- Email: [email protected]
Impact of ApoB-100 expression on cognition and brain pathology in wild-type and hAPPsl mice
Neurobiology of Aging 34 (2013) 2379e2388
Contents lists available at SciVerse ScienceDirect
Neurobiology of Aging journal homepage: www.elsevier.com/locate/neuaging
Impact of ApoB-100 expression on cognition and brain pathology in wild-type and hAPPsl mice Tina Löffler a, b, *, Stefanie Flunkert a, Daniel Havas a, Miklós Sántha c, Birgit Hutter-Paier a, Ernst Steyrer b, Manfred Windisch a a b c
QPS Austria, Grambach, Austria Institute of Molecular Biology and Biochemistry, Medical University Graz, Graz, Austria Institute of Biochemistry, Biological Research Center of the Hungarian Academy of Sciences, Szeged, Hungary
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 January 2013 Received in revised form 12 March 2013 Accepted 3 April 2013 Available online 3 May 2013
During their lifetime, people are commonly exposed to several vascular risk factors that may affect brain ageing and cognitive function. In the last few years, increasing evidence suggests that pathological plasma lipid profiles contribute to the pathogenesis of late-onset Alzheimer’s disease. Importantly, hypercholesterolemia, especially elevated low-density lipoprotein cholesterol values, that is, increased apolipoprotein B-100 (ApoB-100) levels, represents an independent risk factor. In this study, the effects of ApoB-100 overexpression, either alone or in combination with cerebral expression of human amyloid precursor protein (hAPP), on cognitive functions and brain pathology were assessed. Our results show that ApoB-100 overexpression induces memory decline and increases cerebral lipid peroxidation and amyloid beta levels compared to those in wild-type animals. Although double-transgenic ApoBxAPP animals did not develop more distinct behavioral deficits than single-transgenic hAPP littermates, hApoB-100 expression caused additional pathophysiological features, such as high LDL and low HDL-cholesterol levels, increased lipid peroxidation, and pronounced ApoB-100 accumulation in cerebral vessels. Thus, our results indicate that ApoBxAPP mice might better reflect the situation of elderly humans than hAPPsl overexpression alone. Ó 2013 Elsevier Inc. All rights reserved.
Keywords: Alzheimer’s disease Apolipoprotein B-100 Hyperlipidemia Vascular risk factor Transgenic mice Brain
1. Introduction Alzheimer’s disease (AD), a progressive neurodegenerative disease that is highly associated with accumulation and deposition of amyloid b (Abeta) peptides, is the most common form of dementia (Burns et al., 2002; Ferri et al., 2005). Generally, the disease can be separated into (1) heritable early-onset AD (EOAD) linked to mutations in the genes for Abeta Precursor Protein (APP) as well as presenilins (Tanzi and Bertram, 2001); and (2) so-called late-onset AD (LOAD) with unknown origin, which accounts for more than 95% of all AD cases (Harman, 2006). Especially because of the genes found to be affected in EOAD patients, the main research focus was drawn on the investigation of Abeta generation, processing, and clearance. Although the approved “amyloid cascade hypothesis” implies an imbalance between increased production and decreased clearance of Abeta as the triggering factor in AD (Hardy and Selkoe, 2002), the underlying causes for the shifted balance in LOAD remain unclear. * Corresponding author at: Institute of Molecular Biology and Biochemistry, Medical University Graz, Harrachgasse 21, 8010 Graz, Austria. Tel.: þ43 316 258 111 215; fax: þ43 316 258 111 300. E-mail address: Tina.Loeffl[email protected] (T. Löffler). 0197-4580/$ e see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neurobiolaging.2013.04.008
In the last years, increasing evidence suggests that hypercholesterolemia and other vascular factors may contribute to the pathogenesis of LOAD (Humpel, 2011). Notably, AD, atherosclerosis, and especially coronary artery disease (CAD) share many risk factors, such as diabetes, hypertension, hyperlipidemia, and the inheritance of an apolipoprotein E4 (ApoE4) allele (Corder et al., 1993; Jarvik et al., 1995; Skoog et al., 1999). In addition to the ApoE genotype, apolipoprotein B (ApoB) is also associated with both CAD and AD (Kuo et al., 1998; Martins et al., 2009). ApoB is known as the primary apolipoprotein of cholesterol-carrying, lowdensity lipoproteins (LDL) and triglyceride-rich, very-low-density lipoproteins (VLDL). CAD and AD patients share a very similar pathological plasma lipid profile, exhibiting increased levels of LDL along with decreased high-density lipoprotein (HDL) levels (Kuo et al., 1998; Sparks et al., 2000). Intriguingly, a post-mortem study of AD patients showed that LDL and ApoB levels positively correlate with brain Abeta42 levels (Kuo et al., 1998). In addition, several cerebrovascular abnormalities have been identified in AD brains: decreased microvascular density, an impaired bloodebrain barrier, and cerebral amyloid angiopathy (CAA) (Farkas and Luiten, 2001). So far it is still under debate as to what extent these pathologies contribute to the development of LOAD and how they influence the time course of progression. In this study, the effects of ApoB-100
2380
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
expression, either alone or in combination with cerebral expression of human amyloid precursor protein (hAPP), on cognitive functions and brain pathology were assessed. ApoB-100 animals, overexpressing the entire 43-kb human ApoB-100 gene including its natural human promoter (Bjelik et al., 2006; Callow et al., 1994) were originally designed as a model of hyperlipidemia and atherosclerosis. Because the murine plasma lipid profile has a very low LDL to HDL ratio compared to the human profile, wild-type mice are protected against hypercholesterolemia and are usually not affected by atherosclerosis (Breslow, 1996). It has been shown that overexpression of human ApoB-100 in mice shifted the lipoprotein profile to a more atherogenic phenotype. Accordingly, atherosclerotic lesions were detected at branch points of the aorta, especially as a consequence of a high-cholesterol diet (Callow et al., 1995). Recent publications showed that ApoB-100eoverexpressing animals display cerebral microvascular lesions, changes in APP metabolism, and some plaque-like accumulations in their brains (Bereczki et al., 2008; Bjelik et al., 2006; Sule et al., 2009). In contrast, the hAPPsl transgenic mouse line is a well-described model of familial AD. These animals are highly overexpressing human APP751 with London (V717I) and Swedish (KM670671NL) mutations under the control of the murine Thy-1 promoter, as first described by Rockenstein et al. (2001). The hAPPsl mice display severe brain amyloid pathology associated with early-onset progressive memory deficits (Havas et al., 2011). The aim of the present study was to crossbreed hAPPsl with ApoB-100 mice to introduce a vascular risk factor into the existing AD mouse model to better mimic the situation in elderly people and to investigate the interplay between the 2 pathologies. Effects of ApoB-100 overexpression, but also possible synergisms in ApoBxAPP animals, on several biochemical and histological markers as well as the behavioral phenotype, were analyzed in an agedependent manner. 2. Methods
immersion in fresh 4% paraformaldehyde/phosphate-buffered saline, pH 7.4, at room temperature for 1 hour, followed by 24-hour incubation in 15% sucrose solution for cryo-conservation. Hemispheres were stored at 80 C until histological processing. 2.3. Homogenization of frozen brain samples The frozen tissue samples were weighed, and Tissue Homogenization Buffer (THB; 250 mmol/L sucrose, 1 mmol/L ethylenediaminetetraacetic acid (EDTA), 1 mmol/L ethylene glycol tetraacetic acid (EGTA), 20 mmol/L Tris, pH 7.4) including 1x protease inhibitor (Calbiochem, Darmstadt, Germany) was added. For cortex samples, 1 mL THB per 100 mg tissue was added; for hippocampal samples, 2.5 mL was used for 100 mg tissue. The tissue was homogenized with the Tissue Ruptor (Qiagen, Düsseldorf, Germany) for 20 seconds using the highest level. 2.4. Extraction of noneplaque-associated proteins (DEA-Fraction) For extraction of noneplaque-associated proteins, 100 mL of the THB homogenate was mixed with 100 mL diethylamine (DEA) solution (0.4% DEA, 100 mmol/L NaCl). The mixture was centrifuged for 1 hour at 74,200 g, 4 C. A 170-mL quantity of the supernatant was transferred to a 1.5-mL Eppendorf tube and neutralized with 17 mL 0.5 mol/L Tris, pH 6.8. 2.5. Extraction of deposited proteins (FA-Fraction) For extraction of deposited proteins, 100 mL of the THB homogenate was mixed with 220 mL cold formic acid (FA) and sonificated for 1 minute on ice. A 300-mL quantity of this solution was transferred to a centrifugation tube and centrifuged for 1 hour at 74,200 g, 4 C. After centrifugation, 210 mL of the supernatant was transferred to a fresh tube and mixed with 4 mL FA Neutralization Solution (1 mol/L Tris, 0.5 mol/L Na2HPO4, 0.05% NaN3).
2.1. Animals 2.6. Lipid peroxidation measurements By heterozygous crossbreeding of ApoB-100 transgenic (Bjelik et al., 2006) and hAPPsl mice (Rockenstein et al., 2001), both on a C57Bl/6J background, 4 different genotypes were obtained: double transgenic ApoBxAPP, single-transgenic ApoB-100, single-transgenic hAPPsl, and nontransgenic (wild-type [WT]) littermates. From every genotype, 10 male and 10 female mice were tested for behavioral deficits at 4, 6, 9, and 12 months of age. Animals were housed in individually ventilated cages under a constant lightedark cycle (12 hours light/dark). Room temperature and humidity were kept constant at approximately 24 C and 40% to 70%, respectively. Dried, pelleted standard rodent chow (Altromin, Lage, Germany) and normal tap water were available ad libitum. Animal studies conformed to the Austrian guidelines for the care and use of laboratory animals and were approved by the Styrian Government, Austria.
A 50-mL quantity of the THB homogenates was mixed with 50 mL 5% SDS solution including 1x protease inhibitor as well as 1x butylhydroxytoluene (BHT) solution (5 mmol/L) and sonificated for 5 seconds. Malondialdehyde (MDA) was used as a standard at final concentrations between 1 and 20 mmol/L in 1:1 (v/v) mixtures of THB and 5% SDS. The thiobarbituric acid reactive species assay (TBARS-assay) was started by adding 55 mL of 1.33% TBA and 95 mL of 20% acetic acid, pH 3.5, to 100 mL of the prepared samples or standards. After 1 hour incubation at 95 C, 250 mL of n-Butanol/ Pyridin (15:1 v/v) was added, and the solutions were mixed by gently inverting the tubes. After centrifugation for 10 minutes at 4000 g, 200 mL of the upper organic phase was transferred to 1 well of a 96-well plate, and absorbance was measured at 535 nm with a mQuant plate photometer. 2.7. Plasma lipid measurements
2.2. Tissue sampling From every tested mouse, blood and brain tissue were sampled after sedation by standard inhalation anaesthesia. Blood was collected into heparin-coated vials and used to obtain plasma. After blood sampling, mice were transcardially perfused with physiological (0.9%) saline. Thereafter, brains were removed and hemisected. From the left hemisphere, the cortex and hippocampus were dissected and immediately frozen on dry ice and stored at 80 C until biochemical examination. The right hemispheres of all mice were fixed by
For the determination of total cholesterol, triglycerides and HDL-cholesterol Fluitest CHOL, Fluitest TG, and Fluitest HDL direct (Analyticon Biotechnologies, Lichtenfels, Germany) assays, respectively were used. Plasma of all animals was diluted 1:2 with 0.2% NaCl solution and the assays were carried out according to the supplied manuals. LDL cholesterol content was measured with Fluitest LDL direct (Analyticon Biotechnologies, Lichtenfels, Germany) assay, as well as calculated with Friedewald equation (LDL-chol ¼ total chol HDL chol ([TG/5])).
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
2.8. Abeta1e38, Abeta1e40, and Abeta1e42 measurements Samples of 5 female and 5 male animals per genotype were randomly selected for each age. Thus, per age, cortical and hippocampal samples (DEA and FA fraction) of 40 animals were analyzed for Abeta1e38, Abeta1e40, and Abeta1e42 with a 96-well MULTISPOT 4G8 Abeta Triplex Assay (Mesoscale Discovery, Rockville, MD). The immune assay was carried out according to the manual, and plates were analyzed on the Sector Imager. The assay detects human and rodent Abeta species.
2381
of 2 factors (genotype and time) on 1 measured parameter, 2-way ANOVA was carried out. Significance was set at p < 0.05. 3. Results Influences of genotype, age, and sex on all measured parameters were statistically assessed. Because gender differences were not detected in most cases, all graphs display results for a mixed population except as noted otherwise. 3.1. Plasma lipid profile
2.9. Immunohistochemistry After sagittal systematic cryo-cutting of brain hemispheres, five 10-mm-thick mounted sections deriving from 5 different mediolateral levels per animal were stained. For quantification of plaque load sections were stained with an anti-amyloid fibrils LOC antibody recognizing amyloid fibrils and fibrillar oligomers, but not monomers (Millipore, Billerica, MA, AB2287). Region size of hippocampus and cortex was measured and the immunoreactive area within each region was quantified using automated image analysis software (Image Pro Plus, version 6.2). Distribution of ApoB-100 in the brain was visualized by staining with an antibody detecting both human and murine ApoB (Abcam, Cambridge, UK, ab20737). For investigating inflammatory processes, GFAP-IHC (Dako, Glostrup, Denmark, Z0334) for astrocytes as well as CD11b-IHC (Serotec, Kidlington, UK, MCA711) for microglia was performed and quantified, taking account of the outlined region size, with Image Pro Plus. 2.10. Behavioral tests 2.10.1. Contextual fear conditioning The Contextual Fear Conditioning (CFC) task was performed to assess emotional learning and memory. On the first day, the animals were placed in a TSE-System box (TSE Systems, Bad Homburg, Germany). After 5 seconds, a 2-second foot shock [0.5 mA; unconditioned stimulus (US)] was administered. Rodents were removed from the chamber 30 seconds later. On the second day, rodents were returned to the same boxes in which training occurred (context), and freezing behavior was recorded by the automated system. At the end of the 5-minute context test, rodents were returned to their home cages. Low freezing times on the second day are associated with emotional memory impairment. 2.10.2. Morris Water Maze task The Morris Water Maze (MWM) task was used to assess spatial learning (Morris, 1984). Mice have to learn to find a hidden platform in a circular water tank using visual cues. The protocol used in this study contained 4 training days, with 3 trials per day and 1 probe trial 2 hours after the last trial on the fourth day. The water temperature was set to 22 C, and light intensity was dimmed to 40 to 60 lux. After training and probe trial, a visual test was carried out. Mice had to find a marked platform in the middle of the pool. If a mouse did not find the platform within 1 minute, it was excluded from the evaluation. 2.11. Statistical analysis GraphPad Prism 4.03 (GraphPad Software, Inc., La Jolla, CA) was used to graph and analyze data. To compare genotypes, 1-way analysis of variance (ANOVA) was used. To evaluate the influence
Changes in the plasma lipid profile, especially high LDL-to-HDL ratios, are well described risk factors for cardiovascular disease but are currently receiving more and more attention regarding other “age-induced” diseases. HDL- and LDL-cholesterol levels were assessed in all genotypes at 6 months of age. Compared to wild-type animals, LDL cholesterol levels were increased, up to 3.5-fold, in both ApoB100eoverexpressing genotypes, ApoBxAPP and ApoB-100 (Fig. 1A). In contrast, HDL cholesterol (Fig. 1B) was significantly decreased, to 68% and 75% in ApoBxAPP and ApoB-100 animals, respectively compared to wild-type littermates. In addition, plasma triglyceride levels, and also total cholesterol, were increased in ApoBxAPP and ApoB-100 animals compared to wild-type mice (data not shown). The described plasma lipid profile did not significantly change over age. 3.2. Lipid peroxidation Oxidative stress is known to play an important role for the development of atherosclerosis as well as AD. Therefore, malondialdehyde (MDA) levels were measured in cortex and hippocampus as a representative marker for lipid peroxidation. In the cortex, MDA values in ApoBxAPP and ApoB-100 mice increased significantly at 6 months of age (Fig. 1C). Notably, in the hippocampus already at 4 months of age, significantly increased MDA levels were detected in ApoBxAPP and ApoB-100 transgenic animals compared to wild-type mice (Fig. 1D). At 9 months of age, an increase in MDA levels was also seen in hAPPsl single-transgenic mice. Therefore, MDA levels in both cortex and hippocampus of all 3 transgenic models were significantly increased at 9 and 12 months of age compared to wild-type animals. 3.3. Amyloid pathology Abeta quantitations were carried out using 6-, 9-, and 12month-old animals. This assay detects human as well as rodent Abeta species which are merely differing at 3 amino acids close to their N-terminus. While ApoBxAPP and hAPPsl mice display both murine and human Abeta species, ApoB-100 and wild-type animals generate only murine Abeta peptides from endogenous APP. Analysis of plaque-associated Abeta (FA-fraction), from ApoBxAPP and hAPPsl mice showed, because of hAPP overexpression, profoundly increased levels of all Abeta peptides compared to endogenous Abeta in wild-type animals. At 12 months of age, Abeta values increased by 15 times, 300 times, and 100 times for Abeta1e38, Abeta1e40, and Abeta1e42, respectively. To allow visualization of ApoB-100einduced differences between hAPP expressors (ApoBxAPP vs. APP) as well as non-hAPP transgenic mice (ApoB-100 vs. WT), these genotypes were compared separately (Fig. 2). Comparison of Abeta levels in the hippocampus of ApoBxAPP and hAPPsl mice revealed significant differences only in 6-month-old
2382
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
significantly increased in hippocampal samples of ApoB-100 overexpressors compared to wild-type mice, starting at 9 months of age (Fig. 2EeG). This difference persisted up to 12 months. In older hAPP-overexpressing animals (ApoBxAPP as well as hAPPsl), the Abeta1e40 peptide was the most abundant Abeta species, followed by Abeta1e42. Wild-type mice hardly displayed murine Abeta1e42 up to the age of 9 months. Wild-type as well as ApoB-100 animals mainly produced murine Abeta1e38. Similar results were obtained for soluble Abeta measurements (DEA-Fraction), but soluble Abeta levels were about 50 times lower than insoluble species in the FA fraction (data not shown). Immunohistochemical detection of fibrillar Abeta in cortex and hippocampus underlined the results obtained from MSD measurements. Although no differences between ApoBxAPP and hAPPsl mice were obtained, a strong increase in immunoreactivity over age was detected (Fig. 2H), revealing a tremendous plaque burden at 12 months of age. In ApoB-100 and wild-type animals, no plaque formation was detected (data not shown). 3.4. Cerebral distribution of ApoB-100 Brain slices of all genotypes were immunohistochemically analyzed for ApoB with an antibody detecting both murine and human ApoB. Both hApoB-overexpressing genotypes, ApoBxAPP and ApoB100, showed strong accumulation of ApoB in cerebral vessels already at the age of 6 months. Especially in the larger leptomeningeal vessels around the hippocampus, high immunoreactivity was detected (Fig. 3A). Also in deeper vessels, most likely representing venules and capillaries, accumulation and even extravasation of ApoB into brain tissue was discovered (Fig. 3C, E). Because hApoB-100 expression is driven by its natural promoter, expression should occur only in liver and intestine, whereas cerebral ApoB-100 immunoreactivity should develop due to damage in the cerebral vasculature. Expression of hApoB-100 in brains of transgenic animals was excluded by mRNA expression analysis (data not shown). In contrast, wild-type and hAPPsl mice showed almost no cerebral ApoB immunoreactivity (Fig. 3B, D and F). 3.5. Inflammation
Fig. 1. hApoB-100 overexpression leads to an atherogenic plasma lipid profile and increased cerebral lipid peroxidation. (A) Plasma LDL- and (B) HDL-cholesterol of 6-month-old mice. (C and D) Lipid peroxidation shown as malondialdehyde (MDA) levels in cortical (C) and hippocampal (D) samples. n ¼ 18 per group. Statistical analysis plasma lipids: 1-way analysis of variance (ANOVA); lipid peroxidation: 2-way-ANOVA followed by Bonferroni post-test compared to wild-type (WT) mice.
animals. Abeta1e38 levels were significantly higher in ApoBxAPP animals. This difference vanished in older animals (Fig. 2AeC). Sex differences were detected in hAPP expressors (ApoBxAPP and APP) only at the age of 12 months. When evaluated separately, 12-month-old male ApoBxAPP and hAPPsl mice displayed significantly lower Abeta1e40 levels compared to females (Fig. 2D). This gender difference could also be seen for Abeta1e38, but not for Abeta1e42 (data not shown). A direct impact of ApoB-100 overexpression on endogenous Abeta levels was detected by comparison of ApoB-100 to wildtype animals. Intriguingly, all 3 measured Abeta species were
The inflammation status of brains from all genotypes was analyzed by quantitative immunohistochemical detection of glial filbrillary acidic protein (GFAP), a marker for activated astrocytes, and CD11b, implicated in adhesive interactions of microglia in different brain areas. In the cortical region of double transgenic ApoBxAPP and singletransgenic hAPPsl mice, strong GFAP immunoreactivity was detected at 9 and 12 months of age (Fig. 4A). Notably, increased GFAP immunoreactivity in the hippocampus was detected only in 12month-old double transgenic animals (Fig. 4B and KeN). Signs of astrogliosis in the corpus callosum were found only in older ApoBxAPP mice (Fig. 4C). At 12 months of age, an increase in activated microglia could be observed in all investigated brain regions of ApoBxAPP and hAPPsl mice (Fig. 4DeF and GeN). The only statistically significant difference in CD11b immunoreactivity between these 2 genotypes was detected in the cortex at 6 months of age (Fig. 4D). Here, the additional ApoB overexpression in ApoBxAPP mice led to increased microglia activation in the cortex in contrast to 6-month-old hAPPsl mice and nontransgenic littermates. Interestingly, ApoB singletransgenic animals also showed a tendency toward increased microglia activation at 6 months, whereas this difference was not seen in older mice.
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
2383
Fig. 2. Influence of hApoB-100 overexpression on insoluble Abeta levels in the hippocampus (HC) of hAPP expressors, ApoBxAPP, and hAPPsl (AeC) as well as non-hAPP expressors, ApoB, and wild-type (WT) mice (EeG). n ¼ 10 per group. (D) Abeta-40 levels and sex differences in 12-month-old hAPP expressors ApoBxAPP and hAPPsl; 5 male and 5 female animals per genotype. (H) Quantitative immunohistochemical detection of fibrillar Abeta in hAPP expressors ApoBxAPP, and hAPPsl (LOC antibody), n ¼ 6 per group. Statistical analysis: 2-way analysis of variance with Bonferroni post-test.
Activated astrocytes and microglia in the cortex were predominantly found on and around amyloid plaques, as previously described for hAPPsl mice (Rockenstein et al., 2001). 3.6. Behavioral testing To analyze differences in behavior and memory performance among all genotypes, diverse behavioral tests were applied in 4-, 6-, 9-, and 12-month-old animals of both sexes. Before starting with the examination of cognitive performance, general health status as well as activity and anxiety parameters of every animal were assessed in the Irwin test, including vertical pole test and wire suspension time, and Open Field, respectively. With these tests, no
differences between genotypes and sexes could be detected (data not shown). According to these results, bias of cognitive tests by disturbed motor abilities, anxious behavior, or hyperactivity could be excluded. Two cognitive tests were carried out with animals of all genotypes and both sexes, to determine learning and memory abilities. The MWM task was used to assess spatial learning and memory function related to hippocampal conditions. This test showed that ApoBxAPP as well as hAPPsl animals displayed early learning deficits at the age of 6 months, as illustrated by increased escape latencies compared to wild-type animals (Fig. 5B). These significant differences compared to nontransgenic animals were even more
Fig. 3. Cerebral accumulation and extravasation of ApoB in leptomeningeal vessels in the hippocampal area of 6-month-old ApoBxAPP (A) and hAPPsl (B) mice. Accumulated ApoB in deep vessels of ApoBxAPP (C) brains in contrast to those of hAPPsl mice (D). ApoB immunoreactivity of capillaries in the thalamic region of ApoB (E) and wild-type (WT) mice (F). Scale bars ¼ 100 mm.
2384
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
Fig. 4. Immunohistochemical detection of activated astrocytes and microglia by GFAP and CD11b staining. Quantitative evaluation of GFAP immunoreactive area in the cortex (A), hippocampus (B) and corpus callosum (C). Percentage of CD11b immunoreactive area in cortex (D), hippocampus (E), and corpus callosum (F). Statistical analysis: 2-way analysis of variance with Bonferroni post-test. n ¼ 6 per group. Representative pictures of cortex (GeJ) and dentate gyrus (DG) (KeN) of 12-month-old ApoBxAPP (G and K), ApoB (H and L), hAPPsl (I and M), and wild-type (WT) mice (J and N). Red indicates glial fibrillary acidic protein (GFAP); green indicates CD11b. Scale bars ¼ 100 mm.
pronounced in 9- and 12-month-old mice (Fig. 5C and D). At the age of 6 months, ApoB-100 single-transgenic mice did not show any difference compared to wild-type animals in the MWM task. However, 12-month-old ApoB-100 mice showed significantly increased escape latencies on the last training day (Fig. 5D), as well as reduced abidance in the target quadrant during the probe trial (Fig. 5E). This finding is unexpected and underlines the impact of ApoB-100 overexpression on cognitive function. The second cognitive test was the contextual fear conditioning task (CFC), detecting differences in emotional learning. In the CFC, ApoBxAPP and hAPPsl mice also showed first significant memory deficits at the age of 6 months, displayed by
a reduction of the freezing time (Fig. 5F) compared to wild-type animals. This phenotype was stable for both genotypes at 9 and 12 months of age. Most notably, ApoB-100 animals again showed significant differences in emotional learning compared to wild-type mice at the age of 12 months (Fig. 5F). 3.7. Cerebral lipid peroxidation status correlates with memory function To investigate a potential functional link between behavioral data and several biochemical parameters, correlation analyses were carried out.
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
2385
Fig. 5. Assessment of memory decline over age. Morris Water Maze (MWM) test learning curves showing escape latencies during 4 testing days of 4- (A), 6- (B), 9- (C), and 12-monthold (D) mice. (E) Abidance in the target quadrant during the MWM probe trial. (F) Freezing times during CFC task over age. Four-month-old ApoBxAPP n ¼ 18, ApoB n ¼ 19, hAPP n ¼ 16, and wild-type (WT) n ¼ 17; 6-month-old ApoBxAPP n ¼ 18, ApoB n ¼ 20, hAPP n ¼ 19, and WT n ¼ 21; 9-month-old ApoBxAPP n ¼ 20, ApoB n ¼ 22, hAPP n ¼ 21, and WT n ¼ 19; and 12-month-old ApoBxAPP n ¼ 14, ApoB n ¼ 23, hAPP n ¼ 13, and WT n ¼ 22. Statistical analyses: 2-way-ANOVA with Bonferroni post-test compared to WT animals.
Correlation of cortical and hippocampal MDA levels of all 9-month-old animals with their corresponding freezing times obtained from the CFC, revealed significant correlation of these variables (Fig. 6A). These data show that increased MDA levels are associated with a restricted memory performance. Correlations containing Abeta values were done separately for hAPP expressors (ApoBxAPP, hAPPsl) and nonehAPP expressors (ApoB-100 and WT), because of the remarkable differences in Abeta levels between these groups. Correlation analysis of cerebral Abeta values obtained from ApoB100 and wild-type animals to several related behavioral or biochemical variables revealed 2 major findings, both connected to Abeta1e38 in 12-month-old animals. First, a strong proportional link was found between increased insoluble Abeta1e38 and MDA levels, especially in the cortex (Fig. 6B), pointing to an association between Abeta aggregation and lipid peroxidation. Second, a correlation between the CFC freezing time and soluble Abeta1e38 in the DEA fraction became apparent (Fig. 6C). High levels of soluble Abeta1e38 correlated with better memory performance, shown as increased freezing time. The same correlation analyses for ApoBxAPP and hAPPsl mice did not reveal anything of significance. 4. Discussion Humans and mice differ remarkably in regard to plasma lipid profiles. Whereas in mice HDL is the major and LDL the minor cholesterol-carrying lipoprotein, humans display considerably higher amounts of LDL together with low HDL levels. High LDL levels are considered to be a major risk factor for atherosclerosis. Because of their low LDL-to-HDL ratio, wild-type mice are usually not affected
by atherosclerosis and other cardiovascular pathologies (Breslow, 1996). By overexpression of ApoB-100, the murine plasma lipid profile was severely shifted to high LDL and lowered HDL-cholesterol levels, more closely reflecting the human plasma lipid profile (Bereczki et al., 2008; Bjelik et al., 2006). In addition to the well-characterized changes in the plasma lipid profile, the present study shows that hApoB-100 overexpression also plays a role in increasing oxidative stress (lipid peroxidation). In the hippocampus, elevated MDA levels were detected in ApoBxAPP and ApoB-100 mice already at only 4 months of age, and further increased over age. It is known that oxidative modifications of LDL are one of the earliest events in atherosclerosis. Oxidized LDL represents a variety of modifications of both lipid and ApoB components by lipid peroxidation (Bonomini et al., 2008). Most notably, these peroxidation events take place in ApoBxAPP and ApoB-100 mouse brains and are here shown to be associated with memory decline. A synergistic effect of additional APP overexpression on lipid peroxidation could not be observed in ApoBxAPP mice, but also hAPPsl single-transgenic animals displayed increased MDA levels at later time points. In ApoB-100 animals, lipid peroxidation started before the onset of behavioral deficits, whereas in hAPPsl mice the onset of lipid peroxidation appeared after manifestation of memory attenuation. This may indicate that lipid peroxidation plays an initial role in memory decline in ApoB-100 animals but is a downstream effect in hAPPsl transgenic mice. In LOAD patient brains, lipid peroxidation was detected and described as an early event (Markesbery et al., 2005; Yao et al., 2003), rendering ApoBxAPP mice better models than hAPPsl animals regarding oxidative events.
2386
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
Fig. 6. Correlation analyses of behavioral and biochemical data. (A) Negative correlation between cerebral malondialdehyde (MDA) levels and corresponding Contextual Fear Conditioning (CFC) freezing times of all genotypes at 9 months of age; n ¼ 42. (B) Insoluble Abeta1-38 correlates with dedicated cortical MDA levels in 12-month-old non-hAPP expressors, ApoB-100, and WT (pooled); n ¼ 22. (C) Cortical soluble Abeta1-38 correlates with corresponding CFC freezing time of 12-month-old non-hAPP expressors, ApoB-100, and wild-type (WT) mice (pooled); n ¼ 22. Statistical analyses: Pearson correlation.
In AD research, the main theory accounting for neuronal dysfunction, neuronal loss, and as a result also cognitive decline, is the amyloid cascade hypothesis (Hardy and Selkoe, 2002; Kowalska, 2004). Abeta peptides are the main component found in plaques and were shown to be neurotoxic, especially in an oligomerized form (Walsh et al., 2002). Overexpression of hAPPsl led to a profound increase of Abeta secretion and also to a shift from Abeta1e38 towards more toxic Abeta species Abeta1e40 and Abeta1e42. 12 months old female ApoBxAPP and hAPPsl mice showed significantly increased levels of Abeta1e38 and Abeta1e40 compared to males, which is in line with findings from Havas et al. (2011). Male mice seemed to reach a plateau after 9 months, whereas in female brains Abeta accumulation proceeded. That this sex effect was not detected in behavioral examinations is most likely a cause of the nonphysiologically high levels of Abeta. It might be that increased Abeta levels are relevant for cognitive deficits only until a certain concentration is reached, and that every additional accumulation of Abeta does not further change the phenotype. In line with this theory is the fact that 20-fold lower Abeta levels in 6-month-old ApoBxAPP and hAPPsl mice already led to a behavioral phenotype. Also, in human AD patients
alterations in Abeta markers are one of the earliest detectable changes, but seem to have plateaued by the mild cognitive impairment (MCI) stage (Frisoni et al., 2010). This possible “ceiling effect” may also be causative for the lack of correlation between Abeta levels of ApoBxAPP and hAPPsl mice and the corresponding behavioral data. Also a potential effect of additional ApoB-100 overexpression on Abeta levels in hAPPsl mice is probably not detectable, as Abeta levels in hAPP expressors are already very high. Comparison of non-hAPPsl expressors, ApoB-100 and wild-type, revealed a distinct impact of ApoB-100 overexpression on Abeta levels. Although plaque formation was not detected in ApoB-100 animals up to the age of 12 months, all 3 investigated Abeta species were significantly increased compared to wild-type animals at the age of 9 months. Interestingly, Abeta1e38 was elevated not only in young ApoBxAPP animals compared to hAPPsl mice, but was also the most abundant species found in ApoB-100 as well as wild-type brains. In 12-month-old ApoB-100 and wild-type mice, a significant correlation between insoluble Abeta1e38 and MDA levels was detected. It is therefore tempting to assume that increased
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388 Table 1 Onset of different pathologies in all 3 transgenic mouse lines
Increased LDL-cholesterol Decreased HDL-cholesterol Increased Lipid peroxidation Elevated Abeta levels Cerebral ApoB-100 accumulation Increased GFAP IR CTX/HC Increased CD11b IR CTX/HC Cognitive decline
ApoB-100
APP
ApoBxAPP
<6 mo <6 mo 4 mo 9 mo Yes e/e e/e 12 mo
e e 9 mo <6 mo No 9 mo/e 12 mo/12 mo 6 mo
<6 mo <6 mo 4 mo <6 mo Yes 9 mo/12 mo 6 mo/12 mo 6 mo
“Onset” is defined as the time point when the first significant differences in relation to nontransgenic (wild-type) animals were measured. For inflammatory markers, glial fibrillary acidic protein (GFAP), and CD11b, 2 values are shown, the first representing the cortex (CTX), the second the hippocampus (HC).
aggregation of Abeta1e38 is connected to increased lipid peroxidation, either due to an identical underlying cause or to 1 factor being causal for the other. The fact that elevated MDA levels were already detected before the raise of insoluble Abeta1e38 supports the hypothesis that increased lipid peroxidation leads to stronger aggregation of Abeta1e38, but not vice versa. Correlation between high levels of soluble, nonaggregated Abeta1e38 and high CFC freezing times points toward a possible protective role of soluble Abeta1e38, probably against oxidative stress. Notably, Abeta1e38 was poorly investigated in earlier studies, but at least monomeric Abeta1e40 was previously shown to have antioxidative properties against metal-induced oxidative damage (Zou et al., 2002) and to inhibit oxidation of CSF lipoproteins and plasma LDL (Kontush et al., 2001). Findings of the present study for the first time point toward a similar antioxidative effect of soluble Abeta1e38, but this hypothesis needs further evaluation. Possible reasons why the same correlation analyses for ApoBxAPP and hAPPsl mice did not reveal any significance may be that the regulation of hAPP expression is driven by a different promoter (Thy1) or may be caused by differential processing of human mutated APP in contrast to the endogenous murine APP. ApoB-100 overexpression also led to changes in the histological appearance of cerebral vessels. Accumulation and even extravasation of ApoB-100, a 515-kDa protein, was found in cerebral vessels of ApoBxAPP and ApoB animals. This may indicate a disturbance of the bloodebrain barrier. Changes in the lumens of larger arteries due to ApoB accumulation could not be detected (data not shown), but in a previous study it was shown that older ApoB-100 mice exhibit severe changes in the capillary network, which was detected as decreased density and increased capillary lumen diameter (Sule et al., 2009). The authors hypothesized impaired angiogenesis as being responsible for the reduced number of capillaries and the concomitant dilation of capillary lumens as a complementary and compensatory mechanism to maintain the blood supply. In the present study, ApoB accumulation was found in all types of vessels. According to their shape also veins were affected in very high numbers, which suggests a possible impairment in cerebral clearance of metabolites. In general, hApoB-100 overexpression seems to induce perceptible alterations in appearance and probably also function of the cerebral vasculature, partially comparable to those in human AD patients (Farkas and Luiten, 2001; Miyakawa, 2010). In the present study, an increase in activated astrocytes in the corpus callosum was found only in double transgenic ApoBxAPP animals. Notably, increased GFAP immunoreactivity in hippocampus and corpus callosum was the only parameter that pointed toward synergistic effects of both transgenes in ApoBxAPP mice, as it could not be detected in each of the single-transgenic groups. A major outcome of the present study is that hApoB-100 overexpression alone induced cognitive decline. As shown with the CFC
2387
and MWM tests, restricted memory performance of ApoB-100 mice was observed at 12 months of age. Correlation between CFC freezing time and cerebral lipid peroxidation levels suggests that oxidative stress is an important trigger factor for cognitive decline, but also the above described changes in the vasculature may be a cause for worsened memory performance. Changes in the peripheral lipid metabolism and therefore chronic exposure to the risk factor seem to lead to severe oxidative stress, perceptible alterations in the cerebral vasculature, and probably also in the cerebral lipid metabolism, all detrimental to physiological brain function. 5. Conclusion The fact that hApoB-100 overexpression alone was observed to induce memory decline in older animals indicates the importance of a balanced (peripheral) lipid metabolism and an intact vascular network to maintain cognitive function. As no plaque formation could be detected in ApoB single-transgenic animals, they are an interesting model for “hyperlipidemia-induced neurodegeneration” (Lenart et al., 2012) rather than a model for LOAD. Although double transgenic ApoBxAPP animals did not develop more distinct behavioral deficits than single-transgenic hAPPsl littermates, additional pathophysiological features obtained by hApoB-100 expression, such as high LDL and lowered HDL-cholesterol levels, increased cerebral lipid peroxidation, ApoB-100 accumulation in cerebral vessels, and astrogliosis in hippocampus (Table 1), indicate that this combination might better reflect the situation of elderly humans than hAPPsl overexpression alone. Disclosure statement S.F., D.H., B.H.-P., and M.W. are employees of QPS Austria, an international contract research organization in the field of Alzheimer’s disease and other neurodegenerative disorders. This work is part of T.L.’s PhD thesis at the Medical University of Graz. Because the mice were provided by QPS Austria, most parts of the work, especially behavioral tests, were done on the premises of QPS Austria. Acknowledgements The authors greatly thank the whole team of QPS Austria as well as Prof. Wolfgang Sattler and his group from the Medical University of Graz. This work was carried out in the course of the PhD program ‘Neuroscience’ at the Medical University of Graz. References Bereczki, E., Bernat, G., Csont, T., Ferdinandy, P., Scheich, H., Santha, M., 2008. Overexpression of human apolipoprotein B-100 induces severe neurodegeneration in transgenic mice. J. Proteome Res. 7, 2246e2252. Bjelik, A., Bereczki, E., Gonda, S., Juhasz, A., Rimanoczy, A., Zana, M., Csont, T., Pakaski, M., Boda, K., Ferdinandy, P., Dux, L., Janka, Z., Santha, M., Kalman, J., 2006. Human apoB overexpression and a high-cholesterol diet differently modify the brain APP metabolism in the transgenic mouse model of atherosclerosis. Neurochem. Int. 49, 393e400. Bonomini, F., Tengattini, S., Fabiano, A., Bianchi, R., Rezzani, R., 2008. Atherosclerosis and oxidative stress. Histol. Histopathol. 23, 381e390. Breslow, J.L., 1996. Mouse models of atherosclerosis. Science 272, 685e688. Burns, A., Byrne, E.J., Maurer, K., 2002. Alzheimer’s disease. Lancet 360, 163e165. Callow, M.J., Stoltzfus, L.J., Lawn, R.M., Rubin, E.M., 1994. Expression of human apolipoprotein B and assembly of lipoprotein(a) in transgenic mice. Proc. Natl. Acad. Sci. U.S.A. 91, 2130e2134. Callow, M.J., Verstuyft, J., Tangirala, R., Palinski, W., Rubin, E.M., 1995. Atherogenesis in transgenic mice with human apolipoprotein B and lipoprotein (a). J. Clin. Invest. 96, 1639e1646. Corder, E.H., Saunders, A.M., Strittmatter, W.J., Schmechel, D.E., Gaskell, P.C., Small, G.W., Roses, A.D., Haines, J.L., Pericak-Vance, M.A., 1993. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 261, 921e923.
2388
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
Farkas, E., Luiten, P.G., 2001. Cerebral microvascular pathology in aging and Alzheimer’s disease. Prog. Neurobiol. 64, 575e611. Ferri, C.P., Prince, M., Brayne, C., Brodaty, H., Fratiglioni, L., Ganguli, M., Hall, K., Hasegawa, K., Hendrie, H., Huang, Y., Jorm, A., Mathers, C., Menezes, P.R., Rimmer, E., Scazufca, M., 2005. Global prevalence of dementia: a Delphi consensus study. Lancet 366, 2112e2117. Frisoni, G.B., Fox, N.C., Jack Jr., C.R., Scheltens, P., Thompson, P.M., 2010. The clinical use of structural MRI in Alzheimer disease. Nat. Rev. Neurol. 6, 67e77. Hardy, J., Selkoe, D.J., 2002. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 297, 353e356. Harman, D., 2006. Alzheimer’s disease pathogenesis: role of aging. Ann. N. Y. Acad. Sci. 1067, 454e460. Havas, D., Hutter-Paier, B., Ubhi, K., Rockenstein, E., Crailsheim, K., Masliah, E., Windisch, M., 2011. A longitudinal study of behavioral deficits in an AbetaPP transgenic mouse model of Alzheimer’s disease. J. Alzheimers Dis. 25, 231e243. Humpel, C., 2011. Chronic mild cerebrovascular dysfunction as a cause for Alzheimer’s disease? Exp. Gerontol. 46, 225e232. Jarvik, G.P., Wijsman, E.M., Kukull, W.A., Schellenberg, G.D., Yu, C., Larson, E.B., 1995. Interactions of apolipoprotein E genotype, total cholesterol level, age, and sex in prediction of Alzheimer’s disease: a case-control study. Neurology 45, 1092e1096. Kontush, A., Berndt, C., Weber, W., Akopyan, V., Arlt, S., Schippling, S., Beisiegel, U., 2001. Amyloid-beta is an antioxidant for lipoproteins in cerebrospinal fluid and plasma. Free Radic. Biol. Med. 30, 119e128. Kowalska, A., 2004. [The beta-amyloid cascade hypothesis: a sequence of events leading to neurodegeneration in Alzheimer’s disease]. Neurologia i neurochirurgia polska 38, 405e411. Kuo, Y.M., Emmerling, M.R., Bisgaier, C.L., Essenburg, A.D., Lampert, H.C., Drumm, D., Roher, A.E., 1998. Elevated low-density lipoprotein in Alzheimer’s disease correlates with brain abeta 1-42 levels. Biochem. Biophys. Res. Commun. 252, 711e715. Lenart, N., Szegedi, V., Juhasz, G., Kasztner, A., Horvath, J., Bereczki, E., Toth, M.E., Penke, B., Santha, M., 2012. Increased tau phosphorylation and impaired
presynaptic function in hypertriglyceridemic ApoB-100 transgenic mice. PLoS One 7, e46007. Markesbery, W.R., Kryscio, R.J., Lovell, M.A., Morrow, J.D., 2005. Lipid peroxidation is an early event in the brain in amnestic mild cognitive impairment. Ann. Neurol. 58, 730e735. Martins, I.J., Berger, T., Sharman, M.J., Verdile, G., Fuller, S.J., Martins, R.N., 2009. Cholesterol metabolism and transport in the pathogenesis of Alzheimer’s disease. J. Neurochem. 111, 1275e1308. Miyakawa, T., 2010. Vascular pathology in Alzheimer’s disease. Psychogeriatrics 10, 39e44. Morris, R., 1984. Developments of a water-maze procedure for studying spatial learning in the rat. J. Neurosci. Methods 11, 47e60. Rockenstein, E., Mallory, M., Mante, M., Sisk, A., Masliaha, E., 2001. Early formation of mature amyloid-beta protein deposits in a mutant APP transgenic model depends on levels of Abeta(1e42). J. Neurosci. Res. 66, 573e582. Skoog, I., Kalaria, R.N., Breteler, M.M., 1999. Vascular factors and Alzheimer disease. Alzheimer Dis. Assoc. Disord. 13 (suppl 3), S106eS114. Sparks, D.L., Martin, T.A., Gross, D.R., Hunsaker 3rd, J.C., 2000. Link between heart disease, cholesterol, and Alzheimer’s disease: a review. Microsc. Res. Techn. 50, 287e290. Sule, Z., Mracsko, E., Bereczki, E., Santha, M., Csont, T., Ferdinandy, P., Bari, F., Farkas, E., 2009. Capillary injury in the ischemic brain of hyperlipidemic, apolipoprotein B-100 transgenic mice. Life Sci. 84, 935e939. Tanzi, R.E., Bertram, L., 2001. New frontiers in Alzheimer’s disease genetics. Neuron 32, 181e184. Walsh, D.M., Klyubin, I., Fadeeva, J.V., Rowan, M.J., Selkoe, D.J., 2002. Amyloid-beta oligomers: their production, toxicity and therapeutic inhibition. Biochem. Soc. Trans. 30, 552e557. Yao, Y., Zhukareva, V., Sung, S., Clark, C.M., Rokach, J., Lee, V.M., Trojanowski, J.Q., Pratico, D., 2003. Enhanced brain levels of 8,12-iso-iPF2alpha-VI differentiate AD from frontotemporal dementia. Neurology. 61, 475e478. Zou, K., Gong, J.S., Yanagisawa, K., Michikawa, M., 2002. A novel function of monomeric amyloid beta-protein serving as an antioxidant molecule against metal-induced oxidative damage. J.Neurosci. 22, 4833e4841.
Contents lists available at SciVerse ScienceDirect
Neurobiology of Aging journal homepage: www.elsevier.com/locate/neuaging
Impact of ApoB-100 expression on cognition and brain pathology in wild-type and hAPPsl mice Tina Löffler a, b, *, Stefanie Flunkert a, Daniel Havas a, Miklós Sántha c, Birgit Hutter-Paier a, Ernst Steyrer b, Manfred Windisch a a b c
QPS Austria, Grambach, Austria Institute of Molecular Biology and Biochemistry, Medical University Graz, Graz, Austria Institute of Biochemistry, Biological Research Center of the Hungarian Academy of Sciences, Szeged, Hungary
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 January 2013 Received in revised form 12 March 2013 Accepted 3 April 2013 Available online 3 May 2013
During their lifetime, people are commonly exposed to several vascular risk factors that may affect brain ageing and cognitive function. In the last few years, increasing evidence suggests that pathological plasma lipid profiles contribute to the pathogenesis of late-onset Alzheimer’s disease. Importantly, hypercholesterolemia, especially elevated low-density lipoprotein cholesterol values, that is, increased apolipoprotein B-100 (ApoB-100) levels, represents an independent risk factor. In this study, the effects of ApoB-100 overexpression, either alone or in combination with cerebral expression of human amyloid precursor protein (hAPP), on cognitive functions and brain pathology were assessed. Our results show that ApoB-100 overexpression induces memory decline and increases cerebral lipid peroxidation and amyloid beta levels compared to those in wild-type animals. Although double-transgenic ApoBxAPP animals did not develop more distinct behavioral deficits than single-transgenic hAPP littermates, hApoB-100 expression caused additional pathophysiological features, such as high LDL and low HDL-cholesterol levels, increased lipid peroxidation, and pronounced ApoB-100 accumulation in cerebral vessels. Thus, our results indicate that ApoBxAPP mice might better reflect the situation of elderly humans than hAPPsl overexpression alone. Ó 2013 Elsevier Inc. All rights reserved.
Keywords: Alzheimer’s disease Apolipoprotein B-100 Hyperlipidemia Vascular risk factor Transgenic mice Brain
1. Introduction Alzheimer’s disease (AD), a progressive neurodegenerative disease that is highly associated with accumulation and deposition of amyloid b (Abeta) peptides, is the most common form of dementia (Burns et al., 2002; Ferri et al., 2005). Generally, the disease can be separated into (1) heritable early-onset AD (EOAD) linked to mutations in the genes for Abeta Precursor Protein (APP) as well as presenilins (Tanzi and Bertram, 2001); and (2) so-called late-onset AD (LOAD) with unknown origin, which accounts for more than 95% of all AD cases (Harman, 2006). Especially because of the genes found to be affected in EOAD patients, the main research focus was drawn on the investigation of Abeta generation, processing, and clearance. Although the approved “amyloid cascade hypothesis” implies an imbalance between increased production and decreased clearance of Abeta as the triggering factor in AD (Hardy and Selkoe, 2002), the underlying causes for the shifted balance in LOAD remain unclear. * Corresponding author at: Institute of Molecular Biology and Biochemistry, Medical University Graz, Harrachgasse 21, 8010 Graz, Austria. Tel.: þ43 316 258 111 215; fax: þ43 316 258 111 300. E-mail address: Tina.Loeffl[email protected] (T. Löffler). 0197-4580/$ e see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neurobiolaging.2013.04.008
In the last years, increasing evidence suggests that hypercholesterolemia and other vascular factors may contribute to the pathogenesis of LOAD (Humpel, 2011). Notably, AD, atherosclerosis, and especially coronary artery disease (CAD) share many risk factors, such as diabetes, hypertension, hyperlipidemia, and the inheritance of an apolipoprotein E4 (ApoE4) allele (Corder et al., 1993; Jarvik et al., 1995; Skoog et al., 1999). In addition to the ApoE genotype, apolipoprotein B (ApoB) is also associated with both CAD and AD (Kuo et al., 1998; Martins et al., 2009). ApoB is known as the primary apolipoprotein of cholesterol-carrying, lowdensity lipoproteins (LDL) and triglyceride-rich, very-low-density lipoproteins (VLDL). CAD and AD patients share a very similar pathological plasma lipid profile, exhibiting increased levels of LDL along with decreased high-density lipoprotein (HDL) levels (Kuo et al., 1998; Sparks et al., 2000). Intriguingly, a post-mortem study of AD patients showed that LDL and ApoB levels positively correlate with brain Abeta42 levels (Kuo et al., 1998). In addition, several cerebrovascular abnormalities have been identified in AD brains: decreased microvascular density, an impaired bloodebrain barrier, and cerebral amyloid angiopathy (CAA) (Farkas and Luiten, 2001). So far it is still under debate as to what extent these pathologies contribute to the development of LOAD and how they influence the time course of progression. In this study, the effects of ApoB-100
2380
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
expression, either alone or in combination with cerebral expression of human amyloid precursor protein (hAPP), on cognitive functions and brain pathology were assessed. ApoB-100 animals, overexpressing the entire 43-kb human ApoB-100 gene including its natural human promoter (Bjelik et al., 2006; Callow et al., 1994) were originally designed as a model of hyperlipidemia and atherosclerosis. Because the murine plasma lipid profile has a very low LDL to HDL ratio compared to the human profile, wild-type mice are protected against hypercholesterolemia and are usually not affected by atherosclerosis (Breslow, 1996). It has been shown that overexpression of human ApoB-100 in mice shifted the lipoprotein profile to a more atherogenic phenotype. Accordingly, atherosclerotic lesions were detected at branch points of the aorta, especially as a consequence of a high-cholesterol diet (Callow et al., 1995). Recent publications showed that ApoB-100eoverexpressing animals display cerebral microvascular lesions, changes in APP metabolism, and some plaque-like accumulations in their brains (Bereczki et al., 2008; Bjelik et al., 2006; Sule et al., 2009). In contrast, the hAPPsl transgenic mouse line is a well-described model of familial AD. These animals are highly overexpressing human APP751 with London (V717I) and Swedish (KM670671NL) mutations under the control of the murine Thy-1 promoter, as first described by Rockenstein et al. (2001). The hAPPsl mice display severe brain amyloid pathology associated with early-onset progressive memory deficits (Havas et al., 2011). The aim of the present study was to crossbreed hAPPsl with ApoB-100 mice to introduce a vascular risk factor into the existing AD mouse model to better mimic the situation in elderly people and to investigate the interplay between the 2 pathologies. Effects of ApoB-100 overexpression, but also possible synergisms in ApoBxAPP animals, on several biochemical and histological markers as well as the behavioral phenotype, were analyzed in an agedependent manner. 2. Methods
immersion in fresh 4% paraformaldehyde/phosphate-buffered saline, pH 7.4, at room temperature for 1 hour, followed by 24-hour incubation in 15% sucrose solution for cryo-conservation. Hemispheres were stored at 80 C until histological processing. 2.3. Homogenization of frozen brain samples The frozen tissue samples were weighed, and Tissue Homogenization Buffer (THB; 250 mmol/L sucrose, 1 mmol/L ethylenediaminetetraacetic acid (EDTA), 1 mmol/L ethylene glycol tetraacetic acid (EGTA), 20 mmol/L Tris, pH 7.4) including 1x protease inhibitor (Calbiochem, Darmstadt, Germany) was added. For cortex samples, 1 mL THB per 100 mg tissue was added; for hippocampal samples, 2.5 mL was used for 100 mg tissue. The tissue was homogenized with the Tissue Ruptor (Qiagen, Düsseldorf, Germany) for 20 seconds using the highest level. 2.4. Extraction of noneplaque-associated proteins (DEA-Fraction) For extraction of noneplaque-associated proteins, 100 mL of the THB homogenate was mixed with 100 mL diethylamine (DEA) solution (0.4% DEA, 100 mmol/L NaCl). The mixture was centrifuged for 1 hour at 74,200 g, 4 C. A 170-mL quantity of the supernatant was transferred to a 1.5-mL Eppendorf tube and neutralized with 17 mL 0.5 mol/L Tris, pH 6.8. 2.5. Extraction of deposited proteins (FA-Fraction) For extraction of deposited proteins, 100 mL of the THB homogenate was mixed with 220 mL cold formic acid (FA) and sonificated for 1 minute on ice. A 300-mL quantity of this solution was transferred to a centrifugation tube and centrifuged for 1 hour at 74,200 g, 4 C. After centrifugation, 210 mL of the supernatant was transferred to a fresh tube and mixed with 4 mL FA Neutralization Solution (1 mol/L Tris, 0.5 mol/L Na2HPO4, 0.05% NaN3).
2.1. Animals 2.6. Lipid peroxidation measurements By heterozygous crossbreeding of ApoB-100 transgenic (Bjelik et al., 2006) and hAPPsl mice (Rockenstein et al., 2001), both on a C57Bl/6J background, 4 different genotypes were obtained: double transgenic ApoBxAPP, single-transgenic ApoB-100, single-transgenic hAPPsl, and nontransgenic (wild-type [WT]) littermates. From every genotype, 10 male and 10 female mice were tested for behavioral deficits at 4, 6, 9, and 12 months of age. Animals were housed in individually ventilated cages under a constant lightedark cycle (12 hours light/dark). Room temperature and humidity were kept constant at approximately 24 C and 40% to 70%, respectively. Dried, pelleted standard rodent chow (Altromin, Lage, Germany) and normal tap water were available ad libitum. Animal studies conformed to the Austrian guidelines for the care and use of laboratory animals and were approved by the Styrian Government, Austria.
A 50-mL quantity of the THB homogenates was mixed with 50 mL 5% SDS solution including 1x protease inhibitor as well as 1x butylhydroxytoluene (BHT) solution (5 mmol/L) and sonificated for 5 seconds. Malondialdehyde (MDA) was used as a standard at final concentrations between 1 and 20 mmol/L in 1:1 (v/v) mixtures of THB and 5% SDS. The thiobarbituric acid reactive species assay (TBARS-assay) was started by adding 55 mL of 1.33% TBA and 95 mL of 20% acetic acid, pH 3.5, to 100 mL of the prepared samples or standards. After 1 hour incubation at 95 C, 250 mL of n-Butanol/ Pyridin (15:1 v/v) was added, and the solutions were mixed by gently inverting the tubes. After centrifugation for 10 minutes at 4000 g, 200 mL of the upper organic phase was transferred to 1 well of a 96-well plate, and absorbance was measured at 535 nm with a mQuant plate photometer. 2.7. Plasma lipid measurements
2.2. Tissue sampling From every tested mouse, blood and brain tissue were sampled after sedation by standard inhalation anaesthesia. Blood was collected into heparin-coated vials and used to obtain plasma. After blood sampling, mice were transcardially perfused with physiological (0.9%) saline. Thereafter, brains were removed and hemisected. From the left hemisphere, the cortex and hippocampus were dissected and immediately frozen on dry ice and stored at 80 C until biochemical examination. The right hemispheres of all mice were fixed by
For the determination of total cholesterol, triglycerides and HDL-cholesterol Fluitest CHOL, Fluitest TG, and Fluitest HDL direct (Analyticon Biotechnologies, Lichtenfels, Germany) assays, respectively were used. Plasma of all animals was diluted 1:2 with 0.2% NaCl solution and the assays were carried out according to the supplied manuals. LDL cholesterol content was measured with Fluitest LDL direct (Analyticon Biotechnologies, Lichtenfels, Germany) assay, as well as calculated with Friedewald equation (LDL-chol ¼ total chol HDL chol ([TG/5])).
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
2.8. Abeta1e38, Abeta1e40, and Abeta1e42 measurements Samples of 5 female and 5 male animals per genotype were randomly selected for each age. Thus, per age, cortical and hippocampal samples (DEA and FA fraction) of 40 animals were analyzed for Abeta1e38, Abeta1e40, and Abeta1e42 with a 96-well MULTISPOT 4G8 Abeta Triplex Assay (Mesoscale Discovery, Rockville, MD). The immune assay was carried out according to the manual, and plates were analyzed on the Sector Imager. The assay detects human and rodent Abeta species.
2381
of 2 factors (genotype and time) on 1 measured parameter, 2-way ANOVA was carried out. Significance was set at p < 0.05. 3. Results Influences of genotype, age, and sex on all measured parameters were statistically assessed. Because gender differences were not detected in most cases, all graphs display results for a mixed population except as noted otherwise. 3.1. Plasma lipid profile
2.9. Immunohistochemistry After sagittal systematic cryo-cutting of brain hemispheres, five 10-mm-thick mounted sections deriving from 5 different mediolateral levels per animal were stained. For quantification of plaque load sections were stained with an anti-amyloid fibrils LOC antibody recognizing amyloid fibrils and fibrillar oligomers, but not monomers (Millipore, Billerica, MA, AB2287). Region size of hippocampus and cortex was measured and the immunoreactive area within each region was quantified using automated image analysis software (Image Pro Plus, version 6.2). Distribution of ApoB-100 in the brain was visualized by staining with an antibody detecting both human and murine ApoB (Abcam, Cambridge, UK, ab20737). For investigating inflammatory processes, GFAP-IHC (Dako, Glostrup, Denmark, Z0334) for astrocytes as well as CD11b-IHC (Serotec, Kidlington, UK, MCA711) for microglia was performed and quantified, taking account of the outlined region size, with Image Pro Plus. 2.10. Behavioral tests 2.10.1. Contextual fear conditioning The Contextual Fear Conditioning (CFC) task was performed to assess emotional learning and memory. On the first day, the animals were placed in a TSE-System box (TSE Systems, Bad Homburg, Germany). After 5 seconds, a 2-second foot shock [0.5 mA; unconditioned stimulus (US)] was administered. Rodents were removed from the chamber 30 seconds later. On the second day, rodents were returned to the same boxes in which training occurred (context), and freezing behavior was recorded by the automated system. At the end of the 5-minute context test, rodents were returned to their home cages. Low freezing times on the second day are associated with emotional memory impairment. 2.10.2. Morris Water Maze task The Morris Water Maze (MWM) task was used to assess spatial learning (Morris, 1984). Mice have to learn to find a hidden platform in a circular water tank using visual cues. The protocol used in this study contained 4 training days, with 3 trials per day and 1 probe trial 2 hours after the last trial on the fourth day. The water temperature was set to 22 C, and light intensity was dimmed to 40 to 60 lux. After training and probe trial, a visual test was carried out. Mice had to find a marked platform in the middle of the pool. If a mouse did not find the platform within 1 minute, it was excluded from the evaluation. 2.11. Statistical analysis GraphPad Prism 4.03 (GraphPad Software, Inc., La Jolla, CA) was used to graph and analyze data. To compare genotypes, 1-way analysis of variance (ANOVA) was used. To evaluate the influence
Changes in the plasma lipid profile, especially high LDL-to-HDL ratios, are well described risk factors for cardiovascular disease but are currently receiving more and more attention regarding other “age-induced” diseases. HDL- and LDL-cholesterol levels were assessed in all genotypes at 6 months of age. Compared to wild-type animals, LDL cholesterol levels were increased, up to 3.5-fold, in both ApoB100eoverexpressing genotypes, ApoBxAPP and ApoB-100 (Fig. 1A). In contrast, HDL cholesterol (Fig. 1B) was significantly decreased, to 68% and 75% in ApoBxAPP and ApoB-100 animals, respectively compared to wild-type littermates. In addition, plasma triglyceride levels, and also total cholesterol, were increased in ApoBxAPP and ApoB-100 animals compared to wild-type mice (data not shown). The described plasma lipid profile did not significantly change over age. 3.2. Lipid peroxidation Oxidative stress is known to play an important role for the development of atherosclerosis as well as AD. Therefore, malondialdehyde (MDA) levels were measured in cortex and hippocampus as a representative marker for lipid peroxidation. In the cortex, MDA values in ApoBxAPP and ApoB-100 mice increased significantly at 6 months of age (Fig. 1C). Notably, in the hippocampus already at 4 months of age, significantly increased MDA levels were detected in ApoBxAPP and ApoB-100 transgenic animals compared to wild-type mice (Fig. 1D). At 9 months of age, an increase in MDA levels was also seen in hAPPsl single-transgenic mice. Therefore, MDA levels in both cortex and hippocampus of all 3 transgenic models were significantly increased at 9 and 12 months of age compared to wild-type animals. 3.3. Amyloid pathology Abeta quantitations were carried out using 6-, 9-, and 12month-old animals. This assay detects human as well as rodent Abeta species which are merely differing at 3 amino acids close to their N-terminus. While ApoBxAPP and hAPPsl mice display both murine and human Abeta species, ApoB-100 and wild-type animals generate only murine Abeta peptides from endogenous APP. Analysis of plaque-associated Abeta (FA-fraction), from ApoBxAPP and hAPPsl mice showed, because of hAPP overexpression, profoundly increased levels of all Abeta peptides compared to endogenous Abeta in wild-type animals. At 12 months of age, Abeta values increased by 15 times, 300 times, and 100 times for Abeta1e38, Abeta1e40, and Abeta1e42, respectively. To allow visualization of ApoB-100einduced differences between hAPP expressors (ApoBxAPP vs. APP) as well as non-hAPP transgenic mice (ApoB-100 vs. WT), these genotypes were compared separately (Fig. 2). Comparison of Abeta levels in the hippocampus of ApoBxAPP and hAPPsl mice revealed significant differences only in 6-month-old
2382
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
significantly increased in hippocampal samples of ApoB-100 overexpressors compared to wild-type mice, starting at 9 months of age (Fig. 2EeG). This difference persisted up to 12 months. In older hAPP-overexpressing animals (ApoBxAPP as well as hAPPsl), the Abeta1e40 peptide was the most abundant Abeta species, followed by Abeta1e42. Wild-type mice hardly displayed murine Abeta1e42 up to the age of 9 months. Wild-type as well as ApoB-100 animals mainly produced murine Abeta1e38. Similar results were obtained for soluble Abeta measurements (DEA-Fraction), but soluble Abeta levels were about 50 times lower than insoluble species in the FA fraction (data not shown). Immunohistochemical detection of fibrillar Abeta in cortex and hippocampus underlined the results obtained from MSD measurements. Although no differences between ApoBxAPP and hAPPsl mice were obtained, a strong increase in immunoreactivity over age was detected (Fig. 2H), revealing a tremendous plaque burden at 12 months of age. In ApoB-100 and wild-type animals, no plaque formation was detected (data not shown). 3.4. Cerebral distribution of ApoB-100 Brain slices of all genotypes were immunohistochemically analyzed for ApoB with an antibody detecting both murine and human ApoB. Both hApoB-overexpressing genotypes, ApoBxAPP and ApoB100, showed strong accumulation of ApoB in cerebral vessels already at the age of 6 months. Especially in the larger leptomeningeal vessels around the hippocampus, high immunoreactivity was detected (Fig. 3A). Also in deeper vessels, most likely representing venules and capillaries, accumulation and even extravasation of ApoB into brain tissue was discovered (Fig. 3C, E). Because hApoB-100 expression is driven by its natural promoter, expression should occur only in liver and intestine, whereas cerebral ApoB-100 immunoreactivity should develop due to damage in the cerebral vasculature. Expression of hApoB-100 in brains of transgenic animals was excluded by mRNA expression analysis (data not shown). In contrast, wild-type and hAPPsl mice showed almost no cerebral ApoB immunoreactivity (Fig. 3B, D and F). 3.5. Inflammation
Fig. 1. hApoB-100 overexpression leads to an atherogenic plasma lipid profile and increased cerebral lipid peroxidation. (A) Plasma LDL- and (B) HDL-cholesterol of 6-month-old mice. (C and D) Lipid peroxidation shown as malondialdehyde (MDA) levels in cortical (C) and hippocampal (D) samples. n ¼ 18 per group. Statistical analysis plasma lipids: 1-way analysis of variance (ANOVA); lipid peroxidation: 2-way-ANOVA followed by Bonferroni post-test compared to wild-type (WT) mice.
animals. Abeta1e38 levels were significantly higher in ApoBxAPP animals. This difference vanished in older animals (Fig. 2AeC). Sex differences were detected in hAPP expressors (ApoBxAPP and APP) only at the age of 12 months. When evaluated separately, 12-month-old male ApoBxAPP and hAPPsl mice displayed significantly lower Abeta1e40 levels compared to females (Fig. 2D). This gender difference could also be seen for Abeta1e38, but not for Abeta1e42 (data not shown). A direct impact of ApoB-100 overexpression on endogenous Abeta levels was detected by comparison of ApoB-100 to wildtype animals. Intriguingly, all 3 measured Abeta species were
The inflammation status of brains from all genotypes was analyzed by quantitative immunohistochemical detection of glial filbrillary acidic protein (GFAP), a marker for activated astrocytes, and CD11b, implicated in adhesive interactions of microglia in different brain areas. In the cortical region of double transgenic ApoBxAPP and singletransgenic hAPPsl mice, strong GFAP immunoreactivity was detected at 9 and 12 months of age (Fig. 4A). Notably, increased GFAP immunoreactivity in the hippocampus was detected only in 12month-old double transgenic animals (Fig. 4B and KeN). Signs of astrogliosis in the corpus callosum were found only in older ApoBxAPP mice (Fig. 4C). At 12 months of age, an increase in activated microglia could be observed in all investigated brain regions of ApoBxAPP and hAPPsl mice (Fig. 4DeF and GeN). The only statistically significant difference in CD11b immunoreactivity between these 2 genotypes was detected in the cortex at 6 months of age (Fig. 4D). Here, the additional ApoB overexpression in ApoBxAPP mice led to increased microglia activation in the cortex in contrast to 6-month-old hAPPsl mice and nontransgenic littermates. Interestingly, ApoB singletransgenic animals also showed a tendency toward increased microglia activation at 6 months, whereas this difference was not seen in older mice.
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
2383
Fig. 2. Influence of hApoB-100 overexpression on insoluble Abeta levels in the hippocampus (HC) of hAPP expressors, ApoBxAPP, and hAPPsl (AeC) as well as non-hAPP expressors, ApoB, and wild-type (WT) mice (EeG). n ¼ 10 per group. (D) Abeta-40 levels and sex differences in 12-month-old hAPP expressors ApoBxAPP and hAPPsl; 5 male and 5 female animals per genotype. (H) Quantitative immunohistochemical detection of fibrillar Abeta in hAPP expressors ApoBxAPP, and hAPPsl (LOC antibody), n ¼ 6 per group. Statistical analysis: 2-way analysis of variance with Bonferroni post-test.
Activated astrocytes and microglia in the cortex were predominantly found on and around amyloid plaques, as previously described for hAPPsl mice (Rockenstein et al., 2001). 3.6. Behavioral testing To analyze differences in behavior and memory performance among all genotypes, diverse behavioral tests were applied in 4-, 6-, 9-, and 12-month-old animals of both sexes. Before starting with the examination of cognitive performance, general health status as well as activity and anxiety parameters of every animal were assessed in the Irwin test, including vertical pole test and wire suspension time, and Open Field, respectively. With these tests, no
differences between genotypes and sexes could be detected (data not shown). According to these results, bias of cognitive tests by disturbed motor abilities, anxious behavior, or hyperactivity could be excluded. Two cognitive tests were carried out with animals of all genotypes and both sexes, to determine learning and memory abilities. The MWM task was used to assess spatial learning and memory function related to hippocampal conditions. This test showed that ApoBxAPP as well as hAPPsl animals displayed early learning deficits at the age of 6 months, as illustrated by increased escape latencies compared to wild-type animals (Fig. 5B). These significant differences compared to nontransgenic animals were even more
Fig. 3. Cerebral accumulation and extravasation of ApoB in leptomeningeal vessels in the hippocampal area of 6-month-old ApoBxAPP (A) and hAPPsl (B) mice. Accumulated ApoB in deep vessels of ApoBxAPP (C) brains in contrast to those of hAPPsl mice (D). ApoB immunoreactivity of capillaries in the thalamic region of ApoB (E) and wild-type (WT) mice (F). Scale bars ¼ 100 mm.
2384
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
Fig. 4. Immunohistochemical detection of activated astrocytes and microglia by GFAP and CD11b staining. Quantitative evaluation of GFAP immunoreactive area in the cortex (A), hippocampus (B) and corpus callosum (C). Percentage of CD11b immunoreactive area in cortex (D), hippocampus (E), and corpus callosum (F). Statistical analysis: 2-way analysis of variance with Bonferroni post-test. n ¼ 6 per group. Representative pictures of cortex (GeJ) and dentate gyrus (DG) (KeN) of 12-month-old ApoBxAPP (G and K), ApoB (H and L), hAPPsl (I and M), and wild-type (WT) mice (J and N). Red indicates glial fibrillary acidic protein (GFAP); green indicates CD11b. Scale bars ¼ 100 mm.
pronounced in 9- and 12-month-old mice (Fig. 5C and D). At the age of 6 months, ApoB-100 single-transgenic mice did not show any difference compared to wild-type animals in the MWM task. However, 12-month-old ApoB-100 mice showed significantly increased escape latencies on the last training day (Fig. 5D), as well as reduced abidance in the target quadrant during the probe trial (Fig. 5E). This finding is unexpected and underlines the impact of ApoB-100 overexpression on cognitive function. The second cognitive test was the contextual fear conditioning task (CFC), detecting differences in emotional learning. In the CFC, ApoBxAPP and hAPPsl mice also showed first significant memory deficits at the age of 6 months, displayed by
a reduction of the freezing time (Fig. 5F) compared to wild-type animals. This phenotype was stable for both genotypes at 9 and 12 months of age. Most notably, ApoB-100 animals again showed significant differences in emotional learning compared to wild-type mice at the age of 12 months (Fig. 5F). 3.7. Cerebral lipid peroxidation status correlates with memory function To investigate a potential functional link between behavioral data and several biochemical parameters, correlation analyses were carried out.
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
2385
Fig. 5. Assessment of memory decline over age. Morris Water Maze (MWM) test learning curves showing escape latencies during 4 testing days of 4- (A), 6- (B), 9- (C), and 12-monthold (D) mice. (E) Abidance in the target quadrant during the MWM probe trial. (F) Freezing times during CFC task over age. Four-month-old ApoBxAPP n ¼ 18, ApoB n ¼ 19, hAPP n ¼ 16, and wild-type (WT) n ¼ 17; 6-month-old ApoBxAPP n ¼ 18, ApoB n ¼ 20, hAPP n ¼ 19, and WT n ¼ 21; 9-month-old ApoBxAPP n ¼ 20, ApoB n ¼ 22, hAPP n ¼ 21, and WT n ¼ 19; and 12-month-old ApoBxAPP n ¼ 14, ApoB n ¼ 23, hAPP n ¼ 13, and WT n ¼ 22. Statistical analyses: 2-way-ANOVA with Bonferroni post-test compared to WT animals.
Correlation of cortical and hippocampal MDA levels of all 9-month-old animals with their corresponding freezing times obtained from the CFC, revealed significant correlation of these variables (Fig. 6A). These data show that increased MDA levels are associated with a restricted memory performance. Correlations containing Abeta values were done separately for hAPP expressors (ApoBxAPP, hAPPsl) and nonehAPP expressors (ApoB-100 and WT), because of the remarkable differences in Abeta levels between these groups. Correlation analysis of cerebral Abeta values obtained from ApoB100 and wild-type animals to several related behavioral or biochemical variables revealed 2 major findings, both connected to Abeta1e38 in 12-month-old animals. First, a strong proportional link was found between increased insoluble Abeta1e38 and MDA levels, especially in the cortex (Fig. 6B), pointing to an association between Abeta aggregation and lipid peroxidation. Second, a correlation between the CFC freezing time and soluble Abeta1e38 in the DEA fraction became apparent (Fig. 6C). High levels of soluble Abeta1e38 correlated with better memory performance, shown as increased freezing time. The same correlation analyses for ApoBxAPP and hAPPsl mice did not reveal anything of significance. 4. Discussion Humans and mice differ remarkably in regard to plasma lipid profiles. Whereas in mice HDL is the major and LDL the minor cholesterol-carrying lipoprotein, humans display considerably higher amounts of LDL together with low HDL levels. High LDL levels are considered to be a major risk factor for atherosclerosis. Because of their low LDL-to-HDL ratio, wild-type mice are usually not affected
by atherosclerosis and other cardiovascular pathologies (Breslow, 1996). By overexpression of ApoB-100, the murine plasma lipid profile was severely shifted to high LDL and lowered HDL-cholesterol levels, more closely reflecting the human plasma lipid profile (Bereczki et al., 2008; Bjelik et al., 2006). In addition to the well-characterized changes in the plasma lipid profile, the present study shows that hApoB-100 overexpression also plays a role in increasing oxidative stress (lipid peroxidation). In the hippocampus, elevated MDA levels were detected in ApoBxAPP and ApoB-100 mice already at only 4 months of age, and further increased over age. It is known that oxidative modifications of LDL are one of the earliest events in atherosclerosis. Oxidized LDL represents a variety of modifications of both lipid and ApoB components by lipid peroxidation (Bonomini et al., 2008). Most notably, these peroxidation events take place in ApoBxAPP and ApoB-100 mouse brains and are here shown to be associated with memory decline. A synergistic effect of additional APP overexpression on lipid peroxidation could not be observed in ApoBxAPP mice, but also hAPPsl single-transgenic animals displayed increased MDA levels at later time points. In ApoB-100 animals, lipid peroxidation started before the onset of behavioral deficits, whereas in hAPPsl mice the onset of lipid peroxidation appeared after manifestation of memory attenuation. This may indicate that lipid peroxidation plays an initial role in memory decline in ApoB-100 animals but is a downstream effect in hAPPsl transgenic mice. In LOAD patient brains, lipid peroxidation was detected and described as an early event (Markesbery et al., 2005; Yao et al., 2003), rendering ApoBxAPP mice better models than hAPPsl animals regarding oxidative events.
2386
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
Fig. 6. Correlation analyses of behavioral and biochemical data. (A) Negative correlation between cerebral malondialdehyde (MDA) levels and corresponding Contextual Fear Conditioning (CFC) freezing times of all genotypes at 9 months of age; n ¼ 42. (B) Insoluble Abeta1-38 correlates with dedicated cortical MDA levels in 12-month-old non-hAPP expressors, ApoB-100, and WT (pooled); n ¼ 22. (C) Cortical soluble Abeta1-38 correlates with corresponding CFC freezing time of 12-month-old non-hAPP expressors, ApoB-100, and wild-type (WT) mice (pooled); n ¼ 22. Statistical analyses: Pearson correlation.
In AD research, the main theory accounting for neuronal dysfunction, neuronal loss, and as a result also cognitive decline, is the amyloid cascade hypothesis (Hardy and Selkoe, 2002; Kowalska, 2004). Abeta peptides are the main component found in plaques and were shown to be neurotoxic, especially in an oligomerized form (Walsh et al., 2002). Overexpression of hAPPsl led to a profound increase of Abeta secretion and also to a shift from Abeta1e38 towards more toxic Abeta species Abeta1e40 and Abeta1e42. 12 months old female ApoBxAPP and hAPPsl mice showed significantly increased levels of Abeta1e38 and Abeta1e40 compared to males, which is in line with findings from Havas et al. (2011). Male mice seemed to reach a plateau after 9 months, whereas in female brains Abeta accumulation proceeded. That this sex effect was not detected in behavioral examinations is most likely a cause of the nonphysiologically high levels of Abeta. It might be that increased Abeta levels are relevant for cognitive deficits only until a certain concentration is reached, and that every additional accumulation of Abeta does not further change the phenotype. In line with this theory is the fact that 20-fold lower Abeta levels in 6-month-old ApoBxAPP and hAPPsl mice already led to a behavioral phenotype. Also, in human AD patients
alterations in Abeta markers are one of the earliest detectable changes, but seem to have plateaued by the mild cognitive impairment (MCI) stage (Frisoni et al., 2010). This possible “ceiling effect” may also be causative for the lack of correlation between Abeta levels of ApoBxAPP and hAPPsl mice and the corresponding behavioral data. Also a potential effect of additional ApoB-100 overexpression on Abeta levels in hAPPsl mice is probably not detectable, as Abeta levels in hAPP expressors are already very high. Comparison of non-hAPPsl expressors, ApoB-100 and wild-type, revealed a distinct impact of ApoB-100 overexpression on Abeta levels. Although plaque formation was not detected in ApoB-100 animals up to the age of 12 months, all 3 investigated Abeta species were significantly increased compared to wild-type animals at the age of 9 months. Interestingly, Abeta1e38 was elevated not only in young ApoBxAPP animals compared to hAPPsl mice, but was also the most abundant species found in ApoB-100 as well as wild-type brains. In 12-month-old ApoB-100 and wild-type mice, a significant correlation between insoluble Abeta1e38 and MDA levels was detected. It is therefore tempting to assume that increased
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388 Table 1 Onset of different pathologies in all 3 transgenic mouse lines
Increased LDL-cholesterol Decreased HDL-cholesterol Increased Lipid peroxidation Elevated Abeta levels Cerebral ApoB-100 accumulation Increased GFAP IR CTX/HC Increased CD11b IR CTX/HC Cognitive decline
ApoB-100
APP
ApoBxAPP
<6 mo <6 mo 4 mo 9 mo Yes e/e e/e 12 mo
e e 9 mo <6 mo No 9 mo/e 12 mo/12 mo 6 mo
<6 mo <6 mo 4 mo <6 mo Yes 9 mo/12 mo 6 mo/12 mo 6 mo
“Onset” is defined as the time point when the first significant differences in relation to nontransgenic (wild-type) animals were measured. For inflammatory markers, glial fibrillary acidic protein (GFAP), and CD11b, 2 values are shown, the first representing the cortex (CTX), the second the hippocampus (HC).
aggregation of Abeta1e38 is connected to increased lipid peroxidation, either due to an identical underlying cause or to 1 factor being causal for the other. The fact that elevated MDA levels were already detected before the raise of insoluble Abeta1e38 supports the hypothesis that increased lipid peroxidation leads to stronger aggregation of Abeta1e38, but not vice versa. Correlation between high levels of soluble, nonaggregated Abeta1e38 and high CFC freezing times points toward a possible protective role of soluble Abeta1e38, probably against oxidative stress. Notably, Abeta1e38 was poorly investigated in earlier studies, but at least monomeric Abeta1e40 was previously shown to have antioxidative properties against metal-induced oxidative damage (Zou et al., 2002) and to inhibit oxidation of CSF lipoproteins and plasma LDL (Kontush et al., 2001). Findings of the present study for the first time point toward a similar antioxidative effect of soluble Abeta1e38, but this hypothesis needs further evaluation. Possible reasons why the same correlation analyses for ApoBxAPP and hAPPsl mice did not reveal any significance may be that the regulation of hAPP expression is driven by a different promoter (Thy1) or may be caused by differential processing of human mutated APP in contrast to the endogenous murine APP. ApoB-100 overexpression also led to changes in the histological appearance of cerebral vessels. Accumulation and even extravasation of ApoB-100, a 515-kDa protein, was found in cerebral vessels of ApoBxAPP and ApoB animals. This may indicate a disturbance of the bloodebrain barrier. Changes in the lumens of larger arteries due to ApoB accumulation could not be detected (data not shown), but in a previous study it was shown that older ApoB-100 mice exhibit severe changes in the capillary network, which was detected as decreased density and increased capillary lumen diameter (Sule et al., 2009). The authors hypothesized impaired angiogenesis as being responsible for the reduced number of capillaries and the concomitant dilation of capillary lumens as a complementary and compensatory mechanism to maintain the blood supply. In the present study, ApoB accumulation was found in all types of vessels. According to their shape also veins were affected in very high numbers, which suggests a possible impairment in cerebral clearance of metabolites. In general, hApoB-100 overexpression seems to induce perceptible alterations in appearance and probably also function of the cerebral vasculature, partially comparable to those in human AD patients (Farkas and Luiten, 2001; Miyakawa, 2010). In the present study, an increase in activated astrocytes in the corpus callosum was found only in double transgenic ApoBxAPP animals. Notably, increased GFAP immunoreactivity in hippocampus and corpus callosum was the only parameter that pointed toward synergistic effects of both transgenes in ApoBxAPP mice, as it could not be detected in each of the single-transgenic groups. A major outcome of the present study is that hApoB-100 overexpression alone induced cognitive decline. As shown with the CFC
2387
and MWM tests, restricted memory performance of ApoB-100 mice was observed at 12 months of age. Correlation between CFC freezing time and cerebral lipid peroxidation levels suggests that oxidative stress is an important trigger factor for cognitive decline, but also the above described changes in the vasculature may be a cause for worsened memory performance. Changes in the peripheral lipid metabolism and therefore chronic exposure to the risk factor seem to lead to severe oxidative stress, perceptible alterations in the cerebral vasculature, and probably also in the cerebral lipid metabolism, all detrimental to physiological brain function. 5. Conclusion The fact that hApoB-100 overexpression alone was observed to induce memory decline in older animals indicates the importance of a balanced (peripheral) lipid metabolism and an intact vascular network to maintain cognitive function. As no plaque formation could be detected in ApoB single-transgenic animals, they are an interesting model for “hyperlipidemia-induced neurodegeneration” (Lenart et al., 2012) rather than a model for LOAD. Although double transgenic ApoBxAPP animals did not develop more distinct behavioral deficits than single-transgenic hAPPsl littermates, additional pathophysiological features obtained by hApoB-100 expression, such as high LDL and lowered HDL-cholesterol levels, increased cerebral lipid peroxidation, ApoB-100 accumulation in cerebral vessels, and astrogliosis in hippocampus (Table 1), indicate that this combination might better reflect the situation of elderly humans than hAPPsl overexpression alone. Disclosure statement S.F., D.H., B.H.-P., and M.W. are employees of QPS Austria, an international contract research organization in the field of Alzheimer’s disease and other neurodegenerative disorders. This work is part of T.L.’s PhD thesis at the Medical University of Graz. Because the mice were provided by QPS Austria, most parts of the work, especially behavioral tests, were done on the premises of QPS Austria. Acknowledgements The authors greatly thank the whole team of QPS Austria as well as Prof. Wolfgang Sattler and his group from the Medical University of Graz. This work was carried out in the course of the PhD program ‘Neuroscience’ at the Medical University of Graz. References Bereczki, E., Bernat, G., Csont, T., Ferdinandy, P., Scheich, H., Santha, M., 2008. Overexpression of human apolipoprotein B-100 induces severe neurodegeneration in transgenic mice. J. Proteome Res. 7, 2246e2252. Bjelik, A., Bereczki, E., Gonda, S., Juhasz, A., Rimanoczy, A., Zana, M., Csont, T., Pakaski, M., Boda, K., Ferdinandy, P., Dux, L., Janka, Z., Santha, M., Kalman, J., 2006. Human apoB overexpression and a high-cholesterol diet differently modify the brain APP metabolism in the transgenic mouse model of atherosclerosis. Neurochem. Int. 49, 393e400. Bonomini, F., Tengattini, S., Fabiano, A., Bianchi, R., Rezzani, R., 2008. Atherosclerosis and oxidative stress. Histol. Histopathol. 23, 381e390. Breslow, J.L., 1996. Mouse models of atherosclerosis. Science 272, 685e688. Burns, A., Byrne, E.J., Maurer, K., 2002. Alzheimer’s disease. Lancet 360, 163e165. Callow, M.J., Stoltzfus, L.J., Lawn, R.M., Rubin, E.M., 1994. Expression of human apolipoprotein B and assembly of lipoprotein(a) in transgenic mice. Proc. Natl. Acad. Sci. U.S.A. 91, 2130e2134. Callow, M.J., Verstuyft, J., Tangirala, R., Palinski, W., Rubin, E.M., 1995. Atherogenesis in transgenic mice with human apolipoprotein B and lipoprotein (a). J. Clin. Invest. 96, 1639e1646. Corder, E.H., Saunders, A.M., Strittmatter, W.J., Schmechel, D.E., Gaskell, P.C., Small, G.W., Roses, A.D., Haines, J.L., Pericak-Vance, M.A., 1993. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 261, 921e923.
2388
T. Löffler et al. / Neurobiology of Aging 34 (2013) 2379e2388
Farkas, E., Luiten, P.G., 2001. Cerebral microvascular pathology in aging and Alzheimer’s disease. Prog. Neurobiol. 64, 575e611. Ferri, C.P., Prince, M., Brayne, C., Brodaty, H., Fratiglioni, L., Ganguli, M., Hall, K., Hasegawa, K., Hendrie, H., Huang, Y., Jorm, A., Mathers, C., Menezes, P.R., Rimmer, E., Scazufca, M., 2005. Global prevalence of dementia: a Delphi consensus study. Lancet 366, 2112e2117. Frisoni, G.B., Fox, N.C., Jack Jr., C.R., Scheltens, P., Thompson, P.M., 2010. The clinical use of structural MRI in Alzheimer disease. Nat. Rev. Neurol. 6, 67e77. Hardy, J., Selkoe, D.J., 2002. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 297, 353e356. Harman, D., 2006. Alzheimer’s disease pathogenesis: role of aging. Ann. N. Y. Acad. Sci. 1067, 454e460. Havas, D., Hutter-Paier, B., Ubhi, K., Rockenstein, E., Crailsheim, K., Masliah, E., Windisch, M., 2011. A longitudinal study of behavioral deficits in an AbetaPP transgenic mouse model of Alzheimer’s disease. J. Alzheimers Dis. 25, 231e243. Humpel, C., 2011. Chronic mild cerebrovascular dysfunction as a cause for Alzheimer’s disease? Exp. Gerontol. 46, 225e232. Jarvik, G.P., Wijsman, E.M., Kukull, W.A., Schellenberg, G.D., Yu, C., Larson, E.B., 1995. Interactions of apolipoprotein E genotype, total cholesterol level, age, and sex in prediction of Alzheimer’s disease: a case-control study. Neurology 45, 1092e1096. Kontush, A., Berndt, C., Weber, W., Akopyan, V., Arlt, S., Schippling, S., Beisiegel, U., 2001. Amyloid-beta is an antioxidant for lipoproteins in cerebrospinal fluid and plasma. Free Radic. Biol. Med. 30, 119e128. Kowalska, A., 2004. [The beta-amyloid cascade hypothesis: a sequence of events leading to neurodegeneration in Alzheimer’s disease]. Neurologia i neurochirurgia polska 38, 405e411. Kuo, Y.M., Emmerling, M.R., Bisgaier, C.L., Essenburg, A.D., Lampert, H.C., Drumm, D., Roher, A.E., 1998. Elevated low-density lipoprotein in Alzheimer’s disease correlates with brain abeta 1-42 levels. Biochem. Biophys. Res. Commun. 252, 711e715. Lenart, N., Szegedi, V., Juhasz, G., Kasztner, A., Horvath, J., Bereczki, E., Toth, M.E., Penke, B., Santha, M., 2012. Increased tau phosphorylation and impaired
presynaptic function in hypertriglyceridemic ApoB-100 transgenic mice. PLoS One 7, e46007. Markesbery, W.R., Kryscio, R.J., Lovell, M.A., Morrow, J.D., 2005. Lipid peroxidation is an early event in the brain in amnestic mild cognitive impairment. Ann. Neurol. 58, 730e735. Martins, I.J., Berger, T., Sharman, M.J., Verdile, G., Fuller, S.J., Martins, R.N., 2009. Cholesterol metabolism and transport in the pathogenesis of Alzheimer’s disease. J. Neurochem. 111, 1275e1308. Miyakawa, T., 2010. Vascular pathology in Alzheimer’s disease. Psychogeriatrics 10, 39e44. Morris, R., 1984. Developments of a water-maze procedure for studying spatial learning in the rat. J. Neurosci. Methods 11, 47e60. Rockenstein, E., Mallory, M., Mante, M., Sisk, A., Masliaha, E., 2001. Early formation of mature amyloid-beta protein deposits in a mutant APP transgenic model depends on levels of Abeta(1e42). J. Neurosci. Res. 66, 573e582. Skoog, I., Kalaria, R.N., Breteler, M.M., 1999. Vascular factors and Alzheimer disease. Alzheimer Dis. Assoc. Disord. 13 (suppl 3), S106eS114. Sparks, D.L., Martin, T.A., Gross, D.R., Hunsaker 3rd, J.C., 2000. Link between heart disease, cholesterol, and Alzheimer’s disease: a review. Microsc. Res. Techn. 50, 287e290. Sule, Z., Mracsko, E., Bereczki, E., Santha, M., Csont, T., Ferdinandy, P., Bari, F., Farkas, E., 2009. Capillary injury in the ischemic brain of hyperlipidemic, apolipoprotein B-100 transgenic mice. Life Sci. 84, 935e939. Tanzi, R.E., Bertram, L., 2001. New frontiers in Alzheimer’s disease genetics. Neuron 32, 181e184. Walsh, D.M., Klyubin, I., Fadeeva, J.V., Rowan, M.J., Selkoe, D.J., 2002. Amyloid-beta oligomers: their production, toxicity and therapeutic inhibition. Biochem. Soc. Trans. 30, 552e557. Yao, Y., Zhukareva, V., Sung, S., Clark, C.M., Rokach, J., Lee, V.M., Trojanowski, J.Q., Pratico, D., 2003. Enhanced brain levels of 8,12-iso-iPF2alpha-VI differentiate AD from frontotemporal dementia. Neurology. 61, 475e478. Zou, K., Gong, J.S., Yanagisawa, K., Michikawa, M., 2002. A novel function of monomeric amyloid beta-protein serving as an antioxidant molecule against metal-induced oxidative damage. J.Neurosci. 22, 4833e4841.