Insulin Resistance and tau Pathology

Read Also

Besides accumulation of Aβ and plaque formation, AD is characterized by the formation of intracellular neurofibrillary tangles (NFTs) composed of the hyperphosphorylated microtubule associated tau, a protein mainly expressed in neurons. When hyperphosphorylated, tau cannot interact with microtubules, leading to altered microtubule dynamics and compromised axonal transport (Iqbal and Grundke-Iqbal, 2008). Gradual accumulation of hyperphosphorylated tau and formation of NFTs has therefore been associated with neuronal dysfunction and ultimate cell death (Ballatore et al., 2007; Iqbal and Grundke-Iqbal, 2008). In AD, formation of NFTs is thought to be required for clinical expression of the disease and in related tauopathies it can lead to dementia in the absence of Aβ plaques. Also, the degree of cognitive deterioration in AD correlates well with the NFT-load in brains of Alzheimer patients (Nagy et al., 1995), thus it is important to understand whether, and/or how, these major pathophysiological characteristics of AD, i.e. Aβ toxicity and tau pathology, may be mechanistically interrelated. Interestingly, insulin resistance may prove to be a promising link (see fig. 2).  

As tau hyperphosphorylation is central to NFT formation, there has been extensive research on tau kinases, such as GSK-3, regulating this process (Ballatore et al., 2007; Iqbal and Grundke-Iqbal, 2008). Consequently, aberrant activity of GSK-3 is thought to play a pivotal role in AD pathogenesis and the ‘GSK-3 hypothesis of AD’ is based on this presumption (Hooper et al., 2008). GSK-3 is known as a serine-threonine kinase that has been shown to phosphorylate tau at the same sites as in NFTs present in AD. Since GSK-3 is a downstream target of insulin signaling, and several studies have reported that insulin signaling is potentially impaired in the brain of AD patients (Cook et al., 2003; Frolich et al., 1998; Perez et al., 2000; Steen et al., 2005), a resulting overactivation of GSK-3 may provide a link between impaired insulin signaling in the CNS and tau pathology. In line with this assumption, Jolivalt et al. reported that phosphorylation of both AKT and GSK-3 was significantly reduced in a mouse model of type 1 diabetes mellitus (T1DM), i.e. streptozotocin-induced insulin deficient mice, after 9 weeks of diabetes (Jolivalt et al., 2008). GSK-3 is a substrate of AKT and is functionally inactivated following its phosphorylation. Thus reduced phosphorylation and activation of AKT is associated with reduced phosphorylation and therefore increased activation of GSK-3. Most likely as a consequence, a gradual increase in tau phosphorylation was observed, also reaching statistical significance after 9 nine weeks. Interestingly, Aβ levels were also significantly increased in the brain of insulin deficient mice (Jolivalt et al., 2008). However, the question whether increased Aβ levels are the result of impaired insulin signaling or vice versa, remains less clear. Furthermore, it is important to note that these findings were not reproduced in db/db mice, a model of T2DM (Jolivalt et al., 2008). The authors conclude that this discrepancy could indicate that aberrant GSK-3 activity and its consequences may develop earlier and less subtle in insulin-deficient hyperglycemic T1DM mice, than in insulin-resistant hyperglycemic T2DM mice (Jolivalt et al., 2008). For that reason, the short duration of diabetes in this study, i.e. 9 weeks, may provide an explanation for the absence of activity changes in AKT signaling, in db/db mice. In line with that hypothesis, a similar study reported changes in expression of insulin/AKT signaling components and phosphor-tau levels in rat models with spontaneous onset of both T1DM and T2DM (Li et al., 2007). Remarkably, these changes appeared more severe in the rat model of T2DM, leading the authors to believe that these aberrancies may be associated with insulin resistance rather than insulin deficiency.

The association between tau pathology and insulin signaling is further supported by the findings that reduced GSK-3β phosphorylation and thus increased activation, correlated with increased tau phosphorylation in neuron-specific IR knockout mice showing impaired insulin signaling (Schubert et al., 2004). These data demonstrate directly the relative contribution and importance of IR signaling for the activation of this pathway in the CNS and further emphasize how neuronal insulin resistance potentially predisposes for neuronal dysfunction and degeneration. Also, in a previous study the same research group showed that knock-out of the insulin receptor substrate 2 (IRS2) in mice, which leads to diabetes resulting from impaired pancreatic β-cell survival and function (Withers et al., 1998), causes hyperphosphorylated tau to accumulate in the hippocampus of old mice (Schubert et al., 2003). Since IRS2 signaling also mediates peripheral insulin action and promotes pancreatic β-cell function, dysregulation of IRS2 signaling could provide the molecular basis for understanding the relationship between neuronal dysfunction and peripheral insulin resistance and diabetes.

ADDL-mediated IR internalization, may thus promote tau phosphorylation in a similar fashion, by decreasing insulin/AKT signaling. Nevertheless, in the IR knockout mice described above, no memory deficits were reported (Schubert et al., 2004). As Schubert and colleagues note, this could indicate that changes in insulin/AKT signaling must coincide or interact with other mechanisms for development of clinical symptoms of AD. Consistent with ADDL-induced tau phosphorylation, De Felice and colleagues reported that soluble ADDL oligomers, either prepared in vitro or extracted from the AD brain, stimulate tau phosphorylation in mature rat hippocampal neurons and in neuroblastoma cells at amino acid residues that are characteristic of hyperphosphorylated tau in AD (De Felice et al., 2008). Furthermore, it has been reported that insulin depletion by streptozocin treatment, leads to hyperphosphorylation of tau in rodent brains (Clodfelder-Miller et al., 2006; Grunblatt et al., 2007). Streptozotocin enters cells through GLUT2 and is mostly toxic for insulin producing/secreting β-cells of the pancreas, but as insulin-producing neuronal cells are also believed to express GLUT2 (Brant et al., 1993), streptozotocin is also thought to affect insulin signaling in the brain.

 Based on the available data, it is conceivable that in non-insulin resistant or -diabetic individuals, overactivation of GSK-3 and subsequent hyperphosphorylation of tau, is in part a consequence of ADDLs impairing brain insulin/AKT signaling, perhaps in a similar fashion as insulin resistance itself may affect brain insulin signaling and potentially hereby increase the risk for developing AD. In summary, ADDLs appear capable of accounting for key features of AD pathology and cognitive failure, consistent with the unifying hypothesis that soluble oligomers of Aβ act as proximal neurotoxins in AD. ADDLs are thought to bind to extracellular target receptors or intracellular signaling components, thereby contributing to impaired activation of neurotrophic signaling and to an aberrant set of downstream events that lead to tau hyperphosphorylation and neuronal dysfunction. In addition, failure to downregulate GSK-3 may adversely affect neurogenesis and axonogenesis. Several studies show that localized inhibition of GSK-3β is crucial for proper axonogensis (Ciani and Salinas, 2005; Jiang et al., 2005; Shi et al., 2004; Yoshimura et al., 2005). Taken together, regulation of GSK-3 activity and associated tau phosphorylation may be important in maintaining neuronal contacts, potentially by affecting the organization and dynamics of the cytoskeleton.