Endothelial derived miRNA-9 mediated cardiac fibrosis in diabetes and its regulation by ZFAS1

Abstract

Diabetic cardiomyopathy (DCM) is one of the most prevalent causes of morbidity and mortality in diabetic patients. Hyperglycemia induces increased expression/deposition of extracellular matrix (ECM) proteins including fibronectin (FN) and collagen (Col) and plays an important role in fibrosis in diabetic cardiomyopathy (DCM). The roles of RNAs including microRNA (miRNA) and long non-coding RNAs (lncRNA) have begun to be understood in many conditions. In this study, we investigated the role of a specific miRNA, miR-9, and its interactions with lncRNA ZFAS1 in mediating fibrosis in DCM. Treatment with 25 mM glucose (HG) decreased miR-9 expression and increased expressions of ZFAS1, ECM proteins and inflammatory markers, compared to 5 mM glucose (NG) in the HCMECs by using qRT-PCR. Glucose-induced upregulation of ECM proteins can be prevented by ZFAS1 siRNA or miR-9 mimic transfection. Luciferase assay was confirmed miR-9 binding to FN 3'-UTR. miR-9 expression can be regulated by ZFAS1 through polycomb repressive complex 2 (PRC2) components using RNA immunoprecipitation (RIP) and chromatin immunoprecipitation (ChIP) assays. In the in vivo experiment, hyperglycemia-induced the ECM production can be prevented by the miR-9 overexpression in the fibrosis in DCM. These studies showed a novel glucose-induced molecular mechanism in which ZFAS1 participates in the transcriptional regulation of ECM protein production in diabetes through miR-9.

Fig 1. miR-9 regulates glucose-induced production of ECM in vitro in HCMEC.

RT-qPCR analyses indicating HG-induced downregulation of (A) miR-9 transcript, (B, C, D) ECM transcripts (FN, Col4a1, Col1a1) and (E) FN protein expression in HCMECs. Similar glucose induced effects were also seen regarding proinflammatory markers (F), IL-1 β, (G) IL-6, and H) NF-kB1. Such glucose-induced increases of these ECM were prevented following (B-H) miR-9 mimic transfection Luciferase reporter assay demonstrated the interaction between miR-9 and FN(I). (mRNA expressed as a ratio to β-actin; *p < 0.05, compared to Scr HG; Scr = scrambled mimic; NG = 5 mM D-glucose; HG = 25 mM D-glucose; M = 25 mM mannitol; Vwt = wild pMIR vector; Vmut = mutant pMIR vector; n = 6).

Fig 2. Effects of ZFAS1 silencing on mRNA levels of ECM and proinflammatory factors in HCMECs.

(A) ZFAS expression was increased in HG along with B) reduced miR9 expression., Treatment of HCMECs with HG also increased the mRNA expression of ECM proteins C) FN and D) Col 4a1, proinflammatory factors markers E) NF-kB1 (F), IL-1 β and (G) IL-6, compared with NG along with H) IL-6 protein levels. Suppression of ZFAS1 reversed these alterations (B-H). Rescue experiments showed that (C-G) ZFAS1siRNA induced recovery can be reversed by miR-9 antagomir (A-E). (RNA data are expressed as a ratio to β-actin; mean ± SEM. *p<0.05 vs NG+Scr; ξ p<0.05 vs HG+Scr. NG = 5 mM D-glucose; HG = 25 mM D-glucose; M = 25 mM mannitol; Scr = scrambled siRNA; si = siRNA; 9(i) = miR-9 antagomir; n = 6).

Fig 3. ZFAS1–PRC2 complex interaction regulating miR-9 expression in vitro in HCMEC.

Treatment of HCMECs with methylation inhibitors (DZNep) reduced HG induced mRNA upregulation of (A, B, C) EZH2, EED and SUZ12. In conjunction, DZNep also corrected HG induced decreased (D) miR-9 levels. Interestingly, transfection of HCMECs with ZFAS1 siRNA reduced HG induced (E, F) upregulation of EZH2 and EED mRNA, whereas(G) SUZ12 mRNA remained unchanged. (H) RIP assay using anti-EZH2 antibody performed in HCMECs showed glucose-induced increased binding of ZFAS1 with EZH2 (anti-IgG as negative control). (I) ChIP assay demonstrated increased binding of EZH2 on miR-9 promoter as measured by RT-PCR. (NG = 5 mM glucose; HG = 25 mM glucose; RIP = RNA immunoprecipitation; ChIP = chromatin immunoprecipitation RNA; D = DZNep; data expressed as a ratio to β-actin, mean ± SEM; *P < 0.05 compared to NG; ξ < 0.05 compared to HG; n = 6/group).

Fig 4. EC-specific miR-9 transgene (TG) alleviates diabetes-induced ECM (FN, Col1a1) production in vivo.

A) RT-qPCR analyses of the cells from the cardiac tissues of the transgenic mice confirmed 2.5-fold increased miR-9 expression in the ECs but not in other cell types. Following two months of diabetes, reduced miR-9 were seen in the B) whole heart and in the C) ECs isolated from wild type diabetic mouse heart compared to the wild type non-diabetic controls, in association with increased expressions of (D) ZFAS1, ECM transcripts E) FN and F) Col1a1. Such diabetes induced ECM changes were prevented (E, F) in the miR-9 TG mice with diabetes. (G-J) hematoxylin and eosin and (K-N) corresponding trichrome staining showed increased focal fibrosis (arrow, green stain in L) in the hearts of wild diabetic animals compared with K) wild type control animals. Such diabetes-induced changes were not seen in the M.N) TG mice with or without diabetes. (Data expressed as a ratio to β-actin normalised to control; *P < 0.05, compared to B6-Co or Co-ECs; §P < 0.05, compared to B6-Di; B6-Co = Wild type non-diabetic controls; B6-Di = Wild type diabetic; miR9-Co = miR-9 transgenic control; and miR9-Di = miR-9 transgenic diabetic. MH = mouse heart. Myo = myocytes; fib = fibroblast; EC = endothelial cells. Co = control; Di = diabetic; n = 8/group. Same magnification for all micrographs).

Fig 5. A schematic outline of the mechanism related to ZFAS1 in cardiac fibrosis in DCM.