2022 Volume 8 Pages 29-33
Alzheimer’s disease (AD) is a progressive neurological disease that leads to dementia and ultimately death due to cognitive decline caused by the loss of neurons in the brain. It has been reported that type 2 diabetes increases the risk of developing AD dementia, although the molecular mechanism linking type 2 diabetes and AD is unknown. The high incidence of AD in patients with type 2 diabetes may be explained by the effect of diabetes on tau pathology, one of the cardinal neuropathological features of the AD brain. Clinical studies using biofluid biomarkers report that tau in cerebrospinal fluid is increased by diabetic conditions; however, studies using post-mortem brain and positron emission tomography imaging provided controversial results regarding the effect of type 2 diabetes on tau pathology. In animal models, the effects of type 2 diabetes on tau have been extensively examined at the molecular level. The effect of type 2 diabetes increases phosphorylated tau in the brain via multiple intracellular signaling pathways. Therefore, tau phosphorylation may underlie the pathological interplay between diabetes and AD. A better understanding of the molecular mechanisms linking these two diseases may provide novel insights into diagnostic and therapeutic developments for AD.
It is estimated that the number of people with dementia worldwide will nearly triple to more than 152 million by 20501). To extend the healthy life spans of elderly people, the development of preventions and treatments for people with dementia is urgently needed.
About 70% of patients with dementia suffer from Alzheimer’s disease (AD) and about 20% from vascular dementia (VaD)2). AD, a progressive neurodegenerative disease, is characterized by two neuropathological features: senile plaques (SPs), which are extracellular accumulations of β-amyloid (Aβ), and neurofibrillary tangles (NFTs), which are intraneural accumulations of phosphorylated tau protein. AD ultimately leads to death due to cognitive decline caused by the loss of neurons in the brain3). More than 95% of AD cases are a sporadic form of AD, and a genetic risk factor for sporadic AD is the apolipoprotein E (ApoE) epsilon 4 allele, which is believed to promote the development of SPs. A variety of risk factors for dementia have been reported, including aging and lifestyle-related diseases4). Livingston et al . reported that about 35% of dementia cases could potentially be prevented by eliminating these modifiable risk factors5).
Diabetes mellitus, one of the modifiable risk factors for dementia, has a high prevalence among the elderly. It is estimated that in 2021, 537 million people over 20 years of age worldwide had diabetes mellitus6). More than 90% of patients with diabetes mellitus are classified as type 2 diabetes, which is mainly caused by insulin resistance or a relative deficiency. Many epidemiological studies have reported that type 2 diabetes increases the risk of developing dementia due to AD by about two-fold7,8). In the Rotterdam study, the relative risk of developing AD in elderly patients with type 2 diabetes was 1.9-fold, which is comparable to developing VaD with diabetes8). In addition, the relative risk of AD increases about four-fold in patients with diabetes treated with insulin8). The Hisayama study reported that the risk of AD was 2.1 times higher in those with diabetes compared to those without diabetes, and the incidence of AD was significantly associated with increased blood glucose levels in oral glucose tolerance tests7). Many studies support the idea that type 2 diabetes, especially with insulin resistance and impaired glucose tolerance, increases the risk of developing AD.
Evidence from post-mortem brain and functional imaging research
The high incidence of AD due to type 2 diabetes may be explained by the impact of diabetes on tau pathology, as the cerebral accumulation of pathological tau is known to correlate with the severity of cognitive dysfunction and neurodegeneration9). Some studies have reported that type 2 diabetes increased NFTs compared to controls under the presence of ApoE epsilon 4 risk10,11), suggesting the impact of diabetes on tau-related pathology. However, other post-mortem brain studies have reported the absence of a relationship between type 2 diabetes and the development of NFTs12,13) (Table 1). Therefore, studies based on post-mortem brain analyses have reported controversial findings regarding the effect of type 2 diabetes on tau pathology.
AD: Alzheimer’s disease; CSF: Cerebrospinal fluid; SPs: senile plaques; NFTs: neurofibrillary tangles; DM: diabetes mellitus; IR: insulin resistance; Aβ: β-amyloid
| Post-mortem brain | ||||
| Reference | Age | Study size | AD pathology | |
| （year, ref. no.） | mean±SD (DM/Non-DM) | number (DM/Non-DM) | SPs | NFTs |
| Dos Santos Matioli et al. (2017, 10) |
(74±11/75±12) | 1,037 (279/758) | No change | Increase (ApoE4 carrier） |
| Malek-Ahmadi et al. (2013, 11) |
(84±3/83±9) | 362 (40/322) | Increase (ApoE4 carrier） |
Increase (ApoE4 carrier） |
| Abner et al. （2016, 12） |
(88±6/89±7) | 2,365 (507/1,858) | No change | No change |
| Sonnen et al. （2009, 13） |
(86±7/84±7) | 259 (196/63) | No change | No change |
| CSF biomarker | ||||
| Reference | Age | Study size | CSF biomarker | |
| （year, ref. no.） | mean±SD (DM or IR/Non-DM) | number (DM or IR/Non-DM) | Aβ | Tau |
| Motta et al. (2021, 17) |
72±8 | 257 (71/186) | No change | tau↑ |
| Moran et al. (2015, 18) |
(76±6/74±7) | 415(56/359) | No change | tau↑・p-tau↑ |
| Starks et al. （2015, 19） |
61±6 | 113 （ApoE4 carrier: 43） |
No change | tau↑・p-tau↑ （ApoE4 carrier） |
Positron emission tomography (PET) has been used as a novel imaging biomarker for tau pathology. Tau PET imaging enables the visualization of tau deposits in the brains of living patients. It has been well validated that in vivo PET signals tightly correlate with postmortem histopathology in patients with tauopathy14), a class of neurodegenerative diseases characterized by the cerebral accumulation of tau. However, a study using tau PET imaging did not observe any significant impact of diabetes on the tau load in the brain15).
Studies using biofluid biomarkers
Cerebrospinal fluid (CSF) levels of tau predict the amount of NFTs in the brain that was confirmed by correlation analysis using postmortem brain16). Studies using CSF biomarkers reported that type 2 diabetes could have a significant impact on AD-related neuropathology (Table 1). CSF levels of tau were significantly higher in AD patients with type 2 diabetes compared to those without type 2 diabetes17). Type 2 diabetes was associated with elevated levels of phosphorylated tau in the CSF of patients with mild cognitive impairment (MCI)18). The severity of insulin resistance (i.e. HOMA-IR) was correlated with increased phosphorylated and non-phosphorylated tau levels in CSF in ApoE4 carriers, while the effect was not observed in ApoE4-negative patients19). This means that the impact of type 2 diabetes on CSF tau levels could be affected by both AD- and diabetes-related factors.
The discrepancy between the results from post-mortem brain and CSF biomarker studies regarding the effect of type 2 diabetes may be explained by the difference in the nature of the tau examined in each study. Post-mortem brain studies primarily measure highly aggregated intracellular tau, whereas CSF biomarker studies measure soluble extracellular tau. CSF biomarkers can potentially reflect dynamic pathological changes in the AD brain, which might allow for the sensitive detection of the effect of type 2 diabetes on tau.
Tau contains up to 85 potential phosphorylation sites, 44 of which are pathologically phosphorylated in AD brains20). The hyperphosphorylation of tau causes neuronal dysfunction and the aggregation of tau21), leading to the formation of NFTs and neurodegeneration in AD brains22). Since the pathogenesis of type 2 diabetes involves insulin resistance and deficiency, many analyses have focused on the phosphorylation of tau via insulin signaling.
The administration of streptozotocin (STZ) is used as the mouse model of insulin deficiency. Clodfelder-Miller et al. reported that STZ treatment increased the amount of tau phosphorylation at multiple sites in the brains of C57BL/6 mice23). This result was also reproduced in studies using non-tg24–26) and tau-transgenic (Tau-tg) mice27). These results imply that peripheral insulin deficiency increases the phosphorylation of tau in the brain.
In studies using mice fed with a high-fat diet, many studies have reported higher levels of phosphorylated tau in the brain28–30), while a few studies have reported no difference in the levels31). However, these discrepancies may be due to the experimental conditions (e.g. the duration of the loading, composition of the diet, and amount of food intake). The amount of phosphorylated tau is increased in the brains of genetically diabetic (db/db) mice compared to wild-type mice32). The levels of phosphorylated tau in the brains of AD mice crossbred with genetically diabetic mice are increased compared to non-diabetic AD mice33). These two results indicate that the phosphorylation of tau is increased in both type 1 diabetes (i.e. the STZ-administration model) and type 2 diabetes mice brains (i.e. diet-induced obesity and genetically diabetic models).
In addition, some studies have analyzed the aggregation of tau in the brain using diabetic Tau-tg mouse models. In studies using Tau-tg mice fed with a high-fat diet, the amount of aggregated tau was reportedly increased28) and maintained29). In summary, type 2 diabetes increases the phosphorylation of tau in the brains of animal models, the same as in the human brain.
Both observational studies in humans and experimental studies using animal models suggest that type 2 diabetes increases the phosphorylation of tau. It has been reported that many kinases and phosphatases regulate the phosphorylation of tau20) and type 2 diabetes could affect the activity of these kinases34). Alteration of kinase activity due to diabetic conditions may be associated with the phosphorylation of tau. Insulin signaling is associated with kinase activity and may link type 2 diabetes with the phosphorylation of tau.
Insulin is thought to increase the phosphorylation of tau by altering the activity of its downstream kinase35) via Akt, a kinase widely expressed in the brain and involved in various cellular activities, such as cell development, differentiation, and survival36). The activated form of Akt (phosphorylated at Ser473) increases the phosphorylation of GSK3β at Ser9 and inhibits the enzymatic activity of GSK3β. The activation of GSK3β increases the phosphorylation of tau when insulin signaling in the brain is impaired. Many reports have focused on insulin signaling-related factors, such as insulin receptor, Akt, and GSK3β, as the mechanism for the increased phosphorylation of tau in the brain23,25,28,29,31). These studies reported that highly phosphorylated levels of Akt at Ser47328,29) and GSK3β at Ser923,28,29) occurred in association with high phosphorylated tau levels.
The phosphorylation of tau may be induced by mechanisms other than insulin signaling25). Other reports have suggested that the neuroprotective intracellular signaling pathway, BDNF/TrkB may be involved in the increased phosphorylation of tau induced by a high-fat diet30). Despite the absence of insulin resistance, Tau-tg (THY-Tau22) mice fed with a high-fat diet showed increasing amounts of phosphorylated tau, Akt (at Ser473), and GSK3β (at Ser9) in the brain29). Therefore, multiple intracellular signaling pathways may be involved in the mechanism of phosphorylation of tau induced by type 2 diabetes.
Previous reports on the phosphorylation of tau have focused on a specific phosphorylation site of tau (Figure 1) and have not surveyed multiple sites or the interaction between distinct phosphor-sites. Many studies have examined the amount of phosphorylated tau using AT8 (phosphorylated Ser202/205) tau antibody and other antibodies, such as phosphorylated tau at the mid-domain or C-terminal side. Thus, phosphorylation sites without pre-existing antibodies, such as the N-terminal side, are rarely investigated (Figure 1). Type 2 diabetes may induce alterations in multiple intracellular signaling pathways, so the diabetes-induced phosphorylation of tau may also occur at sites that have not been previously examined. Therefore, it is necessary to comprehensively analyze alterations in tau phosphorylation and intracellular signaling pathways that are induced by type 2 diabetes. This will help clarify the relationship between type 2 diabetes and tau phosphorylation, which may lead to the discovery of new targets for the development of the prevention and treatment of AD.
Type 2 diabetes increases the risk of developing dementia, but the detailed mechanisms of this remain unclear. Type 2 diabetes increases tau levels in CSF, although the results of post-mortem brain research appear to be controversial. In animal models, type 2 diabetes increases phosphorylated tau in the brain via multiple intracellular signaling pathways. Therefore, it is crucial to comprehensively analyze the alterations in tau phosphorylation at multiple sites and intracellular signaling pathways related to type 2 diabetes (Figure 2). A better understanding of the mechanisms underlying the relationship between type 2 diabetes and tau phosphorylation could lead to the development of new therapeutic targets for AD.
This work was supported by JSPS KAKENHI Grant Number 21H02828 (grant-in-Aid for Scientific Research (B)) (S.T.) and the research grant from Cell Science Research Foundation (S.T.).
Disclosures: The authors have no potential conflicts of interest to declare.
