The Central Nervous System

Read Also

Physiologically, glucose is the major nutrient for the brain. Therefore, the undisturbed supply and metabolism of glucose in the brain, is of central significance. A role for insulin in regulation of this process by acting in the brain is becoming increasingly acknowledged. In addition to insulin crossing the blood-brain barrier (BBB) by a saturation dependent transport mechanism, a small proportion of brain insulin is produced in the brain itself by pyramidal neurons residing in the hippocampus, prefrontal cortex, entorhinal cortex and the olfactory bulb (Hrytsenko et al., 2007; Madadi et al., 2008; Plata-Salaman, 1991; Singh et al., 1997). Insulin gene expression and insulin synthesis have been demonstrated in both immature and mature mammalian neuronal cells (Schechter and Abboud, 2001; Schechter et al., 1992). Insulin mRNA was found to be distributed in the brain, with the highest density in pyramidal cells of the hippocampus and high densities in the prefrontal cortex, the enthorhinal and perihinal bulb, as well as in the hypothalamus. Two different types of IRs have been found in the adult mammalian brain, a neuron-specific brain type with an α-subunit  of 118 kDa and a β-subunit of 91 kDa which is not downregulated by insulin, and a peripheral type on glial cells with an α-subunit of 130 kDa and a β-subunit of 95kDa which is downregulated by insulin (Adamo et al., 1989). In the brain, insulin’s various functions are dependent on IR distribution. For example, in the hypothalamic regions, signaling through the IR appears to be vitally involved in the regulation of food intake and body weight (Guthoff et al., 2010; Plum et al., 2006). However, in the higher limbic system, including the hippocampus, signaling through the IR has been involved in mediating biochemical stimuli necessary for synaptic activity, plasticity and memory function (Guzowski et al., 2000; Kremerskothen et al., 2002; Lannert and Hoyer, 1998; Park, 2001; Steward et al., 1998; Zhao et al., 1999).

Several studies show that hippocampal cognitive performance may be dependent on an adequate supply of glucose to meet the metabolic demands of cognitive processes, and that a sufficient supply of glucose to the hippocampus can enhance memory performance (Gold, 2005; McNay et al., 2000; McNay and Gold, 2002). Similar to its effects in the periphery, insulin has been thought to stimulate translocation of GLUTs to the membrane of hippocampal neurons, indicating that insulin may increase hippocampal glucose uptake and/or metabolism (McEwen and Reagan, 2004; Reagan, 2005; Thorens and Mueckler, 2010). Several in vitro studies are consistent with a role for insulin in modulation of synaptic plasticity (Izumi et al., 2003; van der Heide et al., 2005; Zhao et al., 2004). Hence, there are substantial reasons for believing that hippocampal functions such as learning and memory may in part be dependent on insulin acting in the brain. In support of this hypothesis, recent animal studies have suggested that delivery of insulin to the hippocampus may be able to modulate hippocampal memory processes. However, these studies have mostly used insulin doses that exceed the physiological situation or aversive methods with elevated systemic glucose and epinephrine, making interpretation of a specific insulin effect difficult (Babri et al., 2007; Moosavi et al., 2007; Moosavi et al., 2006). Nevertheless, a recent study by McNay and colleagues shows that delivery of physiological doses of insulin, but not insulin like growth factor (IGF-1), directly to the hippocampus, specifically enhances spatial working memory via a PI3K-dependent mechanism (McNay et al., 2010). Administration of insulin also increased local glucose removal from interstitial fluid, whereas blockade of endogenous hippocampal insulin action, either with PI3K antagonists or small anti-insulin antibody-like peptides, impaired cognitive performance below baseline. These data are in support of a role for insulin in physiological hippocampal memory processes.

As the brain consumes 18-30% of total body glucose (Magistretti, 2006), disruptions in glucose supply and utilization regulated by insulin signaling, can easily result in neuronal dysfunction or damage. Indeed, in the developing brain, hyperglycemia and hyperinsulemia are accompanied by neuronal death, whereas severe chronic hypoglycemia can lead to permanent brain damage and mental retardation (Dunne et al., 2004). Insulin resistance and the sometimes resulting T2DM have long been associated with impaired cognitive function (Awad et al., 2004; Biessels et al., 2008; McNay et al., 2010; Stewart and Liolitsa, 1999; Strachan et al., 1997), thus it is crucial to know what happens to insulin signaling in the brain, as a consequence of insulin resistance and peripheral hyperinsulemia. Insulin is readily transported into the CNS across the BBB by a saturable, receptor-mediated process (Simpson et al., 2007). Based on a study regarding the kinetics of CNS insulin uptake after dexamethasone treatment (Baura et al., 1996), Craft postulates that peripheral insulin resistance and subsequent hyperinsulinemia does not promote parallel increases of insulin in the brain and actually reduces transport into the brain (Craft, 2006). Although direct evidence for this assumption is still lacking, this could indicate that in insulin resistant subjects, insulin signaling in the brain may be deficient due to reduced levels of insulin. Contradictory, Craft and colleagues have also reported that intravenous insulin administration in humans induces transport of insulin into the CNS (Craft, 2006; Craft et al., 2003), which would be more conceivable, as insulin is transported through the BBB by a saturable mechanism.

Due to loss of lean body mass and an increase in adipose tissue during ageing, less muscle tissue and relatively more adipose tissue is available for glucose clearance, thereby increasing the risk for insulin resistance in genetic susceptible individuals. Consistent with reduced brain insulin levels as a result of increased insulin resistance and hyperinsulemia with due to ageing, Frölich and colleagues reported that brain insulin levels are decreased in elderly, as is the number of IRs (Frolich et al., 1998). Interestingly, they also showed that in SAD patients, compared to age-matched controls, the number of IRs was increased in all regions of the brain cortex, with a significant increase in the occipital cortex. The authors suggest that this finding is due to an upregulation of the IR on top of a general loss of IRs with ageing, potentially because of impaired downstream signal transduction. Accordingly, Frölich and colleagues also reported that IR tyrosine kinase activity was reduced in SAD patients compared to both middle-aged and age-matched controls. These data are in support of the hypothesis that impaired insulin signaling may represent an important pathogenic event in AD.

In contrast to studies reporting that brain insulin levels decrease with age, as well as in AD, some have reported that insulin levels rise with ageing and are strong predictors of cognitive decline in adults without diabetes (Stolk et al., 1997a; Stolk et al., 1997b). As the authors note, these findings appeared to be compatible with a direct effect of insulin on the brain. Also, these results are to be expected as insulin resistance is associated with ageing, and is likely to lead to increased brain insulin levels. An in vivo study by Gerozissis and colleagues, further demonstrated that brain insulin levels are regulated locally and are not always dependent of pancreatic insulin regulation, further emphasizing the complexity of the regulatory process of brain insulin levels and downstream signal transduction (Gerozissis et al., 2001). Correspondingly, different studies characterized and quantified AD-associated abnormalities in insulin, IGF and IR gene expression, and associated downstream signaling mechanisms using human AD and age-matched control post mortem brain samples (Rivera et al., 2005; Steen et al., 2005). They found that insulin-, IGF-I-, IR-, and IGF-I receptor-positive neurons were less abundant in the AD brain compared with the age-matched control samples. Reduced labeling of neurons in AD cases appeared attributable to both loss of neurons as well as reduced neuronal expression. Furthermore, the authors reported that levels of activated AKT (phospho-AKT) were also reduced in the AD brain, in contrast to GSK-3β activity which was increased. Combined, these studies convincingly implicate aberrant insulin signaling in the pathophysiology of AD.

In summary, to elucidate to which extent insulin signaling is misregulated during ageing, or contributes to AD pathology, it is important to understand how insulin signaling is affected as a result of ageing, or as a possible cause or downstream effect in the pathogenesis of AD (see fig. 2). Most data appear to be in agreement with the notion that downstream signaling through the IR decreases with ageing and is further impaired in AD. Therefore it is possible that the increased risk for AD as a result of T2DM or insulin resistance, is related to reduced output of insulin signaling pathways in AD associated brain areas. Interestingly, streptozotocin-induced insulin deficiency promoted cerebral amyloidosis and cognitive deficits in an Alzheimer transgenic mouse model expressing a chimeric mouse/human APP (MO/HuAPP695swe) and a mutant human presenilin1 (PSEN1dE9) gene (Wang et al., 2010b). Nevertheless, knockout of the neuronal IR appeared not sufficient to cause cognitive deficits or neuronal loss in mice, even though tau hyperphosphorylation was observed, indicating that impaired insulin signaling must perhaps interact with other pathogenic mechanisms for the development of AD. Moreover, evidence suggests that downstream insulin signaling is not simply reduced, but that insulin signaling components may be aberrantly translocated (Cole and Frautschy, 2007). Consistent with an aberrant activation and translocation event, phospho-ser473AKT was found to be increased in membrane particulate fractions, but reduced in cytosolic fractions from AD brains (Griffin et al., 2005). In conclusion, it remains difficult to reliably evaluate phosphorylation patterns in postmortem AD and control brain, and more importantly, it remains to be demonstrated that aberrant fluctuations in brain insulin levels and/or signaling in AD, are high enough to be functionally significant to neurons.