Therapeutic knockdown of miR-320 improves deteriorated cardiac function in a pre-clinical model of non-ischemic diabetic heart disease

Abstract

Non-ischemic diabetic heart disease (NiDHD) is characterized by diastolic dysfunction and decreased or preserved systolic function, eventually resulting in heart failure. Accelerated apoptotic cell death because of alteration of molecular signaling pathways due to dysregulation in microRNAs (miRNAs) plays a significant role in the development of NiDHD. Here, we aimed to determine the pathological role of cardiomyocyte-enriched pro-apoptotic miR-320 in the development of NiDHD. We identified a marked upregulation of miR-320 that was associated with downregulation of its target protein insulin growth factor-1 (IGF-1) in human right atrial appendage tissue in the late stages of cardiomyopathy in type 2 diabetic db/db mice and high-glucose-cultured human ventricular cardiomyocytes (AC-16 cells). In vitro knockdown of miR-320 in high-glucose-exposed AC-16 cells using locked nucleic acid (LNA) anti-miR-320 markedly reduced high-glucose-induced apoptosis by restoring IGF-1 and Bcl-2. Finally, in vivo knockdown of miR-320 in 24-week-old type 2 diabetic db/db mice reduced cardiomyocyte apoptosis and interstitial fibrosis while restoring vascular density. This resulted in partial recovery of the impaired diastolic and systolic function. Our study provides evidence that miR-320 is a late-responding miRNA that aggravates apoptosis and cardiac dysfunction in the diabetic heart, and that therapeutic knockdown of miR-320 is beneficial in partially restoring the deteriorated cardiac function.

Keywords: apoptosis; fibrosis; insulin growth factor-1; locked nucleic acid; miR-320; microRNA; microangiopathy; non-ischemic diabetic heart disease.

Graphical abstract

Figure 1

Diabetes upregulates miR-320 in the human heart (A) Quantitative scatterplot bar graph showing miR-320 expression in the RAA tissue by RT-PCR analysis. Samples were collected from diabetic (D) and non-diabetic (ND) patients with ischemic heart disease (IHD) undergoing coronary artery bypass graft surgery. Data are mean ± SEM and expressed as relative DCT. (B–D) Representative western blots and quantitative scatterplot bar graphs showing the expression of IGF-1 (B), Bcl-2 (C), and cleaved caspase-3 (D) in the study groups. Data are represented as the ratio to total protein and are mean ± SEM. Each western blot analysis was repeated three independent times. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

Figure 2

miR-320 upregulates in the later stage of diabetes (A) Quantitative scatterplot bar graph showing miR-320 expression in the heart by RT-PCR analysis. Samples were collected from obese diabetic db/db mice and lean non-diabetic db/+ mice at different time points. Data are mean ± SEM and expressed as relative DCT. (B–D) Representative western blots and quantitative scatterplot bar graphs showing the expression of IGF-1 (B), Bcl-2 (C), and cleaved caspase-3 (D) at different time points in the study groups. Data are represented as the ratio to total protein and are mean ± SEM. Each western blot analysis was repeated three independent times. Bcl-2 and cleaved caspase-3 were tested only at 28 and 32 weeks as the changes in miR-320 and the upstream target IGF-1 was not observed until 28 weeks of age. Other time points for IGF-1 are shown in Figure S2A. ∗p < 0.05 and ∗∗∗p < 0.001. The same image was used in (B) and (C) for total proteins as the membrane used to probe IGF-1 (B) and Bcl-2 (C).

Figure 3

High glucose induces upregulation of miR-320 in cultured human cardiomyocytes (A) Quantitative scatterplot bar graph showing miR-320 expression in the normal glucose (NG) and high glucose (HG)-treated AC-16 human cardiomyocytes by RT-PCR analysis. Samples were collected at different points after exposing the cells to HG or mannitol for osmotic control (NG). Data are mean ± SEM and expressed as relative DCT. (B–D) Representative western blots and quantitative scatterplot bar graphs showing the expression of IGF-1 (B), Bcl-2 (C), and cleaved caspase-3 (D) at different time points in the study groups. Data are represented as the ratio to total protein and are mean ± SEM. Each western blot analysis was repeated three independent times. Bcl-2 and cleaved caspase-3 were tested only at 120 and 140 h of HG exposure as the changes in miR-320 and the upstream target IGF-1 was not observed until 96 h. Other time points for IGF-1 are shown in Figure S2A. (E) Quantitative scatterplot bar graphs showing caspase-3/7 activity after normalizing the cell numbers using CyQUANT assay. Data are represented as arbitrary units (a.u.) and are mean ± SEM. (F) Quantitative scatterplot bar graphs showing absolute number of cells by CyQUANT assay and represented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Figure 4

Inhibition of miR-320 ameliorated apoptosis in HG-treated human ventricular cardiomyocytes (A) Quantitative scatterplot bar graph showing the expression of miR-320 by RT-PCR analysis after transfection of AC-16 cells either LNA-precursor miR-320 (LNA-miR-320) to knock down miR-320 expression or scrambled sequence (Scrambled) as the control in both study groups. Data are mean ± SEM and expressed as relative DCT. (B–D) Representative western blots and quantitative scatterplot bar graphs showing the expression of IGF-1 (B), Bcl-2 (C), and cleaved caspase-3 (D) in both the study groups after treatment with LNA-mIR-320 or scrambled sequence. Data are represented as the ratio to total protein and are mean ± SEM. Each western blot analysis was repeated three independent times. (E) Quantitative scatterplot bar graphs showing caspase-3/7 activity after normalizing the cell numbers using CyQUANT assay. Data are represented as a.u. and are mean ± SEM. (F) Quantitative scatterplot bar graphs showing absolute number of cells by CyQUANT assay and represented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗∗p < 0.001.

Figure 5

Therapeutic inhibition of miR-320 prevented adverse cardiac remodeling in type 2 diabetic db/db mice (A) Quantitative scatterplot bar graph showing the expression of miR-320 expression by RT-PCR analysis after treatment of type 2 diabetic (db/db) mice with either LNA-precursor miR-320 (LNA-miR-320) to knock down miR-320 expression or scrambled sequence (Scrambled) as the control. Lean db/+ mice were only injected with a scrambled sequence. Data are mean ± SEM and expressed as relative DCT. (B–D) Representative western blots and quantitative scatterplot bar graphs showing the expression of IGF-1 (B), Bcl-2 (C), and cleaved caspase-3 (D) in all the study groups after treatment with LNA-mIR-320 or scrambled sequence. Data are represented as the ratio to total protein and are mean ± SEM. Each western blot analysis was repeated three independent times. (E) Representative fluorescent microscopy images and the quantitative scatterplot bar graphs showing the number of TUNEL positive cardiomyocytes (arrowhead) per field. Data are mean ± SEM. (F and G) Representative confocal microscopy images and the quantitative scatterplot bar graphs show the capillaries (F, green) and arterioles (G, red) among the study groups. Data are represented as a number of capillaries (F) and arterioles (G) per mm² and are mean ± SEM. (H) Representative microscopy images captured with polarized lens and the quantitative scatterplot bar graphs showing the percentage of fibrotic area per field among the study groups. Data are mean ± SEM. Five random images were taken at 200× magnification from each section, and three sections were used for each sample. Scale bars, 100 μm. ∗p < 0.05 and ∗∗p < 0.01. n = at least 4 animals in each group.

Figure 6

Therapeutic inhibition of miR-320 improved cardiac function (A–I) Quantitative line graphs showing cardiac functions measured by echocardiography in all the study groups. Data are mean ± SEM. LVAWs, left ventricular anterior wall during systole; LVAWd, LVAW during diastole; LVPWs, LV posterior wall during systole; LVPWs, LVPW wall during diastole; LVIDs, LV internal diameter during systole; LVIDd, LVID during diastole; FS, fractional shortening; EF, ejection fraction. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001 versus non-diabetic (ND) scrambled sequence (Scr)-treated group of the corresponding age; #p < 0.05 and ##p < 0.01 versus diabetic (D) scrambled (Scr)-treated group; φp < 0.05, φφp<0.01, φφφp<0.001, and φφφφp<0.0001 versus 24 weeks (wks) age time point; δδp<0.01 and δδδδp<0.0001 versus 28 weeks time point. n = 7 animals in each group.