Slow flow induces endothelial dysfunction by regulating thioredoxin-interacting protein-mediated oxidative metabolism and vascular inflammation

Abstract

Endothelial cells are highly sensitive to hemodynamic shear stresses, which act in the blood flow's direction on the blood vessel's luminal surface. Thus, endothelial cells on that surface are exposed to various physiological and pathological stimuli, such as disturbed flow-induced shear stress, which may exert effects on adaptive vascular diameter or structural wall remodeling. Here we showed that plasma thioredoxin-interactive protein (TXNIP) and malondialdehyde levels were significantly increased in patients with slow coronary flow. In addition, human endothelial cells exposed to disturbed flow exhibited increased levels of TXNIP in vitro. On the other hand, deletion of human endothelial TXNIP increased capillary formation, nitric oxide production and mitochondrial function, as well as lessened oxidative stress response and endothelial cell inflammation. Additional beneficial impacts from TXNIP deletion were also seen in a glucose utilization study, as reflected by augmented glucose uptake, lactate secretion and extracellular acidification rate. Taken together, our results suggested that TXNIP is a key component involved in mediating shear stress-induced inflammation, energy homeostasis, and glucose utilization, and that TXNIP may serve as a potentially novel endothelial dysfunction regulator.

Keywords: disturbed flow; endothelial dysfunction; mitochondrial dysfunction; oxidative metabolism; thioredoxin-interacting protein (TXNIP).

FIGURE 1

Plasma TXNIP and MDA were elevated in the coronary of patients with coronary slow flow phenomenon. (A) Plasma TXNIP was measured by using ELISA; (B) Plasma MDA was measured by using ELISA; *P < 0.05 vs. Con, ^#P < 0.05 vs. SF-CA.

FIGURE 2

Increased expression of TXNIP in disturbed flow-induced endothelial dysfunction. (A) Thioredoxin-interactive protein (TXNIP) expression was determined by Western Blotting in the treatment groups of laminar flow with negative control siRNA (LF + NC-siRNA), disturbed flow with negative control siRNA (DF + NC-siRNA) and disturbed flow with TXNIP-siRNA (DF + TXNIP-siRNA). (B) Formation of capillaries in Matrigel. Scale bars = 50 μm. Results are shown as mean ± SD. N = 5/group. *P < 0.05 vs. LF + NC-siRNA, ^#P < 0.05 vs. DF + NC-siRNA.

FIGURE 3

Disturbed flow reduced NO production in a TXNIP dependent manner. (A) NO production was assessed with DAF-FM DA, (B) Western Blot analysis of nitrotyrosine modified protein expression, (C) Western Blots comparing the presence of phosphorylated eNOS at Ser1177 and total eNOS, in the LF + NC-siRNA, DF + NC-siRNA, and DF + TXNIP-siRNA treated groups. Results are shown as mean ± SD. N = 5/group. *P < 0.05 vs. LF + NC-siRNA, ^#P < 0.05 vs. DF + NC-siRNA. DAF-FM DA: 4-amino-5-methylamino-2,7’-difluorofluorescein.

FIGURE 4

Disturbed flow induced endothelial mitochondrial dysfunction by regulating TXNIP expression. (A) Reactive oxygen species (ROS) level, (B) Mitochondrial ATP level, (C) Mitochondrial membrane potential, in the LF + NC-siRNA, DF + NC-siRNA, and DF + TXNIP-siRNA treated groups. (D) Representative image of mitochondrial morphologies with transmission electron microscopy of HUVECs, respectively. Results are shown as mean ± SD. N = 5/group. *P < 0.05 vs. LF + NC-siRNA, ^#P < 0.05 vs. DF + NC-siRNA.

FIGURE 5

Disturbed flow decreased glucose utilization through TXNIP-dependent activation. (A) Glucose uptake, (B) Lactate production, (C) Kinetic oxygen consumption rate (OCR) responses of HUVECs to 20 mM glucose and 5 mM oligomycin, (D) Calculated glucose oxidation rate, (E) Kinetic extracellular acidification rate (ECAR) responses of HUVECs to glucose (20 mM), oligomycin (5 μM) and 2-DG (100 mM), in the LF + NC-siRNA, DF + NC-siRNA, and DF + TXNIP-siRNA treatment groups. (F) Calculated glycolytic flux and glycolytic capacity. The glycolytic flux and glycolytic capacity are calculated by ECAR increase normalized with cell protein content. (G) GLUT4 and PDH E1α expression were determined by Western Blotting in the treatment groups of laminar flow with negative control siRNA (LF + NC-siRNA), disturbed flow with negative control siRNA (DF + NC-siRNA), and disturbed flow with TXNIP-siRNA (DF + TXNIP-siRNA). All values are presented as mean ± SD. N = 5/group. *P < 0.05 vs. LF + NC-siRNA, ^#P < 0.05 vs. DF + NC-siRNA.

FIGURE 6

TXNIP activation promoted the disturbed flow-induced pro-inflammatory response. (A) Representative Western Blots, along with quantification of cleaved-IL-1β and NLRP3, (B) Representative Western Blots, along with VCAM1 and ICAM1 quantification, in the LF + NC-siRNA, DF + NC-siRNA, and DF + TXNIP-siRNA groups. The blot shows representative images of five independent experiments. Results are shown as mean ± SD. *P < 0.05 vs. LF + NC-siRNA, ^#P < 0.05 vs. DF + NC-siRNA.

FIGURE 7

Proposed signaling mechanism linking shear stress to endothelial dysfunction in HUVECs.