Atheroprone hemodynamics regulate fibronectin deposition to create positive feedback that sustains endothelial inflammation

Abstract

Rationale: The extracellular matrix protein fibronectin (FN) is focally deposited in regions of atherosclerosis, where it contributes to inflammatory signaling.

Objective: To elucidate the mechanism by which FN deposition is regulated by local shear stress patterns, its dependence on platelet-endothelial cell adhesion molecule (PECAM)-1 mechanotransduction and the role this pathway plays in sustaining an atheroprone/proinflammatory phenotype.

Methods and results: Human endothelial cells were exposed in vitro to atheroprone or atheroprotective shear stress patterns derived from human carotid arteries. Onset of atheroprotective flow induced a transient increase in FN deposition, whereas atheroprone flow caused a steady increase in FN expression and integrin activation over time, leading to a significant and sustained increase in FN deposition relative to atheroprotective conditions. Comparing FN staining in ApoE(-/-) and ApoE(-/-)PECAM(-/-) mice showed that PECAM-1 was essential for FN accumulation in atheroprone regions of the aortic arch. In vitro, small interfering RNA against PECAM-1 blocked the induction of FN and the activation of nuclear factor (NF)-kappaB by atheroprone flow, which was rescued by the addition of exogenous FN. Additionally, blocking NF-kappaB activation attenuated the flow-induced FN expression. Small interfering RNA against FN significantly reduced NF-kappaB activity, which was rescued by the addition of exogenous FN.

Conclusions: These results indicate that FN gene expression and assembly into matrix fibrils is induced by atheroprone fluid shear stress. This effect is mediated at least in part by the transcription factor NF-kappaB. Additionally, because FN promotes activation of NF-kappaB, atheroprone shear stress creates a positive feedback to maintain inflammation.

Figure 1. Physiologic Shear Stress Regulates FN Deposition

A. ECs under atheroprone or atheroprotective shear stress or static conditions were analyzed for FN protein expression at 4, 16, and 24 hours and were normalized to α-tubulin and presented as fold changes relative to the initial condition (t0) A representative blot after 24 hours of flow is shown. B. FN gene expression after 24 hours was assessed and presented as fold change relative to prone flow. C. FN staining after 24 hours at the indicated magnification following atheroprone or atheroprotective shear stress. (Values are means±SE, n=5-10. A. *= p<0.05 between flow types, two-way ANOVA, B. *=p<0.05 students t-test).

Figure 2. PECAM-1 Promotes FN Deposition in ApoE−/− Mice

FN staining was compared between ApoE−/−PECAM+/+ and ApoE−/−PECAM−/− mice in the aortic arch cross sections both prior to lesion formation (chow diet) and in advanced lesions (western diet) (representative images for 3 mice under each condition). Larger images display higher magnification of the full vessel cross-section showed in the inset.

Figure 3. PECAM Promotes Atheroprone Flow-Induced NF-κB Activity via FN Deposition

A. PECAM (siPECAM) or control siRNA (siControl) treated ECs exposed to atheroprone flow were analyzed for protein expression by Western blot and normalized to α-tubulin. Each measurement was presented as fold change relative to the untreated siControl condition represented by the horizontal line at 1 (B-D). FN protein levels (n=4-8) (B.), NF-κB luciferase reporter activity (n=4-7) (C.), and VCAM expression (n=4.6) (D.) was assessed in ECs under atheroprone flow after siPECAM (+) or siControl (−), with or without exogenous FN. E-F. PECAM tyrosine phosphorylation following exposure to atheroprone or atheroprotective shear stress was measured (n=4) (F). Values are means±SE normalized to α-tubulin and presented as fold changes relative to the untreated siControl (A-D) or normalized total PECAM and presented as fold changes relative to the static condition (E-F). †p<0.05, one-sample t-test, compared to untreated siControl (B, C, D) or static conditions (F), *p<0.05, one-way ANOVA.

Figure 4. Atheroprone-Induced FN Deposition is NF-κB-dependent

NF-κB activity was blocked during 24 hours of atheroprone flow using either BAY pharmacological agent or with an IκB dominant negative construct, and FN protein (A) and gene expression (B) were assessed compared to the vehicle control. Values are means±SE (n=4) and reported as fold changes relative to the atheroprone control. *p<0.05, paired t-test.

Figure 5. Atheroprotective Flow Maintains Low, Basal FN via Reduced Integrin Activation

A-B. FN levels and NF-κB activity was assessed after 24 hours of atheroprotective flow with or without the addition of exogenous FN and/or integrin activating antibody TS2/16 for the last 8 hours of flow, and data was presented as fold change relative to the initial condition (t0) or to atheroprone control. C. ECs were treated with 5ng/ml of IL-1β and FN was normalized to α-tubulin and compared to an untreated condition (represented by the horizontal line at 1). D. After 24 hours of shear stress, EC monolayers were incubated with 20 μg/ml GST-FNIII_9-11 for 30 minutes to measure integrin activity/binding and presented as fold change relative to the atheroprone condition. Values are means±SE (n=3-5), A-B *=p<0.05, two-way ANOVA compared to untreated control, C. one-sample t-test compared to untreated, D. *p<0.05 student's t-test.

Figure 6. Increased FN Deposition Under Atheroprone Flow Promotes NF-κB Activity

A. ECs were treated with siRNA against FN (siFN), with or without addition of exogenous FN under static and atheroprone conditions and compared to untreated siRNA control (siControl). B. NF-κB-reporter activity was compared between the siFN and siControl ECs after 24hours of atheroprone flow and presented as fold change relative to the siControl untreated atheroprone condition (represented by the horizontal line at 1). C. Cells treated with siFN or siControl were exposed to atheroprone flow for 24h then analyzed for expression NF-κB-dependent genes presented as fold changes relative to siControl (horizontal line at 1). Values are means±SE (n=6). †p<0.05, one-sample t-test, compared to the untreated siControl atheroprone condition, set to 1, *p<0.05, one-way ANOVA).