Background
Alzheimer’s disease (AD) is the leading cause of dementia in the elderly, which accounts for 60 to 80 percent of dementia cases [
1]. Brain extracellular amyloid plaques and intracellular neurofibrillary tangles are the two pathological hallmarks of AD. Apolipoprotein E (apoE) is a major apolipoprotein and a cholesterol carrier in the brain [
2]. The human
APOE gene exists as three major polymorphic alleles: ε2, ε3 and ε4. Ample evidence indicates that the ε4 allele of the
APOE gene is the strongest genetic risk factor for late-onset AD (LOAD), whereas the ε2 allele is protective [
2-
6].
While the primary function of apoE is to deliver cholesterol and other essential lipids to neurons through binding to cell surface apoE receptors, apoE also regulates brain metabolism of amyloid-β (Aβ) [
7], accumulation of which leads to deposition of amyloid and is considered the initiating event in the pathogenesis of AD [
8-
10]. Previous studies have shown that brain Aβ levels and amyloid plaque loads are apoE isoform-dependent (E4 > E3 > E2) both in humans and in AD transgenic mouse models [
11-
13]. Several reports showed that the
APOE genotype strongly affects apoE levels (E4 < E3 < E2) in human cerebrospinal fluid (CSF), brain parenchyma, and in apoE targeted-replacement (apoE-TR) mice [
14-
16]. Analysis on the levels of apoE and Aβ in different brain regions of non-demented individuals found that apoE levels negatively correlate with Aβ levels both in
APOE4 carriers and non-carriers [
17]. ApoE protein levels in human CSF have also been shown to positively associate with CSF Aβ42 levels [
14]. In addition, liver X receptors (LXRs) or the retinoid X receptor (RXR) agonists facilitate Aβ clearance and reverse the memory deficits in amyloid model mice by increasing apoE levels and its lipidation [
18-
20]. These results suggest that apoE levels and lipidation status likely contribute to Aβ clearance; thus, new therapeutic approaches aimed at increasing apoE expression are actively being pursued. However, to better design mechanism-based therapy, it is critical to understand how increasing apoE expression, in particular apoE4 in
APOE4 carriers, impacts apoE lipidation and Aβ metabolism.
Given that apoE is produced predominantly by astrocytes in the brain [
21], we investigated the impact of overexpressing different human apoE isoforms in astrocytes in apoE-TR mice, in which the coding region of the mouse endogenous
Apoe gene was replaced with one of the three human
APOE alleles without changing the regulatory elements required for modulation of gene expression [
22-
24]. Thus, apoE-TR mice express human apoE isoforms at the physiological levels and respond to regulatory pathways in a physiological setting [
22-
25], providing an excellent
in vivo system to explore the normal function of each
APOE allele and apoE-associated diseases. Here, we used a gene delivery approach by which adeno-associated viral serotype 8 (AAV8) vectors expressing various human
APOE alleles under the control of astrocyte-specific glial fibrillary acidic protein (GFAP) promoter were bilaterally injected into the cerebral lateral ventricles of neonatal apoE3-TR or apoE4-TR mice. We demonstrated that apoE isoforms are specifically expressed in astrocytes in the brain three months after injection. Importantly, we found that increasing apoE4 levels in apoE4-TR mice led to decreased apoE lipidation, lower apoE-associated cholesterol, and increased endogenous Aβ. Conversely, increasing apoE2 expression in apoE4-TR mice had the opposite, beneficial effects. Our findings for the first time reveal the consequential effects of overexpressing apoE isoforms in specific isoform background and should help to provide guidance for the designs of apoE-based targeted therapy to treat AD.
Discussion
The most prevalent genetic risk factor for late-onset AD is the
APOE genotype with an allele-specific risk profile:
APOE4 >
APOE3 >
APOE2 [
33,
34]). Despite progress in the past two decades [
2,
3], it remains unclear how apoE4 increases, and apoE2 decreases, the risk of AD. The amyloid hypothesis proposes that an imbalance between Aβ production and clearance leads to Aβ accumulation in the brain that initiates AD pathogenesis [
8,
10]. Multiple lines of evidence demonstrate that apoE isoforms differentially regulate Aβ metabolism in the brain [
35-
38], likely through modulation of Aβ clearance and aggregation [
3,
12,
32]. These findings suggest that apoE levels, lipidation status and isoform-dependent effects may differently impact Aβ metabolism, thereby contributing to or preventing AD pathogenesis.
In the current study, we used AAV8 constructs driven by an astrocyte-specific GFAP promoter to express three apoE isoforms (AAV8-GFAP-apoE2/3/4) in neonatal apoE3-TR or apoE4-TR mice through intracranial injection. Transduction of AAV8-GFAP-apoE led to sustained expression of apoE isoforms in mouse brain. We found that transduction of apoE4 in apoE4-TR background significantly increased the amounts of small, lipid-poor apoE particles and the levels of endogenous Aβ. Conversely, transduction of apoE2 in the apoE4-TR background led to an increase in the amounts of large, lipid-rich apoE particles and decreased endogenous Aβ. Our findings indicate that modulating the levels of apoE isoforms had differential effects on the lipidation status of apoE and Aβ levels in apoE4-TR, less so in apoE3-TR mice. Specifically, overexpression of apoE2 and apoE4 in apoE4 background leads to opposite effects on the amounts of cholesterol associated with apoE and Aβ accumulation with apoE4 being detrimental and apoE2 beneficial. We did not observe significant changes in APP proteolytic processing products in mice transduced with different apoE isoforms, indicating that APP processing to Aβ is not regulated by the expression levels of apoE isoforms.
A recent study assessed the effects of overexpressing apoE isoforms in the background of amyloid model mice with endogenous murine apoE background [
39]. They found that overexpression of apoE4 leads to higher levels of Aβ and enhanced amyloid deposition, whereas overexpression of apoE2 had the opposite effects. Our current study uniquely address how overexpression of apoE isoforms in different human apoE isoform background impact apoE lipidation and Aβ accumulation, events that are critical for the synaptic functions and the pathogenesis of AD. When combined, these studies should provide mechanism-based guidance on regulating apoE expression and lipidation in the designs of therapeutic approaches to treat AD.
The mechanism underlying the protective effect of apoE2 against AD remains unclear. Our results show that apoE protein levels are significantly higher in mice transduced with apoE2 compared with mice transduced with GFP controls, apoE3 or apoE4 viruses, despite similar mRNA levels. Previous studies have shown that apoE2 may exhibit a more stable conformation [
40] and has lower affinity to the low-density lipoprotein (LDL) receptor, which may lead to decreased catabolism [
15]. Thus, the higher levels of apoE observed in apoE2-transduced mice may be attributed to its increased stability and decreased turnover. Previous studies suggest that formation of protein complexes between apoE and Aβ may facilitate Aβ clearance [
41]. Compared with apoE2 or apoE3, apoE4/Aβ complexes have been shown to be less stable, which is likely attributed to the poor lipidation status of apoE4/lipoprotein particles [
7,
31,
42]. In addition, complex formation between Aβ and apoE isoforms may affect their binding property to their individual or common receptors including the low-density lipoprotein receptor-related protein 1 (LRP1), thereby further affects Aβ clearance [
32,
43]. Thus, it is possible that overexpression of apoE2 increases and apoE4 decreases apoE/Aβ complex formation and therefore, the clearance of Aβ in the brain.
Emerging evidence indicates that apoE lipidation status directly affects its role in Aβ clearance [
20,
44,
45]. Our current study demonstrates that AAV-mediated transductions of apoE isoforms have different effects on the lipidation status of apoE. Compared with controls, transduction of apoE4 in apoE4-TR mice increased the proportion of poorly-lipidated apoE. Several studies suggest that apoE4 is significantly less lipidated than apoE2 and apoE3 in both human and APP transgenic mice expressing human apoE isoforms [
29-
31]. ApoE4 more easily self-aggregates, which may affect its lipidation capacity [
46]. Hence, it is possible that overexpression of apoE4 in apoE4-TR mice results in apoE4 aggregation, which might impair its lipidation. Consistent with this notion, haploinsufficiency of ABCA1, which mediates apoE lipidation, impairs Aβ clearance and exacerbates amyloid deposition in apoE4-TR mice, but not in apoE3-TR mice [
47]. Though we did not observe an alteration in the levels of ABCA1 and ABCG1, boosting apoE4 in apoE4-TR mice may enhance apoE4 aggregation in parallel with an impairment of its lipidation, which in turn decreases Aβ clearance and/or promotes Aβ deposition.
In addition to its effects on Aβ metabolism, apoE mediates neuronal delivery of cholesterol, an essential component for synaptic plasticity required for learning and memory formation [
2]. Given the ongoing clinical trial aimed at increasing apoE levels and its lipidation [
20], it is imperative to understand the consequences of boosting apoE on cholesterol homeostasis and Aβ metabolism in the background of different apoE isoforms. Our current study directly demonstrates the divergent effects of apoE2 and apoE4 isoforms on Aβ accumulation and apoE lipidation status, effects that are more pronounced in apoE4-TR mice.
Materials and methods
Animals
Human apoE3-TR or apoE4-TR mice, which express respective human apoE isoform driven by the endogenous murine ApoE promoter, were obtained from Taconic. All animal experiments were conducted in compliance with the protocols of the Institutional Animal Care and Use Committee at Mayo Clinic.
Viral vector construction and production
The AAV8 vector construction and production were performed by the Hope Center Viral Vectors Core at the Washington University School of Medicine. ApoE2, 3, or 4 cDNA in pcDNA3.1 was amplified by PCR with the forward primer containing a BamHI site and the reverse primer containing a SalI site and an influenza hemagglutinin (HA) tag sequence. The PCR products were digested with BamHI and SalI and inserted into an AAV8-GFAP-GFP backbone in which the GFP was replaced with the PCR products. The resulting AAV8-GFAP-aopE2-HA, AAV8-GFAP-aopE3-HA, and AAV8-GFAP-aopE4-HA contain the respective apoE isoforms with a HA tag under the control of a GFAP promoter. The AAV8 viruses were produced as previously described [
48]. Briefly, the packaging cell line HEK293 is maintained in Dulbecco’s modified Eagles medium (DMEM), supplemented with 5% fetal bovine serum (FBS), 100 units/ml penicillin, 100 μg/ml streptomycin in 37°C incubator with 5% CO
2. The cells were co-transfected with pAAV2/8, pHelper, and rAAV transfer plasmid containing GFP or apoE isoforms using the calcium phosphate precipitation method. The cells were incubated at 37°C for 3 days before harvesting. Cells were lysed by three freeze/thaw cycles. The cell lysate were treated with 50U/ml of Benzonaze followed by iodixanol gradient centrifugation. The iodixanol gradient fraction is further purified by HiTrap Q column chromatography (GE Healthcare) and concentrated with Vivaspin concentrator (Bohemia, NY). The virus titer was determined by dot blot assay.
Intracranial injections in neonatal mice
AAV8 virus intracranial injection in neonatal mice was performed as previously described with minor modifications [
27,
28]. Briefly, postnatal day 2 (P2) mice were cryoanesthetized on ice for 5 min and injected intracerebroventricularly with 2 μl (~10
13 viral particles/ml) of viruses into both hemispheres using a syringe (7642–01, Hamilton) with a 30 gauge needle (7803–07, Hamilton). After injections were complete, pups were placed on a warming pad until they regained normal color and resumed movement. All injected animals were then returned to their mothers for further recovery.
Preparation of mouse brains
Three months post AAV virus injection, mice were anesthetized using isoflurane, perfused with PBS, and their brains were rapidly harvested. The right hemibrains were fixed in 4% paraformaldehyde and used for immunofluorescence analysis, and left hemibrains were dissected and kept frozen at −80°C until further biochemical analysis.
Immunohistochemistry
The immunohistochemistry for the brain sections was performed as previously described [
27]. Briefly, brains were fixed overnight at 4°C in fresh 4% paraformaldehyde and then transferred to 30% sucrose for cryoprotection. Brains were frozen on dry ice and then sectioned at 45 μm using a freezing-sliding microtome (LEICA SM2400, Leica). Sections were permeated and blocked with 3% goat serum containing 0.1% Triton-X 100 in Tris-buffered saline (TBS) buffer (25 mM Tris, 0.15 M NaCl, pH 7.4) for 1 h at room temperature. Then, the sections were incubated with mouse anti-HA (1:1000, Covance), rabbit anti-GFAP (1:1000, Millipore), rabbit anti-NeuN (1:1000, Millipore) or rabbit anti-Iba1 (1:1000, Wako Chemicals) antibodies in blocking solution at 4°C overnight. The sections were washed with TBS/0.5% Tween 20 (TBST) and incubated with Alexa Fluor 488-conjugated goat anti-mouse (1:400; Invitrogen) and Alexa Fluor 568-conjugated donkey anti-rabbit (1:400; Invitrogen) secondary antibodies for 2 h at room temperature. Images were obtained by confocal microscope (LSM 510 META, Zeiss) and analyzed by ZEN software.
Quantitative real-time PCR
Total RNAs were extracted from frozen cortical tissues using Trizol (Invitrogen) and Direct-zol RNA MiniPrep kit (Zymo Research). Reverse transcription was performed using SuperScript III First-Strand Synthesis System (Invitrogen). Real-time qPCR was conducted with Universal SYBR Green Supermix (Bio-Rad) using an iCycler thermocycler (Bio-Rad). The following primers were used: Human apoE mRNA forward primer, TGTCTGAGCAGGTGCAGGAG; and reverse primer, TCCAGTTCCGATTTGTAGG. Mouse β-actin forward primer, AGTGTGACGTTGACATCCGTA; and reverse primer, GCCAGAGCAGTAATCTCCTTC Relative mRNA levels were calculated by ΔΔCt method with β-actin used as a reference.
ELISA for human apoE
Cortical tissues were homogenized with a Polytron homogenizer in ice cold TBS buffer containing protease inhibitor and phosphate inhibitor cocktails (Roche). The homogenates were centrifuged at 100,000 g for 30 min at 4°C and the supernatants were collected. ApoE levels were analyzed using an ELISA. Briefly, 96-well plates were coated overnight with an apoE antibody (AB947, Millipore) in carbonate buffer at 4°C overnight. The plates were blocked with 1% Block Ace in PBS, and then washed with PBS 3 times. Recombinant apoE (Fitzgerald) along with samples were diluted and added at a volume of 100 μl/well incubated at 4°C overnight. The plates were washed and incubated with biotin-conjugated goat anti-apoE antibody (Meridian Life Science) for 2 h at room temperature. After incubation with Horseradish Peroxidase Avidin D (Vector Laboratories) for 90 min at room temperature, the plate was developed by adding tetramethylbenzidine Super Slow substrate (Sigma). The reaction was stopped and read at 450 nm with a microplate reader.
ELISA for mouse endogenous Aβ
Mouse endogenous Aβ was extracted from cortical tissues by diethylamine as previously described [
49]. The Aβ levels in age-matched APP knockout mice were used as a negative control and subtracted as background. For detection of endogenous mouse Aβ40, 96-well plates were coated at 4°C overnight with 13.1.1 mAb (anti-Aβ35-40) [
49,
50]. The plates were blocked with 1% Block Ace in PBS, washed with PBS and loaded with samples. Synthetic rodent Aβ40 (AnaSpec) was used as standards. Following overnight incubation at 4°C, the plates were washed, followed by incubation with HRP-conjugated mAb 32.4.1 (rodent Aβ1-16 specific) [
49] for detection. The plates were developed with tetramethylbenzidine Supersensitive substrate (Sigma). The reaction was stopped and read at 450 nm with a microplate reader.
Analyses of apoE/lipoprotein particles by Blue Native PAGE and Western blot
Cortical tissues were homogenized with a Polytron homogenizer in ice cold TBS buffer containing protease inhibitor and phosphate inhibitor cocktails (Roche). Proteins were separated by Native PAGE™ Novex 4–16% Bis-Tris gels (Invitrogen) under native conditions following the manufacturer’s instructions and transferred to PVDF (Millipore) at 100 V for 1 h using the Trans-Blot Cell (Bio-Rad). Blots were treated with Ponceau S Staining Solution (0.1% (w/v) Ponceau S in 5% (v/v) acetic acid) to visualize the molecular mass markers. The NativeMark Unstained Protein Standard from Invitrogen was used for estimation of particle sizes. After washing in TBS, blots were processed for Western blot. The membrane was incubated with goat anti-apoE antibody (K74180B, Meridian Life Science) overnight at 4°C, followed by peroxidase-labeled donkey anti-goat antibody (Santa Cruz). The membrane was developed with Lumigen ECL Ultra Western Blotting HRP Substrate (Lumigen), and the signals were detected by Fuji film Luminescent Image Analyzer (LAS4000). An antibody that recognizes the C-terminus of APP (18961, IBL-America) was used for detecting APP and its C-terminal fragments. Anti-β-actin (Sigma) was used as a loading control. Western blot bands were quantified by Image J software.
Immunoprecipitation (IP) of apoE particles and cholesterol assay
Streptavidin-conjugated agarose beads (Sigma) were first incubated with biotin-conjugated goat anti-apoE antibody (K74180B, Meridian Life Science) for 2 h at room temperature with shaking. The unbound antibodies were removed by washing with TBS three times. The antibody-bound agarose beads were then mixed with brain lysates in TBS fractions (500 μg total protein) at 4°C overnight. The IP solution was centrifuged for 2 min at 1,000 g to collect agarose beads. The beads were washed with cold TBS three times and re-suspended with TBS/0.1% Triton X-100. The resulting suspension was subjected to Western blot analysis in which equal volumes of samples were loaded. Proteins were separated by 4-20% Mini-PROTEAN TGX Gel (Bio-Rad) and transferred to Immobilon-P PVDF (Millipore). The membranes were incubated with apoE antibody (K74180B, Meridian Life Science), followed by IRDye secondary antibody (LI-COR Biosciences). The results were visualized and quantified by Odyssey infrared imaging system (LI-COR Biosciences). The remaining suspension was used for the detection of apoE particle-associated cholesterol quantified by Amplex Red cholesterol assay (Invitrogen) according to the manufacturer’s protocol.
Statistical analysis
All data were analyzed by one-way analysis of variance (ANOVA) with a Tukey’s post-hoc test using GraphPad Prism 5. Data were presented as average ± SEM. A p value of < 0.05 was considered statistically significant.
Acknowledgements
This work was supported by NIH grants P01NS074969, P01AG030128, P50AG016574, R01AG035355, R01AG027924, R01AG046205, and a grant from the Alzheimer’s Association (to G.B.). We thank Drs. Mingjie Li and Joyce Snyder from the Hope Center Viral Vectors Core at the Washington University School of Medicine for producing the AAV viruses. The Hope Center Viral Vectors Core is supported by a Neuroscience Blueprint Core grant (P30NS057105) from NIH to Washington University. We also thank Caroline Stetler for careful reading of this manuscript.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
Experiments were designed and performed by JH, CL, XC, YZ, HX and GB. Manuscript was written by JH, CL and GB and critically reviewed by XC, YZ and HX. All authors read and approved the final manuscript.