Amyloid-β and Neuronal Insulin Resistance − Rakyat Biologi

It’s becoming increasingly acknowledged that AD may be closely linked to a state of insulin resistance in the brain (de la Monte et al., 2006; Ronnemaa et al., 2008). Although the effects of insulin in the CNS have long been unclear, it is now thought that signaling through this pathway is important for many aspects of neuronal function, including plasticity and memory functions (Gispen and Biessels, 2000; Zhao et al., 1999; Zhao and Alkon, 2001). Why memory is specifically targeted in AD has long been a fundamental mystery, but it is becoming evident that the crucial pathogenic event may be the functional and morphological deterioration of specific memory center synapses induced by potent neurotoxins, such as soluble oligomeric forms of Aβ that accumulate in the AD brain (Ferreira et al., 2007; Haass and Selkoe, 2007; Klein et al., 2001; Selkoe, 2002; Terry et al., 1991; Zhao et al., 2008). Impairment of the insulin/AKT signaling pathway specifically, has emerged as an important consequence of soluble Aβ neurotoxicity and is speculated to result in a disruption of hippocampal synaptic function and memory deficits (De Felice et al., 2009; Lee et al., 2009b; Townsend et al., 2007; Zhao et al., 2008). Accordingly, several studies convincingly show that ADDL neurotoxicity can be linked to memory impairment in rodents (Cleary et al., 2005; Kawarabayashi et al., 2004; Lesne et al., 2006), whereas in human post-mortem AD brains, the presence of apparent ADDL-like oligomeric assemblies correlated with memory loss (Gong et al., 2003). Nevertheless, the molecular mechanisms of how these small Aβ oligomers induce aberrant inhibition of insulin signaling in the brain remains less clear and several hypotheses have been proposed on the subject.

Accumulating evidence suggests that ADDL’s may bind to several molecules on the neuronal surface (Townsend et al., 2007). Due to a sequence similarity between Aβ and insulin, which are both thought to be substrates for the insulin degrading enzyme (Cook et al., 2003; Kurochkin, 1998), Xie and colleagues investigated the effects of Aβ peptides on insulin binding to the IR (Xie et al., 2002). They found that Aβ reduced the affinity of insulin for its receptor, potentially by acting as a direct competitor for insulin binding and action. Correspondingly, another study investigated how soluble Aβ oligomers interfere with some of the signal transduction cascades that mediate activity dependent synaptic plasticity (Townsend et al., 2007). Using cultured hippocampal neurons from wild type mice, Townsend and colleagues showed that ADDL’s co-immunoprecipitate with the IR and inhibit its autophosphorylation and downstream signal transduction in a similar fashion as antagonists of the IR. As AKT activity is thought to be important for long-term potentiation (LTP), a cellular mechanism crucial for memory functions and learning in the hippocampus (Lynch, 2004; Walsh et al., 2002), disruption of this pathway due to extracellular ADDL toxicity may lead to cognitive dysfunction. Consistent with this hypothesis, it has recently been reported that insulin ameliorates ADDL induced inhibition of LTP in the CA1 region of hippocampal slices (Lee et al., 2009a).

In addition, using rat primary hippocampal cultures, De Felice and colleagues showed that insulin can block ADDL binding to synaptic sites, thereby preventing their toxic effects (De Felice et al., 2009). It has previously been reported that ADDLs induce major aberrant loss of IRs from neuronal dendrites by a process involving selective endocytosis of IRs (Zhao et al., 2008), which is potentially mediated by deregulation of casein kinase II (CK2) and Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) (De Felice et al., 2009). Under normal conditions both kinases are involved in activity-dependent internalization of NMDA receptors (Chung et al., 2004), which are thought to be important in neuronal plasticity. Remarkably, De Felice and colleagues found that insulin was able to completely block ADDL-induced loss of dendritic IRs (De Felice et al., 2009). In this study, double-labeling for IRs and ADDLs showed that administration of insulin to hippocampal neurons exposed to ADDLs, significantly reduced ADDL-binding to synaptic sites. The authors note that the ability of insulin to block binding of ADDLs to synaptic sites appears not to be the result of direct competition for a common binding site. In fact, inhibition of IR protein tyrosine kinase activity by AG1024 completely abolished the ability of insulin to block ADDL binding, indicating that inhibition of ADDL binding by insulin involves an IR signaling-dependent downregulation of ADDL binding sites (De Felice et al., 2009). In the same study, addition of the insulin-sensitizing drug rosiglitazone enhanced the ability of submaximal insulin levels to block binding of ADDLs to synaptic sites, emphasizing that ADDL-binding is insulin signaling dependent. In summary, physiological insulin and pathological ADDLs negatively regulate the availability of each other’s binding sites, i.e. insulin signaling appears to downregulate ADDL binding, while ADDLs downregulate IR availability. Interestingly, a marked loss of IRs from dendrites in the brain has also been found in a recent neuropathology study of the AD brain (Moloney et al., 2010), and could thus be in line with the neurotoxic effect of ADDLs discussed above.

In contrast to soluble extracellular Aβ species acting on the insulin signaling pathway, it has also been proposed that intraneuronal Aβ oligomers may exert adverse effects, downstream of the IR and its associated adaptor proteins (Lee et al., 2009b). In a previous study on the effects of intracellular Aβ on insulin signaling, Magrane and colleagues reported that intraneuronal Aβ caused a decrease in levels of phosphorylated AKT (Ser473) in rat primary cortical neurons expressing human Aβ-42, as well as in transgenic mice (Tg2576) overexpressing human FAD mutant APP (KM-670/671-NL) (Magrane et al., 2005). In the same study, extracellular administration of Aβ-42 in vitro did not result in a significant change in phosphorylation of AKT, although it did appear to contribute to cell death. These data indicate that insulin signaling may be impaired as a result of intracellular effects of Aβ. Consistent with reduced AKT activation, levels of both phosphorylated AKT (Thr308) and GSK-3β (Ser9) were decreased in post mortem brain samples of AD patients as compared to healthy controls, indicating that GSK-3β may have been overactive as a result of decreased activity within the insulin signaling pathway (Lee et al., 2009b). In vitro, under insulin-stimulated conditions, AKT1 activity and subsequent GSK-3β phosphorylation were significantly reduced in C₂C₁₂ myotube cultures overexpressing Aβ-42 (Lee et al., 2009b). Importantly, in the same cultures, activation of the IR appeared independent of the presence of intracellular Aβ-42. These results are in line with the presumption that ADDLs, in addition to possible extracellular effects, can act intracellular and downstream of the IR to inhibit the insulin pathway’s output.

Moreover, interaction of Aβ with both AKT1 and PDK1 separately, was found to be increased in the AD brain samples, whereas an Aβ interaction with control proteins such as protein kinase A (PKA) was not found. As in the AD brain, ADDLs appear to interact with PDK and AKT in these C₂C₁₂ myotubes, and were found to impair the PDK-AKT interaction, thus disrupting PDK-mediated phosphorylation and activation of AKT. These data again indicate that Aβ species may have specific affinity for PDK and/or AKT, key components of the insulin pathway. Whether this interaction has a significant contributing effect in AD pathogenesis, remains uncertain. It should also be noted, that analysis of post mortem brain material does not necessarily address primary causative events in a pathogenic cascade. Initial molecular events will have to be studied using model systems, which due to their inherent differences may provide conflicting data.