Targeting Krüppel-Like Factor 2 as a Novel Therapy for Glomerular Endothelial Cell Injury in Diabetic Kidney Disease

Abstract

Key Points:

1. Krüppel-like factor 2 (KLF2) has emerged as a key endoprotective regulator by suppressing inflammatory and oxidative pathways, thrombotic activation, and angiogenesis.

2. Our study now demonstrates that KLF2 protects against glomerular endothelial injury and attenuates diabetic kidney disease progression in mice.

3. Compound 6 is a novel KLF2 activator that can potentially confer dual cardiorenal protection against diabetic complications.

Background: Diabetic kidney disease (DKD) is a microvascular disease, and glomerular endothelial cell injury is a key pathological event in DKD development. Through unbiased screening of glomerular transcriptomes, we previously identified Krüppel-like factor 2 (KLF2) as a highly regulated gene in diabetic kidneys. KLF2 exhibits protective effects in endothelial cells by inhibiting inflammation, thrombotic activation, and angiogenesis, all of which are protective for cardiovascular disease. We previously demonstrated that endothelial cell–specific ablation of Klf2 exacerbated diabetes-induced glomerular endothelial cell injury and DKD in mice. Therefore, in this study, we sought to assess the therapeutic potential of KLF2 activation in murine models of DKD.

Methods: We first examined the effects of endothelial cell–specific inducible overexpression of KLF2 (KLF2^ov) in streptozotocin-induced diabetic mice. We developed small molecule KLF2 activators and tested whether higher KLF2 activity could impede DKD progression in type 2 diabetic db/db and BTBR ob/ob mice.

Results: Diabetic KLF2^ov mice had attenuated albuminuria, glomerular endothelial cell injury, and diabetic glomerulopathy compared with control diabetic mice. A novel KLF2 activator, compound 6 (C-6), effectively induced downstream Nos3 expression and suppressed NF-kB activation in glomerular endothelial cells. The administration of C-6 improved albuminuria and glomerulopathy in db/db and BTBR ob/ob mice, which was associated with improved glomerular endothelial cell and podocyte injury.

Conclusions: These results validate KLF2 as a potential drug target and KLF2 activators, such as C-6, as a novel therapy for DKD.

Graphical abstract

Figure 1

Endothelial cell–specific KLF2 overexpression attenuates glomerular hypertrophy and albuminuria in diabetic mice. (A) KW/BW ratio in control and diabetic mice at 20 weeks after STZ injection. (B) Twelve-hour total urinary albumin excretion in control and diabetic mice. (C) GFR in control and diabetic mice. (D) Representative PAS-stained mouse kidneys at 400× magnification. Scale bar, 20 μm. (E) Average glomerular area and mesangial matrix fraction in control and diabetic mice (n=6 mice per group), *P < 0.05, **P < 0.01, and ***P < 0.001 between indicated groups by one-way ANOVA with Bonferroni's correction. KLF2, Krüppel-like factor 2; KW/BW, kidney weight-to-body weight; PAS, periodic acid–Schiff; STZ, streptozotocin; WT, wild-type.

Figure 2

Endothelial cell–specific overexpression of KLF2 normalizes eNOS and angiogenic genes in GECs and attenuates podocyte injury in diabetic mice. (A) Representative images of eNOS immunofluorescence. Nuclei are counterstained with DAPI. Glomeruli are outlined (original magnification ×400, scale bar, 20 μm). (B) eNOS immunostaining intensity, relative to WT mice (60 glomeruli evaluated per mouse, n=6 mice per group). (C) Nos3 mRNA expression in isolated glomeruli of control and diabetic mice was assessed by qPCR (n=6 mice per group). (D) mRNA expression of vascular factors and receptors is assessed in isolated glomeruli of control and diabetic mice by qPCR (n=3 mice per group). (E) Representative images of WT-1 immunofluorescence. (F) Average WT-1+ cells per glomerular cross-section (60 glomeruli evaluated per mouse, n=6 mice per group). (G) Real-time PCR analysis of Wt1 mRNA in isolated glomeruli of mice (n=6 mice per group). *P < 0.05, **P < 0.01, and ***P < 0.001 between indicated groups by one-way ANOVA with Bonferroni's correction. DAPI, 4',6-diamidino-2-phenylindole; eNOS, endothelial nitric oxide synthase; GEC, glomerular endothelial cell; qPCR, quantitative PCR.

Figure 3

C-6 directly binds to KLF2 and increases its transcriptional activity. (A) HGECs were treated with vehicle (Vh) or 50 nM of potential KLF2 activator C-6, C-7, and C-9 for 24 hours. KLF2 activity was assessed by Western blot analysis of eNOS expression. (B) HGECs were treated with varying concentrations of C-6 as indicated for 24 hours. KLF2 activity was assessed by Western blot analysis of eNOS protein expression. (C) Quantification of Western blot analysis in (A) and (B) (n=3 per group). (D) Dose-dependent actions of C-6 were assessed by induction of eNOS expression by ELISA, repression of p65 by NF-κB luciferase activity, and repression of CCL2 mRNA expression by qPCR in HGECs. (E) DARTS assay shows protection of KLF2 against pronase digestion with an increasing C-6 dose. Quantification of the Western blot is shown on the right (n=3 per group). (F) ChIP-PCR analysis of HGECs treated with vehicle or C-6 for 24 hours (50 nM) with anti-KLF2 antibody. KLF2-overexpressing cells served as a positive control, and control (Ctrl) IgG served as a negative control (n=6 per group). *P < 0.05, **P < 0.01, and ***P < 0.001 between indicated groups by one-way ANOVA with Bonferroni's correction. C-6, compound 6; ChIP, chromatin immunoprecipitation; DARTS, drug affinity responsive target stability; HGEC, human glomerular endothelial cell; nd, not detected.

Figure 4

C-6 attenuates albuminuria and diabetic glomerulopathy in diabetic BTBR ob/ob mice. (A) Urinary albumin-to-creatinine ratio (UACR) of mice for 8 weeks of C-6 or vehicle treatment. P < 0.001 between vehicle- and C-6–treated BTBR ob/ob mice after 8 weeks of treatment. (B) Representative PAS images of control and diabetic kidneys at 400× magnification. Scale bar, 20 μm. (C) Quantification of glomerular area and % mesangial matrix in control and diabetic mice (n=6 mice per group). **P < 0.01 and ***P < 0.001 between indicated groups by one-way ANOVA with Bonferroni's correction.

Figure 5

C-6 treatment normalizes eNOS and angiogenic gene expression in diabetic BTBR ob/ob mice. (A) Representative images of eNOS immunofluorescence. Nuclei are counterstained with DAPI. Glomeruli are outlined (original magnification ×400, scale bar, 20 μm). (B) eNOS immunostaining intensity, relative to WT mice (60 glomeruli evaluated per mouse, n=6). (C) Nos3 mRNA expression in isolated glomeruli of control and diabetic mice was assessed by real-time PCR (n=6 mice per group). (D) Expression of vascular factors and receptors is assessed in isolated glomeruli of control and diabetic mice by qPCR (n=3 mice per group). *P < 0.05, **P < 0.01, and ***P < 0.001 between indicated groups by one-way ANOVA with Bonferroni's correction.

Figure 6

C-6 attenuates podocyte loss and macrophage infiltration in diabetic BTBR ob/ob mice. (A) Representative images of WT-1 immunostaining (left) and quantification of average WT-1+ cells per glomerular cross-section (right). Scale bar, 20 μm. Sixty glomeruli evaluated per mouse, n=6 mice per group. (B) Representative images of F4/80 immunostaining (left) and quantification of average F4/80-positive area per field. Scale bar, 50 μm. Fifteen fields were evaluated per mouse, n=6 mice per group. *P < 0.05 and ***P < 0.001 between compared groups by one-way ANOVA with Bonferroni's correction.