Overexpression of hexokinase 2 reduces mitochondrial calcium overload in coronary endothelial cells of type 2 diabetic mice

Abstract

Coronary microvascular rarefaction, due to endothelial cell (EC) dysfunction, is one of the causes of increased morbidity and mortality in diabetes. Coronary ECs in diabetes are more apoptotic due partly to mitochondrial calcium overload. This study was designed to investigate the role of hexokinase 2 (HK2, an endogenous inhibitor of voltage-dependent anion channel) in coronary endothelial dysfunction in type 2 diabetes. We used mouse coronary ECs (MCECs) isolated from type 2 diabetic mice and human coronary ECs (HCECs) from type 2 diabetic patients to examine protein levels and mitochondrial function. ECs were more apoptotic and capillary density was lower in the left ventricle of diabetic mice than the control. MCECs from diabetic mice exhibited significant increase in mitochondrial Ca²⁺ concentration ([Ca²⁺]_mito) compared with the control. Among several regulatory proteins for [Ca²⁺]_mito, hexokinase 1 (HK1) and HK2 were significantly lower in MCECs from diabetic mice than control MCECs. We also found that the level of HK2 ubiquitination was higher in MCECs from diabetic mice than in control MCECs. In line with the data from MCECs, HCECs from diabetic patients showed lower HK2 protein levels than HCECs from nondiabetic patients. High-glucose treatment, but not high-fat treatment, significantly decreased HK2 protein levels in MCECs. HK2 overexpression in MCECs of diabetic mice not only lowered the level of [Ca²⁺]_mito, but also reduced mitochondrial reactive oxygen species production toward the level seen in control MCECs. These data suggest that HK2 is a potential therapeutic target for coronary microvascular disease in diabetes by restoring mitochondrial function in coronary ECs.

Keywords: HK2; ROS; mitochondrial Ca2+ overload.

Fig. 1.

Endothelial cell apoptosis is increased and capillary density is reduced in the left ventricle (LV) myocardium of type 2 diabetic (T2D) mice. A: representative photographs showing capillaries (green), apoptotic cells (red), and apoptotic endothelial cells (ECs; orange with arrow) in the LV of control and T2D mice. Bar = 50 μm. B: averaged data showing capillary density in control (Cont, n = 6) and T2D (n = 6) LV. C: averaged data showing the apoptotic ECs in control (n = 6) and T2D (n = 6) LV. Data are expressed as means ± SE. *P < 0.05 vs. Cont.

Fig. 2.

Mitochondrial Ca²⁺ concentration ([Ca²⁺]_mito) is increased in mouse coronary endothelial cells (MCECs) isolated from type 2 diabetic (T2D) mice. A: representative images showing [Ca²⁺]_mito in primary cultured MCECs isolated from control (Cont) and T2D mice. Bars, 10 μm. B: summarized data showing average [Ca²⁺]_mito in primary cultured MCECs from control mice (n = 102), and T2D mice (n = 82). Data are means ± SE. *P < 0.05 vs. Cont.

Fig. 3.

Voltage-dependent anion channel (VDAC), mitochondrial Ca²⁺ uniporter (MCU), hexokinase 1 (HK1), and 2 (HK2) protein levels, and ubiquitination levels of HK2 in MCECs from control (Cont) and type 2 diabetic (T2D) mice. A–D: top images are typical Western blot images of protein levels in mouse coronary endothelial cells (MCECs) freshly isolated from control and T2D mice. Actin was used as a loading control. Bottom columns show the summarized data of focus protein levels normalized to actin. A: VDAC protein levels in MCECs of control and T2D mice (n = 6 in each group). B: MCU protein levels in MCECs of control and T2D mice (n = 5 in each group). C: HK1 protein levels in MCECs from control (n = 5) and T2D (n = 6) mice. D: HK2 protein levels in MCECs from control (n = 6) and T2D (n = 7) mice. Data are means ± SE. *P < 0.05 vs. Cont. E: top image shows the typical image of immunoprecipitation (IP) with anti-ubiquitin (Ub) antibody and immunoblotting (IB) with anti-HK2 antibody in coronary endothelial cells (ECs) from control and T2D mice. Bottom image shows IgG heavy chain (IgG-HC) signal from each sample as antibody/beads loading control. Bottom column shows summarized data of ubiquitinated HK2 protein normalized to IgG-HC signal. Data are means ± SE; n = 6 in each group. *P < 0.05 vs. Cont.

Fig. 4.

Effect of high-glucose (HG) and high-fat (HF) treatment on hexokinase 2 (HK2) protein levels in endothelial cells (ECs) and protein level of HK2 in human coronary endothelial cells (HCECs) isolated from control (Cont) and type 2 diabetic (T2D) patients. A: HK2 protein levels (n = 9 in each group) were measured in mouse coronary endothelial cell line (MCEC-CL) treated with normal glucose (NG; 5 mM glucose, open bar) or HG (25 mM glucose, solid bar) for 96 h. Actin was used as a loading control. Data are means ± SE. *P < 0.05 vs. NG. B: HK2 protein levels (n = 6 in each group) were measured in ECs treated with vehicle [normal free fatty acid (NF), 0 mM palmitic acid, open bar] or free fatty acid (HF, 150 mM palmitic acid, solid bar) for 24 h. GAPDH was used as a loading control because actin was significantly affected by free fatty acid (FFA) treatment. Data are means ± SE. *P < 0.05 vs. NF. C: top images are typical Western blot images of HK2 protein levels in HCECs from control and T2D patients. Actin was used as a loading control. Bottom columns show the summarized data of focus protein levels normalized to actin (n = 8). Data are means ± SE. *P < 0.05 vs. Cont.

Fig. 5.

Hexokinase 2 (HK2) overexpression decreases mitochondrial Ca²⁺ concentration ([Ca²⁺]_mito) in diabetic mouse coronary endothelial cells (MCECs). A: HK2-adenovirus (Adv) transduction in mouse coronary endothelial cell line (MCEC-CL) increases HK2 protein level in a dose-dependent manner. Western blots show HK2 and actin protein levels (top). Actin was used as a loading control. Bottom columns show HK2 protein level normalized to actin. Endothelial cells (ECs) with control Adv (Cont-Adv), n = 4; ECs with HK2-Adv, n = 4. Data are means ± SE. *P < 0.05 vs. Cont-Adv. B: a typical recording of [Ca²⁺]_mito in MCECs at rest and after treatment with the sarco/endoplasmic reticulum Ca²⁺-ATPase inhibitor cyclopiazonic acid (CPA; 10 μmol/l), in the absence or presence of extracellular Ca²⁺. C–G: summarized [Ca²⁺]_mito data. Control ECs transduced with control Adv (Cont-EC, Cont-Adv), n = 72; diabetic ECs transduced with control Adv. (T2D-EC, Cont-Adv), n = 77; diabetic ECs transduced with HK2-Adv (T2D-EC, HK2-Adv), n = 56. At least 4 mice/group were used to isolate MCECs. Data are means ± SE *P < 0.05 vs. Cont-EC, Cont-Adv. ^#P < 0.05 vs. T2D-EC, Cont-Adv.

Fig. 6.

Hexokinase 2 (HK2)-adenovirus (Adv) transduction in diabetic mouse coronary endothelial cells (MCECs) decreases mitochondrial reactive oxygen species concentration ([ROS]_mito) toward the level of control (Cont) MCECs. A: representative images showing [ROS]_mito (red) and mitochondrial structure (green) in MCECs isolated from control and type 2 diabetic (T2D) mice transduced with Cont-Adv or HK2-Adv. Bar = 10 μm. B: summarized data of [ROS]_mito. Cont-EC, Cont-Adv, n = 140; T2D-EC, Cont-Adv, n = 139; T2D-EC, HK2-Adv, n = 73. At least 4 mice/group were used to isolate MCECs. Data are means ± SE. *P < 0.05 vs. Cont-EC, Cont-Adv. ^#P < 0.05 vs. T2D-EC, Cont-Adv.