Lipoprotein-mediated brain cholesterol transport: a potential pharmacological target in Alzheimer’s Disease

Abstract

STATE OF THE ART: AD is a multifactorial neurodegenerative disease clinically characterized by a progressive cognitive impairment. Alterations of cholesterol homeostasis in the central nervous system (CNS) have been associated to various neurodegenerative disorders, including Alzheimer's disease (AD), as shown by genomic-wide association studies that have identified several loci involved in lipid metabolism among AD susceptible genes. Among these, the apoE4 gene is the most important risk factor for Ad onset. Cholesterol synthesis in neurons is very high during embryogenesis, but it gets progressively lost in adult neurons, that rely on cholesterol produced from astrocytes for neuronal plasticity maintenance, synaptogenesis and regeneration after injury. CNS cholesterol trafficking between astrocytes and neurons is mediated by HDL-like particles, similar to plasma HDL, identified in human in the cerebrospinal fluid (CSF) and containing mainly apolipoprotein E (apoE). HDL-like particles interact with membrane cholesterol transporters such as the ATP-binding cassette transporters A1 and G1 (ABCA1 and ABCG1) and convey cholesterol from astrocytes to neurons. In this process, nascent discoidal particles are firstly secreted by astrocytes through the activity of the ABCA1 transporter; nascent particles successively undergo maturation through the activity of remodeling enzymes, such as lecithin cholesterol acyltransferase (LCAT) which esterifies cholesterol and catalyses the conversion of nascent discoidal into mature and spherical particles. Newly generated HDL-like particles can be finally uptaken by neurons through the binding of apoE to the LDL receptor and LDLR family receptors (LRP1, LDLR, VLDLR and apoER2). These receptors are sensitive to the degrading activity of the Proprotein convertase subtilisin/kexin type 9 (PCSK9), a serine protease firstly described to target hepatic low-density lipoprotein receptor (LDLr) but also identified in brain and detected in human cerebrospinal fluid. PCSK9 modified activity might in principle be involved in the derangement of brain cholesterol trafficking, in lipoprotein homeostasis and in AD pathogenesis. The genetic studies conducted so far in humans are not conclusive on the impact of PCSK9 mutations on AD. Although Wollmer and colleagues firstly identified PCSK9 among the cholesterol-related genes that has been matched with AD genes listed in the AlzGene database, no association was found between PCSK9 polymorphism and the risk of AD onset neither in a Japanese nor in a Swedish cohort study. Consistently, in a recent Mendelian randomization analysis, PCSK9 loss-of-function mutations were not associated to a raise in the risk of AD (Hazard Ratio (HR) = 0.50; p = 0.37). To a negative conclusion came also the results of genetic studies among African American REGARDS (Reasons for Geographic and Racial Differences in Stroke) participants with and without the PCSK9 loss-of-function variants C697X or Y142X. The presence of these variants did not affect the primary endpoint of the study, i.e. the neurocognitive performance. In another study conducted in French Canadian subjects, carriers of the PCSK9 loss of function mutations, R46L and InsLEU, did not differ from non-carriers as either AD prevalence or age of disease onset. In addition, two PCSK9 SNPs seems to predispose to increased AD risk although with a gender-specificity. Only few observations suggest that in neurodegenerative diseases cholesterol transport between astrocytes and neurons mediated by HDL-like particles may be altered. AIM: The objective of my PhD project was to establish, through a case-control study, whether the HDL-mediated cholesterol transport from astrocytes to neurons is defective in AD. To achieve this goal, we investigated cholesterol trafficking with a dual approach: from the astrocyte side we measured the ability of CSF HDL to promote cell cholesterol efflux through the transporters ABCA1 and ABCG1 (CSF cholesterol efflux capacity, CSF-CEC). From the neuronal point of view first, we measured PCSK9 levels in CSF of AD patients to evidence potential differences. After that, we also test the hypothesis that PCSK9 could be a pathogenetic factor in AD by studying the impact of PCSK9 on AD using an in vitro approach focusing on cholesterol uptake and expression of receptors modulating cholesterol homeostasis in the brain. MATERIAL AND METHODS: Controls (n=39) and sex and age-matched AD patients (n=37) were recruited. In addition, in order to verify if the potential alteration of cholesterol transport is specific for AD, we included in the analysis also subjects with non-AD-related dementia (DEM non-AD, n=16). CSF capacity to promote cholesterol efflux (CSF-CEC) was evaluated thought a radioisotopic technique analyzing transporters ABCA1 and ABCG1 by using specific cell models expressing the single transporters. CSF total apoE, apoE4, apoA-1 and PCSK9 were performed by ELISA kit. Reconstituted HDL (rHDL) containing apoE, were prepared using the cholate dialysis procedure containing ApoE⁄lecithin⁄cholesterol in molar ratio of 1:100:2. In parallel, liposomes made of PL and free cholesterol but without apolipoprotein E were prepared, to evaluate non-apolipoprotein-mediated uptake of cholesterol. Cholesterol uptake was analyzed in the lysates to determine the amount of labelled cholesterol incorporated by scintillation counting and expressed as CPM/mg proteins. LDLr and apoER2 protein expression was evaluated by Western Blot analyses while apoE-FITC internalization into cells was performed using the confocal laser scanning microscopy. RESULTS: The analysis of CSF total apoE and apoA-1 concentrations did not reveal differences between the three groups; conversely, levels of the isoform apoE4 were higher in CSF of AD subjects compared to controls (+3,9 fold; p=0.0068). CSF-CEC through AD did not differ between the three groups except for a slight but significant decrease in DEM non-AD subject compared to controls (- 40%; p=0,002). CSF- CEC through ABCA1 was reduced in AD patients compared to controls (-73%; p=0,001) while no difference was found with DEM non-AD and also CSF-CEC through ABCG1 was significantly reduced in AD subjects (-33; p=0,004). CSF-CEC in DEM non-AD thought these two pathways was similar to controls. We have also stratified the CSF-CEC values according on their apoE phenotype founding no differences in CSF ABCA1-CEC and ABCG1-CEC between carriers and non- carriers of the apoE4 genotype. We found a positive relationship between ABCG1-CEC with Aβ1-42 amyloid levels (p=0,025; r=0,305). Moreover, we found an inverse relationship between ABCA1-CEC and total Tau and phospho-tau levels (p=0,018; r=0,348 and p=0,048; r=0,294). In addition, the analysis of CSF revealed that PCSK9 levels were significantly higher in AD patients than in controls (+1.65 fold; p=0.0106). With respect to the relationship between PCSK9 and apoE in CSF, considering all samples analyzed, PCSK9 did not significantly correlate with total apoE (p=0.285) but it positively correlated with apoE4 levels (r2=0,445; p=0,0065). Interestingly we found that CSF PCSK9 levels were higher in APOE Ɛ4 carriers among both non-AD and AD subjects, reaching statistical significance in the AD group (+1.53 fold; p=0.05). We also evaluated the modulating effect of exogenous PCSK9 on brain cholesterol homeostasis by in vitro studies. For this reason, apoE containing particles (rHDL- apoE) were prepared and based on previous published data, we expected a size of 0.9nm. Although the size was not directly verified, from literature we know that such particle are able to promote cholesterol efflux through ABCG1 and aqueous diffusion but not through ABCA1. For this reason, ApoE containing particles (rHDL- apoE) functionality was measured for their capacity to promote cholesterol efflux through ABCA1, ABCG1 and aqueous diffusion in different cell models. As expected from the predicted size, apoE-containing sHDL were able to produce a significant cholesterol efflux through ABCG1 pathway and aqueous diffusion. On the other hand, these apoE- containing particles were not able to promote cholesterol efflux through ABCA1 as mentioned for apoAI, the favorite acceptor through this pathway. In order to mimic the degrading effect of PCSK9 on LDLR-family, that internalized rHDL-apoE in neurons, we tested the capacity of rHDL-apoE to be internalized by fibroblast from control patients (ctrl) and from patients with homozygous familiar hypercholesterolemia (FH) that are characterized by the absence of the low-density lipoprotein receptor (LDLR), as it would occur as a consequence of the degrading activity of PACK9. We found that rHDL-apoE (30-60nm) uptake from fibroblast FH was markedly reduced compared to fibroblast ctrl (p<0,0001 for both concentrations) while liposomes prepared without apoE display a similar uptake between fibroblast ctrl and FH. We also evaluated the capacity of rHDL-apoE to be internalized in a macrophage cell line overexpressing PCSK9 (J774 CTRL and J774 PCSK9) and we observed that the uptake of rHDL-apoE was significantly reduced in J774 overexpressing PCSK9 compared to the CTRL ones. Also, in this cellular system, liposomes without apoE are not able to produce a reduction in cholesterol uptake in J774 PCSK9 compared to control J774. We investigated the rHDL-apoE uptake into SH-SY5Y human neuroblastoma cells differentiated into neurons and we found that the uptake was significant and occurred in a concentration-dependent manner. After the generation of a cellular model of SH-SY5Y overexpressing or not PCSK9, we first evaluated the protein expression and the secretion of PCSK9 from these cells. In differentiated SH-SY5Y overexpressing PCSK9, the uptake of rHDL-apoE [60 and 120nM] was reduce compared to control cells (p= 0.0232 and p=0.005 respectively) and dependent from apoE since rHDL without apoE displayed a similar uptake. We analyzed the expression of two receptor involved in PCSK9 degrading activity, LDLr and apoER2, finding that differentiated PCSK9 cells displayed a lower expression of both receptors (for both p<0.05). Moreover, in the same cellular model, we observe the interaction between fluorescinated-apoE and neuronal cells thought confocal laser scanning microscopy. The first event observed was the appearance of granular pattern and widespread fluorescence in both SH-SY5Y control and PCSK9 overexpressing cells. After 90 minutes a morphological evidence between the two cell types were observed with fluorescent grains distributed along the cell membrane to outline the cells that become more evident in the SH- SY5Y control compared to PCSK9. CONCLUSIONS: Our results of the case-control study indicate that, despite no change in CSF apoE concentrations, possibly reflecting no change in HDL levels, an impairment of CSF capacity to promote cholesterol efflux specifically occurs in AD. These findings, together with increased PCSK9 levels in AD patients and the correlations found with the typical neurobiomarkers, reinforce the idea that alterations of brain cholesterol transport may play an important role in AD pathogenesis. With respect to CSF PCSK9 evaluation, our results showed for the first time an alteration of CSF PCSK9 levels in AD and suggest a pathophysiological link between PCSK9, apoE4, and AD. To better understand mechanism involved in the pathophysiological role of PCSK9. In addition, our in vitro findings, although preliminary, indicate that PCSK9 can negatively modulate the uptake of cholesterol possibly by reducing the expression of the apoE-receptors. Future experiments will aim to clarify mechanistic insights on the possible pathogenetic role of PCSK9 in the brain.