Soluble apoE/Aβ complex: mechanism and therapeutic target for APOE4-induced AD risk

As soluble Aβ is considered a major neurotoxin in AD, the apparent correlation between APOE4 and increased levels of particularly soluble oAβ suggests that understanding the underlying pathway(s) that mediates this effect may be critical to understanding the APOE4-induced risk for AD. In turn, this fundamental biology could inform rational drug design and development of successful AD therapeutics.

ApoE isoform-specific effects on apoE/Aβ complex levels may mediate the increased soluble Aβ and oAβ levels that correlate with APOE4. Research efforts to determine the effect of apoE isoform on exogenous complex formation (Table 1) or endogenous apoE/Aβ complex levels (Table 2) have been ongoing for the last ~20 years (see Table 1). These studies demonstrate that the method of detection and the source of the components for apoE/Aβ complex are critical parameters for experimental outcomes.

As we seek to define the role of apoE/Aβ complex in the parenchyma of the brain, extracting intact lipoproteins from tissue homogenates is critically important but procedurally problematic. For now, the optimal CNS samples are soluble brain homogenates or CSF. In addition, as discussed in the previous section, multiple factors are known to influence the detection of apoE/Aβ complex levels. One technique is not necessarily superior to another; each has merits and limitations. Thus, future studies may utilize multiple, complementary techniques for sample analysis. Overall, based on the literature and our recent data demonstrating an inverse relationship between the levels of soluble apoE/Aβ complex and oAβ, we hypothesize that complex plays a significant role in modulating oAβ levels. The lower levels and instability of apoE4/Aβ complex compared to apoE3/Aβ suggests a potential mechanism for the APOE4-induced risk for AD.

The general dogma in the field is that apoE4 in the CNS is less lipidated than apoE3. However, it is difficult to isolate and analyze intact CNS lipoproteins, thus limiting direct study of the lipidation state of apoE4-particles versus apoE3-particles. However, the functional consequences of conditions predicted to affect the lipidation state of lipoproteins have provided valuable insights. For example, in vitro data demonstrate that glia-mediated degradation of apoE is increased and cholesterol release is reduced in primary glial cultures expressing apoE4 compared to apoE3 [77, 78]. In vivo study of Tg mice using over-expression or knock-out of the ABC transporters has also provided important information on the functional effects of the lipidation state of CNS lipoproteins [79‐82]. For example, Fitz and co-workers demonstrated that introducing ABCA1^-/+ to an FAD/apoE-Tg mouse results in increased plaque deposition and a general AD phenotype in mice expressing apoE4 but not apoE3 [82]. In addition, we recently developed a three-step sequential protein extraction protocol (TBS, TBSX, formic acid/FA). A non-ionic detergent Triton X-100 (TBSX) is used to in an attempt to release apoE from lipoprotein particles without inducing the formation of new micelles, as can occur with ionic detergents such as SDS [12, 83]. Using EFAD mouse brain extraction fractions, we demonstrated that while total brain apoE4 levels are lower compared to apoE2 and apoE3, this decrease is seen only in the TBSX fraction [12]. These data provide evidence that less apoE4 is associated with lipoproteins and thus apoE4 may be less lipidated than apoE2 and apoE3.

Identifying whether apoE4-containing lipoproteins are less lipidated than apoE3- containing lipoproteins, and which lipids differ between the two, are important considerations. If, as hypothesized, apoE4-containing lipoproteins are less lipidated than apoE3 containing lipoproteins, this may explain the reduced levels and stability of the apoE4/Aβ complex.

The Aβ clearance rate is slower with apoE4 compared to apoE3 in PDAPP/apoE-TR mice [87] and in apoE-TR mice after a bolus brain injection of Aβ [88]. ApoE exhibits an isoform-specific effect on a number of cellular processes that modulate soluble Aβ clearance, and the mechanisms underlying these effects may be influenced by apoE/Aβ complex levels. Compared to apoE3, apoE4 results in a decrease in Aβ levels via: 1) clearance via glia [89‐93], neurons [94‐96], and the blood-brain barrier [88, 97]; 2) intracellular and parenchymal enzymatic degradation [93]; and 3) drainage via the interstitial fluid (ISF) [87] and perivasculature [98]. One potential mechanism is that soluble apoE3/Aβ complex may reduce soluble Aβ levels via clearance, whereas the lower levels of soluble apoE4/Aβ complex result in higher soluble Aβ levels, particularly oAβ. In terms of cellular clearance, the literature appears contradictory, likely due to the variety of models and reagents used, including; glia versus neurons, mouse versus human cells, mouse versus human apoE, source of apoE (purified protein or lipidated particle), source and aggregation state of the Aβ, etc. In a particularly relevant study, Nielson and co-workers used primary human astrocytes to address the role of the human apoE isoforms in binding and uptake of Aβ42. Their results demonstrated that both apoE3 and apoE4 inhibited the uptake of oAβ42 but not fibrillar forms of the peptide [99].

Data are conflicting as to the effect of apoE/Aβ complex on Aβ aggregation. Previous publications have demonstrated that apoE promotes Aβ aggregation, resulting in retention of the peptide; within glia [99], as intraneuronal Aβ [100], or in the parenchyma as either diffuse Aβ deposits or amyloid [12]. APOE4 is associated with an increase in amyloid deposition, suggesting that apoE functions as a "pathological chaperone", an effect that might actually be beneficial [101]. However, as demonstrated in vitro, apoE isolated from human brain increases the oligomerization of Aβ [27], an effect that is greater with apoE4 compared to apoE3. This is consistent with the hypothesis that apoE binds to oAβ and prevents further aggregation [64]. These results would favor the view that lower levels of apoE/Aβ complex are beneficial [102, 103].