Abstract
The foot processes of podocytes possess abundant microfilaments and modulate glomerular filtration. We investigated the actin filament organization of foot processes in adult rat podocytes and the formation of the actin cytoskeletal system of immature podocytes during glomerulogenesis. Electron microscopy revealed two populations of actin cytoskeletons in foot processes of adult podocytes. One is the actin bundle running above the level of slit diaphragms and the other is the cortical actin network located beneath the plasmalemma. Immunogold labeling for actin-binding proteins demonstrated that oí-actinin and synaptopodin were localized in the actin bundle, whereas cortactin was in the cortical actin network. Immunofluorescence labeling for actin-binding proteins in immature podocyte showed that α-actinin was localized at the level of the junctional complex, whereas cortactin was distributed beneath the entire plasmalemma. Synaptopodin was first observed along the basal plasmalemma from the advanced S-shaped body to the capillary loop stage. We conclude that foot processes have specialized actin filamentous organization and that its establishment is associated with the expression and redistribution of actin-binding proteins during development.
T
The actin cytoskeleton of the foot process plays important roles in maintaining its own unique structure (Andrews 1981) and in supporting the glomerular capillary wall (Kriz et al. 1994). The glomerular capillary has higher hydrostatic pressure than that in other tissues, which is crucial for maintaining glomerular filtration. Consequently, the glomerulus requires supporting systems to protect the glomerular capillary wall against such pressure. One such supporting system is the coupling with the mesangial cell and the glomerular basement membrane (GBM) (Sakai and Kriz 1987), and another is the foot processes that cover the urinary surface of the GBM. The foot processes of the podocyte have also been considered to control the extreme expansion of the capillary wall, because some contractile proteins (α-actinin and myosin) are known to exist in the foot processes (Drenckhahn and Franke 1988).
The foot processes are interdigitated with those of neighboring podocytes, and the intercellular space between adjacent foot processes is bridged by a slit diaphragm. The molecular components of the slit diaphragms have gradually been clarified. The tight junction protein ZO-1 was first identified as a slit diaphragm-associated protein and is concentrated at the points of insertion of the slit diaphragm (Schnabel et al. 1990). ZO-1 has two isoforms, and only one of its isoforms, ZO-1α –, is expressed in podocytes (Kurihara et al. 1992a). ZO-1 has been shown to be associated with the actin cytoskeleton (Fanning et al. 1998). Nephrin is a component of the slit diaphragm protein and is encoded by NPHS1, which is a causative gene of Finnish type congenital nephrotic syndrome (Kestilä et al. 1998). This protein has been demonstrated to link with actin cytoskeleton in the foot processes via a mediator protein CD2AP (Yuan et al. 2002).
Furthermore, it has been suggested that the peripheral membrane proteins linking to the actin cytoskeleton are separated by the membrane domain of the podocyte (Kurihara et al. 1995). It has also been suggested that those actin-binding proteins play an important role in the linkage between the transmembrane protein and the actin cytoskeleton (Drenckhahn and Franke 1988; Orlando et al. 2001).
The rearrangement of the cytoskeleton is crucial for tissue remodeling under both developmental (Kurihara et al. 1998) and pathological (Shirato et al. 1996) conditions. In the developing glomerulus, the cellular architecture of the immature podocyte is dramatically rearranged to form the foot processes. The undifferentiated podocytes at S-shaped body stages show a columnar epithelium and connect to each other by the typical junctional complex, which migrates from the apex to the base of the cell during glomerular maturation. In the capillary loop stage, the splay processes cover the urinary surface of the GBM and vigorous interdigitations between the processes occur to form the foot processes (Reeves et al. 1978). Actin reorganization might be involved in foot process formation, but the precise mechanism is still unknown.
In this study we demonstrate morphologically the existence of two populations of actin cytoskeleton in the foot processes of adult rat podocyte for the first time, and demonstrate the kinds of actin-associated proteins that construct these two populations. We chose three actin-associated proteins, α-actinin, cortactin, and synaptopodin, and investigated their participation in the maintenance of these two actin-based cytoskeletal populations. α-Actinin (Trenchev et al. 1976; Drenckhahn and Franke 1988; Lachapelle and Bendayan 1991) and synaptopodin (Mundel et al. 1991) are already reported to be expressed in podocytes. Cortactin, an actin-binding protein, is associated with the formation of the cortical actin meshwork in wide variety of cells (Wu and Parsons 1993; Wu and Montone 1998). In addition, we trace the course of foot process formation during glomerulogenesis from the aspect of expression and localization of three actin-associated proteins.
Materials and Methods
Reagents and Antibodies
Polyvinyl alcohol (MW 10,000), polyvinylpyrrolidone (MW 10,000), antipain, benzamidine, leupeptin, pepstatin A, PMSF, and BSA (fraction V) were obtained from Sigma-Aldrich (St Louis, MO). Two kinds of mouse anti-α-actinin monoclonal antibodies (MAbs) AT6.172 (IgG) and BM75.2 (IgM) were from Chemicon (Temecula, CA) and Sigma-Aldrich, respectively. AT6.172 recognizes all isoforms of α-actinin, while BM75.2 is specific for actinin-1 and the specificity of this antibody has been reported (Honda et al. 1998). Rabbit polyclonal anti-actinin-4 antibody was a kind gift from Dr. Setsuo Hirohashi and Dr. Tesshi Yamada (National Cancer Center Research Institute; Tokyo, Japan). Mouse anti-cortactin MAb (IgG) was from Upstate Biotechnical (Waltham, MA), and rabbit polyclonal anti-ZO-1 antibody was from Zymed (San Francisco, CA). Mouse anti-synaptopodin MAb (IgG) was from Progen (Heidelberg, Germany). Rhodamine (TRITC)-conjugated donkey anti-mouse IgG F(ab')2 fragment, fluorescein (FITC)-conjugated donkey anti-rabbit IgG F(ab')2 fragment, and TRITC-conjugated donkey anti-mouse IgM were from Jackson Immuno-Research Laboratories (West Grove, PA). Goat anti-mouse IgG coupled to 5-nm or 10-nm colloidal gold was obtained from Amersham (Arlington Heights, IL).
Animals
Male Wistar rats (4–6 weeks old) and neonatal rats (1–2 days old) were obtained from Charles River Japan (Kanagawa, Japan). They were kept in an air-conditioned room and maintained on a commercial stock diet and tapwater ad libitum.
Preparation of Glomeruli
Rat kidneys were perfused with PBS, pH 7.4, containing protease inhibitors (1 mM each antipain, benzamidine, leupeptin, pepstatin A, and PMSF) under anesthesia with Nembutal. Rat glomeruli were isolated by graded sieving at 4C in the presence of protease inhibitors. Glomerular fractions were collected by centrifugation (800 × g for 5 min) and washed with PBS containing protease inhibitors (PBS-PI). Isolated glomeruli were extracted with PBS-PI containing 1% Triton X-100 (TX-100) for 3–5 min at room temperature (RT). The materials were then centrifuged at 800 × g for 10 min at 4C. The pellet was processed for conventional electron microscopy.
Confocal Laser Scanning Microscopy
Rat kidneys were perfused with 4% paraformaldehyde fixative buffered with 0.1 M phosphate buffer (PB, pH 7.4) under anesthesia with Nembutal and then cut into small pieces. These samples were immersed in the same fixative for about 30 min. After washing with PBS, the tissue was immersed successively in PBS solution containing 10%, 15%, and 20% sucrose (respectively for 4 hr, 12 hr, and 4 hr). After the tissue was embedded in OCT compound and frozen, cryosections (thickness 5–10 μm) were cut using a Jung Frigocut 2800E (Leica; Wetzlar, Germany) and then mounted on silane-coated glass slides. The cryosections were rinsed with PBS and blocked in blocking solution (0.1% BSA in PBS). The sections were incubated for 2 hr at RT with AT6.172 (1:100 dilution), BM75.2 (1:100), rabbit polyclonal anti-actinin-4 antibody (1:1000), mouse anti-cortactin MAb (1:100), mouse anti-synaptopodin MAb (1:25), or rabbit polyclonal anti-ZO-1 antibody (1:100). Then the sections were incubated for 1 hr at RT with TRITC-conjugated anti-mouse IgG or IgM (1:200), TRITC-conjugated anti-rabbit IgG (1:200), or FITC-conjugated anti-rabbit IgG (1:100). Fluorescence specimens were viewed with a confocal laser scanning microscope LSM510 (Carl Zeiss; Oberkochen, Germany). As the negative control experiment, the primary antibodies were either omitted from the incubation solution or substituted by normal mouse IgG or normal rabbit serum (Vector Laboratories; Burlin-game, CA) at the same concentrations as the primary antibodies. The specificities of primary antibodies used in this study were preliminarily confirmed by immunoblotting analysis.
Cold Dehydration Method for Transmission Electron Microscopy
Rat kidneys were perfused with 2.5% glutaraldehyde fixative buffered with 0.1 M PB (pH 7.4) under anesthesia with Nembutal and immersed in the same fixative for about 12 hr. The specimens were sectioned at 200-μm thickness with a Microslicer and processed further by a modified cold dehydration technique. The original protocol for the cold dehydration technique has been described by Sakai and Kriz (1987). The samples were successively immersed in 0.1% OsO4 in 0.1 M PB for 30 min, 5% extracts of oolong tea (OTE) in 0.05 M maleate buffer for 3 hr, and 1% uranyl acetate in 0.05 M maleate buffer solution. Then the specimens were dehydrated with a graded series of acetone at 0C to −30C before embedding in Epon 812. These procedures enabled detailed morphological observation of the extracellular matrices and intracellular fibrils. Ultrathin sections (80 nm) were processed with a diamond knife, transferred to copper grids (50 mesh) that had been coated with Formvar membrane, stained with uranyl acetate and lead citrate, and observed in a transmission electron microscope JEM1230 (JEOL; Tokyo, Japan).
Immunogold Labeling of Ultrathin Cryosections
The samples fixed with paraformaldehyde as above mentioned were rinsed and infiltrated with 40% polyvinyl alcohol/2.3 M sucrose buffered with 0.1 M PB and embedded on nails for immunogold electron microscopy. Ultrathin cryosections were cut with the Ultracut UCT microtome equipped with the FC-4E cryoattachment (Leica; Vienna, Austria) at −110C. Sections were transferred to nickel grids (150 mesh) that had been coated with a Formvar membrane. Subsequent incubation steps were carried out by floating the grids on droplets of the filtered solution. After quenching free aldehyde groups with PBS-0.01 M glycine, sections were incubated overnight with AT6.172 (1:100), mouse anti-cortactin MAb (1:100) and mouse anti-synaptopodin MAb (1:25). They were then incubated with anti-mouse IgG coupled to 5-nm or 10-nm colloidal gold (1:100) for 1 hr. After immunostaining, they were fixed with 2.5% glutaraldehyde buffered with 0.1 M PB. The sections then were contrasted with 2% uranyl acetate solution for 20 min and absorption-stained with 3% polyvinyl alcohol containing 0.2% uranyl acetate for 10 min. As the negative control experiment, the primary antibodies were either omitted from the incubation solution or substituted by normal mouse IgG (Vector Laboratories) at the same concentrations as the primary antibodies.
Results
Actin Cytoskeleton in the Foot Processes of Adult Rat Podocytes
The major cytoskeleton in the foot processes of podocytes consisted of actin filaments, which were represented either by dense actin bundles or by a cortical network, as visualized by a modified cold dehydration technique using en bloc staining with OTE as described in Materials and Methods. The dense bundles of actin filaments were found predominantly at the central portion of the upper cytoplasm of the foot processes of podocytes, above the level of the slit diaphragm (Figure 1A). The actin bundles ran longitudinally in the foot processes extending from the primary processes, as visualized in specimens sectioned parallel to the base of the cells (Figure 1B). The cortical cytoplasm of foot processes between the actin bundles and cell membranes was occupied by the cortical network of actin filaments, which were sparse in the apical cytoplasm and moderately dense in the basal cytoplasm. Therefore, the actin bundles were separated from the apical cell membranes by a sparse cortical network and did not have any close relation with the slit diaphragm. At the most distal part of the foot processes, the actin bundles terminated, usually at the neck without entering the enlarged portion at the tip of the cell process. At the origin of foot processes, the actin bundles of the two neighboring foot processes were connected with each other by arcuate actin filaments running along the cell membrane between the foot processes.
The organization of the cortical network in foot processes was clearly visualized by treatment of isolated glomeruli with TX-100. The apical cortical cytoplasm between the actin bundles and the apical cell membranes contained sparse filaments that were connected with the bundles and membranes. The basal cortical cytoplasm of foot processes contained a fine meshwork of microfilaments made up of actin (Figures 1C and 1D). The filaments of the basal cortical actin network were associated with dense plaques at the base of the slit diaphragms (Figure 1C). The filaments of the cortical actin network were frequently merged with those of actin bundles, and therefore clear demarcation of these two populations of actin filaments was not always possible.
Localization of Three Actin-associated Proteins in Adult Rat Glomeruli
Electron microscopy suggested that the foot processes contained two distinct populations of actin cytoskeletons: actin bundles at the central portion of the cytoplasm and a cortical actin network at the cell periphery. Therefore, we localized three actin-associated proteins, α-actinin, cortactin, and synaptopodin, in adult rat glomeruli to analyze the two types of actin filaments in the foot processes by using confocal laser scanning microscopy with specific antibodies.
Electron micrographs of podocyte foot processes. Electron microscopic specimens were made by use of either the modified cold dehydration technique (
The remarkable immunoreactivities of all three actin-associated proteins in adult rat glomeruli were observed as a linear staining pattern around the lumen of glomerular capillary (Figures 2A–2C). The staining patterns of three actin-binding proteins in glomeruli were quite similar to that of ZO-1 (Figure 2D). Therefore, we tried to examine the co-localization of actin-binding proteins with ZO-1 by immunofluorescence confocal microscopy. The immunoreactivity of α-actinin corresponded perfectly to that of ZO-1 in the glomerulus (Figure 3D). In adult rat glomeruli, the conspicuous immunoreactivity of cortactin was detected as a linear pattern that corresponded to that of ZO-1. In addition, the notable labeling of cortactin was also observed in the mesangial area (Figure 4D). The cell bodies and primary processes of podocytes and glomerular endothelial cells showed faint staining of cortactin (Figure 4D). The double immunofluorescence study for synaptopodin with ZO-1 demonstrated that the linear staining of synaptopodin corresponded predominantly to that of ZO-1 but that a part of the positive reaction for synaptopodin was independent of that of ZO-1 (Figure 5D).
Confocal laser scanning images showing the localization of α-actinin, cortactin, synaptopodin, and ZO-1 in adult rat glomeruli. The positive stainings for α-actinin (
The immunofluorescence signal for each actin-binding protein corresponded to that of ZO-1 as a linear staining pattern in adult glomeruli, suggesting that the three actin-binding proteins examined in this study were located at the level of the foot processes of the podocyte. To determine the precise localization of three actin-binding proteins in podocytes, we used immunogold electron microscopy with cryoultrathin sections.
Both α-actinin and synaptopodin were concentrated predominantly in the actin bundle running above the level of the slit diaphragm in mature foot processes (Figures 6A and 6C). α-Actinin was localized in the entire region of the actin bundle and was not recognized in the cortical actin network containing the electron-dense deposits at the intracellular basis of slit diaphragm (Figure 6A). On the other hand, the gold particles for synaptopodin were located in the proximal portion of actin bundle (Figure 6C) and were hardly observed in the distal portion. In addition, synaptopodin was also recognized in the major processes.
Cortactin was located in both the cortical actin network of the foot processes and the cell periphery of the cell body and major processes in the podocye (Figure 6B). The gold particles for cortactin were recognized uniformly throughout the entire region of the cortical actin network on the perpendicular section, but were not concentrated in the particular region of the cortical actin network.
Localization of Three Actin-associated Proteins in Immature Podocytes of Newborn Rat Kidney
The localizations of three actin-associated proteins together with that of ZO-1 were visualized in each developmental phase of the podocytes. The immunostaining of ZO-1 corresponding to the location of the tight junction or slit diaphragm represented the border between the apical and basolateral cell membranes in each developmental stage. In the S-shaped body stage, the podocytes were columnar in shape and ZO-1 migrated from the apical (Figures 3A, 4A, and 5A) along the lateral (Figures 3B and 4B) to the basal position (Figure 5B). In the capillary loop stage, the band-like staining of ZO-1 at the base of podocytes was irregularly contoured by the emergence of capillaries (Figures 3C, 4C, and 5C).
Confocal laser scanning images showing the localization of α-actinin and ZO-1 in developing and mature glomeruli. Double immunofluorescence stainings for α-actinin (red) and ZO-1 (green) were performed. In this experiment, we used mouse MAb clone AT6.172 as the α-actinin-specific antibody. During glomerulogenesis, the signal for α-actinin consistently corresponded to that of ZO-1. In the early (
During glomerulogenesis, the signals for α-actinin consistently coincided with that of ZO-1 at the light microscopic level. In the early to middle period of the S-shaped body stage, the immunoreactivity of α-actinin typically showed a spot-like pattern at the location of the tight junction (Figure 3A). The spot-like immunoreactivity of α-actinin moved toward the basal side of the cell as well as ZO-1, which appeared either spotty or linear due to the direction of sectioning (Figure 3B). From the advanced S-shaped body stage to the early capillary loop stage, the positive reaction of ZO-1 was recognized as a band-like pattern along the basal surface of the presumptive podocytes (Figure 3C), suggesting that the membrane folding and the process formation of podocyte might occur vigorously in this term. The signal for α-actinin in the podocyte from the advanced S-shaped body stage to the early capillary loop stage also showed the band-like pattern and corresponded to that of ZO-1 (Figure 3C). In the advanced capillary loop and developing glomerular stages, both α-actinin and ZO-1 indicated the linear staining mode.
Confocal laser scanning images showing the localization of cortactin in developing and mature glomeruli. Double immunofluorescence stainings for cortactin (red) and ZO-1 (green) were performed. A positive reaction for cortactin is consistently detected throughout the plasma membrane of presumptive podocytes at all stages of glomerulogenesis. In the early (
Confocal laser scanning images showing the localization of synaptopodin in developing and mature glomeruli. In the early S-shaped body stage (
Immunogold electron micrographs showing the localization of α-actinin, cortactin, and synaptopodin in adult rat foot processes. (
α-Actinin is known to have four isoforms, i.e., two muscle isoforms, actinin-2 and actinin-3, and two non-muscle isoforms, actinin-1 and actinin-4. Actinin-2 and actinin-3 are limited to localization in the sarcomere of striated muscle. Recently, a mutation in ACTN4 encoding actinin-4 was reported to cause focal segmental glomerulosclerosis in human. Furthermore, actinin-4, but not actinin-1, was demonstrated to be present exclusively in human glomerulus (Kaplan et al. 2000). To investigate the expression of α-actinin isoforms in rat podocytes, we performed immunofluorescence staining by using isoform-specific antibodies against actinin-1 (mouse MAb clone BM75.2) and actinin-4 (rabbit anti-actinin-4 specific polyclonal antibody). The immunolabelings for both actinin-1 and actinin-4 consistently coincided with those of ZO-1 in podocytes during all stages of glomerulogenesis, including the mature stage in adult rat (data not shown).
Cortactin immunoreactivity was consistently detected along the entire plasma membrane of the presumptive podocyte in all stages of glomerulogenesis. In the early S-shaped body stage, the positive reaction of cortactin was recognized along the cell membrane of the immature columnar podocytes (Figures 4A and 4B), and its immunoreactivity at the location of tight junction showed much stronger intensity than at other staining sites (Figure 4B). Furthermore, in the presumptive podocytes at the early S-shaped body stage, the immunoreactive intensity beneath either the apical or basal plasma membrane appeared stronger than that of the lateral plasma membrane. From the advanced S-shaped body stage to the capillary loop stage, the cortactin immunoreactivity along the basolateral membrane of the podocyte showed stronger intensity than that of the apical membrane. In these stages, the positive reaction of cortactin showed the band-like pattern that corresponded to that of ZO-1 as seen in α-actinin (Figure 4C).
No positive reaction of synaptopodin was recognized in the early S-shaped body stage. Synaptopodin immunoreactivity was first detected as a faint linear pattern along the basal membrane of undifferentiated podocytes from the advanced S-shaped body stage to the early capillary loop stage (Figures 5A and 5B). The initial expression of synaptopodin is likely to coincide with the phase when the band-like stainings of α-actinin and cortactin appear. Throughout glomerulogenesis, the positive reaction of synaptopodin fundamentally showed a linear pattern along the basal surface of the presumptive podocytes. Most of the immunoreactive staining for synaptopodin corresponded to that of ZO-1. However, a partial positive reaction of synaptopodin did not coexist with that of ZO-1 (Figures 5B and 5C).
Discussion
In this study we have shown that two distinct actin cytoskeletal systems exist within the foot process by using two methods for visualizing cytoskeletal structures in detail, and revealed what kind of actin-binding proteins organize these specialized microfilament structures. Furthermore, we also demonstrated the formation of the microfilamentous system in the podocyte during glomerulogenesis from the viewpoint of the specific expression of actin-binding proteins.
Two Populations of Actin Cytoskeleton Are Distinguishable in the Foot Processes of Adult Rat Podocytes
Cytoskeletons play a critical role in sustaining cell structure, especially in several cell types with highly complicated morphology, such as podocytes and neurons. Previous studies on the cytoskeletal structures in podocytes have indicated that actin filaments are mainly concentrated in the foot processes, whereas microtubules and vimentin-type intermediate filaments are distributed in the cell body and the primary processes of podocytes (Andrews and Bates 1984; Vasmant et al. 1984; Drenckhahn and Franke 1988). However, the precise distribution of actin-based microfilaments in foot processes remains obscure. We sought to answer how actin filaments are organized in the foot processes of podocytes. To visualize the precise actin cytoskeletal structure, we adopted the modified cold dehydration technique using OTE en bloc staining for natural tissue and detergent treatment for isolated glomeruli (Kurihara et al. 1998). On the basis of the present results, we show that the actin cytoskeletal structure in the foot processes of adult podocytes is divided into two populations: actin bundles and a cortical actin network (Figure 1E). The actin bundle runs along the longitudinal axis of the foot process above the level of the slit diaphragm, while the cortical actin network distributes beneath the plasma membrane of the foot process. The diversity of actin cytoskeletal structures in various cell types is ensured by the multiplicity of actin-binding proteins. Each actin cytoskeletal structure is organized by particular actin-binding protein(s). Our immunoelectron microscopic observations strongly support two different subpopulations of actin filaments. Thus, α-actinin and synaptopodin are observed only in the actin bundles, whereas cortactin is associated with the cortical actin network.
It is believed that actin filaments in the foot processes play a role in cell-cell and cell-matrix adhesion (Drenckhahn and Franke 1988; Fanning et al. 1998; Yuan et al. 2002). α-Actinin is known to localize to cell–cell and cell–matrix contacts in many cell types (Otey et al. 1993; Knudsen et al. 1995). We have shown that α-actinin is one of the components forming the actin bundle but is not associated with the cell-matrix junction in normal podocytes. Increased α-actinin molecules under diseased conditions cause the disorganization of actin bundles and the effacement of foot processes (Smoyer et al. 1997). Therefore, regulated expression of α-actinin is necessary for normal actin bundle formation in foot processes. Synaptopodin, which is exclusively expressed in podocytes, is also associated with the actin bundle. However, the localization of synaptopodin is more confined to the border between foot processes and primary processes, where the arciform actin bundle is observed. The limited expression of synaptopodin suggests that this molecule performs specialized functions for maintaining the foot processes.
In various cultured cells, integrins known as cell-matrix adhesion molecules are distributed at focal contacts at which actin bundles, termed stress fibers, are terminated (Sastry and Burridge 2000). It should be emphasized that the actin bundles of the foot process do not directly contact the basal membrane in normal podocytes. For the foot processes to counteract the expansive forces of the capillary wall, it is reasonable to assume that the actin bundles of the foot processes link to the GBM via the cell-matrix adhesion molecule. Previous reports have suggested that submembranous cytoskeletal proteins (talin, vinculin, and paxillin) are located at the base of foot processes (Drenckhahn and Franke 1988; Kurihara et al. 1995). These proteins are known to mediate the interaction between integrins and actin filaments. Our present findings indicate that integrin and related molecules are associated with the cortical actin network, which is located at the space between the actin bundle and the basal plasma membrane. Therefore, the actin bundle and the integrin-mediated cell adhesion apparatus are presumed to be connected indirectly through the cortical actin-submembranous protein network.
Cortactin is localized in the peripheral actin layer termed the cell cortex in cultured cells (Wu and Parsons 1993), the terminal web of epithelial cell types (Wu and Montone 1998), and the podosomes of macrophages, including osteoclasts (Hiura et al. 1995). In the podocytes, cortactin was localized in the submembranous cytoplasm of the foot processes and the cell periphery of the cell body and the major processes, suggesting that cortactin is a major actin-binding protein in the cortical actin network. In observing the perpendicular section of foot process by electron microscopy, electron-dense materials were visible at the cytoplasmic insertion site of the slit diaphragm. The connection between the electron-dense deposits and actin filaments was also observed in TX-100-treated glomeruli, suggesting that the slit diaphragm binds to the actin filaments that construct the cortical actin network. Nephrin, a slit diaphragm protein encoded by NPHS1, which is a causative gene in Finnish-type congenital nephrotic syndrome (Katsube et al. 1998), has been reported to interact with the actin cytoskeleton via CD2AP (Yuan et al. 2002). A two-hybrid yeast experiment has indicated that cortactin directly binds to ZO-1 (Katsube et al. 1998). This suggests that cortactin plays an important role in the linkage between the slit diaphragm and actin filaments in the cortical actin network, although the direct interaction between nephrin and ZO-1 has not yet been reported.
Formation of the Actin Cytoskeleton in the Podocytes During Glomerulogenesis
During glomerulogenesis, three actin-binding proteins examined in the present study show individually different patterns of localization in immature podocytes. In developing podocyte, α-actinin was localized at the position of the junctional complex at the early S-shaped body stage and migrated with the junctional complex from the apical to the basal side of the cell, whereas cortactin was localized along the entire plasma membrane, where it was especially concentrated at the level of the tight junction. From the advanced S-shaped body stage to the early capillary loop stage, the process formation commenced at the basal side of the podocyte. During this period, α-actinin or cortactin was detected along the basal cell membrane as a bandlike pattern, whereas synaptopodin was first detected linearly along the basal cell membrane. Our previous data indicate that a novel intermediate filament-associated protein p250, exclusively expressed in podocytes, is first observed at the base of immature podocytes at the S-shaped body stage (Kurihara et al. 1998). Podocalyxin, a major sialoglycoprotein in podocytes, is also detectable at this stage (Schnabel et al. 1989). These data indicate that the cell-specific organization on both membranous proteins and cytoskeletons is initiated in the presumptive podocyte during the S-shaped body stage.
α-Actinin is also one of the components of the cell-cell adherens junction and plays a role in linking α-catenin and the annular actin cytoskeleton that serves as a cytoplasmic anchor for the components of the adherens junctions in typical epithelial cells (Youssoufian et al. 1990; Knudsen et al. 1995; Nieset et al. 1997). We have shown that α-actinin is initially associated with the junctional complex in immature podocytes (Figures 3A and 3B) and is located only in actin bundles in foot processes in mature podocytes, but is not associated with slit diaphragms (Figure 6C). Accordingly, during the course of foot process formation, podocytes are presumed to lose their annular actin cytoskeleton lining the intercellular junctional apparatus.
Synaptopodin was initially detected in the beginning period of process formation in immature podocytes, and was localized in the proximal region of actin bundles of the foot processes and in part of the actin cytoskeletons of the major processes in mature ones. It is possible that α-actinin plays a role in the migration of the junctional complex in immature podocytes and that this molecule and synaptopodin are necessary to construct the core actin bundles during foot process formation. On the other hand, the junctional structures between podocytes were consistently associated with actin cytoskeletal structures, including cortactin (i.e., cell cortex and cortical actin network), during glomerulogenesis. Furthermore, because the linkage between cortactin and ZO-1 was revealed on the basis of in vivo and in vitro studies (Fanning et al. 1998; Katsube et al. 1998), tight junctions are likely to possess an affinity for the actin cytoskeletal system built by cortactin. Therefore, the consistent association among the junctional apparatus of podocytes, including the slit diaphragm and the cortactin-related actin cytoskeleton, supports the hypothesis that the slit diaphragm is one of the highly specialized tight junctions (Kurihara et al. 1992b; Schnabel et al. 1990).
In conclusion, foot processes in the podocyte have specialized actin filament organization and its establishment is accompanied by the expression and redistribution of actin-binding proteins during development.
Footnotes
Acknowledgements
This work was supported in part by Grants-in-aid from the Ministry of Education, Science, Sports and Culture of Japan.