The Impact of Apolipoprotein E (APOE) Epigenetics on Aging and Sporadic Alzheimer’s Disease

1. Introduction

The study of complex diseases is based on the association among epigenetics, gene variants, and environmental factors [1,2]. The pathophysiology of Alzheimer’s disease (AD) is a mixture of many pathogenic pathways and gene expression networks. The current model of AD is based on the amyloid-β (Aβ) hypothesis, in which a series of deterministic events leads from Aβ and tau deposition to neurodegeneration and progressive decline of cognitive function. This conceptualization matches autosomal-dominant AD, defined as dominantly inherited AD with pathological confirmation, although it is less appropriate for sporadic AD. A probabilistic AD model connoted by three variants of the disease has been proposed: autosomal-dominant AD, apolipoprotein E (apolipoprotein E gene, APOE) ε4 allele-related sporadic AD, and APOE ε4 allele-unrelated sporadic AD [3]. These three variants suggested a reduced weight of the Aβ hypothesis, giving more importance to environmental factors and lower-risk genes [3].

There were epigenetic modifications also in AD [4]. From a genetic point of view, the known risk loci showed a low penetrance in causing AD, except for Aβ production-related genes, and none of them have been related to different AD pathogenic pathways. On the contrary, epigenetic alterations may modify transcriptional activity globally throughout different genes and multiple biological pathways. Epigenetic mechanisms may also explain the influence of environmental stimuli such as dietary patterns, harmful exposures, and lifestyle factors on phenotypic outcomes in individuals with the same genetic variants [5]. Additionally, genetic sequence and epigenetic code were linked in a clear way. In fact, some single nucleotide polymorphisms (SNPs) are considered a common epigenetic mark because of the rearranging of cytosine-phosphate-guanine (CpG) dinucleotides with C nucleotide methylation. These CpG-altering SNPs may modulate DNA methylation levels in a cis or trans manner or they may modify gene transcription at regions enhanced of CpG known as CpG islands [6,7,8].

Since early 90′s, many studies have showed that APOE could play a central role in AD neurodegeneration. For sporadic AD APOE allele ε4 is a key genetic risk factor [9,10,11], with a semidominant inheritance [12], and associated to the ApoE4 isoform. Conversely, in sporadic AD, the APOE allele ε2, associated to the ApoE2 isoform, could have a protective effect [13,14]. Although the increased risk for sporadic AD in APOE ε4-carriers, the presence of the APOE ε4 allele alone is not a causal factor for AD pathology [15]. In this context, epigenetics may represent a candidate for a point of overlapping among several AD genetic risk factors, such as the APOE ε4 allele, and the AD pathophysiological processes. Human ApoE is a glycoprotein of 299-amino acids, traditionally binding phospholipids and cholesterol. ApoE is produced in 3 common isoforms (ApoE2, ApoE3, and ApoE4) differing in two amino acid residues at positions 112 and 158, and one very uncommon isoform (ApoE3r) [16].

The APOE variants, respectively ε² ε³, ε⁴ and ε³r, are determined by four haplotypes, derived from the allele association of 2 common SNPs rs429358 (C^3,937 → T) and rs7412 (C^4,075 → T) at the APOE locus (19q13.32), coding for the different protein isoforms [16]. These APOE four alleles [17], are considered the most investigated variants in human Caucasian genome. Remarkably, the APOE exon 4 region, encompassing the ε2/ε3/ε4 allele variants, is a well-defined CpG islands rich area. Moreover, the two common SNPs rs429358 and rs7412 are CpG-altering and modify the CpG content of this area. This APOE CpG islands rich area is a transcriptional enhancer with a specificity linked to the ε4 allele and cell-type [18]. In the present review article, we briefly summarized and highlighted the complex epigenetic regulation of APOE gene in aging and sporadic AD.

2. The role of apolipoprotein e in Alzheimer’s disease pathogenesis

For sporadic AD, APOE is the most important genetic risk factor as well as for the earlier stages of cognitive decline represented by mild cognitive impairment (MCI) [19], but its expression is poorly understood. Astrocytes and activated microglia produced the major amount of ApoE in the brain. Having one APOE ε4 allele conducts to a 4-fold increased risk of developing AD, while having two APOE ε4 alleles conducts to a 12-fold increased risk, compared to the APOE ε3-carriers. Conversely, the uncommon heterozygous carriers of the APOE ε2 allele have a risk 40% lower for AD and the homozygous carriers have a further reduced risk [20]. It was showed that APOE ε4-carriers with normal cognition displayed elevated Aβ and tau brain burden than APOE ε3-carriers; conversely, APOE ε2-carriers had reduced global Aβ burden, without differences in regional tau burden or accumulation over time [21]. The contribution in AD pathogenesis from APOE involves not only Aβ aggregation and its clearance, but also tau-mediated neurodegeneration [22], microglia impairment [23,24], astrocyte reactivity [25], and blood-brain barrier disruption [26,27].

The three ApoE isoforms bind and transport Aβ peptides with differential affinity during AD pathogenesis [28,29], being highest for ApoE4, intermediate for ApoE3, and lowest for ApoE2 [30,31]. Therefore, their effects are also different concerning Aβ aggregation and clearance, but not Aβ production [32,33]. ApoE also can affect tau-mediated neurodegeneration and tauopathy by modulating microglial responses to Aβ plaque pathology [34,35,36]. Thus, different ApoE isoforms may increase or reduce the risk for AD [29,31], based on different combined effects of ApoE isoforms on both Aβ deposition and neurofibrillary tangles [37]. APOE and its ε2/ε3/ε4 alleles have been connected by several genetic studies to multiple physiological conditions and disorders. Epigenetic alterations could explain the association between APOE and its associated diseases, considering that genetic signal associated with the disease also reflects a site’s sequence architecture for epigenetic code [18].

3. Apolipoprotein E, human longevity, and Alzheimer’s disease

There was a genetic association of APOE is with both human longevity and AD, but its mechanistic contribution in aging is largely under investigation. APOE pleiotropic roles may be explained by its exceptional epigenetic properties. In AD brain, these epigenetic changes could contribute to neural cell dysfunction. Additionally, several studies showed DNA methylation modifications on specific genes implicated in AD pathology such as APOE. In AD brain, it was showed that APOE CpG islands were differentially methylated in an APOE- and tissue-specific way [38]. In the brain of targeted replacement (TR) mice expressing human ApoE, allele variations within the major APOE CpG island may affect its methylation [39]. Epigenetic changes may link modified gene expression with environmental stimuli such as dietary patterns and physical exercise. In animal models, APOE alleles may have alterations in epigenetic regulation in response to external stimuli reported in studies on APOE TR mice [40].

The differences between mouse and human APOE gene clusters, the complexity of transcriptional control of human ApoE, and the structure of the targeting construct should be considered in the strategy for replacing mouse ApoE in the APOE TR models [41]. Moreover, lifestyle factors like education, alcohol consumption, smoking, and physical activity may attenuate genetic risk in the process of age-related cognitive decline and twelve modifiable risk factors might prevent or delay up to 40% of different dementias [42]. The complex interactions among age-related cognitive decline, genetics, and lifestyle may encourage behaviors maintaining cognitive health in older age [43]. At this regard, ApoE may be important for the pathophysiology of lipid metabolism [44] and central nervous system (CNS), although the role in healthy aging and longevity has seen its value growth [45,46,47].

Studies on longevity and healthy aging are related because subjects who live long tend to be healthy for a greater part of their lives [48]. Healthy aging can be defined as achieving older age maintaining intact cognition and/or mobility and without disabilities or multimorbidity. This last can be defined as the coexistence of two or more chronic diseases in the same subjects [49]. The detrimental effects of the APOE ε4 allele on longevity could influence the probability of a long human lifespan [48]. The APOE ε2 allele is more frequent in long-lived individuals than the ε4 allele [50]. Thus, the main longevity factor is the homozygous APOE ε3/ε3 genotype. The higher frequency of the ε3 allele in older individuals and their offspring than in controls derives from the greater amount of APOE ε3/ε3 genotype compared to the ε2/ε3 or ε3/ε4 genotypes [51].

In the pathophysiology of lipid metabolism, the role of ApoE may be related with normal/pathological aging, while its function in the pathophysiology of CNS needs further clarification [52]. In fact, in the CNS, there was about a quarter of total body cholesterol that may exert an important role in synaptic plasticity [53]. With advancing age, cholesterol metabolism may modify, and its related brain changes may be associated with the pathophysiology of AD [53]. So, in longevity and healthy aging, lipid and cholesterol maintenance are a critical factor also from an interventional point of view. The detrimental effects of APOE ε4 allele might be managed by dietary interventions [54], with a Mediterranean dietary pattern potentially including higher n-3 polyunsaturated fatty acid intakes [55,56].

4. Specific epigenetic modifications of apolipoprotein E in Alzheimer’s disease

In response to environmental stimuli, epigenetic marks and signals may enable temporal combination of regulatory events through mechanisms including DNA methylation, histone modification/chromatin conformation, and noncoding microRNAs (miRNAs). Several studies investigating DNA methylation in the APOE gene suggested an age-dependent flow and APOE DNA methylation specific for brain area. The APOE genomic sequence is approximately 4 kb in size (chromosome19:45408714-45412650, hg19) including its promoter. This region encompasses 172 CpG dinucleotides [57]. In the late 90s and early 2000s, polymorphic sites in the first intron and the proximal promoter the of APOE gene cluster (−1,019 to +407) affecting APOE expression have been identified [58,59,60,61,62,63,64] (Table 1). Notably, these polymorphisms have been related with a differential AD risk [65,66]. However, in AD, the association between these polymorphic sites and the variability of sequence in the proximal promoter with ApoE protein levels were not clearly understood. In fact, among different studies, findings on the levels of expression of APOE RNA and the relationship with the ApoE levels varied. In human postmortem brain, there was elevated methylation in AD frontal lobe of a 5′-C-phosphate-G-3′ (CpG) island overlapping with exon four and downstream [67]. Interestingly, APOE has a well-defined CpG island not residing in the promoter region and overlapping with the APOE 3′-exon. In the human genome, these 3′-CpG islands are very rare representing < 1% of total CpG islands and are also conserved in other mammals [68,69]. However, the APOE CpG island methylation level relates to the expression level of four known APOE transcripts. The majority of the total APOE mRNA, with higher expression in the AD frontal lobe than in the frontal lobe of control subjects, is constituted by circular RNAs, miRNAs, and truncated APOE transcripts. The findings of several studies suggested several changes in epigenome and the regulatory role of epigenomic elements associated with the risk or clinical presentation of different neurological diseases, although the exact clinical significance of these signatures in the quantities of RNA and methylation level of CGI in the APOE 3′-exon was still unclear [67] (Table 1).

At the level of the individual CpG site, epigenetic regulation was showed by up/down patterns in the methylation profiles between samples and tissues. Significant differences in the global methylation levels among several brain regions were discovered across postmortem brain tissues. In brain regions primary affected by AD such as frontal lobe, temporal lobe, and hippocampus, methylation levels were lower. Conversely, in the cerebellum, a region apparently lacking profound pathological changes in AD but with recent important findings, the highest methylation levels were observed, suggesting that a correlation may exist between the methylation levels of the APOE CpG islands and the vulnerability of brain AD regions [70]. In fact, age- and AD-related alterations in several cerebellar subregions may also impact numerous functional domains, especially those affecting cognitive processing [70].

Genetic variants, which consist of CpG-altering SNP, can modify DNA methylation levels. These genetic variations may act like regulatory elements connecting genetic changes with epigenetic variability [71] (Table 1). As previously described, the APOE ε2/ε3/ε4 alleles are produced by two CpG-altering SNPs (rs429358 and rs7412) residing within the core region of the APOE CpG islands. The APOE ε4 allele, if compared with ε2 or ε3 alleles, adds one more CpG, further saturating a small 12 bp region with 4 CpG sites. On the contrary, the APOE ε2 allele eliminates 1 CpG and opens a 33-bp CpG-free region. Therefore, these two SNPs may alter the regional CpG burden and probably influence global DNA methylation of the CpG islands. These CpG load changes might change the binding profiles of methyl CpG-binding domain proteins, connected specifically to methylated DNA through their exclusive amino acid pattern [72].

Furthermore, within the APOE CpG islands, there is evidence of indirect indicators of protein binding which consist of histone marks and a DNase I hypersensitivity cluster. These findings suggested that the APOE CpG islands (and exon 4) may be a site for chromatin remodeling and protein binding. Considering that environmental stimuli could influence DNA methylation gradually with aging, the differences in APOE CpG islands methylation between healthy individuals and AD increased with age [73]. Taken together, different methylation landscapes could be represented by inheritance of different ε2/ε3/ε4 alleles in the APOE CpG islands, which could accumulate or change continuously with age, also modified by environmental factors. Recent results showed that methylation levels for most CpG sites may be in the order of APOE ε4-carriers > APOE ε3/ε3-carriers > APOE ε2-carriers, considering that APOE e4-carriers have the greatest number of CpG sites, while APOE e2 carriers have the smallest number with ε3/ε3 in the middle [74] (Table 1). These changes could potentially alter protein binding, with some consequences on biological systems, even affecting the pathophysiological processes of multiple diseases and plasma lipids levels. APOE methylation could partially mediate the effects of age on plasma lipid (Figure 1).

In the epigenetic landscape, miRNAs are known to be small non-coding RNAs with a length of ~ 22 nucleotides They are also implicated in AD, as showed by the altered expression of miRNA 650 (miR-650) in AD brains [75]. Bioinformatic analysis showed that miR-650 may target the expression of three components associated to AD: APOE, presenilin 1 (PSEN1), and cyclin-dependent kinase 5 (CDK5), with recent findings confirming that miR-650 may reduce in vitro the expression of APOE, PSEN1, and CDK5 [75].

5. Epigenetics of apolipoprotein E and cognitive function: contrasting evidence in Alzheimer’s disease

Several lifestyle and environmental stimuli could explain the effects of APOE genotype on AD and cognitive functioning, such as exercise [76], education [77], and vitamin D status [78]. Among implications for the development and progression of AD, vitamin D supplementation may be another potential strategy to consider for the APOE ε4 allele-carriers. Some reports showed that higher vitamin D concentrations in APOE ε4 homozygous carriers allow to perform better at memory scores [79]. Then, compared to the APOE ε3/ε3-carriers, the APOE ε4-carriers showed earlier onset of cognitive impairment in AD. Although, after the disease onset, the effect of APOE genotype on the progression of cognitive impairment remained debated [80].

For this reason, epigenetic modifications of APOE such as DNA methylation have a central role in maintaining cognitive function in older age. Growing DNA methylation levels at the APOE promoter region were found on postmortem prefrontal cortex samples of sporadic AD individuals by mass spectrometry [81]. Numerous previous studies have investigated the association between APOE DNA methylation and AD or MCI [82,83,84]. Instead, the association between APOE DNA methylation and cognitive function in healthy subjects without cognitive impairment was evaluated by two studies with controversial findings [85,86]. Liu and colleagues found an inverse association between DNA methylation in the APOE gene region and delayed recall capacity among 289 older African Americans people with a mean age of 67 years during normal cognitive aging [85]. Conversely, the other study conducted in a large European cohort, observed no association between general cognitive functioning and APOE DNA methylation [86].

Many reports have suggested that neuroinflammation may have a key role in AD pathogenesis [87]. Dietary habits are known to influence systemic inflammation, neuroinflammation, and inflammaging [88]. A recent study conducted in a cohort of racially diverse middle-aged people (n = 411), pursued to identify DNA methylation sites associated with cognitive function in the genomic region of APOE. About inflammatory potential of the diet, among the dietary inflammatory index, cognitive performance, and the methylation level of several CpG sites have been detected significant relationships [89].

However, studies are contrasting at this regard, and if epigenetic biomarkers could be used for predicting AD is still unclear. In the APOE gene, DNA methylation at two CpG sites (3/13) that are known to show age-dependent changes, was related with the total cholesterol and high-density lipoprotein cholesterol ratio, but not with cognitive status, family history of AD, or the risk of cardiovascular disease in a blood-based DNA methylation study of 5828 people from the Generation Scotland cohort [90]. These findings supported that there is no evidence yet for considering APOE methylation as a biomarker for predicting AD or cardiovascular disease, although APOE methylation was associated with the blood levels of cholesterol [90].

6. Conclusions

In the panorama of current available evidence, the investigation of healthy aging and longevity is currently of remarkable interest. APOE could be considered an epigenetic mediator of senescence considering that different ApoE biochemical pathways in lipid metabolism, neuroinflammation and neurodegeneration may contribute to longevity and healthy aging. Nonetheless, such areas of investigation are still increasing, since ApoE function in neurodegenerative diseases such as AD cannot be uniquely explained by ApoE effects in lipid metabolism. Furthermore, the imbalance in the ApoE isoforms could explain the pathophysiological process of cognitive impairment linked to sporadic AD [91].

Stochastic factors (such environmental, diet, and pollution) may play a significant role in sporadic AD, despite the elevated lifetime risk linked to APOE ε3/ε4 and APOE ε4/ε4 genotypes. Indeed, according to the notion of stochastic risk or protective factors and although it is known that APOE ε4/ε4-carriers developed dementia about 10 years earlier than APOE ε2 carriers [92], there was still significant discrepancy in the age of onset for APOE ε4/ε4-carriers (standard deviation of 6 years) [93]. During the process of aging, the accumulation of molecular alterations driven by genetic and epigenetic events in the organism lead to a loss of phenotypic plasticity over time. Also, epigenetics may be altered during the process of aging, and it is particularly important as age is the greatest risk factor for developing AD [94].