Abstract
Alterations in the cellular architecture, adhesion, and/or loss of glomerular podocytes are causal factors in the development of proteinuria and the progression to end-stage renal failure. With the use of an inducible podocyte differentiation system, it was found that the cellular levels of PINCH-1, integrin linked kinase (ILK), and α-parvin, cytoplasmic components of cell–extracellular matrix adhesions, were significantly increased during podocyte differentiation. Concomitantly, an increased amount of the PINCH-1–ILK–α-parvin complex was detected in the differentiated, foot process–containing podocytes. Overexpression of the PINCH-1–binding ankyrin repeat domain of ILK but not that of a PINCH-1–binding defective mutant form of the ankyrin domain effectively inhibited the formation of the PINCH-1–ILK–α-parvin complex. Disruption of the PINCH-1–ILK–α-parvin complex significantly reduced the podocyte–matrix adhesion and foot process formation. Furthermore, a marked increase of apoptosis in the podocytes in which the assembly of the PINCH-1–ILK–α-parvin complex was compromised was detected. Inhibition of ILK with a small compound inhibitor also altered podocyte cytoskeleton and increased apoptosis. Finally, it is shown that α-parvin is phosphorylated in podocytes. Mutations at the α-parvin N-terminal proline-directed serine phosphorylation sites reduced its complex formation with ILK and resulted in defects in podocyte adhesion, architecture, and survival. These results provide important evidence for a crucial role of the PINCH-1–ILK–α-parvin complex in the control of podocyte adhesion, morphology, and survival.
Renal glomerular podocytes are highly differentiated cells whose complex, heavily branched cellular architecture has long fascinated cell biologists. Recent studies in human patients and animal models have suggested that alterations in the delicate cellular architecture, extracellular matrix adhesion, and/or loss of podocytes are intimately associated with a variety of renal diseases and are important causal factors in the development of proteinuria and the progression to end-stage renal failure (for recent reviews, see references 1–8). Understanding the molecular mechanisms that control podocyte adhesion, cytoarchitecture, and survival therefore is not only a challenge to cell biologists but also of considerable clinical importance.
Integrin linked kinase (ILK) (9) is a multidomain focal adhesion protein that contains both protein–protein interaction and protein kinase activities (reviewed in references 10–13). The ILK N-terminal ankyrin (ANK) repeat domain interacts with PINCH-1, a focal adhesion protein that contains primarily five LIM domains (14,15). The ILK C-terminal region contains a protein kinase catalytic domain and binding sites for several focal adhesion proteins, including members of the CH-ILKBP/actopaxin/parvin (abbreviated as parvin herein) protein family (16–18). ILK, through interactions mediated by its N- and C-terminal domains, respectively, forms a ternary complex with PINCH-1 and α-parvin in a number of cell types, including fibroblasts and myoblasts (16,19). Immunohistochemical analyses of human kidneys have revealed that mature glomerular podocytes express a high level of ILK protein (20). Furthermore, messenger RNA differential display studies in glomeruli from patients with congenital nephritic syndrome of the Finnish type have suggested that alteration of ILK expression is likely critically involved in the pathogenesis of proteinuria (21–23).
Mature podocytes are highly differentiated cells that are unable to proliferate in culture, which had hampered the progress of experimental investigations on podocytes at molecular and cellular levels. To circumvent the experimental difficulty, Mundel et al. (24) successfully developed an inducible podocyte differentiation system in which cells that retain differentiation potential similar to podocytes in vivo were conditionally immortalized. Culturing the conditionally immortalized cells under nonpermissive conditions renders the cells growth arrested and induces many characteristics of differentiated podocytes, including formation of numerous foot processes (21,24). Using this inducible podocyte differentiation system, we investigated the functions of the PINCH-1–ILK–α-parvin complex in podocytes. Our results suggest that the PINCH-1–ILK–α-parvin complex plays at least two important roles in podocytes, namely (1) regulation of podocyte–matrix adhesion and morphology and (2) protection of podocytes from apoptosis. Furthermore, we provide evidence suggesting that phosphorylation of α-parvin is critically involved in the regulation of the PINCH-1–ILK–α-parvin complex formation and consequently the extracellular matrix adhesion, architecture, and survival of podocytes.
Materials and Methods
Podocyte Differentiation
The differentiation of conditionally immortalized mouse podocytes was induced as described previously (21,24). Briefly, the cells were propagated under permissive condition at 33°C in RPMI 1640 medium that contained 10% FCS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 10 U/ml mouse recombinant IFN-γ. For inducing differentiation, the cells were switched to RPMI 1640 medium that contained 10% FCS, 100 U/ml penicillin, and 100 μg/ml streptomycin but lacked IFN-γ (nonpermissive condition) and was cultured at 37°C.
Recombinant Adenoviral Expression Vector Construction and Infection
The adenoviral expressing vector encoding the PINCH-1–binding ILK N-terminal fragment (residues 1 to 230) was generated as described previously (20). For generating the PINCH-1–binding defective mutant form of the N-terminal fragment, a point mutation that substitutes D31 with A was introduced into the ILK cDNA fragment encoding the N-terminal fragment (residues 1 to 230) using a QuickChange site-directed mutagenesis system (Stratagene, La Jolla, CA). The cDNA fragment bearing the mutation was cloned into the SalI/XbaI sites of the pAdTrack-CMV shuttle vector. The shuttle vector plasmid was linearized with PmeI, purified by phenol/chloroform extraction and ethanol precipitation, and mixed with supercoiled pADEsay-1. The vectors were transferred into Escherichia coli BJ5183 by electroporation using a Bio-Rad Gene Pulser electroporator. The bacteria were immediately placed in 1 ml of LB Broth, Lennox (10 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl; Fisher, Pittsburgh, PA) and grown at 37°C for 1 h. The bacteria then were inoculated onto agar that contained LB Broth supplemented with 50 μg/ml kanamycin. After 16 to 20 h of growth, colonies were picked and grown in 2 ml of LB Broth that contained 50 μg/ml kanamycin. Clones were screened by digestions with restriction endonucleases PacI and BamHI. The positive plasmids were transformed into DH10B cells by electroporation for large-scale amplification. The plasmid DNA was digested with PacI, ethanol-precipitated, and was used to transfect 293 cells with LipofectAmine PLUS. The transfected cells were harvested 10 d after transfection. The cells were lysed by three cycles of freezing in a methanol/dry ice bath and rapid thawing at 37°C, and the lysates that contained the recombinant adenovirus were collected. For generating adenoviral expressing vector encoding the α-parvin mutant (ΔN) in which the N-terminal phosphorylation sites were deleted, a cDNA fragment encoding a FLAG-tagged α-parvin mutant (residues 23 to 372) was generated by PCR and inserted into the pAdTrack-CMV shuttle vector. The adenoviral vector encoding the α-parvin mutant then was generated using the protocol as described above. The control adenoviral expression vector encoding β-galactosidase was provided by Drs. Tong-Chuan He and Bert Vogelstein (Howard Hughes Medical Institute, Johns Hopkins Oncology Center, Baltimore, MD).
For expressing the ILK and α-parvin mutants in podocytes, the cells were cultured under the nonpermissive condition for 14 d and then infected with adenoviral vectors encoding β-galactosidase, the FLAG-tagged PINCH-1–binding ILK N-terminal fragment, the FLAG-tagged PINCH-1–binding defective ILK N-terminal fragment bearing the D31A mutation, or the FLAG-tagged α-parvin mutant as specified in each experiment. The infection efficiency was monitored by the expression of GFP encoded by the viral vectors, which typically reached 80 to 90% within 3 d. The expression of the FLAG-tagged ILK and α-parvin mutants in the infected cells was confirmed by Western blotting with monoclonal anti-FLAG antibody M5. The PINCH-1–ILK–α-parvin complex formation, phosphorylation, podocyte adhesion, spreading, and apoptosis were analyzed 3 d after the adenoviral infection.
Immunoprecipitation and Western Blot
The formation of the PINCH-1–ILK–α-parvin complex in podocytes was analyzed by immunoprecipitation based on a previously described method (16,19). Briefly, the cells (as specified in each experiment) were lysed with the lysis buffer (1% Triton X-100 in 50 mM Tris-HCl [pH 7.4] that contained 150 mM NaCl, 5 mM EDTA, 2 mM Na3VO4, 2.5 mM Na4PO7, 100 mM NaF, 200 nM microcystin-LR, and protease inhibitors). The cell lysates (500 μg) were mixed with 500 μl of hybridoma culture supernatant that contained monoclonal anti–CH-ILKBP antibody 1D4. The samples were incubated for 3 h, mixed with 40 μl of UltraLink Immobilized Protein G (Pierce), and then incubated for an additional 2 h. The beads were washed four times, and the proteins bound were released from the beads by boiling in SDS-PAGE sample buffer for 5 min. The samples were analyzed by Western blotting with mouse monoclonal anti–CH-ILKBP antibody 3B5, mouse monoclonal anti-ILK antibody 65.1, rabbit polyclonal anti–PINCH-1, or mouse monoclonal anti–phospho-Ser/Thr antibody MPM2 (Upstate Biotechnology, Lake Placid, NY) antibodies as specified in each experiment. For immunoprecipitating FLAG-tagged ILK mutants, podocytes that were infected with the corresponding adenoviral expression vectors were lysed with the lysis buffer. The lysates (400 μg) were mixed with 30 μl of agarose beads conjugated with anti-FLAG antibody M2 (Sigma, St. Louis, MO), incubated, and washed as described above. The precipitated proteins were released from the beads by boiling in SDS-PAGE sample buffer for 5 min and analyzed by Western blotting as specified in each experiment.
Cell Adhesion and Spreading
Cell adhesion and spreading were analyzed as described previously (16). Briefly, for analyzing cell adhesion, cells (as specified in each experiment) were harvested with trypsin and washed twice with RPMI 1640 that contained 10% FBS and then twice with serum-free Opti-MEM (Life Technologies, Grand Island, NY). The cells (2.25 × 104/well) were seeded in laminin-coated 96-well plates (BD Biosciences, Bedford, MA). After incubation at 37°C under a 5% CO2/95% air atmosphere for 60 min, the wells were washed three times with PBS. The numbers of adhered cells were quantified by measuring N-acetyl-β-D-hexosaminidase activity as described previously (23). For analyzing cell spreading, cells were prepared as described above and then plated in laminin-coated 96-well plates and incubated at 37°C under a 5% CO2/95% air atmosphere for 45 min. The cell morphology was observed under an Olympus IX70 microscope equipped with a Hoffman Modulation Contrast system and recorded with a DVC-1310C Magnafire digital camera (Optronics). Unspread cells were defined as round cells, whereas spread cells were defined as cells with extended processes as described (16,25–27). The percentage of cells that adopted spread morphology was quantified by analyzing at least 300 cells from three randomly selected fields (>100 cells/field).
Apoptosis
Apoptosis was analyzed using both caspase-3 and terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick-end labeling (TUNEL) assays. The caspase 3 activity was measured using fluorogenic caspase-3 substrate VII (Ac-DEVD-AFC) from Calbiochem (San Diego, CA). The TUNEL assay was performed using an In Situ Cell Death Detection kit (Roche) following the manufacturer’s protocol.
Treatment of the Podocytes with ILK Inhibitor MC-5
Cells were propagated under permissive conditions as described above. For inducing differentiation, podocytes were maintained on type I collagen–or laminin-coated surface or coverslips at 37°C without mouse IFN-γ for 10 d. The cells were starved overnight in serum-free medium before treatment with the ILK inhibitor. The next day, cells were incubated for 24 h with ILK inhibitor MC-5 (Kinetek Pharmaceuticals, Vancouver, BC, Canada) (28) at 0, 1, or 5 μM as specified in each experiment and processed as described below.
For phalloidin staining, cells that were grown on collagen I–coated glass coverslips were fixed with 2% paraformaldehyde and 4% sucrose in PBS for 5 min at room temperature. Subsequently, podocytes were permeabilized with 0.3% Triton X-100 in PBS for 10 min. After rinsing three times with PBS, specific binding sites were blocked (2% FCS, 2% BSA, 0.2% fish gelatin in PBS) for 30 min followed by incubation with Phalloidin-TRITC (Sigma, Germany) 1:6000 in blocking solution for 60 min at room temperature. Finally, coverslips were washed five times with PBS and covered in Mowiol (Calbiochem, Merck, Germany). Percentage of cytoskeletal alterations with a reduction in stress fibers accompanied by cortical localization and aggregation was scored in a blinded manner. To do this, the development of stress fiber formation was rated by a blinded observer as follows. One hundred cells on each coverslip were scored intact when extended stress fibers throughout the cell and a normal distribution of actin were encountered. Cells with an accumulation of actin fibers, partially developed stress fibers, or stress fibers located in the subcortical region were counted as negative. Coverslips of three independent experiments were examined (n = 300 cells).
Results
Level of the PINCH-1–ILK–α-Parvin Complex Is Increased in Differentiated, Foot Process–Containing Podocytes
To begin to analyze the functions of the PINCH-1–ILK–α-parvin complex in podocytes, we compared the levels of PINCH-1, ILK, and α-parvin in conditionally immortalized podocytes that were cultured under permissive (33°C with IFN-γ) and nonpermissive (37°C without IFN-γ) conditions, respectively. Consistent with previous studies (21,24), the conditionally immortalized podocytes that were cultured under the permissive condition exhibited cobblestone morphology (Figure 1A), whereas those that were cultured under the nonpermissive condition formed numerous processes that are characteristics of differentiated podocytes (Figure 1B). Western blot analyses of the cells showed that the cellular levels of PINCH-1, ILK, and α-parvin were significantly increased in the differentiated podocytes (Figure 1, C through E, compare lanes 1 and 2). To test whether PINCH-1, ILK, and α-parvin form a complex in podocytes, we immunoprecipitated α-parvin from the cells with a monoclonal anti–α-parvin antibody (clone 1D4) and probed the α-parvin immunoprecipitates with antibodies that recognize α-parvin (Figure 1C), ILK (Figure 1D), and PINCH-1 (Figure 1E), respectively. The results showed that PINCH-1 (Figure 1E, lanes 3 and 4) and ILK (Figure 1D, lanes 3 and 4) were readily co-immunoprecipitated with α-parvin (Figure 1C, lanes 3 and 4), suggesting that PINCH-1, ILK, and α-parvin form a complex in podocytes. Furthermore, consistent with the increase in the cellular levels of PINCH-1, ILK, and α-parvin in the differentiated podocytes, we detected an increased amount of the PINCH-1–ILK–α-parvin complex in the differentiated podocytes (Figure 1, C through E, compare lanes 3 and 4).
The level of the PINCH-1–integrin linked kinase (ILK)–α-parvin complex is increased in differentiated, foot process–containing podocytes. (A and B) Morphology of conditionally immortalized mouse podocytes cultured under the permissive condition (A) or the nonpermissive condition (B) for 14 d. Bar = 50 μM. (C through E) Lysates (10 μg/lane) of the proliferating podocytes that were cultured under the permissive condition (lane 1) or the differentiated podocytes that were cultured under the nonpermissive condition (lane 2) for 14 d were analyzed by Western blotting with monoclonal anti–α-parvin antibody 3B5 (C), anti-ILK antibody 65.1 (D), and a rabbit polyclonal anti–PINCH-1 antibody (E). Lanes 3 and 4, α-parvin was immunoprecipitated from the proliferating (lane 3) or the differentiated podocytes (lane 4) with monoclonal anti–α-parvin antibody 1D4 as described in the Materials and Methods section. The α-parvin immunoprecipitates were analyzed by Western blotting with mouse monoclonal anti–α-parvin antibody 3B5 (C), anti-ILK antibody 65.1 (D), and a rabbit polyclonal anti–PINCH-1 antibody (E).
PINCH-1–ILK–α-Parvin Complex Is Important for Control of Podocyte–Matrix Adhesion and Shape Modulation
The increase of the PINCH-1–ILK–α-parvin complex level in the foot process–containing podocytes that were cultured under the nonpermissive condition (Figure 1) suggests that it may play a role in regulation of podocyte architecture. To test this experimentally, we sought to disrupt the formation of the PINCH-1–ILK–α-parvin complex in the podocytes. Toward this end, we infected the podocytes with an adenoviral vector encoding a FLAG-tagged ILK N-terminal fragment (ANK) that contains the PINCH-1–binding site and an adenoviral vector encoding β-galactosidase as a control, respectively. The expression of the FLAG-tagged ILK N-terminal fragment in the ANK adenoviral infected podocytes (Figure 2A, lane 2) but not in the control β-galactosidase adenoviral infected podocytes (Figure 2A, lane 1) was confirmed by Western blotting. Immunoprecipitation analysis showed that PINCH-1 (Figure 2B, lane 5) was co-immunoprecipitated with the FLAG-tagged ILK ANK fragment (Figure 2A, lane 5), whereas no PINCH-1 was precipitated with the anti-FLAG antibody in the absence of the ILK N-terminal fragment (Figure 2, lane 4). To provide an additional control for the functional studies, we generated an adenoviral vector encoding a FLAG-tagged ILK N-terminal fragment bearing a point mutation (D31→A) at the PINCH-binding site. The expression of the FLAG-tagged ILK N-terminal fragment (D31A) in the adenoviral infected podocytes was confirmed by Western blotting (Figure 2A, lane 3). Immunoprecipitation analysis showed that unlike the ILK N-terminal fragment that contained the wild-type PINCH-1–binding sequence (Figure 2, lane 5), the D31A fragment (Figure 2A, lane 6) failed to bind to PINCH-1 (Figure 2B, lane 6).
Overexpression of ILK mutants in podocytes. Mouse podocytes were cultured under the nonpermissive condition for 14 d and then infected with adenoviral vectors encoding β-galactosidase (lanes 1 and 4), the FLAG-tagged PINCH-1–binding ILK N-terminal fragment (lanes 2 and 5), or the FLAG-tagged PINCH-1–binding defective ILK N-terminal fragment bearing the D31A mutation (lanes 3 and 6). The FLAG-tagged ILK fragments were immunoprecipitated from the podocytes 3 d after infection with mouse monoclonal anti-FLAG antibody M2. The cell lysates (lanes 1 to 3; 10 μg/lane) and the anti-FLAG immunoprecipitates (lanes 4 to 6) were analyzed by Western blotting with mouse monoclonal anti-FLAG antibody M5 (A) and a rabbit polyclonal anti–PINCH-1 antibody (B).
To analyze the effect of the ILK ANK fragment on the PINCH-1–ILK–α-parvin complex formation in podocytes, we immunoprecipitated α-parvin from podocytes overexpressing the PINCH-1 binding ANK fragment (Figure 3A, lane 5), the PINCH-1 binding defective D31A fragment (Figure 3A, lane 6), or β-galactosidase (Figure 3A, lane 4). Analyses of the α-parvin immunoprecipitates showed that, as expected, ILK and PINCH-1 were readily co-immunoprecipitated with α-parvin in podocytes overexpressing β-galactosidase (Figure 3, lane 4) as well as those overexpressing the PINCH-1 binding defective D31A fragment (Figure 3, lane 6). By marked contrast, PINCH-1 (Figure 3C, lane 5) was not detected in the α-parvin immunoprecipitate (Figure 3A, lane 5) derived from the podocytes overexpressing the ILK ANK fragment, despite the presence of ILK in the α-parvin immunoprecipitate (Figure 3B, lane 5). Thus, overexpression of the PINCH-1 binding ANK fragment but not that of the PINCH-1 binding defective D31A fragment effectively disrupted the interaction between PINCH-1 and ILK (and consequently the formation of the PINCH-1–ILK–α-parvin complex).
Overexpression of the ILK N-terminal fragment but not that of the ILK N-terminal fragment bearing the D31A mutation inhibits the PINCH-1–ILK–α-parvin complex formation in podocytes. α-Parvin was immunoprecipitated from mouse podocytes that were infected with adenoviral vectors encoding β-galactosidase (lanes 1 and 4), the FLAG-tagged PINCH-1–binding ILK N-terminal fragment (lanes 2 and 5), or the FLAG-tagged PINCH-1–binding defective ILK N-terminal fragment bearing the D31A mutation (lanes 3 and 6) with monoclonal anti–α-parvin antibody 1D4 as described in the Materials and Methods section. The cell lysates (lanes 1 to 3; 10 μg/lane) and the anti–α-parvin immunoprecipitates (lanes 4 to 6) were analyzed by Western blotting with monoclonal anti–α-parvin antibody 3B5 (A), anti-ILK antibody 65.1 (B), and a rabbit polyclonal anti–PINCH-1 antibody (C).
We next analyzed the effect of disruption of the PINCH-1–ILK–α-parvin complex on podocyte–matrix adhesion and shape modulation. To do this, we plated the podocytes in which the formation of the PINCH-1–ILK–α-parvin complex was inhibited (Figure 3, lane 5) and the control podocytes that assembled a normal amount of the PINCH-1–ILK–α-parvin complex (Figure 3, lanes 4 and 6) on laminin. The results showed that disruption of the PINCH-1–ILK–α-parvin complex significantly reduced podocyte adhesion to laminin (Figure 4A). Furthermore, whereas the majority of the β-galactosidase and the D31A control podocytes (Figure 4, B, D, and E) spread soon after plating, most of the ANK-overexpressing podocytes (Figure 4, C and E) remained round 45 min after the plating, suggesting that the formation of the PINCH-1–ILK–α-parvin complex is crucial for proper control of podocyte morphologic changes.
Inhibition of the PINCH-1–ILK–α-parvin complex impairs podocyte–matrix adhesion and spreading. Mouse podocytes that were infected with adenoviral vectors encoding β-galactosidase (Control), the FLAG-tagged PINCH-binding ILK ankyrin fragment (ANK), and the FLAG-tagged PINCH-binding defective ILK ANK fragment bearing the D31A mutation (D31A) were seeded in laminin-coated 96-well plates. Cell adhesion (A) and spreading (B through D) were analyzed as described in the Materials and Methods section. Bar = 20 μm. The percentage of the podocytes overexpressing β-galactosidase (control), the PINCH-1–binding ILK N-terminal fragment (ANK), and the PINCH-1–binding defective ILK N-terminal fragment bearing the D31A mutation (D31A) that adopted spread morphology was quantified as described in the Materials and Methods section (E). Data represent means ± SD.
PINCH-1–ILK–α-Parvin Complex Protects Podocytes from Apoptosis
During the investigation, we repeatedly observed that many podocytes that were infected with the adenoviral vector encoding the PINCH-1 binding ILK N-terminal fragment were lost during culture. This observation raised an interesting possibility that the formation of the PINCH-1–ILK–α-parvin complex is crucial not only for regulation of podocyte morphology but also for protection of podocytes from apoptosis. To test this, we analyzed the podocytes by TUNEL staining. The results showed that the number of apoptotic (TUNEL-positive) cells was significantly increased in the podocytes overexpressing the PINCH-1–binding ILK N-terminal fragment (Figure 5, A through D, and G). By contrast, no significant increase of the number of apoptotic cells was observed in podocytes overexpressing the PINCH-1–binding defective D31A fragment (Figure 5, E through G). Analyses of caspase-3, a key mediator of apoptosis, showed that its activity was significantly increased in podocytes overexpressing the PINCH-1–binding ILK N-terminal fragment but not in those overexpressing the PINCH-1–binding defective D31A fragment (Figure 5H). These results provide strong evidence suggesting that the formation of the PINCH-1–ILK–α-parvin complex is crucial not only for regulation of podocyte morphology but also for protection of podocytes from apoptosis.
Inhibition of the PINCH-1–ILK–α-parvin complex promotes podocyte apoptosis. Mouse podocytes overexpressing β-galactosidase (A and B), the PINCH-1–binding ILK N-terminal fragment (C and D), or the PINCH-1–binding defective ILK N-terminal fragment bearing the D31A mutation (E and F) were fixed and incubated in the terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick-end labeling (TUNEL) reaction mixture, washed, and stained with DAPI. The apoptotic (TUNEL-positive) cells (A, C, and E) and total cells (DAPI positive; B, D, and F) were detected by fluorescence microscopy. (G) The percentages of the TUNEL-positive cells were calculated by counting at least 800 cells from five randomly selected fields. Data represent means ± SD from three independent experiments. (H) The activities of caspase-3 in the mouse podocytes overexpressing β-galactosidase (Control), the PINCH-1–binding ILK N-terminal fragment (ANK), and the PINCH-1–binding defective ILK N-terminal fragment bearing the D31A mutation (D31A) were quantified as described in the Materials and Methods section. The relative caspase-3 activities were expressed as fold increases over that of the control cells (the caspase-3 activity of the podocytes expressing the ILK mutants/the caspase-3 activity of the control cells). Data represent means ± SD from three independent experiments.
Treatment of the Podocytes with ILK Inhibitor MC-5 Reduces Stress Fiber Formation and Increases Apoptosis
To investigate further the functional significance of ILK in podocyte cytoskeleton organization and survival, we treated the podocytes with a small compound ILK inhibitor (MC-5; Kinetek Pharmaceuticals) (28). The results showed that treatment of the podocytes with the ILK inhibitor resulted in a lower frequency of stress fibers with a subcortical aggregated distribution (Figure 6A). Podocytes that were treated for 24 h with 1 μM ILK inhibitor MC-5 show a slight reduction and cells that were treated with 5 μM MC-5 showed a significant (n = 3; t test P < 0.05) reduction of well-developed actin stress fibers (Figure 6A). Similar alterations of the actin cytoskeleton have been described in murine ILK−/− cells (29). Treatment of the podocytes with the ILK inhibitor also increased the frequency of apoptotic (TUNEL-positive) cells (Figure 6B). The apoptosis frequency increased from 8.7% (basal conditions) to 10.3% (1 μM ILK inhibitor; n = 4; P < 0.05) and 14.7% (5 μM ILK inhibitor; n = 4; P < 0.05). Thus, the results obtained with the small compound ILK inhibitor are consistent with those obtained with the dominant negative ILK ANK fragment. Together, they suggest a crucial role of the ILK kinase activity in podocyte cytoskeleton organization and survival.
Treatment of podocytes with ILK inhibitor MC-5 reduces stress fiber formation and promotes apoptosis. (A) Effect of ILK inhibition on actin stress fiber formation. Mouse podocytes that were cultured on a collagen I–coated surface were treated for 24 h with different doses of an ILK-specific inhibitor (MC-5) as described in the Materials and Methods section. Thereafter, the F-actin was stained with phalloidin. Podocytes usually possess a distinct cytoskeleton with long F-actin stress fibers as can be observed in untreated 0 μM MC-5. Cells that were treated with ILK inhibitor MC-5 show fewer stress fibers and a more cortical localization of the F-actin where it forms aggregates. Cells with long, well-defined stress fibers were evaluated and counted blindly. A significant reduction (n = 3; P < 0.05) of stress fibers can be observed after 24 h of incubation with 5 μM ILK inhibitor MC-5. Treatment of podocytes that were cultured on a laminin-coated surface with MC-5 resulted in a similar alteration of the actin filaments (data not shown). (B) Inhibition of ILK promotes podocyte apoptosis. Mouse podocytes that were cultured on a collagen I–coated surface were incubated for 24 h with the ILK inhibitor MC-5, fixed, and subsequently incubated with the TUNEL reaction mixture and DAPI. Data represent the percentage of TUNEL-positive cells and means ± SD from four independent experiments (1200 cells per experiment were analyzed). An induction of apoptosis can be seen after treatment with 1 μM (+3.6%) and 5 μM ILK inhibitor MC-5 (+6%) (both dosages compared with the control; n = 4; P < 0.05).
N-Terminal Phosphorylation Sites of α-Parvin Are Involved in the Modulation of Complex Formation with ILK and PINCH-1
The N-terminus of α-parvin consists of several proline-directed serine phosphorylation sites. Recent studies have demonstrated that the α-parvin N-terminal sites can be phosphorylated by cdc2 and mitogen-activated protein kinases (MAPK) (30,31). To test the functional significance of α-parvin phosphorylation in podocytes, we generated an adenoviral expression vector encoding a FLAG-tagged α-parvin mutant (ΔN) in which the N-terminal phosphorylation sites are deleted. Podocytes were infected with the adenoviral vector encoding FLAG-ΔN and that encoding β-galactosidase as a control, respectively. The expression of FLAG-ΔN in the FLAG-ΔN adenoviral infected podocytes (Figure 7A, lane 2) but not the control β-gal adenoviral infected podocytes (Figure 7A, lane 1) was confirmed by Western blotting with a monoclonal anti-FLAG antibody. The overexpressed FLAG-tagged α-parvin mutant, together with the endogenous α-parvin, was immunoprecipitated from the podocytes overexpressing FLAG-ΔN with monoclonal anti–α-parvin antibody 1D4 (Figures 7, A and B, lane 4). As expected, α-parvin (Figure 7B, lane 3) but not FLAG-ΔN (Figure 7A, lane 3) was detected in the α-parvin immunoprecipitate derived from the control cells. To test whether α-parvin is phosphorylated in podocytes, we probed the α-parvin immunoprecipitate derived from the control cells with a monoclonal anti–phospho-Ser/Thr antibody (MPM2) that preferentially recognizes certain proline-containing phospho-Ser/Thr epitopes (32). The results showed that α-parvin was readily recognized by the monoclonal anti–phospho-Ser/Thr antibody (Figure 7C, lane 3), indicating that α-parvin is phosphorylated in the podocytes. Consistent with the overexpression of FLAG-ΔN in the FLAG-ΔN adenoviral infectants, more α-parvin proteins (the endogenous α-parvin plus FLAG-ΔN) were detected in the cell lysates (Figure 7B, compare lanes 1 and 2) and the immunoprecipitate derived from the FLAG-ΔN overexpressing podocytes (Figure 7B, compare lanes 3 and 4). Despite the increase of the amount of the α-parvin proteins in the immunoprecipitate derived from the FLAG-ΔN overexpressing podocytes, no increase of the amount of the phospho–α-parvin was detected (Figure 7C, compare lanes 3 and 4). Consistent with this, much less phospho–α-parvin protein was detected in the FLAG-ΔN overexpressing podocytes (Figure 7C, compare lane 5 with lane 3) when the amount of the sample was adjusted so that it contained a similar amount of α-parvin proteins as that derived from the control cells (Figure 7B, lanes 3 and 5). Probing the same samples with a monoclonal anti-ILK antibody revealed that less ILK was associated with α-parvin proteins derived from the FLAG-ΔN overexpressing podocytes (Figure 7D, compare lanes 3 and 5). Taken together, these results suggest that overexpression of the α-parvin mutant in which the N-terminal phosphorylation sites are deleted significantly reduces the phosphorylation of α-parvin and concomitantly decreases its complex formation with ILK in podocytes.
Effect of deletion of α-parvin N-terminal phosphorylation sites on its complex formation with ILK. α-Parvin was immunoprecipitated from the podocytes overexpressing the FLAG-tagged α-parvin mutant in which the N-terminal 22 residues consisting of the proline-containing phosphorylation sites are deleted (lanes 2 and 4) or the control podocytes overexpressing β-galactosidase (lanes 1 and 3) with monoclonal anti–α-parvin antibody 1D4 as described in the Materials and Methods section. The cell lysates (lanes 1 and 2; 10 μg/lane) copy. (G) The percentages of the TUNEL-positive cells were calculated by counting at least 800 cells from five randomly selected fields. Data represent means ± SD from three independent experiments. (H) The activities of caspase-3 in the mouse podocytes overexpressing β-galactosidase (Control), the PINCH-1–binding ILK N-terminal fragment (ANK), and the PINCH-1–binding defective ILK N-terminal fragment bearing the D31A mutation (D31A) were quantified as described in the Materials and Methods section. The relative caspase-3 activities were expressed as fold increases over that of the control cells (the caspase-3 activity of the podocytes expressing the ILK mutants/the caspase-3 activity of the control cells). Data represent means ± SD from three independent experiments.
Overexpression of α-Parvin Lacking N-Terminal Phosphorylation Sites Impairs Podocyte Adhesion, Shape Modulation, and Survival
We next analyzed the effects of overexpression of the α-parvin mutant lacking the N-terminal phosphorylation sites on podocyte behavior. The results showed that it significantly reduced podocyte–matrix adhesion and spreading (Figure 8, A through C). Furthermore, both the caspase-3 activity (Figure 8E) and the percentage of the TUNEL-positive cells (Figure 8F) were markedly increased in the FLAG-ΔN overexpressing podocytes, suggesting that the α-parvin proline-directed serine phosphorylation sites are critically involved in the proper control of podocyte adhesion, shape modulation, and survival.
Overexpression of α-parvin lacking the N-terminal phosphorylation sites impairs podocyte adhesion, shape modulation, and survival. Mouse podocytes were infected with adenoviral vectors encoding β-galactosidase (Control) or the FLAG-tagged α-parvin mutant in which the N-terminal phosphorylation sites are deleted (ΔN). Cell spreading (A through C), adhesion (D), caspase-3 activities (E), and TUNEL (F) were analyzed as described in Figures 3 and 4. Bar = 30 μm in A.
Discussion
Proper podocyte adhesion to extracellular matrix, morphology, and survival are crucial to normal renal functions (1,2,4,6,8). There is ample evidence now that injury to podocytes leads to proteinuria, a hallmark of most glomerular diseases. Despite the well-documented importance of podocytes in renal physiology and pathology, the molecular mechanisms that control podocyte–matrix adhesion, morphology, and survival are inadequately understood. Recent studies have implicated ILK as an important component of cell–matrix adhesions (10–12,15). The studies presented in this article demonstrate that PINCH-1, ILK, and α-parvin form a ternary complex in podocytes. Furthermore, using an inducible podocyte differentiation system, we show that the level of the PINCH-1–ILK–α-parvin complex is elevated accompanying the formation of the foot processes during podocyte differentiation. Importantly, either disruption of the PINCH-1–ILK–α-parvin complex with dominant negative ILK inhibitor or inhibition of ILK activity with a small compound ILK inhibitor impairs podocyte cytoskeleton organization and survival. Thus, the PINCH-1–ILK–α-parvin complex likely serves as a key regulator of podocyte behavior, and its dysregulation results in podocyte architecture alteration and apoptosis and consequently the breakdown of the filtration barrier and proteinuria.
In addition to demonstrating that the PINCH-1–ILK–α-parvin complex is crucial for podocyte functions, the studies described in this article provide important evidence for a regulatory role of the α-parvin N-terminus in podocyte–matrix adhesion, morphology, and survival. How does the α-parvin N-terminus regulate podocyte functions? That (1) the α-parvin N-terminus contains several phosphorylation sites (30), (2) deletion of the N-terminal phosphorylation sites reduces the phosphorylation and concomitantly the association of α-parvin with ILK (Figure 7), and (3) the phenotypical changes induced by the overexpression of the α-parvin phosphorylation mutant (Figure 8) are remarkably similar to those of disruption of the PINCH-1–ILK–α-parvin complex (Figures 4 and 5) strongly suggest that the α-parvin N-terminus regulates podocyte functions through, at least in part, influencing the PINCH-1–ILK–α-parvin complex formation.
Previous studies have demonstrated that the CH2 domain of α-parvin is both necessary and sufficient for interaction with ILK (16,18). How does the α-parvin N-terminus influence its association with ILK? Although a complete answer must await detailed information on the tertiary structure of the ILK–α-parvin complex, several possibilities could be delineated. One possibility is that although the N-terminal region and the CH2 domain of α-parvin are located at two termini at the amino acid sequence level, they could be proximal at the tertiary structure level and therefore alterations of the N-terminal phosphorylation could influence the ILK binding activity mediated by the C-terminal CH2 domain of α-parvin. A second possibility is that phosphorylation at the N-terminal region could influence additional phosphorylation at or near the ILK binding site within the CH2 domain and thereby regulate the association with ILK. Finally, the N-terminal region of α-parvin could interact with other molecules that modulate the ILK binding activity of α-parvin and consequently phosphorylation at the α-parvin N-terminal region could alter its interaction with these molecules and indirectly influence the complex formation between α-parvin and ILK.
Irrespective of the specific mechanism, the finding that the α-parvin N-terminus is involved in the regulation of the phosphorylation and association of α-parvin with ILK is significant, as it suggests a novel pathway that links Ser/Thr-phosphorylation to the formation of the PINCH-1–ILK–α-parvin complex and consequently the regulation of podocyte–matrix adhesion, morphology, and survival. The N-terminus of α-parvin contains consensus sequences (PX[S/T]P and [S/T]PXR/K) that are essential for phosphorylation by families of proline-directed kinases, including families of MAPK and cdc2. Recent biochemical studies have demonstrated that the N-terminus of α-parvin can be phosphorylated by cdc2 and MAPK (30,31). Thus, proline-directed kinases such as those of the cdc2 and/or MAPK families are strong candidates for catalyzing the phosphorylation of α-parvin and consequently regulating the PINCH-1–ILK–α-parvin complex formation and the adhesion, architecture, and survival of podocytes. Under normal condition, cdc2 is crucially involved in the regulation of the G2/M transition of cell-cycle progression (33), a process accompanying extensive changes of cell–matrix adhesion and morphology. Thus, the α-parvin phosphorylation-mediated regulation of the PINCH-1–ILK–α-parvin complex formation could play a role in the regulation of podocyte morphology during glomerulogenesis in which presumptive podocytes undergo active cell-cycle progression. In addition, although mature podocytes are quiescent under normal condition, they do proliferate under certain pathologic conditions such as collapsing forms of focal segmental glomerulosclerosis, including idiopathic focal segmental glomerulosclerosis and HIV-associated nephropathy, which are known to associate with dramatic morphologic alterations including loss of foot processes, apoptosis, and a rapid decline in renal function (3,34–36). Furthermore, it has been shown that podocyte cdc2 level is significantly increased in the passive Heymann nephritis model of membranous nephropathy despite the lack of proliferation of the podocytes (37,38), suggesting that cdc2 could play an important role in regulation of podocyte architecture and survival in nonproliferating podocytes. It will be interesting to test in future studies whether cdc2 participates in the phosphorylation of α-parvin and regulation of the PINCH-1–ILK–α-parvin complex formation in podocytes. Finally, other proline-directed kinases, such as members of MAPK family, can also phosphorylate the N-terminus of α-parvin, and phosphorylation of the α-parvin N-terminal region by these protein kinases could contribute to the regulation of the PINCH-1–ILK–α-parvin complex formation and consequently the adhesion, architecture, and survival of podocytes. The PINCH-1–ILK–α-parvin complex, therefore, likely serves as an important converging point mediating the effects of multiple upstream stimuli that alter podocyte–matrix adhesion, cytoarchitecture, and survival.
Acknowledgments
This work was supported by National Institutes of Health Grants DK54639 and GM65188 to C.W. and DFG Kr1492/6-3+4 to M.K.
We thank Drs. Tong-Chuan He and Bert Vogelstein for providing the pAdTrack-CMV and pADEsay-1 vectors.
- © 2005 American Society of Nephrology