Lipoproteins in the Central Nervous System: From Biology to Pathobiology

2.1.1. Types of lipoproteins in the periphery.

Most current knowledge of lipoproteins comes from studies in the periphery, due to their involvement in diseases such as atherosclerosis and other cardiovascular diseases (19). There are four major classes of lipoproteins that differ in size, composition, and density, each having a distinct function. Chylomicrons are the largest in size, with a diameter greater than 100 nm, followed by very-low-density lipoproteins (VLDL), with diameters 30–90 nm; intermediate low-density lipoproteins (IDL) and low-density lipoproteins (LDL) are slightly smaller, with diameters of approximately 20 nm; and high-density lipoproteins (HDL) are the smallest, with diameters of 8–12 nm (19) (Figure 1).

2.1.2. Chylomicron metabolism in plasma.

Plasma lipoprotein function is tightly linked to dietary lipid processing, from their absorption to transport to target sites and their metabolism (2). Dietary triglycerides and cholesteryl esters absorbed in the intestine are assembled into chylomicron particles containing various apolipoproteins, including apoB and apoE (19, 20). Hydrolysis of triglycerides in chylomicrons generates fatty acids that either serve as fuel for different cell types or are stored as triglycerides in adipocytes. Chylomicron remnants are taken up by hepatic cells through binding to apoE receptors, members of the LDL receptor (LDLR) family, where they are further metabolized (19). Lipid transport from the liver to peripheral tissues is initially mediated by VLDL containing both apoE and apoB. These lipoproteins are rich in triglycerides, but gradual hydrolysis by lipoprotein lipase generates smaller, cholesterol-rich IDL and ultimately LDL. The apolipoprotein composition of lipoproteins in plasma changes during their metabolism (21).

2.1.3. Reverse cholesterol transport.

The main function of HDL is to clear excess cholesterol from peripheral tissues for elimination by the liver through a process known as reverse cholesterol transport (19). While VLDL, IDL, and LDL originate from the sequential catabolism of chylomicrons, HDL formation follows a different pathway that starts in hepatocytes and intestinal enterocytes that synthesize apoA1 (2). Cholesterol is incorporated into HDL discs composed of apoA1 and phospholipids by ATP-binding cassette transporter A1 and G1 (ABCA1/ABCG1), forming mature HDL (3). HDL particles are commonly divided into two major subfractions, HDL₂ (1.063–1.125 g/ml) and HDL₃ (1.125–1.21 g/ml), that differ in apolipoprotein composition and metabolic function (22). HDL₂ represents the majority (~85%) of HDL and can be subdivided into apoE-containing and apoE-lacking lipoproteins. HDL particles act as a reservoir for apoE, which can be transferred to chylomicrons and VLDL (2, 22).

2.1.4. Lipid delivery in the periphery.

Lipid uptake into hepatic and extrahepatic cells is mediated by interactions between lipoprotein particles containing apolipoproteins such as apoB or apoE and cell surface receptors, including LDLR family members, scavenger receptors, and heparan sulfate proteoglycans (HSPG) (21, 23). The LDLR family contains many receptors such as LDLR, LDLR-related protein 1 (LRP1), VLDL receptor (VLDLR) and apoE receptor 2 (apoER2) (Figure 2) (24).

3.1.1. Discovery of apoE in the periphery.

ApoE was first discovered in the context of human hyperlipidemia, a group of diseases characterized by the accumulation of abnormally cholesterol-rich VLDL in the plasma (68). These cholesterol-rich particles were identified as chylomicron remnants containing high levels of an arginine-rich protein (68) that had previously been described as a constituent of VLDL (69) and later designated as apoE (70). ApoE was quickly found to be heterogeneous following analysis of VLDL in familial hyperlipoproteinemia type III (HLP-III), a disease associated with premature atherosclerosis. The existence of three APOE gene alleles, ε2, ε3 and ε4, results in six apoE genotypes: APOE2/2, APOE2/3, APOE2/4, APOE3/3, APOE3/4, and APOE4/4 (71). The APOE gene was later identified and mapped to chromosome 19 (72). APOE encodes a 317-amino-acid protein that includes an 18-residue signal peptide (73). Cleavage of the signal sequence and subsequent glycosylation result in mature apoE, a 299-amino-acid glycoprotein of approximately 34 kDa (21).

The frequency of APOE alleles in the general population differs by ethnicity. Nevertheless, APOE3 is the most common allele, followed by APOE4 and then APOE2 (74). A summary of the APOE allele frequencies for African Americans and European Americans is shown in Table 1.

3.1.2. ApoE isoforms differ at two amino acid residues in their primary sequence.

Each apoE isoform, apoE2, apoE3, and apoE4, displays a unique isoelectric focusing point between 5.4 and 6.1. ApoE2 corresponds to the most acidic isoform, followed by apoE3; apoE4 is the most basic isoform (75). This biochemical difference is caused by a Cys-to-Arg interchange at positions 112 and 158. ApoE2 harbors two Cys residues, apoE3 contains Cys-112 and Arg-158, and apoE4 has two Arg residues (76) (Figure 5). In addition to the major apoE isoforms, several rare apoE variants have also been detected in humans. Of particular interest are two mutations identified in apoE3 that protect against AD. ApoE3-Christchurch (apoE3-Ch), corresponding to an Arg-to-Ser mutation at position 136, was recently associated with resistance against autosomal dominant AD in a homozygote carrier (77). ApoE3-Jacksonville, corresponding to a Val-to-Glu mutation at position 236, was linked to a reduced risk of AD (78, 79) (Figure 5). Additional apoE mutations, such as apoE4-L28P and apoE3-R145C, have been associated with lipid disorders and heart disease (80, 81).

There is no X-ray crystal structure of full-length apoE to date due to its highly aggregative nature. Nevertheless, apoE has been characterized at the secondary, tertiary, and quaternary structural levels using different biophysical techniques. ApoE is rich in α-helices in both its lipidated and nonlipidated forms (82, 83). Crystallization of the N-terminal domain of apoE3 isolated from plasma suggested that it is composed of five α-helices, four of which are arranged in an antiparallel, elongated four-helix bundle (84); this result was supported by nuclear magnetic resonance (NMR) spectroscopy (85). ApoE is formed of two independently folded domains separated by a hinge region (residues 168–205) that can undergo thrombin cleavage. The N-terminal domain (residues 1–167) corresponds to a 22-kDa fragment, and the C-terminal region (residues 206–299) to a 10-kDa peptide (85, 86). Analysis of H-bonds and salt bridges identified domain–domain interactions, particularly between the N-terminal domain and the C-terminal domain of apoE3 (85). Sedimentation equilibrium experiments performed on delipidated apoE isolated from human plasma showed that apoE forms tetramers in aqueous solution, with tetramerization of apoE mediated by the 10-kDa C-terminal fragment (83, 87).

3.1.3. ApoE production in the brain.

Next to the liver, the brain is the second largest production site for apoE (33). Immunocytochemical staining of apoE in rat brains initially ascribed apoE production solely to astrocytes (24). Further studies have shown that apoE is also highly expressed by the choroid plexus, which secretes CSF (88), and in microglia (89), particularly in reactive microglia (88). ApoE expression was also detected in pericytes and oligodendrocytes, albeit to a lesser extent (90, 91). Neurons have also been reported to produce apoE in response to injury (88). Levels of CNS apoE are in part modulated by ABCA1, likely because of its effect on apoE lipidation status, with poorly lipidated apoE more likely to be catabolized (40).

Plasma levels of apoE in humans are isoform dependent (apoE2 > apoE3 > apoE4) (92). ApoE levels in the CNS also appear to be isoform dependent, following the same trend as in plasma (93-95). For instance, in apoE-targeted replacement (TR) mice, in which the Apoe murine gene was replaced by the human gene, apoE2-TR mice have higher levels of apoE, followed by apoE3-TR and then apoE4-TR mice, in brain parenchyma, CSF, and ISF (94, 95). The impact of APOE genotype on apoE abundance in human brain is less evident. In vitro human cell models such as induced-pluripotent stem cells (iPSC)-derived astrocytes and cerebral organoids have shown that APOE4 is associated with lower levels of cellular and secreted apoE when compared to APOE3 (93). However, when assessing apoE abundance in human CSF, results varied with the quantification method used. Levels of CSF apoE were shown to be genotype dependent when quantified by enzyme-linked immunosorbent assay (ELISA), where APOE2 was associated with higher, and APOE4 with lower, apoE levels compared to APOE3 (96). This observation was not supported by quantification using mass spectrometry, as no isoform-dependent difference was detected (97). Measurements of apoE levels in human postmortem cortex by ELISA were consistent with iPSC-based in vitro studies (APOE2 > APOE3 > APOE4) (98).

3.1.4. ApoE-containing lipoproteins in the central nervous system are isoform and cell type dependent.

The cellular origin of apoE-containing lipoprotein particles affects their structure and composition. For instance, microglia-produced apoE-containing lipoproteins are poorly lipidated compared to apoE particles secreted by astrocytes isolated from human apoE knockin (KI) mice (89). Whereas astrocytic apoE particles are approximately 15 nm in diameter (38), microglial apoE particles are approximately 8 nm (89). Importantly, apoE is significantly upregulated upon microglial activation (99, 100). This suggests that microglial apoE is more involved in responding to pathogenic events, whereas astrocytic apoE is important for maintaining brain homeostasis.

The nature of apoE-containing lipoprotein particles is also significantly impacted by apoE isoforms in the CNS. APOE2 carriers displayed larger apoE-containing lipoprotein particles compared to APOE3 homozygotes, and APOE4 was associated with smaller particles (101). Moreover, fractionation of CSF samples collected from APOE2/3, APOE3/3, and APOE3/4 individuals showed that apoE2 and apoE3 were associated with formation of denser particles than apoE4 in specific fractions. ApoE2-containing particles are richer in total cholesterol content than particles isolated from E3/E3 individuals, and apoE4-containing particles are poorer. ApoE4 particles also have the lowest phospholipid content (102).

ApoE2 hyperlipidation compared to apoE3 and apoE4 has been demonstrated in mice expressing human apoE isoforms and in in vitro models (103). In a study by Hu et al. (104), adeno-associated virus–mediated overexpression of apoE2 in astrocytes of apoE4-TR mice led to increased size and lipidation of apoE-containing particles, with higher cholesterol content. The opposite was observed with overexpression of apoE4 in apoE4-TR mice. ApoE secreted by primary astrocytes and microglia derived from human APOE KI mice also follows this pattern, with apoE2 producing larger lipoproteins (89). An isoform-dependent effect on lipoprotein size was also observed in iPSC-derived astrocytes, with apoE3 overall forming larger and more cholesterol-rich lipoprotein particles than apoE4 (105).

3.1.5. Structural differences between apoE isoforms result in distinct lipid-binding preferences.

Unilamellar liposome turbidity clearance kinetics by recombinant apoE produced in Escherichia coli suggest that apoE tetramer dissociation into monomers is required prior to binding lipids. As such, differential self-association propensity and stability of apoE isoforms impact their lipid-binding ability (106). In vitro, recombinant, nonlipidated apoE4 has a higher tendency to form aggregates compared to apoE2 and apoE3 (82). Moreover, lipidation protects apoE from forming high molecular weight oligomers (107).

Lipid binding is likely associated with additional structural rearrangements of apoE, as suggested by the NMR structure of apoE3. Hydrophobic residues in the C-terminal region exposed at its surface attract lipids. Lipid binding may result in disruption of interactions between the N- and C-terminal domains (85, 108). A mechanism for lipid binding to apoE was proposed based on experimental and computational observations: A charged lipid molecule would disrupt domain–domain interactions by interfering with salt bridges, resulting in further opening of the gap as the N- and C-terminal domains separate, allowing binding to lipoprotein receptors through the N-terminal domain (108).

3.1.6. Glycosylation and its effects on lipidation.

ApoE is a glycoprotein, with posttranslational O-glycosylation through galactosylation and subsequent sialylation (109, 110). Thr-194 in the hinge region of apoE was the first residue identified as an O-glycosylation site (110). Mass spectrometry identified additional major and minor glycosylation sites in both the N-terminal (Thr-8, Thr-18) and C-terminal (Thr-289, Ser-290) domains (111, 112) (Figure 5). The extent of apoE glycosylation and its glycosylation patterns are dependent on its source. ApoE glycoproteomic analyses showed CSF apoE to be more heavily glycosylated than plasma apoE. Furthermore, CSF apoE was more sialylated than plasma apoE in its C-terminal domain, whereas the N-terminal domain of plasma apoE was more sialylated compared to its CSF counterpart (111).

Posttranslational modifications modulate biological functions in proteins. Examination of CSF-derived apoE-lipoprotein particles showed an inverse correlation between the extent of apoE sialylation and total lipid concentrations. This suggests that heavier sialylation of apoE results in lipid-poor particle formation (113). How apoE glycosylation patterns affect its lipidation and whether there are isoform-dependent effects on glycosylation and related functions require more in-depth studies to determine.

3.1.7. Isoform-dependent and lipidation-dependent receptor binding.

The receptor-binding domain of apoE is contained within a region rich in basic amino acids in the N terminus of the protein that spans residues 136–150 (84, 114). Binding of apoE to receptors is isoform and lipidation dependent (103). Lipidation of apoE has been proposed to enhance binding to LDLR due to exposure of the receptor-binding region (108). However, lipidation of apoE does not seem to be necessary for binding to other receptors such as LRP1, VLDLR, or HSPGs (103).

Defective binding of apoE2 to LDLR is linked to familial HLP-III. This deficiency is due to Cys-158 in apoE2; cysteamine treatment resulting in the generation of a Lys analogue enhanced the affinity of apoE2 for LDLR in fibroblasts to levels similar to apoE3 and apoE4 (115). Residue 136 also seems to affect apoE binding to LDLR, with the Arg-to-Ser mutation in apoE3-Ch resulting in a 60% reduction in affinity for LDLR, consistent with the involvement of basic amino acids in receptor binding (114). Binding of lipidated apoE2 to LRP1 in vitro is also impaired when compared to apoE3 and apoE4 (40% of apoE3 binding) (116). Binding to heparin and HSPGs is mediated by amino acids 142–147 in the N-terminal domain and residues around Lys-233 in the C terminus (117). The affinity of apoE for heparin is also isoform dependent; Cys-158 in apoE2 reduces its affinity for heparin by 40% compared to apoE3 (118), whereas apoE4 has higher affinity. Additionally, Ser-136 in apoE3-Ch recombinantly produced in bacteria impairs heparin binding to 10% of apoE3 binding (77). ApoE is also a ligand for TREM2, but binding appears to be isoform independent (119).

3.1.8. ApoE isoform-dependent lipid efflux, lipid droplets, and lipid metabolism.

Accumulating evidence demonstrates a role for apoE in promoting lipid efflux. In rat primary astrocytes and neuronal cultures, apoE-mediated lipid efflux appears to be isoform dependent. Addition of recombinant apoE led to cholesterol and phosphatidylcholine efflux to a greater extent in the presence of apoE2, followed by apoE3 and then apoE4 (120). ApoE3 dimerization through disulfide bridge formation, which apoE4 cannot achieve without a Cys residue, appeared to further enhance lipid efflux (121).

Lipid efflux capacity is also isoform dependent. APOE4 iPSC-derived astrocytes accumulate cholesterol intracellularly while concomitantly increasing de novo cholesterol synthesis and secretion compared to APOE3 astrocytes (93). APOE4 in astrocytes is also associated with increased formation and accumulation of lipid droplets compared to APOE3 astrocytes. Lipid droplets are intracellular, lipid-rich organelles that sequester cytotoxic fatty acids and serve as a reservoir for cholesterol by entrapping it as cholesteryl esters. Lipid droplets have been shown to over accumulate during aging and in several disorders, including AD (122). For instance, lipid droplets build up in a subset of aged microglia designated as lipid-droplet-accumulating microglia (LDAM) (Figure 4, step ⑤). LDAM are characterized by defective phagocytosis, as well as enhanced oxidative stress and proinflammatory status, which could contribute to neurodegeneration (123).

APOE4 is associated with numerous cell-type-specific transcriptomic variations compared to APOE3. A comparison between the APOE3 and APOE4 transcriptomes in astrocytes identified over 1,000 differentially expressed genes linked to lipid metabolism and transport, further supporting the differential involvement of apoE isoforms in lipid homeostasis (93).

3.1.9. Neurite outgrowth, neuronal survival, and synaptic integrity and function.

Neuronal uptake of apoE lipoprotein particles via apoE receptors is critical to brain homeostasis, synaptic integrity, and synaptic function. For instance, apoE isoforms differentially modulate neurite outgrowth of mouse cortical neurons, with apoE3 promoting and apoE4 diminishing neurite extension (124). This effect was shown to be LRP1 dependent (124). The same effect has been reported using different cell models, including neuronal cell lines and primary hippocampal cultures (125, 126). In vivo, young apoE3-TR mice exhibit enhanced neuronal dendritic arborization, more complex dendritic branching patterns, and overall higher total dendritic length in the lateral amygdala compared to apoE4-TR mice (127). Differences in spine morphology were also observed in the cortex of 4-week-old apoE-TR mice, with apoE2-TR animals displaying significantly longer spines compared to apoE3-TR and apoE4-TR mice (128).

ApoE-containing lipoproteins also differentially participate in maintaining synaptic integrity and in supporting synaptic function. In vitro, coculturing APOE3/3 iPSC-derived neurons with APOE3/3 iPSC-derived astrocytes leads to an increase in the expression of synaptic proteins such as PSD95 and vGlut1, significantly more than that observed in cocultures with APOE4/4 iPSC-derived astrocytes (105). Interestingly, synaptic deficits have been reported in apoE4-TR mice compared to apoE3-mice, with apoE4-TR mice displaying decreased excitatory transmission in the lateral amygdala (127).

ApoE was also shown to impact neuronal viability. Addition of apoE-containing lipoproteins to rat retinal ganglion cells improved neuronal survival when cultured under growth factor–free conditions, through a mechanism dependent on LRP1. ApoE4-containing lipoproteins were less efficient in protecting against neuronal apoptosis than apoE3-containing particles (10). More recently, this effect was also reported in human iPSC-derived neurons cultured in growth factor–free medium. Cocultures of neurons with APOE3/3 iPSC-derived astrocytes led to significantly improved neuronal survival compared to coculture with APOE4/4 astrocytes (105).

3.1.10. Immune responses.

The APOE genotype has also been linked to differential effects on immune function. For instance, microglial cultures established from apoE4-TR mice respond to immune stimulation by producing an increased amount of nitric oxide and cytokines compared to those from apoE3-TR mice (129). Reports on the effect of apoE2 on microglial activation have been more controversial, with some studies reporting no difference (130) or a reduced response (131) compared to apoE3. Transcriptomic analysis in iPSC-derived microglia indicated that the major differences between APOE4 and APOE3 pertain to the immune response–related genes, providing further clues for potentially altered immune responses associated with APOE4 (93).

3.2.1. Discovery of apoJ.

ApoJ, also known as clusterin, is an acidic glycoprotein first isolated and purified from ram rete testis fluid (132). It is also referred to as an extracellular chaperone because of its function in preventing aggregation of nonnative proteins (133). ApoJ has been isolated in a wide range of tissue and fluids in the brain and periphery, highlighting its complexity and multifunctionality in various biological pathways (134). It is encoded by a highly conserved gene, CLU, localized to chromosome 8 in humans (134, 135).

3.2.2. Structure of apoJ.

The crystal structure of purified apoJ has been only partially elucidated. ApoJ is a heterodimer of a ~70-kDa glycoprotein originating from a 58-kDa precursor protein (136). The apoJ precursor protein is first N-glycosylated with high-mannose oligosaccharide and sialic acid moieties. Proteolytic cleavage prior to secretion generates two separate subunits, apoJα (34–36 kDa) and apoJβ (36–39 kDa) (137). Both subunits contain five Cys residues through which disulfide bonds are formed to yield mature apoJ (138). Human apoJ purified from plasma contains α-helices, β-sheets, and random coils in both its lipidated and nonlipidated forms (139). Further characterization of its secondary structure has postulated the presence of both lipid-binding and heparin-binding domains in apoJ (138). Recent data suggest that the apoJ–heparin interaction may serve as a protein quality control system for preserving extracellular proteostasis (140).

3.2.3. Production in the brain.

Distribution of apoJ mRNA in human tissue revealed that apoJ is widely expressed in several peripheral tissues as well as in the brain (138). In addition, secreted apoJ was detected in a wide range of peripheral fluids including breast milk, plasma, and CSF (138, 141, 142).

In the human brain, apoJ mRNA was detected in astrocytes within the hippocampus (37). In situ hybridization studies in adult rat brain showed that apoJ mRNA was also detected in a subset of hippocampal neurons. Despite high levels of apoJ mRNA in astrocytes, apoJ was predominantly detected in neurons, suggesting potential posttranscriptional regulation events such as protein turnover, secretion, and trafficking (37). Further analysis of apoJ mRNA profiles in vitro showed that mouse primary neurons and astrocytes exhibit distinct apoJ mRNA profiles, resulting in different cell-type-specific and organelle-specific apoJ isoforms. The differential functions of these various isoforms remain unclear; however, a nuclear isoform has been suggested to regulate apoptosis (143).

3.2.4. ApoJ-containing lipoproteins in the central nervous system.

ApoJ-containing lipoproteins were first isolated from human plasma (141). Characterization of apoJ-containing particles revealed that apoJ is associated with lipid-poor, protein-rich apoA1-containing HDL, as well as other plasma lipoproteins (141, 144-146). Consistent with this observation, apoJ-containing lipoproteins have a diameter from 5 to 16 nm, as shown by negative stain TEM (141).

Different distribution patterns of apoJ in human CSF lipoprotein particles have been reported. ApoJ was initially reported as being uniformly distributed across CSF lipoproteins of varying sizes (38). However, CSF fractionation by density ultracentrifugation followed by immunoblotting showed that the majority of apoJ in CSF is found in a fraction similar in size to plasma very-high-density lipoproteins (147). CSF fractionation by anti-apoA1 and anti-apoE immunoaffinity chromatography showed that apoJ is associated with both apoA1-containing and apoE-containing lipoprotein particles, as well as with particles containing neither apoA1 nor apoE and as free protein (38, 142).

Rat primary astrocytes secrete apoJ in the form of discoidal lipoproteins devoid of a cholesteryl ester core, suggesting that astrocyte-secreted apoJ-containing particles acquire additional lipids before reaching CSF (38). Additionally, cultured astrocytes from WT, Apoe-KO, and apoE-TR mice all secreted apoJ-containing particles comparable in size, suggesting that apoE does not alter the astrocytic lipidation of apoJ (148). ApoJ secreted from primary astrocytes from apoE-TR mice is relatively lipid poor compared to secreted apoE, with a protein-to-lipid ratio of three to one (149).

3.2.5. Role of ApoJ in lipid transport in the central nervous system.

Numerous functional properties have been attributed to apoJ, in both physiological and pathological conditions. For instance, apoJ has a role in innate immune responses as an inhibitor of the complement system (150). It also acts as an extracellular chaperone with increased expression in response to cellular stress (151) and displays antiapoptotic properties (152). Roles for apoJ in lipid transport and redistribution, as well as in reverse cholesterol transport, have been suggested in the periphery (141, 144), given that apoJ was identified as a minor component of lipoproteins (141, 145, 146) and that apoJ-containing lipoprotein particles in plasma display cholesteryl ester transferase activity (141). However, these functions remain poorly characterized to date.

Little is known about apoJ and its role in lipid and cholesterol transport in the brain. ApoJ binds to several receptors in the brain that belong to the LDLR family, such as LRP1, LRP2, VLDLR, and ApoER2 (153), suggesting a potential role for apoJ in lipid trafficking in the CNS. ApoJ binds LRP2 with high affinity (62). LRP2, also known as megalin, is a transmembrane protein expressed in epithelial cells of ventricular ependyma, capillaries, and choroid plexus (61, 62). LRP2 mediates apoJ endocytosis (62); however, it is still unclear whether apoJ facilitates lipid transport and distribution in the brain and whether receptor-mediated endocytosis or transcytosis plays a role in this process. Lipidated apoJ, but not delipidated apoJ, is also a ligand for TREM2, suggesting a role for apoJ-mediated lipid-related function in microglia (65).

3.3.1. ApoA1.

Lipoproteins isolated from CSF are rich in apoA1, which is one of the most abundant apolipoproteins in the CSF. Since its expression has been ascribed to the liver and intestines, it is believed that the presence of apoA1 in the CNS originates from the periphery (4). For instance, knocking down Apoa1 in both intestine and liver in mice drastically reduces HDL cholesterol in plasma and results in a lack of apoA1 in CSF (154). Injection of fluorescently labeled human apoA1 intravenously in mice showed that it can enter the CNS via the choroid plexus. ApoA1 can also be taken up by human epithelial and endothelial cells in in vitro models of the BBB (155). The exact roles of apoA1 in brain homeostasis and pathogenic events require further investigation.

3.3.2. ApoD.

ApoD is one of the few apolipoproteins synthesized in the brain and is produced by astrocytes and oligodendrocytes. However, despite belonging to the apolipoprotein family, it is structurally closer to proteins from the lipocalin family and does not appear to be involved in the synthesis of nascent lipoprotein particles. Nevertheless, apoD is capable of binding lipids, but its role in lipoprotein metabolism in the CNS is unclear (7).

3.3.3. Other apolipoproteins in the brain.

Additional apolipoproteins such as apoA2, apoA4, apoC, and apoH have been detected in the brain. Their origin and role in lipoprotein metabolism in the CNS remain unclear to date (7).