SSeCKS-controlled mesangial cell cytoskeleton - Journal of Cell ...

361

Journal of Cell Science 112, 361-370 (1999) Printed in Great Britain © The Company of Biologists Limited 1999 JCS7302

Involvement of the protein kinase C substrate, SSeCKS, in the actin-based stellate morphology of mesangial cells Peter J. Nelson2, Konstadinos Moissoglu1, Jesus Vargas, Jr1, Paul E. Klotman2 and Irwin H. Gelman1,* 1Department

of Microbiology and 2Division of Nephrology, Mount Sinai School of Medicine, 1 Gustave L. Levy Place, New York, NY 10029-6574, USA *Author for correspondence (e-mail: [email protected])

Accepted 13 November 1998; published on WWW 13 January 1999

SUMMARY Activation of protein kinase C is a key signal transduction event in mesangial cell dedifferentiation and proliferation, yet little is known about downstream substrates or their roles in normal or diseased glomeruli. SSeCKS, a novel protein kinase C substrate originally isolated as a srcsuppressed negative mitogenic regulator in fibroblasts, controls actin-based cytoskeletal architecture and scaffolds key signaling kinases such as protein kinase C and protein kinase A. Based on the morphologic similarity between SSeCKS-overexpressing fibroblasts and stellate mesangial cells, we hypothesized that SSeCKS might play a role in mesangial cell morphology in a protein kinase C-dependent manner. Immunoblotting, in situ staining and northern blotting detected abundant expression of SSeCKS in human and rodent mesangial cells and glomerular parietal cells but not in renal tubular epithelia. Immunofluorescence analysis showed enrichment of

SSeCKS in mesangial cell podosomes and along a cytoskeletal network distinct from F-actin. Activation of protein kinase C by phorbol ester resulted in a rapid serine phosphorylation of SSeCKS and its subsequent translocation to perinuclear sites, coincident with the retraction of stellate processes. These effects were blocked by concentrations of bis-indolylmaleimide that selectively inhibit protein kinase C. Finally, ablation of SSeCKS expression using retroviral anti-sense vectors induced (1) an elongated, fibroblastic cell morphology, (2) production of thick, longitudinal stress fibers and (3) repositioning of vinculin-associated focal complexes away from the cell edges. These data suggest a role for SSeCKS as a downstream mediator of protein kinase C-controlled, actin-based mesangial cell cytoskeletal architecture.

INTRODUCTION

intracellular signaling pathways controlling cytoskeletal reorganization and subsequent MC dedifferentiation remains incompletely understood. Activation of protein kinase C (PKC) is a common pathway through which disparate mitogenic and pro-inflammatory factors signal, including those causing phenotypic plasticity. Constitutively expressed (α, βI, ε, δ, ζ, γ) and inducible (βII) PKC isotypes have been identified in MC (Pfeilschifter, 1994; Ganz et al., 1996). PKC-dependent morphological alterations of MC occur after signaling by vaso-active peptides, growth factors, cytokines, amines, prostanoids and bio-active lipids (Ganz et al., 1996; Mene and Cinotti, 1992), and by diacylglycerol from hyperglycemic metabolism (Derubertis and Craven, 1994; Heilig et al., 1997). In many cell types, activated PKC directly promotes cytoskeletal reorganization through phosphorylation of proteins at plasma membranes sites, such as the focal adhesion proteins vinculin and tensin (reviewed in Jaken, 1996; Rozengurt, 1995). Other reports suggest that activated PKC may control cytoskeletal architecture by dominating over other signaling pathways, such as RhoA-mediated stress fiber formation (Rozengurt, 1995). However, little information exists on downstream targets of

Mesangial cells (MC) are unique contractile pericytes of the renal glomerulus with fibroblast and smooth muscle-like characteristics (Mene et al., 1989) that undergo phenotypic plasticity in response to appropriate stimuli. Upregulated or de novo expression of numerous cytoskeletal proteins, such as αsmooth muscle actin (Johnson et al., 1991), myosin heavy chain isoforms (Hiroi et al., 1996), profilin (Tamura et al., 1996), and moesin and radixin (Hugo et al., 1996), have been identified as markers of MC dedifferentiation that occurs after mitogenesis. Detailed microscopic analysis of MC following activation by insulin or insulin-like growth factor shows that significant cytoskeletal reorganization precedes the expression of these markers (Hugo et al., 1996; Berfield et al., 1996, 1997). Several groups have suggested that this cytoskeletal change from a quiescent to an activated phenotype is required for adhesion, migration and proliferation of MC during repopulation or hyperplasia in the glomerulus (Berfield et al., 1997; Johnson et al., 1992). Specific mitogens have been implicated in this process (Berfield et al., 1996, 1997); however, correlation of these extracellular ligands with

Key words: SSeCKS, PKC, Mesangial cell, Cytoskeleton

362

P. J. Nelson and others

PKC in MC, and the assumption that active PKC will phosphorylate substrates in common with fibroblasts and/or smooth muscle cells may not be accurate. Some PKC substrates control cellular processes required in all cells, but many substrates determine cell- or tissue-specific characteristics congruent with their cell- or tissue-specific expression. We isolated and characterized a novel PKC substrate, named SSeCKS (Lin et al., 1996), that functions as a negative regulator of mitogenesis in fibroblasts (Lin et al., 1995; Nelson and Gelman, 1997), as a cytoskeletal differentiating factor (Lin et al., 1996; Lin and Gelman, 1997), and as a scaffolder of protein kinases C and A (Nauert et al., 1997) (Table 1). SSeCKS’ mitogenic regulatory role is based on our observation that inducible overexpression of this src- and ras-suppressed gene disengages SSeCKS from cell cycle control, causing growth arrest (without apoptosis) in fibroblasts (Lin and Gelman, 1997). This result recapitulated the finding of increased SSeCKS gene transcription and decreased SSeCKS protein phosphorylation in contact-inhibited, quiescent cells, but transcriptional downregulation and hyperphosphorylation following activation of ts-src or addition of serum growth factors (Lin et al., 1996; Lin and Gelman, 1997). SSeCKS localizes to membrane and cortical cytoskeletal sites in fibroblasts (Lin et al., 1996), and overexpression of SSeCKS causes a profound change in cellular morphology similar to that seen during differentiation of MC: stellate elaboration of filopodia, lamellipodia and podosomal structures, accompanied by cell flattening and increased cellular adhesion (Lin and Gelman, 1997). SSeCKS’ scaffolding functions also include a phosphatidylserine-dependent binding of PKC (Lin et al., 1996; Nauert et al., 1997) and an A kinase anchoring protein (AKAP) activity based on a PKA binding motif shared with a related human protein, Gravin (Nauert et al., 1997), and with AKAP79 (Klauck et al., 1996). Based on similarities in the morphology of SSeCKS overexpressor cells and MC, and the importance of PKC Table 1. Characterization of SSeCKS Protein 290 kDa (major species), 280 kDa and 240 kDa (minor species)a Similarity to human Gravin, an autoantigen in myasthenia gravis (also known as AKAP250) Four PKC phosphorylation sites (in vitro) Phospholipid and plasma membrane binding Putative zinc-finger domain Predicted filamentous, rod-like structure Four predicted nuclear localization (SV40 Tag-like) motifs Transcriptional regulation Downregulation in src- and ras-, but not in raf-transformed fibroblasts Upregulation from G0 to G1 Upregulation induced by contact inhibition Post-translational regulation PKC phosphorylation (in vivo): hyperphosphorylated during G1/S, hypophosphorylated during G2/M and G0 N-terminal myristylation Multivalent scaffolding protein PKC binding; affinity lost on PKC phosphorylation Ca2+/calmodulin binding (unpublished) Predicted PKA bindingb aIn rodent fibroblasts. bNauert et al. (1997).

signaling in MC activation and phenotypic alteration, we hypothesized that SSeCKS may be a major defining PKC substrate in MC mitogenic and cytoskeletal control. In this study, we characterize SSeCKS as an important downstream target of activated PKC in MC. Our data suggest that SSeCKS is integrally involved in MC morphological differentiation, but not in the phenotypic development of most other renal cell types. Modulation of SSeCKS function by PKC phosphorylation may be an early event in MC dedifferentiation, and may play a role in mitogenic and cytoskeletal control of MC.

MATERIALS AND METHODS Cell culture Primary human adult and fetal MC, isolated as previously described (Schnaper et al., 1996), were gifts of Jeffrey Kopp (NIH/NIDDK) and William Schnaper (Northwestern University Medical School). Briefly, human fetal and adult MC were isolated by differential sieving of minced glomerular tissue from a fetal kidney after elective abortion (14-15 weeks gestation) and from an adult kidney after nephrectomy, respectively. MC phenotype was confirmed by the following criteria: morphology, presence of abundant actin filaments and absence of staining for cytokeratin and factor VIII-related antigen. Stock cultures were maintained in Dulbecco’s modified Eagle’s medium (DMEM/F12) supplemented with 20% heat-inactivated fetal calf serum (Gibco), Insulin-Transferrin-Selenium-A (Gibco), 1 mM L-glutamine, 10 units/ml penicillin and streptomycin and 250 ng/ml amphotericin B. All studies utilized cells between passages 6 and 20. SV40immortalized mouse MC (SV40-MES 13; American Type Culture Collection, number: CRL-1927) and Rat-6 fibroblasts were maintained in DMEM supplemented with 10% heat-inactivated calf serum, 10 units/ml penicillin and streptomycin, and 250 ng/ml amphotericin B. Human organ transcriptional expression assay Poly(A)-selected mRNA organ blots (2.0 µg per lane; Clontech) comparing multiple adult human organs (spleen, thymus, prostate, testes, ovary, small intestine, colon, peripheral blood lymphocytes, heart, brain, placenta, lung, liver, skeletal muscle, kidney and pancreas) were incubated with Rapid Hybridization Buffer (Amersham International) and 32P-labeled rat SSeCKS cDNA probe, washed under semi-permissive conditions (final wash with 0.1× SSC/0.5% SDS at 56°C), and then autoradiographed. Renal cell protein expression assay Frozen pellets from confluent cultures of micro-dissected primary adult human renal cells representing proximal tubule, thick ascending loop of Henle and collecting tubule epithelia were prepared as previously described (Wilson et al., 1985). Briefly, normal human renal tubules from an adult nephrectomized kidney were microdissected at proximal, thick ascending limb, and distal tubular epithelial sites and individually plated onto collagen-coated tissue culture plates. Cells were cultured in Hepes-buffered RPMI medium supplemented with 5 µg/ml transferrin, dexamethasone (5×10−8 M), 5 µg/ml insulin, triiodothyronine (10−12 M), 1% fetal bovine serum, 200 units/ml penicillin and 100 µg/ml streptomycin. Cells were grown to confluency in 25 cm2 flasks, washed with PBS containing protease inhibitors, scraped, centrifuged and the pellet snap-frozen in liquid nitrogen. Total cellular protein from these frozen pellets, as well as from confluent cultures of SV40-immortalized mouse mesangial cells, primary adult human mesangial cells, primary human fetal mesangial cells and Rat-6 fibroblasts, was extracted using RIPA buffer as previously described (Lin et al., 1996). Protein concentrations were

SSeCKS-controlled mesangial cell cytoskeleton determined by the Bradford method (Bradford, 1976) (BioRad Protein Reagent). 10 µg of protein from SV40-immortalized mouse mesangial cells, and 20 µg of protein from the remaining samples, were analyzed on stacking gels by SDS-PAGE and immunoblotting as previously described (Lin et al., 1996) using rabbit polyclonal anti-SSeCKS sera followed by alkaline phosphatase-labeled anti-rabbit Ig (BoehringerMannheim) and Western Blue detection reagent (Promega Corp.). Glomerular immunohistochemistry To avoid loss of antigenic epitopes in the kidney as a result of standard tissue fixing procedures, the following protocol (Department of Pathology, Mount Sinai Medical Center) was done: kidneys from one killed Sprague-Dawley rat were removed by dissection, bisected longitudinally, and immediately frozen with a liquid nitrogen-cooled heat extractor after embedding the tissue in OCT (Tissue-Tek; Miles). Frozen 6 µm thick sections were immediately UV-fixed using the Cryostat Frozen Sectioning Aid (Instrumedics, Inc., New Jersey). Sections were then freeze-dried in the cryostat for 10 minutes. DAB peroxidase immunohistochemical localization of SSeCKS in rat kidney was performed with a 1:250 dilution of immuno-affinity purified rabbit polyclonal anti-SSeCKS antibody using a Super Sensitive Animal Detection Kit (BioGenex) and SigmaFast diaminobenzidine tetrahydrochloride substrate (Sigma BioSciences) without pre-blocking of endogenous peroxidases. Sections were counterstained with Hematoxylin and then mounted (Permount; Sigma Biosciences). SSeCKS immunohistological staining was also performed on sections from human kidney core biopsies and from rat kidneys processed by standard tissue fixing procedures (2% formaldehyde) with endogenous peroxidase blocking. SSeCKS phosphorylation Human mesangial cells were plated onto 10 cm plates at 70% confluency and then treated with phorbol 12-myristate 13-acetate (PMA, 200 nM) or dimethylsulfoxide carrier (mock treatment). Some cells were pre-treated for 30 minutes with 10 µM bisindolylmaleimide (Boehringer-Mannheim), a PKC-specific inhibitor, prior to PMA treatment. After washing twice in room-temperature phosphate-buffered saline (PBS), cells were lysed in RIPA buffer as described (Lin et al., 1996). SSeCKS protein was immunoprecipitated from 150 µg of protein lysate, and electrophoresed in 5% SDSpolyacrylamide gels and then blotted onto Immobilon-P (Millipore) membranes. Immunoblotting was performed using rabbit antiphosphoserine antibody (Zymed; 1:500 dilution) preincubated with phosphotyrosine and phosphothreonine peptides (Nelson and Gelman, 1997), followed by horseradish peroxidase-labeled anti-rabbit Ig (Chemicon) and ECL luminescence reagent (NEN). After exposing, the same blot was probed for SSeCKS protein levels using antiSSeCKS serum, alkaline phosphatase-labeled anti-rabbit Ig and Western Blue reagent as described above. Bands were quantified by densitometric scanning of duplicate samples. To measure the phosphorylation of SSeCKS in response to glucose, SV40-MES or Rat-6 cells were seeded onto 10 cm dishes at 80% confluency and, the next day, starved of phosphate for 4 hours by incubating in two changes of phosphate-free MEM (ICN) containing 0.5% dialyzed calf serum (Gibco). The cells were labeled in vivo with [32P]orthophosphate (NEN), as described previously (Lin et al., 1996), while incubating for 4 hours in phosphate-free MEM/dialyzed calf serum supplemented with low glucose (5 mM) or high glucose (30 mM). SSeCKS protein immunoprecipitated from cell lysates (Lin et al., 1996) was subjected to SDS-PAGE, blotted onto PVDF membrane and probed with anti-SSeCKS serum as described above. The blots were then autoradiographed to detect phosphorylated SSeCKS protein. Phospho-amino acid analysis After labeling cells in vivo with [32P]orthophosphate as described above, SSeCKS was precipitated from RIPA lysates and then

363

subjected to SDS-PAGE and blotting onto PVDF membranes. Radiolabeled SSeCKS bands were excised, incubated with 6 N HCl for 1 hour at 120°C, cooled and then lyophilized. The products were electrophoresed as described in Boyle et al. (1991) in parallel to phosphoserine, -threonine and -tyrosine controls (Sigma). Control spots were visualized by 0.25% Ninhydrin in acetone spray followed by 2 minutes incubation at 100°C, and the SSeCKS phospho-amino acids identified by autoradiography. Mesangial cell immunofluorescence Primary human MC were seeded onto sterile 22 mm2 coverslips at a density of approximately 50% and allowed to grow for 18-24 hours. In some cases, cells were starved of serum overnight, then treated with 200 nM PMA in the presence or absence of 10 µM bisindolylmaleimide before fixing. The cells were fixed at −20°C for 20 minutes with pre-cooled 60% acetone/3.7% formaldehyde. After washing in PBS, the cells were incubated for 1 hour with immunoaffinity-purified rabbit polyclonal anti-SSeCKS antibody at a 1:250 dilution and rhodamine-labeled phalloidin (1:800; Sigma). Secondary antibodies to detect SSeCKS were fluorescein-labeled antirabbit Ig (Boehringer-Mannheim). The coverslips were mounted in Prolong antifade solution (Molecular Probes, Inc.), photographed on either a Zeiss Planopar fluorescence microscope with Kodak Elite II film (ASA 100), scanned (Argus II; Agfa, Inc.), or on a Zeiss Confocal Laser Scanning Microscope. Images were processed on a Silicon Graphics Indigo2XZ workstation with Vital Images’ Microscopy Workbench software (VoxelView, VoxelMath, VoxelAnimator) and on a Macintosh PowerPC 8100/100 AV with Adobe Photoshop (v4.0), and printed on a Codonics NP1600 dye sublimation printer. Anti-sense SSeCKS expression An EcoRI fragment containing the full-length SSeCKS cDNA (Lin et al., 1996) was cloned in the anti-sense orientation into the pBABEhygromycin retroviral vector. MC cells were infected with transiently produced (2 days post-transfection) anti-sense SSeCKS (or control, empty vector) virus packaged in amphotropic φNX cells, a gift of Garry Nolan (Stanford University Medical School), and then selected for growth in MC medium supplemented with hygromycin (85 µg/ml). Five individual clones (taken from different dishes, and thus not siblings) containing anti-sense SSeCKS or control virus were expanded, and tested by immunoblotting for SSeCKS protein levels (above). To ensure that these clones did not represent carryover of packaging cells, lysates of the anti-sense clones were shown to be negative when probed with polyclonal anti-Moloney leukemia virus immune serum (Quality Biotech, Inc.; data not shown). Additionally, these clones did not grow in media selective for φNX cells (diphtheria toxin).

RESULTS Abundant expression of SSeCKS is cell-specific in the kidney Data from our laboratory indicates that the major PKC substrate, SSeCKS, functions as a negative mitogenic regulator (Lin et al., 1995; Gelman et al., 1998), possibly by working at the junction of signaling and cytoskeletal control pathways. This suggests that cells or tissues with the highest steady state levels of SSeCKS expression might have low replication frequencies. In analyzing the tissue-specific expression of SSeCKS in rodents (J. Vargas Jr, E. Tombler and I. H. Gelman, unpublished), we noted that kidney expression of SSeCKS is focused in glomerular mesangial cells and parietal cells of Bowman’s capsule, cells types showing significant differentiation. Based on the similarity in the stellate

364


morphology of rodent fibroblasts overexpressing SSeCKS (Lin and Gelman, 1996; Gelman et al., 1998) and mesangial cells, we characterized the expression and localization of SSeCKS in primary human and rodent mesangial cells before and after PKC-mediated cytoskeletal remodelling. As in rodents (Lin et al., 1995; Chapline et al., 1996), humans transcribe SSeCKS in a wide variety of organs such as the heart, skeletal muscle, testes, ovary, and less so in brain, placenta, small intestine, colon and lung (Fig. 1). Unlike SSeCKS’ steady state mRNA expression in adult mouse and rat kidney (Lin et al., 1995; Chapline et al., 1996), levels of SSeCKS’ mRNA expression in whole adult human kidney are significantly less than in most other organs (Fig. 1). Nonetheless, rodent kidneys express very low levels of the major SSeCKS 280/290 kDa protein isoforms (Chapline et al., 1996; I. H. Gelman and E. Tombler, unpublished) indicating that SSeCKS is not expressed abundantly in most human and rodent adult renal cells. Additionally, the 6.3 Kb transcript in human kidney corresponds to the 6.1 kb transcript in rat kidney (Chapline et al., 1996), but differs from the 3.0 Kb transcript in mouse kidney (Lin et al., 1995). Although the smaller, most likely alternatively spliced, transcripts are found in some other organs in both humans and rodents, and may be responsible for encoding smaller SSeCKS-related proteins with organ-specific function, no such products were detected in lysates of either human or rodent kidneys. To determine if the low level of SSeCKS mRNA expression

Fig. 1. Transcription of SSeCKS in human tissues: SSeCKS is weakly expressed in the kidney. A northern blot of poly(A)-selected mRNA (2.0 µg per lane; Clontech) from multiple human organs was hybridized with 32P-labeled full-length rat SSeCKS cDNA and then washed under semi-permissive conditions as described in Materials and methods. The predominant transcript in most organs that express the SSeCKS gene is 6.3 Kb (arrow). Alternative sized transcripts (1.3, 1.35, 3.0, 5.2, 5.8, 6.8 and 8.0 Kb) are detected in some organs. Transcript sizes were determined relative to RNA markers (BRL) shown at left.

in whole human kidney reflects a low level of protein expression across differing renal cell types, we assayed for SSeCKS protein expression in primary MC in comparison to primary renal epithelial cells. A 305/287 kDa doublet of SSeCKS in human and mouse MC is expressed 2- to 5-fold above levels of the 290 kDa SSeCKS species in Rat-6 fibroblasts, and >ten-fold above levels in human renal epithelia from either the proximal tubule, thick ascending limb (of Henle’s loop), and collecting tubule (Fig. 2). A 250 kDa species of SSeCKS is more readily detectable in tubular epithelial cells than the larger SSeCKS species, suggesting that an alternative form of SSeCKS is specifically expressed across the differing renal epithelial cell types, albeit at a much reduced level compared to the MC and fibroblast. Thus, abundant protein expression of SSeCKS in human kidney is restricted to specific cell types. SSeCKS is rapidly serine-phosphorylated following activation of PKC in mesangial cells SSeCKS binds PKC in a phosphatidylserine-dependent manner and also is a major PKC substrate in vitro and in vivo (Lin et al., 1996; Chapline et al., 1996). We investigated whether SSeCKS could serve as a PKC substrate in MC. Fig. 3A shows that SSeCKS is rapidly serine-phosphorylated in MC by PMA treatment and that pre-incubation with the PKC-specific inhibitor, bis-indolylmaleimide, abrogates this phosphorylation. Unlike Rat-6 cells, in which the 280/290 kDa protein forms are predominant, in MC the 250 kDa form is as equally abundant as the slower mobility doublet. Although the phosphoserine-specific SSeCKS bands are never as sharp as those identified by the anti-SSeCKS polyclonal serum, the data in Fig. 3A suggest that PKC selectively phosphorylates the 250 and 305 kDa forms in MC. We confirmed that PMA induces serine-phosphorylation of SSeCKS using phospho-amino acid analysis (Fig. 3B).

Fig. 2. SSeCKS is abundantly expressed in primary human mesangial cells but not in primary human renal tubular epithelia. A western blot containing 20 µg/lane of protein from Rat-6 fibroblasts, primary human adult mesangial cells (MC), primary human adult proximal tubule epithelia (PT), primary human adult thick ascending limb (of Henle’s loop) epithelia (TAL), primary human adult collecting tubule epithelia (CT), primary human fetal mesangial cells, and 10 µg/lane from SV40-immortalized mouse MC was probed with anti-SSeCKS primary antibody, AP-labeled anti-rabbit Ig and stained with Western Blue reagent as described in Materials and methods. Note that anti-SSeCKS serum recognizes 290, 280 and 240 kDa SSeCKS forms in Rat-6 lysates (the latter not visible at 20 µg/lane of protein) whereas in human cells, 305, 287 and 250 kDa forms are recognized (relative to the 220 kDa marker protein at left).

SSeCKS-controlled mesangial cell cytoskeleton

365

Fig. 3. Phorbol ester-induced serine phosphorylation of SSeCKS in MC. (A) SSeCKS immunoprecipitates from PMAtreated (or mock-treated with dimethyl sulfoxide) primary human MC were probed with antiphosphoserine primary antibody, HRP-labeled anti-rabbit Ig and ECL chemiluminescence reagent, and then exposed to film (right). Some cells were pre-treated with the PKC-specific inhibitor, bisindolylmaleimide, for 30 minutes. The blots were then probed for levels of SSeCKS protein as in Fig. 2 (left). (B) Phosphoamino acid analysis of SSeCKS following PMA treatment. MC grown in the presence of [32P]orthophosphate were treated with PMA or DMSO for 10 minutes. After immunoprecipitation and SDS-PAGE, the labeled protein was isolated, digested with acetic acid and chromatographed as described in Materials and methods. Control phosphoserine, phosphothreonine and phosphotyrosine, labeled by Ninhydrin, are shown at right.

High glucose levels have been shown to activate PKC in MC, and it has been suggested that this contributes to diabetesinduced nephropathy (Derubertis and Craven, 1994; Heilig et al., 1997). We determined whether high glucose levels could induce phosphorylation of SSeCKS. Dialyzed calf serum (5%) was added to minimize the osmolarity effects of varying the glucose concentrations. Fig. 4 shows that SV40-MES grown either in high glucose for 4 hours or in low glucose for 2 hours then in high glucose for 2 hours have increased levels of SSeCKS phosphorylation relative to cells grown in low glucose for 4 hours. These data suggest that SSeCKS serves as a marker for PKC activation in MC. Staining of SSeCKS in the kidney is limited to mesangial cells, interstitial fibrocytes, and to the parietal epithelium of Bowman’s capsule Given that SSeCKS is abundant in primary mesangial cells but not in renal tubular epithelia, we next determined whether kidney immunohistochemical staining of SSeCKS would be limited to MC in situ. In frozen rat kidney sections processed to retain antigenic epitopes as well as in formaldehyde-fixed mouse kidney sections, SSeCKS stains prominently in MC, interstitial fibrocytes, and epithelial cells lining the parietal surface of Bowman’s capsule (Fig. 5), although in the latter case, it cannot be excluded that staining was coincident with extracellular matrix. Abundant staining of SSeCKS was not detected in glomerular endothelial cells or podocytes. In less antigenic sections from rat kidney and from human kidney core biopsies, processed by standard fixing protocols followed by block of endogenous peroxidase, SSeCKS staining in tubular epithelia was not appreciable, but remained detectable in the cells types listed above (data not shown). SSeCKS concentrates in the perikaryon and at the leading end of podosomal extensions in mesangial cells, elaborating a stellate morphology Immunofluorescence staining of MC (Fig. 6) shows that SSeCKS associates with a cortical cytoskeletal matrix as described previously in fibroblasts (Lin et al., 1996; Gelman et

al., 1998). In rodent fibroblasts, the cortical SSeCKS staining pattern is resistant to Triton X-100 permeabilization (Gelman et al., 1998). Both primary adult and fetal MC contain enrichments of SSeCKS in the termini of filopodia and in podosomes. These concentrations are likely sites of cytoplasmic growth and/or movement as they extend beyond the anchorage of F-actin stress fibers. Additionally, the subtle edge staining in some cells is probably plasma membraneassociation of myristylated SSeCKS isoforms (Gelman et al., 1998). It is unlikely to be due to extracellular SSeCKS because unfixed cells do not stain with our polyclonal antibody preparation, and SSeCKS cannot be labeled by extracellular iodination (P. Nelson and I. H. Gelman, unpublished data).

Fig. 4. High glucose levels induce SSeCKS phosphorylation. Serumstarved SV40-MES were labeled in vivo with [32P]orthophosphate during 4 hours growth in either high glucose (HG = 30 mM) or low glucose (LG = 5 mM), or 2 hours in low then 2 hours in high glucose. SSeCKS was then immunoprecipitated and immunoblotted as in Fig. 2 (B) and then autoradiographed (A). As a comparison, the left lane shows the phosphorylation of the 290 kDa SSeCKS form induced by serum addition to serum-starved Rat-6 cells, as we demonstrated previously (Nelson and Gelman, 1997).

366


Fig. 5. SSeCKS stains in mesangial cells and in the parietal epithelia of Bowman’s space. Glomerular expression of SSeCKS was detected by immunohistochemical staining of frozen rat (a,b) or formalin-fixed mouse (c,d) kidney sections. (a) and (c) were treated with normal rabbit serum whereas (b) and (d) were incubated with immunoaffinity-purified polyclonal antiSSeCKS antibody. Staining in mesangial cells (m) and parietal epithelia (p) of Bowman’s space are shown. Note that the tubule staining is not reproducible and probably reflects background staining (a). Bar, 25 µm; 630×.

We next determined whether short-term treatment with PMA affected the localization of SSeCKS in primary human MC cultures and in SV40-MES. We previously showed that PMAinduced PKC activation in Rat-6 fibroblasts caused SSeCKS to translocate rapidly from plasma membrane and cytoskeletal sites to the perinucleus (Lin et al., 1996). Fig. 7 shows that the podosome and filopodia enrichments of SSeCKS in primary MC disappear after 5 minutes of PMA treatment, followed by a change in SSeCKS cytoskeletal staining from cortical to filamentous. In contrast to the mock-treated cells, which

Fig. 6. Localization of SSeCKS in primary human adult and fetal MC. Adult and fetal primary MC were fixed and stained for SSeCKS or F-actin. Note the podosome enrichment of SSeCKS in untreated adult and fetal cells (triangles) extending past F-actin stress fibers (long arrows). In some cells, SSeCKS concentrates in punctate granules at the cell edge (short arrows). 630×.

contain fewer yet thicker stress fibers, 5 minutes of PMA treatment seems to induce the proliferation of greater numbers of thinner, longitudinal stress fibers. By 10 minutes of PMA treatment, after which most F-actin stress fibers are lost and SSeCKS is significantly serine-phosphorylated (Fig. 3), SSeCKS begins to translocate to the perinucleus. SSeCKS translocation continues following 30 and 60 minutes PMA treatment. Finally, after 60 minutes of PMA treatment a small fraction of SSeCKS is detected in membrane ruffles. Most significantly, PMA-induced translocation of SSeCKS

SSeCKS-controlled mesangial cell cytoskeleton

367

Fig. 7. Phorbol ester induces loss of SSeCKS in podosomes and translocation to the perinucleus in MC. Primary adult human MC mock-treated (0′) or treated with 200 nM PMA for 5, 10, 30 or 60 minutes were fixed and stained for SSeCKS protein (left panel of pair) or F-actin (right panel of pair) as in Fig. 6. Note that the concentrations of SSeCKS (arrows) are lost after 5 minutes of PMA treatment. Also note that SSeCKS staining becomes more filamentous and stress fiber numbers increase after 5 minutes of PMA treatment, whereas after longer treatments with PMA, SSeCKS translocates to perinuclear sites and stress fibers are lost. 630×.

coincides with an overall loss in stellate morphology and conversion to increasingly refractile cells. Although the morphology of untreated SV40-MES is less stellate than primary MC (Fig. 8a,a′), and SSeCKS is not concentrated in podosome structures, PMA induces a similar effect on SSeCKS translocation and stress fiber assembly. For example, 5 minutes of PMA results in an increase in stress fiber number (Fig. 8; compare a′ to b′) followed by a loss of stress fibers after longer treatment (c′-e′). SSeCKS translocation to the perinucleus is apparent after 10 minutes of PMA treatment, and association of a small fraction of SSeCKS with membrane ruffles is apparent after 30 and 60 minutes of treatment (d,d′ and e,e′). These data clearly show a link between PKC activation, SSeCKS translocation and control of stellate morphology in MC.

Loss of SSeCKS expression alters actin-based stellate MC morphology In order to assess the role of SSeCKS in defining MC stellate morphology, we attempted to ablate endogenous SSeCKS protein levels by expressing full-length SSeCKS anti-sense cDNA (ASN) via retroviral vectors. Fig. 9C shows a loss of the 305 and 287 kDa SSeCKS protein forms in several typical clones containing the ASN construct compared to controls. Control cells containing empty vector exhibit stellate, polygonal morphologies (Fig. 9Ba,b). IFA analysis shows typical patches of thin stress fibers emanating towards vinculin-associated focal complexes enriched at cell edges (Fig. 9Aa-c). In contrast, cells containing SSeCKS ASN are elongated and fibroblastic, some even exhibiting increased refractility reminiscent of oncogenic

Fig. 8. Phorbol ester-induced translocation of SSeCKS in SV40-MES. SV40-MES mock-treated (a,a′) or treated with 200 nM PMA for 5 (b,b′), 10 (c,c′), 30 (d,d′) or 60 minutes (e,e′) were fixed and stained for SSeCKS protein and F-actin as in Fig. 6. Although these cells are more refractile and less stellate than primary MC, SSeCKS is somewhat enriched in podocytes, filopodia and at cell edges in untreated cells. As with MC, short-term (5 minutes) treatment with PMA causes an increase in filamentous SSeCKS staining and F-actin formation (b,b′). Longer treatments with PMA result in translocation of SSeCKS to perinuclear sites (pn) and membrane ruffles (mr). 630× (bar, 25 µm).

368


Fig. 9. Ablation of SSeCKS expression in MC results in alteration of actin-based stellate morphology. (A) MC cells were infected with packaged amphotropic retrovirus containing pBABEhygro control vector (a-c) or pBABEhygro containing the ASN orientation of full-length rat SSeCKS cDNA (d-i), and then hygromycin-resistant colonies were selected as described in Materials and methods. Control (a-c) or ASN clones (clone 12, d-f; clone 3, g-i) were fixed and stained for vinculin (a,d,g) or F-actin (b,e,h); (c,f,i) are computer-generated composite images. (B) Control (a,b) or ASN clones (clone 12, c; clone 3, d) were photographed with Hoffman optics (250×) on an Olympus IX-70 using Kodak T-MAX (ASA 400) film. (C) Immunoblot analysis showing loss of SSeCKS 287/305 kDa protein forms in MC clones transduced with SSeCKS anti-sense cDNA relative to MC controls (vector-transduced). The lower part of the blot was probed independently for β-actin as a control.

transformation (Fig. 9Bc,d). In ASN clones, stress fibers are thick and longitudinal, and focal complexes are distributed along stress fibers away from the cell edge. ASN clones with the lowest SSeCKS level (clones 3 and 12) showed the greatest degree of stress fiber thickening, whereas clones with higher levels of SSeCKS (e.g. clone 5) showed only partial thickening (data not shown). We ruled out the possibility that these cell lines were carryover contaminants of the φNX packaging cells by showing a lack of Moloney virus-derived Env, Pol and Gag proteins in the MC clones (by immunoblotting with polyclonal anti-MLV sera) and by showing that the MC clones failed to proliferate in the selective media for the φNX cells (data not shown). These data clearly indicate a role for SSeCKS in maintaining actin-based stellate MC morphology. DISCUSSION Mesangial cells maintain the structure and function of the glomerulus by elaboration of supportive cellular structures, secretion of extracellular matrices, endocytosis of debris, and

through cellular contraction (Mene et al., 1989; Johnson et al., 1992). Numerous studies have demonstrated PKC-dependent alteration of these MC functions. However, few reports have identified downstream targets of PKC in MC that may mediate these changes in function. Here, we report that SSeCKS is a novel PKC substrate of MC that likely plays a role in MC differentiation and, based on our previous results in fibroblasts (Lin et al., 1995, 1996; Lin and Gelman, 1997; Nelson and Gelman, 1997), may also function in controlling MC mitogenesis and proliferation. In human and rodent kidneys, abundant expression of SSeCKS protein is limited to the small MC and fibrocyte population. This correlates well in humans with the low abundance of SSeCKS mRNA from whole kidney. In contrast, the significant levels of 3 and 6 Kb SSeCKS messages in mouse and rat kidney, respectively, do not correlate with the selective protein expression in MC. This may be a vestige of the abundant SSeCKS protein expression we find in embryonic rodent kidneys (I. H. Gelman, E. Tombler and J. Vargas Jr, unpublished). Alternatively, these messages might encode proteins in adult rodent kidneys not recognized by our sera.

SSeCKS-controlled mesangial cell cytoskeleton The finding that SSeCKS stains in the parietal epithelia of Bowmans’ space suggests that SSeCKS plays a role in establishing the functional and morphological differences between this spindle-like cell and its contiguous columnar partner in the proximal tubule. The expression in Bowman’s capsule is analogous to the phenotypic change induced by FSP1, a fibroblast marker expressed de novo by renal tubular epithelia during persistent inflammation, causing tubular cells to flatten, elongate, reduce cytokeratin expression and express vimentin (Strutz et al., 1995). Our data strongly suggest that SSeCKS functions downstream of PKC to mediate changes in cytoskeletal architecture and cell morphology. As we showed in fibroblasts, SSeCKS is rapidly phosphorylated following the treatment of MC with phorbol esters. We also show that PMA induced translocation of SSeCKS to perinuclear and membrane ruffling sites, is concomitant with induced cell-shape changes and loss of actin stress fibers. Interestingly, the shortest PMA treatments induced the transient formation of longitudinal stress fibers in both MC and SV40-MES, and caused a parallel polarization of SSeCKS-associated fibers. This correlates with previous evidence showing a dependence of integrin-mediated mitogenic signaling on the presence of F-actin fibers and their physical linkage to focal contact complexes (Defilippi et al., 1997; Niu and Nachmias, 1994). MC proliferation and morphological change have been documented in diabetesinduced nephropathy (Heilig et al., 1997), and several studies have shown PKC activation by high levels of glucose (Derubertis and Craven, 1994; Heilig et al., 1997; Studer et al., 1997; Zhou et al., 1997; Koya et al., 1997; Cosio, 1995). Shortterm treatment of MC with high glucose induced both loss of stellate morphology (data not shown) and SSeCKS phosphorylation, strengthening the notion that SSeCKS functions downstream of PKC in controlling cell shape. We also show evidence that SSeCKS is required for maintenance of the stellate morphology in MC. Specifically, expression of anti-sense SSeCKS cDNA via retroviral vectors caused remodeling of MC actin-based cytoskeletons. The thickening and polarization of F-actin stress fibers in SSeCKSdeficient cells is reminiscent of the earliest effects of PMA treatment (compare Fig. 9Ae,h to Fig. 7d′). Thick actin fibers associate with α-actinin and talin (Zand and Buehler, 1989; Katoh et al., 1995), cytoskeletal proteins that are recruited to focal contacts after mitogenic stimulation (Ben Ze’ev, 1997). Thus, loss of SSeCKS seems to induce the initial cytoskeletal state of mitogenic activation, suggesting that SSeCKS functions to modulate recruitment of signaling proteins to focal complexes. As SSeCKS expression in fibroblasts is cell-cycle regulated, with highest expression in early G1 (or in contact-inhibited cells) and a rapid loss of expression following growth factor induced G1→S progression (Lin et al., 1995; Nelson and Gelman, 1997), our finding in MC suggests that SSeCKS is required for maintenance of the stellate morphology in G1. Also, since the ASN-MC clones proliferate faster than controls (data not shown), ablation of SSeCKS expression may alter cell cycle phase length. As SSeCKS was originally identified as being transcriptionally downregulated in src- and rastransformed cells (Lin et al., 1995), we are investigating whether the ASN-MC (which appear somewhat oncogenically transformed) gain oncogenic-like growth characteristics such

369

as anchorage-independence observed in hyperproliferative mesangial cell lesions. The concentration of SSeCKS at the growing end of podosomes, and the translocation of SSeCKS from these cortical actin extensions following PKC activation, suggest that SSeCKS plays an important role in the establishment of MC morphological differentiation. This notion is supported by the stellate morphology seen in SSeCKS overexpressing fibroblasts (Lin and Gelman, 1997), and the relative lack of expression of SSeCKS in cuboidal and columnar renal tubular epithelial cells (this paper; I. H. Gelman, E. Tombler and J. Vargas Jr, unpublished). SSeCKS’ role in MC morphological differentiation may be in organizing cortical cytoskeletal matrices, including those involved in the formation of podosomal extensions, suggesting that SSeCKS has some similarities to the ezrin/radixin/moesin (ERM) family of proteins (reviewed in Tsukita and Yonemura, 1997). Another intriguing similarity of ERM proteins to SSeCKS is the recent discovery that at least one ERM protein, ezrin, functions as an AKAP (Dransfield et al., 1997). AKAPs are thought to control PKA function by sequestering type II PKA to distinct subcellular compartments, and the AKAP superfamily consists of several protein families of varying sizes and tissue-specific expression (reviewed in Scott and McCartney, 1994; Dell’Acqua and Scott, 1997). Both SSeCKS and a highly related protein, Gravin, bind type II PKA via common Cterminal amphipathic helical domains (Nauert et al., 1997). Thus, SSeCKS may sequester type II PKA to specific membrane and cytoskeletal sites in MC. Interestingly, PKA activation in MC inhibits mitogenesis in vitro and in vivo (Dell’Acqua and Scott, 1997; Tsuboi et al., 1996), but it remains to be determined how SSeCKS’ AKAP function relates to mitogenic control or cell differentiation. One possibility is suggested by studies of dedifferentiated thyroid cells, where simultaneous inhibition of PKC and activation of PKA restored cellular differentiation (Gallo et al., 1995), whereas activation of PKC caused a loss of PKA activity through translocation of type II PKA from membranes to cytosol, presumably through PKC phosphorylation of AKAPs (Gallo et al., 1995; Feliciello et al., 1996). These data suggest that AKAPs such as SSeCKS play critical regulatory roles at the junction of PKA and PKC signaling pathways. We thank the following investigators for their contribution to this work: William Schnaper and Jeffrey Kopp for primary adult and fetal human mesangial cells; Patricia Wilson for microdissected primary human kidney epithelial cells; Michael Lipkowitz for SV-40 immortalized mouse mesangial cells; Basil Hans for help with rat kidney histology preparation; and Steven Dikman for histological reading of SSeCKS staining in the kidney. This work was supported by the National Cancer Institute grant R29-CA65787 (I. H. G.), NCI Minority Pre-Doctoral Supplement (J. V., Jr) and NIH-NIDDK grant 5PO1-DK50795 (P. E. K). Confocal laser scanning microscopy was performed at the MSSM-CLSM core facility, supported with funding from NIH shared instrumentation grant (1 S10 RR0 9145-01) and NSF Major Research Instrumentation grant (DBI-9724504).

REFERENCES Ben Ze’ev, A. (1997). Cytoskeletal and adhesion proteins as tumor suppressors. Curr. Opin. Cell Biol. 9, 99-108. Berfield, A. K., Douglas, S. and Abrass, C. K. (1997). Insulin-like growth

370


factor (IGF-1) induces unique effects in the cytoskeleton of culture rat glomerular mesangial cells. J. Histochem. Cytochem. 45, 583-593. Berfield, A. K., Raugi, G. J. and Abrass, C. K. (1996). Insulin induces rapid and specific rearrangement of the cytoskeleton of rat mesangial cells in vitro. J. Histochem. Cytochem. 44, 91-101. Boyle, W. J., van der Geer, P. and Hunter, T. (1991). Phosphopeptide mapping and phosphoaminoacid analysis by two-dimensional separation on thin-layer cellulose plates. Meth. Enzymol. 201, 110-152. Bradford, M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analyt. Biochem. 72, 248-254. Chapline, C., Mousseau, B., Ramsay, K., Duddy, S., Li, Y., Kiley, S. C. and Jaken, S. (1996). Identification of a major protein kinase C-binding protein and substrate in rat embryo fibroblasts – decreased expression in transformed cells. J. Biol. Chem. 271, 6417-6422. Cosio, F. G. (1995). Effects of high glucose concentrations on human mesangial cell proliferation. J. Am. Soc. Nephrol. 5, 1600-1609. Defilippi, P., Venturino, M., Gulino, D., Duperray, A., Boquet, P., Fiorentini, C., Volpe, G., Palmieri, M., Silengo, L. and Tarone, G. (1997). Dissection of pathways implicated in integrin-mediated actin cytoskeleton assembly. Involvement of protein kinase C, Rho GTPase, and tyrosine phosphorylation. J. Biol. Chem. 272, 21726-21734. Dell’Acqua, M. L. and Scott, J. D. (1997). Protein kinase A anchoring. J. Biol. Chem. 272, 12881-12884. Derubertis, F. and Craven, P. (1994). Activation of protein kinase C in glomerular cells in diabetes. Diabetes 43, 1-8. Dransfield, D. T., Bradford, A. J., Smith, J., Martin, M., Roy, C., Mangeat, P. H. and Goldenring, J. R. (1997). Ezrin is a cyclic AMP-dependent protein kinase anchoring protein. EMBO J. 16, 35-43. Feliciello, A., Giuliano, P., Porcellini, A., Garbi, C., Obici, S., Mele, E., Angotti, E., Grieco, D., Amabile, G., Cassano, S., Li, Y., Musti, A. M., Rubin, C. S., Gottesman, M. E. and Avvedimento, E. V. (1996). The vKi-Ras oncogene alters cAMP nuclear signaling by regulating the location and the expression of cAMP-dependent protein kinase IIβ. J. Biol. Chem. 271, 25350-25359. Gallo, A., Feliciello, A., Varrone, A., Cerillo, R., Gottesman, M. E. and Avvedimento, V. E. (1995). Ki-ras oncogene interferes with the expression of cyclic AMP- dependent promoters. Cell Growth Differ. 6, 91-95. Ganz, M., Saska, B., Saxena, R., Hawkins, K. and Sedor, J. (1996). PDGF and IL-1 induce and activate specific protein kinase C isoforms in mesangial cells. Am. J. Physiol. Renal Physiol. 271, F108-F113. Gelman, I. H., Lee, K., Tombler, E., Gordon, R. and Lin, X. (1998). Control of cytoskeletal architecture by the src-suppressed C kinase substrate, SSeCKS. Cell Motil. Cytoskel. 41, 1-17, 1998. Heilig, C. W., Brosius, F. C.-3rd and Henry, D. N. (1997). Glucose transporters of the glomerulus and the implications for diabetic nephropathy. Kidney Int. Suppl. 60, S91-9. Hiroi, J., Kimura, K., Aikawa, M., Tojo, A., Suzuki, Y., Nagamatsu, T., Omata, M., Yazaki, Y. and Nagai, R. (1996). Expression of a non-muscle myosin heavy chain in glomerular cells differentiates various types of glomerular diseases in rats. Kidney Int. 49, 1231-1241. Hugo, C., Pichler, R., Gordon, K., Schmidt, R., Amieva, M., Couser, W. G., Furthmayr, H. and Johnson, R. J. (1996). The cytoskeletal linking proteins, moesin and radixin, are upregulated by platelet-derived growth factor, but not basic fibroblast growth factor in experimental mesangial proliferative glomerulonephritis. J. Clin. Invest. 97, 2499-2508. Jaken, S. (1996). Protein kinase C isozymes and substrates. Curr. Opin. Cell Biol. 8, 168-173. Johnson, R. J., Floege, J., Yoshimura, A., Hiroyuki, I., Couser, W. G. and Alpers, C. E. (1992). The activated mesangial cell: a glomerular ‘myofibroblast’. J. Am. Soc. Nephrol. 2, S109-S197. Johnson, R. J., Iida, H., Alpers, C. E., Majesky, M. W., Schwartz, S. M., Pritzi, P., Gordon, K. and Gown, K. M. (1991). Alpha-smooth muscle actin is a marker of mesangial cell proliferation. J. Clin. Invest. 87, 847858. Katoh, K., Masuda, M., Kano, Y., Jinguji, Y. and Fujiwara, K. (1995). Focal adhesion proteins associated with apical stress fibers of human fibroblasts. Cell Motil. Cytoskel. 31, 177-195. Klauck, T., Faux, M. C., Labudda, K., Langeberg, L. K., Jaken, S. and

Scott, J. D. (1996). Coordination of three signaling enzymes by AKAP79, a mammalian scaffold protein. Science 271, 1589-1592. Koya, D., Jirousek, M. R., Lin, Y. W., Ishii, H., Kuboki, K. and King, G. L. (1997). Characterization of protein kinase C beta isoform activation on the gene expression of transforming growth factor-beta, extracellular matrix components, and prostanoids in the glomeruli of diabetic rats. J. Clin. Invest. 100, 115-126. Lin, X. and Gelman, I. H. (1997). Re-expression of the major protein kinase C substrate, SSeCKS, suppresses v-src-induced morphological transformation and tumorigenesis. Cancer Res. 57, 2304-2312. Lin, X., Nelson, P. J., Frankfort, B., Tombler, E., Johnson, R. and Gelman, I. H. (1995). Isolation and characterization of a novel mitogenic regulatory gene, 322, which is transcriptionally suppressed in cells transformed by src and ras. Mol. Cell. Biol. 15, 2754-2762. Lin, X., Tombler, E., Nelson, P. J., Ross, M. and Gelman, I. H. (1996). A novel src- and ras-suppressed protein kinase C substrate associated with cytoskeletal architecture. J. Biol. Chem. 271, 28,430-28,438. Mene, P. and Cinotti, M. T. (1992). Signal transduciton in mesangial cells. J. Am. Soc. Nephrol. 2, S100-S106. Mene, P., Simonson, M. S. and Dunn, M. J. (1989). Physiology of the mesangial cell. Physiol. Revs 69, 1347-1427. Nauert, J., Klauck, T., Langeberg, L. K. and Scott, J. D. (1997). Gravin, an autoantigen recognized by serum from myasthenia gravis patients, is a kinase scaffolding protein. Curr. Biol. 7, 52-62. Nelson, P. and Gelman, I. H. (1997). Cell-cycle regulated expression and serine phosphorylation of the myristylated protein kinase C substrate, SSeCKS: correlation with cell confluency, G0 phase and serum response. Mol. Cell. Biochem. 175, 233-241. Niu, M. Y. and Nachmias, V. T. (1994). Two-step mechanism for actin polymerization in human erythroleukemia cells induced by phorbol ester [published erratum appears in Cell Motil. Cytoskel. 1994; 28(4): 359]. Cell Motil. Cytoskel. 27, 327-336. Pfeilschifter, J. M. (1994). Regulatory functions of protein kinase C isoenzymes in glomerular mesangial cells [editorial]. Nephrol. Dial. Transplant. 9, 1061-1062. Rozengurt, E. (1995). Convergent signalling in the action of integrins, neuropeptides, growth factors and oncogenes. Cancer Surv. 24, 81-96. Schnaper, H. W., Kopp, J. B., Poncelet, A. C., Hubchak, S. C., Stetler Stevenson, W. G., Klotman, P. E. and Kleinman, H. K. (1996). Increased expression of extracellular matrix proteins and decreased expression of matrix proteases after serial passage of glomerular mesangial cells. J. Cell Sci. 109, 2521-2528. Scott, J. D. and McCartney, S. (1994). Localization of A-kinase through anchoring proteins. Mol. Endocrinol. 8, 5-11. Strutz, F., Okada, H., Lo, C. W., Danoff, T., Carone, R. L., Tomaszewski, J. E. and Neilson, E. G. (1995). Identification and characterization of a fibroblast marker: FSP1. J. Cell Biol. 130, 393-405. Studer, R. K., Craven, P. A. and DeRubertis, F. R. (1997). Antioxidant inhibition of protein kinase C-signaled increases in transforming growth factor-beta in mesangial cells. Metabolism 46, 918-925. Tamura, M., Tanaka, H., Hirona, T., Ueta, Y., Higashi, K. and Hirona, H. (1996). Enhanced glomerular profilin gene and protein expression in experimental mesangial proliferative glomerulonephritis. Biochem. Biophys. Res. Commun. 222, 683-687. Tsuboi, Y., Shankland, S. J., Grande, J. P., Walker, H. J., Johnson, R. J. and Dousa, T. P. (1996). Suppression of mesangial proliferative glomerulonephritis development in rats by inhibitors of cAMP phosphodiesterase isozymes types III and IV. J. Clin. Invest. 98, 262-270. Tsukita, S. and Yonemura, S. (1997). ERM (ezrin/radixin/moesin) family: From cytoskeleton to signal transduction. Curr. Opin. Cell Biol. 9, 70-75. Wilson, P., Dillingham, M. A., Breckon, R. and Anderson, R. J. (1985). Defined human renal tubular epithelia in culture: growth, characterization, and hormonal response. Am. J. Physiol. 248, F436-443. Zand, M. S. and Albrecht Buehler, G. (1989). What structures, besides adhesions, prevent spread cells from rounding up? Cell Motil. Cytoskel. 13, 195-211. Zhou, X., Li, C., Dlugosz, J., Kapor Drezgic, J., Munk, S. and Whiteside, C. (1997). Mesangial cell actin disassembly in high glucose mediated by protein kinase C and the polyol pathway. Kidney Int. 51, 1797-1808.