Therapeutic miR-21 Silencing Ameliorates Diabetic Kidney Disease in Mice

Abstract

Diabetic nephropathy is the main cause of end-stage renal disease. MicroRNAs are powerful regulators of the genome, and global expression profiling revealed miR-21 to be among the most highly regulated microRNAs in kidneys of mice with diabetic nephropathy. In kidney biopsies of diabetic patients, miR-21 correlated with tubulointerstitial injury. In situ PCR analysis showed a specific enrichment of miR-21 in glomerular cells. We identified cell division cycle 25a (Cdc25a) and cyclin-dependent kinase 6 (Cdk6) as novel miR-21 targets in mesangial cells. miR-21-mediated repression of Cdc25a and Cdk6 resulted in impaired cell cycle progression and subsequent mesangial cell hypertrophy. miR-21 increased podocyte motility by regulating phosphatase and tensin homolog (Pten). miR-21 antagonism in vitro and in vivo in streptozotocin-induced diabetic mice decreased mesangial expansion, interstitial fibrosis, macrophage infiltration, podocyte loss, albuminuria, and fibrotic- and inflammatory gene expression. In conclusion, miR-21 antagonism rescued various functional and structural parameters in mice with diabetic nephropathy and, thus, might be a viable option in the treatment of patients with diabetic kidney disease.

Keywords: TGF-β; cell-cycle regulators; diabetic nephropathy; mesangial hypertrophy; miR-21; microRNA; podocyte motility.

Figure 1

miR-21 in Diabetic Mice and Diabetic Patients (A and B) miRNA-array analysis: arrow indicates miR-21 (A), miR-21 array quantification (B), and qPCR validation of miR-21 (C). miR-21 visualization by in situ PCR on kidney sections of non-diabetic (D) and diabetic (E) mice. (F) Quantification of in situ PCR miR-21 positive (red) staining. (G) Serum miR-21 expression is increased in diabetic patients compared to healthy controls. (H) Urinary miR-21 expression correlated with proteinuria in diabetic patients. (I) miR-21 expression in human kidney biopsies of diabetic patients correlated with chronic tubulointerstitial injury (percentage of injury was translated into the following arbitrary scale for further correlation studies: 1%–10% = 1; 11%–20% = 2; 21%–30% = 3; 31%–40% = 4; 41%–50% = 5; 51%–60% = 6; 61%–70% = 7; 71%–80% = 8; 81%–90% = 9; 91%–100% = 10) (I). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 2

Mesangial Cell Hypertrophy Is Induced by miR-21-Mediated Repression of Cdc25a and Cdk6 (A and B) Mesangial cell hypertrophy displayed by F-actin staining (red) after transfection with LNA-scr and either BSA (A) as control (LNA-scr+BSA) or TGF-β (B) treatment (LNA-scr+TGFβ). (C) Inhibition of miR-21 by transfection of LNA-21 and additional TGF-β treatment (LNA-21+TGFβ). (D) Quantification of the cell size measurements (n = 5 each). (E–G) DNA histograms showing mesangial cells detected in G0/G1, S, or M/G2 phase of the cell cycle after LNAscr+BSA (E), LNA-scr+TGF-β (F), and LNA-21+TGF-β (G). (H–J) Quantification of cells detected in G0/G1 phase (H), S phase (I), and M/G2 phase (J) (n = 4). (K and L) Predicted consequential pairing of Cdc25a target region and miR-21 (K) and of Cdk6 target region and miR-21 (L), respectively (http://www.targetscan.org). (M and P) Western blot (WB) analysis of cytosolic Cdc25a (M) and Cdk6 (P) after overexpression of miR-21 by transfection with precursor-miR21 (pre-miR-21) in comparison to a scrambled control (pre-neg) oligonucleotide (n = 3). (N and Q) Western blot (WB) analysis of cytosolic Cdc25a (N) and Cdk6 (Q) after LNA-scr+TGFβ and LNA-21+TGFβ (n = 3). (O and R) Validation of miR-21 targeting Cdc25a (O) and Cdk6 (R) by Luciferase activity in relation to β-galactosidase activity after overexpression of miR-21 by transfection with pre-miR-21 in comparison to pre-neg with either wild-type or mutated target sites of Cdc25a and Cdk6 (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 3

AP-1 as Key Transcriptional Activator of miR-21 in Mesangial Cells (A) Electrophoretic mobility shift assay showing AP-1 activation as an AP-1/DNA complex (shifted band) after TGF-β treatment for 2 hr and 24 hr. (B) Every treatment was assessed in comparison to a competition assay (CA), respectively. qPCR quantification of primary-miR-21 after TGF-β treatment (n = 3). (C) AP-1 Luciferase activity in relation to β-galactosidase activity after an initial increase of AP-1 activity by phorbol myristate acetate (PMA) and subsequent AP-1 inhibition by synthetic retinoid 11302 (SR11302) (n = 4). (D–F) qPCR quantification of miR-21 (D), CTGF (E), and Col1a2 (F) after co-treatment with TGF-β and SR11302 (n = 4). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 4

miR-21-Mediated Podocyte Phenotype (A and B) qPCR quantification of miR-21 (A) and Col1a2 (B) after TGF-β treatment (n = 5). (C–H) Scratch migration for 8 hr in podocytes after pre-miR-21 treatment compared to pre-neg (C and D) and LNA-21 treatment compared to LNA-scr (F-G); quantification of the cell free area after 8 hr migration (E and H) (n = 6). (I and J) Western blot analysis of Pten after LNA-21 treatment compared to LNA-scr. (K) Predicted consequential pairing of Pten target region and miR-21 (miRTaBase). (L–Q) Annexin V apoptosis staining and FACS analysis of cells undergoing early apoptosis (quadrant 3 = Q3) after pre-miR-21 treatment compared to pre-negative controls (L–N) and LNA-21 treatment compared to LNA-scr (O–Q). RED fluorescence detects the level of cell viability using 7-amino actinomycin D (7-AAD), RED2 fluorescence represents annexin V positivity. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 5

Molecular Therapeutic Inhibition of miR-21 Reduced Structural Complications in Diabetic Mice Histological evaluation of therapeutic miR-21 inhibition in diabetic (STZ) mice. (A–D) Mesangial index analysis after periodic acid Schiff (PAS) staining in non-diabetic mice (A), diabetic mice with LNA-scrambled (B), and LNA-21 treatment (C) as well as quantification of results (D). (E–I) Detection of interstitial collagen by Sirius Red staining in non-diabetic mice (E), diabetic mice with LNA-scrambled (F), and LNA-21 treatment (G) as well as quantification of results (H). (I–L) Analysis of F4/80 (red) positive macrophage infiltration in non-diabetic mice (I), diabetic mice with LNA-scrambled (J), and LNA-21 treatment (K) as well as quantification of results (L). (M–P) Urinary albumin:creatinine ratio (M), qPCR quantification of Col1a2 (N), CTGF (O) and MCP-1 (P) in non-diabetic control mice as well as diabetic mice subjected to LNA-21 and LNA-scr treatment. Non-diabetic mice, n = 10. Diabetic (STZ) mice treated with LNA-scr, n = 6. Diabetic (STZ) mice treated with LNA-21, n = 5. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 6

Molecular Therapeutic Inhibition of miR-21 Reduced Functional Complications in Diabetic Mice (A–D) Determination of WT-1 (red) positive podocyte loss in non-diabetic mice (A), diabetic mice with LNA-scrambled (B), and LNA-21 treatment (C) as well as quantification of results (D). (E–H) Cdc25a immunofluorescence staining in non-diabetic mice (E), diabetic mice with LNA-scrambled (F), and LNA-21 treatment (G) as well as quantification of results (H). (I–L) Cdk6 immunofluorescence staining in non-diabetic mice (I), diabetic mice with LNA-scrambled (J), and LNA-21 treatment (K) as well as quantification of results (L). Non-diabetic mice, n = 10. Diabetic (STZ) mice treated with LNA-scr, n = 6. Diabetic (STZ) mice treated with LNA-21, n = 5. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 7

Summary of the Main Results High glucose under diabetic conditions leads to an increase in TGF-β. miR-21 targets Cdc25a and Cdk6 in mesangial cells (turquoise), leading to mesangial cell hypertrophy by inducing a G1-phase arrest. miR-21 targets Pten, leading to increased podocyte motility and extracellular matrix production (orange). miR-21 leads to extracellular matrix accumulation (purple). These mechanisms contribute to diabetic kidney disease. miR-21 silencing ameliorates these phenomena.