The Role of Pyroptosis in Alzheimer's Disease

Pyroptosis is a type of regulated cell death that relies on caspases, vesicles, and the cleavage of gasdermin proteins (which create pores in the cell membrane). The nucleotide-binding oligomerization domain-like receptor protein 3 (NLRP3) inflammasome, which is involved in this process, is the most widely studied inflammasome. Caspase-1 activates pro-inflammatory cytokines, such as IL-1 $\beta{}$ and IL-18. Gasdermin D (GSDMD) is the most important executive protein. GSDMD, a substrate rather than an upstream protease, determines the occurrence of pyroptosis. Pyroptosis is essential for maintaining body homeostasis, but excessive or poorly regulated cell death can aggravate the inflammatory response. Undoubtedly, this will be an important direction for future research on Alzheimer’s disease (AD). Here, we review recent research progress on the morphological characteristics, molecular mechanisms, and role of pyroptosis in the context of AD, thereby providing new directions for identifying potential disease biomarkers and treatment strategies for AD.

Keywords

Alzheimer's disease

gasdermin D

inflammasome

pyroptosis

therapy

1. Introduction

Alzheimer’s disease (AD) is a disorder that causes people to gradually lose their memory and cognition. The pathophysiological process of AD is thought to begin years, or even decades, before symptoms arise [1]. Despite decades of research, our knowledge of AD pathogenesis remains unclear, and our ability to intervene via the prevention or treatment of dementia is still limited. There is still much we don’t know about the molecular mechanisms that lead to AD, and this is important because it would help us develop more sensitive diagnostic markers and find new ways to treat the disease.

Pyroptosis is an inflammatory form of regulated cell death and a critical and necessary host-defense immune mechanism. Increasing evidence shows that dysfunctional pyroptosis participates in neurological disorders such as Parkinson’s disease [2], amyotrophic lateral sclerosis [3], and Huntington’s disease [4]. There are emerging evidences that the inflammatory responses in the central nervous system (CNS) may be a major cause and common feature of AD [5, 6]. Pyroptosis is also involved in $\beta{}$ -amyloid (A $\beta{}$ ) protein deposition and the hyperphosphorylation of tau [5, 7]. Thus, pyroptosis is important to the development of neuropathological lesions in AD. In our review, we summarize the process of pyroptosis, particularly in AD, to better understand the pathogenesis of AD and provide a novel strategy for more effective prevention, diagnosis, and treatment measures for AD.

2. Overview of Pyroptosis

2.1 Morphological Characteristics of Pyroptosis

There are two main ways for cell death to occur: ordered (programmed-like, regulated) and nonordered (necrosis). To date, over 20 forms of regulated cell death have been identified and studied, including apoptosis, autophagy or autophagic death, pyroptosis, and ferroptosis, but they are not all equally well characterized. Studies have shown that substrates, rather than their upstream proteases, determine the nature of cell death. Because gasdermin family members are indispensable executors of pyroptosis, pyroptosis is also known as gasdermin-mediated regulated cell death [8, 9].

Pyroptotic cells display DNA fragmentation, which can be detected by terminal deoxynucleotidyl transferase dUTP nick-end labeling, but at a lower intensity than apoptotic cells. Chromatin condensation also occurs in pyroptosis, but the nucleus remains intact. Furthermore, pyroptotic cells become annexin-V-positive because, in early membrane rupture, the inner leaflet of membrane is exposed to the outside [10, 11]. During pyroptosis, the cell membrane breaks down to create small holes with a diameter of 11–24 nm; this causes increased cell permeability and the release of inflammatory cytokines, lactate dehydrogenase, and other intracellular substances. The small pores in the cell membrane cause the cell to lose its salt and water balance, and this causes it to swell. Finally, the cell membrane is destroyed, and the cellular contents are released into the extracellular environment. This causes the body’s immune system to become active, drawing in more inflammatory cells, and inducing a serious inflammatory reaction [12, 13]. Unlike necrosis, pyroptotic cell death and the consequent inflammatory responses are reversible and controllable. Thus, pyroptosis has attracted increasing attention in the study of infectious diseases, various neoplastic diseases and metabolic diseases, and represents a new research direction.

2.2 Gasdermins in Pyroptosis

Members of the gasdermin family have been recently identified as having pore-forming activity and are found in many different cells and tissues. Currently, this family comprises six homologous genes in humans: gasdermin A (GSDMA); GSDMB; GSDMC; GSDMD; GSDME; and pejvakin. All gasdermins except for pejvakin contain a cytotoxic N-terminal (GSDM ${}^{NT}$ ) domain and a C-terminal (GSDM ${}^{CT}$ ) repressor domain. The GSDM ${}^{NT}$ fragment has the ability to form pores in the cell membrane, which can disrupt its integrity. The expression of GSDM ${}^{NT}$ alone can induce pyroptosis [8, 14, 15]. The GSDM ${}^{CT}$ fragment can bind to the GSDM ${}^{NT}$ domain and act as a repressor, whereas overexpression of GSDM ${}^{CT}$ can block cell death [8].

Pore formation by the gasdermin family is a characteristic of pyroptosis [16, 17], the binding of the GSDM ${}^{NT}$ domain to membrane lipids causes it to change shape, which leads to the formation of pores. Experimental evidence indicates that the GSDM ${}^{NT}$ domain can directly interact with lipid molecules in cells. The GSDMD ${}^{NT}$ domain preferentially targets acidic phospholipids such as phosphoinositides and cardiolipin [18, 19, 20, 21]. The N-terminal domains of other gasdermins, such as GSDME and GSDMA, have a similar way of forming small holes in the cell membrane [18].

At present, the mechanism by which GSDMD induces pore formation in the membrane (which constitutes cell pyroptosis) is relatively clear, but it is still a matter of debate how caspases recognize and cleave GSDMD. In studies on apoptosis, caspases have been shown to activate the substrate protein by recognizing a tetrapeptide sequence in the substrate and cleaving after the aspartate residue [22, 23]. However, this process is different in pyroptosis. Recently, it was found that autocleavage at the Asp289/Asp285 location in caspase-4/11 creates p10. To carry out a tetrapeptide sequence-independent cleavage, the p10-form promotes the binding of caspase-4/11 to the GSDMD ${}^{CT}$ domain; caspase-1 employs a similar structural mechanism for targeting GSDMD [24].

2.3 Molecular Mechanisms of Pyroptosis

2.3.1 Regulatory Mechanism of Canonical Pyroptosis

The canonical pyroptosis pathway is mediated by caspase-1. Proteins called pattern recognition receptors (PRRs) assist the body in recognizing two groups of substances: damage-associated molecular patterns (DAMPs), which are linked to cell components, and pathogen-associated molecular patterns (PAMPs), which are linked to microbial infections. Inflammatory signaling cascades are thereby initiated. Several families of PRRs, including NOD-like receptors (NLRs), toll-like receptors (TLRs), and RIG-I-like receptors, play vital roles in the immune system [25]. The inflammasome is activated by various stimuli, which induces the activation of caspase-1. On the one hand, GSDMD which is cleaved and activated by activate caspase-1 causes cell membrane to form pores, and lead to pyroptosis; on the other hand, caspase-1 can rapidly process pro-interleukin-1 $\beta{}$ (pro-IL-1 $\beta{}$ ) and pro-IL-18 into their active proinflammatory cytokines, while IL-1 and IL-18 flow out of the pores and trigger inflammation [26] (Fig. 1).

Fig. 1.

Molecular mechanisms of pyroptosis. The canonical pyroptosis pathway depends on the inflammasome and GSDMD by caspase-1. Activated inflammasome promotes the activation of caspase-1, which cleaves the pore-forming factor GSDMD. Active caspase-1 also cleaves the proinflammatory cytokines such as IL-1 $\beta{}$ and IL-18. NLRP1 initiates inflammasome activation upon anthrax toxin. NLRP3 needs to be primed prior to activation. The activators of NLRC4 are flagellin and components of T3SS, the adaptor protein ASC is not necessary for the assembly of NLRC4 inflammasome. The activators of NLRP6 are microbial metabolites and lipoteichoic acid. Rotavirus infection lead to the activation of NLRP9b. AIM2 recognizes dsDNA in the cytosol. The non-canonical pyroptosis pathway requires directly binding of LPS to caspase-11 (murine) or caspase-4/-5 (human) and release of GSDMD N-terminus. Caspase-8 mediates the cleavage of GSDMC and GSDMD. Chemotherapy drugs and TNF- $\alpha{}$ promotes the activation of caspase-3, which cleaves the GSDME. Gzm B in NK cells and cytotoxic T lymphocytes can also directly cleave GSDME. Gzm A from cytotoxic lymphocytes could cleaves GSDMB. Human pathogen group A Streptococcus secretes SpeB, which induces GSDMA-dependent pyroptosis. DAMP, damage-associated molecular pattern; PAMP, pathogenassociated molecular patterns; ASC, apoptosis associated speck-like protein containing a caspase recruitment domain; LPS, lipopolysaccharide; GSDM, Gasdermin.

A collection of multiprotein signaling complexes called inflammasomes are found in the cytoplasm. Several inflammasomes, including NLRP1, NLRP2, NLRP3, AIM2, and NLRC4, have been discovered [27]. The immune receptor protein NLRP3, adaptor protein ASC (apoptosis associated speck-like protein containing a caspase recruitment domain), and inflammatory protease caspase-1 make up the most extensively studied inflammasome, which reacts to microbial infections, endogenous danger signals, metabolic risk factors, and environmental stimuli [28, 29]. An earlier investigation revealed that the gene expression of NLRP3 was so low that it was insufficient to activate the assembly of an inflammasome during resting conditions. Therefore, the canonical NLRP3 inflammasome is activated when two steps are completed [30, 31, 32]: the first step is the priming step, through the myd88-NF- $\kappa{}$ B pathway, the ligands of TLRs, NLRs, and cytokine receptors cause the production of pro-IL-1 and NLRP3 [33, 34, 35]. The second step is the activation step, which happens when a variety of foreign objects (PAMPs or DAMPs) cause NLRP3 inflammasome to assemble and activate [36].

2.3.2 Regulatory Mechanism of Noncanonical Pyroptosis

The noncanonical pyroptosis pathway is mediated by caspase-4/5/11. Bacterial lipopolysaccharide stimulates and activates caspase family proteases (caspase-4/5 in humans and caspase-11 in mice), and activate caspase-4/5/11 cleaves GSDMD, which forms pores in the cell membrane and eventually leads to pyroptosis [14, 37]. Unlike the canonical signaling pathway, IL-1 $\alpha{}$ and high mobility group box 1, not IL-1 $\beta{}$ or IL-18, are released by caspase-4/5/11-mediated pyroptosis. Caspase-4/5/11 activates the transmembrane channel pannexin-1, causing an efflux of cellular ATP, which promotes P2X7 receptor-dependent pyroptosis [38]; the activation of pannexin-1 channels causes potassium efflux, which is essential for the activation of NLRP3 inflammasome. Therefore, noncanonical pyroptosis can also induce an inflammatory cascade.

2.3.3 Other Caspase-Induced Pyroptosis

In addition to the two different pyroptosis pathways described above, recent research has revealed some new mechanisms. GSDME is specifically cleaved by caspase-3, and thereby induces the switching of caspase-3-mediated apoptosis to pyroptosis [39]. Caspases-3/7 which are thought to play a role in apoptosis, could induce GSDMD-associated microglial pyroptosis under neuroinflammatory conditions [40]. Caspase-8 is a caspase that is involved in apoptosis and necroptosis, however, recent evidence suggests that caspase-8 plays a critical role in pyroptosis [41]. The RIPK1- and caspase-8-dependent cleavage of GSDMD results in cell death; no other caspase is involved in the whole process [42]. Additionally, the lysosomal Ragulator-Rag complex initiates caspase-8-mediated pyroptosis by Yersinia [43].

3. Pyroptosis in CNS cell types

Pyroptosis has been assessed in the CNS using a variety of assays; here, we review the evidence for pyroptosis in each CNS cell type (Fig. 2; Table 1, Ref. [5, 7, 44, 45, 46, 47]).

Fig. 2.

Neuron-glia crosstalk in Alzheimer’s disease. Microglia, working as both phagocytic cells and innate immunocyte, can recognize and clear A $\beta{}$ . However, in AD, the canonical pyroptosis pathway is activated in A $\beta{}$ -induced reactive microglia. Meanwhile, A $\beta{}$ species can also induce pyroptosis in neurons. Activate NLRP3, caspase-1 and downstream IL-1 $\beta{}$ , IL-18 promote the aggregation of hyperphosphorylated tau, neurofibrillary tangles in turn trigger the activation of NLRP3 inflammasome. The interaction between NLRP3 and hyperphosphorylated tau promotes neuroinflammation and neuronal injury. Activated microglia induces A1 astrocytes by secreting IL-1 $\alpha{}$ , TNF- $\alpha{}$ and C1q, and A $\beta{}$ species induce the activation of NLRP3 infammasome and caspase-1 in astrocytes, which futher triggering the release of IL-1 $\beta{}$ and IL-18.

Table 1.Studies on pyroptosis in CNS cell types in AD.

Experimental model	Cellular location	Comments	Reference
AD mice	microglia	Activation of NLRP3-caspase-1-GSDMD axis by ASC-A $\beta$ composites	[44]
AD mice	astrocytes	Activation of NLRP3 and caspase-1 by A $\beta$ ${}_{1-42}$ , but GSDMD was not detected	[45]
AD rats and PC12 cells	neurons	Activation of NLRP1-caspase-1-GSDMD axis by hyperphosphorylated tau	[7]
AD mice	neurons	Activation of NLRP3-caspase-1-GSDMD axis by A $\beta$ ${}_{1-42}$	[5]
AD mice	neurons	Activation of NLRP1-caspase-1-GSDMD axis by A $\beta$	[46]
AD mice	oligodendrocytes	Activation of NLRP3-caspase-1-GSDMD axis by A $\beta$ ${}_{1-42}$	[47]

3.1 Microglia

Microglia is a type of resident macrophage present in the CNS. But the mechanism by which it works in the resting state is still poorly understood. Activated microglia can not only promote the repair of tissue damage, but also promote the inflammatory response of the CNS. However, the effect of inflammatory microglia on brain injury and damage is closely related to the pathogenesis of AD [6, 48]. The NLRP1, NLRP3, and AIM2 inflammasomes have been reported to be activated in microglia, astrocytes, and neurons [49, 50, 51], and the NLRP3 inflammasome is highly activated in microglia [52]. Activation of the NLRP3 inflammasome triggers the release of several proinflammatory cytokines, including IL-1 $\beta{}$ and IL18. Importantly, studies on microglia have shown that gasdermins are recognized and cleaved by caspases, leading to cellular swelling, membrane rupture, and other features of pyroptosis [44, 53, 54, 55, 56, 57, 58].

3.2 Astrocytes

Astrocytes are the most widely distributed in the brain and play an important role in normal central activities. Studies have shown that astrocytes play an important role in the inflammatory response of the CNS. Together with other glial cells (e.g., microglia, oligodendrocytes), astrocytes compose and maintain a regulated microenvironment [59, 60]. Their activation states vary, ranging from neuroprotective (decreases inflammatory response, promotes repair) to neurotoxic (intensifies inflammatory response, causing neurodegeneration) [61]. A $\beta{}$ and extracellular ATP, which are capable of activating LPS-induced astrocytes and interacting with the NLRP3 inflammasome, create a neuroinflammatory environment by the excessive production and release of proinflammatory cytokines and promote pyroptosis in astrocytes in vitro and in vivo [45, 62, 63, 64]. However, inconsistent results have been reported in studies with human astrocytes [65]; astrocytes in the brain exhibited NLRP3, cleaved GSDMD and strong caspase-8 immunoreactivity, but not ASC, caspase-1, or IL-18 [66].

3.3 Neurons

Pyroptosis is of great significance for the pathogenesis of AD. However, most studies have focused on glial cells, and there have been few experiments on the interaction between neurons and pyroptosis. The latest research suggests that pyroptosis is not restricted to glial cells; neurons are immunoreactive to cleaved GSDMD [66]. Studies have showed that the inflammasomes of NLRP3 and NLRP1 may induce neuroinflammatory processes by pyroptosis in neurons. A $\beta{}$ ${}_{1-42}$ can induce pyroptosis via the GSDMD protein in neurons, and NLRP3/caspase-1 signaling is important for mediating GSDMD cleavage [5, 67]. Tan and colleagues [46] have showed that the NLRP1 inflammasome drives neuronal pyroptosis in AD mice, suggesting that NLRP1/caspase-1 signaling is a key pathways responsible for A $\beta{}$ neurotoxicity. In addition, one study demonstrated that hyperphosphorylated tau could induce pyroptosis in PC12 cells, the release of IL-1 $\beta{}$ and IL-18 in turn increased hyperphosphorylated tau while spreading neuroinflammation [7].

3.4 Oligodendrocytes

Oligodendrocytes (OLs) are important for the functioning of the CNS. Although there is increasing evidence that OL damage and white matter degeneration are important pathological changes in AD, their roles in the occurrence and development of AD are still unclear [68]. Zhang et al. [47] reported that mature OLs in both AD patients and AD mice undergo NLRP3-dependent GSDMD-associated inflammatory injury, accompanied by demyelination and neurodegeneration. In mature OLs, overactivation of Drp1 leads to impaired glucose metabolism, leading to NLRP3-related inflammation and pyroptosis.

4. Association between Pyroptosis and AD

AD is a neurodegenerative disorder clinically defined by gradually increasing cognitive impairment and alterations in executive functions. The neuropathological hallmarks of AD are the accumulation of A $\beta{}$ and neurofibrillary tau tangles (NFTs) [69]. Studies have confirmed that the inflammatory response mediated by inflammasomes play a key role in AD pathology. Rui et al. [70] have shown that GSDMD-mediated inflammasomes and pyroptosis were activated in peripheral blood mononuclear cells of patients with amnesiac mild cognitive impairment and AD. However, the interaction between pyroptosis and the pathophysiology of AD is not clear yet. Here, we summarize the relationship between pyroptosis and AD, focusing on A $\beta{}$ plaques and NFTs.

4.1 A

\beta{}

Activation of the NLRP3 inflammasome by fibrillar A $\beta{}$ and soluble A $\beta{}$ has been described previously [71, 72]. Heneka et al. [28, 73] demonstrated that AD mice exhibit obvious inflammatory phenotypes in the cerebral cortex and hippocampus. It was characterized by the activation of microglia associated with A $\beta{}$ plaques, accompanied by extensive vascular endothelial damage. NLRP3–/– or caspase-1–/– mice, mainly due to reduced activation of caspase-1 and IL-1 $\beta{}$ in the brain, increased A $\beta{}$ content, thereby attenuating the loss of spatial memory and other AD symptoms [28, 73]. In addition, in 5 xFAD mice aged 7–8 months, whose brains contained the ASC+/– genotype, the amyloid content in the brain was significantly reduced; inhibition of inflammasome activation enhanced phagocytosis capability of astrocytes and improves learning and memory [74]. Han et al. [5] found that A $\beta{}$ ${}_{1-42}$ can cause pyroptosis through GSDMD protein, and the NLRP3-Caspase-1 signaling pathway is the key to mediate GSDMD cleavage. The role of the NLRP3 inflammasome in AD has also been confirmed in clinical research [75]. A $\beta{}$ triggers the activation of inflammasomes and mediates pyroptosis in the brain; conversely, pyroptosis also accelerates the formation of neuritic plaques and is crucial for the development of AD.

4.2 Tau

Pyroptosis has been reported to act as a component in the progression of AD by its interaction with A $\beta{}$ , but does pyroptosis affect tau pathology? Data on the correlation between tau hyperphosphorylation and pyroptosis are scarce. Previous studies have showed that the overexpression of proinflammatory cytokines increases neurofibrillary tangles [76], but recent studies have found that this phenomenon is caused by activation of the NLRP3 inflammasome and pyroptosis. Inhibition of the NLRP3 inflammasome reduced neurofibrillary tangles and significantly improved memory and cognition in AD mouse models [77]. Li et al. [7] used two hyperphosphorylated tau rat models and PC12 cells to study the correlation between tau protein and pyroptosis. The authors found that the high level of hyperphosphorylated tau induce the release of caspase-1, IL-1 $\beta{}$ and IL-18, and the degree of cell injury [7]. However, in AD brain tissues, GSDMD-positive neurons had no NFTs, but were found in close proximity to A $\beta{}$ plaques [66].

5. Anti-Pyroptotic Therapies

With continuing research on pyroptosis, increasing evidence shows that pyroptosis can be used as a new therapeutic target for the treatment of AD (e.g., using NLRP3 inflammasome inhibitors such as MCC950 and JC-124 [78, 79] and proinflammatory caspase inhibitors such as VX765) [80, 81]. However, none of them have been applied in the clinic. At present, there is no relevant report on GSDMD inhibitors as an AD treatment strategy. However, because GSDMD is the main executor of cell pyroptosis, it is reasonable to speculate that inhibitors of GSDMD might have a place in the treatment of AD.

In recent years, Traditional Chinese Medicine (TCM) has been widely used to treat AD, and research on the anti-pyroptotic effects of the active ingredients of TCM preparations has gained increasing attention. Because it inhibits NF- $\kappa{}$ B activity and NALP3 inflammasome activation, artemisinin has protective effects on the pathology of AD [82]. Dl-3-n-butylphthalide, also known as apigenin, suppresses the TXNIP-NLRP3 interaction, inhibits NLRP3 inflammasome activation, reduces proinflammatory cytokine levels, and prevents A $\beta{}$ production [83]. Resveratrol has protective effects on AD pathology through suppressing the inflammatory response [84]. Treatment with Ginkgo biloba extract EGb 761 has been found to decrease microglial secretion of TNF- $\alpha{}$ and IL-1 $\beta{}$ , NLRP3 and caspase-1, and inhibit inflammatory activation, thereby significantly improving cognitive function in mice [85]. Recent findings suggest that scutellarin inhibits neuroinflammation and microglial activation via regulation of the reactive oxygen species/NLRP3 signaling pathway [86]. Scutellarin may be regarded as a caspase-11 inhibitor that inhibits the generation of GSDMD ${}^{NT}$ , leading to reduced pyroptosis [87]. Ginsenoside, triptolide, epigallocatechin-3-gallate, curcumin, andrographolide, gastrodin, and the combination of Panax ginseng and Angelica sinensis, all have effect on inhibiting pyroptosis and alleviating the inflammatory response [88, 89, 90, 91, 92, 93, 94, 95]. However, the therapeutic effects of these medications on alleviating pyroptosis in AD have not yet been reported.

Despite the substantial research effort focused on finding drugs for the treatment of AD, some therapeutics are single-target drugs, and most of these clinical trials have ended in failure. Therefore, a new approach to developing AD drugs is urgently needed. Considering the complex multifactorial etiology of AD, TCMs, which have the synergistic effects of binding multiple targets and activating multiple pathways, may be safe and ideal candidates as therapies for AD. Therefore, TCMs have broad application prospects in the treatment of AD.

6. Conclusions

All in all, Pyroptosis plays a pivotal role in the progression of AD. Our review highlights key developments in understanding pyroptosis in different CNS cells and AD pathology. Although the pyroptotic machinery has been studied in great detail, it is still a novel research topic, and there are many gaps and challenges in the regulatory mechanisms of pyroptosis during the AD process that should be investigated in future studies. In this review, we considered several critical points, including key pyroptosis-associated regulatory genes, noncoding RNAs, and even results from multiomics analyses. Perhaps pyroptosis-related molecules, such as GSDMD, can be used as biomarkers for diagnosis and prognosis. The exact pathogenetic mechanisms underlying AD remain uncertain, as there are still no drugs that can slow the progression of AD, let alone offer a cure. The exploration of pyroptosis may lead to new ways to treat AD. Overall, pyroptosis is a new perspective on the pathogenesis of AD.

Abbreviations

AD, Alzheimer’s disease; CNS, central nervous system; A $\beta{}$ , $\beta{}$ -amyloid protein; NFTs, neurofibrillary tau tangles; NOD, nucleotide-binding oligomerization domain; IL, interleukin; OL, oligodendrocyte; TCM, traditional Chinese medicine.

Author Contributions

YJ and QZ designed the review. YJ, LZ and SL collected the data. YJ and QZ wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Ethics Approval and Consent to Participate

Not applicable.

Acknowledgment

Thanks to all the peer reviewers for their opinions and suggestions.

Funding

This research was funded by Fund of Science and Technology Development Project of Jilin Province, grant number 20200404076YY.

Conflict of Interest

The authors declare no conflict of interest.

References

[1]

Jack CR, Jr, Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimer’s & Dementia. 2018; 14: 535–562.