MicroRNA-146a decreases high glucose/thrombin-induced endothelial inflammation by inhibiting NAPDH oxidase 4 expression

Abstract

Diabetes is associated with hyperglycemia and increased thrombin production. However, it is unknown whether a combination of high glucose and thrombin can modulate the expression of NAPDH oxidase (Nox) subtypes in human aortic endothelial cells (HAECs). Moreover, we investigated the role of a diabetes-associated microRNA (miR-146a) in a diabetic atherothrombosis model. We showed that high glucose (HG) exerted a synergistic effect with thrombin to induce a 10.69-fold increase in Nox4 mRNA level in HAECs. Increased Nox4 mRNA expression was associated with increased Nox4 protein expression and ROS production. Inflammatory cytokine kit identified that the treatment increased IL-8 and IL-6 levels. Moreover, HG/thrombin treatment caused an 11.43-fold increase of THP-1 adhesion to HAECs. In silico analysis identified the homology between miR-146a and the 3'-untranslated region of the Nox4 mRNA, and a luciferase reporter assay confirmed that the miR-146a mimic bound to this Nox4 regulatory region. Additionally, miR-146a expression was decreased to 58% of that in the control, indicating impaired feedback restraint of HG/thrombin-induced endothelial inflammation. In contrast, miR-146a mimic transfection attenuated HG/thrombin-induced upregulation of Nox4 expression, ROS generation, and inflammatory phenotypes. In conclusion, miR-146a is involved in the regulation of endothelial inflammation via modulation of Nox4 expression in a diabetic atherothrombosis model.

Figure 1

Endothelial Nox4 expression and intracellular ROS production were significantly induced by HG/thrombin. (a) HAECs were stimulated with HG (25 mmol/L), L-glucose (25 mmol/L), and thrombin (2 U/mL) for 5 h. Real-time PCR showed that stimulation of HAECs with D-glucose/thrombin induced a 10.69-fold synergistic increase of Nox4 mRNA levels compared with the control, while L-glucose/thrombin had no such effect. The bars represent the means ± SEM from 3 experiments. *P < 0.05, as compared with the control. **P < 0.01, as compared with the control. (b) Stimulation of HAECs with HG/thrombin for 12 h increased Nox4 protein expression 1.35-fold compared with the control; alternatively, L-glucose/thrombin had no discernible effect on Nox4 expression. The bars represent the means ± SEM from 4 experiments. *P < 0.05, as compared with the control. (c) Stimulation of HAECs with HG/thrombin for 5 h significantly increased ROS production 2.81-fold compared with the control. The bars represent the means ± SEM from 3 experiments. *P < 0.05, as compared with the control. **P < 0.01, as compared with the control.

Figure 2

Endothelial inflammation was induced by HG/thrombin through an ROS-dependent pathway. (a) From top to bottom, multiple small dots accumulated into 6 horizontal lines, which represented the expression levels of IL-8, IL-1β, IL-6, IL-10, TNF-α, and IL-12p70 proteins in conditioned medium. The rightward shift of the dotted lines represents the relative protein expression levels. Analysis using an inflammatory cytokine kit showed that HG/thrombin stimulation increased IL-8 and IL-6 expression in HAECs, while L-1β, IL-10, TNF-α, and IL-12p70 had no discernible effect on its protein expression levels. The images shown represent the results from 3 independent experiments. (b) Stimulation of HAECs with HG/thrombin for 5 h increased caused 4.06-fold and 5.05-fold increases of IL-6 and IL-8 mRNA gene expression levels, respectively. The bars represent the means ± SEM from 3 experiments. *P < 0.05, as compared with the control. (c) Stimulation of HAECs with HG/thrombin for 5 h induced an 11.43-fold synergistic increase of THP-1 adhesiveness to HAECs compared with the control. The bars represent the means ± SEM from 4 experiments. **P < 0.01, as compared with the control. (d) Stimulation of HAECs with HG/thrombin for 5 h induced significant 8.72-, 11.67-, and 67.21-fold increases of VCAM-1, ICAM-1, and SELE gene expression, respectively. The bars represent the means ± SEM from 3 experiments. *P < 0.05, as compared with the control. (e) HAECs were pretreated with 5 mM N-acetylcysteine (NAC) for 1 h and then stimulated with HG/thrombin for 5 h. Real-time PCR showed that NAC pretreatment caused significant decreases in the gene expression of VCAM-1, ICAM-1, and SELE to 51%, 70%, and 49% of that in the HG/thrombin-stimulated HAECs. The bars represent the means ± SEM from 4 experiments. *P < 0.05 and **P < 0.01, as compared with the HG/thrombin.

Figure 3

Nox4 was identified as a target of miR-146a. (a) Bioinformatic miR target analysis identified homology between miR-146a and the 3′-UTR of the human Nox4 mRNA, indicating potential regulation of Nox4 by miR-146a. (b) A luciferase reporter assay showed that the miR-146a mimic could downregulate relative luciferase activity of pGL2-Nox4-3′-UTR compared with the miR-control. The bars represent the means ± SEM from 6 experiments. **P < 0.01, as compared with the miR-control.

Figure 4

HG/thrombin downregulated miR-146a expression in HAECs. miR-146a expression was measured by real-time PCR. HG/thrombin stimulation for 5 h decreased miR-146a expression to 58% of that in the control. The bars represent the means ± SEM from 5 experiments. *P < 0.05 and **P < 0.01, as compared with the control. miR levels are expressed as the ratio of RNU6B levels.

Figure 5

The stimulatory effects of HG/thrombin on Nox4 expression and ROS production were attenuated in miR-146a mimic-transfected HAECs. (a) The stimulatory effect of HG/thrombin on Nox4 mRNA expression was preserved in miR-control-transfected HAECs, with an increase of 8.95-fold Nox4 mRNA levels compared with the control. In contrast, HG/thrombin did not increase Nox4 mRNA expression in miR-146a mimic-transfected HAECs. The bars represent the means ± SEM from 3 experiments. *P < 0.05, as compared with the miR-control. (b) The stimulatory effect of HG/thrombin on Nox4 protein expression was significantly attenuated in miR-146a mimic-transfected HAECs. The bars represent the means ± SEM from 4 experiments. **P < 0.01, as compared with the miR-control. (c) The stimulatory effect of HG/thrombin on intracellular ROS production was significantly attenuated in miR-146a mimic-transfected, HG/thrombin-stimulated HAECs. The bars represent the means ± SEM from 3 experiments. *P < 0.05, as compared with the miR-control.

Figure 6

The proinflammatory phenotypes of HG/thrombin-stimulated HAECs were attenuated by a miR-146a mimic. (a) Analysis using an inflammatory cytokine kit showed that the stimulatory effect of HG/thrombin on IL-8 and IL-6 protein expression in conditioned mediums was preserved in miR-control-transfected HAECs, while the stimulatory effect was attenuated in miR-146a mimic-transfected, HG/thrombin-stimulated HAECs. The images shown represent the results from 3 independent experiments. (b) The stimulatory effect of HG/thrombin on THP-1 adhesion to HAECs was preserved in miR-control-transfected HAECs, while the THP-1 adhesiveness was attenuated in miR-146a mimic-transfected, HG/thrombin-stimulated HAECs to 34% of that in the miR-control-transfected HACEs. The bars represent the means ± SEM from 4 experiments. **P < 0.01, as compared with the miR-control. (c) The stimulatory effect of HG/thrombin on the gene expression of VCAM-1, ICAM-1, SELE, IL-6, and IL-8 significantly decreased in miR-146a mimic-transfected, HG/thrombin-stimulated HAECs to 42%, 51%, 71%, 40%, and 34% of that in miR-control-transfected, HG/thrombin-stimulated HAECs. The bars represent the means ± SEM from 3 experiments. *P < 0.05 and **P < 0.01, as compared with the miR-control.

Figure 7

The proposed role of miR-146a in regulating HG/thrombin-induced endothelial inflammation.