Abstract 1648 Perinatal Brain Injury: Pattern and Mechanisms Platform, Tuesday, 5/4

Postnatally acquired dietary iron deficiency is associated with decreased cognitive rather than motor function. This implies differential prioritization of iron among brain circuits during iron deficiency. Erikson (Journal of Nutrition, 1997) demonstrated the heterogeneous distribution of brain iron in iron sufficient post-weanling animals as well as the differential loss of iron among certain brain regions in response to an iron deficient diet. Diabetes mellitus during pregnancy and intrauterine growth retardation are two gestational conditions that can results in up to a 40% reduction in brain iron. Iron, its regulatory protein transferrin, and CytOx have not been mapped in either the iron sufficient or deficient neonatal rat brain. CytOx is the terminal iron-containing enzyme in the oxidative phosphorylation reaction and has a critical role in ATP generation. Its activity in the brain reflects iron dependent neuronal metabolism. Because perinatal iron deficient rats demonstrate abnormalities in hippocampally based recognition memory tasks, we hypothesized that perinatal iron deficiency decreases CytOx activity with more pronounced losses in structures associated with recognition memory. Pregnant Sprague-Dawley rats were fed iron deficient or sufficient chow from GD 1 until PND 10; 25 rat pups from each group were studied. Iron staining and CytOx activity were mapped histochemically using diaminobenzidine in 21 brain structures starting at 6.2 mm through 0.8 mm anterior to the external auditory meatus. Brain iron staining was reduced by a mean of 86% (range: 75-100%) and CytOx activity was decreased by a mean of 23% (range: 0-42%) in the iron deficient group (p<0.001). Despite the homogeneous loss of iron staining (structures with significant CytOx activity loss had 84.8% iron loss vs. structures without CytOx activity loss had 88.3% iron loss). CytOx activity loss was regionalized, suggesting differential regulation of uptake and utilization of iron by neurons. Structures with the greatest reductions in CytOx activity (p<0.001) included the hippocampal formation (CA1: 42%, CA3ab: 34%, CA3c: 33%, dentate gyrus: 32%, and the subiculum: 32%), the cingulate cortex (41%), the medial dorsal thalamic nucleus (28%) and the orbital cortex (37%). In contrast the anterior thalamic nucleus, the lateral nucleus of the amygdala and the medial habenula, structures not involved in recognition memory, did not have significantly reduced CytOx activity (0%, 10%, and 16% respectively). We conclude that perinatal iron deficiency differentially reduces neuronal metabolic activity, specifically targeting the hippocampus and the prefrontal projections of the limbic system. These data support previous behavioral studies in rats and in human infants which suggest abnormal hippocampally based memory function during iron deficiency. (Supported by NICHD).