D Systems) on supernatants that were collected into silicon-coated tubes, according to product specifications. Active TGF-1 or TGF-2 were measured directly in aliquots of culture supernatants, respectively. For measuring total (active + latent) TGF-1 or TGF-2, separate aliquots of culture supernatants were pretreated with 1 N HCl and then neutralized with an equal volume of 1.2 N NaOH/0.5 M HEPES before the assay measurements to activate latent forms of TGF- (17).
Western Blot Analyses
Total and phosphorylated proteins were detected by Western blotting as described previously (18), using the following primary antibodies: Goat anti-Smad3 (Zymed Laboratories); rabbit anti-Cdkn2b (Santa Cruz Biotechnology, Santa Cruz, CA); mouse anti-Cdkn1a (Chemicon, Temecula, CA); rabbit anti–phospho-Smad2/3, mouse anti-Smad2/3, rabbit anti–phospho-, and mouse anti-total p38 (all from Cell Signaling Technology, Beverly, MA); and rabbit anti–guanosine diphosphate (GDP) dissociation inhibitor (GDI, gift of Dr. Philipp Scherer, Albert Einstein College of Medicine, Bronx, NY). Anti–GDP dissociation inhibitor is used as loading control.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Function Is Required for Cell-Cycle Arrest and Differentiation of Conditionally-Immortalized Podocytes is well characterized in the regulation of cell-cycle arrest in epithelial cells, we reasoned that autocrine TGF- may have a role in growth arrest and/or differentiation of murine podocytes. To establish a quantitative assay for podocyte differentiation, we adapted standard flow cytometric methods to monitor the expression of the differentiation marker synaptopodin in podocytes that were cultured under NPC for up to 2 wk. Both the measurements of synaptopodin levels per cell and the number of synaptopodin-positive cells increased consistently after day 2 and reached steady-state level by day 7 under NPC, defining a time window for further analysis of signaling mechanisms of growth arrest and differentiation in podocytes between day 1 and day 7 (Figure 1A). In contrast, the characteristic increase of synaptopodin protein levels (flow cytometry product) between day 2 and day 7 was blocked by a pan-neutralizing anti–TGF- antibody 2G7 but not by the isotype-matched negative control IgG2b (Figure 1B).
|
2G7–treated group (P < 0.01; Figure 1C). At later time points (day 6 and 8), the percentage change of G0-G1 cells was further increased but not significantly different among all three groups (Figure 1C). The loss of anti–TGF- antibody effect at late time points was possibly due to loss of anti–TGF- antibody activity, because these studies were conducted using only a single dose of 2G7 treatment.
To determine whether TGF- was essential for the manifestation of morphologic hallmarks of podocyte differentiation under NPC, we used immunofluorescence labeling for synaptopodin and ZO-1. ZO-1 is a 220-kD peripheral membrane phosphoprotein that is expressed in cell–cell junctions and shows a differentiation-dependent redistribution in podocytes (15,19). After 5 d under control (no antibody or control IgG2b) NPC, synaptopodin was localized in a typical fiber-associated pattern that extended through cell processes (Figure 1D), as described before (7). ZO-1 protein was redistributed from an epithelial linear pattern in undifferentiated podocytes to a "sawtooth" pattern (Figure 1D) that has previously been characterized with de novo formation of cellular extensions and slit diaphragm–like structures (15). Anti–TGF- antibody 2G7 blocked synaptopodin expression and inhibited the redistribution of ZO-1 (Figure 1D). These findings demonstrate that TGF- function is required for the manifestation of characteristic features of podocyte differentiation, including G0/G1 cell-cycle arrest, and the expression and localization of the differentiation markers synaptopodin and ZO-1.
Increased Synthesis and Activation of Latent TGF-2 Is Associated with Smad Signaling and Podocyte Differentiation
The functional requirement of TGF- in podocyte differentiation suggested that autocrine synthesis and/or activation of TGF- isoforms may be induced under NPC. TGF- isoform–specific sandwich ELISA showed significant increases in both total (latent + active) and active TGF-2 in culture media within 2 d of NPC (Figure 2, A and B, respectively). In contrast, the amount of TGF-1 in culture medium was not significantly changed under NPC (data not shown). Similarly, TGF-2 mRNA but not TGF-1 mRNA synthesis increased during NPC (data not shown). TGF-3 mRNA was not detectable under these conditions (data not shown). In addition, signaling via the ligand-activated TGF- receptor complex, measured here by analysis of phosphorylated Smad2 and Smad3 (phospho-Smad2/3), was increased on day 2 and day 4 of NPC. Anti–TGF- antibody significantly reduced phospho-Smad2/3 levels on days 2 and 4 when compared with control IgG cells (Figure 2C). These results suggest that a moderate increase of approximately 10 pg/ml active TGF-2, observed between day 1 and day 2 under NPC, was sufficient to activate Smad signaling with subsequent G0/G1 cell-cycle arrest during podocyte differentiation.
|
2 Induces Smad3-Dependent G0/G1 Cell-Cycle Arrest and Differentiation signaling (20,21). To explore the role of Smad3 in podocyte differentiation, we generated conditionally immortalized podocyte cell lines that carry targeted deletions of exon 8 of Smad3 (described in Materials and Methods). Elimination of Smad3 protein expression in Smad3-knockout (Smad3–/–) mice was confirmed in two podocyte cell lines that were derived from individual Smad3–/– mice (Figure 3A). The increase in synaptopodin product, typically observed between days 4 and 7 during NPC in wild-type podocytes, was significantly reduced and delayed in both Smad3–/– podocyte clones (Figure 3B). In addition, transcript levels of the late podocyte differentiation markers synaptopodin, podocalyxin, and nephrin (3,22) were significantly decreased in Smad3–/– cells after day 5 of NPC (Figure 3C), indicating impaired differentiation. In contrast, mRNA levels of the early podocyte lineage marker, Wilm’s tumor antigen 1 (WT-1), were increased in Smad3–/– cells compared with control podocytes (Figure 3C). Cell-cycle analysis revealed two distinct phases of increase of G0/G1 population and decreases of G2/M phase cell populations of podocytes. We detected an early G0/G1 increase and G2/M decrease that was detectable by day 2 regardless of podocyte genotype and clone and that accounted for less than half of the total changes in each cell-cycle phase, respectively (Figure 3D). In contrast, a major contribution of G0/G1 arrest and G2/M decrease induced by NPC occurred after day 2 in wild-type podocytes but was not detectable in both Smad3KO clones (Figure 3D). In addition, the sawtooth pattern of ZO-1 typically observed by day 5 of NPC in wild-type cells was reduced in Smad3–/– clones (Figure 3E). These findings indicate that Smad3 is an essential mediator of G0/G1 cell-cycle arrest and differentiation of podocytes induced by autocrine TGF-2 under NPC.
|
2/Smad3-Dependent Induction of Cdkn2b Is Associated with G0/G1 Arrest and Differentiation of Podocytes2/Smad3-dependent G0/G1 arrest, we examined protein levels of Cdkn2b (p15INK4b), Cdkn1a (p21Waf/Cip1), Cdkn1b (p27Kip1), and Cdkn1c (p57Kip2) in wild-type and Smad3–/– podocytes under permissive conditions (day 0) and NPC (days 1, 2, and 4). Baseline Cdkn2b protein (day 0 and day 1) became strongly increased on day 2 and day 4 in wild-type cells (Figure 4A). In contrast, baseline Cdkn2b protein was decreased in Smad3–/– podocytes compared with wild-type, and the Cdkn2b increase that was observed in wild-type cells was prevented by neutralizing anti–TGF- antibody 2G7 and dramatically reduced in Smad3–/– podocytes (Figure 4A). Cdkn1a protein expression was moderately increased on day 4 of NPC in wild-type cells (Figure 4B). It is interesting that Cdkn1a protein levels were increased in wild-type cells that were pretreated with neutralizing anti–TGF- antibody, compared with untreated cells (Figure 4B). Similarly, Cdkn1a protein levels were also increased in Smad3–/– podocytes compared with wild-type podocytes under NPC (Figure 4B). Cdkn1c expression was not significantly regulated at any time under NPC (data not shown), and Cdkn1b was not detectable by quantitative real-time PCR in this system under these conditions. Taken together, these results indicate that Smad3-dependent Cdkn2b synthesis coincides with an increase in active TGF-2 and subsequent G0/G1 cell-cycle arrest in podocytes under NPC. In contrast, moderate induction of Cdkn1a was observed at later time points and was independent of TGF-2 and of Smad3 function. These observations suggest that upregulation of Cdkn2b but not Cdkn1a is the primary cyclin-dependent kinase inhibitor target gene of autocrine TGF-2/Smad3 signaling during G0/G1 arrest and podocyte differentiation induced by NPC.
|
Activates a Concentration-Dependent Switch of Signaling Signature and Cell-Cycle Profile in Podocytes That Is Associated with Inhibition of Differentiation1 is increased in podocytes in glomerular injury models in situ (12), we exposed podocytes under NPC to increasing concentrations of active recombinant TGF-1, described here as low (0.125 ng/ml), medium (1.0 ng/ml), and high (4.0 ng/ml). The low concentration TGF-1 had no significant effect on the increase of synaptopodin (Figure 5A) or on G0/G1 arrest induced by NPC (Figure 5, B and C). However, high concentrations significantly reduced synaptopodin induction (Figure 5A), abrogated the G0/G1 cell-cycle arrest (Figure 5B), and enhanced the G2/M phase populations (Figure 5C). To test whether these effects were isoform specific, we performed the analogous experiments using equivalent concentrations of recombinant TGF-2. TGF-2 produced similar results compared with TGF-1, indicating that the dose-dependent TGF- effect was not isoform specific (Figure 5). The medium concentration induced an intermediate phenotype (Figure 5). TGF-1 caused a moderate increase of phospho-Smad2/3 levels at the low concentration, with a stronger increase at the medium and high concentrations on days 1 and 2, compared with untreated NPC (Figure 6A). However, by days 4 and 5, only the high concentration further augmented Smad2 levels (Figure 6A). Cdkn2b protein synthesis was induced equally by all three TGF-1 concentrations after 1 d of NPC (Figure 6B), whereas high concentrations significantly suppressed the NPC-induced increase of Cdkn2b protein on day 4 (Figure 6B). Low dose TGF-1 had no significant effect on Cdkn1a protein levels during NPC (Figure 6C). Medium and high dose caused a decrease of Cdkn1a on day 4 of NPC (Figure 6C). In Smad3–/– podocytes, recombinant TGF- had no effect on cell-cycle profiles, differentiation, and Cdkn2b or Cdkn1a protein levels (data not shown). We conclude that a moderate increase in the synthesis and activation of autocrine TGF-2 mediates G0/G1 cell-cycle arrest with TGF-/Smad3-dependent induction of Cdkn2b, permitting differentiation of podocytes under NPC. In contrast, a continuous presence of higher concentrations of TGF- results in suppression of Cdkn2b protein synthesis with mitigation of G0/G1 cell-cycle arrest, activation of G2/M checkpoints, and inhibition of differentiation.
|
|
Activate Apoptosis and p38 MAPK Signaling1 under NPC. Beginning at day 4, the high concentration of TGF-1 significantly increased the rate of apoptosis in wild-type podocytes, whereas the medium and the low concentrations had minimal and no effects, respectively (Figure 7A). Recombinant TGF-2 produced similar results (data not shown). In contrast with wild-type cells, Smad3–/– podocyte clones were resistant to TGF-–induced apoptosis irrespective of TGF- concentration (Figure 7B). Consistent with increased apoptosis beginning on day 4, TGF- activated p38 MAPK in a concentration-dependent manner (Figure 7C). In particular, only the medium and high concentrations but not the low concentration of TGF-1 increased phospho-p38 MAPK compared with untreated NPC on day 4 (Figure 7C).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 synthesis and activity provides an essential extracellular signal to induce G0/G1 arrest and to express protein markers and morphology of podocyte differentiation. This conclusion is supported by the following key observations. First, neutralization of TGF- blocked G0/G1 arrest and inhibited markers and morphologic hallmarks of differentiation typically induced after 2 d of NPC podocytes in vitro. Second, a significant increase of autocrine secretion and activation of TGF-2 but not TGF-1 was kinetically associated with TGF-–dependent Smad phosphorylation and preceded G0/G1 arrest and differentiation on day 2. Third, G0/G1 arrest and differentiation were abrogated in Smad3–/– podocytes. Fourth, induction of Cdkn2b mRNA and protein coincided kinetically with autocrine TGF-2 activation and Smad phosphorylation, beginning on day 2, and was inhibited by neutralization of TGF- and Smad3 deficiency. Thus, we have identified a functional autocrine TGF-2 signaling signature of TGF-2/R-Smad/Cdkn2b activation that can be induced by relatively small increases (10 to 20 pg/ml) of active TGF-2 concentrations. Of note, podocytes express beta-glycan, a cell surface accessory binding protein that selectively binds TGF-2 and is required for effective binding and activation of type II TGF- signaling receptors by TGF-2. The autocrine TGF-2 signaling signature may provide a molecular readout for a "physiologic context," including cell-cycle exit of proliferating podocyte progenitors during glomerulogenesis and maintenance of a quiescent, homeostatic state of differentiated podocytes in mature mammalian kidneys.
Several in vivo observations support this interpretation. For example, TGF- isoforms and TGF- receptor type II are expressed in podocyte progenitors in early epithelial nephron structures (25). Specifically, expression of TGF-2 has been described in newly formed epithelial structures, including S-shaped bodies (26). The embryonic lethal deletion of the Tgfb2 gene in mice is characterized by multiple organ defects, including proteinaceous casts and dilated tubules in the kidney (27). Incompletely penetrant kidney agenesis in the female mice implicates Tgfb2 in the induction of the metanephros (27). In contrast, kidney development and function at birth seem normal in Tgfb1 and Tgfb3 knockout mice (28,29). TGF-2, in combination with fibroblast growth factor-2, has also been shown to induce formation of glomerular-like structures in explant cultures (26). Finally, loss of the common TGF- signal transducer Smad4 in the metanephric mesenchyme leads to impaired condensation of nephrogenic mesenchyme, suggesting a role for TGF- signaling in the early stages of nephrogenesis (30).
In contrast with physiologic context, pathophysiologic contexts of glomerular and podocyte injury are typically associated with increased expression of TGF- ligands, predominantly TGF-1. For example, podocytes in diseased kidneys showed increased expression of predominantly TGF-1 ligand and TGF- receptors in animal models such as the rat anti-Thy1 model (31), experimental membranous nephropathy (32), and CD2AP-deficient mice (12). The pathophysiologic context is further distinguished from the physiologic context by the occurrence of podocyte apoptosis in vivo (10,12,33). We previously reported that chemical inhibition of p38 MAPK prevented apoptosis that was induced by a high dose (5.0 ng/ml) of TGF-1 in murine podocytes, demonstrating that p38 MAPK activation exerts an essential and causal role in TGF-–induced podocyte apoptosis (10,12). In this context, our findings demonstrate in a single cell culture model that increasing TGF- concentration is sufficient to adjust its biologic functions from physiologic G0/G1 growth arrest and differentiation to a pathophysiologic induction of G2/M arrest and apoptosis. In addition to defining a new concentration threshold that demarcates normal cytostatic and differentiating functions from abnormal apoptotic functions of TGF-, our results delineate distinct molecular signaling profiles that may specify the concentration-dependent switch. At the high, proapoptotic concentrations, we observed a significantly increased and sustained phosphorylation state of Smad2, suggesting increased activity of ligand-induced TGF- receptor complexes. This was correlated with selective activation of p38 MAPK. Because the depletion of podocytes that is typically associated with progressive glomerulosclerosis results at least in part from podocyte apoptosis, the results presented here suggest that escalation of TGF- activity may be a critical event that determines a point of no return in diseases of the podocyte.
Another distinctive feature of the proapoptotic signaling profile is the repression of the cyclin-dependent kinase inhibitor Cdkn2b. Cdkn2b is a specific inhibitor of cyclin D-cdk4 and -cdk6 activities in G1, where it mediates G1 arrest in many cells that are growth inhibited by TGF- (34,35). TGF- has been shown to induce rapidly the expression of Cdkn2b in a variety of different cell types through Smad-mediated transactivation (36) and relief of Myc-mediated repression (37). However, our results demonstrate that Cdkn2b regulation is bimodal in podocytes, depending on TGF- concentration. Thus, a low increase of autocrine TGF-2 activity may signal via R-Smads to induce Cdkn2b synthesis, and this was associated with G0/G1 arrest and differentiation. In contrast, sustained proapoptotic TGF- concentrations suppressed Cdkn2b and were associated with failure to induce G0/G1 arrest in NPC podocytes. It is interesting that short-term treatment at any concentration induced Cdkn2b protein equally within the first day of NPC, suggesting that duration and context of treatment may contribute to the bimodal Cdkn2b response. The molecular determinants that underlie the switch from induction to repression of Cdkn2b remain unclear at this time and require further investigation.
Finally, our findings suggest that Smad3 is a central and essential mediator of concentration-dependent TGF- signaling signatures, Cdkn2b-associated G0/G1 growth arrest, and p38 MAPK-associated apoptosis. Thus, any model of concentration-dependent differential target gene activation must accommodate the apparent universal requirement of a single modulator, Smad3. Recent reports demonstrate that alterations in Smad affinity alone can generate a threshold response, but cooperative DNA binding with different Smad partners may also be important (38,39). Recently demonstrated heterotrimeric Smad complexes, which may include two different R-Smads (40), may perform different tasks by recognizing and activating specific gene promoters. The unique model system established in our work should provide a suitable experimental tool to begin to investigate these mechanisms.
It is interesting that by light microscopic analysis, podocytes and glomeruli in kidneys of Smad3–/– mice seem normal and urinary albumin excretion is normal. In contrast, on the basis of our in vitro findings, one would anticipate a developmental podocyte defect in Smad3–/– mice. A possible explanation may be provided by recent elegant studies providing new insights into relative functions of full-length Smad2, alternatively spliced short form of Smad2 lacking exon 3 (41), and Smad3 (42). On the basis of genetic isoform replacement experiment in mice, short form of Smad2 and Smad3 but not full-length Smad2 is sufficient to mediate all transcriptional responses of TGF- required for normal development in the absence of Smad3 (42). Thus, it is possible that Smad3–/– of developing kidneys is compensated in differentiating podocytes in vivo by alternatively spliced Smad2 (Smad2
exon3), which may be sufficient to back up Smad3 function during mouse embryo development. It is possible that the short form Smad2exon3 is not sufficient to back up for Smad3 deficiency in physiologic responses such as cell-cycle control in conditionally immortalized podocytes in vitro. Further studies are under way to evaluate this possibility.
|
Acknowledgments |
|---|
This study was presented in part at the American Society of Nephrology 35th Annual Meeting and Scientific Exposition, Philadelphia, PA, November 1 to 4, 2002, and has been published in abstract form (J Am Soc Nephrol 13: 92A-93A, 2002).
We thank the members of the Böttinger laboratory for discussion and review of this article. We thank Liping Yu for technical assistance. We are grateful to Dr. John Letterio for providing TGF- neutralizing antibody.
|
Footnotes |
|---|
Published online ahead of print. Publication date available at www.jasn.org.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Q. Wang, Y. Wang, A. W. Minto, J. Wang, Q. Shi, X. Li, and R. J. Quigg MicroRNA-377 is up-regulated and can lead to increased fibronectin production in diabetic nephropathy FASEB J, December 1, 2008; 22(12): 4126 - 4135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Shi, H. Qu, M. Kretzler, and C. Wu Roles of PINCH-2 in regulation of glomerular cell shape change and fibronectin matrix deposition Am J Physiol Renal Physiol, July 1, 2008; 295(1): F253 - F263. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Li, Y. S. Kang, C. Dai, L. P. Kiss, X. Wen, and Y. Liu Epithelial-to-Mesenchymal Transition Is a Potential Pathway Leading to Podocyte Dysfunction and Proteinuria Am. J. Pathol., February 1, 2008; 172(2): 299 - 308. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Puthawala, N. Hadjiangelis, S. C. Jacoby, E. Bayongan, Z. Zhao, Z. Yang, M. L. Devitt, G. S. Horan, P. H. Weinreb, M. E. Lukashev, et al. Inhibition of Integrin {alpha}v 6, an Activator of Latent Transforming Growth Factor- , Prevents Radiation-induced Lung Fibrosis Am. J. Respir. Crit. Care Med., January 1, 2008; 177(1): 82 - 90. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Wang, F. N. Ziyadeh, E. Y. Lee, P. E. Pyagay, S. H. Sung, S. A. Sheardown, N. J. Laping, and S. Chen Interference with TGF-beta signaling by Smad3-knockout in mice limits diabetic glomerulosclerosis without affecting albuminuria Am J Physiol Renal Physiol, November 1, 2007; 293(5): F1657 - F1665. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Asanuma, K. N. Campbell, K. Kim, C. Faul, and P. Mundel Nuclear relocation of the nephrin and CD2AP-binding protein dendrin promotes apoptosis of podocytes PNAS, June 12, 2007; 104(24): 10134 - 10139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Theilig, W. Kriz, T. Jerichow, P. Schrade, B. Hahnel, T. Willnow, M. Le Hir, and S. Bachmann Abrogation of Protein Uptake through Megalin-Deficient Proximal Tubules Does Not Safeguard against Tubulointerstitial Injury J. Am. Soc. Nephrol., June 1, 2007; 18(6): 1824 - 1834. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Y. Jung, K. Chen, M. Kretzler, and C. Wu TGF-beta1 Regulates the PINCH-1-Integrin-Linked Kinase-{alpha}-Parvin Complex in Glomerular Cells J. Am. Soc. Nephrol., January 1, 2007; 18(1): 66 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Sugiyama and T. Yokoyama Sustained cell proliferation of renal epithelial cells in mice with inv mutation. Genes Cells, October 1, 2006; 11(10): 1213 - 1224. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. G. Docherty, O. E. O'Sullivan, D. A. Healy, M. Murphy, A. J. O'Neill, J. M. Fitzpatrick, and R. W. G. Watson TGF-beta1-induced EMT can occur independently of its proapoptotic effects and is aided by EGF receptor activation Am J Physiol Renal Physiol, May 1, 2006; 290(5): F1202 - F1212. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||
) or isotype-matched negative control IgG2b (
) at 200 µg/ml for up to 8 d were analyzed for flow cytometry product (FC product = %positive cells x mean fluorescence intensity) of synaptopodin (mean ± SEM; n = 2; B) and percentage of cells that were exiting from cell cycle (%change of G0-G1 cells; C) by days 3, 6, and 8 under nonpermissive culture (NPC) in each of the experimental groups relative to the day 0 baseline (ModFIT software; mean ± SEM; n = 2). (D) Indirect immunofluorescence images localize synaptopodin (top) and zona occludens-1 (ZO-1) protein (bottom) in podocytes that were cultured at 37°C in the absence (No Ab) or in the presence of antibody (Control Ab = IgG2b, 
) relative to wild-type podocytes (
), where levels in wild-type podocytes are set to 1 (mean ± SEM; n = 2). (D) Histograms showing cell population distribution in G0/G1 and G2/M phases determined by flow cytometric analysis of propidium iodide–labeled DNA content in wild-type and Smad3–/– podocyte cell lines (S3KO#1 and S3KO#2), as indicated after NPC for up to 9 d (mean ± SEM; n = 3; *P < 0.05 by unpaired t test). All profiles were normalized to day 1 values. (E) Indirect immunofluorescence images detect ZO-1 protein localization in wild-type and Smad3–/– podocyte cell lines after NPC for 5 d. Magnification, x40.
), 0.125 ng/ml (
), 4.0 ng/ml (
) (mean ± SEM; n = 3; *P < 0.05 by unpaired t test).
