A growing body of evidence supports an intriguing clinical/epidemiological connection between Alzheimer disease (AD) and type 2 diabetes (T2D). T2D patients have significantly increased risk of developing AD and vice versa. Recent studies have begun to reveal common pathogenic mechanisms shared by AD and metabolic disorders, notably obesity and T2D. In T2D and obesity, low-grade chronic inflammation is a key mechanism leading to peripheral insulin resistance, which progressively causes tissue deterioration and overall health decline. In the brain, proinflammatory signaling was recently found to mediate impaired neuronal insulin signaling, synapse deterioration, and memory loss. Here, we review evidence indicating that inflammation, insulin resistance, and mitochondrial dysfunction are common features in AD and T2D. We further propose the hypothesis that dementia and its underlying neuronal dysfunction are exacerbated or driven by peripheral inflammation. Identification of central and peripheral inflammation as potential mediators of brain dysfunction in AD may lead to the development of effective treatments for this devastating disease.
The incidence of metabolic disorders, including type 2 diabetes (T2D) and obesity-related insulin resistance, is increasing at alarming rates worldwide, largely due to poor lifestyle habits. Diabetes, for example, is now estimated to affect 382 million people worldwide, and this appalling figure may further rise to almost 600 million people in 2035 (1). In parallel, the prevalence of Alzheimer disease (AD), the most common form of dementia in the elderly, also increases as the world population ages. Thirty-five million people are now thought to be affected by AD worldwide, and this number is expected to double in the next few decades (2). Besides being major causes of morbidity and mortality, AD and metabolic diseases appear to be intimately connected. Since the original Rotterdam study (3), epidemiological/clinical observations have accumulated showing that diabetic patients are significantly more likely to develop cognitive deterioration and exhibit increased susceptibility to dementia, particularly AD (4). Indeed, studies by Ott and colleagues (3,5) have shown that dementia is twice as frequent in diabetic patients as in normal subjects. More generally, the incidence of other neurological disorders, such as vascular pathology and stroke, also appears to be doubled in T2D individuals compared with nondiabetic subjects (6,7).
Importantly, impaired metabolic parameters, such as hyperglycemia and hyperinsulinemia, positively correlate with development of AD-related pathology (8,9). Elevated blood glucose levels increase the hazard risk of dementia in both diabetic and nondiabetic individuals (by 40 and 18%, respectively) (4) and are associated with cognitive decline and reduced hippocampal volume (10). These findings indicate that persistently elevated blood glucose levels negatively impact the brain, even in the absence of overt T2D or impaired glucose tolerance. A new view is thus emerging according to which even a prediabetes state maintained throughout a long period of life may constitute a significant risk factor for dementia. In harmony with this concept, AD was recently proposed to be a form of dementia caused by metabolic dyshomeostasis that manifests in the elderly as a result of a cumulative, lifelong impact on peripheral tissues and on the brain (11,12). In addition to this metabolic hypothesis, it is important to note that prediabetic hyperglycemia may be accompanied by a state of low-grade peripheral inflammation, which might directly impact brain insulin signaling (as discussed below). Thus, from a mechanistic point of view, the connection between AD and T2D may comprise both inflammatory and metabolic components. A corollary of this proposal is that poor lifestyle habits (e.g., lack of or insufficient physical activity or inadequate nutrition) known to predispose to T2D and obesity are increasingly thought to play important roles in susceptibility to AD later in life (11,13).
Obesity is recognized as the most important epidemiological factor predisposing to T2D. Several studies have shown that high-fat feeding leads to important alterations in hypothalamic and peripheral systems that regulate food intake and energy expenditure (14). A recent study showed that the hypothalamus of obese humans and of experimental models of obesity presents signs of inflammation and dysfunction that precede the installation of increased adiposity and T2D (15). Obesity, often associated with consumption of diets that are high in fat, has also been proposed to increase the risk of dementia and AD later in life (16–18). Epidemiologic studies further suggest that diets high in saturated fats during midlife are a risk factor for development of AD later in life (19,20). Interestingly, induction of diabetes or obesity in animals triggers and/or accelerates AD-like pathology (21–23). Therefore, adopting a healthy lifestyle and strategies aimed at lifelong control of blood glucose levels, which are thought to prevent the development of metabolic diseases, may be of further benefit to preserve cognition in the elderly and to prevent AD development.
In the following sections, we discuss molecular events and pathways recently implicated in disrupted brain insulin signaling, inflammation, and cellular/metabolic stress in AD and review recent evidence indicating that inflammation, defective insulin signaling, and mitochondrial dysfunction are common molecular denominators connecting T2D to AD. We further highlight the need to unravel how metabolic disorders, especially T2D, negatively influence brain function, ultimately leading to dementia, and propose that peripheral inflammation and insulin resistance may be linked to clinical manifestation of AD.
Defective Insulin Signaling in T2D and AD
Insulin resistance, defined as a smaller than expected response to a given dose of insulin, is a pathological state characteristic of aging and of several disorders, including cancer, polycystic ovarian disease, infections, and trauma. Most significantly, insulin resistance is a major hallmark of obesity and T2D, two highly prevalent metabolic disorders. Knowledge of the molecular mechanism(s) of defective insulin signaling has increased tremendously during the past decades (24,25). Insulin exerts its actions via a complex signaling network that has multiple effects on metabolism and growth. Impaired insulin signaling can result, for example, from mutations or aberrant posttranslational modifications in molecular components of the insulin signaling pathway (25–27). Major insight into the mechanisms by which insulin resistance develops in peripheral tissues has come from studies performed in the laboratories of Ronald Kahn, Bruce Spiegelman, and Gökhan Hotamisligil, among others (28–34). Seminal articles from these groups have demonstrated that prolonged metabolic stress and activation of proinflammatory pathways lead to attenuated insulin signaling and decreased cellular responsiveness to insulin.
Insulin resistance acutely impairs the ability of cells to maintain energy homeostasis. Intriguingly, AD brains present similar abnormalities as peripheral tissues in T2D, including metabolic stress and neuroinflammation (35–39). Thus, it is conceivable that similar mechanisms account for peripheral insulin resistance in T2D and impaired brain insulin signaling in AD. Recent studies have indeed demonstrated substantial commonalities between neuropathogenic mechanisms triggered by amyloid-β oligomers (AβOs), toxins that accumulate in the AD brain and are increasingly thought to underlie synapse failure and memory loss, and mechanisms involved in peripheral insulin resistance in diabetes (12,37,40).
It is important to consider the toxic role of AβOs in AD, as their impact in the brain appears intimately related to defective neuronal insulin signaling. AβOs are soluble aggregates of Aβ, a 4 kDa peptide that exhibits a high propensity to self-associate in aqueous medium. Abnormal production, processing, and/or clearance of Aβ may lead to its accumulation and aggregation in the brain parenchyma and interstitial fluid. Early histopathological investigation of AD brains revealed the presence of fibrillar Aβ aggregates forming large, insoluble deposits known as amyloid plaques. Further in vitro biochemical and cell biology studies, as well as studies using a number of transgenic mouse models of AD, provided strong support to what initially seemed to be a solid concept, namely that Aβ fibrils/plaques played crucial roles in AD pathogenesis (reviewed in 41,42). However, a large body of evidence accumulated during the past 15 years indicates that fibrils are probably not the most harmful structures generated by self-association of Aβ. Of direct clinical relevance, postmortem analysis of the brains of individuals who died without signs of significant cognitive deterioration has revealed abundant brain amyloid deposits, whereas individuals lacking such deposits have been found to exhibit various extents of cognitive deterioration (e.g., 43). Moreover, the best correlate of the extent of dementia is not brain amyloid burden but rather synapse loss (44,45), suggesting that synapse deterioration and cognitive impairment are caused by a toxin other than fibrillar Aβ.
Landmark studies by William L. Klein and colleagues (46) first addressed this controversy, showing that Aβ self-aggregates to form neurotoxic soluble oligomers, aggregates much smaller than fibrils that are not detected in classical neuropathological examination. Oligomers were recently detected at postsynaptic sites in AD hippocampi (47), and their levels are elevated in the brain and cerebrospinal fluid of AD patients (48,49). Interestingly, the absence of AβOs at the postsynapse was reported in cognitively intact elderly individuals whose brains presented amyloid deposits (47). Klein’s discovery, confirmed and expanded by several groups, led to a novel hypothesis on how AD progression leads to dementia: synapse failure and neuronal dysfunction are now considered to derive from the accumulation and impact of AβOs in AD brains (41,42,50).
Recent evidence indicates that AD can be considered a brain-specific form of diabetes. AD brains exhibit defective insulin signaling, altered levels and/or aberrant activation of components of the insulin signaling pathway, and, importantly, decreased responsiveness to insulin (35,36,38). Molecular clues into how the brain becomes insulin resistant in AD came from studies demonstrating that AβOs trigger the removal of insulin receptors (IRs) from the plasma membrane in cultured hippocampal neurons (51,52), leading to reduced IR protein tyrosine kinase activity (52). IRs are widely distributed in the central nervous system (CNS), suggesting that insulin has important physiological roles in the brain. The hippocampus, a region that is fundamentally involved in the acquisition, consolidation, and recollection of new memories, presents particularly high levels of IRs. Insulin has been shown to be neuroprotective (37,51,53,54) and to modulate synapse plasticity mechanisms (55). IR signaling further regulates circuit function and plasticity by controlling synapse density (56). In cultured hippocampal neurons, IRs exhibit a punctate dendritic distribution (51,52) consistent with synaptic localization. Nonetheless, knowledge of the precise roles of brain IRs is still limited. Although a number of studies have reported learning-associated changes in IR pathways and beneficial actions of insulin on memory (57), specific deletion of brain IRs did not lead to major learning and memory impairment in mice (58). It is, nonetheless, possible that compensatory mechanisms may operate to prevent memory deficits in such mice by stimulation of insulin signaling–related pathways via other receptors, including IGF-1 and GLP-1 receptors (GLP-1Rs). Altered neuronal IR function thus appears to be an important aspect of the overall synaptic and neuronal pathology induced by AβOs.
In peripheral tissues, activated IRs recruit and phosphorylate members of a conserved family of adaptor proteins called IR substrates 1–4 (IRS1–4) (59). Upon phosphorylation at tyrosine residues, IRS proteins act as scaffolds that couple IR stimulation to downstream effectors, such as PI3K, Akt/PKB, and mTORC1 (59), allowing cellular metabolic and transcriptional reprogramming (60). On the other hand, inhibitory serine phosphorylation (pSer) of IRS-1 and IRS-2, the best studied components of the IRS family, causes their dissociation from the IR and decreases tyrosine phosphorylation (pTyr) (60). Therefore, the balance between IRS phosphorylation at serine or tyrosine residues (IRS-1pSer vs. IRS-1pTyr) determines the extent of insulin actions.
In T2D, activation of the stress kinase c-Jun NH2-terminal kinase (JNK) phosphorylates IRS-1 at serine residues (IRS-1pSer), blocking downstream insulin signaling and causing peripheral insulin resistance (28). Similarly, AβOs instigate JNK activation and IRS-1 inhibition in primary hippocampal neurons (37,40) and in the hippocampi of cynomolgus monkeys that received intracerebroventricular infusions of AβOs (37). IRS-1 inhibition was also demonstrated in the brains of a transgenic mouse model of AD (37). Most important in establishing the clinical relevance of these findings was the demonstration of elevated IRS-1pSer (37,38) and activated JNK (37) in postmortem AD brains. Because AβOs trigger internalization and redistribution of neuronal IRs (52), it is possible that removal of IRs from the cell surface facilitates IRS-1pSer, a view consistent with our finding that insulin blocks both neuronal IR downregulation (51) and IRS-1pSer induced by AβOs (37).
Inflammation as a Major Link Between the Pathogenesis of AD and Metabolic Disorders
Inflammation is an important feature of diabetes and AD and is thought to play critical roles in the pathogenesis of both disorders (28,39,61). Inflammation is part of the body’s mechanisms of defense against multiple challenges, including infections and injury, and involves both soluble factors and specialized cells that are mobilized to neutralize such threats and restore normal body physiology (62). Similar inflammatory processes are thought to occur in the brain and in peripheral tissues. Several studies have established the presence of inflammatory markers in the AD brain, including elevated levels of cytokines/chemokines and gliosis (notably microgliosis) (63). Moreover, blood concentrations of inflammatory mediators, including tumor necrosis factor-α (TNF-α), interleukin 6 (IL-6), and IL-1β, are increased in AD patients (64). On the other hand, macrophage activation/infiltration into adipose tissue and overproduction of proinflammatory cytokines, including TNF-α, are key features of the pathophysiology of metabolic disorders. Elevated levels of TNF-α, overexpressed in adipose tissue of obese individuals, cause peripheral insulin resistance (30). Thus, both in the brain and in peripheral tissues, unchecked or chronic inflammation becomes deleterious, leading to progressive tissue damage in degenerative diseases (Fig. 1).
It is interesting to note that inflammation also underlies hypothalamic dysfunction in obesity. The hypothalamus plays a key role in neuroendocrine interaction between the CNS and the periphery (65). Emerging evidence indicates that inflammation and endoplasmic reticulum (ER) stress are critical pathogenic events in the establishment of hypothalamic and peripheral insulin resistance in metabolic disorders (15,66–68). In animal models of T2D and obesity, an inflammatory response in the hypothalamus, notably via activation of TNF-α and the IκBα kinase (IKK)-β/nuclear factor-κB pathway, is an important part of the mechanism underlying pathogenesis (69,70). Therefore, hippocampal dysfunction in AD and hypothalamic deregulation in obesity seem to share common inflammatory pathogenic pathways.
Activation of Proinflammatory and Cell Stress Signaling Pathways Impairs Neuronal Insulin Signaling in AD
In peripheral insulin resistance, aberrant TNF-α signaling leads to activation of JNK (71). Activation of the TNF-α/JNK pathway is linked to major inflammatory/stress signaling networks, including ER stress and the stress kinases IKK (IκBα kinase) and PKR (double-stranded RNA-dependent protein kinase) (72). In T2D, elevated TNF-α levels trigger serine phosphorylation of IRS-1 by stress kinases (30), blocking insulin signaling (71,72). In the brain, TNF-α is mainly secreted by microglial cells in response to trauma, infection, or abnormal accumulation of protein aggregates. TNF-α levels are elevated in brain microvessels and cerebrospinal fluid in AD (73), as well as in the brains of transgenic mouse models of AD (74). Initial evidence that impaired neuronal insulin signaling in AD was linked to proinflammatory signaling came from the finding that AβOs cause IRS-1 inhibition through TNF-α/JNK activation (37). Reinforcing the notion that common mechanisms underlie impaired peripheral insulin signaling in T2D and brain insulin resistance in AD, we recently showed that IKK and PKR, as well as ER stress, reported to be elevated in AD brains (75,76), mediate AβO-induced IRS-1 inhibition in hippocampal neurons (37,54).
IKK, a stress kinase activated by TNF-α signaling in peripheral insulin resistance (77), also mediates AβO-induced neuronal IRS-1 inhibition (37). Overnutrition induces an inflammatory response in peripheral metabolic tissues, a process referred to as “metaflammation” (78), which causes metabolic defects underlying T2D and obesity, including IKK activation (78). The recently established involvement of IKK in IRS-1 inhibition in AD provides additional evidence for a close parallel between inflammation-associated defective brain insulin signaling in AD and chronic inflammation–induced insulin resistance in peripheral tissues. Further studies aimed at exploring the role of IKK in neuronal dysfunction are warranted and may bring novel clues on mechanisms underlying AD pathology.
The double-stranded RNA-dependent protein kinase (PKR), originally identified as a pathogen sensor and a regulator of the innate immune response against viral infections in higher eukaryotes (79), can regulate or act in conjunction with major inflammatory kinases/signaling pathways implicated in metabolic homeostasis, including JNK and IKK (80). Interestingly, PKR is involved in AβO-induced neuronal IRS-1 inhibition (37), further underlining the parallelism between peripheral insulin resistance in T2D and impaired brain insulin signaling in AD.
Current evidence thus indicates that proinflammatory TNF-α signaling and activation of cell stress pathways play key roles in peripheral and central IRS-1 inhibition and in neuronal dysfunction in AD (Fig. 2).
Antidiabetes Agents as Novel Therapeutics in AD: Unraveling the Mechanisms of Neuroprotection
Evidence linking pathogenic mechanisms in the AD brain to mechanisms present in metabolic diseases provides a rationale for using antidiabetes agents as novel therapeutics in AD. As noted above, insulin actions seem to be important for proper function of the hippocampus, a memory-linked brain region that presents high levels of IRs (81). In cultured hippocampal neurons, IRs show a punctate dendritic synaptic distribution (51,52), and IR signaling regulates synaptic plasticity by controlling synapse density (56). In rodents, IR signaling contributes to long-term memory consolidation and improves spatial learning (81–84). Insulin has also been proposed to regulate neuronal survival and to act as a growth factor (85), possibly by activating either its own receptor or IGF receptors (86).
Supporting the notion that insulin-related signaling plays a central role in learning and memory, impaired insulin sensitivity has been linked to cognitive deficits and structural and functional brain deficits in the elderly (87). Therefore, restoring brain insulin signaling might be beneficial to circumvent age-related dementia. Since to date most approaches proposed as possible treatments for AD have disappointingly failed in clinical trials (88), diabetes-based strategies have emerged as potentially effective therapies for AD. Intranasal insulin administration, a preferential route for CNS delivery, improves memory in healthy adults without affecting circulating levels of insulin or glucose (89). Intranasal insulin also enhances verbal memory in memory-impaired subjects and improves cognitive performance in early AD patients (12,90).
More recently, GLP-1R agonists have been proposed as an alternative therapeutic approach or as an addition to insulin-based therapies in AD. GLP-1R agonists, such as exendin-4 and liraglutide, are indeed an attractive option because they activate pathways common to insulin signaling through G protein–dependent signaling (91). GLP-1Rs are present and functional in cultured neurons as well as in rodent and human brains (92,93). In mice, GLP-1R analogs are stable in blood, with most of the injected peptide reaching the brain intact. Furthermore, recent evidence indicates that GLP-1R stimulation facilitates hippocampal synaptic plasticity, cognition, and cell survival (94–96).
An important step now is to understand if and how stimulation of insulin signaling in the brain might facilitate neuroprotection in AD, thereby preserving normal brain function. Molecular mechanisms underlying the protective actions of insulin signaling in the brain have recently started to be unraveled. Insulin and GLP-1R agonists were found to protect neurons against damage induced by AβOs in cellular and animal models of AD. Insulin was found to block IR downregulation (51), IRS-1pSer (37), eIF2α-P (54), and oxidative stress (51) induced by AβOs. Remarkably, insulin protects against synapse loss induced by AβOs (51). Insulin further favors nonamyloidogenic processing of the amyloid precursor protein in vitro (97) and reduces tau hyperphosphorylation in a rat model of T2D (98). Exendin-4 was recently found to block AβO-induced impairment in insulin signaling in hippocampal cultured neurons (37). Exendin-4 and liraglutide also restore impaired insulin signaling in the brains of a transgenic mouse model of AD, improving cognition, decreasing Aβ accumulation, and attenuating eIF2α-P (37,54,99). Importantly, liraglutide further protected the brain of nonhuman primates from AβO-induced eIF2α-P (54). Collectively, those studies provide a rational basis for use of insulin and/or GLP-1R agonists as therapeutics in AD.
Insulin and GLP-1R activation may thus provide novel strategies to resensitize impaired brain insulin signaling and to prevent or halt neurodegeneration in AD. Larger-scale clinical trials are now planned to determine the efficacy of insulin and GLP-1R agonists in improving memory and cognition in AD patients (www.adcs.org/studies/SNIFF.aspx; clinicaltrials.gov/ct2/show/NCT01843075), and results from such studies are highly expected in the field.
Possible Impact of Peripheral Inflammation and Metabolic Deregulation in AD
Aging, the single most important risk factor for AD, is often associated with a chronic state of low-grade inflammation resulting from deregulated levels of pro- and anti-inflammatory cytokines, a condition referred to as “inflamm-aging” (100). Increased levels of proinflammatory cytokines and markers, such as IL-1β, IL-6, TNF-α, and C-reactive protein, have been associated with “normal” aging. However, it is not yet clear whether increased levels of proinflammatory cytokines precede and play causal roles in the biochemical changes associated with aging, or whether they are a consequence of the aging process (101). Interestingly, it has been proposed that inflamm-aging is a result of lifetime exposure to acute and chronic infections (102,103), with human longevity at least partly related to the capacity to maintain inflammatory response at low levels (102).
Low-grade inflammation may persist throughout periods of life due to recurrent or persistent infections, and it is tempting to ask if chronically elevated peripheral inflammatory mediators could be associated with accelerated neuronal dysfunction and cognitive decline (11). Significantly, T2D induces changes in blood-brain barrier (BBB) permeability (104), and postmortem analysis of diabetic AD brains showed increased levels of IL-6 compared with nondiabetic AD brains (105). Moreover, the BBB of a transgenic mouse model of AD has been reported to be more permeable to peripheral inflammatory cytokines (106). These findings raise the possibility that AD brains could be more susceptible to peripheral inflammatory dyshomeostasis.
Peripheral (adipose tissue) inflammation is a major trait of diabetes and obesity (107), and both adipocytes and adipose-resident macrophages appear to participate in a cross-talk between the periphery and CNS. In obese patients, adipocytes react by producing proinflammatory cytokines, adipokines, and chemokines, while resident macrophages undergo a phenotypic change to a so-called M1 proinflammatory state (108). This leads to increased TNF-α, IL-1β, and IL-6 production, all of which can cross the BBB (109). Fat-derived inflammatory mediators could thus be an important addition to cytokines locally produced by CNS-resident microglia to generate a state of brain inflammation.
In addition to cytokines, altered levels of fat-derived leptin and adiponectin have been reported in AD. Both leptin and adiponectin receptors are ubiquitously distributed, including in the brain, consistent with important central actions of these hormones in addition to their roles in hypothalamic metabolic control (110,111). Initially described for its role in satiety and long-term body weight maintenance, leptin was recently proposed to regulate cognition, axonal growth, and synaptogenesis in extrahypothalamic regions (112). Moreover, leptin induces hippocampal neurogenesis (113), reduces Aβ generation in vitro (114), and improves cognitive performance (114) in transgenic mouse models of AD. Lower plasma levels of leptin have further been associated with increased risk of development of AD (115).
Plasma adiponectin levels are decreased in animal models of obesity and in obese patients (116,117). Surprisingly, recent studies found increased levels of adiponectin in mild cognitive impairment and AD patients (118). Although the possible action(s) of adiponectin in neuronal physiology are still unclear, it will be interesting to determine whether it plays any role in neuronal dysfunction in AD.
Dyslipidemia is another important trait of metabolic disorders. Cholesterol- and sphingolipid-enriched specialized cell membrane domains, called lipid rafts, appear to be preferential sites for Aβ generation via amyloidogenic cleavage of the amyloid precursor protein (119). Particular types of sphingolipids, namely ceramide and its metabolites, cause inflammation (120) and have been increasingly associated with T2D (121). Peripherally generated ceramides cross the BBB (122) and could contribute to AD pathogenesis in two ways: 1) by changing the microenvironment of lipid rafts, thereby favoring Aβ generation, and 2) by inducing central inflammation and disruption in neuronal insulin signaling (121).
Preclinical and clinical observations thus support the notion that peripheral mediators (cytokines, adipokines, and lipids) link peripheral and central metabolic/inflammatory dyshomeostasis in AD. Although effective therapeutic approaches to combat such deregulated signaling events are not yet available, current evidence supports encouragement of a healthier lifestyle and long-term metabolic control as a preventative measure to reduce the risk of AD (11,13).
Mitochondrial Dysfunction as a Link Between Peripheral/Brain Inflammation and Defective Insulin Signaling
Mitochondrial energy–transducing capacity is essential for maintenance of cellular function, and impaired mitochondrial energy metabolism/redox homeostasis is a hallmark of both brain and peripheral aging (123,124). Reactive oxygen species (ROS) are cytotoxic byproducts of normal mitochondrial metabolism generated by monoelectronic reduction of oxygen in the respiratory chain. In peripheral tissues, transient ROS generation in response to physiological stimuli such as insulin facilitates insulin signaling by inhibiting protein phosphatases, including PTEN (125). In the brain, transient production of ROS is implicated in synaptic signaling and facilitates long-term potentiation and memory-related mechanisms (126). However, chronically elevated ROS levels and/or an imbalance between ROS production and intracellular levels of antioxidant defenses lead to oxidative stress and mitochondrial dysfunction, a condition that has been associated with both T2D and AD (127,128).
Consumption of a high-fat diet causes increased mitochondrial ROS production and oxidative stress in skeletal muscle, leading to the development of peripheral insulin resistance in T2D (129). Similarly, elevated ROS levels in the brain can be detrimental. Excessive ROS levels are implicated in the molecular etiology of AD, with elevated markers of oxidative stress, including oxidized forms of lipids, proteins, and DNA present in an AD brain (128). In AD, aberrant stimulation of excitatory N-methyl-d-aspartate receptors (NMDA-Rs) by AβOs has been proposed as a key mechanism leading to excessive ROS production, likely as a consequence of Ca2+-related mitochondrial dysfunction (130). Brain insulin signaling and oxidative stress appear to be intimately connected, as AβO-induced neuronal oxidative stress is blocked by insulin (51,131). The mechanism of protection by insulin appears to involve activation of Akt (131) and prevention of abnormal NMDA-R activation (132). NMDA-R deregulation indeed seems to play a role in oxidative stress and defective neuronal insulin signaling in AD, as AβO-induced inhibition of IR signaling is prevented by the NMDA-R blocker memantine (52). It is possible that mitochondrial dysfunction and chronically elevated ROS levels sustain a vicious cycle that impairs insulin signaling in AD.
Interestingly, mitochondrial dysfunction and increased ROS generation were recently found to activate JNK and to lead to insulin resistance in skeletal muscle and liver (133). As discussed in previous sections, AD and metabolic disorders have been associated with JNK activation and with reduced levels of insulin/IGF-1 and their receptors (37,38,40,121). Therefore, it is likely that a complex signaling network closely connected to mitochondrial dysfunction leads to impaired insulin signaling in both AD and diabetes. In fact, it has been recently proposed that a coordinated metabolic triad comprising mitochondria, insulin, and JNK signaling plays a key role in neuronal dysfunction in brain aging and AD (134). Since both the IR and the JNK signaling pathways affect mitochondrial function, it is conceivable that a similar triad operates in metabolic disorders, impairing mitochondrial bioenergetics and biogenesis and leading to redox dyshomeostasis.
AD and T2D are chronic, debilitating, and extremely costly for health programs in developed and developing countries. The emergence of molecular links between inflammatory, deregulated insulin signaling and mitochondrial dysfunction in AD and diabetes raises the prospect for development of novel therapeutic strategies for AD based on antidiabetes and/or anti-inflammatory agents (11,61). Further studies aimed to unravel novel pathways and mechanisms implicated in brain inflammation and defective insulin signaling in AD as well as the effects of antidiabetes agents are highly anticipated in the field.
Acknowledgments. The authors thank Natalia Lyra e Silva (Federal University of Rio de Janeiro) for assistance in preparing the figures and Mychael Lourenco (Federal University of Rio de Janeiro) for assistance in editing the manuscript.
Funding. Work in the authors’ laboratory was supported by grants from the Human Frontiers Science Program (to F.G.D.F.), the National Institute for Translational Neuroscience (Brazil) (to S.T.F.), and the Brazilian funding agencies Conselho Nacional de Desenvolvimento Científico e Tecnológico and Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (to F.G.D.F. and S.T.F.).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. F.G.D.F. and S.T.F. wrote the manuscript. F.G.D.F. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. Parts of this study were presented at the 74th Scientific Sessions of the American Diabetes Association, San Francisco, CA, 13–17 June 2014.