TGF-β1 induces EMT-like changes in proximal tubular epithelium-like cells. (A) NRK52E cells were cultured in the presence of TGF-β1 (10 ng/ml, 3 days). After treatment with TGF-β1, collagen I and IV gene expression were significantly increased by real-time quantitative PCR, whereas E-cadherin expression was significantly decreased (*P<0.05 compared with control). (B) Western blot analysis demonstrated that collagen I and collagen IV were both elevated at the protein level after TGF-β1 treatment and (C) these increases were significant (*P<0.05 compared with control). (D) Expression of miR-29a and miR-29b were both significantly decreased in cells treated with TGF-β1 for 3 days (*P<0.05 compared with control). The decrease in miR-29-c was not significant. (E) In longer-term TGF-β1–treated cells (10 days), all three members of miR-29 were significantly reduced (*P<0.05 compared with control). (F) Relative expression levels of the miR-29 family members in control NRK52E cells compared with miR-29b. (*P<0.05 compared with miR-29b).

miR-29 represses the expression of collagen genes. (A) NRK52E cells were transfected with miR-29a/b/c (50 nM each) and RNA was harvested after 3 days for real-time quantitative PCR analysis. miR-29 significantly decreased expression of collagens I, III, and IV (*P<0.05 compared with control transfected cells). (B) Western blot analysis demonstrated a significant decrease in collagen I and collagen IV protein expression (*P<0.05 compared with control transfected cells), which was consistent with the RNA expression analysis. (C) The quantified Western blot data shown in graph format (*P<0.05 compared with control). (D) Conditionally immortalized human podocytes differentiated for 10 days at 33°C before treatment with TGF-β1 (3 days, 5 ng/ml). Treatment with TGF-β1 reduced expression of miR-29a/b/c (*P<0.05 compared with control). (E) Human podocytes were transfected with miR-29a/b/c or miR-C, with or without treatment with TGF-β (5 ng/ml, 3 days) and RNA isolated for expression analysis. TGF-β1 treatment increased expression of αSMA, vimentin, and collagens I, III, and IV (*P<0.05 compared with control). Transfection with miR-29a/b/c significantly decreased basal level expression of collagen IV relative to miR-C, and significantly attenuated TGF-β1–induced expression of αSMA, vimentin (Vim), and collagens I, III, and IV (^#P<0.05 compared with control). (F) Collagen I protein expression as assessed by Western blot analysis in podocytes transfected with miR-29a/b/c or miR-C, with or without TGF-β1 treatment. Protein expression was significantly attenuated by miR-29a/b/c in TGF-β1–treated cells as shown in the graph below (*P<0.05 compared with control).

miR-29 repression of collagen 1 expression in primary mouse mesangial cells. (A) Primary mouse mesangial cells were treated with TGF-β1 for 3 days demonstrated significantly increased collagen I, III, and IV expression (P<0.01 compared with control). (B) TGF-β1 also resulted in significantly decreased miR-29a, miR-29b, and miR-29c expression (P<0.05 compared with control). (C) Mesangial cells were transfected with miR-29a/b/c (50 nM each) and harvested after 3 days. Western blot analysis demonstrated significantly decreased collagen I expression in transfected cells (graphed below, P<0.05 compared with control). (D) Similar experiments with TGF-β1 treatment demonstrated that miR-29 transfection could prevent the induction of collagen I by TGF-β1 treatment (graphed below, P<0.05 compared with control). (E) Real-time qPCR analysis of miR-29 transfected cells shows significantly decreased collagen I mRNA in control but also in TGF-β1–treated cells (*P<0.05 and ^#P<0.05 compared with control).

The 3′UTR of collagens I and IV is regulated by miR-29a/b/c. NRK52E cells were transfected with collagen 3′UTR luciferase reporter plasmids (1 μg), β-galactosidase plasmid (0.2 μg), and either miR-C (150 nM) or miR-29a/b/c (50 nM each), and cells were analyzed for β-galactosidase and luciferase activity after 3 days. TGF-β1 significantly increased luciferase activity in cells containing constructs with the 3′UTR of (A) collagen I, (B) collagen IVa1, and (C) collagen IVa3 (*P<0.05 compared with control transfected cells). In each case, miR-29a/b/c prevented this increase and even reduced luciferase activity below control transfected cells (*P<0.05 compared with control transfected cells). (D) Similar experiments were carried out using luciferase constructs with collagen Ia1 and IVa3 3′UTRs containing mutated miR-29 binding sites. miR-29a/b/c had no effect in those constructs in which the miR-29 binding sites were mutated. (E) Ectopic transfection of NRK52E cells with miR-29a/b/c also resulted in significant elevation of E-cadherin mRNA levels (*P<0.05 compared with control transfected cells). (F) Ectopic expression of miR-29a/b/c also prevented the decrease in E-cadherin expression induced by TGF-β1, as observed by immunofluorescence.

Changes in miR-29a/b/c expression in diabetic mouse kidney. (A) Total RNA was extracted from kidney cortex of control and 10-week diabetic apoE mice (n=7 per group). The expression of miR-29 family members was assessed by real-time quantitative PCR, revealing significantly decreased expression of miR-29a and miR-29c in renal cortex of diabetic animals (*P<0.05 compared with control). The decrease in miR-29b was not significant. (B) Gene expression analysis revealed significantly increased levels of collagen IV and CTGF in diabetic animals (*P<0.05 compared with control). (C) Immunohistochemical analysis confirmed increased collagen IV expression in the kidneys of diabetic animals. (D) The relative expression of the miR-29 family members in control animals is shown relative to miR-29b. (E) Endogenous miR-29a and miR-29c were detected in proximal tubular glomerular cells in control and 10-week diabetic apoE^−/− mouse kidney using LNA probes (Exiqon).

Renal expression of ECM proteins in diabetic rats. (A) Representative microphotographs (×200) show cortical sections from nondiabetic animals (C-VE), vehicle-treated diabetic rats (D-VE), fasudil-treated diabetic animals (D-FA), and diabetic animals treated with losartan (D-LOS) probed with primary antibodies raised against collagens I (Col) and IV (Col IV), fibronectin (×100), and laminin (×200). Arrows show immunoreactivity of proteins of interest in the tubulointerstitial compartment, and red arrowheads show examples of positively stained glomeruli. (B) Protein abundance of collagen 1, fibronectin, and laminin in renal cortical homogenates was further quantified by Western blotting. (C) Representative blots are shown and densitometric analysis is shown (*P<0.05, ^†P<0.01 versus C-VE; ^aP<0.05 versus D-VE.) (D) The expression of miR-29a/b/c was assessed in this model by real-time quantitative PCR. miR-29a and miR-29c are significantly reduced in diabetic kidney cortex in this model. Fasudil treatment restores expression to that of control animals. Losartan treatment partially restores expression (*P<0.05 compared with control; ^#P<0.05 compared with diabetic animals).

Changes in miR-29a/b/c expression in the adenine-induced renal fibrosis model. (A) Trichrome staining of tissue sections from renal cortex from control and adenine-fed C57bl6 mice after 4 weeks of treatment (n=3 per group). Blue staining indicates high levels of collagen in the adenine-fed mouse kidney compared with control. (B) mRNA was extracted from the renal cortex of control and adenine-fed C57BL6 mice. Gene expression was assessed by real-time quantitative PCR, revealing significantly increased expression of collagen I (*P<0.05 compared with control). (C) The expression of miR-29a/b/c, as assessed by real-time qPCR, showed significant reduction for all miR-29 family members (*P<0.05 compared with control). The expression of the housekeeping gene used in these studies, RN6UB, was not altered.

Schema of miRNAs and their potential role in renal fibrosis.