Hemodynamic shear stress via ROS modulates PCSK9 expression in human vascular endothelial and smooth muscle cells and along the mouse aorta

Abstract

Aims: To investigate a possible link between hemodynamic shear stress, reactive oxygen species (ROS) generation, and proprotein convertase subtilisin/kexin type 9 (PCSK9) expression.

Results: Using a parallel-plate flow chamber, we observed that PCSK9 expression in vascular smooth muscle cells (SMCs) and endothelial cells (ECs) reached maximal value at low shear stress (3-6 dynes/cm(2)), and then began to decline with an increase in shear stress. PCSK9 expression increased when cells were treated with lipopolysaccharide. PCSK9 expression was always greater in SMCs than in ECs. ROS generation followed the same pattern as PCSK9 expression. Aortic branching and aorta-iliac bifurcation regions of mouse aorta that express low shear stress were also found to have greater PCSK9 expression (vs. other regions). To determine a relationship between ROS and PCSK9 expression, ECs and SMCs were treated with ROS inhibitors diphenylene-iodonium chloride and apocynin, and both markedly reduced PCSK9 expression. Relationship between PCSK9 and ROS was further studied in p47(phox) and gp91(phox) knockout mice; both mice strains revealed low PCSK9 levels in serum and mRNA levels in aorta-iliac bifurcation regions (vs. wild-type mice). Other studies showed that ROS and NF-κB activation plays a bridging role in PCSK9 expression via lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1).

Innovation: Low shear stress induces PCSK9 expression, which is mediated by NADPH oxidase-dependent ROS production.

Conclusions: This study provides evidence that low shear stress enhances PCSK9 expression in concert with ROS generation in vascular ECs and SMCs. ROS seem to regulate PCSK9 expression. We propose that PCSK9-ROS interaction may be important in the development of atherosclerosis in arterial channels with low shear stress.

Figures

FIG. 1.

Effect of shear stress on PCSK9 expression in ECs and SMCs. (A) Schematic drawing of the experimental flow chamber system. The overflow upper-tank provides a steady flow and pressure to ECs and SMCs. (B) Low shear stress (3–6 dynes/cm2) markedly induced PCSK9 expression, while higher shear stress had much less effect in both ECs and SMCs. At all levels of shear stress, PCSK9 expression was much higher in SMCs compared with ECs. *p<0.05 compared with control in ECs, †p<0.05 compared with control in SMCs. (C) At normal shear stress (15 dynes/cm2), LPS induced PCSK9 expression in a dose-dependent manner in both ECs and SMCs. Again, PCSK9 expression was much greater in SMCs compared with ECs. (D) In LPS-treated cells, low shear stress induced more PCSK9 expression in ECs and SMCs. Bar graphs represent data in mean±SD based on five independent experiments, *p<0.05. ECs, endothelial cells; PCSK9, proprotein convertase subtilisin/kexin type 9; SMCs, smooth muscle cells.

FIG. 2.

Distribution of PCSK9 and VCAM-1 along the aorta at baseline and after LPS administration. (A) LPS administration induced a systemic pro-inflammatory state measured as IL-1β levels (ELISA). (B) LPS administration also induced PCSK9 release in serum measured as mPCSK9 (ELISA). (C) Representative Western blots of PCSK9 expression in saline and LPS-treated mice. PCSK9 expression was higher at the level of aortic arch branching and aorta–iliac bifurcation regions where shear stress is low and disturbed, and PCSK9 expression was much less in ascending aorta, thoracic aorta, and iliac artery where the shear stress in high and steady. (D) Representative Western blots of VCAM-1 expression along the aorta in saline and LPS-treated mice. Each aortic segment was measured in duplicate, number of mice in each group (n=5). Bar graphs represent data in mean±SD, *, #, †p<0.05. ELISA, enzyme-linked immunosorbant assay; IL, interleukin; VCAM-1, vascular cell adhesion molecule-1.

FIG. 3.

Effect of shear stress on ROS production. (A, B) Low shear stress induced ROS production in ECs and SMCs (measured by two different methods). ROS generation declined as shear stress was increased. ROS generation was greater in SMCs compared with ECs. Bar graphs represent data in mean±SD based on five independent experiments, *p<0.05. ROS, reactive oxygen species.

FIG. 4.

ROS inhibition and PCSK9 expression. (A) Pretreatment with NADPH oxidase inhibitors DPI (20 μM) or apocynin (10 μM) markedly inhibited ROS production in both ECs and SMCs. (B) Pretreatment with DPI or apocynin also reduced PCSK9 expression in both ECs and SMCs. Both ECs and SMCs were exposed to shear stress at 15 dynes/cm2. Bar graphs represent data in mean±SD based on five independent experiments, *p<0.05. DPI, diphenylene-iodonium chloride. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

FIG. 5.

PCSK9 inhibition and ROS production. (A) PCSK9 was significantly inhibited by its siRNA transfection in ECs and SMCs. (B, C) PCSK9 knockdown markedly inhibited ROS production in ECs and SMCs (DCFDA staining measured by flow cytometry). Both ECs and SMCs were exposed to shear stress at 15 dynes/cm2. Bar graphs represent data in mean±SD based on five independent experiments, *p<0.05. (D) PCSK9 abrogation (KO mice) inhibited LPS induced IL-1β release in serum (measured by ELISA). (E) PCSK9 abrogation inhibited expression of NADPH oxidases subunits p47phox and gp91phox in aortic tissues. (F) p47phox and gp91phox KO mice showed significant reduction in PCSK9 secretion in serum. Each aorta or serum was measured in duplicate using 5 mice in each group (n=5). Bar graphs represent data in mean±SD, *p<0.05. DCFDA, 2,7-dichlorofluorescin diacetate; KO, knockout.

FIG. 6.

PCSK9 treatment induces ROS production. (A, B) hPCSK9 induced ROS production in a dose-dependent manner in ECs and SMCs. Both ECs and SMCs were exposed to shear stress at 15 dynes/cm2. Bar graphs represent data in mean±SD based on five independent experiments, *p<0.05. (C) Serum levels of hPCSK9 over a 4 h period after PCSK9 administration. (D, E) Compared with WT, hPCSK9 administration increased expression of p47phox and gp91phox in aorta. Each aorta or serum was measured in duplicate using 5 mice in each group (n=5). Bar graphs represent data in mean±SD, *p<0.05. hPCSK9, recombinant human PCSK9 protein.

FIG. 7.

The interaction between PCSK9, LOX-1, superoxide, and NF-κB. (A) hPCSK9 induced LOX-1 expression in a dose-dependent manner in ECs and SMCs. (B) PCSK9 siRNA transfection significantly inhibited expression of LOX-1 and p-NF-κB in ECs and SMCs. (C, D) LOX-1 siRNA transfection markedly inhibited superoxide production and expression of PCSK9 and p-NF-κB in ECs and SMCs. (E) Inhibition of p47phox and gp91phox by siRNA transfection significantly decreased expression of LOX-1 and p-NF-κB in ECs and SMCs. Both ECs and SMCs were exposed to shear stress of 15 dynes/cm2. Bar graphs represent data in mean±SD based on three to five independent experiments, *p<0.05. LOX-1, lectin-like oxidized low-density lipoprotein receptor-1; p-NF-κB, phospho-NF-κB.