Loss of MicroRNA-192 promotes fibrogenesis in diabetic nephropathy

Abstract

The role of microRNAs (miRs), which are endogenous RNA oligonucleotides that regulate gene expression, in diabetic nephropathy is unknown. Here, we performed expression profiling of cultured proximal tubular cells (PTCs) under high-glucose and control conditions. We identified expression of 103 of 328 microRNAs but did not observe glucose-induced changes in expression. Next, we performed miR expression profiling in pooled RNA from formalin-fixed, paraffin-embedded tissue from renal biopsies. We studied three groups of patients with established diabetic nephropathy and detected 103 of 365 miRs. Two miRs differed by more than two-fold between progressors and nonprogressors, and 12 miRs differed between late presenters and other biopsies. We noted the greatest change in miR-192 expression, which was significantly lower in late presenters. Furthermore, in individual biopsies, low expression of miR-192 correlated with tubulointerstitial fibrosis and low estimated GFR. In vitro, treatment of PTCs with TGF-beta1 decreased miR-192 expression. Overexpression of miR-192 suppressed expression of the E-Box repressors ZEB1 and ZEB2, thereby opposing TGF-beta-mediated downregulation of E-cadherin. In summary, loss of miR-192 expression associates with increased fibrosis and decreased estimated GFR in diabetic nephropathy in vivo, perhaps by enhancing TGF-beta-mediated downregulation of E-cadherin in PTCs.

Figure 1.

miR profiling in PTCs in shown. HK-2 cells were cultured in 25 mM glucose or control medium for 48 hours before RNA extraction and miR profiling using Ambion mirVana miRNA Bioarray. (A) Expression of 328 human miRs in HK-2 cells is represented on a heat map scale, with dark blue being low/no expression and red being maximum expression. Samples and miRs were grouped according to unsupervised Bicluster Analysis. (B) Volcano plot shows the relationship between change in miR expression between the two groups (25 mM glucose and control) and significance (−log10 of uncorrected P values). Horizontal red line represents P = 0.05 threshold.

Figure 2.

Taqman Low Density Array determination of miR expression in renal biopsy tissue from patients with diabetic nephropathy. Renal biopsies from patients with diabetic nephropathy were separated into three groups on the basis of progression after biopsy: Progressors (change in eGFR >5 ml/min per yr; n = 9), nonprogressors (change in eGFR <5 ml/min per yr; n = 9), and late presenters (eGFR ≤15 ml/min at biopsy; n = 4). After RNA extraction, expression of 365 miRs was quantified in pooled RNA for each of the groups using Taqman Low Density Array V1. Fold expression values are displayed for miRs expressed (threshold cycle <33). (A) Comparison of progressors versus nonprogressors. (B) Comparison of combined earlier stage biopsies (progressors+nonprogressors) versus late presenters.

Figure 3.

miR-192 expression in individual kidney biopsies and its correlation with severity of kidney disease. (A) miR-192 was examined in individual samples pooled for use in miR profiling (nine nonprogressors, nine progressors, and four late presenters) by qRT-PCR, using miR-16 as an endogenous control. Data are means ± SEM. (B) Plot of miR-192 expression versus fibrosis score (Pearson coefficient −0.704; P < 0.001). (C) Plot of miR-192 expression versus four-variable MDRD eGFR (Pearson coefficient 0.768; P < 0.001).

Figure 4.

miR-200b expression in individual kidney biopsies and its correlation with severity of kidney disease are shown. (A) miR-200b was examined in individual samples pooled for use in miR profiling (nine nonprogressors, nine progressors, and 4 late presenters) by qRT-PCR, using miR-16 as an endogenous control. Data are means ± SEM. (B) Plot of miR-192 expression versus fibrosis score (Pearson coefficient 0.321; P = 0.146). (C) Plot of miR-192 expression versus four-variable MDRD eGFR (Pearson coefficient −0.331; P = 0.133).

Figure 5.

In situ hybridization for miR-192 is shown. Sections from two biopsies in each group (nonprogressors, progressors, and late presenters) were stained for miR-192 using a double-digoxigenin–labeled locked nucleic acid probe. No staining was seen in control experiments performed using secondary antibody in the absence of probe (data not shown). Sections were examined and photographed under phase contrast microscopy. Representative ×100 and ×200 images are shown.

Figure 6.

PTC responses to TGF-β were examined by qRT-PCR. (A through L) HK-2 cells were cultured without serum for 48 hours and then incubated with 10 ng/ml TGF-β (■) or control medium (□) for indicated time points (A through F) or 96 hours (G through L). Expression of E-cadherin (A and G), ZEB1 (B and H), ZEB2 (C and I), vimentin (D and J), PAI-1 (E and K), and miR-192 (F and L) was measured by qRT-PCR and normalized to GAPDH expression. miR-192 was normalized to miR-16. Error bars (G through L) represent SEM (n = 3). *P < 0.05; **P < 0.005; ***P < 0.0005.

Figure 7.

Altered E-cadherin and E-Box repressors expression in HK-2 cells overexpressing miR-192 by qRT-PCR is shown. Stable lines overexpressing miR-192 or a control short hairpin (Scrambled) were generated as described in the Concise Methods section. After 48 hours without serum, cells were incubated with TGF-β (■) or control medium (□) for 96 hours. (A through D) Subsequently, expression of miR-192 (A), ZEB1 (B), ZEB2 (C), and E-cadherin (D) was examined by qRT-PCR. miR-192 was normalized to miR-16, and the other genes were normalized to GAPDH. The experiment was performed for the control line and three independent cell lines overexpressing miR-192, all three giving similar results (results from one line are shown). Error bars represent SEM (n = 3).

Figure 8.

Altered E-cadherin expression in cells overexpressing miR-192 by immunoblot and fluorescence microscopy is shown. Stable lines overexpressing miR-192 or a control short hairpin (Scrambled) were growth-arrested in serum-free medium for 48 hours before incubation with 10 ng/ml TGF-β (■) or control medium (□) for 96 hours. (A and B) Then E-cadherin protein was examined by Western blot (A) or fluorescence microscopy (B). The representative results of at least three independent experiments are shown.

Figure 9.

Expression of other EMT markers in HK-2 cells overexpressing miR-192 by qRT-PCR is shown. Stable lines overexpressing miR-192 or a control short hairpin (Scrambled) were growth-arrested in serum-free medium for 48 hours before incubation with 10 ng/ml TGF-β (■) or control medium (□) for 96 hours. (A through F) Expression of PAI-1 (A), vimentin (B), Snail (C), Slug (D), Twist (E), and HMGA2 (F) was examined by qRT-PCR. The expression was normalized to GAPDH. Error bars represent SEM (n = 3).