Non-Invasive Molecular Imaging of Endothelial Injury in Diabetic Animals

Abstract

Purpose: : Diabetic retinopathy (DR), a microvascular manifestation of diabetes, is a leading cause of adult vision loss. There is an urgent need to diagnose DR before the onset of irreversible signs of the disease, as early detection and treatment may prevent loss of vision. We aim to develop a non-invasive technique for molecular imaging of endothelial injury in diabetic animals.

Methods: : Long-Evans rats were injected with 12 mg Streptozotocin to induce Type I Diabetes. Anti-Intercellular-Adhesion Molecule-1 (Anti-ICAM-1) was conjugated to fluorescent microspheres (2µm diam.). After three weeks of diabetes conjugated microspheres (0.35 ml) were injected systemically into anesthetized rats and their rolling and adhesion in the choriocapillaris were investigated under physiologic flow conditions by scanning laser ophthalmoscopy (SLO). Subsequently, animals were perfused and retinal and choroidal flatmounts were prepared to count the number of firmly adhering microspheres.

Results: : A significantly higher number of anti-ICAM-I conjugated microspheres adhered to the choriocapillaris endothelium of the diabetic rats (45.9±2, n=8), compared to normal non-diabetic controls (30.6±1, n=8, p=0.03). Microspheres conjugated with mouse-IgG, used as a negative control, did not show a significant difference in adhesion to the choriocapillars in diabetic and normal rats. In contrast to microspheres that target ICAM-I, conjugated microspheres that bind to endothelial P-selectin did not show a significant difference between diabetic animals and normal controls.

Conclusions: : We introduce a novel non-invasive method for detection of choroidal endothelial surface molecules in vivo, allowing quantitative assessment of endothelial injury during diabetes. ICAM-I provides a sensitive molecular target for detection of early changes during experimental DR. In contrast, adhesion of microspheres to endothelial P-selectin did not show a difference between diabetic and normal animals. This technique could be further developed to detect subclinical signs of DR in diabetic patients.

Keywords: diabetic retinopathy • imaging/image analysis: non-clinical • inflammation