Search

Aims/hypothesis: The retina is embryologically similar to cerebral cortex and the tissues of both are exposed to similar blood glucose concentrations. Nevertheless, in diabetes the retina develops metabolic abnormalities and microvascular lesions from which cerebrum seems relatively protected. We directly compared glucose concentrations and expression of GLUT-1 (the major carrier transporting glucose from blood into the neural retina and cerebrum) in the two tissues from normal and diabetic rats., Methods: Tissue and intracellular glucose were measured using two methods: direct assay of glucose and assay of Amadori products on intracellular proteins. The expression of GLUT-1 was measured using western blots in tissue and in the isolated endothelial luminal membrane of the two vascular beds., Results: Both methods assessing intracellular glucose indicate that intracellular concentrations of glucose in diabetes increased significantly in the retina but not in cerebral cortex. Concentrations of free glucose and Amadori product in retinas of diabetic animals were increased above normal by 334% and 122%, respectively, whereas there was no statistically significant increase in either parameter in the cerebral cortex of diabetic animals. In contrast to the observed increase in glucose in the retina in diabetes, expression of GLUT-1 on the luminal plasmalemma of the retinal vascular endothelium and in homogenates of whole retina decreased to a statistically significant extent (55% and 36%, respectively compared to normal). In the luminal cell membrane of the cerebral vasculature, diabetes did not decrease expression of GLUT-1 but tended to increase it slightly., Conclusions/interpretation: Even among tissues that do not require insulin for glucose uptake, tissue glucose concentration varies in diabetes. The greater increase in glucose concentration in retina than in cerebrum in diabetes probably contributes to the tissue differences in biochemical and histopathologic sequelae of the disease. The expression of GLUT-1 in the microvasculature is unlikely to account for the differences in tissue glucose between retina and cerebrum.

Freezing lesions have been shown to cause a depression in glucose use, particularly in cortical areas of the brain ipsilateral to the lesion, and this effect was interpreted to be caused by a depressed functional activity in these regions. The metabolic status of the affected areas has not been previously examined and could be a factor in the observed changes in local CMRglc. In frozen-cut and dried sections taken from brains 3 days after freeze lesioning, discrete pieces of the median and lateral parietal cortex, striatum, hippocampus, and hypothalamus were dissected and analyzed for ATP, P-creatine, glucose, and lactate. CMRglc measurements were also made in the same animals. The concentrations of the four metabolites were significantly increased in the lesioned hemisphere, with the most predominant effects observed in the cortical areas that exhibited the greatest depression in CMRglc. The enriched metabolite profile, particularly in the cortical areas, is consistent with the hypothesis that decreased glucose use in the traumatized brain is caused by diminished need rather than by decreased supply of energy. Because the lumped constant in the operational equation of the deoxyglucose method for determination of CMRglc is a function of brain glucose content and decreases gradually in hyperglycemia, the degree of metabolic depression in cortical areas of lesioned hemisphere probably have been somewhat overestimated in this and previous publications. However, provisionally recalculated local CMRglc in the lesioned hemisphere remain significantly lower than in the contralateral hemisphere and in the normal brain.

The common carotid arteries were occluded in gerbils for 5 min and the metabolic rate was estimated by measuring the loss of high-energy phosphate equivalents at 4 days of reperfusion in the cerebral cortex, hippocampus, and striatum. Metabolites values at 4 days of reperfusion were not different from those of controls with the exception of glycogen, which was significantly elevated in the hippocampus. The metabolic rate, as determined by the "closed-box" method at 4 days of reflow, was decreased by more than 50% in all three regions after 5 min of bilateral ischemia. The ischemic time necessary to elicit the hypometabolic response at 4 days of reflow was 2, 3 and 4 min for the striatum, hippocampus, and cortex, respectively. It is suggested that delayed postischemic hypometabolism may be a component of an adaptive process which counteracts, to varying degrees, the deleterious effects of ischemia depending on the region examined.