Apolipoprotein E and Alzheimer disease: risk, mechanisms, and therapy

Abstract

Apolipoprotein E (ApoE) is a major cholesterol carrier that supports lipid transport and injury repair in the brain. APOE polymorphic alleles are the main genetic determinants of Alzheimer disease (AD) risk: individuals carrying the ε4 allele are at increased risk of AD compared with those carrying the more common ε3 allele, whereas the ε2 allele decreases risk. Presence of the APOE ε4 allele is also associated with increased risk for cerebral amyloid angiopathy and age-related cognitive decline during normal ageing. ApoE–lipoproteins bind to several cell-surface receptors to deliver lipids and also to hydrophobic amyloid-β (Aβ) peptide, which is thought to initiate toxic events that lead to synaptic dysfunction and neurodegeneration in AD. ApoE isoforms differentially regulate Aβ aggregation and clearance in the brain, and have distinct functions in regulating brain lipid transport, glucose metabolism, neuronal signalling, neuroinflammation, and mitochondrial function. In this Review, we describe current knowledge on ApoE in the CNS, with a particular emphasis on the clinical and pathological features associated with carriers of different ApoE isoforms. We also discuss Aβ-dependent and Aβ-independent mechanisms that link ApoE4 status with AD risk, and consider how to design effective strategies for AD therapy by targeting ApoE.

Introduction

Alzheimer disease (AD) is a progressive neurodegenerative disease associated with cognitive decline and is the most common form of dementia in the elderly. Approximately 13% of people over the age of 65 and 45% over the age of 85 are estimated to have AD.¹ Mounting evidence from genetic, pathological, and functional studies has shown that an imbalance between the production and clearance of amyloid-β (Aβ) peptides in the brain results in accumulation and aggregation of Aβ. The toxic Aβ aggregates in the form of soluble Aβ oligomers, intraneuronal Aβ, and amyloid plaques injure synapses and ultimately cause neurodegeneration and dementia.^{2, 3} The toxicity of Aβ seems to depend on the presence of microtubule-associated protein tau,⁴ the hyperphosphorylated forms of which aggregate and deposit in AD brains as neurofibrillary tangles. Aβ is composed of 40 or 42 amino acids and is generated through proteolytic cleavage of amyloid precursor protein (APP).⁵

Early-onset familial AD, which typically develops before the age of 65 years and accounts for only a small portion (<1%) of AD cases,^{2, 3} is primarily caused by overproduction of Aβ owing to mutations in either the APP gene or genes encoding presenilin 1 (PSEN1) or presenilin 2 (PSEN2), essential components of the γ-secretase complexes responsible for cleavage and release of Aβ. The majority of AD cases occur late in life (>65 years) and are commonly referred to as late-onset AD (LOAD). Although multiple genetic and environmental risk factors are involved in LOAD pathogenesis, overall impairment in Aβ clearance is probably a major contributor to disease development.⁶ Genetically, the ε4 allele of the apolipoprotein E (APOE) gene is the strongest risk factor for LOAD.^7–9 The human APOE gene exists as three polymorphic alleles—ε2, ε3 and ε4—which have a worldwide frequency of 8.4%, 77.9% and 13.7%, respectively.¹⁰ However, the frequency of the ε4 allele is dramatically increased to ~40% in patients with AD.¹⁰

ApoE regulates lipid homeostasis by mediating lipid transport from one tissue or cell type to another.¹¹ In peripheral tissues, ApoE is primarily produced by the liver and macrophages, and mediates cholesterol metabolism in an isoform-dependent manner. ApoE4 is associated with hyperlipidaemia and hypercholesterolemia, which lead to atherosclerosis, coronary heart disease and stroke.^{11, 12} In the CNS, ApoE is mainly produced by astrocytes, and transports cholesterol to neurons via ApoE receptors, which are members of the low-density lipoprotein receptor (LDLR) family.⁸

ApoE is composed of 299 amino acids and has a molecular mass of ~34 kDa.¹¹ Differences between the three ApoE isoforms are limited to amino acids 112 and 158, where either cysteine or arginine is present (Figure 1a): ApoE2 (Cys112, Cys158), ApoE3 (Cys112, Arg158), and ApoE4 (Arg112, Arg158).¹¹ The single amino acid differences at these two positions affect the structure of ApoE isoforms and influence their ability to bind lipids, receptors and Aβ.^13–15 Human and animal studies clearly indicate that ApoE isoforms differentially affect Aβ aggregation and clearance. Several Aβ-independent functions are also associated with ApoE isoforms. In this Review, we provide an overview of clinical evidence for the association between APOE genotypes and the risk of cognitive decline in AD, mild cognitive impairment (MCI) and other CNS diseases with a cognitive component, and discuss our current understanding of the mechanisms underlying ApoE actions and ApoE-targeted therapies.

Figure 1. APOE ε4 is a major genetic risk factor for Alzheimer disease.

(a) The ApoE2, E3, and E4 isoforms, which are encoded by the ε2, ε3 and ε4 alleles of the APOE gene, respectively, differ from one another at amino acid residues 112 and/or 158 (red circles). ApoE has two structural domains: the N-terminal domain, which contains the receptor-binding region (residues 136–150), and the C-terminal domain, which contains the lipid-binding region (residues 244–272); the two domains are joined by a hinge region. A meta-analysis demonstrated a significant association between the ε4 allele of APOE and AD.¹⁰ (b) APOE ε4 increases the risk of AD and lowers the age of disease onset in a gene-dose-dependent manner.^{7, 20} Abbreviations: AD, Alzheimer disease; ApoE, Apolipoprotein E.

APOE genotypes, AD and cognition

APOE ε4 as a strong risk factor for AD

Genome-wide association studies have confirmed that the ε4 allele of APOE is the strongest genetic risk factor for AD.^{16, 17} The presence of this allele is associated with increased risk for both early-onset AD and LOAD.^{18, 19} A meta-analysis of clinical and autopsy-based studies demonstrated that, compared with individuals with an ε3/ε3 genotype, risk of AD was increased in individuals with one copy of the ε4 allele (ε2/ε4, OR 2.6; ε3/ε4, OR 3.2) or two copies (ε4/ε4, OR 14.9) among Caucasian subjects.¹⁰ The ε2 allele of APOE has protective effects against AD: the risk of AD in individuals carrying APOE ε2/ε2 (OR 0.6) or ε2/ε3 (OR 0.6) are lower than those of ε3/ε3.¹⁰ In population-based studies, the APOE4–AD association was weaker among African Americans (ε4/ε4, OR 5.7) and Hispanics (ε4/ε4, OR 2.2) and was stronger in Japanese people (ε4/ε4, OR 33.1) compared with Caucasian cases (ε4/ε4, OR 12.5).¹⁰ APOE ε4 is associated with increased prevalence of AD and lower age of onset.^{7, 10, 20} The frequency of AD and mean age at clinical onset are 91% and 68 years of age in ε4 homozygotes, 47% and 76 years of age in ε4 heterozygotes, and 20% and 84 years in ε4 noncarriers,^{7, 20} indicating that APOE ε4 confers dramatically increased risk of development of AD with an earlier age of onset in a gene dose-dependent manner (Figure 1b).

Genetic variants in the TOMM40 (translocase of outer mitochondrial membrane 40 homologue) gene, which lies adjacent to the APOE gene on chromosome 19, have been implicated as a modulator of AD age-of-onset in APOE ε3 carriers.²¹ A more recent study, however, has cast doubt on the strength of this association.²² Whether the effects of APOE and TOMM40 on AD risk, both genetically and functionally, are synergistic requires further investigation.

APOE and Aβ deposition

ApoE has an important role in Aβ metabolism (Figure 2). Studies show that APOE genotypes strongly affect deposition of Aβ to form senile plaques and cause cerebral amyloid angiopathy (CAA), two major hallmarks of amyloid pathology in AD brains.²³ Immunohistological evidence demonstrates that ApoE is co-deposited in senile plaques in the brains of AD patients.²⁴ The Aβ deposition in the form of senile plaques is more abundant in APOE ε4 carriers compared with noncarriers.^25–27 The difference was most evident among individuals aged 50–59 years: 40.7% of APOE ε4 carriers had senile plaques compared with 8.2% of noncarriers.²⁵ In individuals with positive Pittsburgh compound B (PiB)-PET images, which indicate fibrillar aggregates of Aβ,²⁸ APOE ε4 was more common than in those with negative scans (65% versus 22%) in patients with AD.²⁹

Figure 2. Apolipoprotein E and amyloid-β metabolism in the brain.

The main Aβ clearance pathways include receptor-mediated uptake by neurons and glia, drainage into interstitial fluid or through the BBB, and proteolytic degradation by IDE and neprilysin. Impaired clearance of Aβ can cause Aβ accumulation in brain parenchyma, leading to formation of neurotoxic Aβ oligomers and amyloid plaques. Aβ accumulation in the perivascular region leads to CAA, which disrupts blood vessel function. ApoE is primarily synthesized by astrocytes and microglia, and is lipidated by the ABCA1 transporter to form lipoprotein particles. Lipidated ApoE binds to soluble Aβ and facilitates Aβ uptake through cell surface receptors, including LRP1, LDLR, and HSPG^{175, 177} in a manner that probably depends on ApoE isoform and its level of lipidation. ApoE facilitates binding and internalization of soluble Aβ by glial cells, disrupts Aβ clearance at the BBB in an isoform-dependent manner (ApoE4 > ApoE3 > ApoE2) and influences CAA pathogenesis. Abbreviations: Aβ, amyloid-β; ABCA1, ATP-binding cassette A1; BBB, blood–brain barrier; CAA, cerebral amyloid angiopathy; HSPG, heparan sulphate proteoglycan; IDE, insulin-degrading enzyme; LDLR, low-density lipoprotein receptor; LRP1, low-density lipoprotein receptor-related protein 1; LXR, liver X receptor.

Fibrillar Aβ deposition is often detected in the brains of elderly, cognitively normal individuals in a manner that depends on the presence of APOE ε4, although such an association is weaker than that in patients with AD.³⁰ In addition, APOE ε4 carriers have lower cerebrospinal fluid (CSF) Aβ₄₂ levels and higher PiB-positive imaging, which reflect the presence of cerebral amyloid deposition and serve as potential biomarkers for AD.^{31, 32} Cognitively normal APOE ε4 carriers exhibit PiB-positive imaging about 56 years of age, compared with about 76 years of age in noncarriers.³³ This difference suggests that APOE ε4 probably increases the risk of AD by initiating and accelerating Aβ accumulation, aggregation and deposition in the brain. Although APOE ε2 reduces the risk of dementia,³⁴ in individuals older than 90 years, both the ε2 and ε4 alleles of APOE increase amyloid burden compared with APOE ε3, suggesting that the protective effects of APOE ε2 against AD might not be associated with Aβ deposition.

APOE ε4 also shows an association with CAA and CAA-related haemorrhages.^{35, 36} CAA refers to the pathological condition in which amyloid spreads and deposits throughout the cerebral blood vessel walls³⁷ and is frequently detected in AD.²³ Interestingly, although APOE ε2 is protective against AD, it is a risk factor for CAA-related haemorrhage, independently of AD, possibly by predisposing vessels to vasculopathic complications of CAA.³⁶

Prediction of AD in MCI

MCI is a transitional stage between normal ageing and dementia, and is associated with increased risk of AD.³⁸ The rate at which patients with amnesic MCI (aMCI) progress to clinically diagnosable AD is 10–15% per year, in contrast to a rate of 1–2% per year among healthy elderly individuals.³⁹ The prevalence of APOE ε4 is substantially higher in both aMCI and dys-executive MCI than in control individuals.⁴⁰ Patients with MCI who harbour APOE ε4 exhibit distinct cognitive profiles, which seem to resemble those of patients in the early stages of AD.⁴¹ A case–control study reported poorer memory performance among patients with MCI who were carriers of APOE ε4 compared with noncarriers.⁴² APOE ε4 is associated with impaired memory performance and increased risk of memory decline in middle-aged (40–59 years) and elderly (60–85 years) people with MCI.^{43, 44} Furthermore, patients with MCI who are carriers of APOE ε4 experience more-rapid decline in several cognitive and functional assessments, and severity of the deficits is strongly associated with the APOE ε4 gene dose.^{41, 45, 46} Importantly, the presence of APOE ε4 is associated with increased risk of progression from MCI to AD-type dementia.^47–49 Among individuals with aMCI, APOE ε4 carriers tend to be younger than noncarriers, consistent with younger age of AD onset in individuals with APOE ε4.⁵⁰ These findings indicate that the APOE ε4 genotype in patients with MCI can serve as a predictive factor for determination of clinical outcome and the risk of conversion to AD.

In patients with MCI, the adverse effects of APOE ε4 on cognitive functions correlate with the severity of neuronal pathology. Those who are carriers of APOE ε4 have lower CSF Aβ₄₂ levels, higher tau levels and greater brain atrophy than do noncarriers.⁵⁰ Furthermore, patients with MCI who are PiB-positive are more likely to be APOE ε4 carriers and exhibit worse memory performance than are PiB-negative patients.⁵¹ Other finding suggest, although not without controversy,⁵² that APOE ε4 has considerable deleterious effects on memory performance⁴² and might be used to predict disease progression in combination with AD biomarkers and neuroimaging approaches.⁵³

Predicting cognitive decline in healthy cases

Healthy APOE ε4 carriers not diagnosed with MCI or AD show an accelerated longitudinal decline in memory tests, which starts around the age of 55–60 years, revealing a possible pre-MCI state in this genetic subset of individuals.^{54, 55} This memory decline, despite ongoing normal clinical status, suggests that pathological changes in AD might manifest in the brain as early as the sixth decade of life.^{56, 57} Thus, APOE ε4 is associated with cognitive decline many years before cognitive impairment becomes clinically apparent.^{56, 58} Interestingly, APOE ε4 has differential effects on memory performance depending on age. Some studies in young adults and children have found evidence of better cognitive performance in APOE ε4 carriers than in noncarriers, which could suggest antagonistic pleiotropy,^59–61 in which APOE ε4 might offer benefits during development and early adulthood at the expense of more-rapid decline in cognitive function with ageing.⁶²

Similar to the situation in patients with MCI, APOE ε4 is associated with enhanced amyloid pathology in cognitively normal people. The proportion of PiB-positive individuals follows a strong APOE allele-dependent pattern (ε4 > ε3 > ε2),^{25, 63, 64} and APOE ε4 increases the amount of amyloid deposition in a gene-dose-dependent manner.³⁰

APOE ε4 and other AD risk factors

The APOE ε4 genotype combines synergistically with atherosclerosis, peripheral vascular disease, or type 2 diabetes in contributing to an increased risk of AD.^{65, 66} APOE ε4 is a risk factor for cardiovascular disease, suggesting that this allele and cerebrovascular disease might have compounding effects on cognitive decline in AD.⁶⁷ Diabetes also increases the risk of AD, and the association is particularly strong among APOE ε4 carriers.^{66, 68, 69} Patients with diabetes who are carriers of APOE ε4 have more neuritic plaques, neurofibrillary tangles and CAA than do noncarriers.⁶⁶ The combination of a diabetes-related factor—that is, hyperglycaemia, hyperinsulinaemia, and insulin resistance—and the APOE ε4 allele promotes neuritic plaque formation.⁶⁹ APOE ε4 seems to modify the risk of AD in patients with diabetes—a disease that directly or indirectly causes vascular and neuronal damage and further exacerbates AD pathology. Furthermore, recent research demonstrated that, independently of Aβ, ApoE4 triggers inflammatory cascades that cause neurovascular dysfunction, including blood–brain barrier breakdown, leakage of blood-derived toxic proteins into the brain and reduction in the length of small vessels.⁷⁰ This result suggests that ApoE4-associated damage to vascular systems in brain could have a key role in AD pathogenesis.

APOE and traumatic brain injury

Increasing evidence has shown that APOE ε4 is associated with poorer outcomes following traumatic brain injury (TBI), regardless of the severity of initial injury.⁷¹ A meta-analysis demonstrated that the outcome of TBI at 6 months after injury is worse in APOE ε4 carriers.⁷² TBI is associated with increased risk of AD,⁷³ and such a risk is more evident in patients with APOE ε4.⁷⁴ Only 10% of APOE ε4 noncarriers with TBI have Aβ plaque pathology, whereas 35% and 100% of TBI patients with one or two APOE ε4 alleles, respectively, possess Aβ pathology.⁷⁵ The poorer outcomes associated with ApoE4 might relate to its reduced ability to repair and remodel synapses and protect neurons upon injury compared with ApoE3.⁸ These possibilities are currently under investigation.

APOE and vascular diseases

Vascular cognitive impairment, which comprises clinical conditions with cerebrovasculature-derived cognitive disturbances including vascular dementia, is observed in approximately 8–15% of aged individuals with cognitive dysfunction in Western clinic-based series.⁷⁶ A recent meta-analysis has shown evidence of increased risk of vascular dementia in individuals with APOE ε4 compared with APOE ε3 (OR 1.72).⁷⁷ Several studies suggest that the contribution of APOE ε4 to risk of vascular cognitive impairment is independent of other vascular risk factors including hypertension, dyslipidaemia and atherogenesis,⁷⁸ whereas another report shows that age-related cognitive decline among APOE ε4 carriers is induced by brain damage owing to increased blood pressure.⁷⁹ In addition, APOE ε4 is associated with poor outcome after subarachnoid haemorrhage,⁸⁰ and is a strong risk factor for CAA-related intracranial haemorrhage.⁸¹ These results suggest that APOE ε4 is closely associated with neurovascular dysfunctions.

APOE and other types of dementia

Lewy body disease is thought to be the second most common kind of dementia comprised of a spectrum of diseases that includes Parkinson disease (PD), PD-associated dementia and dementia with Lewy bodies (DLB). Clinical and pathological features of PD and AD frequently overlap. Most studies, however, have failed to report associations between APOE ε4 and susceptibility to PD and PD-associated dementia.^{82, 83} DLB also shares clinical and pathological characteristics with AD and PD,⁸⁴ and several reports have shown that APOE ε4 increases risk of DLB.⁸⁵ Immunohistochemical analysis showed that deposition of Lewy bodies in patients with DLB who are APOE ε4 carriers is substantially more abundant than those who are noncarriers.⁸⁶ As Lewy bodies are considerably increased in the cerebral cortex of DLB patients with Aβ deposition,⁸⁷ the strong association between amyloid pathology and the pathology of Lewy body disease could explain why APOE ε4 increases risk of DLB. APOE ε4 might also be a risk factor for frontotemporal dementia,⁸⁸ although the pathophysiological role of ApoE in this disease requires further investigation. The APOE genotypes do not seem to influence the risks of Huntington disease ⁸⁹ nor amyotrophic lateral sclerosis.⁹⁰

Mechanisms of ApoE isoforms in AD

APOE ε4 confers a gain of toxic functions, a loss of neuroprotective functions or both in the pathogenesis of AD (Figure 3).

Figure 3. The role of Apolipoprotein E4 in Alzheimer disease pathogenesis.

ApoE4 confers toxic gain of function, loss of neuroprotective function or both in the pathogenesis of Alzheimer disease. Key functional differences between ApoE4 and ApoE3 are illustrated. Abbreviations: Aβ, amyloid-β; ApoE, Apolipoprotein E.

Aβ metabolism and aggregation

Studies in humans and transgenic mice showed that brain Aβ levels and amyloid plaque loads are ApoE isoform-dependent (ε4 > ε3 > ε2),^{30, 63, 91} suggesting an important role of ApoE in modulating Aβ metabolism, aggregation, and deposition. ApoE4 is less efficient in Aβ clearance than is ApoE3 in young and old amyloid mouse models that express human ApoE isoforms.⁶³ Additionally, ApoE isoforms differentially regulate cholesterol levels, which have been shown to modulate γ-secretase activity and Aβ production.⁹² Several studies reported an APOE genotype-dependent effect on CSF and brain ApoE levels (ε4 < ε3 < ε2) in ApoE-targeted replacement (ApoE-TR) mice, in which the mouse Apoe gene is replaced with human APOE isoforms.^{91, 93, 94} This result suggests that lower levels of total ApoE exhibited by APOE ε4 carriers might contribute to disease progression. However, whether human ApoE isoform status affects CSF and brain ApoE protein levels in healthy individuals and patients with AD remains to be established.^{95, 96}

ApoE-knockout mice clear Aβ from the brain faster than do control mice.⁹⁷ Stimulation of liver X receptors (LXRs)^{98, 99} or the retinoid X receptor (RXR)¹⁰⁰ facilitates Aβ clearance, probably by increasing ApoE levels and lipidation. Further investigation is needed to determine whether ApoE levels are directly associated with Aβ clearance. In addition, a recent study showed that lack of one copy of ATP-binding cassette transporter A1 (ABCA1), which shuttles lipids to ApoE, impairs Aβ clearance and exacerbates amyloid deposition and memory deficits in ApoE4-TR mice, but not in ApoE3-TR mice.¹⁰¹ This result suggests that ApoE isoforms exhibit differential lipidation status, which affects Aβ clearance in an isoform-dependent manner. Alternatively, ApoE–lipoprotein particles may sequester Aβ and promote cellular uptake and degradation of ApoE–Aβ complexes.¹⁰²

ApoE4–lipoproteins bind Aβ with lower affinity than do ApoE3–lipoproteins,¹⁰³ suggesting that ApoE4 might be less efficient in mediating Aβ clearance. In addition, ApoE might modulate Aβ removal from the brain to the systemic circulation by transporting Aβ across the blood–brain barrier. In this respect, ApoE impedes Aβ clearance at the blood–brain barrier in an isoform-specific fashion (ApoE4 > ApoE3 and ApoE2).¹⁰⁴ Finally, studies in microglia have shown that ApoE3 promotes enzyme-mediated degradation of Aβ more efficiently than does ApoE4.¹⁰⁵ Together, these studies suggest that ApoE4 inhibits Aβ clearance and/or is less efficient in mediating Aβ clearance compared with ApoE3 and ApoE2.

ApoE also seems to regulate Aβ aggregation and deposition. An important study showed that deletion of the mouse Apoe gene essentially eliminates deposition of fibrillar Aβ in amyloid model mice.¹⁰⁶ Given that ApoE is co-deposited with Aβ in human AD brains,²⁴ it is possible that ApoE promotes Aβ aggregation and deposition in an isoform-dependent manner. The exact mechanisms by which ApoE isoforms differentially regulate Aβ aggregation and deposition require further investigation.

Brain activity and atrophy

AD is associated with both functional abnormalities of the hippocampus and cortical atrophy in the memory network.^{107, 108} Patients with AD or MCI who are APOE ε4 carriers exhibit greater medial temporal lobe atrophy, particularly in the hippocampal area.^{41, 109, 110} Structural MRI studies found that, compared with noncarriers, APOE ε4 carriers have accelerated age-related loss in cortical thickness and hippocampal volume that are tightly coupled to decline in cognitive performance.^111–113

Functional MRI (fMRI) studies reported that ApoE4 disrupts resting state fMRI connectivity and the balance between brain networks, in the absence of amyloid pathology.^{114, 115} Furthermore, cognitively normal APOE ε4 carriers have elevated resting-state activity in the default mode network—a network that is preferentially affected early in AD—and higher hippocampal activation during memory tasks.^116–118 Such changes have been hypothesized to represent a compensatory response by APOE ε4 carriers in which increased cognitive effort is required to achieve an equivalent level of performance to that of noncarriers.^{116, 118}

Elevated baseline activity in brain networks of APOE ε4 carriers could potentially contribute to increased Aβ production, as Aβ levels are regulated by neuronal activity.^{119, 120} Interestingly, in adults who do not have dementia, increased hippocampal activity was associated with reduced cortical thickness in the medial temporal lobe and brain regions that are vulnerable to AD pathology.¹²¹ Studies suggested that hippocampal hyperactivity might represent impending synaptic dysfunction and incipient cognitive decline.¹²² Interestingly, another study showed a reduction of posterior default mode network connectivity in APOE ε4 carriers in cognitively normal elderly people, implying that APOE ε4 carriers may exhibit a more rapid decline in connectivity of this network than do noncarriers as they age.¹¹⁵

¹⁸F-fluorodeoxyglucose PET imaging, which measures cerebral metabolic rates of glucose as a proxy for neuronal activity, correlates with disease progression and predicts histopathological diagnosis in AD.¹²³ Mounting evidence suggests that APOE ε4 carriers exhibit lower cerebral glucose metabolism.^124–126 Healthy adults with APOE ε4 show altered patterns of brain metabolism both at rest and during cognitive challenges compared with noncarriers.^{126, 127} Representative studies illustrating the association of ApoE4 isoform with altered brain metabolism and activity, memory decline, and amyloid pathology in cognitively normal people are shown in Figure 4. Improved understanding of the mechanisms of ApoE4-related brain activity changes, brain atrophy and reduced metabolism should help to explain why ApoE4 is a risk factor for cognitive decline and AD.

Figure 4. Abnormal brain function and enhanced neuropathology and memory decline in cognitively normal APOE ε4 carriers.

(a) ¹⁸F-fluorodeoxyglucose PET images show that cognitively normal APOE ε4 carriers have lower glucose metabolism than do noncarriers. (b) APOE ε4 carriers exhibit a greater increase in functional MRI signal in brain regions associated with task performance, and show increases in additional regions compared with APOE ε3 carriers. (c) Age-related memory decline occurs more rapidly in APOE ε4 carriers than noncarriers, starting from age 55–60 years. (d) APOE ε4 carriers show increased cerebral Aβ deposition which persists in greater frequencies with age compared with noncarriers. Increased PiB binding and reduced CSF Aβ₄₂ levels reflect cerebral amyloid deposition. Abbreviations: Aβ, amyloid-β; APOE, apolipoprotein E; CSF, cerebrospinal fluid; PiB, Pittsburg compund B. Part a, is modified, with permission from the National Academy of Science, USA © Small, G. W. et al. Proc. Natl Acad. Sci. USA 97, 6037–6042 (2000). Part b, is modified, with permission from the Massachusetts Medical Society © Bookheimer, S. Y. et al. N. Engl. J. Med. 343, 450–456 (2000). Part c, is modified, with permission from the Massachusetts Medical Society © Caselli et al. N. Engl. J. Med. 361, 255–263 (2009). Part d is modified, with permission, from John Wiley and Sons © Morris, J. C. et al. Ann. Neurol. 67, 122–131 (2010).

Tau phosphorylation and neurotoxicity

ApoE is produced primarily by astrocytes and microglia. Neuronal ApoE expression can, however, be induced in response to stress or injury, probably for the purpose of neuronal repair and remodelling.^{128, 129} A truncated fragment of ApoE4, resulting from proteolytic cleavage of ApoE following stress or injury, increases tau hyperphosphorylation, cytoskeletal disruption and mitochondrial dysfunction.^{128, 130, 131} ApoE4 also exacerbates neurotoxicity triggered by Aβ and other insults.^{128, 131}

A recent study showed that neurons in patients with temporal lobe epilepsy who harbour APOE ε4 are less resilient to the damaging hyperexcitability and more susceptible to Aβ toxicity than are those in APOE ε4 carriers,¹³² suggesting that ApoE3 might confer a neuroprotective advantage over ApoE4 against neuronal stress. Interestingly, astrocyte-derived ApoE4 has neuroprotective effects against excitotoxic injuries, whereas neuronal expression of ApoE4 promotes excitotoxic cell death. This result suggests that ApoE derived from various cellular sources might exhibit different physiological and pathological activity.¹³³

Lipid metabolism

Abnormal lipid metabolism is strongly related to the pathogenesis of AD. In the CNS, ApoE mediates neuronal delivery of cholesterol, which is an essential component for axonal growth, synaptic formation and remodelling—events that are crucial for learning, memory formation and neuronal repair.^{134, 135} Brain cholesterol levels are substantially reduced in hippocampal and cortical areas in patients with AD compared with age-matched controls.¹³⁶ Preferential degradation of ApoE4 relative to ApoE3 in astrocytes has been proposed to result in low levels of ApoE in the brain and CSF and reduced capacity for neuronal delivery of cholesterol, suggesting that low levels of total ApoE exhibited by APOE ε4 carriers may directly contribute to the disease progression.⁹³ ApoE4 is also less efficient than ApoE3 in transporting brain cholesterol.¹³⁷ Moreover, ApoE4-TR mice have abnormal cholesterol levels and impaired lipid metabolism.¹³⁸ Insufficient levels of ApoE and/or impaired ApoE function in carriers of the ε4 allele might, therefore, lead to aberrant CNS cholesterol homeostasis and neuronal health, which contribute to AD risk.

Synaptic plasticity and spine integrity

Synaptic failure is an early pathological feature of AD.^{139, 140} Increasing evidence demonstrates that ApoE isoforms differentially regulate synaptic plasticity and repair.^{141, 142} In AD and healthy aged controls, APOE ε4 gene dosage correlates inversely with dendritic spine density in the hippocampus.¹⁴³ ApoE4-TR mice also have lower dendritic spine density and length compared with ApoE3-TR mice.^{144, 145} ApoE3, but not ApoE4, prevents loss of synaptic networks induced by Aβ oligomers.¹⁴⁶ ApoE isoforms also differentially regulate dendritic spines during ageing.^{143, 147} The age-dependence of these differences implies that the effects of ApoE isoforms on neuronal integrity might relate to increased risk of dementia in aged APOE ε4 carriers.

Reduced synaptic transmission was observed in 1-month-old ApoE4-TR mice compared with ApoE3-TR mice, suggesting that ApoE4 may also contribute to functional deficits early in development, which could account for alteration of neuronal circuitry that eventually results in cognitive disorders later in life.¹⁴⁷ In addition, ApoE4 selectively impairs ApoE receptor trafficking and signalling, as well as glutamate receptor function and synaptic plasticity.¹⁴¹ Together, these findings suggest that the effect of APOE ε4 genotype on risk of AD might be mediated, at least in part, through direct effects on synaptic function.

Neuroinflammation

Neuroinflammation contributes to neuronal damage in the brain and is implicated in AD pathogenesis.¹⁴⁸ ApoE colocalizes with plaque-associated amyloid and microglia, suggesting a role for ApoE in the innate immune response in AD. Lack of ApoE in mice is associated with increased inflammation in response to Aβ,^{149, 150} but ApoE isoforms might differently regulate the innate immune response.¹⁵¹ ApoE4 seems to have proinflammatory and/or reduced anti-inflammatory functions, which could further exacerbate AD pathology. For example, ApoE4-TR mice exhibit greater inflammatory responses to lipopolysaccharide compared with ApoE3-TR mice.¹⁵² In addition, young APOE ε4 carriers show an increased inflammatory response that may relate to AD risk later in life.¹⁵³ Consistent with this notion, non-steroidal anti-inflammatory drugs were shown to reduce AD risk only in APOE ε4 carriers,¹⁵⁴ suggesting that APOE genotype might determine the effect of anti-inflammatory medications for AD.

Neurogenesis

Hippocampal neurogenesis has an important role in structural plasticity and maintenance of brain networks. Dysfunctional neurogenesis resulting from early disease manifestations could, therefore, exacerbate neuronal vulnerability to AD and contribute to memory impairment.¹⁵⁵ ApoE is required for maintenance of the neural stem or progenitor cell pool in the adult dentate gyrus region of the hippocampus.¹⁵⁶ In ApoE-TR mice, ApoE4 inhibits hippocampal neurogenesis by impairing maturation of hilar γ-aminobutyric acid-containing interneurons, which contributes to learning and memory deficits.^{157, 158} These results demonstrate an important pathological role of ApoE4 in impairment of neurogenesis, which might contribute to AD pathogenesis.

ApoE as a therapeutic target for AD

Most therapeutic approaches for AD target the Aβ pathway. With the recent failure of clinical trials of drugs targeting solely Aβ, an urgent need exists to define new targets and develop alternative therapeutic strategies to treat AD. As APOE genotype determines AD risk, and ApoE has crucial roles in cognition, ApoE might offer an attractive alternative target for AD therapy. APOE genotype status could be included in clinical trial enrolment criteria, as some therapies might be effective only in specific APOE genotypes. Here, we briefly discuss several approaches that are currently being explored (Table 1).

Table 1.

ApoE-targeted strategies for treatment of Alzheimer disease

Strategy	Rationale	Examples
Pharmacological approaches
Modulate ApoE levels	Promotes Aβ clearance, lipid homeostasis and synaptic function	LXR/RXR agonists, small molecules
Increase ABCA1 expression	Promotes ApoE lipidation and stabilizes ApoE, thereby decreasing amyloid deposition	LXR/RXR agonists, small molecules
Disrupt ApoE–Aβ interaction	Reduces Aβ aggregation and deposition	Aβ12–28P, small molecule inhibitors, ApoE-specific antibody
Conversion of ApoE4 to ApoE3	Increases ApoE3-associated protective functions and decreases ApoE4-related toxic effects	Small compounds (i.e., disulphonate and (monosulphoalkyl)
Restore ApoE functions	Increases ApoE protective functions and decreases neuroinflammation	ApoE-mimetic peptide
Blockade of ApoE fragmentation	Decreases tau pathology, and toxicity to mitochondria	Inhibitors for specific proteinases involved in ApoE fragmentation
Increase LRP1 and/or LDLR levels	Enhances Aβ clearance, cholesterol transport and synaptic plasticity	Small molecules
Increase apoE receptor 2 and/or VLDLR levels	Increases ApoE signalling and synaptic plasticity	Small molecules
Restore brain vascular integrity	Eliminates ApoE4-mediated blood–brain barrier breakdown and leakage of blood-derived toxic molecules	Cyclosporine A
Nonpharmacological approaches
APOE genotyping prior to immunotherapy	Helps to predict clinical outcome for Aβ-targeted or other therapies	Aβ immunotherapy might be more effective in APOE ε4 carriers or noncarriers
Preventive care	Maintains healthy brain vasculature function	Physical exercise, intellectual activities (e.g., puzzles), social connections (e.g., calling family and friends), healthy diet

Abbreviations: Aβ, amyloid-β; ABCA1, ATP-binding cassette transporter A1; AD, Alzheimer disease; ApoE, apolipoprotein E; LDLR, low-density lipoprotein receptor; LRP1, low-density lipoprotein receptor-related protein 1; LXR, liver X receptor; RXR, retinoid X receptor; VLDLR, very-low-density lipoprotein receptor.

APOE genotype and Aβ immunotherapy

Recent phase III clinical trials for immunotherapy have shown that bapineuzumab, an antibody that targets the N-terminus of Aβ, prevents Aβ deposition in the brains of APOE ε4 carriers with mild or moderate AD, but not noncarriers.^{159, 160} Bapineuzumab also lowers levels of phosphorylated tau in the CSF of both APOE ε4 carriers and noncarriers.^{159, 160} These reports suggest that Aβ immunotherapy is useful to eliminate Aβ from the brains of patients with AD and that its effect is likely to depend on ApoE isoforms. Major adverse effects of bapineuzumab—vasogenic cerebral oedema and microhaemorrhage—occur more frequently in APOE ε4 carriers than in noncarriers.^{159, 160} Although bapineuzumab failed to prevent cognitive and functional decline in these clinical trials, a combination of Aβ immunotherapy and an ApoE-targeted approach might lead to more effective therapeutic strategies.

Preventive of cognitive decline in APOE ε4 carriers

A prospective study of a cognitively normal cohort showed that risk of dementia in APOE ε4 carriers is negatively associated with high education, high level of leisure activities, and absence of vascular risk factors.¹⁶¹ A recent study demonstrated that physical exercise was strongly associated with reduced PiB-positivity in cognitively normal APOE ε4 carriers,³¹ indicating that a sedentary lifestyle in APOE ε4 carriers might increase the risk of amyloid deposition. Such studies indicate that high education, active leisure activities and exercise, and maintenance of vascular health could be beneficial in reducing the risk of AD and cognitive decline, particularly in APOE ε4 carriers.

Regulation of ApoE expression

ApoE levels in CSF and plasma tend to be lower in patients with AD than in healthy individuals, although such findings remain controversial.^{162, 163} Thus, increasing the expression of ApoE in all APOE genotypes may prevent or slow progression of AD through acceleration of Aβ metabolism and promotion of ApoE functions in lipid metabolism and synaptic support. Compounds that increase brain ApoE expression can be identified through comprehensive drug screening. Given that expression of ApoE is controlled by peroxisome proliferator-activated receptor-γ and LXRs, which form complexes with RXRs,^{100, 164} agonists or antagonists of these nuclear receptors are potential candidates as ApoE modulators. Indeed, recent work has demonstrated that oral administration of an RXR agonist, bexarotene, to an amyloid mouse model decreases Aβ plaque deposition and improves cognitive function in an ApoE-dependent manner.¹⁰⁰ The LXR agonist TO901317 also increases ApoE levels in the brain, facilitates clearance of Aβ₄₂, and reverses contextual memory deficit in amyloid mouse models.^{98, 99}

In addition to ApoE, LXRs also regulate ABCA1, which promotes cholesterol efflux.¹⁶⁵ Consequently, reduction of amyloid burden by the LXR agonist GW3965 depends on expression of ABCA1 in amyloid mouse models.¹⁶⁶ These results suggest that upregulation of lipidated ApoE might be necessary to maximize therapeutic effects in AD. These studies did not, however, assess the effect of increasing human ApoE3 or ApoE4 specifically. Because Aβ deposition is greater in APP-transgenic mice expressing mouse ApoE than in those expressing human ApoE isoforms,¹⁶⁷ further studies are needed to confirm the therapeutic effect of modulating the level of human ApoE isoforms. In addition, whether increasing ApoE4 is beneficial or harmful in AD brains remains unclear, and the effects might depend on age and disease status. Toxic functions associated with ApoE4 suggest that lowing ApoE4 expression might be beneficial in APOE ε4 carriers with cognitive decline during MCI and AD. Additional preclinical studies are needed to test potential beneficial or harmful effects of increasing or decreasing ApoE expression, particularly with regard to ApoE isoforms.

Blocking of ApoE-Aβ interaction

ApoE is required for deposition of Aβ fibrils in amyloid mouse models.¹⁰⁶ Recent studies have demonstrated that haploinsufficiency of human APOE results in significantly decreased amyloid plaque deposition in amyloid mouse models regardless of APOE isoform status.^{168, 169} Thus, disruption of the interaction between ApoE and Aβ might reduce Aβ aggregation and deposition, and should be considered as a therapeutic approach. Aβ interacts with ApoE through amino acid residues 12–28. A synthetic peptide mimicking this sequence, Aβ12–28P, reduces Aβ deposition and ameliorates memory deficits in amyloid mouse models.¹⁷⁰ Blocking the ApoE–Aβ interaction using Aβ-mimicking peptides could, therefore, be an effective approach for treatment of AD. Screening assays can also be used to identify compounds or ApoE-specific antibodies that block ApoE–Aβ interaction. These approaches should be assessed carefully because they could disrupt ApoE–lipid interactions and the associated beneficial functions of ApoE.

Other ApoE-based therapeutic approaches

ApoE4 is structurally different from ApoE2 and ApoE3 owing to different domain interactions,¹³¹ and this difference probably contributes to ApoE4 isoform-specific harmful effects. Modification of the structure of ApoE4 to form an ApoE3-like molecule might, therefore, be an interesting approach to ameliorate these harmful effects. Indeed, several molecules that bind to ApoE4 and interfere with domain interactions between N- and C-termini have been found. GIND–25 (disulphonate) and GIND–105 (monosulphoalkyl) are good candidates because they decrease Aβ production induced by ApoE4 to a similar level induced by ApoE3.¹³¹ CB9032258 (a phthalazinone analogue) and its derivatives disrupt ApoE4 domain interaction and restore functional activities of ApoE4 in neurons.¹⁷¹

An ApoE-mimetic peptide containing the receptor-binding region suppresses neuronal cell death and calcium influx associated with N-methyl-D-aspartate exposure in vitro.¹⁷² COG112, a chimeric peptide containing the receptor-binding region, is also reported to improve symptoms in mouse models of multiple sclerosis ¹⁷³ and sciatic nerve crush¹⁷⁴ through modulation of inflammatory responses. The effects of these peptides on AD pathogenesis are unknown, however, because they do not contain Aβ-interacting nor lipid-binding regions.¹³

ApoE receptors are also potential targets for AD therapy. For example, low-density lipoprotein receptor-related protein 1 and low-density lipoprotein receptor have crucial roles in brain lipid metabolism and Aβ clearance.^175–177 ApoE receptor 2 and very low-density lipoprotein receptor are essential for reelin signalling, which is important for neuronal migration during development and synaptic plasticity in adult brains.¹⁷⁸ Modulation of the expression of these ApoE receptors in AD brains might, therefore, restore lipid homeostasis and synaptic plasticity, and augment Aβ clearance.^{8, 178} Although ApoE-based therapies are still in the early stages of development, they offer great promises in the fight against AD. Clinical trials to further evaluate therapeutic potential of ApoE-based strategies are needed, with an eventual goal to develop curative and/or protective treatments for AD.

Conclusions

Work summarized in this Review highlights clinical evidence for the association between APOE ε4, AD and cognitive decline. Although the presence of APOE ε4 does not necessarily entail disease development, this genetic isoform probably accelerates the rate of disease conversion and progression. In particular, the effects of APOE ε4 on brain network connectivity, memory performance, and rate of cognitive decline are age-dependent in patients with AD and cognitively normal individuals. Thus, understanding the potential pathogenic link between APOE ε4 and cognitive function might allow for earlier identification of people at risk of developing AD. In combination with other putative AD biomarkers—such as MRI scans, PiB scans, and measurements of CSF Aβ and tau—APOE allele status could add predictive value to clinical diagnosis and evaluation of treatment efficacy.

Mechanistically, ApoE4 seems to increase risk of AD and cognitive decline through both Aβ-dependent and Aβ-independent pathways. ApoE isoforms differentially regulate Aβ production, aggregation and clearance. Independent of Aβ, ApoE4 might be less efficient than ApoE3 and ApoE2 in delivering cholesterol and essential lipids for the maintenance of synaptic integrity and plasticity. In addition, ApoE is a crucial regulator of the innate immune system, and in which ApoE4 promotes proinflammatory responses that could exacerbate AD pathogenesis. Finally, ApoE isoforms have differential roles in maintaining vascular health—roles that are crucial given that defects in vascular health are strongly associated with AD. Elucidating the contribution of ApoE4 to AD pathogenesis is a considerable challenge, but one that affords the potential to assist in combating AD.

Key Points.

• The ε4 allele of the apolipoprotein E (APOE) gene is the main genetic risk factor for Alzheimer disease (AD)

• APOE ε4 carriers have enhanced AD pathology, accelerated age-dependent cognitive decline and worse memory performance than do noncarriers.

• Numerous structural and functional brain changes associated with AD pathogenesis are detected in APOE ε4 carriers before clinical symptoms become evident.

• ApoE affects amyloid-β (Aβ) clearance, aggregation and deposition in an isoform-dependent manner.

• ApoE4 also contributes to AD pathogenesis by Aβ-independent mechanisms that involve synaptic plasticity, cholesterol homeostasis, neurovascular functions, and neuroinflammation.

• ApoE-targeted AD therapy should focus on restoration of the physiological function of ApoE through increased expression and lipidation, and inhibition of the detrimental effects of ApoE4.

Review criteria.

This Review was based on searches of the PubMed database using the following terms: “apolipoprotein E”, “cognitive decline”, “Alzheimer disease”, “amyloid beta”, “synaptic plasticity”, “cerebral amyloid angiopathy”, “mild cognitive impairment”, “cholesterol”, “brain activity”, “cerebrovascular diseases”, “brain metabolism”, “neurogenesis”, “brain atrophy”, “neuroinflammation”, “tau” and “traumatic brain injury”. Only articles published in English were retrieved. Full-text papers were available for most of the articles that were chosen for review, and the references of these articles were searched for further relevant material.

Biographies

Chia-Chen Liu is a Postdoctoral Research fellow in the Department of Neuroscience at Mayo Clinic College of Medicine, Jacksonville, FL, USA. She is also a visiting scientist in the Institute of Neuroscience at Xiamen University in Xiamen, China. She gained her Ph.D. from Washington University School of Medicine, St. Louis, MO, USA. Her current research interests focus on apolipoprotein E and its receptors in Alzheimer diseases and vascular dementia.

Takahisa Kanekiyo is an Instructor in the Department of Neuroscience at Mayo Clinic College of Medicine, Jacksonville, FL, USA. He gained his M.D. and Ph.D. from Osaka University Graduate School of Medicine, Japan, and engaged in postdoctoral training at Washington University School of Medicine, St. Louis, MO, USA. His research interests focus on apolipoprotein E and its receptors in neurovascular diseases including Alzheimer diseases and cerebral amyloid angiopathy.

Huaxi Xu is an adjunct Professor in the Institute of Neuroscience at Xiamen University in Xiamen, China. His research focuses on the molecular and cellular mechanisms underlying the pathogenesis of Alzheimer disease. His major contributions include dissecting the cellular trafficking pathways for amyloid precursor protein (APP) and presenilins, and identifying cellular factors that regulate APP processing and neuronal apoptosis. He serves as Co-Editor-in-Chief of Molecular Neurodegeneration.

Guojun Bu is a Professor of Neuroscience at Mayo Clinic College of Medicine, Jacksonville, FL, USA. He is also an adjunct Professor in the Institute of Neuroscience at Xiamen University in Xiamen, China. Since 1994, he has led a research programme studying the biological and pathological functions of Apolipoprotein E (ApoE) and ApoE receptors using cellular and animal models, with a specific focus on the pathogenesis of Alzheimer disease. His group has defined a critical role for the ApoE receptor lipoprotein receptor-related protein 1 in brain ApoE metabolism and in the clearance of amyloid peptides. He founded and serves as the Editor-in-Chief of Molecular Neurodegeneration.