Both neurons and astrocytes synthesize cholesterol in situ
, but differ considerably in terms of its metabolism, in particular, regarding the mechanisms of its synthesis, maintaining of its homeostasis, as well as the brain cholesterol efflux and influx (Björkhem et al.
2010). Single peptide modulators, as well as complex systems of receptors are involved in regulation of cholesterol transport. The first group includes apolipoprotein E and ATP-binding cassette transporter, while the second includes liver X receptors and LDL receptors. The coordinated cooperation between them ensures the maintenance of cholesterol homeostasis in the brain.
Apolipoprotein E
Apolipoprotein E (APOE) is a lipid carrier of cholesterol, involved in its transport in both physiological and pathological conditions (Mahley and Rall
2000). APOE is a 299 amino acid apolipoprotein, encoded by the
APOE gene, which is highly expressed in many human organs particularly on the liver and the brain (Mahley and Rall
2000). In physiological state APOE is highly expressed in the astrocytes, but not in the neurons. However, in the brain pathological processes such as excitotoxic injury, the neurones can also express APOE (Xu et al.
1999). The astrocytes synthesize APOE and secret it to the extracellular space in the form of APOE-containing lipoproteins and phospholipids, that transport cholesterol, which is then absorbed by neurons during receptor-mediated endocytosis (Fig.
2). The cholesterol taken up by the neurons can be further converted into the 24S–OHC, in order to eliminate it from the brain (Hayashi et al.
2004). Alternatively, it can be utilized for the formation of new cellular membranes and microdomains in the form of lipid rafts (Lahiri
2004).
The APOE is stable only when it is bound to the lipids, whereas in the unbound state it is degraded and its level drastically decreases (Hirsch-Reinshagen et al.
2004). It has been previously reported that APOE level is reduced in
abca1-gene-knockout mice (Wahrle et al.
2008). The product of this gene is necessary for the APOE lipidation. Studies have shown that mice with a reduced level of APOE do not display significant changes in cholesterol levels, whereas in mice lacking APOE there is observed a reduction in both cholesterol and its precursors compared to control animals (Levi et al.
2005; Jansen et al.
2009). It is not clear whether changes in the cholesterol level result directly from the APOE deficit. Interestingly, a reduced level of APOE does not influence the cholesterol homeostasis, which could indicate an existence of alternative regulatory mechanism (Anderson et al.
1998). Studies showed that the APOE deficit during brain development coincided with abnormal synaptic and dendritic densities in the hippocampus, but without significant functional consequences. In the adult mice, however, it has been accompanied by serious behavioural changes and neurological symptoms (Anderson et al.
1998). A relationship between the level of the APOE in brain and development of AD has been suggested by some authors (Corder et al.
1993; Castellano et al.
2011). However, the exact mechanism remains unknown.
In the human brain APOE is expressed in three isoforms: ApoE2, ApoE3 and ApoE4 (Corder et al.
1993). Whereas in the mouse brain, ApoE3 and ApoE4 are expressed by astrocytes both during development and in adult mice (Sun et al.
1998). It seems intriguing that the cellular source of APOE can have a decisive influence on regulating neurite overgrowth. The results obtained by the Holtzmans group show that apolipoproteins containing ApoE3 have a different neurobiological activity than ApoE4-containing lipoproteins. It is evident that hippocampal neurons are defined by increased neurite growth in the presence of ApoE3-secreting astrocytes compared to ApoE4-secreting astrocytes or ApoE knock-out astrocytes in primary cultures (Sun et al.
1998).
Recent studies show the role of the APOE genotype in modulating the astrocyte phagocytic capacity and in controlling the rate of synapse pruning and turnover by astrocytes in vitro and in vivo (Chung et al.
2016). ApoE2 allel increases the rate of synaptic phagocytosis by astrocytes, while Apo4 prevents it, which in consequence leads to excessive accumulation of senescent synapses. Moreover, recent research shows that the APOE allele also affect the amount of C1q protein accumulated in the hippocampus, which may be the determinant of the accumulation of senescent synapses (Chung et al.
2016). ApoE2 knock-in animals are characterized by a reduction in C1q protein accumulation, while in ApoE4 animals the level is significantly increased relative to ApoE3 animals. Therefore, the regulation of the phagocytosis rate, conditioned by APOE isoforms, occurs by modulating one or more phagocytic pathways (Chung et al.
2016).
Corder et al. showed that presence of ApoE2 was related to reduced risk of AD (Corder et al.
1993). The ApoE4 increases the risk of AD and its presence correlates with the lower age of its onset. This protective relationship between ApoE4 and AD may be related to the partially defective phagocytic capacity of astrocytes (Chung et al.
2016). Interestingly, there is evidence that APOE binds to Aβ and plays role in β-amyloid plaques formation (Vance and Hayashi
2010). In addition, various APOE isoforms have different capacity for binding with Aβ, which seems to be related with different lipidation capacity. The APOE2 and APOE3 isoforms become lipidated more easily than ApoE4, so it seems that the APOE lipidation can be an important process involved in AD (Castellano et al.
2011).
ATP-binding cassette (ABC) transporters
ATP-binding cassette (ABC) transporters are key regulators of lipid homeostasis in the brain (Hirsch-Reinshagen et al.
2004). The ABC transporters, that regulate cholesterol homeostasis, belong to two families: ABCA and ABCG. They are responsible for lipid transport across the cell membrane (Oram and Heinecke
2005). Both the astrocytes and the neurons express: ABCA1, ABCG1 and ABCG4 transporters. The ABCA1 transporter has a broad substrate specificity which enables it to transport cholesterol and phospholipids (Oram and Heinecke
2005). Both ABCA1 and ABCG1 are involved in the cholesterol efflux from astrocytes (Fig.
2). Whereas silencing or increasing expression of ABCA1 and ABCG1 has no effect on the cholesterol efflux from the neurons, it reduces and increases cholesterol efflux from the glial cells, respectively.
ABCA1 is an important factor maintaining the proper level of APOE and thus cholesterol homeostasis in the brain (Hirsch-Reinshagen et al.
2004). Deficiency in this transporter may result in poor lipidation and degradation of APOE and significant reduction of cholesterol levels (Hirsch-Reinshagen et al.
2004). In brain, the low expression of ABCA1 increases accumulation of Aβ, whereas its high expression reduces the amyloid deposition in the murine model brain of AD (Wahrle et al.
2005,
2008). The levels of ABCA1, ABCG1 and APOE may be increased by the 24S–OHC of neuronal origin, which indirectly regulates the cholesterol efflux from the glial cells (Abildayeva et al.
2006).
Kim et al. showed that the expression of ABCG4 was much higher in neurons than in glial cells (Kim et al.
2006). Consequently, ABCG4 is involved in the cholesterol efflux from neurons, but not from astrocytes (Chen et al.
2013). Thus, the increase in expression of this transporter activates the efflux of cholesterol only from the neurons, not from glial cells.
LDL receptors
There are two functionally important LDL receptors: the prototypic low-density lipoprotein receptor (LDLR) and the low-density lipoprotein receptor-related protein 1 (LRP1) that are involved in the cholesterol transport in the human brain (Fig.
2) (Herz
2009). Although both of them are expressed in astrocytes and neurons, the LDLR is highly expressed in the astrocytes, whereas LRP1 is expressed mainly in the neurons (William Rebeck et al.
1993). The LDL receptors play an important role in lipid homeostasis and intracellular signalling (Pfrieger and Ungerer
2011). They bind different ligands, including apoE-containing lipoproteins, and regulate the integrity of the BBB, what makes them important factors regulating cholesterol efflux from the cells (Strickland et al.
2014). Liu et al. showed increased level of APOE in LDL-receptor knockout mice, which resulted in decrease of the cholesterol levels in the brain (Liu et al.
2010b). Due to the involvement of the LDL receptors in the clearance of Aβ associated with BBB permeability, it was suggested that these modulating agents could be potentially considered for treatment of AD (Martiskainen et al.
2013).
Liver X receptors
Liver X type α and β receptors (LXRs) belong to a family of nuclear receptors involved in regulation of cholesterol metabolism. Furthermore, stimulation of LXRs is responsible for astrocytic recruitment to the inflammatory response, in the course of pathological processes (Edwards et al.
2002). Previous studies have shown that LXRs inhibit cholesterol synthesis, induce expression of the ABC transporters and enhance the APOE lipidation (Abildayeva et al.
2006). Consequently, in vitro studies show that the cholesterol efflux is stimulated, which results in the exhaustion of cholesterol pool in the astrocytic culture. In addition, LXRs activity in the astrocytes can be regulated by the level of 24S–OHC released by the neurons, which augments the effect of cholesterol efflux (Abildayeva et al.
2006).
Deficiency of either LXRα or LXRβ in the APP/PS1 transgenic mouse model of AD increases the deposition of Aβ in the brain, through negative regulation of microglial phagocytosis, resulting in dramatic increase in lipid accumulation and development of neurodegenerative disorders (Zelcer et al.
2007). The LXRs inhibit the processing of amyloid precursor protein (APP) by modulating membrane cholesterol levels via ABCA transporters and accelerate Aβ clearance by inducing APOE lipoprotein secretion (Riddell et al.
2007). Due to the ability to modulate metabolic factors involved in the pathogenesis of AD and to regulate the inflammatory response by inhibiting glial cells activity, the LXRs seem to be an attractive therapeutic target for neurodegenerative diseases (Zelcer et al.
2007).