Advances in the Study of APOE and Innate Immunity in Alzheimer’s Disease

INTRODUCTION

Originally, Alzheimer’s disease (AD) was described by a doctor named Alois Alzheimer while describing his patients’ conditions [1]. As early as 1911, the medical community began using the term AD to refer to patients with similar symptoms in Europe and the United States [2]. AD is the most common cause of dementia, accounting for about 60–80% of all cases. The hallmark pathology of AD is an accumulation of protein fragments amyloid-β (Aβ) (plaques) outside neurons and twisted strands of tau (tangles) inside neurons [3, 4]. According to official statistics, AD was the sixth leading cause of death among Americans in 2018. In addition, it was the fifth leading cause of death among Americans over the age of 65. In the mid-20th century, there were expected to be more than 13.8 million people with the disease. 121,499 people died from AD in the United States in one year, according to the latest official death records [5]. The latest statistics indicate a rise of 117% in the number of people with dementia worldwide during the 26 years between 1990 and 2016. As a result, dementia is the fifth leading cause of death globally, and by the middle of this century, it will have reached 100 million individuals worldwide [6]. Recent studies have indicated that, as a consequence of globalization trends of irreversible population aging [7], an increasing relationship is now evident between AD and aging [8]. Based on age intervals of five years, the specific incidence rates of AD from 70 to 95 years of age were: 0.31%, 0.77%, 1.77%, 3.54%, 6.65%, and 10.33% [9], respectively. As medical advancements extend human life in the future, it is likely that humans will also become untimely victims of AD.

The causative factors of AD are categorized into two categories, acquired risk factors and genetic risk factors. The acquired risk factors are cerebrovascular disease, hypertension [10], type II diabetes [11], dyslipidemia [12, 13], obesity [14, 15], stress [16], depression [17], smoking [18], and sleep [19]. Genetic factors are amyloid precursor protein (APP), presenilin1, presenilin2 [20, 21], and APOE genes [22]. Genetic factors are responsible for approximately 70% of the causative factors [23]. At the age of 65, AD is classified as early-onset and late-onset, with early-onset accounting for only 4–6% of all cases [23]. It is estimated that more than 90% of AD patients are present after the age of 65 [24]. APP genes, presenilin1 and presenilin2, are primarily associated with early-onset AD [25], while APOE genes are mainly associated with late-onset AD [26]. In summary, genetic factors are the leading cause, most notably the APOE gene.

There is a close relationship between innate immunity and AD [27]. Firstly, innate immunity serves as the body’s first defense against foreign pathogens, and it plays an integral role in tissue repair, wound healing, cell cycle apoptosis, and debris removal. The brain falls within this category [28]. Secondly, AD represents a typical neurodegenerative lesion in which inflammatory lesions of neuronal cells are one of the major causes [29]. In addition, innate immune cells, such as microglia, are instrumental in causing neuroinflammatory lesions. Furthermore, as neuronal cells in the brain become senile, various inflammatory factors are released, producing an imbalance between pro-inflammatory and anti-inflammatory cytokines [30]. This imbalance may lead to abnormal activation of inflammatory factors in the brain, which induces inflammatory lesions in neurons, ultimately leading to the development of AD.

STRUCTURE AND COMPOSITION OF APOE GENES

Among human genes, the APOE gene, located on chromosome 19, contains 299 amino acid residues. APOE is a 34-kDa protein that transports cholesterol and other lipids from the plasma and the CNS by binding to APOE receptors on cell surfaces [31, 32]. The APOE gene comprises three polymorphic alleles, APOE2, APOE3, and APOE4. They differ only at residues 112 and 158. APOE2 contains cysteine residues at both positions; APOE3 contains cysteine at position 112 and arginine at position 158, and APOE4 contains arginine at both positions [33]. Therefore, six genotypes can be formed (APOE2/2, APOE3/3, APOE4/4, APOE2/3, APOE2/4, APOE3/4). Their proportion varies, APOE2 being the least, with only about 8%, APOE3 being as high as nearly 77%, and the remaining 15% or so being APOE4 [32]. The APOE4 gene is the oldest ancestral gene [34], followed by the APOE3 gene, but the APOE2 gene appears after APOE3, making it the youngest gene (originating in East Asia around 8000 years ago) [35]. The APOE2 gene, however, has a significant competitive advantage under natural selection [36]. But as the variants continue to be discovered and studied in depth, things seem to be moving in an uncertain direction.

APOE4 AND AD

Since 1993, after decades of observational experiments, APOE4 has been recognized as the most potent genetic risk factor for AD [37]. On the one hand, APOE4 is closely associated with Aβ in the brain [38]. Moreover, studies have shown that whenever APOE4 is present in the genotype of AD subjects, there is a high probability that Aβ will be present in the subject’s brain. In other words, cerebral amyloid angiopathy (CAA) is significantly more prevalent than any other genotype of APOE [39]. Although the mechanism of action underlying this genetic susceptibility is not fully understood, recent experiments have revealed the role of peripheral cells in CAA mediated by APOE4 through reconstructing the human blood-brain barrier in vitro. And it is highlighted that the nuclear factor of activated T cells signaling regulated by calcineurin might be a therapeutic target for CAA and AD [40]. On the other hand, APOE4 is closely associated with tau. Studies have shown that APOE4 causes more severe brain damage in experimental mice, while APOE4 deficiency provides some protection [41]. In addition to animal studies, a positron emission tomography (PET) study in adults showed that: medial temporal lobe tau proteins are associated with all cognitive impairments. And carriers of the APOE4 gene have a more robust modulation of tau proteins, provided that they are adjusted for PET of Aβ [42]. It provides a direction for elucidating the potential mechanism of APOE action and developing corresponding preventive or therapeutic measures accordingly. Interestingly, the latest research shows that R251G always coinherited with ɛ4 on the APOE gene, which mitigates the ɛ4-associated AD risk [43]. Furthermore, the latest study shows that LilrB3 has been identified as a specific cell surface receptor for APOE4 and the structural basis for this specific ligand–receptor recognition has been elucidated. These experimental findings reveal clues to APOE isoform-dependent functions and AD [44]. Therefore, the future direction of APOE4 is still variable.

APOE3 AND AD

APOE4 is the most potent genetic risk factor, while APOE2 is the most vital genetic protective factor. In addition, APOE3, although it accounts for the most significant percentage of genetic composition (up to 77%), has been characterized as the blandest gene and is generally assumed to play a neutral role in the development of human disease [45, 46]. Nevertheless, a recent case study appears to challenge this conventional wisdom. The patient was a female in her 70 s who presented with cognitive impairment. In contrast, the average age of cognitive impairment in her family was around 45 years, more than 30 years later than the age of her family members. The discovery of this case has aroused great interest among researchers. Testing on her and her relatives showed that her uniqueness was due to the presence of APOE3-Christchurch (R136S) homozygote, despite her high Aβ plaque burden, her tau burden and neurodegeneration were relatively limited [47]. Furthermore, researchers who performed an autopsy on her brain concluded that the presence of APOE3-Christchurch (R136S) homozygote caused the different distribution of tau in the brain [48], A study further showed that Cadherin and Wnt pathways as signaling mechanisms regulated by the APOE3-Christchurch (R136S) homozygote through single cell RNA sequencing in cerebral organoids and an elevation of β-catenin protein, a regulator of tau phosphorylation, as a candidate mediator of APOE3-Christchurch (R136S) homozygote resistance to tauopathy [49]. Another study showed that mutations in APOE3-Jac (V236E), which lies in the backbone of APOE3 carriers, can prevent AD and related dementia. The specific process can promote healthy brain aging by reducing self-aggregation, thereby reducing amyloid plaques and the associated toxicity and thus enhancing the lipidation process. Moreover, the protective APOE3-Jac (V236E) variant substitution reduces APOE4 aggregation and increases cholesterol efflux [50]. These studies may validate the potential protective role of APOE3 and indicate that APOE3 is not neutral for AD but may represent an excellent candidate for genetic protection [51]. In addition, in immunology, bionanoparticles doped with ApoE3 have been demonstrated as a viable, effective, and safe nanovaccine for cancer immunotherapy through the macropinocytosis pathway in recent years [52]. It is expected to provide novel perspectives on preventing and treating AD.

APOE2 AND AD

APOE2 is considered to be the most robust protection gene [37]. Changes in cholesterol and phospholipid composition in bilayers and non-bilayers play a role in the pathogenesis and progression of AD [53]. Through the application of mass spectrometry platforms, histological approaches, and massive data analysis, the APOE2 genotype confers genomically relevant metabolic pathways that have protective effects and may slow the progression of AD and neurodegenerative diseases [54]. A cross-sectional study was designed to investigate the protective effect of APOE2. A total of 4,432 adults were selected from 65 to 85 years of age without cognitive impairment with advanced information regarding APOE types and a clinical dementia score of 0. After nearly three years of data collection, analysis and interpretation, the results showed that the accumulation of Aβ in brains with the APOE2 genotype was significantly lower than that with the APOE4 genotype, both overall and at age. The protective effect of APOE2 was not affected even by the APOE4 genotype. The genotypes already had significant variability at age 65, and APOE4 carriers made up as much as 67% of patients with AD. However, the protective effect of APOE2 also suggested a therapeutic potential for APOE4 carriers [55]. In addition to the cross-sectional study, a pilot cohort study of over 5,000 people was conducted to collect and analyze the data from patients with various genotype combinations of APOE. The trial results showed that after adjusting for age, APOE genotype was associated with a more significant accumulation of Aβ (plaques) and severity of tau protein entanglement, with APOE2/2 being the least severe, APOE2/4 the most severe, and APOE2/3 among them. And more deeply, it was concluded that APOE2/2 had a protective effect that was more than three times that of APOE2/3. Despite the relatively few clinical records of APOE2/2 patients, it is equally possible to conclude that the purity of the APOE2 allele is correlated with low chances of developing AD [56]. This also provides the basis for the promise of future therapeutic approaches through gene editing, protein reduction, protein modification, or other safe and effective replication of the APOE2/2 genotype to prevent clinical AD episodes. APOE2 is associated with human longevity in addition to the prevalence of deficient levels of AD and can even exist independently of AD [57]. Data analysis of over 379,000 individuals supports the notion that APOE2 may serve as an anti-aging gene [58]. The research and development of APOE2 is full of significance in the whole course of human life.

DIFFERENCES BETWEEN APOES

Two main hallmarks of pathological changes leading to AD are the formation of Aβ oligomers and plaques, and the tangling of neuronal fibers. Nevertheless, many years of experimental studies have revealed that APOE correlates more directly and significantly with Aβ [59–61]. The brain contains 20% of the body’s total lipids but only 2% of its mass [62], so it is a lipid-rich environment. According to the previous discussion, APOE is primarily produced by astrocytes and microglia in the brain for the metabolism and transport of lipoproteins. One of the mechanisms that has been investigated is the activation of the mitogen-activated protein kinase signaling cascade, which stimulates immediate early gene phosphorylation and APP gene transcription, resulting in increased APP and AP synthesis. Studies have been conducted on glucose uptake by astrocytes. There is also a significant decrease in glucose levels in the brain when AD occurs [63]. According to the experimental results, glucose uptake by APOE4 astrocytes is reduced compared to glucose uptake by APOE3 astrocytes. However, glucose uptake by APOE2 astrocytes is increased. Individuals expressing APOE4 may exhibit similar glucose metabolism decades before the onset of the disease [64]. A study of astrocytes also revealed that APOE4 astrocytes have fewer protrusions and secrete less laminin and collagen but secrete the most significant amounts of fibronectin than APOE3 astrocytes. APOE4 astrocytes are more sensitive to Aβ because of the lack of a basement membrane. It may lead to slow clearance of Aβ along the internal drainage pathway of the arterial wall and thus promote CAA [65]. Additionally, as Ca2+ signaling is the basis of astrocyte excitability, the experiment measured changes in Ca2+ content in APOE3 and APOE4 astrocytes. It concluded that the APOE gene affects the excitability of astrocytes in a sex-dependent manner [66]. Nevertheless, men and women with the APOE3/4 genotype are equally susceptible to developing AD between 55 and 85 years of age [67]. However, women are more at risk from a younger age. Since APOE2, APOE3, and APOE4 have differential surface lipid and receptor binding efficiencies for receptors [68], it was concluded that there is a hierarchical order of potency between APOE, with APOE4 being stronger than APOE3 and APOE3 being stronger than APOE2 when it comes to stimulating neuronal Aβ production [43]. According to the comparison, APOE3 and APOE4 have been discussed the most [69, 70], primarily concerning their structure and function. Although the results showed overall structural similarity between APOE3 and APOE4, a single amino acid difference at position 112 appears to produce significant differences in the overall functional and physiological behavior of the two proteins concerning plasma lipoprotein metabolism [71]. APOE3 showed a strong affinity for high-density lipoprotein, whereas APOE4 preferred very low-density lipoprotein. APOE3 is stronger than APOE4 in terms of thermal stability [72, 73]. Apart from the genetic differences between APOE2 and APOE4, it is surprising that APOE2 can be used as a medicine to change APOE4 [55]. An interesting experimental study confirmed that genetic modification using the AAVrh. 10hAPOE2-HA vector delivered in the ventricles encoding APOE2 can effectively mediate APOE2 expression and widespread distribution within AD-related regions in a way that reverses or prevents the progressive neurological damage caused by APOE4 [74]. APOE subtypes differ significantly in receptor binding, lipid binding, etc., thus determining the variability of their respective pathways of action and outcomes. And it is a genius idea that good genes can be saved by human intervention through relevant pathways to save the bad ones.

APOE AND INNATE IMMUNITY

Innate immunity, an inherent immunomodulatory function, plays an essential role in the etiology and treatment of AD [75]. It recognizes pathogens and tissue damage via a germline-encoded pattern recognition receptor [76]. This receptor undergoes very rapid activation, which includes responses such as phagocytosis, cell motility, and cytokine production [77, 78]. Typically, these mechanisms are very rapid and effective at eliminating invading pathogens [79]. AD is a lesion in the CNS [23], whereas microglia are the innate immune cells of the CNS [80]. Furthermore, microglia are responsible for regulating the innate immune function of astrocytes [81]. According to current research, microglia and astrocytes are the primary producers of APOE in the brain [82–84]. Thus, there appears to be some association between APOE and innate immunity in AD development [85]. The discovery of the triggering receptor expressed on myeloid cells (TREM) reinforces this conjecture [86, 87]. TREM2, as an immune receptor, functions to sense lipids and mediate myelin phagocytosis in microglia [88]. According to further studies, TREM2-deficient microglia phagocytose myelin debris cannot remove myelin cholesterol, resulting in cholesteryl ester (CE) accumulation. As a result of impaired transport of brain cholesterol, increased CE was observed in APOE-deficient glial cells (Fig. 1). Relevant animal experiments demonstrated that TREM2-deficient microglia are less prone to express APOE [89]. However, the latest research now shows that TREM2-independent microgliosis, facilitates tau-mediated neurodegeneration in the presence of ApoE4 [90]. Therefore, it might be more helpful to begin with microglia to determine the intriguing association between TREM2 and APOE.

OPEN IN VIEWER

Fig. 1

The top half of the illustration shows the APOE or TREM2 deficit situations. The lack of either APOE or TREM2 can cause impaired lipid transport. The bottom part shows the approximate effects of microglia and astrocytes on neuronal cells after activation. Cytokines secreted under moderately stimulating conditions have a protective effect on nerve cells. However, cytokines secreted under conditions of excessive stimulation have toxic effects on nerve cells, which ultimately contribute to the onset and development of AD. Image created with Adobe Illustrator.

As mentioned above: microglia are the primary innate immune cells of the CNS, which are involved in its development [91], normal function, aging, and damage [92]. The brain performs three essential functions: a pre-post function, which is responsible for constantly detecting changes in the brain environment [93]; a housekeeping function, which ensures that neurons remain healthy and functioning correctly [94, 95]; and a defense function, which is responsible for providing neuroprotection and response to sudden changes [96]. It has been shown that TREM2 regulates the three functions of microglia [97]. Additionally, APOE is now recognized as one of its receptors [98–100], allowing TREM2 to form a TREM2-APOE axis, thus linking APOE to microglial function [101]. A recent RNAsep analysis identified a distinct neurodegeneration and disease-associated microglial phenotype controlled by the triggering receptor expressed on myeloid cells 2-apolipoproteinE (TREM2-APOE) pathway [102, 103]. Aside from being synergistic, TREM2 and APOE are also juxtaposed concerning late-onset AD triggers. It has been observed that people with the TREM2 variant R47H are 2–4 times more likely to develop AD than individuals without the TREM2 variant [104–106]. As a result, TREM2 appears to be one of the most important genetic risk factors associated with late-onset AD after APOE. Furthermore, animal experiments for TREM2 [107] and even new drug development have significantly improved [108]. This is exciting news for AD patients.

Based on the above description, one of the innate immune cells, microglia, are closely related to APOE. Despite this, current experiments continue to indicate that astrocytes, which also possess innate immune functions, are the principal source of APOE [109]. Assuming that APOE acts as a mediator, what is the connection between microglia and astrocytes? How do innate immune cells in the brain contribute to the development of AD?

Microglia are macrophages in the brain [110, 111] and prominent members of the brain’s innate immunity [112]. There are many ways in which astrocytes can be linked to microglia [113, 114]. Even though many seemingly interesting scientific studies have been conducted [115–117], they ultimately examine neurogenic inflammatory lesions [118]. In short, when injury and neuropathy occur in the brain, microglia are the first to be activated. They shift from their original steady state to reactive microglia, referred to as M1 reactive microglia and M2 reactive microglia [119–121]. The astrocytes are similar to microglia with changes in A1-reactive astrocytes and A2-reactive astrocytes [122], respectively. It is noteworthy that A1-reactive astrocytes are induced by activated microglia [123]. The scientific community has continuously emphasized the theoretical study of astrocyte-microglia crosstalk [124], reflecting the complexity of the close relationship between microglia and astrocytes from the side. In addition to this, astrocyte-to-microglia relationships are not purely dependent. It has been shown that when the phagocytosis of microglia is impaired, astrocytes take the place of microglia and exercise phagocytosis to remove cellular debris [125] (Fig. 1, cell section).

The complexity of microglia and astrocytes’ phenotypes is expressed simultaneously in AD patients’ brains. Several studies have demonstrated that significant glial gene expression differences correlate with Aβ or tau expression. Microglia are enriched for phagocytosis, inflammation, and proteostasis pathways, while astrocytes are enriched for proteostasis, inflammation, and metal ion homeostasis pathways [126]. The activation of microglia and astrocytes is involved in the clearance of Aβ in the initial AD pathological setting when it is beneficial to the brain environment. As the disease progresses, activated microglia induce a deleterious response through the overexpression of pro-inflammatory factors, resulting in a decrease in Aβ clearance. As a result, an increased accumulation of Aβ in the brain leads to neuroinflammation and degenerative neuropathy. At the same time, reactive astrocytes exhibit neurotoxic effects and loss of neurotrophic function. Astrocyte dysfunction results in increased release of cytokines and inflammatory mediators, neurodegeneration, reduced glutamate uptake, and loss of neuronal synapses, ultimately leading to AD cognitive deficits [102, 127–129] (Fig. 1, cell section). Microglia and astrocytes are a double-edged sword for the brain, but the best use of this sword requires continuous and in-depth investigation by researchers.

APOE, CLINICAL STUDY OF INNATE IMMUNITY AND AD

Innate immunity, APOE, and AD have a very close relationship, so it has been expected that targeting APOE or innate immunity can hinder AD. As a result of current research findings, it has been determined that strategies targeting APOE for AD can broadly be categorized into three categories: regulating the amount [130, 131] and lipidation [132, 133] of APOE; focusing on the specificity of APOE structure [134] and its interaction with Aβ [135]; targeting the receptor of APOE [136–138]. As a result, most clinical trials have also been conducted according to these categories. Data show that clinical trials for APOE and AD were first conducted in 2005, followed by studies involving drugs, diet, cohort studies, research, and so forth. Although clinical trials on innate immunity and AD currently involve only prospective cohort studies, the importance of innate immunity in AD is evident. Therefore, there is a possibility that more useful clinical trials will emerge shortly for the ultimate benefit of society as a whole (Table 1). In the current table, there are only three completed experiments. The numbers are NCT02198586, NCT02061878, and NCT02707458. The results of the experiment belonging to NCT02198586 showed that brain MRIs of healthy middle-aged participants show a relatively high prevalence of incidental findings and most of participants are first-degree descendants of patients with AD, therefore these results are of special relevance for novel imaging studies in the context of AD prevention in cognitively healthy middle-aged participants [139]. Further result of the analysis in terms of APOE showed a correlation between APOE4 and vulnerable cerebral vessels [140]. Bexarotene increased cerebrospinal fluid ApoE by 25% but had no effect on metabolism of Aβ peptides, so the NCT02061878 study argues that treatment of individuals with an intact blood-brain barrier are unlikely to benefit from bexarotene administration [141]. No results from the NCT02707458 were published. The results of current clinical studies are not yet surprising, and the difficult development of clinical studies reflect more clearly that have long way to go to AD treatment.

CONCLUSION AND FUTURE DIRECTIONS

Considering the real world of slow human aging [142], AD will undoubtedly become one of the greatest hidden dangers threatening the health of all human beings. Scientists have been striving to achieve a breakthrough in AD from a genomic perspective for decades [143–145]. APOE has been the star gene of interest since 1993 when it was first proposed to be associated with AD [146]. In nearly a third of a century, approximately 10,000 publications have been published in which APOE has been associated with AD [147]. As identified by the current association between APOE and AD, APOE is one of the most crucial genetic factors involved in lipid transport and cholesterol transport throughout the CNS by encoding polymorphic proteins that bind to corresponding receptors. It is possible to trigger a range of neuropathies in the brain if this transport balance is disturbed in an environment rich in lipids. APOE is linked to AD, innate immunity and AD are closely related, and APOE is derived from innate immune cells; it seems like a promising approach that should bear further investigation. When the appropriate conditions are met, APOE, innate immunity and AD can be grouped into an APOE-innate immunity-AD axis, ultimately contributing to the effective prevention of AD or hindering disease progression.

According to the consensus, AD is an insidious neurodegenerative disease whose complex molecular and cellular mechanisms are still poorly understood despite many years of research. Single-cell transcriptomic analysis techniques may provide the blueprint for understanding AD’s cellular and molecular basis [148]. In addition to geriatric animal models, more sophisticated human 3D brain modeling techniques have been developed. It can facilitate not only the identification and dissection of AD-specific pathological mechanisms but also the way to study neuroinflammatory responses due to Aβ deposits, such as testing microglia or other immune cells in co-culture conditions [149]. As far as an AD treatment is concerned, targeted therapies for APOE, gene editing, and gene recombination [145] continue progressing from theory to practice, which is positive news. Moreover, a series of immunosuppressive therapies [150] and even future modulations of immune genes [151] will contribute to the further expansion of therapeutic options for AD. At some point in the future, AD progression may be hindered if APOE and innate immunity work together to combat AD.