Vildagliptin improves high glucose-induced endothelial mitochondrial dysfunction via inhibiting mitochondrial fission

Abstract

The dipeptidyl peptidase 4 inhibitor vildagliptin (VLD), a widely used anti-diabetic drug, exerts favourable effects on vascular endothelium in diabetes. We determined for the first time the improving effects of VLD on mitochondrial dysfunction in diabetic mice and human umbilical vein endothelial cells (HUVECs) cultured under hyperglycaemic conditions, and further explored the mechanism behind the anti-diabetic activity. Mitochondrial ROS (mtROS) production was detected by fluorescent microscope and flow cytometry. Mitochondrial DNA damage and ATP synthesis were analysed by real time PCR and ATPlite assay, respectively. Mitochondrial network stained with MitoTracker Red to identify mitochondrial fragmentation was visualized under confocal microscopy. The expression levels of dynamin-related proteins (Drp1 and Fis1) were determined by immunoblotting. We found that VLD significantly reduced mtROS production and mitochondrial DNA damage, but enhanced ATP synthesis in endothelium under diabetic conditions. Moreover, VLD reduced the expression of Drp1 and Fis1, blocked Drp1 translocation into mitochondria, and blunted mitochondrial fragmentation induced by hyperglycaemia. As a result, mitochondrial dysfunction was alleviated and mitochondrial morphology was restored by VLD. Additionally, VLD promoted the phosphorylation of AMPK and its target acetyl-CoA carboxylase in the setting of high glucose, and AMPK activation led to a decreased expression and activation of Drp1. In conclusion, VLD improves endothelial mitochondrial dysfunction in diabetes, possibly through inhibiting Drp1-mediated mitochondrial fission in an AMPK-dependent manner.

Keywords: AMPK; Drp1; mitochondrial dysfunction; mitochondrial fission; mitochondrial reactive oxygen species; vildagliptin.

Figure 1

Vildagliptin (VLD) alleviates ROS, especially mtROS in the endothelium under diabetic conditions. A‐B, Effect of VLD on vascular ROS and mtROS in mice. Freshly isolated aortic rings were incubated with 10 μM DCF‐DA or 5 μM MitoSox Red at 37°C for 1 h. After cutting into segments with the thickness of 10 μm and staining with DAPI, images were captured by a fluorescent microscope at ×200 magnification. C‐F, Effect of VLD on ROS and mtROS in HUVECs. Cells were incubated with 10 μM DCF‐DA or 5 μM MitoSox Red for 30 min at 37°C and then captured using a fluorescent microscope or analysed by flow cytometry. G, Effect of VLD on 8‐OHdG and 3‐NT levels in mice aortas detected by immunohistochemical staining. Data shown are mean ± SD (n = 3‐6). *P<0.05 vs the normal group; ^# P < 0.05 vs the HG group or the diabetic control group. VLD, vildagliptin; ROS, reactive oxygen species; mtROS, mitochondrial reactive oxygen species; HUVECs, human umbilical vein endothelial cells; HG, high glucose; Man, mannitol

Figure 2

VLD promotes NO generation and expression of peNOS in the endothelium under diabetic conditions. Serum or cell supernatant was collected and total NO production was measured by a modified Griess reaction method. The activation of eNOS was evaluated by assaying eNOS phosphorylation by Western blot. A‐B, Effects of VLD on NO production and eNOS activation in the aortas of mice. C‐D, Effect of VLD on NO production and eNOS activation in HUVECs. Data shown are mean ± SD (n = 3‐6). *P < 0.05 vs the normal group; ^# P < 0.05 vs the HG group or the diabetic control group. Cell supernatant was collected for total NO production. VLD, vildagliptin; NO, nitric oxide; eNOS, endothelial nitric oxide synthase; peNOS, phosphorylation of eNOS at Ser1177; HUVECs, human umbilical vein endothelial cells; HG, high glucose; Man, mannitol

Figure 3

VLD improves mitochondrial dysfunction and mitochondrial fragmentation induced by high glucose in endothelial cells. A, Effect of VLD on mtDNA damage in HUVECs under hyperglycemic conditions. Total DNA was extracted and mitochondrial DNA damage determined by quantitative PCR in HUVECs. B, Effect of VLD on ATP production in HUVECs in response to hyperglycemia. ATP detection was performed by an ATPlite assay. C, Effect of VLD on mitochondrial morphology in HUVECs. Cells were stained with MitoTracker Red for 20 min at 37°C and images captured using a confocal microscopy. D‐E, Quantitative image analysis of mitochondrial interconnectivity and mitochondrial elongation by Image J. Data shown are mean ± SD (n = 3). ^* P < 0.05 vs the normal group; ^# P < 0.05 vs the HG group. VLD, vildagliptin; mtDNA, mitochondrial DNA; HUVECs, human umbilical vein endothelial cells; HG, high glucose; Man, mannitol; LM, local magnification

Figure 4

VLD inhibits high glucose‐induced mitochondrial fission by regulation of dynamin‐related proteins. A, Effects of VLD on the expression of Drp1 and Fis1 in mice aortas. B, Effect of vildagliptin on the expression of Drp1 and Fis1 in HUVECs. C, Effect of VLD on Drp1 recruitment from cytoplasm to mitochondria. Data shown are mean ± SD (n = 3‐6). *P<0.05 vs the normal group; ^# P < 0.05 vs the HG group or the diabetic control group. VLD, vildagliptin; HUVECs, human umbilical vein endothelial cells; HG, high glucose; Man, mannitol

Figure 5

VLD inhibits Drp1 activation via the regulation of AMPK activation. A‐B, Effects of VLD on the phosphorylation of AMPK, ACC and Drp1 in mice and in HUVECs. The levels of pAMPK, pACC and pDrp1 were evaluated by Western blot. C, AMPK activation was essential for Drp1 expression and activation in HUVECs. The phosphorylation levels of AMPK and ACC as well as Drp1 and pDrp1 expression in HUVECs were evaluated by Western blot following cell exposure to GSK621, a specific AMPK activator. Data shown represent mean ± SD (n = 3‐6). *P<0.05 vs the normal group; ^# P < 0.05 vs the HG group or the diabetic control group. VLD, vildagliptin; HUVECs, human umbilical vein endothelial cells; HG, high glucose; Man, mannitol; pAMPK, phosphorylation of AMPK at Thr172; pACC, phosphorylation of ACC at Ser79; pDrp1, phosphorylation of Drp1 at Ser637

Figure 6

Inhibition of Drp1 attenuates ROS and mtROS production and increases NO generation in diabetic mice. MtROS, ROS, and NO production in vascular were determined as described in Figures 1 and 2. A‐D, Effect of mdivi‐1 on vascular mtROS, ROS, and NO production in mice. E, Effects of Drp1 knockdown on the levels of 8‐OHdG and 3‐NT in mice aortas. Data shown represent mean ± SD (n = 6). *P < 0.05 vs the diabetic control group. ROS, reactive oxygen species; mtROS, mitochondrial reactive oxygen species; NO, nitric oxide; eNOS, endothelial nitric oxide synthase; peNOS, phosphorylation of eNOS at Ser1177

Figure 7

Drp1 knockdown attenuates ROS and mtROS production and increases NO generation in HG‐treated HUVECs. MtROS, ROS and NO production in HUVECs were determined as described in Figures 1 and 2. Data shown represent mean ± SD (n = 3). *P < 0.05 vs the HG group. ROS, reactive oxygen species; mtROS, mitochondrial reactive oxygen species; HUVECs, human umbilical vein endothelial cells; NO, nitric oxide; eNOS, endothelial nitric oxide synthase; peNOS, phosphorylation of eNOS at Ser1177; HG, high glucose; NC, negative control

Figure 8

Inhibition of Drp1 improves mitochondrial dysfunction and mitochondrial fragmentation induced by high glucose in HUVECs. A, Effects of Drp1 knockdown on mtDNA damage in HUVECs under hyperglycemic conditions. Total DNA was extracted and mitochondrial DNA damage was evaluated by quantitative PCR in HUVECs. B, Effects of Drp1 knockdown on ATP production in HUVECs under hyperglycaemic conditions. ATP detection was performed by an ATPlite assay. C, Effects of Drp1 knockdown on mitochondrial morphology in HUVECs. Mitochondrial staining and imaging were performed as described in Figure 4. D‐E, Quantitative image analysis of mitochondrial interconnectivity and mitochondrial elongation by Image J. Data shown are mean ± SD (n = 3), *P < 0.05 vs the HG group. mtDNA, mitochondrial DNA; HUVECs, human umbilical vein endothelial cells; HG, high glucose; NC, negative control; LM, local magnification