|Home | About | Journals | Submit | Contact Us | Français|
Hypercholesterolemia increases levels of β-amyloid (Aβ), a peptide that accumulates in Alzheimer’s disease brains. Because cholesterol in the blood does not cross the blood brain barrier (BBB), the link between circulating cholesterol and Aβ accumulation is not understood. In contrast to cholesterol, the oxidized cholesterol metabolite 27-hydroxycholestrol can cross the BBB, potentially increasing Aβ levels. However, the mechanisms by which cholesterol or 27-hydroxycholesterol regulate Aβ levels are not known. The insulin-like growth factor-1 (IGF-1) regulates the glycogen-synthase kinase-3α (GSK-3α) and the insulin degrading enzyme (IDE). While GSK-3α increases Aβ production, IDE is a major Aβ-degrading enzyme. We report here that feeding rabbits with a cholesterol-enriched diet increases Aβ levels in the hippocampus, an effect that is associated with reduced IGF-1 levels. 27-hydroxycholestrol also increases Aβ and reduces IGF-1 levels in organotypic hippocampal slices from adult rabbits. We suggest that hypercholesterolemia-induced Aβ accumulation may be mediated by 27-hydroxycholesterol, involving IGF-1 signaling.
Cholesterol is suggested to increase the risk of Alzheimer’s disease (Kivipelto et al. 2001; Pappolla et al. 2003; Wolozin 2004) and has been shown in animal and cellular models to increase the production of β-amyloid (Aβ) peptide (Sparks et al. 1994; Refolo et al. 2000; Shie et al. 2002; Racchi et al. 1997; Simons et al. 1998; Austen et al. 2003; Frears et al. 1999), a major pathological hallmark of Alzheimer’s disease. There is also evidence that cholesterol co-localizes with fibrillar Aβ in the amyloid plaques of transgenic mice (Burns et al. 2003). Despite the large number of studies linking cholesterol to Aβ production, the mechanisms by which cholesterol increases Aβ levels are still to be determined. Cholesterol homeostasis in the brain is regulated through de novo synthesis, with no or very poor transfer from the peripheral circulation due to the impermeability of the BBB to the lipoproteins that carry cholesterol (Lange, et al., 1999). In contrast to cholesterol, 27-hydroxycholesterol, a product of cholesterol oxidation (oxysterol), has been shown to cross the BBB into the brain (Heverin et al. 2005). It may be possible that increased entrance of this oxysterol into the brain following hypercholesterolemia places the brain at risk for neurodegeneration.
Evidence suggests that IGF-1, a neurotrophic factor that promotes neurogenesis and has neuroprotective effects, plays an important role in regulating Aβ peptide levels (Costantini et al. 2006; Puglielli 2008). The IGF-1 signaling involves the binding of IGF-1 to its receptor, IGF-1R, thereby activating protein kinase B (Akt) through phosphorylation. Phosphorylated Akt (p-Akt) modulates IDE expression (Zhao et al. 2004) and regulates the phosphorylation of GSK-3α and β. IDE is a major Aβ-degrading enzyme (Kurochkin and Goto 1994; Farris et al. 2004) due to its ability to cleave the Leu16-Leu17 bond within the Aβ region of the β-APP (Bernstein et al. 1999). The GSK-3α enzyme has been demonstrated to be required for the maximal processing of β-APP and subsequent Aβ production (Phiel et al. 2003). Similarly to GSK-3α, GSK-3β isoform is also an IGF-1 downstream-regulated protein, but is involved in the phosphorylation of tau protein as well as the transcriptional factor cAMP responsive element-binding protein (CREB). While hyperphosphorylation of tau can lead to neurofibrillary tangle formation, phosphorylated CREB (p-CREB) can promote cell survival by up-regulating the expression of the anti-apoptotic proteins Bcl-2 (Ji et al. 1996; Riccio et al. 1999).
Currently, the effect of cholesterol or 27-hydroxycholestrol on IGF-1, IDE, GSK-3α /β, CREB, and Bcl-2 levels and the extent to which changes in levels of these proteins are associated with increased Aβ levels are not clear. The aim of this work was to determine the effect of a cholesterol-enriched diet and 27-hydroxycholestrol on the IGF-1 signaling pathway in rabbit hippocampus. Results of our work would add important insights into the cellular mechanisms by which high cholesterol levels may be associated with Alzheimer’s disease-like pathological hallmarks in rabbit brains.
New Zealand white female retired breeder rabbits (4 ± 0.4 kg and 3 ± 0.25 years old) were used in this study. Animals were randomly assigned to 2 groups as follows: Group 1, normal chow (n=6) and group 2, chow supplemented with 2% cholesterol (n=6). Diets were kept frozen at - 10°C to reduce the risk of oxidation. The animals were allowed water filtered through activated carbon filters. Cholesterol-treated animals and their matched controls were euthanized 12 weeks later. At necropsy animals were perfused with Dulbecco’s phosphate buffered saline at 37°C and the brains were promptly removed. All animal procedures were carried out in accordance with the U.S. Public Health Service Policy on the Humane Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at the University of North Dakota.
We have recently succeeded in growing organotypic slices from adult animals following method optimization (Schrag et al. , 2008). One of the advantages of the organotypic slice system is that local connectivity between neurons, interneurons, and glia is maintained. Organotypic hippocampal slices were prepared as follows. Hippocampi from adult rabbits (n = 3; 2.5–3 years old) were dissected, trimmed of excess white matter and placed into chilled dissection media composed of hibernate A (BrainBits) containing 2% B27 supplement and 2 mM L-glutamine (Invitrogen). Isolated tissue was placed on a wetted filter paper on the Teflon stage of a MacIlwain chopper for coronal sectioning (300 µm thick). From each rabbit hippocampi, 6 inserts of 5–8 sections were prepared. Sections were placed in new dissection media and allowed to rest five minutes on ice before separating and plating on membrane inserts (Millipore). Inserts were placed in 35 mm culture dishes containing 1.1 ml growth media (Neurobasal A with 20% horse serum, 2 mM L-glutamine, 100 U/ml penicillin, and 0.05 µM/ml streptomycin), and warmed 30 minutes prior to plating to ensure complete equilibration. Slices were exposed to a humidified incubator atmosphere (4.5% CO2 and 35°C). Media was changed at DIV1 and at DIV4, slices were switched to a defined medium consisting of Neurobasal A, 2% B27 supplement and 2 mM L-glutamine. Organotypic slices were incubated at DIV 10 with 25 µM of 27-hydroxycholestrol (3 inserts with 5–8 sections from each of the 3 rabbits) or vehicle (0.1% ethanol; 3 inserts with 5–8 sections from each of the 3 rabbits) and harvested after 72 h of treatment. 27-hydroxycholestrol was purchased from Medical Isotopes (Pelham, NH). Circulating 27-hydroxycholestrol levels are 0.15–0.73 µM, and these concentrations can be in the millimolar range in some pathological situations such as atherosclerosis (Brown and Jessup 1999).
Total serum cholesterol was measured in blood collected after overnight fasting from an ear vein immediately before euthanasia with a Flex reagent cartridge and Dimension clinical chemistry system (Dade Behring Inc). Total brain cholesterol levels were measured in all animals as we have previously described (Ghribi et al. 2006a; Ghribi et al. 2006b). In brief, 10 mg of frozen tissue obtained from the hippocampus were homogenized in PBS and sonicated 5 × 5 seconds (Branson Sonifier 150). Chloroform (200 µl) and methanol (200 µl) were added to 100 µl of the tissue homogenates. The bottom chloroform layer was transferred to a new tube, subjected to vacuum, and 50 µl of methanol was added to the residue. Cholesterol concentrations were measured in 50 µl duplicates for all animals using an Amplex Red kit (Molecular Probes Inc).
To determine the effects of the cholesterol-enriched diet on IGF-1 and the downstream target proteins IGF-1R, Akt, IDE, GSK-3α and β, CREB, and Bcl-2, tissue obtained from the hippocampus of control and cholesterol-fed rabbits or from the organotypic slices was gently homogenized using a Teflon homogenizer (Thomas) in T-PER Mammalian protein extraction reagent (Pierce) containing protease (Roche) and phosphatase (Sigma) inhibitors. Protein concentrations were determined with the BCA protein assay reagent (Pierce). For Western blot analyses, proteins (10 µg) were separated by SDS-PAGE (10 % gel), followed by transfer to a polyvinylidene difluoride membrane (Millipore), and were incubated overnight at 4°C with antibodies to IGF-1 (1:100, Santa Cruz Laboratories), IGF-1R (1:100, Novus Biologicals), Akt (1:100, BD Biosciences Pharmingen), p-Akt (1: 100, Abcam), IDE (1:200, Abcam), GSK-3α and β (1: 100, BD Biosciences Pharmingen), p-Tyr GSK-3α and β (1: 200, BD Biosciences Pharmingen), CREB (1:500, Abcam), p-CREB (1:200, Upstate), and Bcl-2 (1:100, Santa Cruz Laboratories). PHF-1 (1:500; gift from Dr. Peter Davis, Albert Einstein College of Medicine), AT8 (1:500, Pierce), or Tau5 (1:500, Pierce) antibodies were used to detect tau phaophorylation. β-actin (1:5000) was used as a gel loading control. For the immunoprecipitation procedure, homogenates from the organotypic slices were diluted to 1mg/ml in PBS and incubated at 4°C overnight with anti-IGF-1 antibody (1:100) using the Catch and Release Reversible Immunoprecipitation System (Upstate) and samples were run on SDS-PAGE (12.5% gel) and incubated overnight at 4° C with antibody to IGF-1 (1: 100 dilution). All the blots were developed with enhanced chemiluminescence (Immun-Star goat anti-mouse IgG detection kit, Bio-Rad). Bands were visualized on a polyvinylidene difluoride membrane and analyzed by LabWorks 4.5 software on a UVP Bioimaging System (Upland). Quantification of results was performed by densitometry and the results analyzed as total integrated densitometric volume values (arbitrary units).
Coronal frozen sections (14 µm) cut at the level of the hippocampus from control and cholesterol-fed rabbits were air-dried, fixed in formalin for 10 minutes, treated with 1% hydrogen peroxide in methanol and incubated with a blocking solution of 5% normal goat serum. Subsequently, sections were reacted overnight at 4°C with p-CREB mAb. After washing with PBS and incubating with the biotinylated secondary antibody, sections were incubated with liquid diaminobenzidine/hydrogen peroxide and rinsed for 3 minutes in PBS, counterstained with hematoxylin, rinsed in distilled water, and dehydrated through 70%, 95% and 100% alcohol. Slides were cleared in xylene and mounted with resinous mounting medium.
Levels of Aβ1–40, Aβ 1–42, and aggregated Aβ in hippocampus as well as in organotypic slices were determined using Immunoassay kits Human β Amyloid 1–40 Colorimetric, Human β Amyloid HS 1–42 Colorimetric, and Human Aggregated Beta Amyloid respectively (Biosource International, Inc.,) according to the manufacturer’s instructions. The quantity of Aβ in each sample was measured in duplicates. Protein concentrations of all samples were determined by standard BCA assay (Pierce). Aβ levels were normalized to total protein content in the samples and expressed as mean ± standard error.
Quantitative data are presented as mean values ± SEM. The significance of differences between the control and cholesterol-fed group was assessed by using the Student's t test, with p < 0.05 considered statistically significant.
Cholesterol concentrations are shown in Table 1. Serum total cholesterol concentrations were dramatically increased in the cholesterol-treated animals in comparison to control rabbits. Conversely to blood cholesterol levels, free cholesterol concentrations in hippocampus did not differ between control and cholesterol-fed rabbits.
IGF-1 signals through IGF-1 R and requires the activation of Akt which involves phosphorylation of its Ser473 residue (Cross et al. 1995). Western blot and densitometric analysis shows a substantial decrease in levels of IGF-1 and IGF-1 R as well as a concomitant decrease in levels of phophorylated Akt protein in the hippocampus of cholesterol-fed rabbits in comparison to control rabbits (Fig. 1).
IGF-1-induced Akt phosphorylation can modulate expression levels of IDE and GSK-3α and β. While IDE and GSK-3α are involved in regulation of Aβ levels, GSK-3β is suggested to phophorylate tau in Alzheimer’s disease. GSK-3α and β are activated by phosphorylation at Tyr276 and Tyr216 residues respectively. Our results show that reduction in phosphorylated Akt levels in hippocampus of cholesterol-fed rabbits was accompanied by a decrease in expression levels of IDE and an increase in levels of active p-Tyr GSK-3α and β (Fig.2).
ELISA assay was used to determine the extent to which decreased IDE and increased p-Tyr276 GSK-3α levels correlate with increased Aβ levels. We found that the total amount of guanidine-solubilized Aβ1–40 and Aβ1–42 was significantly increased in cholesterol-fed rabbits in comparison to control rabbits (Fig. 3). There was also a two-fold increase in Aβ42:Aβ40 ratio (Fig. 3). Aggregated Aβ levels were significantly higher in cholesterol-fed rabbits in comparison to control rabbits (Fig. 3). Our data suggests that cholesterol-enriched diet increases Aβ levels in rabbit by mechanisms that may involve reduction in the degradation of Aβ peptide by IDE as well as a p-Tyr276 GSK-3α-dependent increased production.
In order to determine the extent to which increased levels of p-Tyr216 GSK-3β may increase phosphorylation of tau protein, we used two antibodies: PHF-1 mAb which detects tau phosphorylated at Ser396, and Ser404, and AT8 which detects tau phosphorylated at Ser202 and Thr205. Western blotting demonstrated no difference in immunoreactivity to both PHF-1 and AT8 in the hippocampus from either the control or cholesterol-treated animals (Fig. 3). These results suggest that activation of GSK-3β is not associated with increased phosphorylation of tau following 12 weeks of a 2% cholesterol-enriched diet in rabbit hippocampus.
We applied immunohistochemistry analysis to determine the regional distribution of p-CREB in the hippocampus (Fig. 4a). The immunoreactivity for p-CREB antibody is found predominantly in the dentate gyrus (arrow) of cholesterol-fed rabbits (Fig. 4a, panel B). In control rabbits, only a few scattered cells exhibit staining with p-CREB antibody (Fig. 4a, panel A). As shown in the high magnification microphotographs, p-CREB staining is predominantly nuclear (Fig. 4a, panel D).
It is known that activation of CREB and its translocation into the nucleus increases levels of the anti-apoptotic Bcl-2 (Ji et al. 1996; Riccio et al. 1999). Our Western blot results demonstrated an increase in both p-CREB and BCl-2 protein levels in the hippocampus of cholesterol-fed rabbits in comparison to control rabbits (Fig. 4b).
While the levels of Aβ40 and Aβ42 were undetectable in the slices using ELISA assay, there was a tendency toward increased levels of these peptides in the media of slices treated with 27-hydroxycholesterol in comparison to untreated slices (Table II). Notably, levels of aggregated Aβ were significantly higher in slices treated with cholesterol metabolite 27-hydroxycholesterol than control slices (Table II).
Incubation of organotypic hippocampal slices with 27-hydroxycholesterol reduces levels of IGF-1 and p-Akt, and increases levels of p-Tyr GSK-3α (Fig. 5). These results are similar to those obtained from hippocampus in rabbits fed with the cholesterol-enriched diet (Fig. 1 and Fig. 2). In contrast to the cholesterol-enriched diet, 27-hydroxycholesterol did not significantly reduce levels of IDE in organotypic slices; although a trend to a decrease in IDE is noticeable (Fig. 5). These results strongly indicate that 27-hydroxycholesterol reproduce most of the effects of the cholesterol-enriched diet on the IGF-1 signaling pathways.
The present study was designed to characterize hypercholesterolemia-associated alterations in levels of IGF-1 and related-downstream proteins that regulate Aβ production, tau phosphorylation, and cell survival in the hippocampus of rabbits fed with a cholesterol-enriched diet, a model system for sporadic AD (Sparks et al. 1994). These studies demonstrated for the first time that a cholesterol-enriched diet and subsequent hypercholesterolemia alter the IGF-1 signaling pathway, decrease IDE, increase active p-Tyr276 GSK-3α levels, and lead to increased levels of Aβ in rabbit hippocampus. These changes were associated with the phosphorylation of CREB and the up-regulation of the anti-apoptotic protein Bcl-2, events that may represent a defensive mechanism to prevent cell death.
Cholesterol in the circulation normally does not enter the brain because of the impermeability of the BBB to lipoproteins that carry cholesterol. We also show that increased levels of circulating cholesterol did not alter brain cholesterol levels. The above reasons make it therefore difficult to understand how increased cholesterol levels in the blood trigger Alzheimer’s disease-like pathological hallmarks in the brain.
Studies from our laboratory (Ghribi et al., 2006b) have provided evidence that hypercholesterolemia compromises BBB in rabbits fed with a cholesterol-enriched diet. Conversely to cholesterol, some of its circulating metabolites, such as 27-hydroxycholesterol, have been shown to cross into the brain (Heverin et al. 2005). It is tempting to speculate that levels of cholesterol metabolites are expected to increase during hypercholesterolemia and their entry into the brain may be further facilitated by a leaky BBB. Excess levels of circulating cholesterol metabolites in the brain may therefore trigger a series of events that enhance the production and/or decrease the degradation of Aβ peptides. Our results in organotypic hippocampal slices show that 27-hydroxycholesterol, similary to the cholesterol-enriched diet, increased aggregated Aβ levels. While both the cholesterol-enriched diet and 27-hydroxycholesterol reduced IGF-1 and p-Akt levels and increased active p-Tyr276 GSK-3α, only the former significantly reduced IDE levels. These results indicate that 27-hydroxycholesterol reproduces most of the effects of hypercholesterolemia, and suggest that this oxysterol might be the link between high circulating cholesterol levels and AD-like pathology in the brain. However, levels of 27-hydroxycholesterol in the blood as well as in the brains of cholesterol-fed rabbits remain to be measured.
Various proteins are suggested to play a role in the production and degradation of Aβ in the brain including IDE and GSK-3α. While IDE participates in the degradation of Aβ (for review see Qiu and Folstein 2006), GSK-3α is required for Aβ generation (Phiel et al. 2003). IDE activity has been demonstrated to decrease in a transgenic mouse models for Alzheimer’s disease fed a high fat diet (Ho et al. 2004), as well as in people with a greater risk of developing Alzheimer disease (Zhao et al. 2007). IDE and GSK-3α are both target proteins for the neurotrophic factor IGF-1. IGF-1 levels were shown to dramatically decline in Alzheimer’s disease, and IGF-positive neurons were less abundant in Alzheimer’s disease brains (Steen et al. 2005). IGF-1 has been demonstrated to reduce Aβ production in human SH-SY5Y neuroblastoma cells by mechanisms involving increases in α-secretase processing of endogenous APP (Adlerz et al., 2007). Reduction in IGF-1 levels may therefore limit processing of APP to the non-amyloidogenic products. We demonstrate in the present study that both the cholesterol-enriched diet and 27-hydroxycholesterol reduce levels of IGF-1. Reduction in IGF-1 levels is accompanied by increased levels of GSK-3α and reduced levels of IDE (although not statistically significant with 27-hydroxycholesterol). Our results suggest that reduced levels of IGF-1, increased levels of GSK-3α, and reduced levels of IDE may play a major role in the accumulation of Aβ we showed in vivo and in the organotypic slices.
Active p-Tyr216GSK-3β has multifaceted roles in cellular signaling including pro-apoptotic effects (Jope and Bijur 2002; Li et al. 2002; Song et al. 2002). p-Tyr216GSK-3β can phosphorylate CREB which is stimulated after phosphorylation (Frame and Cohen 2001). Phosphorylation of CREB plays a major role in promoting cell survival and has been demonstrated to protect cells following various insults including ischemia (Sasaki et al. 2007). The mechanisms by which CREB protects against toxic stimuli are diverse and may include overexpression of the anti-apoptotic Bcl-2 (Wilson et al. 1996). Cells with enhanced CREB activity have increased Bcl-2 promoter activity and high cell survival levels (Pugazhenthi et al. 2000). Our results demonstrate that hypercholesterolemia increases phosphorylation of CREB as well as Bcl-2 levels. The increase in levels of p-CREB may have recruited Bcl-2 to prevent apoptosis that may result from increased levels of active p-GSK-3β.
In summary, our data demonstrated that hypercholesterolemia alters the IGF-1 signaling pathway, decreases IDE and increases active GSK-3α levels. These effects were accompanied by increased Aβ levels, suggesting that hypercholesterolemia-induced Aβ accumulation may involve reduced degradation by IDE and increased production by GSK-3α of Aβ peptide. We further demonstrated that 27-hydroxycholesterol reproduces most of the effect of hypercholesterolemia, suggesting that this oxysterol may be the intermediate that links high blood cholesterol levels to AD-like pathology in the brain. Our results add new insight into the cellular mechanisms by which a cholesterol-enriched diet causes pathological hallmarks in rabbit brains.
This work was supported by a Grant from the National Center for Research Resources (5P20RR017699, Centers of Biomedical Research Excellence).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.