Krüppel-like factor 6 regulates mitochondrial function in the kidney

Abstract

Maintenance of mitochondrial structure and function is critical for preventing podocyte apoptosis and eventual glomerulosclerosis in the kidney; however, the transcription factors that regulate mitochondrial function in podocyte injury remain to be identified. Here, we identified Krüppel-like factor 6 (KLF6), a zinc finger domain transcription factor, as an essential regulator of mitochondrial function in podocyte apoptosis. We observed that podocyte-specific deletion of Klf6 increased the susceptibility of a resistant mouse strain to adriamycin-induced (ADR-induced) focal segmental glomerulosclerosis (FSGS). KLF6 expression was induced early in response to ADR in mice and cultured human podocytes, and prevented mitochondrial dysfunction and activation of intrinsic apoptotic pathways in these podocytes. Promoter analysis and chromatin immunoprecipitation studies revealed that putative KLF6 transcriptional binding sites are present in the promoter of the mitochondrial cytochrome c oxidase assembly gene (SCO2), which is critical for preventing cytochrome c release and activation of the intrinsic apoptotic pathway. Additionally, KLF6 expression was reduced in podocytes from HIV-1 transgenic mice as well as in renal biopsies from patients with HIV-associated nephropathy (HIVAN) and FSGS. Together, these findings indicate that KLF6-dependent regulation of the cytochrome c oxidase assembly gene is critical for maintaining mitochondrial function and preventing podocyte apoptosis.

Figure 11. Reduced KLF6 expression in human HIVAN and FSGS.

(A) Immunostaining for KLF6 performed on healthy donor nephrectomy specimens shows a nuclear distribution in podocytes and parietal cells, with a nuclear and cytosolic distribution in tubular cells. By immunohistochemistry, KLF6 expression in the podocytes (arrows) is shown in the biopsy specimens from healthy donors and in patients with diagnosed HIVAN and idiopathic noncollapsing FSGS. Representative images of six subjects in each group are shown (original magnification, ×20). (B) Immunofluorescence was performed using WT1 (podocyte-specific marker) to colocalize for podocyte-specific KLF6 expression in renal biopsy specimens from 10 healthy donor nephrectomies and in 10 patients with idiopathic noncollapsing FSGS. Representative images from 10 subjects in each group are shown (original magnification, ×20). Arrows show colocalization of KLF6 to WT1. Arrowheads show a lack of colocalization of KLF6 to WT1. (C) Twenty glomeruli per biopsy specimen were selected, and quantification of KLF6 staining in the podocytes was determined as the ratio of KLF6⁺WT1⁺ cells to WT1⁺ cells (n = 10). Mann-Whitney U test, ***P < 0.001.

Figure 10. SCO2, a downstream target of KLF6, is critical to preventing the activation of intrinsic apoptotic pathway.

Primary podocytes were isolated from WT and Tg26 mice (FVB/N). RNA was extracted, and real-time PCR was performed. (A) Sco2 mRNA expression was compared between WT and Tg26 mice (n = 5). Mann-Whitney U test, **P < 0.01. (B) SCO2 expression was confirmed using immunofluorescence. Representative images of six independent experiments are shown in the left panel (original magnification, ×20). A total of 30 glomeruli per mouse were selected, and SCO2 expression was quantified in the glomerular region (n = 6) as shown in the right panel. Unpaired t test ***P < 0.001. (C) To assess the role of SCO2 in mitochondrial injury, shRNA-mediated SCO2 knockdown was performed in human podocytes. Western blot analysis was performed for SCO2 and cleaved caspase-9. Representative blots from three independent experiments are shown in the top panel. Quantification by densitometry (n = 3) is shown in the bottom panel. Mann-Whitney U test, **P < 0.01. (D) Immunofluorescence images of cytochrome c staining in EV-shRNA and SCO2-shRNA podocytes are shown. Representative images of four independent experiments are shown to demonstrate the distribution of cytochrome c staining (original magnification, ×20). Arrows show mitochondrial cytochrome c distribution. Arrowheads show cytosolic distribution of cytochrome c.

Figure 9. shRNA-mediated KLF6 knockdown resulted in activation of the intrinsic apoptotic pathway.

Cultured human podocytes (EV-shRNA and KLF6-shRNA) were treated with and without ADR for 24 hours. (A) Immunofluorescence images using Hoechst staining was performed to assess for apoptotic bodies. Representative images of six independent experiments are shown (original magnification, ×20). Arrows show apoptotic bodies. (B) To quantify apoptosis, annexin V/propidium iodide staining in combination with FACS was performed (n = 3). Kruskal-Wallis test with Dunn’s post-hoc test, *P < 0.05. (C) Immunofluorescence images of cytochrome c staining in EV-shRNA and KLF6-shRNA podocytes treated with and without ADR are shown. Representative images of six independent experiments are shown to demonstrate the distribution of cytochrome c staining (original magnification, ×20). Arrows show mitochondrial cytochrome c distribution. Arrowhead shows cytosolic distribution of cytochrome c. (D) Activation of the intrinsic apoptotic pathway was assessed using Western blot analysis for cleaved caspase-9, pro–caspase-3, and cleaved caspase-3 and is shown with β-actin as a loading marker. The representative images of six independent experiments are shown in the top panel. Quantification by densitometry (n = 6) is shown in the bottom panel. Kruskal-Wallis test with Dunn’s post-hoc test, *P < 0.05 vs. treated and untreated EV-shRNA, ^#P < 0.05 vs. all groups, †P < 0.01 vs. all groups. (E) To confirm whether the preservation of KLF6 prevents apoptosis, Western blot analysis was performed on human podocyte lysates from ADR-treated podocytes with (LentiORF-KLF6) and without (LentiORF-control) KLF6 overexpression. Representative images of three independent experiments are shown in the top panel (cleaved caspase-3 and pro–caspase-3 are from the same samples run on parallel gels). Quantification by densitometry (n = 3) is shown in the bottom panel. Mann-Whitney U test, ***P < 0.001, **P < 0.01.

Figure 8. shRNA-mediated KLF6 knockdown increased the susceptibility to mitochondrial injury.

EV-shRNA and KLF6-shRNA human podocytes were treated with and without ADR for 24 hours. (A) Mitochondrial membrane potential was quantified (n = 6). Mann-Whitney U test, **P < 0.01. (B) Top panel: Western blot analysis for SCO2 was performed, and representative images of six independent experiments are shown (SCO2 and β-actin are from the same samples on the same blot, with β-actin being developed after SCO2). Bottom panel: Quantification of SCO2 by densitometry (n = 6). Kruskal-Wallis test with Dunn’s post-hoc test, **P < 0.01. (C) ATP levels were quantified (n = 8). Kruskal-Wallis test with Dunn’s post-hoc test, **P < 0.01, ***P < 0.001. (D) Extracellular oxygen consumption rate (OCR) was measured and expressed as fold change relative to untreated EV-shRNA podocytes (n = 6). Kruskal-Wallis test with Dunn’s post-hoc test, **P < 0.01. (E) Rosamine-based MitoTracker probe was used to assess mitochondrial structure and fragmentation. Top panel: Representative images of six independent experiments (original magnification, ×20). Mitochondrial staining is indicated by tubular (arrows), intermediate (arrowheads), and fragmented (asterisks within images) pattern. Bottom panel: Scoring of mitochondrial morphology from 100 podocytes in each group (n = 6). Two-way ANOVA test with Tukey’s post-test, *P < 0.05 vs. tubular EV-shRNA, **P < 0.01 vs. tubular EV-shRNA, ^#P < 0.05 vs. all intermediate groups, *^#P < 0.05 vs. all fragmented groups. (F) To assess whether the reintroduction of KLF6 can rescue the cells from mitochondrial injury, LentiORF-KLF6 was transiently transfected in EVshRNA-KLF6 human podocytes, and MitoTracker probe was used to assess mitochondrial fragmentation. Top panel: Representative images of three independent experiments (original magnification, ×20). Bottom panel: Scoring of mitochondrial morphology from 100 podocytes in each group (n = 3). Two-way ANOVA test with Tukey’s post-hoc test, ^#P < 0.0001 vs. tubular KLF6-shRNA+LentiORF-control groups, **P < 0.01 vs. intermediate untreated tubular KLF6-shRNA+LentiORF-control group; ***P < 0.001 vs. all fragmented groups.

Figure 7. KLF6 expression is increased early with ADR treatment.

Primary podocytes isolated from WT mice were treated with and without ADR for 12 hours. RNA was extracted, and real-time PCR was performed. (A) Klf6 mRNA expression was compared between primary cultured murine podocytes treated with and without ADR (n = 5). Mann-Whitney U test, **P < 0.01. (B) This was confirmed by immunofluorescence using tissue from WT mice treated with and without ADR. Representative pictures of six mice in each group are shown in the top panel (original magnification, ×20). Bottom panel: 30 glomeruli per mouse were selected, and KLF6 expression was quantified in the glomerular region (n = 6). Unpaired t test, ***P < 0.001. (C) Cultured human podocytes were treated with ADR for 6, 12, 18, and 24 hours. Protein was extracted, and Western blot analysis was performed for KLF6. A representative blot of three independent experiments is shown in the top panel. Densitometry analysis is shown in the bottom panel (n = 6). Kruskal-Wallis test with Dunn’s post-hoc test, **P < 0.01 vs. untreated cells. (D) Immunofluorescence staining for KLF6 with and without ADR for 12 hours is shown. Representative images of six independent experiments are shown in the top panel (original magnification, ×20, scale bar: 100 μm). In the bottom panel, the intensity of KLF6 expression was quantified (n = 6). Unpaired t test, ***P < 0.001.

Figure 6. KLF6 binds to the promoter region of SCO2.

To confirm that binding sites for KLF6 are present in the promoter region of SCO2. KLF6 was initially overexpressed in HEK 293 cells using LentiORF transfection, and ChIP was performed. (A) Overexpression of KLF6 was confirmed with Western blot analysis (left panel), and the presence of putative KLF6 binding sites in the promoter of SCO2 in HEK 293 cells is shown (right panel) (n = 4). Kruskal-Wallis test with Dunn’s post-hoc test, **P < 0.01 vs. the two other groups. (B) Cultured podocytes were treated with ADR for 12 hours, RNA was extracted for real-time PCR, and Sco2 mRNA expression is shown (n = 4). Kruskal-Wallis test with Dunn’s post-hoc test, *P < 0.05, **P < 0.01. (C) SCO2 expression level was also confirmed by immunofluorescence in Podocin-Cre Klf6^fl/fl and Podocin-Cre Klf6^+/+ mice treated with and without ADR. Representative pictures of six mice in each group are shown in the top panel (original magnification, ×20). A total of 30 glomeruli per mouse were selected, and SCO2 expression was quantified in the glomerular region (n = 6) shown in the bottom panel. Kruskal-Wallis test with Dunn’s post-hoc test, **P < 0.01. (D) MT-CO2 levels were assessed in Podocin-Cre Klf6^fl/fl and Podocin-Cre Klf6^+/+ mice treated with and without ADR. Representative immunofluorescence images of six mice in each group are shown in the top panel (original magnification, ×20). A total of 30 glomeruli per mouse were selected, and MT-CO2 expression was quantified in the glomerular region (n = 6) in the bottom panel. Kruskal-Wallis test with Dunn’s post-hoc test, **P < 0.01, ***P < 0.001.

Figure 5. Podocin-Cre Klf6^fl/fl mice exhibit dysmorphic mitochondria in podocytes with ADR treatment.

Podocin-Cre Klf6^fl/fl and Podocin-Cre Klf6^+/+mice were treated with ADR at 12 weeks of age. All mice were sacrificed, and renal cortex was fixed for histology 5 weeks after ADR treatment. (A) Electron microscopy was performed to assess ultrastructural changes in the mitochondria of the podocyte cell body (top panels: original magnification, ×50,000; bottom panels: original magnification, ×75,000). Representative images are shown from six mice in each group. Red arrows show elongated mitochondria with preserved cristae and membrane. Red arrowheads show the loss of cristae and the loss of elongated morphology in the mitochondria. (B) To quantify dysmorphic mitochondria, a total of 100 mitochondrial podocytes per mouse were selected from each group to identify the percentage of mitochondria with a focal loss of visible cristae, clustering of residual cristae at the peripheral mitochondrial membrane, and length of <2 μm (n = 4 mice). Unpaired t test, ***P < 0.001.

Figure 4. Podocin-Cre Klf6^fl/fl mice exhibit substantial podocyte injury with ADR treatment.

Podocin-Cre Klf6^fl/fl and Podocin-Cre Klf6^+/+mice were treated with ADR at 12 weeks of age. All mice were sacrificed, and renal cortex was fixed for histology 5 weeks after ADR treatment. Electron microscopy was performed to assess ultrastructural changes in podocyte morphology (top panels: original magnification, ×3,000; bottom panels: original magnification, ×10,000). Representative images are shown from four mice in each group. Red arrows indicate the change in podocyte foot process morphology. Red asterisks show microvillus transformation.

Figure 3. ADR-treated Podocin-Cre Klf6^fl/fl mice exhibit a significant increase in albuminuria with glomerulosclerosis and tubulointerstitial injury.

Podocin-Cre Klf6^fl/fl and Podocin-Cre Klf6^+/+ mice were treated with ADR at 12 weeks of age. Urine was collected weekly, mice were sacrificed, and renal cortex was fixed for histology 5 weeks after ADR treatment. (A) Albuminuria (urine albumin/creatinine [cr]) was measured (n = 10). Kruskal-Wallis test with Dunn’s post-hoc test, *P < 0.05 as compared with untreated Podocin-Cre Klf6^+/+ mice, ***P < 0.001 as compared with all groups. (B) PAS was used to evaluate glomerular or tubulointerstitial changes (top panel: original magnification, ×10; bottom panel: original magnification, ×40). Representative images from six mice in each group are shown. Arrows show sclerotic glomeruli. Arrowheads show interstitial inflammation. Asterisks within images show tubular casts and dilatation.

Figure 2. Podocyte-specific knockdown of Klf6 in Podocin-Cre Klf6^fl/fl mice was confirmed.

(A) Primary podocytes were isolated from 10-week-old Podocin-Cre Klf6^fl/fl and Podocin-Cre Klf6^+/+ mice and cultured at 37°C for 1 week. RNA was extracted, and real-time PCR was performed for Klf6 mRNA expression (n = 6). Mann-Whitney U test, ***P < 0.001. (B) Protein was also extracted, and Western blot analysis was performed for KLF6. A representative blot of four independent experiments is shown in the left panel (KLF6 and β-actin are from the same samples on the same blot, with β-actin being developed after KLF6). The right panel shows the quantification of KLF6 by densitometry (n = 4). Mann-Whitney U test, **P < 0.01. (C) Immunofluorescence staining for KLF6 and WT1 was performed in 10-week-old Podocin-Cre Klf6^fl/fl and Podocin-Cre Klf6^+/+ mice. Representative images from six mice in each group are shown in the left panel (original magnification, ×20). Arrows show colocalization of KLF6 and WT1. Arrowheads show a lack of colocalization. Right panel: 30 glomeruli were selected in each mouse, and quantification of KLF6 staining in the podocytes was determined by the ratio of KLF6⁺ and WT1⁺ cells to WT1⁺ cells (n = 6 mice). Unpaired t test, ***P < 0.001.

Figure 1. Strong association between KLF6 and the glomerulosclerosis was observed in HIVAN.

(A) mRNA levels of KLFs, previously shown to be expressed in epithelial cells, were measured in differentiated WT and HIV-1–infected human podocytes in culture. Real-time PCR, using primers for KLF4, KLF5, KLF6, KLF7, KLF10, KLF11, KLF13, KLF15, and KLF16, was performed using RNA isolated from WT and HIV-1–infected human podocytes. Inset: Relative mRNA expression for KLF6 (n = 4). Mann-Whitney U test, *P < 0.05. (B) Age- and sex-matched 6-week-old HIV-1 transgenic mice (Tg26) and WT mice on the FVB/N background were used to assess KLF6 expression in the glomeruli. Western blot analysis was performed on the glomerular lysates from WT and Tg26 mice for KLF6 and GAPDH. A representative blot of six independent experiments is shown in the top panel. The bottom panel shows the quantification of KLF6 by densitometry (n = 6). Mann-Whitney U test, *P < 0.05. (C) The top panel shows representative images of immunohistochemistry for KLF6 in kidney sections from WT and Tg26 mice (n = 8 mice) (original magnification, ×20). Bottom panel: 30 glomeruli were selected in each mouse, and the intensity of KLF6 expression was quantified in the glomerular region (n = 8 mice). Unpaired t test, ***P < 0.001.