Oxidative stress and HNE conjugation of GLUT3 are increased in the hippocampus of diabetic rats subjected to stress.

Recent studies demonstrate that cellular, molecular and morphological changes induced by stress in rats are accelerated when there is a pre-existing strain upon their already compromised adaptive responses to internal or external stimuli, such as may occur with uncontrolled diabetes mellitus. The deleterious actions of diabetes and stress may increase oxidative stress in the brain, leading to increases in neuronal vulnerability. In an attempt to determine if stress, diabetes or stress+diabetes increases oxidative stress in the hippocampus, radioimmunocytochemistry was performed using polyclonal antisera that recognize proteins conjugated by the lipid peroxidation product 4-hydroxy-2-nonenal (HNE). Radioimmunocytochemistry revealed that HNE protein conjugation is increased in all subregions of the hippocampus of streptozotocin (STZ) diabetic rats, rats subjected to restraint stress and STZ diabetic rats subjected to stress. Such increases were not significant in the cortex. Because increases in oxidative stress may contribute to stress- and diabetes-mediated decreases in hippocampal neuronal glucose utilization, we examined the stress/diabetes mediated HNE protein conjugation of the neuron specific glucose transporter, GLUT3. GLUT3 immunoprecipitated from hippocampal membranes of diabetic rats subjected to stress exhibited significant increases in HNE immunolabeling compared to control rats, suggesting that HNE protein conjugation of GLUT3 contributes to decreases in neuronal glucose utilization observed during diabetes and exposure to stress. Collectively, these results demonstrate that the hippocampus is vulnerable to increases in oxidative stress produced by diabetes and stress. In addition, increases in HNE protein conjugation of GLUT3 provide a potential mechanism for stress- and diabetes-mediated decreases in hippocampal neuronal glucose utilization.

Regulation of GLUT-3 glucose transporter in the hippocampus of diabetic rats subjected to stress.

Previous studies from our laboratory have demonstrated that chronic stress produces molecular, morphological, and ultrastructural changes in the rat hippocampus that are accompanied by cognitive deficits. Glucocorticoid attenuation of glucose utilization is proposed to be one of the causative factors involved in stress-induced changes in the hippocampus, producing an energy-compromised environment that may make hippocampal neuronal populations more vulnerable to neurotoxic insults. Similarly, diabetes potentiates neuronal damage in acute neurotoxic events, such as ischemia and stroke. Accordingly, the current study examined the regulation of the neuron-specific glucose transporter, GLUT-3, in the hippocampus of streptozotocin-induced diabetic rats subjected to restraint stress. Diabetes leads to significant increases in GLUT-3 mRNA and protein expression in the hippocampus, increases that are not affected by stress. Collectively, these results suggest that streptozotocin-induced increases in GLUT-3 mRNA and protein expression in the hippocampus may represent a compensatory mechanism to increase glucose utilization during diabetes and also suggest that modulation of GLUT-3 expression is not responsible for glucocorticoid impairment of glucose utilization.

Expression of leptin receptor isoforms in rat brain microvessels.

Leptin acts on specific brain regions to affect body weight regulation. As leptin is made by white adipose tissue, it is thought that leptin must cross the blood-brain barrier or the blood-cerebrospinal fluid barrier to reach key sites of action within the brain. High expression of a short form leptin receptor has been reported in the choroid plexus. However, whether one or more of the known leptin receptor isoforms is expressed in brain capillaries is unknown. To identify and quantitate leptin receptor isoforms in rat brain microvessels, we applied quantitative RT-PCR to RNA from purified rat brain microvessels in parallel with in situ hybridization. The results show that the amount of short form leptin receptor messenger RNA (mRNA) in brain microvessels is extremely high, exceeding that in choroid plexus. In contrast, low levels of this mRNA were detected in the cerebellum, hypothalamus, and meninges. The long form leptin receptor mRNA is only present at low levels in the microvessels, but surprisingly, its level in cerebellum is 5 times higher than that in the hypothalamus. In situ hybridization experiments confirmed strong expression of short leptin receptors in microvessels, choroid plexus, and leptomeninges. The distribution and type of leptin receptor mRNA isoforms in brain microvessels are consistent with the possibility that receptor-mediated transport of leptin across the blood-brain barrier is mediated by the short leptin receptor isoform.

Effects of dexamethasone in vivo and in vitro on hexose transport in brain microvasculature.

Glucocorticoids induce hyperinsulinemia, hyperglycemia, and depress glucose transport by aortic endothelium. High glucocorticoid doses are used for many diseases, but with unknown effects on brain glucose transport or metabolism. This study tested the hypothesis that glucocorticoids affect glucose transport or metabolism by brain microvascular endothelium. Male rats received dexamethasone (DEX) s.c. with sucrose feeding for up to seven days. Cerebral microvessels from rats treated with DEX/sucrose demonstrated increased GLUT1 and brain glucose extraction compared to controls. Glucose transport in vivo correlated with hyperinsulinemia. Pre-treatment with low doses of streptozotocin blunted hyperinsulinemia and prevented increased glucose extraction induced by DEX. In contrast, isolated brain microvessels exposed to DEX in vitro demonstrated suppression of 2-deoxyglucose uptake and glucose oxidation. We conclude that DEX/sucrose treatment in vivo increases blood-brain glucose transport in a manner that requires the effects of chronic hyperinsulinemia. These effects override any direct inhibitory effects of either hyperglycemia or DEX.

Ultrastructural localization of GLUT 1 and GLUT 3 glucose transporters in rat brain.

Precise localization of glucose transport proteins in the brain has proved difficult, especially at the ultrastructural level. This has limited further insights into their cellular specificity, subcellular distribution, and function. In the present study, preembedding ultrastructural immunocytochemistry was used to localize the major brain glucose transporters, GLUTs 1 and 3, in vibratome sections of rat brain. Our results support the view that, besides being present in endothelial cells of central nervous system (CNS) blood vessels, GLUT 1 is present in astrocytes. GLUT 1 was detected in astrocytic end feet around blood vessels, and in astrocytic cell bodies and processes in both gray and white matter. GLUT 3, the neuronal glucose transporter, was located primarily in pre- and postsynaptic nerve endings and in small neuronal processes. This study: (1) affirms that GLUT 3 is neuron-specific, (2) shows that GLUT 1 is not normally expressed in detectable quantities by neurons, (3) suggests that glucose is readily available for synaptic energy metabolism based on the high concentration of GLUT 3 in membranes of synaptic terminals, and (4) demonstrates significant intracellular and mitochondrial localization of glucose transport proteins.

Forebrain endothelium expresses GLUT4, the insulin-responsive glucose transporter.

The presence of GLUT4, the insulin-responsive glucose transporter, in microvascular endothelium and the responsiveness of glucose transport at the blood-brain barrier to insulin have been matters of controversy. To address these issues, we examined GLUT4 mRNA and protein expression in isolated brain microvessels and in cultured calf vascular cells derived from brain microvessels and aorta. We report here that GLUT4 mRNA can be detected in rat forebrain and its microvasculature using high stringency hybridization of poly(A)+ RNA isolated from these sources. This mRNA is identical to that found in adipose cells from rat. Immunoblot analysis of isolate brain microvessels reveals that GLUT4 protein is also present. Peptide preadsorption studies and absence of our antibody reaction to human red cells suggest these findings are specific. Immunohistochemical staining of cultured calf vascular cells reveals that GLUT4 is expressed in brain endothelial cells but not pericytes, nor in aortic endothelium or smooth muscle cells. The sensitivity of the methods required to detect GLUT4 in brain and comparison to its abundance in low density microsomes from rat adipose cells indicate that GLUT4 is expressed in relatively low abundance in brain microvascular endothelium. No significant differences are observed in steady state levels of GLUT4 mRNA in brain from streptozotocin diabetic compared to control rats. This last finding supports the concept of tissue-specific regulation of GLUT4. We conclude that brain microvascular endothelium specifically expressed GLUT4 while other vascular cells do not.

Forebrain ischemia increases GLUT1 protein in brain microvessels and parenchyma.

Glucose transport into nonneuronal brain cells uses differently glycosylated forms of the glucose transport protein, GLUT1. Microvascular GLUT1 is readily seen on immunocytochemistry, although its parenchymal localization has been difficult. Following ischemia, GLUT1 mRNA increases, but whether GLUT1 protein also changes is uncertain. Therefore, we examined the immunocytochemical distribution of GLUT1 in normal rat brain and after transient global forebrain ischemia. A novel immunocytochemical finding was peptide-inhibitable GLUT1 immunoreactive staining in parenchyma as well as in cerebral microvessels. In nonischemic rats, parenchymal GLUT1 staining co-localizes with glial fibrillary acidic protein (GFAP) in perivascular foot processes of astrocytes. By 24 h after ischemia, both microvascular and nonmicrovascular GLUT1 immunoreactivity increased widely, persisting at 4 days postischemia. Vascularity within sections of brain similarly increased after ischemia. Increased parenchymal GLUT1 expression was paralleled by staining for GFAP, suggesting that nonvascular GLUT1 overexpression may occur in reactive astrocytes. A final observation was a rapid expression of inducible heat shock protein (HSP)70 in hippocampus and cortex by 24 h after ischemia. We conclude that GLUT1 is normally immunocytochemically detectable in cerebral microvessels and parenchyma and that parenchymal expression occurs in some astroglia. After global cerebral ischemia, GLUT1 overexpression occurs rapidly and widely in microvessels and parenchyma; its overexpression may be related to an immediate early-gene form of response to cellular stress.

Immunohistochemical localization of the neuron-specific glucose transporter (GLUT3) to neuropil in adult rat brain.

The precise histologic localization of GLUT3, a glucose transporter thought to be restricted to neurons, is unknown. Using a high-affinity, specific antiserum against rodent GLUT3 for immunocytochemistry, light microscopic staining concentrates heterogeneously in the neuropil in a region- and lamina-specific manner; intense staining characterizes areas with high rates of glucose utilization such as inferior colliculus and pyriform cortex. Neuropil localization with little perikaryal staining suggests that GLUT3 may provide the energy needed locally for synaptic transmission.

An immunization method for generation of high affinity antisera against glucose transporters useful in immunohistochemistry.

Antipeptide antisera from unique amino acid sequences of proteins, predicted from their cDNA, are useful to study cellular distribution of these proteins, but such peptides often are poorly immunogenic. We describe a secondary immunization method with repeated intravenous administration of KLH-conjugated peptides to boost the immune response rapidly and transiently as high as 60-fold to peptides of low immunogenicity (glucose transporters, GLUT3 and GLUT4). Such antisera are suitable for immunocytochemistry with excellent anatomic detail. This method may be generally useful as it is reproducible and can yield progressive titer increases when repeated.

Reaction patterns of islet-cell autoantibodies in Macaca nigra.

Circulating islet-cell autoantibodies (ICAAs) that reacted specifically with cytoplasmic components have been found in the blood of prediabetic Macaca nigra. The three distinct reaction patterns observed involved the majority of islet cells throughout the islet; a moderate number of cells, mainly at the islet periphery and around the vasculature; and a few cells scattered throughout the islet. Pancreas sections incubated with sera containing ICAAs followed with peroxidase-conjugated antibody were then reacted with anti-insulin, antiglucagon, or antisomatostatin antisera. The pattern associated with most of the islet cells was shown to be reactive to beta cells and was termed B-ICAA; the pattern with cells at the periphery was identified as alpha cells (A-ICAA); and the scattered cells contained somatostatin (D-ICAA). None of the three islet hormones were able to block ICAA reaction after overnight incubation, so the ICAAs are not anti-islet hormone antibodies. The varied reactions with antigens of different secretory cells indicate release of a variety of immunogens from islet cells as they necrose and cause the formation of different ICAAs.

Lipoprotein patterns in nondiabetic, borderline diabetic, and diabetic Macaca nigra.

Lipoproteins were isolated by sequential ultracentrifugation, and the concentrations and compositions were determined in nondiabetic (ND), borderline diabetic (BD), and diabetic (D) Macaca nigra males consuming a chow ration. The total concentrations and components of the VLDL and IDL increased significantly with metabolic deterioration (P less than 0.01). Concentrations and components of LDL increased in the BD and D monkeys, but changes were not statistically significant. The HDL2 and HDL3 particles were virtually unchanged among the three different metabolic groups. The VLDL was the major carrier of the triglycerides, especially in D monkeys. Cholesterol was present predominantly in the LDL. The LDL-cholesterol to HDL-cholesterol ratio increased in the BD and D monkeys, owing mainly to increases in the LDL-cholesterol content. Apoprotein antisera showed apoprotein B in the VLDL, IDL, and LDL, apoprotein E in the VLDL and IDL, and apoprotein A-I in the HDL2 and HDL3 fractions. Because Macaca nigra consume a nonatherogenic, low-cholesterol, low-fat ration, the changes in lipoproteins, particularly in VLDL and IDL, are attributable to metabolic alterations associated with diabetes.

Changes in islet cell composition during development of diabetes in Macaca nigra.

The islets of Langerhans in sections from the pancreas tail of Macaca nigra were stained by antiserum to insulin, glucagon, or somatostatin. The area of stained cells per total area of the islets was determined by a computerized photometric method. Insulin of the beta cells occupied 77% of the islet area in nondiabetic (ND) monkeys and decreased to 62% in monkeys in the earliest stages of metabolic deterioration, i.e., hormonally impaired (HI) monkeys. At the later stage of borderline diabetes (BD), monkeys had only 39% of the islet area occupied by insulin and the area was diminished to less than 1% in diabetic (D) monkeys. Islets in HI monkeys had an unusual pattern in which only the beta cells in the periphery of islets were stained. Glucagon in the alpha cells stained 7% of the islet area in ND monkeys, but the area was almost doubled to 13% in HI monkeys; the percentage decreased to about 5% in BD and 3% in D monkeys. Somatostatin accounted for 5% of the islet area in ND monkeys, was slightly greater at 7% in HI monkeys, and decreased to 3% in BD and 2% in D monkeys. Alterations in percentages of secretory cells correlated with several of the metabolic and clinical changes.

Immunoreactive glucagon in nondiabetic and diabetic Macaca nigra.

The primary form of immunoreactive glucagon (IRG) in Macaca nigra has been identified as pancreatic, alpha-cell-size glucagon (IRG3500), with a molecular weight of about 3500. Assays with 30K and K-964 glucagon antibodies gave virtually identical results. Column chromatography of plasma on Bio-Gel P-30 indicated ony minimal amounts of high-molecular-weight IRG. Levels of IRG decrease during a glucose infusion, a response expected of IRG3500. IRG concentrations apparently greater than human values appear to be characteristic of nonhuman primates. Nondiabetic Macaca nigra average 641 pg of IRG3500/ml. Borderline diabetic monkeys with moderately increased glucose and impaired glucose clearance average 2,938 pg/ml. Diabetic monkeys with hyperglycemia and diminished glucose clearance have 375 pg of IRG3500/ml. Changes in IRG3500 are related to a lesion in the islets of Langerhans.

Solute flux coupling in a homopore membrane.

Our previous studies on solute drag on frog skin and synthetic heteropore membranes have been extended to a synthetic homopore membrane. The 150-A radius pores of this membrane are formed by irradiation and etching of polycarbonate films. The membrane is 6-microm thick and it has 6 x 10(8) pores cm(-2). In this study, sucrose has been used as the driver solute with bulk flow blocked by hydrostatic pressure. As before on heteroporous membranes, the transmembrane asymmetry of tracer solute is dependent on the concentration of the driver solute. Tracer sucrose shows no solute drag while maltotriose shows appreciable solute drag at 1.5 M sucrose. With tracer inulin and dextran, solute drag is detectable at 0.5 M sucrose. These results are in keeping with the previous findings on heteropore membranes. Transmembrane solute drag is the result of kinetic and frictional interaction of the driver and tracer solutes as the driver flows down its concentration gradient. The magnitude of the tracer flux asymmetry is also dependent on the size of the transmembrane pores.