Flow patterns regulate hyperglycemia-induced subendothelial matrix remodeling during early atherogenesis

Abstract

Objective: Altered subendothelial matrix composition regulates endothelial dysfunction and early atherosclerotic plaque formation. Hyperglycemia promotes endothelial matrix remodeling associated with multiple microvascular complications of diabetes, but a role for altered matrix composition in diabetic atherogenesis has not been described. Therefore, we sought to characterize the alterations in matrix composition during diabetic atherogenesis using both in vitro and in vivo model systems.

Methods and results: Streptozotocin-induced diabetes in atherosclerosis-prone ApoE knockout mice promoted transitional matrix expression (fibronectin, thrombospondin-1) and deposition in intima of the aortic arch as determined by qRT-PCR array and immunohistochemistry. Early plaque formation occurs at discrete vascular sites exposed to disturbed blood flow patterns, whereas regions exposed to laminar flow are protected. Consistent with this pattern, hyperglycemia-induced transitional matrix deposition was restricted to regions of disturbed blood flow. Laminar flow significantly blunted high glucose-induced fibronectin expression (mRNA and protein) and fibronectin fibrillogenesis in endothelial cell culture models, whereas high glucose-induced fibronectin deposition was similar between disturbed flow and static conditions.

Conclusions: Taken together, these data demonstrate that flow patterns and hyperglycemia coordinately regulate subendothelial fibronectin deposition during early atherogenesis.

Keywords: Atherosclerosis; Endothelial; Fibronectin; Hyperglycemia; Shear stress.

Figure 1. Altered expression of extracellular matrix genes in early diabetic atherogenesis

ApoE null mice at 6 to 8 weeks of age received either low-dose streptozotocin (50 mg·kg⁻¹) or Na-Citrate buffer injections. After 6 weeks of diabetes, the aorta was harvested and analyzed by qRT-PCR for gene expression of inflammatory and matrix-related genes. Gene expression was normalized to housekeeping genes and to nondiabetic controls. n = 4 mice per condition. * p < 0.05.

Figure 2. Hyperglycemia stimulates fibronectin deposition at sites of disturbed flow in the aortic arch

The aortic arch was collected from diabetic (6 weeks) and nondiabetic ApoE knockout mice and stained for A) VCAM-1, B) thrombospondin-1, or C) fibronectin deposition. Regions exposed to either laminar flow (greater curvature) or disturbed flow (lesser curvature, carotid branch point) are shown at 20× magnification with 4× magnification inserts of the entire artery. n = 4 – 7 mice per group.

Figure 3. Hyperglycemia stimulates fibronectin deposition at sites of disturbed flow in the carotid artery

The carotid sinus and common carotid was collected from diabetic (4 and 6 weeks) and nondiabetic ApoE knockout mice and stained for fibronectin deposition. Representative images from A) citrate-buffer treated (nondiabetic) carotid sinus, B) STZ-treated (diabetic) carotid sinus, and C) STZ-treated common carotid are shown. Images are at 40× magnification and the insets show the 10× magnification of the entire artery. (D,E) Fibronectin staining at different vascular sites (innominate, carotid sinus, or common carotid) was quantified using Nikon Elements software. n =4–7 mice per group, * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 4. Laminar flow significantly inhibits hyperglycemia-induced FN deposition in vitro

HAECs exposed to laminar shear stress (LSS), oscillatory shear stress (OSS), or static conditions for 18 hours in the presence of either 5 mM or 25 mM glucose. Mannose was added to the 5 mM glucose treatments to control for osmolarity. Cells were extracted using Triton X-100 and deoxycholic acid (DOC), and (A) the DOC-insoluble and soluble fractions were collected and the presence of fibronectin in the endothelial matrix was quantified by Western blotting. n = 4–5. (B) Alternatively, the underlying matrix was fluorescently stained for fibronectin and quantified as percent positive area. At least 10 random images were taken for each treatment. n = 3. C) Fibronectin mRNA expression was determined by qRT-PCR and normalized to β2-microglobin mRNA levels. n = 3–6. (D) Total cellular fibronectin protein expression was determined by Western blotting cell lysates. n =4–5. # p < 0.05 and ### p < 0.001 compared to low glucose controls. * p < 0.05, ** p < 0.01, and *** p < 0.001 compared to static controls.