|Home | About | Journals | Submit | Contact Us | Français|
The ε4 allele of ApoE is associated with an earlier onset and faster progression of Alzheimer’s disease in patients with the familial form of this neurodegenerative condition. Although ApoE4 has been repeatedly associated with altered sphingomyelin and cholesterol levels in tissue culture and rodent models, there has not been a direct quantification of sphingomyelin or sterol levels in the brains of patients with different forms of ApoE. We measured the sphingolipid and sterol content of human brain tissues and found no evidence of perturbed sterol or sphingolipid biochemistry in the brains of individuals expressing ApoE4 who did not have a preexisting neurodegenerative condition. Nevertheless, ApoE4 was associated with gross abnormalities in the sterol and sphingolipid content of numerous brain regions in patients with Alzheimer’s diseease. The findings suggest that ApoE4 may not by itself alter sterol or sphingolipid metabolism in the brain under normal conditions, but that other neuropathologic changes of Alzheimer’s are required to unmask the effect of ApoE4, and to perturb sterol and sphingolipid biochemistry.
ApoE4 has been identified as a major risk factor for Alzheimer’s disease (AD). There are three major isoforms of ApoE that result from single amino acid substitutions: ApoE2 has cysteines at positions 112 and 158; ApoE3 has a cysteine at position 112 and an arginine at position 158; and ApoE4 has arginines at positions 112 and 158. These single amino acid changes result in functional differences in ApoE isoforms, including their relative binding affinities for lipoproteins and ApoE receptors. Although a number of ApoE-initiated receptor signaling events have been identified, the primary function for ApoE is sterol transport. Deficits in sterol transport associated with the ε4 allele of ApoE are thought to be responsible for the enhanced rate of cell death and greater volume loss in hippocampus and amygdala structures of AD patients expressing ApoE4 (Lehtovirta et al. 1996). In tissue culture and rodent models it has been demonstrated that ApoE4 is associated with dysfunctions in neurite outgrowth and synaptogenesis, disordered placement of committed neurons during development, disrupted neurotrophic effects, abnormal regulation of nitric oxide and reduced production of antioxidants (Ignatius et al. 1987; Nathan et al. 1994; Fagan et al. 1996; Narita et al. 1997; Trommsdorff et al. 1999; Mauch et al. 2001; Veinbergs et al. 2001; Brown et al. 2002; Colton et al. 2002; Nathan et al. 2002; Levi et al. 2003; Love et al. 2005). Despite numerous observations suggesting that ApoE4-associated deficits in sterol transport can result in neuronal dysfunction, there has not been direct evidence for abnormal sterol metabolism in the brain parenchyma of AD patients who express ApoE4, although several studies have used indirect measures to correlate brain cholesterol with ApoE polymorphisms. In one study, ApoE4 was associated with reduced levels of cholesterol, phospholipids, and fatty acids in the CSF of AD patients (Mulder et al. 1998) and in a second study, 24S-hydroxycholesterol, a major cholesterol ester that is transported out of brain, was found to be elevated in the CSF and serum of AD patients with an apparent gene dosing effect of ApoE4 (Lutjohann et al. 2000; Papassotiropoulos et al. 2002). Although these findings suggest that ApoE4 perturbs cholesterol metabolism, it is not clear if more or less cholesterol is retained in the brain of AD patients expressing ApoE4. In the present study we directly measured the levels of cholesterol and cholesterol esters, sphingomyelin, ceramide, and 4-hydroxynonenal (4HNE) levels in normal and AD brain to determine whether ApoE polymorphism modifies brain sterol and lipid metabolism. We found that ApoE4 was not associated with changes in brain sterol or sphingolipid levels in brain tissues from cognitively normal subjects, but was associated with large accumulations of cholesterol and sphingolipids in particular brain regions of AD patients. These findings suggest that ApoE4 alone may not appreciably modify brain sterol and sphingolipid metabolism, instead, the effect of ApoE4 in disrupting sterol and sphingolipid balance may be manifest only in the presence of additional perturbations that increase free cholesterol.
All brain tissues were obtained from the Alzheimer's Disease Research Center at Johns Hopkins University School of Medicine. Fresh frozen tissue blocks of middle frontal gyrus (MFG), middle temporal gyrus (MTG), and cerebellum were obtained from normal subjects and cases of definite AD diagnosed according to CERAD criteria (Mirra et al. 1991). Patients were divided into four groups on the basis of neurological status and ApoE genotype determined by polymerase chain reaction (PCR) (Hixson and Vernier 1990): In the AD group, E3 (ApoE 2/3, n = 2 and ApoE 3/3, n = 13) and E4 (ApoE 3/4, n = 10 and ApoE 4/4, n = 5). In the normal group, E3 (ApoE 2/3 n = 5 and ApoE 3/3 n = 15) and ApoE4, (ApoE3/4 n = 4 and ApoE 4/4 n = 2). Subject demographics are listed in Table 1. All tissues were frozen without fixative at −80°C until homogenization.
Total lipids from samples were prepared according to a modified Bligh and Dyer procedure (Haughey et al. 2004). Briefly, each sample was homogenized at room temperature in 10 volumes of deionized water containing 300 nM EDTA, followed by 3 volumes of 100% methanol containing 53 mM ammonium formate and vortexed. Four volumes of chloroform then were added, and the mixture was vortexed and then centrifuged at 1,000 g for 10 minutes. The bottom (chloroform) layer was removed and analyzed by direct injection into a tandem mass spectrometer. Borosilicate-coated glass tubes, pipettes, and injectors were used for lipid extraction.
ESI/MS/MS analyses were performed by methods similar to those used in previous studies (Cutler et al. 2004b; Haughey et al. 2004). Samples were injected using a Harvard Apparatus pump at 5 µl/min into an electrospray ionization (Turbo Ion Spray) Sciex API 3,000 triple stage quadrupole tandem mass spectrometer (ESIMS/MS) from Sciex Inc. (Thornhill, Ontario, Canada) operated in the positive mode. The ion spray voltage (V) was 5,500 at a temperature of 80°C with a nebulizer gas of 8 psi, curtain gas of 8 psi, and the collision gas set at 4 psi. The declustering potential was 80 V, the focusing potential 400 V, the entrance potential –10 V, the collision energy 30 V, and the collision cell exit potential was 18 V. MS/MS scanned from 300 to 2,000 atomic mass units (amu) per second with steps of 0.1 amu. Each species of sphingolipids, cholesterol esters, and lipid peroxides initially was identified by a Q1 mass scan, followed by precursor ion scanning or neutral loss scanning of a purified standard. Samples were injected into the ESI/MS/MS for 3 minutes, where the mass counts accumulated and the sum of the total counts under each peak was used to quantify each species. Cholesterol, and cholesterol ester standards C16:0, C18:0, and C18:1 were purchased from Sigma. Sphingomyelin and ceramides C16:0, C18:0, C20:0, C22:0, C24:0, C24:1 were purchased from Avanti Polar Lipids (Alabaster, AL). 4-Hydroxynonenol and adducts were purchased from Cayman Chemicals (Ann Arbor, MI).
We first grouped data based on disease status (without regard to the patients’ ApoE genotype) and found prominent changes of sphingolipid and sterol levels in the middle frontal gyrus (MFG) of AD patients compared with those of neurologically normal subjects. In the MFG grey matter, we found significant increases of sphingomyelin C16:0, C18:0, C22:0 and C24:0, steraoyl, cholesterol, and the cholesterol ester C18:1 (range 1.8- to 4.9-fold). Exact levels are given in the legend to Figure 1. In contrast, in MFG whie matter, we found no difference between sphingomyelin and sterol levels between AD and normal subjects, but significant decreases of ceramide C16:0, C22:0, C24:1, sterolyl, and sulfatide (range 3.7- to 9.2-fold), Figure 2. These data indicate that in the MFG, ceramide and sterol levels were increased in the cell bodies (grey matter) and ceramide was decreased in the fiber tracks (white matter) of AD brain compared with similar tissues from normal subjects. In the middle temporal gyrus grey matter of AD patients there were trends toward decreased levels of sphingomyelin C24:0, and increased levels of ceramide C24:0 and cholesterol esters, although these values were not significantly different from those in normal subjects (data not shown). There were likewise no significant differences in levels of sphingomyelin, ceramide, or sterols in the temporal white matter or cerebellum of AD compared with levels in normal subjects (data not shown).
We next separated patients based on their ApoE genotype and compared AD patients expressing ApoE3 with AD patients expressing ApoE4 and normal subjects expressing ApoE3 with normal subjects expressing ApoE4. During data analysis a gene dosing effect of ApoE4 was not apparent in AD or normal (see supplementary Figures 1–3), and data were combined into two groups: ApoE3 (ApoE3/3), and ApoE4 (ApoE3/4 & ApoE4/4). Cholesterol accumulated in grey matter and relatively few changes were observed in the white matter of AD patients expressing ApoE4 compared with AD patients expressing ApoE3. In the MFG grey matter of AD patients with an ApoE4 genotype, there were increased levels of cholesterol, the cholesterol esters C16:0, and C18:1 (range 2.6- to 3.4-fold), and in the MTG grey matter, only the cholesterol ester C18:1 was elevated (2.4-fold) in AD patients expressing ApoE4 compared with AD patients expressing ApoE3 (Fig 3B, C). There were no differences in the levels of cholesterol, or cholesterol esters in the MFG or MTG white matter or cerebellum of AD patients expressing either the ε3 or ε4 allele (data not shown). Thus, in the MFG and MTG of AD patients expressing ApoE4, cholesterol and esterfied cholesterol accumulated in the cell bodies but not in the fiber tracks, and no ApoE allele-dependent differences were observed in the white matter of AD patients.
In normal subjects, there were no ApoE allele-associated differences in the levels of cholesterol or esters of cholesterol in any brain region tested (Fig. 3D–F). These unexpected findings suggest that ApoE4 may disrupt brain sterol metabolism only in the setting of a neurodegenerative condition.
When we compared AD patients expressing ApoE4 with AD patients expressing ApoE3, we found ApoE-allele associated changes in the sphingolipid content of both grey and white matter. In the MFG grey matter of ApoE4 AD patients there were decreases in sphingomylein C22:0 (12.5-fold), and C24:0 (6.2-fold; Fig 4A), compared with ApoE4 AD patients. In contrast, ceramides C18:0, C24:1 and sulfatide were increased in the MFG grey matter of ApoE4 AD patients (1.8- to 2.0-fold; Fig 4B). In the MFG white matter, sphingomeylin content was similar in ApoE3 and ApoE4 AD patients (Fig 4C), but ceramide C22:0 and sulfatide were increased (2.6- and 1.7-fold respectively), and ceramide C24:0 was decreased (2.1-fold; Fig 4D). Sphingomyelin and ceramide levels of the MTG grey and white matter were similar in ApoE3 and ApoE4 AD patients and there were likewise no differences in the cerebellar content of sphingomyelin or ceramide in either patient population (data not shown).
In brain tissues from normal subjects, there were no ApoE allele-associated differences in sphingomyelin (Fig 4E, G) or ceramide (Fig 4F, H) levels of MFG grey or white matter. There were likewise no ApoE allele-dependent differences in the MTG grey or white matter or in the cerebellum of normal patients (data not shown).
Disruptions in sterol and sphingolipid metabolism can increase cellular oxidation levels and increase the amounts of lipid peroxidation products, including 4-HNE. Coincident with increased cholesterol, cholesterol esters, and ceramide in the MGF grey matter of AD patients with an ApoE4 genotype, we found a 4.9-fold increase in the lysine adduct of 4-HNE compared to that inAD patients expressing ApoE3 (Fig 5A). Consistent with increased ceramide levels of the MFG white matter of ApoE4 vs. ApoE3 AD patients, there were trends toward increased levels of the lysine and histidine adducts of 4-HNE in ApoE4 AD (Fig 5B). Concentrations of 4-HNE were similar in the MTG white and grey matter of ApoE4 and ApoE3 AD patients (data not shown). In normal subjects there were no significant differences in levels of cellular lipid peroxidation products in any brain region tested of ApoE3 compared with subjects expressing ApoE4 (Fig 5C, D and data not shown).
Our results suggest that the ε4 allele of ApoE is not associated with disordered brain sterol or sphingolipid biochemistry in patients without an underlying neurological disorder. Although ApoE4 binds lipoprotein receptors and clears triglyceride-rich lipoprotein remnants with similar efficacy to ApoE3 (Gregg et al. 1986; Bohnet et al. 1996; Knouff et al. 1999), compared with ApoE3, ApoE4 is less efficient at promoting cholesterol efflux in fibroblasts and astrocytes (Huang et al. 1995; Michikawa et al. 2000), binds preferentially very low density lipoproteins (Weisgraber 1990; Dong et al. 1994) and is recycled inefficiently after internalization (Heeren et al. 2004). Despite these deficits, ApoE4 may effectively transport cholesterol in brain under steady state conditions when cholesterol turnover is low. Perturbations that increase the amount of free cholesterol, however, may overwhelm the ability of ApoE4 to transport cholesterol and result in the accumulation of sterols.
Several lines of evidence suggest that dysregulated sphingolipid metabolism in AD could increase free cholesterol. In AD there is a disruption in brain sphingolipid balance, with increased levels of long chain ceramides ((Han et al. 2002; Cutler et al. 2004a);Fig 1). This shift in sphingolipid balance is thought to be the result of an enhanced sphingomyelinase (Smase) mediated catabolism of sphingomyelin to ceramide (Cutler et al. 2004a). There are at least five types of Smases, including the ubiquitous lysosomal acid SMase, the zinc-dependent secreted acid SMase, a neutral, Mg2+-dependent SMase, a neutral Mg2+-independent SMase, and an alkaline SMase. Neutral sphingomyelinase 2 (nSMase2) is a brain-specific nSMase that is thought to be located on the inner leaflet of the plasma membrane and is activated by phosphatidylserine and cellular stressors including the inflammatory cytokines TNFα, IL-1, Fas-L, UV irradiation, the neurotoxic HIV-proteins gp120 and Tat, and amyloid β (Brann et al. 1999; Sortino et al. 1999; Haughey et al. 2003; Castiglione et al. 2004; Lee et al. 2004; Sanchez-Alavez et al. 2006). Once nSMase2 is active, sphingomyelin shifts to the innerleaflet of the plasma membrane, where it can be catabolized to ceramide. Cholesterol that is normally tightly associated with sphingomyelin is released during the membrane translocation of sphingomyelin, thereby creating a pool of free cholesterol. Under normal conditions, free cholesterol would be esterified or cleared by cholesterol binding proteins, of which ApoE is the most abundant in brain. Because ApoE4 has altered binding capacity, and reduced affinity for low-density lipoproteins (Perugini, et al. 2002; Guillaume, et al. 1996), it may be unable to bind effectively and to transport this enlarged pool of cholesterol. Sterols may then accumulate in the brain parenchyma where they could disrupt biophysical properties of membranes and perturb cellular functions, including amyloid processing.
Cholesterol, sphingomyelin, ceramide, and gangliosides are the primary components of specialized membrane domains known as lipid rafts. Cholesterol levels are thought to modulate the processing of amyloid precursor protein by controlling secretase activity. Current evidence suggests that α-secretase resides in phospholipid-rich domains, while γ and β-secretases are found in lipid rafts (Riddell et al. 2001; Wahrle et al. 2002; Cordy et al. 2003). Experimentally reducing cholesterol in tissue culture has been shown to enhance α-secretase activity and to inhibit γ-secretase activity and reduce the production of pathogenic forms of Aβ (Bodovitz and Klein 1996; Kojro et al. 2001; Ehehalt et al. 2003). Use of a glycosylphosphatidylinositol (GPI) anchor to target β-secretase to lipid rafts increased Aβ formation, and this amyloidogenic effect was inhibited when lipid rafts were disrupted by depleting cholesterol from the membrane (Cordy et al. 2003). Nonetheless, there have been reports that lowering cholesterol increases Aβ production (Abad-Rodriguez et al. 2004), suggesting that maintaining homeostatic levels of cholesterol in brain may be important for normal neuronal function. In addition, Aβ may enhance neutral sphingomyelinase activity, increase ceramide levels, and further the production of pathogenic forms of Aβ (Lee et al. 2004). Thus, one potential mechanism for the earlier onset and faster progression of AD in patients with an ApoE4 genotype may involve a positive feed-back effect where Aβ-induced disruption of sphingolipid metabolism increases free cholesterol that enhances the production of pathogenic forms of Aβ. These observations suggest that pharmacologic modulation of sphingomyelin metabolism may protect neurons and reduce the Aβ load in AD patients with ApoE4 genotype.
The authors would like to thank Dr. Pamela Knapp for editorial assistance. This work was supported by NIH/NIA grants AG023471 and AG24984 to N.J.H., and NIH/NIA grant AG05146 to J.T., and by the Johns Hopkins University Alzheimer’s Disease Research Center.
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.
Disclosure statement. The authors have no actual or potential conflicts of interest related to this manuscript.