|Home | About | Journals | Submit | Contact Us | Français|
Trophoblast glycoprotein (Tpbg), a 72-kDa transmembrane glycoprotein, is known to regulate the phenotypes of epithelial cells by modifying actin organization and cell motility. Recently, a microarray study showed that Tpbg is upregulated in Thy1 glomerulonephritis (Thy1 GN). We hypothesized that Tpbg regulates cytoskeletal rearrangement and modulates phenotypic alteration in podocytes under pathological conditions.
We examined Tpbg expression in Thy1 GN and Tpbg function in mouse podocytes.
We demonstrated that Tpbg is upregulated in the injured podocytes of Thy1 GN. In vitro, immunofluorescence studies revealed that Tpbg colocalized with the focal adhesion protein, vinculin, in parallel with stress fiber formation. This colocalization was observed even when actin filaments were depolymerized with cytochalasin D. Tpbg localization at focal adhesions was induced by dominant-active RhoA and suppressed by the ROCK1 inhibitor Y-26732. In addition, transforming growth factor-β increased Tpbg expression at focal adhesions concurrently with rearrangement of stress fibers. Stress fiber formation was suppressed in differentiated podocytes transfected with full-length Tpbg. Furthermore, knockdown of Tpbg using small interfering RNA decreased podocyte motility.
Our findings suggest a novel role of Tpbg in the phenotypic alteration of injured podocytes, and we accordingly propose a new mechanism of glomerular injury in glomerulonephritis.
Podocyte injury is a critical determinant of the development of glomerular damage in glomerulonephritis, although mechanisms are not well understood [1,2,3,4]. Thy1 glomerulonephritis (Thy1 GN), a widely used model of human mesangial proliferative glomerulonephritis, involves mesangial cell death, mesangiolysis and aneurysmal dilatation of intraglomerular capillaries. Dilated capillaries may induce podocyte damage via altered physical forces  or changes in glomerular basement membrane composition . In human IgA nephropathy, podocyte injury is induced by mesangial cell-derived cytokines .
Campean et al.  reported that trophoblast glycoprotein (Tpbg) was upregulated in glomeruli by an RNA microarray study using the Thy1 model. Tpbg is a 72-kDa leucine-rich repeat transmembrane glycoprotein. In normal adult tissues, Tpbg expression is low and restricted to certain epithelia. In many human carcinomas, Tpbg expression is upregulated and strongly associated with metastasis [8,9,10]. In fibroblasts and epithelial cells, Tpbg has marked effects on the actin cytoskeleton, cell motility, and cell attachment [11,12]. In addition, Tpbg is associated with differentiation and epithelial-mesenchymal transition of mouse and human embryonic stem cells [13,14,15]. We hypothesized that Tpbg serves to link mesangial proliferation and podocyte injury in mesangio-proliferative GN, as modeled by Thy1 GN.
Male Wistar rats (CLEA Japan, Tokyo, Japan) weighing 180–200 g were used. Rats were housed under specific pathogen-free conditions. All animal experiments were performed in accordance with institutional guidelines, and the Review Board of Tokushima University granted ethical permission for this study.
Thy1 GN was induced by a single intravenous injection of mouse anti-rat Thy-1.1 monoclonal antibody (1 mg/kg; Cedarlane Laboratories, Ont., Canada). These rats were sacrificed at day 3, 6, and 14 (n = 5–8) after administration of anti-Thy-1.1 antibody.
Before sacrifice, the rats were individually housed in metabolic cages with free access to water for 24-hour urine collection. Urinary protein concentration was determined by the Bradford method (Bio-Rad, Oakland, Calif., USA). Negative controls were six age-matched rats injected with vehicle only. Glomeruli were isolated from the renal cortex of the rats by differential thieving, with a purity of >90%.
The total RNA was extracted from isolated rat glomeruli or cultured podocytes by TRIzol (Takara, Ootsu, Japan). The following PCR primers were used: rat Tpbg: sense 5′-GTCCTTCACAACTCCACCTTG-3′, antisense 5′-CGGCACCACCTCTGTCTCTTTA-3′, yielding a 135-bp PCR product; rat nephrin: sense 5′-TGAAGACACAGACCACCAGC-3′, antisense 5′-GGAGAGCAGCAGAAGACCAC-3′, yielding a 102-bp PCR product; rat WT1: sense 5′-TTGAAGGGAATGGCTGCTGG-3′, antisense 5′-GAGGATGGGGGTTGTGTGG-3′, yielding a 105-bp PCR product; rat transforming growth factor-β (TGF-β): sense 5′-CTGACCCCCACTGATACG-3′, antisense 5′-CACTGAAGCGAAAGCCCTG-3′, yielding a 103-bp PCR product; 18SrRNA: sense 5′-GACTCAACACGGGAAACC-3′, antisense 5′-CGGACATCTAAGGGCATCAC-3′, yielding a 270-bp PCR product; murine Tpbg: sense 5′-CTACTGCTGCTTTGCTCACGC-3′, antisense 5′-CACCTCCTCAACTCCTTTGTTG-3′, yielding a 160-bp PCR product; murine GAPDH: sense 5′-AAAATGGTGAAGGTCGGTGTG-3′, antisense 5′-AATGAAGGGGTCGTTGATGG-3′, yielding a 110-bp PCR product).
PCR products were polymerized using cDNA solution and GoTaq (Promega, Madison, Wisc., USA). Quantitative PCR was performed on an ABI PRISM 7700 sequence detector (Applied Biosystems, Carlsbad, Calif., USA) using SYBR Green PCR Master Mix (Applied Biosystems) and specific primers following the manufacturer's protocols. For quantitative PCR of mouse Tpbg, a TaqMan probe of Tpbg and universal PCR Master Mix were used.
Isolated glomeruli and cultured podocytes were incubated in a lysis buffer (10 mM Tris, 150 mM NaCl, 1.0% NP-40, proteinase inhibitors) for 30 min at 4°C. Protein concentrations were measured by DC protein assay (Bio-Rad Laboratories, Hercules, Calif., USA). Protein samples were heated to 100°C for 3 min in SDS gel-loading buffer, 20 μg of each glomerular sample was applied to SDS gel electrophoresis and proteins were transferred to nitrocellulose filters (GE Healthcare, Little Chalfont, UK). The blots were incubated with anti-Tpbg antibody, followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG (Zymed, San Francisco, Calif., USA).
Polyclonal anti-Tpbg antibody was raised in a rabbit against peptides corresponding to carboxyl-terminal region of mouse Tpbg (INADPRLTNLSSNSDV), and the IgG fraction was purified using protein A sepharose. This peptide sequence corresponds to 93% of carboxyl-terminal region of rat Tpbg. This antibody recognizes specifically a band about 72 kDa in rat glomerular lysate and in mouse podocyte lysate. Antibody specificity was confirmed by peptide blocking assay in in vivo immunostaining. Tpbg antibody was preabsorbed overnight with 25 times the concentration of Tpbg peptide.
For light microscopy, tissues were fixed in methyl Carnoy's solution, and 2-μm paraffin sections were stained with periodic acid-Schiff. Glomerulosclerosis score was semiquantitatively analyzed. The percentage of each glomerulus occupied by mesangial matrix was estimated and assigned a code as follows: 0 = absent; 0.5 = 1–5%; 1 = 5–25%; 2 = 25–50%; 3 = 50–75%, or 4 = 75–100%. The total number of cells in the glomeruli was counted in a blind protocol and computed for 20 full-sized glomeruli (80–100 μm) for each kidney. For immunofluorescence microscopy, frozen 4-μm sections were fixed in acetone for 10 min at 4°C. For double-labeled immunofluorescence microscopy for Tpbg with synaptopodin or nephrin, mouse monoclonal anti-synaptopodin antibody (Progen, Heidelberg, Germany), goat polyclonal anti-nephrin antibody (Santa Cruz, Santa Cruz, Calif., USA) and rabbit polyclonal anti-Tpbg antibody were used as primary antibody. Alexa Fluor 488- or Alexa Fluor 594-conjugated secondary antibodies (Invitrogen, Carlsbad, Calif., USA) were used. For double immunostaining for Tpbg with WT1, sections were incubated with rabbit polyclonal anti-WT1 antibody (Santa Cruz). Next, after blocking with the avidin solution and biotin solutions, sections were incubated with biotin-conjugated rabbit polyclonal anti-Tpbg antibody, followed by Texas Red-conjugated streptavidin (Zymed).
Conditionally immortalized murine podocytes were provided by Dr. Peter Mundel. Podocytes were cultured in a RPMI-1640 (Sigma) medium containing 10% fetal bovine serum, 100 U/ml penicillin, 0.1 mg/ml streptomycin and 2 mM l-glutamine. For propagation, cells were cultivated with a culture medium supplemented with 50 U/ml of recombinant mouse γ-interferon (PeproTech, London, UK) at 33°C with 5% CO2 (permissive conditions). To induce differentiation, cells were cultured on a type I collagen at 37°C without γ-interferon for 14 days (nonpermissive conditions). Podocytes between passage 20 and 25 were used for all experiments.
Differentiated podocytes were treated with 5 μM cytochalasin D (Sigma) for 2 h to disrupt the actin cytoskeleton. Stress fibers were disrupted by treating cells with 10 μM of the specific Rho kinase (ROCK) inhibitor Y-27632 (Wako, Osaka, Japan). To examine the effect of TGF-β on Tpbg expression and actin filament, differentiated podocytes were treated with TGF-β (PeproTech) for 24 h, after serum starvation in 1% RPMI. To inhibit the effect of TGF-β type I receptor, the activin-like kinase receptor 5 (ALK5), podocytes were treated with 0, 1.0 and 10 μM of the ALK5 inhibitor SB 431542 (Sigma).
For immunofluorescence, cells were fixed using 2% paraformaldehyde and permeabilized with 0.1% Triton X. Actin filaments were visualized with Texas Red-conjugated phalloidin (Invitrogen). Rabbit polyclonal anti-Tpbg antibody at 80 μg/ml, mouse monoclonal anti-vinculin antibody (Sigma), mouse monoclonal anti-myc antibody (MBL, Nagano, Japan) and mouse monoclonal anti-FLAG antibody were used as primary antibody. Propidium iodide was used as nuclear staining. For peptide blocking assay, Tpbg antibody was preabsorbed overnight with 25 times the concentration of Tpbg peptide. Specimens were viewed with a confocal laser scanning microscopy (Leica, Wetzlar, Germany).
Podocytes were transfected with Myc-tagged dominant active (pEFBOS-Myc-RhoA-Q14N) RhoA, dominant negative (pEFBOS-Myc-RhoA-T18N) RhoA and full-length Tpbg constructs. Full-length Tpbg cDNA were amplified by RT-PCR from podocyte RNA and inserted into p3 × FLAG-CMV™-14 expression vector (Sigma). The authenticity of these expression plasmids was confirmed by DNA sequencing. Transient transfection of podocytes was performed using FuGene 6 reagent (Roche, Indianapolis, Ind., USA).
Areas and signal intensities of actin fibers, Tpbg and vinculin were identified by immunofluorescence. Positive areas were then converted to black pixels and traced, and the size and signal intensity were measured. The intensity of stress fibers, the ratio of stress fibers to total cell size, and the intensity of Tpbg and vinculin at the focal adhesion were quantified using Multi Gauge v2.2 software (Fujifilm, Tokyo, Japan).
Small interfering RNA (siRNA) against Tpbg (si-Tpbg) and scrambled control siRNA (si- scramble) were purchased from Sigma. Three si-Tpbgs targeted to different sites in the message were used. Differentiated podocytes were transfected with siRNA using N-TER Nanoparticle siRNA transfection system (Sigma). To improve the transfection efficiency, repeated transfection of siRNA was performed at 0 and 24 h. Tpbg expression in podocytes was analyzed by real-time PCR 48 h after first transfection.
Differentiated podocyte monolayer culture on coverslips was wounded with a 200-μl pipette tip after transfection with si-scramble or si-Tpbg and incubated in the absence or presence of 10 ng/ml TGF-β for 24 h. Cell layers were stained with DAPI, photographed using BZ-8000 (Keyence, Osaka, Japan), and the wound width was measured. Migratory rates were calculated as (wound width 0 h – wound width 24 h)/wound width 0 h × 100%.
Data are expressed as mean ± SD. Statistical significance was defined as p < 0.05 and was evaluated using one-factor analysis of variance or Kruskal-Wallis test followed by a post-hoc test.
Expression of Tpbg mRNA was observed in rat isolated glomeruli by RT-PCR (fig. (fig.1a).1a). Specific reaction of the anti-Tpbg antibody with the 72-kDa Tpbg was also demonstrated in glomerular whole lysate by Western blot analysis (fig. (fig.1b1b).
We examined the expression of Tpbg in Thy1 GN. Histological examination showed mesangiolysis at day 3 and the increase of mesangial matrix at day 6. Significant increase in total cell count and extracellular matrix was observed at day 6 (fig. (fig.2a).2a). Urinary protein excretion significantly increased at both day 3 and day 6. Quantitative PCR showed that Tpbg expression in glomeruli significantly increased by 130% at day 6 (p < 0.01; fig. fig.2c).2c). Glomerular expression of Tpbg protein also significantly increased at day 6 compared with control (fig. (fig.2d).2d). By double immunofluorescence of control rat glomeruli, Tpbg was faintly stained and partially colocalized with nephrin (fig. (fig.2e).2e). Tpbg was stained in cell bodies of podocyte labeled with WT1 at day 6 (fig. (fig.2e).2e). Tpbg partially colocalized with synaptopodin (fig. (fig.2h)2h) or nephrin (fig. (fig.2i),2i), perhaps due to expression in the central portions of podocytes. At day 6, no positive staining was detected with control IgG (fig. (fig.2e)2e) and Tpbg antibody preabsorbed with the peptide that was used for immunization (online suppl. fig. 1; for all online supplementary material, see www.karger.com/doi/10.1159/000321366).
Quantitative PCR showed that both nephrin and WT1 significantly decreased concomitantly with mesangial cell proliferation (fig. (fig.2j).2j). These results indicate that Tpbg expression is upregulated in podocytes damaged in Thy1 GN. We hypothesized that TGF-β mediates podocyte injury in glomerulonephritis. Quantitative PCR showed that TGF-β expression began to increase at day 3 and peaked at day 6 (fig. (fig.2k2k).
To investigate the function of Tpbg in vitro, we used conditionally immortalized murine podocytes. Quantitative PCR showed that mRNA expression in differentiated cells was significantly decreased compared with undifferentiated cells (fig. (fig.3a).3a). By immunoblot analysis, we detected a 72-kDa protein (Tpbg) in the whole lysate of both differentiated and undifferentiated cells. Protein expression also significantly decreased in differentiated cells (fig. (fig.3b3b).
Immunofluorescence microscopy failed to detect any stress fibers in undifferentiated cells, but Tpbg colocalized with actin in the cell margins of lamellipodia (fig. (fig.3c).3c). In differentiated cells, stress fibers were present and Tpbg was localized to the tips of stress fibers (fig. 3d, e).
We performed colocalization experiments for Tpbg with vinculin, which links integrins to the actin cytoskeleton at focal adhesions; we observed considerable colocalization of Tpbg and vinculin (fig. (fig.3f).3f). Nuclear staining with propidium iodide demonstrated that nuclei were also labeled with Tpbg antibody (fig. (fig.3g).3g). An excess of Tpbg-blocking peptide was able to abolish the cytoplasmic staining, but did not completely block the nuclear staining (fig. (fig.3h,3h, arrows); this result suggests that the cytoplasmic staining is specific and that the nuclear staining might be not specific.
We also investigated whether Tpbg localization requires stress fibers. In the presence of cytochalasin D, which inhibits actin polymerization, Tpbg still colocalized with short F-actin filaments and with vinculin (fig. (fig.3i).3i). This experiment suggests that stress fiber formation is not required for Tpbg localization to focal adhesions.
RhoA and ROCK1, a downstream effector of RhoA, play prominent roles in regulating actin organization and focal adhesions in podocytes. In transient transfection studies, dominant active-RhoA (DA-RhoA) induced stress fibers (fig. (fig.4c),4c), while the negative controls Myc empty vector (fig. (fig.4a)4a) and the dominant negative-RhoA (DN-RhoA; fig. fig.4b)4b) did not induce stress fibers. Further, in undifferentiated podocytes, DA-RhoA induced the formation of focal adhesions containing vinculin and Tpbg (fig. (fig.55).
In the presence of ROCK1 inhibitor Y-27632 (10 μM), differentiated podocytes lost stress fibers concomitantly with the decrease in Tpbg localization at focal adhesions. The re-formation of stress fibers after removal of Y-27632 was also accompanied by Tpbg localization at focal adhesions (fig. (fig.6).6). These results demonstrate that Tpbg localizes to focal adhesions in parallel with the formation of stress fibers and focal adhesions regulated by RhoA.
TGF-β is involved in regulation of the actin cytoskeleton in epithelial cells [16,17]. Following TGF-β treatment for 24 h, expression of Tpbg mRNA (fig. (fig.7a)7a) and protein (fig. (fig.7b)7b) was increased in a dose-dependent manner.
Treatment of differentiated podocytes with TGF-β for 24 h was associated with thicker stress fibers and more prominent cortical actin bundles, together with more prominent Tpbg localization to the tips of both stress fibers and cortical actin filaments (fig. (fig.7c).7c). Quantitative analysis of the intensity of Tpbg staining (fig. (fig.7d)7d) and F-actin staining (fig. (fig.7e)7e) demonstrated that both increased in the presence of TGF-β.
TGF-β functions by binding to the heteromeric complex of serine/threonine kinase, type I receptor (ALK5) and the type II receptors. We confirmed the expression of ALK5 in podocytes (fig. (fig.7f).7f). SB431542, an ALK5 inhibitor, inhibited Tpbg staining intensity (fig. (fig.7g)7g) and the rearrangement of actin filaments (fig. (fig.7h)7h) induced by TGF-β. Furthermore, in the presence of SB431542 alone, the immunofluorescence intensity of Tpbg (fig. (fig.7i)7i) and F-actin (fig. (fig.7j)7j) decreased. Taken together, these results suggest that TGF-β promotes Tpbg expression in differentiated podocytes, with localization to stress fibers and cortical actin.
We next carried out functional studies to directly assess the contribution of Tpbg to organization of the podocyte actin-based cytoskeleton and to podocyte migration. In differentiated podocytes transfected with control plasmid, stress fibers were maintained. In contrast, stress fiber formation was inhibited in podocytes transfected with full-length Tpbg (fig. (fig.8a8a).
To confirm Tpbg knockdown, we demonstrated that podocytes transfected with si-Tpbg expressed less Tpbg mRNA compared to control podocytes transfected with si-scramble (fig. (fig.8b).8b). In the absence of TGF-β, no significant difference in wound closure rate at 24 h was observed between control and Tpbg knockdown podocytes. TGF-β treatment promoted podocyte migration in control cells, but this effect was nearly abrogated in the presence of si-Tpbg (fig. (fig.8c).8c). We confirmed the result of knockdown experiments with wound closure assay by using three Tpbg siRNAs (si-Tpbg-1, −2 and −3; online suppl. fig. 2). We show the representative data in figure figure8.8. These results demonstrate that Tpbg modifies the phenotype of podocytes, including cell motility and actin cytoskeletal arrangement.
The present study had several principal findings. First, Tpbg expression was increased mainly in injured podocytes concurrently with mesangial cell proliferation in Thy1 GN. The expression of podocyte proteins, which are required for maintaining the slit diaphragm structure, decreases in experimental podocyte injury models  and in human nephrotic syndrome . Podocyte injury is a frequent finding in glomerulonephritis [2,4,6,20,21]. Humoral  and/or mechanical [5,22] factors derived from proliferating mesangial cells are thought to damage the podocytes. We focused on the effect of TGF-β as a possible mediator linking glomerular injury and Tpbg expression. As previously reported , we found an increase in the expression of TGF-β mRNA concurrent with mesangial proliferation and increased Tpbg expression. In human carcinomas and embryonic stem cells, Tpbg overexpression is associated with reorganization of the actin cytoskeleton, decreased cell adhesion and increased cell motility [9,10,11,12,13,14,24]. Taken together, these findings suggest that in the setting of glomerulonephritis, TGF-β upregulates Tpbg expression in glomerular cells, particularly podocytes, which leads to alteration in the podocyte cytoskeleton, foot process architecture, adhesion to the glomerular basement and mobility. In vitro experiments using podocytes produced additional findings that lend support to this model.
Second, we found that Tpbg localization into focal adhesions was concurrent with the dynamic reorganization of stress fibers regulated by the RhoA signal [25,26]. Focal adhesions are specialized sites of cell attachment to the extracellular matrix , and are dynamic structures that assemble, disperse and recycle as cells migrate through cross-talk between Rho family GTPase, integrin and numerous adhesion proteins. The foot process architecture of podocytes critically depends on integrin-mediated cell-glomerular basement membrane interaction at focal adhesions. We considered that the tight regulation of focal adhesion assembly and disassembly is essential for the stability of podocyte architecture and that imbalance of focal adhesion proteins causes the impairment of contractile actin filaments. α3-integrin-deficient podocytes are unable to form mature foot processes, suggesting the importance of outside-in signaling . Integrin-linked kinase, a focal adhesion kinase, plays a crucial role in podocyte adhesion, morphology and survival, and activation of integrin-linked kinase activation causes podocyte damage . Cathepsin L, a lysosomal protease, is upregulated in a puromycin aminonucleoside-induced nephrosis model and modulates podocyte migration by interacting with α3-integrin . Tpbg might modulate the regulation of focal adhesions when at low levels under normal conditions. In contrast, Tpbg overexpression and localization under pathological conditions might disassemble focal adhesions, which causes the motile phenotype of podocytes.
Third, we found that TGF-β promoted Tpbg expression and localization at focal adhesions and that Tpbg modified actin organization and motility regulated by TGF-β. TGF-β mediates podocyte injuries in many types of kidney disease [31,32,33]. Foot process effacement, which is observed in injured podocytes, requires a precise interplay of multiple cellular functions including structural alterations of the cytoskeleton, movement of foot processes over the basement membrane and reconstruction of the slit diaphragm . There has been evidence that foot process effacement of podocytes is a migratory event. Lee et al.  reported that TGF-β increases podocyte motility and albumin permeability through the MCP-1/CCR2 pathway. Wei et al.  reported that urokinase receptor signaling is required to activate integrin in podocytes, promoting cell motility and activation of Cdc42 and Rac1. The present data raise the possibility that, in response to increased TGF-β, Tpbg protein accumulates and localizes to focal adhesions, where it alters cytoskeleton function and enhances podocyte motility.
In conclusion, our study demonstrates that in response to TGF-β, Tpbg upregulation at focal adhesions enhances motility and might be involved in the cytoskeleton rearrangement in proliferative glomerulonephritis. Tpbg may be a novel therapeutic target to limit glomerular injury in proliferative glomerulonephritis.
We greatly appreciate the gift of conditionally immortalized murine podocytes from Dr. Peter Mundel (University of Miami). We would like to give our thanks to Naoto Kobayashi (Ehime University School of Medicine) and Akemi Shono (Kyoto University) for technical advice; Hideo Uchiyama (Taigenkai Hospital), Akiko Sakurai (Tokushima University Hospital), and Michael Hann (Florida International University) for technical assistance.
This study was supported by the Kidney Foundation, Japan (JKFB09-41), Mitsui Sumitomo Welfare Foundation and Grant-in-Aid for Scientific Research (19590973) and the NIDDK Intramural Research Program.