Linking type 2 diabetes and Alzheimer's disease

Since the initial Rotterdam study suggesting an increased risk to develop dementia and Alzheimer's disease (AD) in patients with type 2 diabetes mellitus (T2D), a number of clinical and epidemiological studies have provided further direct evidence to strengthen the link between T2D and AD (1, 2). The possibility that T2D patients might be at increased risk in developing AD has serious societal implications. It is estimated that 5.3 million people in the United States and more than 30 million worldwide are living with AD, and the number is destined to increase dramatically over the next two decades as the population ages. The direct and indirect cost associated with AD in the United States alone has reached 172 billion US dollars annually (3). The situation for T2D is hardly reassuring: There are more than 23 million T2D patients in the United States alone, and ≈18 million of those might have as much as 50% higher risk than a nondiabetic to develop AD. T2D often leads to cardiovascular complications and other conditions that kill the patients before AD typically strikes, usually when the patients are in their 70s. However, as the therapeutic treatment advances for diabetes in the near future, T2D patients will most likely live longer, and the world may soon be facing the daunting challenge of dealing with a new population of AD sufferers with T2D and, possibly, other metabolic complications. Over the last decade, researchers have been busy investigating the biological basis underlying the link between T2D and AD. In PNAS, the study by Takeda et al. (4) provides further understanding of potential underlying mechanisms that link AD and T2D by establishing two new mouse models and studying these mice by using a wide range of well-defined techniques.

The pathogenesis of AD begins with impaired synaptic function, which may result from the accumulation of amyloid-β (Aβ) peptide (5–8). For many years, researchers have focused on the insoluble deposits of amyloid fibrils as the leading cause of memory loss and as the culprit of AD. More recent findings, however, suggest soluble Aβ oligomers may be the cause of memory loss, especially in the early stages of AD, because Aβ oligomers inhibit long-term potentiation in neurons, a well-adopted experimental paradigm for learning and memory (6, 9). Aβ is produced from amyloid precursor protein (APP) by serial proteolytic reactions catalyzed by β-site of APP cleaving enzyme (BACE) and a multiprotein complex γ-secretase, in which presenilin is likely the catalytic component (10).

T2D is caused by relative insulin deficiency, which may be the result of insufficient insulin supply due to defective insulin secretion and/or reduced insulin-secreting β-cell mass, and/or impaired insulin sensitivity in peripheral metabolic organs, such as liver and muscle. Obesity is believed to be caused by impaired central response to adipose tissue-derived hormone leptin, whose major function is to reduce food intake and to promote energy expenditure (11). Insulin and leptin exert their effects through complex signaling cascades and share some common components in the two pathways. Because obesity often fuels the development of T2D, T2D and obesity are sometimes collectively referred to as diabesity.

In addition to their peripheral functions in the regulation and maintenance of physiological homeostasis, insulin and leptin also have profound effects on brain functions. Indeed, both have been shown to regulate neuronal and synaptic functions within hippocampus, cortex, and cerebellum, to protect neurons against neurodegeneration and cell death, and to affect cognition and behavior (12–15). Furthermore, insulin and leptin also regulate Aβ levels by modulating Aβ production via their actions on BACE and/or γ-secretase, as well as by modulating Aβ degradation via Aβ-degrading enzymes, such as insulin-degrading enzyme (16–19). It is conceivable that insulin actions or their lack may be a link between T2D and AD. Some have gone so far as to refer to AD as type 3 diabetes to emphasize the potential endocrine links between these diseases (20).

Takeda et al. (4) explore insulin action and insulin resistance in the development of cognitive impairment by crossing the ob/ob or NSY (Nagoya-Shibata-Yasuda) mouse, two well-established T2D mouse models, into the APP23 transgenic mouse background, a well-studied AD mouse model. The ob/ob mouse is leptin deficient, whereas the NSY mouse is an inbred strain with spontaneous diabetes that resembles human T2D. In contrast to the ob/ob mice, the NSY mice do not develop severe obesity, allowing the study of T2D in the absence of obesity (21). The APP23 mouse line, generated by overexpressing APP₇₅₁ harboring the Swedish double mutation (K670N and M671L) under the murine Thy1 promoter, has an ≈7-fold overexpression of the mutant APP above endogenous APP (22). These mice show Aβ deposits at 6 months of age, and the deposits increase in size and number with age and eventually occupy a substantial area of the neocortex and hippocampus in 24-month-old mice.

Takeda et al. (4) performed metabolic and behavioral tests on the two new mouse lines, APP⁺-ob/ob and APP⁺-NSY. They find that in both lines, diabetes exacerbated cognitive dysfunction without a significant increase in Aβ levels. Furthermore, they find that brain insulin levels were reduced, and Akt phosphorylation, a key step in insulin signaling, was severely impaired. These findings provide experimental evidence to support the notion that impairment of insulin signaling might be a mechanistic link underlying T2D and AD: Peripheral impairment causes T2D and central impairment leads to AD (Fig. 1A).

Fig. 1.

The underlying links between AD and T2D. (A) Insulin resistance as a mechanistic link between AD and T2D. Insulin resistance in peripheral tissues and organs, when coupled with relative insulin deficiency, causes T2D. On the other hand, central insulin resistance, together with reduced brain insulin levels that might have resulted from T2D, leads to accumulation of β-amyloid and, consequently, AD. (B) Inflammation as a mechanistic link between AD and T2D. Through its influence on islet function and peripheral insulin sensitivity, inflammation accelerates the development of T2D. Cerebrovascular and central inflammation, along with increased accumulation of β-amyloid, disrupts normal synaptic function, a starting point of AD pathological progression.

In addition to impaired insulin signaling, Takeda et al. (4) find that diabetes accelerated the appearance of cerebrovascular inflammation and Aβ deposition, as evidenced by increased levels of proinflammatory cytokines IL-6 and TNF-α, as well as dense amyloid deposits in blood vessels. In recent years, inflammatory pathways, including that of NFκB, have been linked to metabolic syndrome and neurodegenerative diseases, including AD. As inflammation is often not restricted to central or peripheral tissues or organs, it is tempting to hypothesize that inflammation may be another mechanistic link underlying T2D and AD (Fig. 1B). Interestingly, in these new mouse lines, the relationship did not end with just T2D's effects on AD; Takeda et al. (4) were able to show that AD could also accelerate the development of diabetic phenotype.

AD and T2D may share common cellular and molecular mechanisms.

The reciprocal actions between AD and T2D thus form a vicious cycle, further illustrating the possibility that AD and T2D may share common cellular and molecular mechanisms.

Although the study by Takeda et al. (4) provides significant insights and experimental evidence on the mechanistic link between AD and T2D, there remain some important questions to be addressed. First, although APP⁺-ob/ob mice demonstrated more severe cognitive deficit than APP23 mice, both APP23 and APP⁺-ob/ob mice also exhibited significant and comparable reductions in insulin levels, thus insulin and insulin action alone could not fully account for the observed difference in cognition. Considering the importance of leptin signaling on brain functions, it is conceivable that leptin may also be involved. Future studies investigating defective leptin signaling in the absence of perturbed insulin signaling will shed light on the significance of leptin signaling in AD development and on the link between obesity and AD. Second, APP transgene appeared to have weight-reducing effects because APP⁺-ob/ob and APP⁺-NSY mice were leaner than ob/ob or NSY mice, respectively. Because there was no difference in food intake between each pair of mouse lines, APP transgene may regulate body weight by increasing energy expenditure and/or by reducing food absorption. Further experimental data will be needed to determine the mechanisms involved. Finally, does the T2D-AD connection observed in mice translate to humans? If so, aggressive therapy to control glucose levels in AD patients could prove to be beneficial in maintaining cognitive function, despite studies where rosiglitazone appeared to be without effect in several phase III AD trials. Alternatively, pharmacological or lifestyle management of AD patients could also lessen their diabetes burden that, in turn, may slow down AD progression. Epidemiological data might provide answers to these tantalizing possibilities.