The effect of IGF-I administration on cognitive impairment and biochemical pathology in a rat model of diabetic dementia

Abstract

Diabetic patients have impaired learning/memory, brain atrophy, and two-fold increased risk of dementia. The cause of cognitive disturbances that progress to dementia and the biochemical pathology in diabetic brain atrophy is unknown. It has been shown that apoptosis is increased in both the hippocampus and retinal cells of streptozotocin diabetic rats. Because neurotrophic insulin-like growth factor (IGF) levels are reduced in diabetic patients and rodents, and IGF can cross the blood-central nervous system barrier (B-CNS-B), the hypotheses that systemically administered IGF can prevent cognitive disturbances and reduce brain atrophy, independently of hyperglycemia and a generalized catabolic state, were tested. Latency to escape to a hidden platform in the Morris Water Maze is widely used to test spatial memory, a hippocampus-dependent task. Adult rats were rendered diabetic with streptozotocin and implanted 4 weeks later with subcutaneous pumps that released either vehicle (D + Veh) or 20 μg/day IGF-I (D + IGF). Ten-and-a-half weeks after the onset of diabetes, the latency to escape to the hidden platform was prolonged in (D + Veh) vs. nondiabetic rats (P < 0.003). Such prolongation was prevented in (D + IGF) vs. (D + Veh) rats (P < 0.03). After 12 weeks of diabetes, rats were euthanized and brains were excised, weighed, and the biochemical pathology was investigated. Wet brain weights (P < 0.001), 18S rRNA per brain (P < 0.002) and poly(A)+ RNA (P < 0.04) per brain were significantly reduced in (D+Veh) vs. nondiabetic rats, and IGF-I treatment had no effect. The (mg protein)/(wet weight brain) as well as (mg protein)/(brain) were significantly reduced in (D+Veh) vs. Nondiabetic rats (P < 0.001), and IGF-I treatment prevented these reductions (D+IGF-I) vs. (D+Veh) (P < 0.03) independently of ongoing hyperglycemia. To examine apoptosis in neuroretinal cells, the eyes were dissected from the STZ rats and placed in 4% parafomaldehyde, sectioned and stained for the apoptotic markers, TUNEL and Phospho-Akt. In diabetic rat retina, the number of TUNEL-immunoreactive cells increased approximately 6-fold in the photoreceptor layer (P < 0.001) and 8-fold in the inner nuclear layer (P < 0.001); phospho-Akt (Thr 308) immunoreactivity increased 8- fold in the ganglion cell layer (P < 0.001) and 3-fold in the inner nuclear layer (P < 0.01). Subcutaneous IGF-I treatment significantly reduced the number of TUNEL (P < 0.001) and phospho-Akt immunoreactive retinal cells (P < 0.05) in diabetic rats approximately to the level of the non-diabetic group. Elevated TUNEL and phospho-Akt immunoreactivities were localized to distinct cell layers in the retina of diabetic rats. The data show that IGF-I can act across the B-CNS-B to prevent loss of cognition related performance in the water maze, prevent reduced protein content and prevent increased apoptosis independently of ongoing hyperglycemia in diabetic rats. The data are consistent with a model in which a loss of IGF activity due to diabetes may contribute to cognitive disturbances. These data also show that brain atrophy may be due in part to reduced protein, mRNA and rRNA contents. Because IGFs are involved in synapse formation and axon elongation in peripheral nerves as well as mitosis in neuroblasts, we tested the hypothesis that brain IGF is essential for learning and memory. A cannula was implanted into the left lateral ventricle of adult rats to infuse IGF-II antibody (IGF-II Ab) or pre-immune serum. Rats were subjected to a passive avoidance test of their ability to learn and remember not to instinctively enter a dark chamber after 10 days of infusion. Rats were allowed to habituated for 2 days and given an electric shock on day. On days 4-6, the latency to enter the dark chamber tested learning and memory (avoidance of shock). The rats were given a reinforcement shock upon entry. The mean latencies for all groups on days 1 -3 were not significantly different. On days 4, 5, and 6, the mean latencies of the IGF-II Ab group were significantly shorter than that of the pre-immune serum group (P < 0.04, 0.02, 0.004, respectively). These data show that endogenous IGF in cerebrospinal fluid regulates or supports learning and memory. Learning and memory deficiencies are observed in aging humans and rodents, and IGF-II gene expression is reduced in the brain of aged rats.