INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE VISCERAL GLOMERULAR epithelial cells or podocytes of the renal glomerulus are uniquely characterized by a large cell body and numerous foot processes. These foot processes are kept wide open to facilitate passage of glomerular filtrate and are held together by tenuous slit diaphragms that bridge the filtration slits (42). The structural integrity of the podocytic epithelium participates in the maintenance of the filtration barrier. The glomerular capillary wall is continuously exposed to a high transcapillary pressure gradient of ~35 mmHg. Undoubtedly, the glomerular epithelial cells (GECs) play a role in supporting the capillary wall against mechanical force (23). However, the mechanisms that maintain the shape of podocytes exposed to hydraulic pressure are poorly understood.

Agents that disrupt cytoskeletal elements such as microtubules and microfilaments cause morphological changes of podocytes, suggesting that the cytoskeleton plays an important role in the maintenance and alteration of the cell shape (20). The cytoskeletal changes in rat GECs may be regulated by vasoactive substances (44). Electron microscopic (EM) observations have demonstrated that podocytes, like typical epithelial cells, contain three major cytoskeletal elements (microfilaments, microtubules, and intermediate filaments) but with a different distribution. Actin-based microfilaments concentrate in the cytoplasm of the foot processes (1). Immunocytochemical observations have shown that actin-binding proteins (8, 33) and junctional proteins are located in the foot processes (26). On the other hand, intermediate filaments and microtubules are distributed throughout the cell body and the major processes but are virtually absent from the foot processes (2, 8, 50).

Among three fibrillar cytoskeletal systems, the expression of intermediate filaments depends on cell type at the phase of cellular differentiation. In general, epithelial cells express cytokeratin-type intermediate filaments, whereas mesenchymal cells express vimentin-type ones (12, 28). In adult kidney, visceral GECs express vimentin-type intermediate filaments, whereas proximal and distal tubular epithelial cells and Bowman's capsular epithelial cells have those of the cytokeratin-type (2, 47). However, all of them are derived from the polarized epithelial cell mass, which is converted from the undifferentiated metanephric mesenchyme at the first stage of glomerular development (9). It is suggested that the met protooncogene, which is a receptor of hepatocyte growth factor/scatter factor (HGF/SF), may play a role in mesenchymal-to-epithelial cell conversion during embryonic kidney development (49). During differentiation of the kidney, vimentin is first detected in immature podocytes at the capillary loop stage (15, 34), when the junctional complex containing tight junctions between GECs migrates from the apex to the base of the cells (24, 39, 41). These findings suggest that the expression of vimentin is coupled with the final differentiation of podocytes, although the regulatory mechanism of its gene expression remains unknown. It is likely that vimentin-type intermediate filaments expressed at the final stage in the development of podocytes also contribute to the unique morphology of those cells.

Recently, we have produced a number of monoclonal antibodies against rat GECs. Clone P-31 recognizes a protein which colocalizes with the intermediate filaments in podocytes. No information is yet available concerning intermediate filament-associated proteins (IFAP) in podocytes.

This is the first report of a 250-kDa protein associated with intermediate filaments in the rat GEC. This molecule, expressed exclusively in podocytes, might provide a clue to understanding the function of intermediate filaments and the regulation of interfilament networks in podocytes.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. Polyvinyl alcohol (mol wt 10,000), polyvinylpyrrolidone (mol wt 10,000), antipain, pepstatin A, leupeptin, benzamidine, diisopropylfluorophosphate (DFP), phenylmethylsulfonyl fluoride (PMSF), BSA (fraction V), and monoclonal anti-plectin antibody (clone 7A8) were obtained from Sigma (St. Louis, MO). Sheep anti-mouse Ig conjugated to FITC, goat anti-mouse IgG coupled to 10-nm colloidal gold, and goat anti-rabbit IgG coupled to 5-nm colloidal gold were obtained from Amersham (Arlington Heights, IL). Mouse monoclonal anti-vimentin antibody (clone V9) and mouse monoclonal anti-desmin antibody were from Boehringer-Mannheim Biochemicals (Indianapolis, IN), rabbit anti-vimentin antibody was from Calbiochem (San Diego, CA), and horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG was from Bio-Rad (Hercules, CA). Mouse monoclonal anti-podocalyxin antibody (5A) was a gift from Dr. Marilyn G. Farquhar (University of California, San Diego, CA).

Animals. Male Sprague-Dawley rats (4 to 6 wk old) were obtained from Japan CLEA (Tokyo, Japan). Female BALB/c mice were from Charles River Japan (Kanagawa, Japan). They were kept in an air-conditioned room and maintained on a commercial stock diet and tap water ad libitum.

Preparation of glomeruli. Rat glomeruli were isolated by graded sieving at 4°C in the presence of protease inhibitors (1 mM each antipain, benzamidine, DFP, leupeptin, pepstatin A, and PMSF dissolved in PBS, pH 7.4) Glomerular fractions were collected by centrifugation (800 g for 5 min), washed with phosphate-buffered saline containing protease inhibitors (PBS-PI), then either used for immunoblotting analysis or resuspended for 10 min in 4% paraformaldehyde fixative buffered with 0.1 M sodium phosphate buffer (PB), pelleted in a microcentrifuge, and prepared for cryosectioning. In some cases, isolated glomeruli were extracted with PBS-PI containing 1% Triton X-100 (TX-100) for 10 min at room temperature. The materials were then centrifuged at 800 g for 10 min at 4°C. The supernatant (TX soluble) and pellet (TX insoluble) were processed for biochemical and immunocytochemical analysis.

Preparation of monoclonal antibodies. Production of monoclonal antibody was carried out essentially as described by Kohler and Milstein (22) using myeloma cells. Briefly, female BALB/c mice were immunized with isolated glomeruli obtained from male rats (8-10 wk old). The immunization was performed with intraperitoneal injections of the mixture of isolated glomeruli (2.8 mg protein/mouse) with complete Freund's adjuvant (Difco Laboratories, Detroit, MI). Mice were then given two booster injections of the mixture of isolated glomeruli and incomplete Freund's adjuvant at an interval of 2 wk. Mice were tested for serum antibodies to rat glomeruli by ELISA. After a further 2 wk, the same volume of glomeruli without adjuvant was injected intraperitoneally, and the spleen was removed 4 days later. Spleen cells of immunized mouse were fused with mouse myeloma cells (P3X63Ag8.653) at a 3:1 ratio of splenocytes to myeloma cells. Antibody screening was performed on rat kidney frozen sections by indirect immunofluorescence technique. The monoclonal antibody (clone P-31) was classified as an IgG1 using a mouse monoclonal isotyping kit (Amersham). IgG fraction was purified from ascites fluid by protein G-Sepharose (Pharmacia, Uppsala, Sweden) affinity chromatography according to the manufacturer's instructions.

Immunoblot analysis. Isolated glomeruli from normal or puromycin aminonucleoside (PAN)-treated rat kidneys were solubilized in PBS containing protease inhibitors, 1% SDS, and 5 mM EDTA, electrophoresed on 7.5% polyacrylamide gels, and transferred to nitrocellulose membranes. Blots were incubated with primary antibody and then with HRP-conjugated goat anti-mouse IgG and detected using the ECL Western blotting detection system (Amersham). In some cases, blots were incubated with primary antibody and then with alkaline phosphatase-conjugated anti-mouse IgG (Promega, Madison, WI). Bound antibodies were detected by color reaction (Bio-Rad). Protein content was determined by the bicinchoninic acid assay (Pierce, Rockford, IL).

Immunoprecipitation. Isolated glomeruli were solubilized with 1% SDS in PBS containing protease inhibitors and 5 mM EDTA. The lysates were diluted 10 times in RIPA buffer (10 mM Tris, 150 mM NaCl, 1% TX-100, 0.5% deoxycholate, and 0.1% SDS, pH 7.2) and incubated with P-31 antibody (18 µg/ml) or with normal mouse IgG for 3 h at 4°C. The mixtures were then incubated with preswollen protein A-Sepharose beads overnight at 4°C. The beads were washed with Tris-buffered saline containing 5% -mercaptoethanol for 5 min. The released proteins were separated by SDS-PAGE, transferred to nitrocellulose, and processed for immunoblot analysis with primary antibodies (P-31 or anti-plectin IgG) then with HRP-conjugated goat anti-mouse IgG and detected using the ECL Western blotting detection system.

Immunocytochemistry. Rat kidneys were perfused with 4% paraformaldehyde fixative buffered with 0.1 M PB (pH 7.4) and immersed in the same fixative. The samples were rinsed and infiltrated with 40% polyvinyl alcohol/2.3 M sucrose buffered with 0.1 M PB and embedded on nails. Semithin sections (thickness, 1 µm) were cut at 70°C on a MT-7000 Ultramicrotome (RMC, Tucson, AZ) equipped with the cryoattachment CR-21. The sections were mounted on gelatin-coated glass slides and processed for staining of P-31 or vimentin. The cryosections were incubated for 2 h at room temperature with P-31 (1:1,000 dilution of ascites fluid), mouse monoclonal anti-plectin antibody (1:300 dilution), and mouse monoclonal anti-vimentin antibody (1:50 dilution), respectively. Next, the sections were incubated with FITC-labeled sheep anti-mouse Ig (diluted 1:100) for 1 h at room temperature. In control experiments, incubation with the primary antibody was omitted. Some aldehyde-fixed tissue samples were rinsed successively in PBS solution containing 10%, 15%, and 20% sucrose (4 h each). Cryosections (thickness, 2-4 µm) were cut using a Tissue Tek-II Microtome/Cryostat (Miles Scientific, Naperville, IL), then mounted on gelatin-coated glass slides and rinsed three times in PBS (10 min each). The sections were processed for immunofluorescence as described above. All sections were examined using an FX-S RFL fluorescence microscope (Nikon, Tokyo, Japan).

Ultrathin cryosections were cut with a MT-7000 Ultramicrotome equipped with the CR-21 cryoattachment at 110°C following the techniques of Tokuyasu (48). Sections were transferred to nickel grids (150 mesh), which had been coated with Formvar and carbon. 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 P-31 (1:500 dilution of ascites fluid with PBS containing 10% FCS). They were then incubated with anti-mouse IgG coupled to 5- or 10-nm gold (diluted 1:100 with PBS containing 10% FCS) for 1 h. After immunostaining, they were fixed with 2.5% glutaraldehyde containing 1% tannic acid buffered with 0.1 M phosphate buffer (PB, pH 7.4). The sections then were postfixed with 1% OsO₄ buffered with PB for 30 min, contrasted with 2% uranyl acetate solution for 20 min, and absorption stained with 3% polyvinyl alcohol containing 0.2% uranyl acetate for 10 min.

Immunogold staining of TX-insoluble fractions of isolated rat glomeruli for P-31. Isolated glomeruli were obtained as described above and treated with 0.2% TX-100 for 10 min at room temperature, then washed with PBS-PI buffer three times and fixed with 0.5% paraformaldehyde buffered with 0.1 M PB (pH 7.4). Next, they were incubated with P-31 antibody (1:200 dilution) overnight at 4°C, washed five times and then incubated with 10-nm gold-conjugated goat anti-mouse IgG (1:10 dilution in 1% BSA/PBS) for 3 h at room temperature. After washing five times, they were fixed with 2.5% glutaraldehyde buffered with 0.1 M PB for 1 h at room temperature. The fixed materials were prepared for routine EM or freeze-etching replica. Double immunogold staining with P-31 and vimentin antibody were also done. In brief, ultrathin cryosections of aldehyde-fixed isolated glomeruli were cut as described above. The sections were incubated with P-31 and rabbit anti-vimentin antibody and then incubated with 5-nm gold-conjugated goat anti-rabbit IgG and 10-nm gold-conjugated goat anti-mouse IgG. Finally, the sections were prepared as described above.

Deep-etched freeze-fracture. Isolated glomeruli treated with TX-100 were labeled with P-31 then with gold-conjugated anti-mouse IgG as described above. They were rinsed with 60% ethanol and quickly frozen in liquid nitrogen. They were transferred to a JFD-7000 freeze-etching equipment (JEOL), fractured with a blade at 130°C, etched deeply at 95°C, 3 × 10⁶ Torr for 15 min, and rotary shadowed with platinum and carbon. The tissues were dissolved in sodium hydrochloride. The replica membranes were put on copper grids (300 mesh) and observed with a JEOL 100-CX electron microscope.

Culture of GECs. Isolated glomeruli from male rats (4 wk old) were transferred to a medium consisting of a mixture of equal parts of Dulbecco's modified Eagle's medium and Ham's F-12 medium supplemented with 1% FCS and were explanted in tissue culture flasks coated with collagen type-I (Collaborative Biomedical Products, Bedford, MA). Cultured cells were fixed with 4% paraformaldehyde for 30 min at room temperature, permeabilized with 0.1% TX-100 for 10 min at room temperature, and subjected to the immunofluorescence technique.

Induction of PAN nephrosis. Male rats (6 wk old) were injected daily with PAN (1.6 mg · rat¹ · day¹) for 2, 6, or 10 days and were killed 24 h after the last injection as described (25).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

P-31 recognizes 250-kDa polypeptide in the TX-100-insoluble fraction of rat glomerular lysate. Screening of monoclonal antibodies against rat glomeruli led to several antibodies that specifically recognize the cells of the rat glomerulus. One of those antibodies, called P-31, recognized the approximate molecular mass of 250-kDa polypeptide in the isolated glomerular lysate but not in the kidney cortex and medulla lysate under both reduced and nonreduced conditions, suggesting that the antigen recognized by this antibody is apparently concentrated in the glomerular fraction (Fig. 1A). To determine whether the antigen is cytoskeletal associated, the immunoblot analysis with P-31 was done by using TX-100-soluble or -insoluble fraction from isolated glomeruli. As shown in Fig. 1B, the single 250-kDa band was only detected for the TX-100-insoluble fraction of rat glomeruli with the antibody.

View larger version (67K): [in this window] [in a new window]   Fig. 1.   Biochemical characterization of P-31 antigen. A: similar amounts of protein from medulla (M), cortex (C), and isolated glomeruli (G) were separated by SDS-PAGE and probed with P-31 monoclonal antibody. A single band was only observed in the glomerular fraction. B: isolated glomeruli were extracted with 1% Triton X-100 (TX-100) in PBS containing protease inhibitors. Materials were then centrifuged at 10,000 rpm for 10 min at 4°C. Supernatant (S) and pellet (P) were lysed with SDS-sample buffer and separated by SDS-PAGE, transferred to nitrocellulose and probed with monoclonal antibody P-31. The band was observed in the pellet fraction. Positions of the molecular mass markers are indicated in kDa.

P-31 antigen is located in GECs. Semithin cryosections (1-2 µm) were prepared from unfixed or aldehyde-fixed rat kidneys. The immunoreactivity for P-31 antibody was specifically located at the glomeruli in the rat kidney cortex. The ductular epithelial and endothelial cells were totally negative (Fig. 2A). High-magnification view of the rat glomerulus showed that the signal of P-31 was only detected in the cytoplasm of GECs. The nuclei of GECs were negative. Dotlike signals facing the urinary space were found along the glomerular basement membranes; they are a part of the major processes of GECs. Other types of cells including Bowman's epithelial cells, capillary endothelial cells, and mesangial cells were negative in the normal adult rat (Fig. 2, B and C).

View larger version (114K): [in this window] [in a new window]   Fig. 2.   Immunofluorescence localization of P-31 antigen in semithin cryosections of adult rat kidney cortex. Semithin cryosections (1-2 µm) were prepared from aldehyde-fixed rat kidneys. Kidneys were perfused with 4% paraformaldehyde fixative and fixed for an additional hour by immersion in the same fixative at room temperature. Sections were incubated with P-31 and followed by incubation with FITC-conjugated sheep anti-mouse Ig. Immunoreactivity for P-31 antibody was exclusively detected in the glomerular epithelial cells (GECs). G, glomeruli. Bars: A, 100 µm; B, 20 µm.

P-31 antigen colocalizes with intermediate filaments in GECs. Immunofluorescence with P-31 showed that the antigen was located in the cytoplasm of the podocytes. To obtain more precise information on the localization of the antigen, we localized the antibody in cryoultrathin sections of the rat kidney cortex. In Fig. 3A, gold particles associated with p250 were found in the cytoplasm of podocyte cell body. The foot processes of GECs were not labeled with this antibody. High-magnification micrograph of the podocyte showed that the gold particles were not associated with membranous organelles (Fig. 3B). Immunoblot analysis of detergent-treated glomerular fractions suggested that the antigen was associated with the cytoskeleton. Therefore, we tried to determine the subcellular localization of the antigen in TX-treated glomeruli. Isolated glomeruli were treated with TX-100, fixed, labeled with P-31, then incubated with gold-conjugated anti-mouse IgG. The labeled glomeruli were then processed for routine EM. TX-100 treatment of the isolated glomeruli made the cells lose the cytoplasmic proteins. As a result, the cytoskeletal elements could be easily made visible. Gold particles in the sections were only associated with intermediate-sized filaments of GEC, not with actin filament bundles located in the foot processes. The particles were especially associated with the interconnections of intermediate filaments (Fig. 4A). Podocytes are known to express vimentin-type intermediate filaments in the cytoplasm. Therefore, we tried to examine the colocalization of P-31 antigen and vimentin by immunogold staining of cryoultrathin sections. The problem with such sections is the difficulty of observing the cytoskeletal elements. When tannic acid was used to contrast the ultrathin cryosections, the cytoskeletons could be clearly detected. Ultrathin cryosections of TX-100-treated isolated glomeruli were doubly stained with P-31 and polyclonal anti-vimentin antibody, followed by incubation with the secondary antibodies conjugated with gold particles of different sizes. Double immunogold labeling of the cryoultrathin sections confirmed that P-31 antigen labeled with 10-nm-gold particles was located on the vimentin filaments labeled with 5-nm-gold particles (Fig. 4, B and C).

View larger version (117K): [in this window] [in a new window]   Fig. 3.   Immunogold staining for P-31 antigen in normal rat glomerulus. A: gold particles (10 nm) were found in the cytoplasm of podocyte cell body, but foot processes (fp) were not labeled. Endothelial cells (E) were not labeled with P-31 monoclonal antibody. B: high magnification of the cytoplasm of podocyte cell body labeled with P-31. Gold particles (5 nm) were sparsely distributed in the cytoplasm. Membranous organelle such as Golgi apparatus (G), mitochondria (M) and nucleus (N) were not labeled with P-31. Bars: A, 0.2 µm; B, 0.1 µm.

View larger version (69K): [in this window] [in a new window]   Fig. 4.   Immunogold staining for P-31 antigen in TX-100-treated glomeruli. A: isolated glomeruli were treated with TX-100, fixed, labeled with P-31 followed by incubation with gold-conjugated anti-mouse IgG. Labeled glomeruli were then processed for routine electron microscopy. Note that gold particles are associated with intermediate filaments, not with actin filament bundles (mf ) and microtubules. B and C: sections of isolated glomeruli were doubly stained with P-31 (10 nm gold) and rabbit anti-vimentin antibody (5 nm gold). In the podocytes, P-31 is colocalized with vimentin filaments. Bars = 0.1 µm.

Freeze-etching replica of P-31-labeled glomeruli. Immunogold staining of P-31 in renal cryosections showed that P-31 antigen is associated with vimentin-type intermediate filaments. Freeze-etching replica technique can present a three-dimensional image of cytoskeletal elements. Therefore, we used it to determine the precise binding site of P-31 antibody on the intermediate filaments. Isolated glomeruli stained with P-31 monoclonal antibody then with gold-conjugated secondary antibody were quickly frozen and processed to obtain the deep-etching replica. The gold particles visible on the replica were located only at the crossing points of intermediate filaments, with none on the thin filaments between intermediate filaments (Fig. 5).

View larger version (137K): [in this window] [in a new window]   Fig. 5.   High-power image of P-31-gold labeling intermediate filaments. Isolated glomeruli were extracted with TX-100, then labeled with P-31 followed by gold-conjugated anti-mouse IgG, quick-frozen, and then deep-etched. Gold particles (arrowheads) are located at the intersection of intermediate filaments. Bar = 0.1 µm.

P-31 antigen is a vimentin filament-associated protein, which is exclusively expressed in podocytes. Immunocytochemical data clearly demonstrated that P-31 antigen is a protein associated with vimentin-type intermediate filaments. As the next step, we investigated whether P-31 antigen is located in the various cells where vimentin is expressed. Immunoblot analysis of P-31 and vimentin in several tissues showed that P-31 antigen was exclusively expressed in rat glomeruli, whereas various amounts of vimentin were expressed in tissues observed in this experiment (Fig. 6).

View larger version (55K): [in this window] [in a new window]   Fig. 6.   Tissue specificity of P-31 antigen expression vs. vimentin determined by immunoblot analysis. Equal amounts of proteins (50 µg/lane) from several tissues were loaded in each lane. P-31 antigen (p250) was only detected in the glomerular fraction, whereas vimentin expression varied in the tissues.

Plectin is an abundant, high-molecular-weight cytomatrix protein (300 kDa) found in a wide variety of tissues and cells. Based on biochemical and immunocytochemical studies, plectin has been suggested to play a role in the cross-linking of intermediate filaments, the interlinking of intermediate filaments with microtubules and microfilaments, and the anchoring of intermediate filaments to the plasma membrane and the nuclear membrane. We compared plectin with P-31 antigen in the glomerular fraction with immunoblot analysis, because of the similarity of molecular size and distribution between P-31 antigen and plectin. However, the mobility of plectin (approximate molecular mass of 300 kDa) was apparently different from that of P-31 antigen (Fig. 7A). Also, the immunoprecipitated protein with P-31 antibody was not detected with anti-plectin antibody (Fig. 7B).

View larger version (20K): [in this window] [in a new window]   Fig. 7.   A: immunoblot analysis of glomerular lysate using P-31 and anti-plectin antibody. Glomeruli were isolated, and equal amounts of material were loaded in each lane. Both plectin (300 kDa) and P-31 antigen (250 kDa) were detected in the glomerular lysate by immunoblotting, but their mobilities differed each other. B: p250 precipitated from glomeruli is not detected by anti-plectin antibody. p250 was immunoprecipitated from a glomerular lysate with P-31 monoclonal antibody. Immunoprecipitated proteins were electrophoresed, transferred to nitrocellulose, and immunoblotted with P-31 antibody (lane 1) or anti-plectin antibody (lane 2). Anti-plectin antibody does not react with p250 protein, which is immunoprecipitated with P-31. Arrow indicates the immunoprecipitated p250.

P-31 antigen is expressed in immature podocytes in developing kidney. The kidneys in newborn rats are not fully developed and still contain immature glomeruli and tubular epithelial cells. Therefore, they are very convenient for checking the expression of P-31 antigen during renal development. The immunofluorescence staining for P-31 antigen demonstrated the labeling of presumptive GECs at the S-shaped body stage and the later capillary loop stage. The signals for P-31 antigen were mainly observed along the basement membrane of these immature polarized cells at the S-shaped body stage (Fig. 8, A and B). The localization of P-31 to the basal portion of developing GECs at the S-shaped body stage suggests that this molecule may play an important role in the cell-matrix contact. On the other hand, P-31 antigen was detected in the whole cytoplasm of the immature GECs, where the foot processes were not developed (Fig. 8, C and D). Immunogold staining of the immature podocytes with P-31 showed that the antigen was distributed in the cytoplasm, but the basal periphery along the basement membrane, where microfilament bundles were visible, was not labeled with P-31 (Fig. 9). Other types of cells were negative against P-31 antibody through all of the stages.

View larger version (136K): [in this window] [in a new window]   Fig. 8.   Immunofluorescence labeling for P-31 antigen in developing rat kidney. Cryosemithin sections were cut from newborn (2 day old) rat kidney and stained with monoclonal P-31 for 2 h at room temperature, followed by FITC-conjugated sheep anti-mouse Ig for 1 h at room temperature. Phase-contrast (A and C) and fluorescence (B and D) micrographs of each section are shown. Immunofluorescence staining for P-31 antigen can be seen in the presumed GECs (G) at the S-shaped body stage (A and B) and capillary loop stage (C and D). P, proximal tubule. Bars = 100 µm.

View larger version (150K): [in this window] [in a new window]   Fig. 9.   Immunogold staining for P-31 antigen in developing GECs. Ultrathin cryosections of the developing kidney cortex were located with P-31 antigen, which was diffusely distributed in the cytoplasm of the podocyte at the capillary loop stage. Gold particles with P-31 were not detected at the periphery of the basal membrane, which were abundant in the microfilament bundles. B, basement membrane. Bar = 0.2 µm.

P-31 antigen is present in cultured cells. When isolated glomeruli were explanted on collagen type I and cultivated in 1% FCS containing medium, GECs selectively spread out from the isolated glomeruli and proliferated. Under these conditions, the cells growing out from the glomerulus reacted with P-31 monoclonal antibody with a filamentous pattern. Nuclei were totally negative against this antibody (Fig. 10). The shape of P-31-positive cells varied, but most of the cells were large and polygonal in the monolayers. P-31-positive cells were also labeled with anti-vimentin antibody. Western blot analysis of these cells with P-31 had the same molecular mass band, the 250-kDa band, as in vivo. On the other hand, cultured mesangial cells identified by Thy-1.1 antigen or endothelial cells identified by factor VIII antigen were not stained with P-31 by immunofluorescence (data not shown).

View larger version (87K): [in this window] [in a new window]   Fig. 10.   Immunofluorescence microscopy of the staining pattern of P-31 in cultured GECs. Isolated glomeruli from male rats (4 wk old) were explanted in tissue culture flasks coated with collagen type I. Cultured cells were fixed with 4% paraformaldehyde and treated for immunofluorescence with P-31 antibody. Cells growing out from the glomerulus (G) reacted with this antibody in a filamentous pattern. Glomerulus was also positive. Bar = 100 µm.

Expression of P-31 antigen increases in injured podocytes. PAN specifically injures podocytes and induces morphological changes including foot process fusion and effacement of the cytoplasm of podocytes followed by massive proteinuria. Such changes are accompanied by changes of the cytoskeletal elements. In particular, desmin-type intermediate filaments, which are not expressed in normal podocytes but are in mesangial cells, are induced in injured podocytes after PAN injection. Therefore, we examined whether the expression of P-31 antigen is altered in PAN-induced nephrotic kidney. The expression of podocalyxin, a major sialoglycoprotein of podocytes, remained unchanged after PAN treatment as described before (21). The expression of vimentin did not change throughout the experimental period, but that of desmin dramatically increased after the initial injection of PAN. Increased P-31 antigen was detected at 7 days when the proteinuria was obvious and desmin expression was enhanced (Fig. 11).

View larger version (46K): [in this window] [in a new window]   Fig. 11.   Immunoblot analysis of glomerular lysates from puromycin aminonucleoside (PAN)-treated rats using P-31, anti-desmin antibody, anti-vimentin antibody and anti-podocalyxin antibody. PAN nephrosis was induced by daily injections of PAN (1.6 mg/rat) for 2, 6, or 10 days, and the animals were killed 24 h after the last injection. Isolated glomeruli from normal and PAN-treated rats were solubilized in SDS sample buffer, and the same amount of protein from each was loaded onto a gel, electrophoresed, and transferred to nitrocellulose. Membrane was incubated with P-31, anti-desmin antibody, anti-vimentin antibody, or anti-podocalyxin monoclonal antibody (5A) followed by a horseradish peroxidase-conjugated goat anti-mouse IgG and ECL detection system. P-31 antigen increased at 7 days after the first PAN injection. Desmin dramatically increased after 11 days. On the other hand, the amounts of vimentin and podocalyxin did not change throughout the experimental period.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Monoclonal antibody technique is a powerful tool for targeting a new molecule, whose localization can be predicted. Recently, it has been successfully used to obtain monoclonal antibodies against membrane proteins of the rat glomerulus (32). To examine the cellular structure and function of GECs, we designed monoclonal antibodies against GECs which have a unique morphology. In the present study, we show that a novel monoclonal antibody, P-31, recognizes a protein which is exclusively expressed in rat visceral GECs. Both biochemical and immunocytochemical data suggest that the 250-kDa molecule (p250) recognized by P-31 is associated with vimentin-type intermediate filaments, which are located in the cell body and the major processes. The absence of this molecule from the foot processes indicates that it is not associated with actin filaments, which are abundant there. These findings suggest that p250 is specifically colocalized with intermediate filaments.

A number of proteins associated with microtubules and microfilaments have been known so far (14, 36). Specific binding proteins seem to be largely responsible for the functional diversity of these types of filaments. Recently, proteins associated with intermediate filaments have also been identified in tissues and cultured cells (11). Proteins specifically interacting with intermediate filaments would seem to have a great potential as regulatory elements which control the organization of intermediate filaments, although their function has been poorly understood. Some proteins are widely distributed in tissues and cultured cells. On the other hand, a limited distribution of IFAP has also been reported (4, 5, 19, 27, 30, 38, 51). In the IFAP family, plectin (300 kDa) seems to be the most characterized member thus far (10). It has been identified over a wide range of tissues and cell types using immunofluorescence and immunoelectron microscopy as well as immunoblot analyses (53, 55). In the glomerulus, plectin is located in tubular epithelial cells, Bowman's capsular epithelial cells, podocytes, and mesangial cells (53, 57). Our immunoblot data show that both p250 and plectin are detected in the glomerular lysate but differ from each other for the following reasons. First, immunoblot analysis shows that they have different molecular weights. Second, their distribution patterns are totally different. The localization of p250 is greatly limited in podocytes compared with the widespread distribution of plectin. In addition, plectin is located in the perinuclear region of GECs (57), whereas p250 is diffusely distributed in the cytoplasm of GECs. Third, the immunoprecipitation experiments with P-31 demonstrate that p250 is not recognized by the plectin antibody. These data strongly suggest that p250 is a molecule distinct from plectin, although further examination of the molecular nature of p250, including cDNA cloning, will be necessary.

Expression of P-31 antigen in developing kidney. Our data on the localization of p250 during glomerular development suggest that its cytoplasmic localization varies at different stages. During development, this molecule is first identified at the S-shaped body stage and is located at the base of presumptive podocytes. Later, this molecule is widely distributed in the whole cytoplasm of podocytes. Schnabel et al. (41) have shown that the glomerular anionic coat protein, podocalyxin, can be detected in the S-shaped body only on the plasmalemma of cells of the glomerular anlage. At this stage, the podocytes are distinguishable from other types of cells such as Bowman's capsular epithelial cells and proximal tubular epithelial cells. Therefore, the expression of p250 in podocytes at the S-shaped body stage strongly indicates that there is a close correlation between the expression of p250 and the differentiation of podocytes. On the other hand, vimentin is first detected in immature GEC at a later stage (e.g., capillary loop stage), when the formation of foot processes has occurred and filtration has started (15). The change of p250 distribution from the basal cell surface to the whole cytoplasm may depend on vimentin expression. Some IFAPs are involved in intermediate filament-plasma membrane interactions (18, 43, 45, 54). Therefore, p250 might interact with unknown proteins and form specialized structures (e.g., hemidesmosomal structure or adherens-type junctions) at the S-shaped body stage and then distribute in the cytoplasm to produce an intermediate filament network when vimentin filaments are expressed at the capillary loop stage, although we have not yet clarified the specialized structure associated with p250 at the basal membrane of the presumptive GECs. P-31 antigen is located in the cytoplasm over the microfilament bundles along the basal membrane in immature podocytes, which have their flattened cell surface attached to the basement membrane. Therefore, it should be emphasized that p250 is not directly attached to the basal membrane. Study of the expression of p250 might offer a clue to the vimentin expression in podocytes, because it occurs earlier than that of vimentin.

Possible functions of P-31 antigen. We did not find localization of p250 in cell types other than podocytes by immunoblot analysis or immunofluorescence microscopy, which suggests that this molecule has a highly specialized function in podocytes. These cells have vimentin-type intermediate filaments, which are usually expressed in mesenchymal cells, whereas keratin-type intermediate filaments are present in Bowman's capsular epithelial cells and tubular epithelial cells. The difference of the intermediate filament type expressed is very interesting, because all of these epithelial cells originate from the same precursor cells, which are generated by conversion of mesodermal mesenchymal cells to polarized epithelial cells at the early developmental stage of the kidney (40). Why do podocytes not express cytokeratin, but do express vimentin-type filaments? Visceral GECs have features that are distinct from those of typical epithelial cells. Podocytes are characterized by a large cell body and numerous foot processes attached to each other by a tenuous slit diaphragm (42). As podocytes are constantly exposed to mechanical force, i.e., high capillary pressure, they need steady and tidy skeletal elements to maintain their unique structure in such a severe environment. Vimentin synthesis responds to changes in cell shape, while cytokeratin synthesis is involved in cell-cell contact in cultured cells (3). In addition, vimentin-type intermediate filaments display unique viscoelastic properties allowing them to resist breakage and become even stronger under mechanical stress conditions that would rupture other cytoskeletal networks (17). However, the lack of vimentin in mice leads to no obvious phenotypic changes (7). On the other hand, loss of the vimentin-associated protein causes epidermolysis bullossa with muscular dystrophy, suggesting that the IFAPs are more important in the regulation of cellular function (13, 46). As many of the IFAPs are cell-type specific, it is likely that IFAP regulates the special function of intermediate filaments. Our data demonstrate that podocytes express the specific IFAP binding to vimentin filaments, which are abundant in numerous tissues. The evidence indicates that p250 has an important role in podocytes. Our immunocytochemical observations and freeze-etching replica study suggest that p250 localization is restricted to the interconnection between intermediate filaments. Therefore, it could be argued that p250 plays a role in the construction of the three-dimensional network of the intermediate filaments providing a mechanical framework for visceral GECs.

Selective depolymerization of microtubules causes collapse of type III intermediate filaments, such as vimentin and desmin, to a perinuclear cap (12). It is likely that the cytoarchitecture of intermediate filaments is influenced by interactions with other IFAPs and other cytoskeletal elements, particularly microtubules. Thus far there is no evidence that intermediate filaments connect with other filament systems or membrane organelles through p250 in the podocytes. What should be examined next is the role of P-31 antigen in the interaction between intermediate filaments and other filament systems, especially microtubules.

The localization of p250 at the border between cell body and foot processes suggests that p250 may play a role in the interaction between the intermediate filament network in the cell body and the microfilament bundles concentrated in the foot processes to maintain the unique structure of the podocytes. A similar interaction between p250 and actin bundles is observed at the basal cortical region of the immature GEC, which has a flattened basal cell surface. The formation of foot processes containing numerous actin filaments is observed after p250 expression in the presumptive GECs. These suggest that p250 might also play a role in the formation of foot processes.

PAN is a specific toxin for GECs and induces nephrosis in rats. In the PAN nephrotic condition, GECs lose their foot processes and have flattened basal cell surfaces (6, 31). Under these conditions, the expression of p250 dramatically increases on days 7 and 14 after starting PAN injection when proteinuria is obvious. Desmin also increases at the same time, whereas the contents of vimentin-type intermediate filaments are not changed after induction of nephrosis. Previous data suggest that the increase of desmin in GECs is associated with glomerular epithelial damage (56). When the muscle-specific intermediate filament desmin from hamster was expressed in nonmuscle cells under the control of a vimentin promoter in mice, the desmin protein colocalized with mouse vimentin filaments (37). Although the colocalization of desmin with vimentin in the same filaments has not been examined for podocytes, increased p250 may be shared by both desmin and vimentin filaments under pathological conditions.

P-31 antigen, a good differentiation marker of podocytes. GECs in culture have become an important in vitro model for investigating the cellular physiology and biochemistry of the glomerulus. However, there is an unsolved problem regarding the origin of epithelial cells expanding from the glomerulus. In cultured cells, it is very difficult to determine whether the origin of GECs is visceral or parietal (35, 52). The major reason is the lack of a cell type-specific marker for each type of cell. The expression of p250 is maintained in the primary cultured cells from isolated glomeruli. The p250-containing cells also express vimentin, but neither the mesangial cell marker, Thy-1.1, nor the endothelial cell marker, factor VIII, were detected. Holthofer et al. (16) suggested that GECs rapidly lose the ability to express the cell surface marker proteins, especially podocalyxin. Our preliminary data show that the expression of p250 is stable compared with that of podocalyxin but is lost at the 5th generation.

The findings reported here confirm the existence of podocyte-specific IFAP recognized by P-31 monoclonal antibody. Its limited distribution in vivo and in vitro demonstrates that the P-31 antigen is specifically expressed in the GEC. This molecule appears to be a candidate not only for a specific marker of GECs but also a potential regulator of intermediate filaments for the maintenance of the specific morphology of this type of cell.