Overexpression of endothelial nitric oxide synthase improves endothelium-dependent vasodilation in arteries infused with helper-dependent adenovirus

Abstract

Adenoviral vectors (Ad) are useful tools for in vivo gene transfer into endothelial cells. However, endothelium-dependent vasodilation is impaired after Ad infusion, and this impairment is not prevented by use of advanced-generation "helper-dependent" (HD) Ad that lack all viral genes. We hypothesized that endothelium-dependent vasodilation could be improved in Ad-infused arteries by overexpression of endothelial nitric oxide synthase (eNOS). We tested this hypothesis in hyperlipidemic, atherosclerosis-prone rabbits because HDAd will likely be used for treating and preventing atherosclerosis. Moreover, the consequences of eNOS overexpression might differ in normal and atherosclerosis-prone arteries and could include atherogenic effects, as reported in transgenic mice. We cloned rabbit eNOS and constructed an HDAd that expresses it. HDAdeNOS increased NO production by cultured endothelial cells and increased arterial eNOS mRNA in vivo by ∼10-fold. Compared to arteries infused with a control HDAd, HDAdeNOS-infused arteries of hyperlipidemic rabbits had significantly improved endothelium-dependent vasodilation, and similar responses to phenylephrine and nitroprusside. Moreover, infusion of HDAdeNOS had local atheroprotective effects including large, significant decreases in intimal lipid accumulation and arterial tumor necrosis factor (TNF)-α expression (p≤0.04 for both). HDAdeNOS infusion yields a durable (≥2 weeks) increase in arterial eNOS expression, improves vasomotor function, and reduces artery wall inflammation and lipid accumulation. Addition of an eNOS expression cassette improves the performance of HDAd, has no harmful effects, and may reduce atherosclerotic lesion growth.

FIG. 1.

HDAdeNOS increases eNOS expression and activity in vitro and ex vivo. (A) Western blot of eNOS in extracts of cultured bovine aortic endothelial cells (BAEC) that were transduced either with HDAdNull or HDAdeNOS. Each lane is from a separate well of cells. Size markers are in kDa. (B) Nitric oxide (NO) activity measured by spin trapping in BAEC transduced either with HDAdNull or HDAdeNOS. Values are mean±SE of three independent transductions. The experiments in A and B were repeated once with similar results. ESR, electron spin resonance; AU, arbitrary units; HDAdeNOS, helper-dependent adenoviral vector endothelial nitric oxide synthase.

FIG. 2.

HDAdeNOS increases eNOS expression in vivo. Carotid arteries of cholesterol-fed rabbits were infused with either HDAdNull or HDAdeNOS. Two weeks later the arteries were harvested, RNA was extracted, and eNOS mRNA was measured by quantitative reverse-transcription (RT)-PCR of eNOS mRNA with normalization to GAPDH mRNA. Data points are from individual arteries; bars are group means. AU, arbitrary units.

FIG. 3.

HDAdeNOS increases endothelium-dependent vascular relaxation. Carotid arteries of cholesterol-fed rabbits were infused with either HDAdNull or HDAdeNOS and harvested 2 weeks later. Rings cut from the transduced arteries were tested for reactivity to (A) phenylephrine, (B) sodium nitroprusside, and (C) acetylcholine. Data are mean±SE of 15–16 arteries per group. *p<0.05 versus HDAdNull over entire range of concentrations (two-way repeated measures ANOVA).

FIG. 4.

Impact of HDAdeNOS on intimal lesion size, macrophage accumulation, and lipid content. Arteries of fat-fed rabbits were infused with either HDAdNull or HDAdeNOS, harvested 2 weeks later, sectioned, and stained either with an antibody to rabbit macrophages (RAM-11) or with oil red O (to detect lipids). (A) Intimal lesion area, measured on oil red O-stained slides. (B) Percentage of intimal lesion area occupied by macrophages. (C) Percentage of intimal lesion area occupied by lipid. Multiple sections per artery were stained and measured. Data points are mean values for each artery; bars are group medians.

FIG. 5.

Adhesion molecule expression in arteries transduced with HDAdNull or HDAdeNOS. Arteries of fat-fed rabbits were infused with either HDAdNull or HDAdeNOS, harvested 2 weeks later, sectioned, and stained with antibodies that detect VCAM-1 or ICAM-1 proteins. Multiple sections per artery were scored for staining intensity. Data points are mean values for each artery; bars are group medians.

FIG. 6.

Regulation of atherogenic cytokine expression by HDAdeNOS. Arteries of fat-fed rabbits were infused with either HDAdNull or HDAdeNOS and harvested 2 weeks later. Arterial RNA was extracted and mRNA for (A) TNF-α, (B) IL-6, and (C) MCP-1 were measured by quantitative RT-PCR with normalization to GAPDH mRNA measured in the same extracts. Data are mean±SE, with n=16 per group. AU, arbitrary units; TNF, tumor necrosis factor; IL, interleukin; MCP, monocyte chemotactic protein.