Hypermethylation and suppression of microRNA219a-2 activates the ALDH1L2/GSH/PAI-1 pathway for fibronectin degradation in renal fibrosis

Abstract

Epigenetic regulations, such as DNA methylation and microRNAs, play an important role in renal fibrosis. Here, we report the regulation of microRNA219a-2 by DNA methylation in fibrotic kidneys, unveiling the crosstalk between these epigenetic mechanisms. Through genome-wide DNA methylation analysis and pyrosequencing, we detected the hypermethylation of microRNA219a-2 in renal fibrosis induced by unilateral ureteral obstruction (UUO) or renal ischemia/reperfusion, which was accompanied by a significant decrease in microRNA-219a-5p expression. Functionally, overexpression of microRNA219a-2 enhanced fibronectin induction during hypoxia or TGF-β1 treatment of cultured renal cells. In mice, inhibition of microRNA-219a-5p suppressed fibronectin accumulation in UUO and ischemic/reperfused kidneys. Aldehyde dehydrogenase 1 family member L2 (ALDH1L2) was identified to be the direct target gene of microRNA-219a-5p in renal fibrotic models. MicroRNA-219a-5p suppressed ALDH1L2 expression in cultured renal cells, while inhibition of microRNA-219a-5p prevented the decrease of ALDH1L2 in injured kidneys. Knockdown of ALDH1L2 enhanced plasminogen activator inhibitor-1 (PAI-1) induction during TGF-β1 treatment of renal cells, which was associated with fibronectin expression. In conclusion, the hypermethylation of microRNA219a-2 in response to fibrotic stress may attenuate microRNA-219a-5p expression and induce the upregulation of its target gene ALDH1L2, which reduces fibronectin deposition by suppressing PAI-1.

Keywords: DNA methylation; fibronectin; kidney fibrosis; microRNA; oxidative stress.

Graphical abstract

Figure 1

Hypermethylation of mir219a-2 is accompanied by decreased mir-219a-5p expression in fibrotic kidneys (A and B) DNA methylation level in C57BL/6J mouse kidneys after 25 min of bilateral ischemia with 1 week reperfusion (I25/1wk), or 25 min of bilateral ischemia with 1 month reperfusion (I25/1mth), or unilateral ureteral obstruction 7 days (UUO7D) was compared with sham control kidneys. Pooled DNA samples from 3 mice/group (I25/1wk, I25/1mth, vs. sham) or 2 mice/group (UUO7D vs. control) were used for genome-wide DNA methylation analysis. (A) The list of differentially methylated microRNA genes. The microRNA genes were identified with >20% methylation level difference in the promotor region compared with control or sham. (B) The methylation level of mir219a-2 gene was examined by genome-wide DNA methylation sequencing and pyrosequencing. The percentage of DNA methylation induction was calculated compared with sham or control. (C) qPCR analysis of mir-219a-5p in control and UUO7D mouse kidneys. ∗p = 0.0058, t = 7.099, df = 3, n = 4, paired t test. (D) qPCR analysis of mir-219a-5p in sham control and I30/1wk mouse kidneys. ∗p = 0.0070, t = 5.102, df = 4, n = 5, paired t test. (E) In situ hybridization of mir-219a-5p in sham control (Sham-7D) or 7 days of UUO mouse kidneys (UUO7D). Scale bar, 0.1 mm. Representative images from two experiments. (F) mir-219a-5p in control BUMPT cells or cells treated with hypoxia (1% O₂) for 24, 48, and 72 h. One-way ANOVA (F = 10.05) with uncorrected Fisher’s LSD for multiple comparisons. ∗p = 0.0051, t = 3.413, df = 12; ∗∗p = 0.0009, t = 4.356, df = 12. (G) mir-219a-5p level in control BUMPT cells or cells treated with 10 ng/mL of TGF-β1 for 24, 48, and 72 h.

Figure 2

Mir219a-2 enhances fibronectin expression during hypoxia or TGF-β1 treatment of renal tubular cells (A and B) BUMPT cells with stable transfection of mir219a-2 or empty vector (control) were treated with (H 72h) or without (N 72h) hypoxia (1% O₂) for 72 h to collect lysate for immunoblot analysis. (A) Representative immunoblots. (B) Densitometry analysis of fibronectin by normalization with cyclophilin b (loading control). n = 5; ∗two-way ANOVA with uncorrected Fisher’s LSD for multiple comparisons, p = 0.0436, t = 2.190, df = 16. (C and D) BUMPT cells with stable transfection of mir219a-2 or empty vector (control) were treated with or without 10 ng/mL TGF-β1 for 72 h to collect lysate for immunoblot analysis. (C) Representative images of immunoblots. (D) Densitometry analysis of fibronectin by normalization with cyclophilin b. n = 6; ∗two-way ANOVA with uncorrected Fisher’s LSD for multiple comparisons, p = 0.0027, t = 3.414, df = 20.

Figure 3

Antagonism of mir-219a-5p suppresses fibronectin expression in UUO-induced renal fibrosis C57BL/6J male mice were treated with negative control (NC) or anti-mir-219a-5p LNAs, and then subjected to unilateral UUO for 2 weeks. The whole lysates of the contralateral (control) and obstructive kidneys (UUO 2 weeks) were examined by immunoblotting. (A) Schematic diagram showing the mouse treatment procedure. (B) qPCR analysis of mir-219a-5p in mouse kidneys. n = 4, ∗unpaired t test, p < 0.0001, t = 108.8, df = 6. (C) Representative immunoblots of fibrotic proteins. A219, anti-mir-219a-5p. (D) Densitometry analysis of fibronectin in mouse kidneys. The value was normalized by cyclophilin b. n = 5, two-way ANOVA with uncorrected Fisher’s LSD for multiple comparisons. ∗p = 0.0001, t = 5, df = 16. (E) Immunohistochemical staining of fibronectin in control and UUO kidneys (UUO 2 weeks). Scale bar, 0.1 mm.

Figure 4

Identification of potential targets of mir-219a-5p HEK293 cells were co-transfected with Flag-Ago2 plasmid and mir-219a-5p mimics or its negative control oligos (NC). RNA samples were extracted from whole cells (total mRNA) or Ago2 immunoprecipitates (RISC mRNA) for deep RNA sequencing (RNA-seq), n = 3. (A) Schematics indicating the procedure to identify potential mir-219a-5p targets. (B) The list of potential targets with predicted mir-219a-5p binding sites in both human and mouse. (C) qPCR to confirm the mRNA levels of potential targets in RISC. n = 3, p values (vs. NC): ∗p = 0.0013, t = 8.100, df = 4; ∗∗p = 0.0030, t = 6.414, df = 4; #p = 0.0088, t = 4.776, df = 4; ##p = 0.0027, t = 6.651, df = 4; Δp = 0.0008, t = 9.178, df = 4. Unpaired t test. (D) qPCR analysis of total RNA samples. The mRNA levels were compared between the cells transfected with mir-219a-5p mimics and those with negative control oligos.

Figure 5

Confirmation of ALDH1L2 as a mir-219a-5p direct target (A) Representative immunoblots of ALDH1L2 in BUMPT cells stably transfected with mir-219a-2 or empty vector (control) with GAPDH as internal loading control. (B) Densitometry analysis of ALDH1L2 in BUMPT cells (normalized by GAPDH). ∗Unpaired t test, p = 0.0013, t = 4.423, df = 10, n = 6. (C) Immunohistochemical staining of ALDH1L2 in mouse kidney samples with negative control (NC) or anti-mir-219-5p LNA (anti-mir-219) treatment (three repeats). Scale bar, 0.1 mm. (D) Immunoblots of ALDH1L2 in HEK293 cells transiently transfected with negative control (NC) or mir-219a mimics with cyclophilin b as internal control. (E) Densitometry analysis of ALDH1L2 in HEK293 cells normalized with cyclophilin b level. ∗Unpaired t test, p = 0.0024, t = 4.361, df = 8, n = 5. (F) Luciferase assay to confirm the binding of mir-219a-5p to the predicted binding site in ALDH1L2 3′ UTR. HEK293 cells were co-transfected with RNA oligos (negative control or mir-219a-5p mimics) and luciferase plasmids without insertion (empty vector) or with insertion (predicted mir-219a-5p binding site of ALDH1L2 [ALDH1L2], or mutated binding site [mutated]) in the 3′ UTR of luciferase. The luciferase ratio between mir-219a-5p vs. negative control was calculated for each group and then normalized by values from empty vector group. One-way ANOVA (F = 5.474) with uncorrected Fisher’s LSD for multiple comparisons. n = 4; ∗p = 0.0173, t = 2.910, df = 9; ∗∗p = 0.0201, t = 2.819, df = 9.

Figure 6

ALDH1L2 suppresses fibronectin protein induction (A–E) HEK293 cells were transfected with ALDH1L2 siRNA (SiRNA) or negative control oligos (NC). (A) Immunoblot analysis verifying ALDH1L2 knockdown in siRNA-transfected cells. (B) Densitometry analysis of ALDH1L2 normalized by cyclophilin b. ∗Unpaired t test, p = 0.0024, t = 4.358, df = 8. (C–E) ALDH1L2 siRNA or negative control oligo-transfected HEK293 cells were treated with 20 ng/mL TGF-β1 for 24 h. (C) Representative immunoblots of fibronectin with cyclophilin b as loading control. (D) Densitometry analysis of fibronectin normalized by cyclophilin b or β-actin. Two-way ANOVA with uncorrected Fisher’s LSD for multiple comparisons. n = 6; ∗p = 0.0021, t = 3.530, df = 20. (E) qPCR analysis of fibronectin mRNA. n = 3. (F–H) BUMPT stable cell lines were established by transfection with negative control (NC) or ALDH1L2 knockdown shRNAs (shRNA). The cells were treated with (tgfb1) or without (NT) 10 ng/mL TGF-β1 for 24 h. (F) Representative immunoblots of ALDH1L2 and fibronectin with cyclophilin b or GAPDH as loading control, respectively. (G) Densitometry analysis of ALDH1L2 in NT condition normalized by cyclophilin b, n = 3. ∗p = 0.0017, t = 23.90, df = 2, paired t test. (H) Densitometry analysis of fibronectin normalized by GAPDH, n = 4. ∗Two-way ANOVA with uncorrected Fisher’s LSD for multiple comparisons. p = 0.0111, t = 3.621, df = 6.

Figure 7

ALDH1L2 promotes GSH accumulation and inhibits PAI-1 (A) GSH concentrations in BUMPT cells stably transfected with mir-219-2 or empty vector (control). ∗Unpaired t test, p = 0.0003, t = 7.681, df = 5.754, n = 4. (B) GSH concentrations in HEK293 cells with negative control (NC) or mir-219a-5p mimic transfection. ∗Unpaired t test, p = 0.0334, t = 3.079, df = 4.311, n = 4. (C) GSH concentrations in HEK293 cells with negative control (NC) or ALDH1L2 siRNA transfection. ∗Unpaired t test, p = 0.0351, t = 2.889, df = 4.895, n = 4. (D) HEK293 cells were transfected with ALDH1L2 siRNA (SiRNA) or negative control sequence (NC). PAI-1 mRNA was examined by qPCR. Two-way ANOVA with uncorrected Fisher’s LSD for multiple comparisons, n = 3. p = 0.004, t = 5.805, df = 8. (E–G) Mice were treated with negative control (NC) or anti-mir-219a-5p LNA and subjected to UUO for 2 weeks. (E) GSH concentrations in mouse kidney tissue, n = 3. Two-way ANOVA with uncorrected Fisher’s LSD for multiple comparisons. ∗p = 0.0055, t = 3.769, df = 8. (F) Percentage of the renal tubular area with significant PAI-1 induction after 2 weeks of UUO. Unpaired t test, p = 0.0498, t = 2.781, df = 4, n = 3. (G) Immunohistochemical images of PAI-1 in mouse kidney samples with or without UUO. Scale bar, 0.2 mm. ∗Renal tubules with significant induction of PAI-1.