miR-29c is downregulated in renal interstitial fibrosis in humans and rats and restored by HIF-α activation

Abstract

Treatment with L-mimosine, which activates hypoxia-inducible factor-α (HIF-α), attenuates renal tubulointerstitial injury and improves renal function in a rat remnant kidney model. The miR-29 family of microRNAs directly targets a large number of extracellular matrix genes and reduces renal interstitial fibrosis. We analyzed microRNA expression profiles in rat remnant kidneys with or without treatment with L-mimosine. The expression of miR-29c was downregulated in rat remnant kidneys compared with sham control and significantly restored by the L-mimosine treatment. In cultured human kidney epithelial HK2 cells, cobalt chloride activated HIF-α and upregulated miR-29c expression. The upregulation of miR-29c expression was significantly attenuated by knockdown of HIF-1α or HIF-2α. Downregulation of miR-29c was associated with significant increases in interstitial fibrosis, collagen type II α1 (COL2A1) protein, and tropomyosin 1α (TPM1) protein in rat remnant kidneys and in kidneys from IgA nephropathy patients. The increases in rat remnant kidneys were attenuated by the L-mimosine treatment. COL2A1 and TPM1 were confirmed to be new, direct targets of miR-29c. In conclusion, miR-29c, an antifibrotic microRNA, is upregulated by HIF-α activation. MiR-29c is downregulated in renal interstitial fibrosis in humans and rats and restored by activation of HIF-α that attenuates fibrosis.

Keywords: HIF; collagen type II α1; miR-29c; tropomyosin 1α; tubulointerstitial fibrosis.

Fig. 1.

l-Mimosin (l-Mim) partially restored microRNA miR-29c expression in the remnant kidneys of rats with ⅚ nephrectomy. A: 9 of the 350 miRNAs examined with an Agilent miRNA microarray were differentially expressed in remnant kidneys in rats treated with l-Mim, an inhibitor of prolyl 4-hydroxylase (PHD) activity and an activator of hypoxia-inducible factor (HIF)-α, and vehicle. Rats with remnant kidneys were treated with intraperitoneal injections of vehicle or l-Mim at a dosage of 50 mg/kg every other day from week 5 to week 12 following surgery (n = 4). See materials and methods for the criteria of differential expression. B: Taqman quantitative (q) PCR analysis indicated that miR-29c was downregulated in the rat remnant kidneys, which was partially restored by the l-Mim treatment. RNA samples used for the individual Taqman qPCR analysis were independent of those used in the high-throughput profiling with the Agilent miRNA microarray (n = 4). *P < 0.05 vs. sham-operated group. #P < 0.05 vs. l-Mim-treated group. C: Rno-pri-miR-29b2/c was also upregulated after l-Mim treatment compared with the vehicle control group. RNA samples used for the individual Taqman qPCR analysis were independent of those used in the high-throughput profiling with the Agilent miRNA microarray (n = 4). *P < 0.05 vs. vehicle control. D: MiR-29c abundance and protein and mRNA levels of HIF-α during the progression time course in the remnant kidney model. MiR-29c was decreased in the early stage of the remnant kidney (RK) model (week 2), and then gradually and partially returned to the baseline value thereafter. *P < 0.05 vs. week 2. #P < 0.01 vs. week 2. §P < 0.01 vs. week 0. Protein levels of HIF-1α and -2α, which were replotted from previously published data (49) to visualize the comparison with miR-29c, were gradually upregulated in the early stage of the remnant kidney model, reached the peak at weeks 4 and 6, and quickly decreased to the baseline value thereafter. ^ΔP < 0.05 vs. week 0 (HIF-1α protein level). ^▲P < 0.05 vs. week 0 (HIF-2α protein level; n = 4).

Fig. 2.

Activation of HIF-α upregulated miR-29c. HK2 cells were preincubated with CoCl₂ [200 μM: negative control group); HIF-1α small interfering (si) RNA group; HIF-2α siRNA group; and CoCl₂ alone group] or vehicle (vehicle control group) for 24 h before siRNA treatment. Samples were collected 24 h (for RT-PCR) or 48 h (for Western blotting) after siRNA treatment. siRNA sequences against HIF-1α, HIF-2α, and HIF-a scrambled control (negative control) were designed and synthesized by GenePharma (Shanghai, China) and used at a final concentration of 100 nM. A: miR-29c was upregulated by CoCl₂, and the upregulation was blunted by siRNA targeting HIF-1α or HIF-2α (n = 6). B: mRNA levels of HIF-1α and HIF-2α were increased by CoCl₂ and knocked down by respective siRNAs (n = 3). C: protein levels of HIF-1α and HIF-2α were increased by CoCl₂ and knocked down by respective siRNAs. D: quantitation of Western blotting results (n = 3). E: relative abundance of pri-miR-29b2/c in HK2 cells was upregulated 24 h after CoCl₂ treatment (200 μM; n = 6). Negative control group: pretreated with CoCl₂ and the nonspecific siRNA; HIF-1α siRNA group: pretreated with CoCl₂ and the HIF-1α siRNA; HIF-2α siRNA group: pretreated with CoCl₂ and the HIF-2α siRNA; CoCl₂ alone: pretreated with CoCl₂, no siRNA; vehicle control group: pretreated with saline and nonspecific siRNA. *P < 0.05 vs. negative control. #P < 0.01 vs. negative control. ^ΔP < 0.01 vs. CoCl₂ alone. ^▲P < 0.05 vs. CoCl₂ alone. §P < 0.01 vs. vehicle control. ^▫P < 0.05 vs. vehicle control. ^■P < 0.05 vs. HIF-2α siRNA.

Fig. 3.

Interstitial fibrosis in rat remnant kidneys was attenuated by l-Mim. The relative abundance of collagens was analyzed using Masson's trichrome staining. The positive-stained area of collagens was quantitatively measured using a computer-aided image system (IMS; ShentengIT, Shanghai, China) on digitized images that were transformed from analog images taken by a video camera (Panasonic, MV-CP410). Each field was 72,800 μm². Magnification ×200. A: sham-operated rats. B: remnant kidneys from l-Mim-treated rats. C: remnant kidneys from vehicle-treated rats. D: summary of the quantitation (n = 5). *P < 0.01 vs. A. ^ΔP < 0.01 vs. B.

Fig. 4.

miR-29c expression was inversely correlated with renal interstitial fibrosis in humans. A: Masson's trichrome staining of renal biopsy specimens from patients with IgA nephropathy showing subjects with or without substantial interstitial fibrosis. Each field was 72,800 μm², magnification ×200. B: immunohistochemical analysis for HIF-1α abundance in renal biopsy specimens from IgA nephropathy (IgAN) patients. HIF-1α protein was present in both group (brown nuclear staining represents positive HIF-1α staining; black arrows). Magnification ×100. C: positive trichrome-stained areas were quantified as described in Fig. 3. D: positive HIF-1α-stained areas were quantified as described in Fig. 3. E: miR-29c abundance was decreased in kidneys with a high degree of interstitial fibrosis from patients with IgAN; TIF, tubulointerstitial fibrosis. *P < 0.05 vs. IgAN without TIF. #P < 0.01 vs. IgAN without TIF; n = 5.

Fig. 5.

Protein expression of tropomyosin 1α (TPM1 or TM), a potential target gene for miR-29c, in IgAN and rat remnant kidneys. A: immunohistochemical analysis for TM abundance in renal biopsy specimens from IgAN patients with or without fibrotic lesions in the interstitium. Arrows show positive staining of TM. B: quantification of TM staining in renal biopsy specimens. The quantification was performed as described in Fig. 3. *P < 0.05 vs. IgAN with TIF; n = 5. C: protein expression of TMs in rat remnant kidneys was upregulated in the ⅚ nephrectomy model, which was attenuated by the l-Mim treatment. Lanes 1–3, remnant kidneys from l-Mim-treated rats; lanes 4–6, remnant kidneys from vehicle-treated rats; lanes 7–9, sham-operated rats. See the text and Fig. 1 for explanation of the animal model and the antibodies.

Fig. 6.

Immunohistochemical analysis of collagen type II α1 (COL2A1), a potential target gene for miR-29c, in rat remnant kidneys and IgAN patients. The positively stained area of COL2A1 protein was quantified as described in Fig. 3. Only the renal interstitial area was selected for analysis. Stained regions were thresholded using an IMS (ShentengIT). Each field was 72,800 μm², magnification ×200. A: sham-operated rats. B: remnant kidneys from l-Mim-treated rats. C: remnant kidneys from vehicle-treated rats. D: IgAN patients without significant tubulointerstitial fibrosis. E: IgAN patients, with tubulointerstitial fibrosis. *P < 0.01 vs. A. ^ΔP < 0.01 vs. B. #P < 0.01 vs. D.

Fig. 7.

MiR-29c interacted with the 3′-untranslated region (UTR) of TPM1 and COL2A1. HeLa cells at 80–90% confluency were cotransfected with a firefly luciferase reporter construct containing the 3′-UTR of TPM1 or COL2A1, a pRL-TK internal control plasmid, and control pre-miR oligonucleotides or the miR-29c mimic. In additional experiments, mutations were introduced into the miR-29c seed region binding site within the 3′-UTR of TPM1 and COL2A1 in the reporter construct. The miR-29c mimic decreased luciferase activity when the reporter gene was linked to a 3′-UTR segment of COL2A1 (A) or TPM1 (B). The effect of miR-29c mimics was not significant when the binding site was mutated. *P < 0.01 vs. pre-miR negative control. #P < 0.05 vs. pre-miR negative control; n = 5.

Fig. 8.

Downregulation of miR-29c contributes to transforming growth factor (TGF)-β1-induced upregulation of TPM1 in HK2 cells. A: TGF-β1 downregulated miR-29c. HK2 cells were treated with TGF-β1 (3 ng/ml) or vehicle for 24 h. B: CoCl₂ restored miR-29c abundance that was suppressed by TGF-β1 treatment. HK2 cells were preincubated with or without CoCl₂ (200 μM) for 24 h before TGF-β1 (3 ng/ml) treatment; n = 6. *P < 0.05 vs. TGF-β1 alone. C: miR-29c mimics reduced TPM1 protein levels in HK2 cells treated with TGF-β1. The density of the TM band was normalized by Coomassie blue staining of the entire lane; n = 5. *P < 0.05 vs. vehicle-treated. #P < 0.05 vs. pre-miR negative control.