Prevention of apoptosis averts glomerular tubular disconnection and podocyte loss in proteinuric kidney disease

Abstract

There is a great need for treatment that arrests progression of chronic kidney disease. Increased albumin in urine leads to apoptosis and fibrosis of podocytes and tubular cells and is a major cause of functional deterioration. There have been many attempts to target fibrosis, but because of the lack of appropriate agents, few have targeted apoptosis. Our group has described an ouabain-activated Na,K-ATPase/IP3R signalosome, which protects from apoptosis. Here we show that albumin uptake in primary rat renal epithelial cells is accompanied by a time- and dose-dependent mitochondrial accumulation of the apoptotic factor Bax, down-regulation of the antiapoptotic factor Bcl-xL and mitochondrial membrane depolarization. Ouabain opposes these effects and protects from apoptosis in albumin-exposed proximal tubule cells and podocytes. The efficacy of ouabain as an antiapoptotic and kidney-protective therapeutic tool was then tested in rats with passive Heymann nephritis, a model of proteinuric chronic kidney disease. Chronic ouabain treatment preserved renal function, protected from renal cortical apoptosis, up-regulated Bax, down-regulated Bcl-xL, and rescued from glomerular tubular disconnection and podocyte loss. Thus we have identified a novel clinically feasible therapeutic tool, which has the potential to protect from apoptosis and rescue from loss of functional tissue in chronic proteinuric kidney disease.

Keywords: a-tubular glomeruli; albuminuria; apoptosis; cell signaling; chronic kidney disease; ouabain; podocyte; proximal tubule; sodium potassium adenosine triphosphatase.

Figure 1.. Albumin uptake into primary renal cells triggers apoptosis followed by increased expression of TGF-beta 1. Protective effect of ouabain.

A) Albumin-induced apoptosis in RPTC incubated with 10 mg/mL albumin for 2, 4 or 8 hours and vehicle control. RPTC were TUNEL-stained (red) to detect apoptotic cells and counterstained with DAPI (blue). Scale bar for all images = 40μm. B) Quantification of albumin-induced apoptosis in RPTC treated with 10 mg/mL albumin (open bars) or vehicle (grey bars) for 2, 4, 8 or 18 hours. Histograms show mean ± SEM. Experiments were repeated four times. The AI in vehicle-treated (control) cells was stable at 2% for the duration of the 18 hours. AI for 10mg/mL albumin was significantly increased vs control *p<0.001, ^#p<0.01. C) Albumin-induced TGF-beta 1 expression in RPTC detected by immunoblot. RPTC were incubated with 10 mg/ml albumin in serum-free medium for 2, 4 or 8 hours. D) Quantification of albumin-induced TGF-beta 1 expression in RPTC incubated with 10 mg/mL albumin (open bars) or vehicle (grey bars) for 2, 4, 8 or 18 hours. Experiments were repeated four times and vehicle control for each time point was set to 100%. Histograms show mean ± SEM, *p<0.001 vs control. E, F) Dose dependence of albumin-induced apoptosis in RPTC. RPTC were incubated with 0 (vehicle control), 5, 10 or 20 mg/mL albumin alone (solid line), and in combination with 5 nM ouabain (dashed line) for 8 hours in E) or 18 hours in F). Plots represent mean ± SEM. Experiments were repeated four times. p<0.001 for 5, 10 and 20 mg/mL albumin vs vehicle control (0 mg/mL), at both 8 and 18 hours. p<0.01 for 5nM ouabain with 5, 10 and 20 mg/mL albumin vs the corresponding albumin treatment alone, at both 8 and 18 hours. G) Primary culture of podocytes; images show isolated rat glomeruli plated and cultured for three days. Staining with the podocyte specific transcriptional factor WT1 shown in green and DAPI stained nuclei shown in blue. Podocytes that migrate out from the glomerulus were used for the apoptosis study shown in H). Note that a subset of cells (indicated by arrow) are not WT1 positive and therefore not included in the analysis. All scale bars = 20μm. H) Dose dependence of albumin-induced apoptosis in podocytes. Podocytes were incubated with 0 (vehicle control), 1, 5 or 10 mg/ml albumin alone (solid line), and in combination with 5 nM ouabain (dashed line) for 18 hours. Only cells positive for the podocyte marker WT1 was included in the analysis. Plots represent mean ± SEM. Experiments were repeated four times. p<0.001 for 5 and 10 mg/mL albumin vs vehicle control (0 mg/mL), p<0.01 for ouabain treatment together with 5 and 10 mg/mL albumin vs the corresponding albumin treatment alone.

Figure 2.. Ouabain intervenes with the intrinsic apoptotic pathway triggered by albumin, but not with LPS stimulated cytokine expression.

A) Cartoon illustrating the ouabain/Na,K-ATPase/IP3R signaling pathway. Ouabain binds to Na/K-ATPase which triggers interaction between the Na/K-ATPase and the IP3R through the amino acid residues LKK in the N-terminus of the catalytic α subunit of the Na/K-ATPase. The interaction activates the IP3R causing calcium to release from the ER. The slow calcium oscillations promote activation the NF-κB p65 subunit which leads to protection from apoptosis through stimulation of the anti-apoptotic protein Bcl-xL. B) Western blot showing expression of Bax, Bcl-xL, caspase-3 and cleaved caspase-3 in RPTC after 8 hours incubation with 10 mg/mL albumin, 5 nM ouabain, 1 μM Helenalin and 1μM CPA as indicated. C) Densitometric quantification of Bcl-xL (left), Bax (center) and cleaved caspase-3 (right) after 8 hours incubation with 10 mg/mL albumin, 5 nM ouabain, 1 μM Helenalin and 1μM CPA as indicated. The density of the band from control cells was set to 100%. Histograms show the mean ± SEM. Experiments were repeated five times. *p<0.001, ^#p<0.01. D) Expression of the inflammatory cytokines, IL-1β and IL-6, after incubation with LPS 1 μg/mL for 0, 1, 2, 4 or 8 hours, and for 8 hours together with 5 nM ouabain as indicated.

Figure 3.. Ouabain protects from albumin triggered mitochondrial dysfunction.

A) Changes in mitochondrial membrane potential in RPTC incubated with vehicle, 2.5 mg/ml albumin or 2.5 mg/ml albumin and 5nM of ouabain for 8h, visualized using JC-1 dye. Aggregation of JC-1 indicate high-Δψ mitochondria in red fluorescence. Monomeric dye indicate low-Δψ mitochondria in green fluorescence. Mitochondrial depolarization is shown as a decrease in the red/green fluorescence intensity ratio. All scale bars = 10μm. B) Localization of Bax to mitochondria in RPTC incubated with vehicle, 2.5 mg/ml albumin or 2.5 mg/ml albumin and 5nM of ouabain for 8h. Representative confocal images of Bax immunofluorescence staining shown in red and mitochondria in green, by using the mitochondrial marker BacMam 2.0. All scale bars = 5μm. C) Bcl-xL expression in RPTC incubated with vehicle, 2.5 mg/ml albumin or 2.5 mg/ml albumin and 5nM of ouabain for 8h. Representative confocal images of Bcl-xL immunofluorescence signal in red and mitochondria in green, by using the mitochondrial marker BacMam 2.0. All scale bars = 5μm. D) Quantification mitochondrial membrane potential change in RPTC incubated with vehicle (control) or 2.5 mg/ml albumin in the presence and absence of 5nM ouabain for 8h. *p<0.001. Data are shown as % of control, mean ± SEM. Experiments were repeated three times. E) Quantification of Bax localization to mitochondria in RPTC incubated with vehicle (control) or 2.5 mg/ml albumin in the presence and absence of 5nM ouabain for 8h. Co-localization was assessed along two perpendicular line traces across the nucleus. Overlap of mitochondria and Bax fluorescence signal peaks along the lines were analyzed. Peaks were considered to overlap if spaced by no more than 140 nm. *p<0.001. Data are shown as % overlapping Bax/mitochondrial peaks, mean ± SEM. Experiments were repeated four times. F) Quantification of Bcl-xL expression in RPTC incubated with vehicle (control) or 2.5 mg/ml albumin in the presence and absence of 5nM ouabain for 8h. ^#p<0.01. All data are shown as % Bcl-xL immune reactivity when control was set to 100%, mean ± SEM. Experiments were repeated four times. G) Plot shows time dependent change as % of control for mitochondrial membrane potential (solid line), localization of Bax to mitochondria (dashed line) and Bcl-xL expression (dotted line) in response to 2.5mg/ml albumin. H, I) Quantification of mitochondrial membrane potential change in RPTC incubated with vehicle (control) or 0.2 mg/ml albumin in presence and absence of 5nM of ouabain for 8 hours H) and 18 hours I). Data are shown as % of control, mean ± SEM ^#p<0.05. Experiments were repeated three times. For all experiments Mann-Whitney U test was used to determine whether differences were statistically significant.

Figure 4.. Long-term treatment with ouabain attenuates apoptosis of renal cortical cells in the PHN rat

A) Representative DAPI staining of nuclei in renal cortex of control rats, PHN rats and ouabain-treated PHN rats at four months after PHN induction. The arrows indicate typical condensed nuclei. For quantification control was set to 100%. All scale bars = 40μm. B) Representative immunostaining for Bcl-xL in renal cortex of control rats, PHN rats and ouabain-treated PHN rats at four months after PHN induction. For semi-quantitative evaluation of Bcl-xL immunoreactivity signal, control was set to 100%. All scale bars = 40μm. C) Representative immunostaining for Bax in renal cortex of control rats, PHN rats and ouabain-treated PHN rats at four months after PHN induction. For semi-quantitative evaluation of Bax immunoreactivity signal, control was set to 100%. All scale bars = 40μm. For all experiments, analysis was done in two sections from each kidney and in five (condensed nuclei) or six (Bcl-xl and Bax) randomly selected areas of the outer cortex. Histograms show the mean ± SEM. Statistical analysis was performed using ANOVA followed by t-test. *p<0.05

Figure 5.. Kidneys from ouabain-treated PHN rats have fewer disconnected proximal tubules than kidneys from vehicle-treated PHN rats.

A) Representative TUNEL staining of early proximal tubules from vehicle- and ouabain-treated PHN rats at four months after PHN induction. Control rats rarely display any positive TUNEL stain in this region. Histograms show quantitative determination of apoptotic PTC, shown as mean ± SEM. *p<0.01. B) Typical PAS staining of slices for morphometric studies, evaluating the glomerular-tubular connections in individual glomeruli. Pictures illustrate the pattern of glomerular-tubular connections, each glomerulus in the slide is counted (left) and evaluated as connected to a normal proximal tubule (middle left), to an atrophic proximal tubule (middle right), or without a tubular connection, i.e a-tubular glomeruli (right). C) Summary of morphometric studies. Quantitative determination of glomeruli connected to an atrophic proximal tubule (left) and a-tubular glomeruli (right). Histograms represent mean ± SEM. *p<0.001, ^#p<0.01 For all experiments Mann-Whitney U test was used to determine significant differences.

Figure 6.. Kidneys from ouabain-treated PHN rats have more preserved WT1 positive glomerular cells than kidneys from vehicle-treated PHN rats.

A) Representative micrographs showing TUNEL stain in WT1 positive glomerular cells from control rats, vehicle-treated and ouabain-treated PHN rats. Sections were TUNEL-stained (red) to detect apoptotic cells and stained for WT1 (green) to detect podocytes, DAPI stained nuclei are shown in blue. All scale bars = 20μm. Histogram shows quantification of apoptotic podocytes, for control rats, PHN rats and ouabain-treated PHN rats. Histograms show mean ± SEM. B) Histogram shows quantification of WT1 positive cells per glomerulus for control rats, PHN rats and ouabain-treated PHN rats. Histograms show mean ± SEM. For all experiments Mann-Whitney U test was used to determine significant differences *p<0.01, ^#p<0.05