MRI Phenotyping of Murine Diabetic Retinopathy
Bruce Berkowitz   (Detroit, MI)
Prognostic imaging biomarkers are a powerful non-destructive approach for characterizing pathophysiology linked with retinal complications of type 1 diabetes in both mice and humans. We have already validated a clinically-relevant MRI biomarker, oxygenation response during a provocation (i.e., DeltaP02), as a measure of retinal vascular function. Retinal DeitaP02 is altered before the appearance of microangiopathy in both preclinical and clinical studies, and is prognostic in preclinical studies of drug treatments which correct vascular histopathology. However, it has not yet been possible to use MRI to monitor functional abnormalities in vision-critical non-vascular receptoral (i.e., outer) retina and post-receptoral (inner) retina, which are also a feature of diabetic retinopathy. Our long term goal is to use prognostic MRI biomarkers to quantitatively phenotype vascular and non-vascular retinal dysfunction associated with diabetes. Healthy vision is-linked with proper cellular demand for ions. Perturbed ion homeostasis occurs early in the course of diabetic retinopathy but imaging methods that can simultaneously measure retinal layer thickness and intraretinal ion demand have not yet been available. We have developed a new functional MRI method for measuring non-vascular layer-specific retinal thickness and ion demand during light and dark adaptation in rodents. This new method, manganese-enhanced MRI (MEMRI), takes advantage of the facts that manganese (Mn2+) ion acts as a surrogate for cellular demand for ions, including calcium, and is a strong MRI contrast agent. The availability of an FDA-approved manganese-based contrast agent raises the possibility of future clinical applications of MEMRI. We confirmed that type 1 diabetes-induced MEMRI and DeitaP02 defects in rats are important endpoints because both abnormalities were corrected with alpha-lipoic acid treatment, a drug which inhibits the appearance of vascular histopathology. Our overlying hypothesis is that early in the course of type 1 diabetes, abnormal ion demand in non-vascular retina (as assessed by MEMRI), appears before that of dysfunctional oxygenation response (measured by DeltaP02). Aim 1: To test the prediction that diabetes-induced abnormalities in retinal ion demand are present before perturbations in oxygenation response. At various times after conversion to diabetes, we will compare MEMRI-based structural and functional patterns in inner and outer retina, and retinal DeitaP02 in mouse models of diabetic retinopathy.