Glycogen synthase kinase‑3β is required for epithelial‑mesenchymal transition and barrier dysfunction in mouse podocytes under high glucose conditions.

Epithelial-mesenchymal transition (EMT) is important for diabetic nephropathy (DN). Podocytes are specialized epithelial cells, which form a major component of the glomerular filtration barrier. Podocyte damage has been suggested to be the primary mechanism behind the albuminuria associated with DN. The present study aimed to determine the function of glycogen synthase kinase (GSK)‑3β in EMT and barrier dysfunction of mouse podocytes exposed to high glucose (HG) conditions. Matured and differentiated podocytes were treated with normal glucose (NG), HG or NG + mannitol. Podocytes were also transfected with a small interfering RNA (siRNA) against GSK‑3β or a scrambled siRNA, or were treated with lithium chloride (LiCl), a GSK‑3β inhibitor, under NG or HG conditions. The expression levels of the epithelial cell markers, nephrin and podocin, and the myofibroblast cell markers, α‑smooth muscle actin (SMA) and fibronectin, in podocytes by western blot analysis and immunofluorescence staining, respectively. The monolayer barrier function was assessed by albumin inflow. The phosphorylation and activity levels of GSK‑3β were also quantified. It was observed that HG promotes EMT in podocytes, due to the increased levels of podocin and nephrin expression and the reduced α‑SMA and fibronectin expression levels. HG also induced barrier dysfunction and increased the expression level of total GSK‑3β, Try216‑phosphorylated‑GSK‑3β and the GSK‑3β activity in podocytes. Transfection of GSK‑3β siRNA or treatment with LiCl reversed the HG‑induced EMT and barrier dysfunction in podocytes. In conclusion, the present study determined that GSK‑3β is required for EMT and barrier dysfunction in podocytes under HG conditions; therefore, GSK‑3β may be a novel target for the treatment of DN.

Introduction

Diabetic nephropathy (DN) is one of the most frequent microvascular complications in patients with diabetes and often leads to end-stage renal disease (1). Albuminuria is a common clinical manifestation of DN and may be associated with kidney disease and its progression (2). Routine therapeutic agents such as angiotensin II receptor blockers and angiotensin-converting enzyme inhibitors reduce urine protein levels and confer limited benefits for the renal function of patients with DN (3,4). However, these therapeutic agents are not sufficient for the prevention of kidney damage. Therefore, it is imperative to identify novel therapeutic targets for DN.

Podocytes are specialized epithelial cells, which are key glomerular endothelial cells and form a major component of the glomerular filtration barrier. Structural and functional alterations in these cells are crucial for the development of albuminuria (5). Epithelial-mesenchymal transition (EMT) is a biological characteristic of epithelial cells under physiological and pathological conditions, which is important during DN. Podocytes undergoing EMT lose the phenotypic characteristics of epithelial cells, including reduced P-cadherin, nephrin and podocin expression levels and express phenotypic markers of mesenchymal cells, including α-smooth muscle actin (α-SMA) and fibronectin (6). However, the mechanisms underlying the EMT of podocytes under high glucose (HG) conditions remain to be elucidated.

Glycogen synthase kinase-3β (GSK-3β), a serine (Ser)/threonine (Thr) kinase, is ubiquitously expressed in eukaryotic cells and is hypothesized to act on signal transduction proteins, structural proteins and transcription factors to regulate cell differentiation, proliferation and apoptosis (7). GSK-3β has two phosphorylation sites, one is a Ser9 inhibition site and the other a tyrosine (Tyr)216 activation site (8). GSK-3β is also suggested to participate in the Wnt/β-catenin pathway, which is important for podocyte EMT (9).

The objective of the present study was to investigate the function of GSK-3β in podocyte EMT and barrier dysfunction under HG conditions in order to identify a novel therapeutic target for DN. Podocytes were transfected with GSK-3β small interfering RNA (siRNA) or treated with lithium chloride (LiCl), a selective inhibitor of GSK-3β (7,10), to inhibit the GSK-3β expression and activity. The alterations in the phenotypic characteristics and barrier function of podocytes following treatment were observed. The results of the present study indicate that GSK-3β is required for HG-induced EMT and barrier dysfunction in podocytes, implying GSK-3β is a novel potentially therapeutic target for the treatment of DN.

Materials and methods

Podocyte cell culture and transfection

Conditionally immortalized mouse podocytes were provided by Professor Nie Jing (Southern Medical University, Guangzhou, China). Undifferentiated podocytes were cultured and differentiated in Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) containing 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.), 5.6 mM glucose (Dingguo Changsheng Biotechnology Co., Ltd., Beijing, China) and 10 U/ml recombinant mouse interferon-γ (Shanghai Sangon Biotech Co., Ltd., Shanghai, China), in an incubator at 33°C in 5% CO2. Following differentiation, podocytes were cultured at 37°C in RPMI 1640 medium without recombinant mouse interferon-γ. They were cultured for 12–14 days with the medium replaced every 1–2 days until they were mature and differentiated.

Matured and differentiated podocytes were seeded onto 6-well or Transwell plates. When 80–90% confluence was reached, podocytes were supplied with serum-free RPMI 1640 medium for another 6–8 h for synchronization. Thereafter, cells were treated as follows: i) Normal glucose (NG, 5.6 mM glucose); ii) HG (12.5 HG, 12.5 mM; 25 HG, 25 mM; and 50 HG, 50 mM glucose); and iii) mannitol as an osmotic control [NG + M, 5.6 mM glucose and 44.4 mM mannitol (Dingguo Changsheng Biotechnology Co., Ltd.), which has an osmotic pressure comparable with the 25 HG group]. Podocytes were also transfected with a siRNA, sequence: 5′-CCACTCAAGAACTGTCAAGTA-3′ (GeneChem Co., Ltd., Shanghai, China) against GSK-3β based on the GSK-3β full-length mouse gene (GenBank accession number: NM-019827.6) or a scrambled siRNA (5′-UUCUCCGAACGUGUCACGUTT-3′; GeneChem Co., Ltd.) when treated NG or HG (25 mM). Podocytes were transfected with 30 nM GSK-3β siRNA for 36 h using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) transfection reagent according to the manufacturer's protocol (11). Podocytes were grouped in the following treatments: i) NG; ii) NG + GSK-3β scrambled siRNA; iii) NG + GSK-3β siRNA; iv) HG (25 mM glucose); v) HG + GSK-3β scrambled siRNA; and vi) HG + GSK-3β siRNA. Additionally, podocytes were treated with water-soluble LiCl (Dingguo Changsheng Biotechnology Co., Ltd.) at a final concentration of 10 mM under NG and HG conditions. Podocytes were grouped as follows: i) NG (NG, 5.6 mM glucose); ii) NG + LiCl (10 mM); iii) HG; and iv) HG + LiCl (10 mM). After 36 h the podocytes were collected by centrifugation at 13,400 g for 5 min at room temperature in order to be used for various assays.

Immunoblotting

Immunoblotting was performed as previously described (12). The primary antibodies used were as follows: β-actin (cat. no. TA-09; 1:1,000; OriGene Technologies, Inc., Beijing, China), nephrin(cat. no. ab58968; 1:1,000), podocin (cat. no. ab50993; 1:1,000), α-SMA (cat. no. ab7817; 1:2,000), fibronectin (cat. no. ab2413; 1:2,000), GSK-3β (cat. no. ab32391; 1:2,000), phosphorylated (p)-Tyr216-GSK-3β (cat. no. ab75745; 1:2,000; Abcam, Cambridge, UK), p-Ser9-GSK-3β (cat. no. 5558, 1:1,000; Cell Signaling Technology, Inc., Boston, USA). Membranes were incubated with primary antibodies at 4°C overnight. Subsequently, the membranes were washed phosphate-buffered saline Tween-20 and the alkaline phosphatase-conjugated secondary antibody (cat. no. IA-0082; 1:2,000; Dingguo Changsheng Biotechnology Co., Ltd.) was incubated at 37°C for 2 h. The blots were developed using a 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium color development kit (Boster Biological Technology, Ltd., Wuhan, China). ImageJ version 2.1.4.7 software (National Institutes of Health, Bethesda, MD, USA) was used for quantitative analysis of the relative grayscale intensity of each protein expression band.

Indirect immunofluorescence staining

Immunofluorescence staining was performed as previously described (13). Mature and differentiated podocytes seeded on coverslips were serum-starved for 8 h, subsequently, graded concentrations of glucose were added (12.5 HG, 12.5 mM; 25 HG, 25 mM; and 50 HG, 50 mM glucose). Following 24 h exposure to the different HG concentrations, the cells were fixed in 4% paraformaldehyde, then incubated with the primary antibodies for nephrin, podocin, α-SMA and fibronectin. The cells were then exposed to the Alexa Fluor 488 secondary antibody (Invitrogen; Thermo Fisher Scientific, Inc.). The coverslips were mounted with antifade mounting medium (Beyotime Institute of Biotechnology, Shanghai, China) and then were observed and photographed using an inverted fluorescence microscope (Olympus Corporation, Tokyo, Japan). The immunofluorescence intensity was measured with Image J software. The expression levels of nephrin, podocin, α-SMA and fibronectin were calculated as the immunofluorescence intensity normalized to that of nephrin under NG conditions.

Detection of monolayer barrier function in podocytes

Podocyte monolayer barrier function was measured using the surrogate measure, podocyte permeability. A modification of a previously described protocol (14) was adopted for the assessment of podocyte permeability, where albumin influx was used as an indicator (14). Differentiated podocytes at a density of 4×105 cells/well were plated onto 12-well Transwell plates (3 µm pore; Corning, Corning, NY, USA) and were serum-starved overnight. Upon reaching 70–80% confluence, the cells were exposed to different concentrations of glucose (NG; HG, 12.5 mM; HG, 25 mM and HG, 50 mM). The cells were then washed twice with phosphate-buffered saline supplemented with magnesium chloride and calcium chloride (both 1 mM). The upper compartment was refilled with 0.25 ml RPMI 1640 medium and the lower compartment was refilled with 0.5 ml RPMI 1640 supplemented with 40 mg/ml FBS. Next, the cells were incubated at 37°C for 2 h. The total albumin influx was determined by quantifying the concentration in the upper compartment using a bicinchoninic acid protein assay kit (Dingguo Changsheng Biotechnology Co., Ltd.).

Detection of GSK-3β activity

This assay was conducted using a GSK-3β activity assay kit (Genmed Scientifics USA, Inc., Shanghai, China) according to the manufacturer's protocol. The optical density of each treatment group was detected using a Nanodrop 2000 spectrophotometer at 280 nm (Thermo Fisher Scientific, Inc.) and GSK-3β activity was calculated in accordance with the formula provided by the manufacturer of the kit.

Statistical analysis

Statistical analysis was performed using SPSS software, version 17.0 (SPSS, Inc., Chicago, IL, USA). Data are presented as the arithmetic mean ± standard error. Differences between groups were evaluated using one-way analysis of variance followed by a Student-Newman-Keuls post-hoc test. P<0.05 was considered to indicate a statistically significant difference.

Results

Phenotypic conversion of podocytes exposed to HG conditions

To verify the trans-differentiation of podocytes into mesenchymal cells during HG conditions, podocytes were cultured for 36 h with different concentrations of glucose. Western blot analysis (Fig. 1A and B) determined that the expression levels of the epithelial cell markers nephrin and podocin were significantly reduced when exposed to increased HG concentrations when compared with the NG control group (P<0.05; Fig. 1A). By contrast, expression of the myofibroblast cell marker α-SMA was significantly increased in cells treated with increased HG concentrations compared with the NG control group (P<0.05; Fig. 1B). No significant difference was observed between the NG and the NG + M groups for all of the epithelial cell markers investigated.

Figure 1.

HG-induced phenotypic conversion and barrier dysfunction of podocytes. Western blot analysis determined that the protein expression levels of (A) nephrin and podocin were significantly reduced in the HG group.*P<0.05 vs. podocin expression in NG group, §P<0.05 vs. nephrin expression in NG group; n=4. (B) Protein expression of α-SMA was increased with increased glucose concentrations. *P<0.05 vs. NG group; n=4. Immunofluorescence staining presented reduced expression levels of (C) nephrin and podocin in cells treated with HG. (D) Increased expression of α-SMA and fibronectin was identified in the HG group compared with the NG group. Fluorescence indicated the protein expression of the relevant proteins. Scale bar, 20 µm. (E) Protein expression levels of nephrin, podocin, α-SMA and fibronectin from the immunofluorescence intensity normalized to that of nephrin in the NG group. *P<0.05 vs. podocin expression in the NG group, &P<0.05 vs. nephrin expression in the NG group, §P<0.05 vs. α-SMA expression in NG group, ##P<0.01 vs. fibronectin expression in NG group; n=4. (F) Monolayer permeability of differentiated and mature podocytes under HG conditions. Quantitative analysis determined that HG increased albumin inflow compared with NG treatment. *P<0.05 vs. NG group; **P<0.01 vs. NG group; n=4. HG, high glucose; NG, normal glucose; α-SMA, α-smooth muscle actin; NG + M, NG + mannitol.

Immunofluorescence analysis (Fig. 1C-E) of podocytes following exposure to 25 mM D-glucose for 36 h revealed significantly reduced nephrin and podocin expression (P<0.05; Fig. 1C and E); however, α-SMA and fibronectin expression levels were significantly increased (P<0.05; Fig. 1D and E) compared with that observed in the NG control group.

Monolayer barrier dysfunction of podocytes exposed to HG conditions

The Transwell chamber assay was conducted to determine whether HG treatment resulted in barrier dysfunction in podocytes. Fig. 1F indicated that there was a significantly increased albumin inflow in the presence of 25 mM D-glucose compared with that in the NG control group (5.6 mM D-glucose) group (P<0.05). In addition, increased albumin inflow was observed in the 50 mM HG group compared with the 12.5 mM HG group (P<0.01; Fig. 1F). These observations suggested that monolayer barrier dysfunction occurs in podocytes following HG treatment.

HG-induced GSK-3β expression and activity in podocytes

GSK-3β expression, phosphorylation levels and activity were evaluated in podocytes. Increased HG concentrations significantly induced total GSK-3β expression (P<0.05; Fig. 2A) and pTry216-GSK-3β; however, reduced p-Ser9GSK-3β expression was observed with increased HG concentrations (P<0.05; Fig. 2B). GSK-3β activity was significantly increased when compared with the NG control group (P<0.05; Fig. 2C).

Figure 2.

HG conditions increased GSK-3β expression, phosphorylation and activity in podocytes. (A) GSK-3β expression, (B) phosphorylation and (C) activity were significantly increased in the HG group compared with NG group. *P<0.05 vs. NG group; n=3. p-Try216-GSK-3β expression was significantly increased; however, p-Ser9-GSK-3β expression was reduced in the HG group. *P<0.05 vs. p-Try216-GSK-3β expression in NG group. &P<0.05 vs. p-Ser9-GSK-3β expression in NG group; n=3. HG, high glucose; GSK-3β, glycogen synthase kinase-3β; NG, normal glucose; p, phosphorylated; Tyr, tyrosine; Ser, serine; NG + M, NG + mannitol; NADH, nicotinamide adenine dinucleotide.

GSK-3β siRNA reversed podocyte EMT and monolayer barrier dysfunction under HG conditions

To confirm whether GSK-3β participated in the trans-differentiation and barrier dysfunction of podocytes during HG conditions, the cells were transfected with GSK-3β or scrambled siRNA. Western blotting in Fig. 3A demonstrated that siRNA-mediated inhibition of the expression of GSK-3β had occurred. When GSK-3β siRNA-transfected podocytes were exposed to HG (25 mM), the expression levels of nephrin and podocin were significantly increased compared with the scrambled siRNA control group (P<0.05; Fig. 3B). The expression levels of the mesenchymal protein α-SMA were significantly reduced (P<0.05; Fig. 3C) compared with those transfected with scrambled siRNA. In addition, the albumin inflow was significantly reduced in GSK-3β siRNA-transfected podocytes compared with scrambled siRNA-transfected group (P<0.05; Fig. 3D), indicating that the monolayer barrier function of podocytes was improved.

Figure 3.

GSK-3β siRNA reversed the podocyte epithelial-mesenchymal transition and monolayer barrier dysfunction under HG conditions. (A) GSK-3β siRNA suppressed the expression of GSK-3β in podocytes. *P<0.05 vs. scrambled siRNA 36 h group. **P<0.05 vs. scrambled siRNA 36 h group; n=4). Podocytes transfected with GSK-3β siRNA expressed higher levels of (B) nephrin and podocin (*P<0.05 vs. podocin expression in NG group, §P<0.05 vs. nephrin expression in NG group, #P<0.05 vs. podocin expression in HG group, &P<0.05 vs. nephrin expression in HG group), lower levels of (C) α-SMA (§P<0.05 vs. α-SMA expression in NG group, *P<0.05 vs. α-SMA expression in HG group) and (D) reduced the albumin inflow of podocytes under HG conditions compared with podocytes transfected with scrambled siRNA and the NG group. **P<0.05 vs. NG group, *P<0.05 vs. HG group; n=4. GSK-3β, glycogen synthase kinase-3β; siRNA, small interfering RNA; HG, high glucose; NG, normal glucose; α-SMA, α-smooth muscle actin.

LiCl reversed podocyte EMT and monolayer barrier dysfunction under HG conditions

In addition to siRNA-mediated downregulation of GSK-3β expression levels, the GSK-3β inhibitor, LiCl was used to inhibit GSK-3β activity. Following GSK-3β inhibition, nephrin and podocin expression levels were significantly increased in podocytes under HG conditions when compared with the HG only group (P<0.05; Fig. 4A), whereas α-SMA and fibronectin expression levels were significantly reduced when compared with the HG only group (P<0.05, Fig. 4B and C). LiCl was significantly associated with the improvement in monolayer barrier function in podocytes exposed to HG, which was reflected by the significant reduction of the albumin inflow when compared with HG only group (P<0.05; Fig. 4D).

Figure 4.

LiCl reversed the podocyte epithelial-mesenchymal transition and monolayer barrier dysfunction under HG conditions. (A) The HG + LiCl group had increased nephrin and podocin expression compared with the HG group. *P<0.05 vs. podocin expression in the NG group; §P<0.05 vs. nephrin expression in the NG group, #P<0.05 vs. podocin expression in the HG group, &P<0.05 vs. nephrin expression in the HG group; n=4. (B) LiCl treatment also reduced α-SMA expression levels. §P<0.05 vs. NG group, *P<0.05 vs. HG group; n=4. (C) Immunofluorescence staining determined that fibronectin expression in the HG + LiCl group was lower compared with the HG group. Scale bar, 40 µm. (D) Albumin inflow in podocytes in the HG + LiCl group was reduced compared with the HG only group. **P<0.05 vs. the NG group, *P<0.05 vs. the HG group; n=4. LiCl, lithium chloride; HG, high glucose; NG, normal glucose; α-SMA, α-smooth muscle actin.

Discussion

Podocyte damage is closely associated with the development of albuminuria and is an important factor in the occurrence and development of various kidney diseases (15). Previous studies have determined that exfoliation and apoptosis of podocytes, and the fusion and disappearance of podocyte processes contribute to the development of DN (16–18). A previous study indicated that podocyte exfoliation and apoptosis occurred following the development of albuminuria (19). Consequently, podocyte loss may be unlikely to initiate a process during DN, which may lead to albuminuria (19). Therefore, earlier cellular events may be involved. It has been proposed that stimulation of EMT in podocytes is a reversible process that occurs prior to exfoliation and apoptosis (9). Phenotypic trans-differentiation of podocytes may lead to disordered cell function, ultimately result in albuminuria.

Nephrin and podocin have been observed to interact with other glomerular slit diaphragm components, including CD2-associated protein and cytoskeleton-associated proteins (for example, zonula occluden 1 and actin). These proteins are important for maintaining podocyte integrity and preserving the normal function of the glomerular slit diaphragm. A previous study has reported that reduced expression of both nephrin and podocin is closely associated with the occurrence of albuminuria (20). Fibronectin and α-SMA are phenotypic markers of mesenchymal cells and increased expression of α-SMA in non-smooth muscle cells has been established as an important basis for cell activation and trans-differentiation (21). The increased expression levels of these proteins act as potential markers of the mesenchymal trans-differentiation process.

The results of the present study demonstrated that HG conditions reduced nephrin and podocin expression, whereas α-SMA and fibronectin expression was increased in cultured mouse podocytes, indicating that HG induced EMT in podocytes. These results were consistent with previous studies (22,23). In addition, the current study demonstrated glucose concentration-dependent impairment of the barrier function of podocytes using Transwell experiments. These observations suggest that high glucose-induced EMT led to the loss of functional proteins, which in turn damaged the integrity of the slit diaphragm and led to abnormal glomerular filtration. This process may represent an important mechanism in the development of albuminuria. However, these results require validation in vivo. Further research is additionally required to investigate whether it is possible to delay or minimize the effects of EMT in order to suppress the progress of DN.

GSK-3β is a Ser/Thr adenosine 5′-monophosphate-activated protein kinase with highly conservative sequence. GSK-3β may act as the primary regulatory enzyme of numerous cellular signal transduction channels and has been identified to influence cell growth and apoptosis (24). The influence of HG on GSK-3β expression and activity in podocytes remains unclear. A previous study reported that the activity of GSK-3β is increased in HG conditions, resulting in the trans-differentiation of renal tubular epithelial cells. Increased GSK-3β activity has also been reported to be associated with HG-induced apoptosis of renal mesangial cells (25). The results of the current study indicated that HG induced GSK-3β activity, which was associated with EMT and barrier dysfunction in podocytes. In addition, HG-induced phosphorylation of GSK-3β at the Tyr216 activation site was observed in podocytes, which was consistent with the study performed by Paeng et al (26). Therefore, this indicated the importance of the expression, phosphorylation and activity of GSK-3β in EMT and barrier dysfunction under HG conditions.

The importance of GSK-3β in DN remains to be fully elucidated. Paeng et al (26) reported that enhanced GSK-3β activity within podocytes under HG conditions was associated with podocyte apoptosis, which has been suggested to be crucial for albuminuria and progression of DN (26). Shang et al (27) reported that sulforaphane partially ameliorated experimental DN by the inhibition of the GSK3β signaling pathway (27). These results implied that GSK-3β may be involved in DN. However, Mariappan et al (28) determined that the activation of GSK3β by sodium nitroprusside (SNP) reduced the HG-induced laminin increase (in contrast to the present study), an important characteristic of DN progression in kidney-proximal tubular epithelial cells. In addition, diabetes led to the inactivation of GSK3β by activation of the Src proto-oncogene, pyruvate kinase 2, protein kinase B and extracellular signal-related kinase. Furthermore, GSK3β activation by SNP mitigated kidney injury induced by diabetes in vivo (28), indicating that GSK3β may suppress the progress of DN. To confirm the importance of GSK-3β in HG-induced EMT and barrier dysfunction in podocytes, GSK-3β siRNA was transfected in podocytes in order to downregulate GSK-3β expression. LiCl was also used to inhibit GSK-3β activity under NG and HG conditions. The results of the present study indicated that GSK-3β expression and activity were essential for HG-induced EMT and barrier dysfunction. As podocyte EMT and barrier dysfunction are important for DN, the current results confirm the involvement of GSK-3β in the pathogenesis and development of DN in vitro.

A previous study on mouse podocytes reported that treatment with a GSK-3β inhibitor and GSK-3β siRNA reduced β-catenin and Snail expression levels and reversed HG-induced upregulation of α-SMA expression levels; however, nephrin expression levels were downregulated (11). The present study used further markers of EMT, including the epithelial cell markers nephrin and podocin, and the myofibroblast cell markers α-SMA and fibronectin in order to confirm the occurrence of EMT in podocytes under HG conditions. In addition to GSK-3β expression, the importance of GSK-3β activity and phosphorylation in HG-induced EMT was investigated in the present study. Furthermore, the effect of GSK-3β expression and activity was observed on the barrier dysfunction of podocytes exposed to HG conditions. Therefore, using the recent study as the basis (11), the present study clarified the importance of GSK-3β in DN progression by using in vitro experiments.

In conclusion, the present study determined that GSK-3β participated in HG-induced EMT and barrier dysfunction in podocytes, which may suggest that GSK-3β is a potential therapeutic target for novel DN treatment. Further experiments using an animal of model of DN are required in order to characterize the effects of GSK-3β expression and activity on podocyte EMT and function and to fully determine the function of GSK-3β in the development of albuminuria in patients with DN.