Beta-amyloid expression, release and extracellular ... - Nature

Molecular Psychiatry (2008) 13, 939–952 & 2008 Nature Publishing Group All rights reserved 1359-4184/08 $30.00 www.nature.com/mp

ORIGINAL ARTICLE

Beta-amyloid expression, release and extracellular deposition in aged rat brain slices J Marksteiner and C Humpel Laboratory of Psychiatry and Exp. Alzheimer’s Research, Department of General Psychiatry, Innsbruck Medical University, Innsbruck, Austria Alzheimer’s disease (AD) is characterized by b-amyloid plaques, tau pathology, cholinergic cell death and inflammation. The aim of this study was to investigate whether b-amyloid is generated, released and extracellularly deposited in organotypic brain slices. In developing slices, no amyloid-precursor protein (APP) was detectable; however, there was a strong upregulation in aging slices. In such slices, rat b-amyloid(1–42) and -(1–40) peptides were found using four sequence-specific antibodies. APP and b-amyloid were expressed in neurons and to a lesser extent in astrocytes. Beta-amyloid was secreted into the medium. Beta-amyloid was located extracellularly when aging slices were incubated with medium at pH 6.0 including apolipoprotein E4 (ApoE4). It is concluded that aging organotypic brain slices express bamyloid and that acidosis induces cell death with efflux of b-amyloid and extracellular depositions, which is triggered by ApoE4. This novel in vitro model may enable us to investigate further the pathological cascade for AD and may be useful to explore future therapeutics. Molecular Psychiatry (2008) 13, 939–952; doi:10.1038/sj.mp.4002072; published online 21 August 2007 Keywords: Alzheimer; b-amyloid; brain slice; ApoE4

Introduction Amyloid-precursor protein (APP) is a transmembrane glycoprotein that is the precursor of b-amyloid. APP is expressed in a variety of cells both within and outside the nervous system. Its physiological role in the brain is not well understood. In neurons, APP undergoes fast anterograde transport, is then integrated in the plasma membrane and cleaved by asecretase into a 90–100 kDa secreted APP (sAPP).1 The sAPP is secreted into the extracellular space and plays a role in neuroprotection and neuroplasticity by reducing intracellular calcium.2 APP, however, can also be cleaved in an alternative pathway by b- and g-secretases, leading to b-amyloid(1–42) and b-amyloid(1–40), which by itself have different physiological effects in the central nervous system.3,4 Beta-amyloid can aggregate and can induce neurotoxicity. In Alzheimer’s disease (AD) extracellular b-amyloid accumulates in amyloid plaques, which damage nerve fibers and lead to neurodegeneration.1,2,5,6 Trophic or toxic properties of b-amyloid depend on the used concentration or on the form of soluble b-amyloid.7 Most of the in vivo models are based on transgenic human APP overexpressing mice. TransCorrespondence: Dr C Humpel, Department of General Psychiatry, Laboratory of Psychiatry and Exp. Alzheimer’s Research, Medical University, Anichstr. 35, Innsbruck A-6020, Austria. E-mail: [email protected] Received 31 October 2006; revised 14 June 2007; accepted 15 July 2007; published online 21 August 2007

genic mice with mutant genes that predispose to AD provide a model for testing the outcome of single factors in AD neuropathology. In addition, in different in vivo models aggregated b-amyloid is injected into rats or mice to study b-amyloid-induced effects on behavior or expression of genes and proteins. The investigation of b-amyloid in rats is very limited, because rats do not develop b-amyloid plaques, and rat b-amyloid has not been found to aggregate. Organotypic co-cultures provide a possibility to investigate how the cholinergic system interacts with b-amyloid formation. Cholinergic neurons in the human basal forebrain play a key role in neurodegenerative disorders such as AD, in other types of dementia or normal aging. Neurons of the basal nucleus of Meynert (nBM) project to the cerebral cortex, and a cerebrocortical hypofunction stands in a causal relationship to cholinergic neuronal loss.8–10 It seems well established that AD is caused by a synaptic dysfunction.11–13 and secretion of b-amyloid and subsequent plaque formation may result in retrograde-induced cell death of cholinergic neurons.11,14 To investigate mechanisms leading to APP dysregulation, b-amyloid formation and initiation of cholinergic cell death is an important issue and is focused on the development of strategies to rescue neurons from progressive neurodegeneration. In the organotypic brain slice model, we will now investigate whether APP/b-amyloid is expressed, and released by taking advantage of this physiological in vitro model that provides a possibility to study

Beta-amyloid and brain slices J Marksteiner and C Humpel

940

age-dependent effects in an isolated three-dimensional tissue for up to 50 weeks. We will also test whether the presence of apolipoprotein E4 (ApoE4) interacts with b-amyloid pathology. As b-amyloid deposition may also depend on changes in the microenvironment, we will also examine whether low pH alone or in combination with ApoE4 leads to extracellular b-amyloid deposits. This study will also reveal whether b-amyloid pathology is associated with deficits in the cholinergic system. It will turn out whether the organotypic slice model is a valid model for AD that has in common at least some neuropathological changes with the human pathology. It will provide the possibility to explore further the mechanisms of b-amyloid formation in an integrated system.

Materials and methods Organotypic brain slice cultures Organotypic brain slice co-cultures of the nBM and parietal cortex were established as described by us in detail.15,16 Briefly, the nBM of postnatal day 8–10 rats was dissected under aseptic conditions, 400 mm slices were cut with a tissue chopper (McIlwain, Mickle Lab Eng., Guildford, UK), and the slices were placed on a 30 mm diameter Millicell-CM 0.4 mm membrane insert (Millipore, Vienna, Austria). Since the nBM is not a discrete nucleus in rodents, a protocol was established to optimize the dissection of the scattered neurons in this brain region. A detailed schematic dissection drawing is given by us.16 It is important to note that the ipsilateral as well as contralateral nBM of four rats was dissected at the same time, cut on the tissue chopper and all slices pooled in medium. Then the nBM slices were distributed randomly on the membranes across all groups. Next, the parietal cortex was dissected from the same rats, chopped, and the slices were randomly put in direct connection to the nBM slices on the membrane. Such co-slices (4–5 per membrane) were cultured in Petri dishes at 371C and 5% CO2 with 1.2 ml/Petri dish of the following culture medium: 50% MEM/HEPES (Gibco, Lofer, Austria); 10% heat-inactivated horse serum (Gibco, Lifetech, Austria); 25% Hanks’ solution (Gibco); 2 mM NaHCO3 (Merck, Vienna, Austria); 6.5 mg/ml glucose (Merck), 2 mM glutamine, pH 7.2 (Merck). All slices were incubated with 10 ng/ml nerve growth factor (NGF; Sigma-Aldrich, Vienna, Austria) to maintain survival of cholinergic neurons and were cultured for up to 50 weeks. Medium was changed once a week. Twelve-week-old brain slices were further incubated for 4 weeks with or without 10 ng/ml human ApoE4 (Sigma) at pH 7.2 or pH 6.0 (acidosis). In addition, also single slices of nBM or cortex were cultured with or without NGF (6–8 slices/well for immunohistochemistry or 15 slices/well for enzyme-linked immunosorbent assay (ELISA)/Western blot). All experiments conformed to ethical guidelines for the use of animals, as all efforts were made to minimize the number of animals used. Molecular Psychiatry

Postmortem Alzheimer brain Alzheimer brains were collected from postmortem patients, who fitted the NINCDS-ADRA clinical criteria of AD; the average age was 8178 years, the postmortem interval was 1574 h.17 Tissue samples were taken from the dorsolateral, the orbifrontal and entorhinal cortex, and then immediately fixed by immersion in cold 4% paraformaldehyde in phosphate-buffered saline (PBS). Immunohistochemistry Immunohistochemistry was performed as described previously.16 All incubations were performed freefloating at 41C for 2 days including 0.1% Triton, such that the antibodies can penetrate from both sides of the slices and which allows good penetration of the antibody into the brain slices. At the end of the experiment slices were fixed for 3 h at 41C in 4% paraformaldehyde in PBS. Slices were then washed with 0.1% Triton/PBS at room temperature for 30 min and pretreated for 20 min with 20% methanol/ 1%H2O2/PBS (only for 3,30 -diaminobenzidine labeling). After thorough rinsing, the slices were blocked with 20% horse serum/0.2% bovine serum albumin/ PBS and then incubated for 2 days at 41C with primary antibodies. Antibodies used were as follows: goat anti-choline acetyltransferase (Chemicon AB144P; 1:750); mouse anti-APP(643–695) (Chemicon MAB343; 1:250); mouse anti-amyloid-b(17–42) (Chemicon AB5366; 1:100); rabbit anti-amyloid(3–16) (Abcam ab14220; 1:100); rabbit anti-amyloid(37–42) (Chemicon AB5306; 1:100); rabbit anti-amyloid(17– 24) (Chemicon AB5366; 1:500); or mouse-anti-neuronal nuclei (neuN; Chemicon MAB377; 1:100). Since human and rat b-amyloid differ in amino acids 5, 10 and 13 (Figure 5A), the four used anti-b-amyloid antibodies were selected to be highly specific for rat b-amyloid. Then the slices were again washed with PBS and incubated with secondary biotinylated antigoat, or anti-rabbit or anti-mouse antibodies (1:200; Vector Lab., Szabo, Vienna, Austria) for 1 h at room temperature. After being washed, slices were incubated in an avidin–biotin complex solution (ABCElite Vectastain reagent; Vector Lab., USA) for 1 h, then washed in 50 mM Tris-buffered saline, and then the signal was detected by using 0.5 mg/ml 3,30 -diaminobenzidine including 0.003% H2O2 as a substrate in Tris-buffered saline. The slices were mounted on glass slides, air-dried and coverslipped with Entellan (Merck, Darmstadt, Germany). Unspecific labeling was defined by omitting the primary antibody. When fluorescence immunohistochemistry was performed, the methanol pretreatment was omitted and as secondary antibodies Alexa-488 or Alexa-546 (both Invitrogen, Lofer, Austria; 1:400) were used. Rabbit anti-microtubule-associated protein-2 (1:1000; Chemicon) or chicken anti-glial fibrillary acidic protein (GFAP, 1:1000; Chemicon) were used for colocalization experiments. To label nuclei the slices were incubated with 4,6-diamidino-2phenylindole (DAPI, 1:10 000; Sigma) for 30 min.


Immunolabeling was visualized with a Leica DMIRB fluorescence inverse microscope equipped with an Apple computer and Improvison DarkLab software or some sections were visualized using scanning confocal microscopy (LSM 510, Zeiss). In situ hybridization In situ hybridization was performed as described by us.18 Briefly, brain slices were transferred onto slides (ProbeOn slides, Fisher Biotech, Pittsburgh, PA, USA) and frozen in a CO2 stream and stored at 201C until use. Antisense oligonucleotides complementary to nucleotide base pairs 1818–1860 of the choline acyteltransferase (ChAT) gene (43mer)19 and the APP695 gene (40mer)20 were labeled at the 30 end with [a-35S]dATP using terminal deoxyribonucleotidyl transferase (New England Nuclear, Vienna, Austria) and purified using Qiagen nucleotide removal kit (Qiagen, Vienna, Austria). Sections were hybridized at 421C overnight in a humidified chamber with 0.1 ml per slide of a hybridization solution (50% formamide, 4 SSC, 0.02% polyvinylpyrrolidone, 0.02% Ficoll, 0.02% bovine serum albumin, 10% dextran sulfate, 0.5 mg/ml sheared salmon sperm DNA, 1% sarcosyl (N-lauroyl sarcosine), 0.02 M NaPO4 (pH 7.0), 50 mM dithiothreitol) containing 1 107 c.p.m./ml probe. Sections were subsequently rinsed, washed four times (15 min each) at 541C in 1 SSC, cooled to room temperature, dehydrated through 70, 90 and 99% ethanol and air-dried. Sections were dipped in Amersham Hypercoat LM-1 photo emulsion, exposed for 8 weeks at 201C, developed, fixed, and lightly labeled with cresyl violet. Western blot analysis Rat b-amyloid(1–42) (100 ng, Calbiochem, Biotrade, Vienna, Austria) or human secreted b-APP (sbAPP; Sigma) or aggregates or brain slice extracts were analyzed by Western Blot. Brain slices were scraped from the membrane inserts, pooled (15 slices), sonicated (5 pulses) in 150 ml PBS þ protease inhibitor cocktail, centrifuged and the supernatant loaded onto gels. Samples were either analyzed under non-reducing or reducing conditions (samples were heated at 701C, 10 min). For the aggregation experiments, 500 ng b-amyloid was incubated in 100 ml buffer (150 mM NaCl, 20 mM Tris, 6.7 mg/ml BSA710 ng/ml ApoE4, and the pH adjusted to pH 7.2 or 6.0) for 2 weeks at 371C. Samples were either loaded directly onto the gels or centrifuged (15 min, 15 000 g, 201C) and the pellet resuspended in 20 ml running buffer and then loaded. Samples were separated in 10% BisTris (NuPage, Invitrogen) sodium dodecyl sulfatepolyacrylamide gels (Invitrogen) for 35 min at 200 V, and electrotransferred to nylon-PVDF ImmobilonPSQ membranes (Millipore) for 60 min at 30 V with a 40% methanol blotting buffer (Invitrogen). For the detection the Western Breeze Chromogenic System (Invitrogen) was used. Briefly, blots were blocked for 30 min with blocking buffer, then incubated for

90–180 min with the primary anti-b-amyloid (17–24) antibody (Chemicon; 1:1000) or mouse anti-APP A4 (Chemicon MAB348; 1:2000), washed and incubated with alkaline phosphatase-conjugated anti-mouse IgG (Invitrogen) for 30–60 min at room temperature. After being washed, bound antibodies were visualized by p-nitro blue tetrazolium chloride and 5-bromo-4chloro-3-indolyl phosphate (both Roche Molecular Biochemicals, Vienna, Austria).

941

ELISA for rat b-amyloid A novel rat-specific sandwich ELISA for b-amyloid was established. Briefly, DELFIA yellow Immuno Plates (DuPont NEN, Vienna, Austria) were coated for 4 days at 41C with anti-b-amyloid(17–24) antibody (Chemicon; 50 ng/well) in carbonate-coating buffer (pH 9.7). After being washed and blocked, 100 ml standards or brain slice extracts or conditioned media were applied to the wells and incubated for 6 h at room. Conditioned media were collected from three wells at week 2 or 12 or 16, pooled (3 1.2 ml), lyophilized, dissolved in 300 ml assay buffer and neutralized. Wells were again washed and the secondary anti-b-amyloid (3–16) antibody (Abcam; 50 ng/well) was added and incubated overnight. Wells were washed and incubated for 1.5 h with anti-rabbit Europium antibody (DuPont NEN; 20 ng/ well). Then wells were washed and after addition of DELFIA enhancer (Dupont NEN), the Europium fluorescence was measured in a Zenyth 3100 ELISA reader by time-resolved fluorescence. Samples were calculated from the standard curve in the linear range. Measurements and statistics The number of ChAT immunoreactive neurons per slice was counted under the microscope in the nBM part and the APP-positive cells were counted separately in the cortex and nBM. Raw data per slice were averaged consisting of at least six slices and expressed as mean number of cells7s.e.m. per slice. Fluorescence cell counting was performed for DAPI-positive nuclei and neuN-positive neurons on a Leica DMIRB fluorescence inverse microscope equipped with an Apple computer and Improvison DarkLab software. Cell counting was performed on six random fields (200 200 mm each) per slice. Raw data from these field counts were averaged and expressed as cells/ mm2. One experiment consisted of at least four slices per condition. Multistatistical analysis was obtained by one-way analysis of variance, followed by Fisher protected least significant difference post hoc test by comparing controls against the respective treatments, where P < 0.05 represents statistical significance.

Results Cholinergic neurons in brain co-slices of the nBM and cerebral cortex It is well established that the 400 mm brain slices flatten down during the first 2 weeks of incubation and have a thickness of approximately 100 mm. This Molecular Psychiatry


942

flattening is also an internal mean for a good preparation and dissection. Our immunohistochemical analysis showed 98716 (mean7s.e.m.; n = 24; n = represents the number of slices from which mean number is obtained) ChAT-immunopositive neurons in 4-week-old co-slices incubated with NGF (Figures 1a and b) and only 1473 (n = 17) neurons in slices incubated without NGF. Neurons incubated without NGF were clearly smaller and did not have processes. When co-slices were cultured with NGF for up to 32 weeks, the number of cholinergic neurons did not change: 6 weeks: 4478, n = 23; 8 weeks: 38717, n = 12; 12 weeks: 51710, n = 16; 24 weeks: 52723, n = 4 or 32 weeks: 62718, n = 15. In 24-week-old slices, some elongated neurons with small perikarya and few neuronal processes were observed (Figures 1c and d). In 50-week-old co-slices, only a few degenerated ‘ghost-like neurons’ were visible (170.5, n = 8, Figures 1e and f). The number of cholinergic neurons in single nBM slices (54710, n = 11) and co-slices (49710, n = 20) was not signifi-

cantly different (Figures 2a and m) when incubated with NGF for 12 weeks. Very few small cholinergic neurons were found in single slices and co-slices incubated without NGF (Figures 2d and p). No ChAT immunoreactive neurons were found in cortex slices incubated with (Figure 2g) or without (Figure 2j) NGF. However, all slices (single or co-slices) displayed a high number of DAPI-positive nuclei independent of NGF incubation (Figures 2b, e, h, k, n and q). Similarly, all slices contained neuN-positive cells irrespective of whether NGF was or was not administered (Figures 2c, f, i, I, o and r). APP- and b-amyloid-like immunoreactivity in brain slices In 2-week-old co-slices incubated with or without NGF, no APP(643–695)-like immunoreactivity was detected (Figures 3a and c). However, when slices were incubated for 12 or 24 weeks with NGF, a strong immunoreactivity for APP(643–695) was found in cortex (12 weeks: 217726 cells/slice, n = 19; 24

Figure 1 Cholinergic neurons in organotypic brain slices were cultured for 4 (a, b), 24 (c, d) or 50 (e, f) weeks with 10 ng/ml NGF and then analyzed by immunohistochemistry for choline acetyltransferase. The nBM (left) and cortex (right, ctx) of a co-slice are separated by a dashed line. Some ‘immunonegative patches’ (stars) are visible (c, e). Two degenerating neurons are marked in a 50-week-old brain slice (f, arrows). Scale bar in f = 500 mm (a, c, e); 125 mm (b, d, f). nBM, basal nucleus of Meynert; NGF, nerve growth factor. Molecular Psychiatry


943

Figure 2 Single slices of the nBM (a–f), or cortex (g–l) or co-slices (m–r) were cultured for 12 weeks with (a–c, g–i, m-o) or without (d–f, j–l, p–r) 10 ng/ml NGF and then analyzed immunohistochemically for ChAT (all pictures) and then counterlabeled with the nuclear dye DAPI (b, e, h, k, n, q) and the marker neuN (c, f, i, l, o, r). Scale bar in a = 250 mm (a, b, d, e, g, h, j, k, m, n, p, q) and 60 mm (c, f, i, l, o, r). ChAT, choline acetyltransferase; DAPI, 4,6-diamidino-2-phenylindole; nBM, basal nucleus of Meynert; NGF, nerve growth factor; neuN, neuronal nuclei.

weeks: 14175, n = 12) or in nBM (12 weeks: 7879, n = 20; 24 weeks: 153715, n = 22) slices (Figures 3b and d). The APP(643–695)-immunolabeling appeared as granular and was located in the cytoplasm (Figures 3d). Colocalization studies in 12-week-old slices revealed that most of the APP(643–695)-positive cells colocalized with microtubule-associated protein2-positive neurons (Figures 3e–g). A comparatively smaller number of APP(643–695)-immunopositive cells also colocalized with astroglial GFAP (Figures 3h–j). In situ hybridization was used to detect ChAT mRNA (Figure 4a) as well as APP mRNA in 12-weekold slices (Figures 4b–d). High-power bright-field microscopy revealed a clustering of silver grains over cresyl violet-labeled cells (Figure 4d). The human and rat b-amyloid sequence differs in three amino acids at positions 5, 10 and 13 (Figure 5A). An antibody against amino acids 17–42 showed several intensely b-amyloid-immunopositive cells in 12-week-old brain slices (Figure 5Ba). These b-amyloid-containing cells colocalized with APPimmunopositive cells (Figures 5Bb, c). A similar labeling pattern was seen when slices were incubated with antibodies against rat-specific b-amyloid (3–16), the amino acids 17–24, or amino acids 37–42 (Figures 5Bd–f). Omitting any of the primary antibodies showed only background labeling (Figure 5Bg). Labeling with nuclear DAPI displayed a nuclear labeling (Figures 5Bh–j). Effects of acidosis on brain slices When 12-week-old brain slices were further incubated for 4 weeks with medium at pH 7.2 þ NGF, the distribution of b-amyloid(1–42)-positive cells in the slices did not change in shape and number

(Figures 6a–c). Co-labeling with nuclear DAPI showed a cytoplasmic labeling (Figure 6b). When 12-week-old brain slices were incubated with medium at pH 6.0, the number of ChAT-positive neurons decreased within 2–4 days (Table 1). The number of DAPI-positive cells gradually decreased within 4 weeks of acidosis treatment (Table 1). The number of neuN-positive neurons also markedly decreased after acidosis resulting in a few neuN-positive neurons after 4 weeks (Table 1). When 12-week-old slices were further incubated with medium at pH 6.0 þ NGF, no or very low granular b-amyloid-like immunoreactivity was seen (Figure 6d). The number of DAPI-positive nuclei was reduced (Figures 6e and f). However, when 12-week-old slices were incubated for further 4 weeks with medium at pH 6.0 þ NGF including ApoE4, the decrease in DAPI nuclei was significantly more pronounced (Figure 6h). Betaamyloid immunoreactivity was strongly granular (Figures 6g–i). Co-labeling with DAPI revealed that b-amyloid immunoreactivity was localized extracellularly (Figure 6i). Western blot of rat b-amyloid(1–42) and APP A4 Western Blot analysis for rat b-amyloid(17–24) and sbAPP is shown in Figure 7. Western blot analysis of 100 ng rat b-amyloid(1–42) showed a strong single band of 4 kDa and a weaker band at 12 kDa under reducing or non-reducing conditions. Western blot analysis of 100 ng sbAPP protein showed a strong band of approximately 100 kDa under non-reducing conditions with the anti-APP A4 antibody, whereas no sbAPP-protein (1000 ng) was detectable with the anti-b-amyloid(17–24) antibody. Brain slices incubated for 12 or 16 weeks at physiological pH Molecular Psychiatry


944

Figure 3 Co-slices of the nBM and cortex were cultured for 2 (a, c) or 12 (b, d–j) weeks with 10 ng/ml NGF and immunohistochemically labeled with antibodies against APP(643–695) (a–d, f, g, i, j). Note that APP(643–695)-like immunoreactivity is nearly absent in 2-week-old brain slices (a, c), whereas in 12-week-old mature brain slices, a strong intracellular granular immunoreactivity for APP(643–695) was found (b, d). Colocalization of APP(643–695) (Alexa-546, red, f, i) with neuronal MAP-2 (Alexa-488, green, e) or with GFAP (Alexa-488, green, h) is shown. Note that APP(643–695) colocalized with neuronal MAP-2 (arrow in g) and also with GFAP (arrow in j) as seen in the merged pictures. Scale bar in a = 75 mm (a, b); 15 mm (c, d), 80 mm (e–j). APP, amyloid-precursor protein; GFAP, glial fibrillary acidic protein; MAP-2, microtubule-associated protein-2; nBM, basal nucleus of Meynert; NGF, nerve growth factor.

displayed a positive band at approximately 25, 60 kDa and approximately 180 kDa. The 60 and 180 kDa bands disappeared in slices treated with pH 6.0. When ApoE4 was added to the slices at pH 6.0 the pattern did not change. Aggregation experiments showed a single 4 kDa band for b-amyloid under all Molecular Psychiatry

tested conditions (pH 7.2 or pH 6.0, 7ApoE4, 2 weeks 371C) when the mix was loaded directly onto the gel. When the samples were centrifuged and the pellet was loaded onto the gel, small oligomeric forms of bamyloid were seen after incubation at pH 7.2. This intensity of oligomeric forms was markedly increased


945

Figure 4 In situ hybridization shows detection of mRNAs for ChAT (a) or APP (b, d) in 12-week-old brain slices cultured with 10 ng/ml NGF. The dark field microscopic pictures show several ChAT mRNA (a) and APP mRNA (b, c) expressing cells (the box in b correlates to the higher magnificiation in c). Bright-field microscopy (d) shows a high power picture of cresyl violet counterlabeled cell. Scale bar in d = 400 mm (a–b); 50 mm (c), 25 mm (d). APP, amyloid-precursor protein; ChAT, choline acetyltransferase; NGF, nerve growth factor.

after incubation at pH 6.0. No obvious effect of ApoE4 was seen on the aggregation at pH 7.2 or pH 6.0. An additional weaker band was observed at approximately 50 kDa. ELISA for rat b-amyloid A rat b-amyloid-specific ELISA-detected rat b-amyloid in a dose-dependent manner, whereas human bamyloid was not detected (Figure 8a). Omission of the secondary antibody abolished the detection (Figure 8a). The ELISA showed a linear range between 0.5 and 25 ng/well (Figure 8a). Beta-amyloid was below the detection limit in brain slices (Figure 8b) and media (Figure 8c) when incubated for 2 or 12 weeks with NGF. However, when 12-week-old slices were incubated for further 4 weeks with NGF and ApoE4 at pH 7.2, there was a marked increase in b-amyloid tissue levels (Figure 8b) as well as secreted b-amyloid (Figure 8c). When 12-week-old brain slices were further incubated for 4 weeks with medium at pH 6.0 with NGF and with or without ApoE4, the bamyloid tissue levels (Figure 8b) and secretion (Figure 8c) was markedly decreased. Beta-amyloid labeling in a human Alzheimer brain Using the anti-b-amyloid(17–42) antibody, immunolabeling of a human Alzheimer brain revealed a

similar granular labeling as seen in brain slices after incubation at pH 6.0 with ApoE4 (Figures 6j–l). The labeling again appeared extracellular as seen by DAPI co-labeling (Figure 6l).

Discussion This study shows that organotypic brain slices cultured for 12 weeks express APP- and b-amyloidlike immunoreactivity localized in neurons and to a lesser extent in astrocytes. At this stage, b-amyloid is found intracellularly, and is released into the culture medium. When slices are treated with a culture medium at pH 6.0 (acidosis), the neurons degenerate, and b-amyloid aggregates extracellularly, which is triggered by ApoE4. Cholinergic neurons in the organotypic brain slice model This study extends our previous findings and shows for the first time that cholinergic neurons can survive in organotypic brain co-slices for at least 32 weeks when incubated continuously with NGF. The organotypic brain slice model has been introduced by Ga¨hwiler et al.,21 modified by Stoppini et al.,22 and meanwhile is well established in our research group since several years.15,16 We have shown16 that Molecular Psychiatry


946

Figure 5 (A) Comparison of the peptide sequence of human and rat b-amyloid(1–42); three amino acids differ within the region 5–13. (B) Immunohistochemistry shows b-amyloid-like immunoreactivity in rat organotypic brain slices. The antibody b-amyloid(17–42) labeled several cells in 12-week-old brain slices (a, h), which fully colocalized with the APP cells (b) as seen in the merged picture (c). A similar intracellular labeling was seen with a rat-specific b-amyloid(3–16) antibody (d) and two human antibodies against the amino-acid sequence (37–42) (e) or (17–24) (f). Omission of the specific antibodies resulted in background labeling (g). The b-amyloid labeling (h) was cytoplasmic and of neuronal phenotype, leaving the nuclei free, as seen by nuclear DAPI labeling (i–j). Scale bar in j = 60 mm (a–c); 90 mm (d–g); 30 mm (h–j). APP, amyloidprecursor protein.

cholinergic neurons survive in nBM brain slices when cultured with NGF for at least 2 weeks, but withdrawal of NGF from brain slice cultures results in massive cell death of cholinergic neurons.16 The enzyme ChAT labels cholinergic neurons and correlates well with the cell death of cholinergic neurons.15 The major advantage of organotypic brain slices is that two functionally related brain regions can be connected and in co-cultures of the nBM and cortex cholinergic neurons survive and send their nerve fibers into the adjacent target area, the cortex.15 Molecular Psychiatry

APP expression in aging brain slices Up to now, in vivo AD models only partly reflect the neuropathological features of AD. Especially, APP or tau overexpressing mice have been well characterized.23 Recently, it has been reported that brain slices prepared from mice overexpressing genes of relevance for AD may be a useful in vitro model.24 Our present study is now the first showing that 12- to 24-week-old rat organotypic brain slices express endogenous rat APP. The APP-like immunoreactivity was found to be expressed mainly in neurons and displayed mostly granular cytoplasmic staining. The APP- and


947

Figure 6 Beta-amyloid(1–42)-like immunreactivity in organotypic brain slices (a–i) and a postmortem Alzheimer brain (j–l). Twelve-week-old brain slices incubated with 10 ng/ml NGF were further incubated for 4 weeks with medium at pH 7.2 þ NGF (a–c), or medium at pH 6.0 þ NGF (d–f) or medium at pH 6.0 þ ApoE4 þ NGF (g–i). Several b-amyloid(1–42)-positive cells were visible as labeled with the antibody against b-amyloid(17–42) and Alexa-488 (a, c). Counterlabeling with DAPI showed several nuclei (b), which displayed intracellular labeling of b-amyloid(1–42) (c, merged, arrow). When slices were exposed for 4 weeks with acidic medium no or very diffuse granular b-amyloid immunoreactivity was observed (d, f) and the number of DAPI-positive nuclei was reduced (e). When slices were incubated with medium at pH 6.0 þ ApoE4 þ NGF a strong granular labeling was found with the anti-b-amyloid(17–42) antibody (g). Only a few DAPI-positive nuclei were seen after the acidosis treatment (h). The merged picture revealed that the b-amyloid(17–42) labeling was extracellular (i, arrow). Immunolabeling of a human Alzheimer postmortem brain showed strong granular labeling for b-amyloid(17–42) (j), strong DAPI-positive nuclei (k), which indicates extracellular depositions (l, arrow) similar as seen in the brain slices. Note that the pictures showing postmortem brains (j–l) are given in a lower magnification. Scale bar in l = 60 mm (a–i); 100 mm (j–l). ApoE4, apolipoprotein E4; DAPI, 4,6-diamidino-2-phenylindole; NGF, nerve growth factor.

b-amyloid-like immunoreactivity were co-expressed in the same cells. In situ hybridization with ratspecific oligonucleotides confirmed the expression of APP mRNA in rat brain slices. APP was also expressed in astroglial cells, but to a lower extent than in neurons. The expression and localization of different isoforms of APP has been studied in rat brain. There is good evidence that APP is found in neurons.25,26 In addition, it has been shown that also cultured rat astrocytes express APP, and in fimbriafornix lesioned rat, APP was mainly observed in GFAP-positive reactive hippocampal astrocytes,26

pointing to astrocytic APP induction after brain insults.27 In our brain slice model astrocytes also expressed APP, but to a lower extent, which may indicate that the organotypic brain slice model represents a stable non-activated in vitro system. The normal functions of APP are not fully understood, but increasing evidence suggests that APP has important roles in regulating neuronal survival, neurite outgrowth, synaptic plasticity and cell adhesion.28 The increased expression of APP overtime with a maximum in 12-week-old brain slices, may support its role in normal physiological neuronal functions. Molecular Psychiatry


948

Table 1 Days 0 1 2 4 7 14 28

Quantification of DAPI, neuN and ChAT after acidosis

DAPI þ

neuN þ

ChAT þ

4128972534 (5) 3063072443 (8)** 2294873417 (4)*** 124057736 (5)*** 1053972331 (5)*** 40657945 (5)*** 7987235 (5)***

1892972589 (5) 95787451 (3)*** 23057327 (4)*** 20497581 (3)*** 1424789 (3)*** 4357232 (5)**** 115721 (4)***

49710 (20) 49713 (11) NS 3578 (4) NS 372 (4)*** 572 (4)*** 0 (9)*** 0 (6)***

Abbreviations: ANOVA, analysis of variance; ChAT, choline acetyltransferase; DAPI, 4,6-diamidino-2-phenylindole; neuN, neuronal nuclei; PLSD, protected least significant difference. Twelve-week-old brain slices were incubated for 0–28 days with medium at pH 6.0, then fixed and immunohistochemically labeled for neuN þ or ChAT þ and counterlabeled with the nuclear dye DAPI. Quantification was performed as described in Materials and Methods. Values are expressed as mean number7s.e.m. of DAPI nuclei or neuN þ neurons per mm2 and as ChAT þ neurons per slice (values in parentheses give the number of analyzed samples). Statistical analysis was performed by one-way ANOVA with a subsequent Fisher PLSD post hoc test compared to time point 0 (NS, not significant; **P < 0.01; ***P < 0.001).

Figure 7 Western blot analysis for rat b-amyloid(17–24) (lanes 1–4 and 9–23) and sbAPP (APP A4) (lanes 5–8) under reducing (R) or non-reducing (nR) conditions. The following standards were loaded: 1000 ng APP (lanes 1 and 3), 1000 ng rat b-amyloid (lanes 6 and 8), 100 ng APP (lanes 5, 7 and 15) or 100 ng rat b-amyloid (lanes 2, 4 and 14). Slices were pooled, extracted and loaded onto the gels: 2 week slices þ NGF (lane 9); 12 week slices þ NGF (lane 10); 12-week slices þ NGF incubated for 4 weeks with medium pH 7.2 þ NGF (lane 11) or 4 weeks with medium pH 6.0 þ NGF (lane 12) or 4 weeks with medium pH 6.0 þ NGF þ ApoE4 (lane 13). Arrows point to changes between the different treatments. Rat b-amyloid (500 ng/ 100 ml) was incubated for 2 weeks at pH 7.2 without ApoE4 (lanes 16 and 20) or 10 ng/ml with ApoE4 (lanes 18 and 22) or at pH 6.0 without ApoE4 (lanes 17 and 21) or with ApoE4 (lanes 19 and 23). Samples were either directly loaded onto the gel (lanes 16–19) or were centrifuged for 15 min at 15 000 g and then the pellet was loaded onto the gel (lanes 20–23). Molecular weights are given at the left and right side in kDa. APP, amyloid-precursor protein; ApoE4, apolipoprotein E4; DAPI, 4,6-diamidino-2-phenylindole; NGF, nerve growth factor; sbAPP, secreted b-APP.

Beta-amyloid expression in aging brain slices Most of the published studies have been performed with human b-amyloid, because of the relevance in AD. Presynaptic terminals in AD are critically dependent on cortical b-amyloid levels, but not on b-amyloid plaque deposition.11,29 In fact, presynaptic terminals already significantly deplete in 2- to 4-month-old APP transgenic mice, at a time where their soluble b-amyloid levels increase but before b-amyloid deposition begins.11 The b-amyloid peptide shows 100% homology in amino acids 14–42 between human and rat. Thus, we have used four Molecular Psychiatry

different sequence-specific antibodies to detect rat b-amyloid, three human antibodies directed against amino acids 17–42 or 37–42 or 17–24 and one rat-specific antibody against amino acids 3–16. All antibodies reacted well for rat b-amyloid and show specific expression of rat b-amyloid(1–42) or -(1–40) in rat brain slices. The antibody AB5366 does not crossreact with secretory APP as stated by Chemicon (see datasheet). Our Western blot data show that the antibody MAB1561 did not recognize sAPP as well, pointing to a specific detection of b-amyloid peptide.


949

Figure 8 A representative ELISA standard curve (a) shows specific labeling for rat b-amyloid (filled circles), which was abolished when the secondary antibody was omitted (filled triangles). Human b-amyloid was not detectable (open circles). The assay was linear between 0.5 and 25 ng/well (small insert). b-Amyloid was measured in brain slices (b) or in medium (c) after different treatments: A, 2 weeks þ NGF; B, 12 weeks þ NGF; C, 12 weeks þ 4 weeks with medium pH 7.2 with NGF and ApoE4; D, 12 weeks þ 4 weeks with medium pH 6.0 with NGF; E, 12 weeks þ 4 weeks with medium pH 6.0 with NGF and with ApoE4. Brain slices (15 slices) were pooled, extracted, centrifuged and the supernatant analyzed. The medium was collected (after incubation for 1 week with 15 slices), pooled (3 1.2 ml), lyophilized, resuspended and analyzed. Values are given as mean7s.e.m. pg rat b-amyloid/mg protein (b) or as pg rat b-amyloid/ml collected in 1 week from 15 cortex slices (c). Values in parenthesis give the number of analyzed slices. The detection limit (DL) is shown by a line and some measurements were below the detection limit. n.d., not detectable; APP, amyloid-precursor protein; ApoE4, apolipoprotein E4; ELISA, enzyme-linked immunosorbent assay; NGF, nerve growth factor.

Beta-amyloid is produced from APP through sequential proteolytic cleavages by secretases.30 APP is axonally transported and b-amyloid accumulates at synapses in high amounts and is then secreted into the extracellular space. The secreted forms are cleared via the blood–brain barrier or intracellular b-amyloid is degraded by specific proteases. We cannot distinguish in our assay between b-amyloid(1–40) and -(1–42), but it is likely that both forms are expressed in the brain slices. Our data show preferentially a cytoplasmic granular labeling; however, we cannot exclude also nuclear staining with one of the antibodies. Although it was shown that b-amyloid(1–40) is produced by rat neuronal tissue,31 our data show for the first time that endogenous expression of APP and b-amyloid does not directly induce cell death of cholinergic neurons in vitro. In fact, it is well established that b-amyloid has neurotrophic and neuroprotective properties when present at low nanomolar physiological concentrations displaying antiapoptotic and antioxidant activity.7 However, only at high (mM) concentrations b-amyloid has toxic properties, which is linked to methionine-35 possibly mediating radical generation.7 Thus, our data indicate that aging brain slices physiologically express and secrete low amounts of b-amyloid(1–42) or (1–40), which does not exert any toxic effects on cholinergic neurons that are the most vulnerable neurons in AD. Even when brain slices are exposed for further 4 weeks with ApoE4 at neutral pH, the b-amyloid immunoreactivity remained intraneuronal and did not exhibit any toxicity. Beta-amyloid and acidosis in brain slices Acidosis occurs in the brain during infarction, such as stroke, ischemia, inflammation or ‘silent stroke’ and

plays a role in damaging neuronal environments, which is of importance in AD.32,33.It is now widely accepted that acidosis is an important component of the pathological event that leads to ischemic brain damage.34,35 We have recently shown, that acidosis (pH < 6.6) rapidly (within 4 days) killed cholinergic neurons.36 Our present data show that in fact, not only cholinergic neurons, but neurons in general degenerate, as seen by a marked loss of neuN-positive neurons in the brain slices after acidosis, accompanied by a gradual decrease in DAPI-positive cells. However, the incubation at low pH enhances at the same time the extracellular aggregation of b-amyloid. It is suggested, that low pH induces atrophy and bursting of neurons and subsequent fast cell death, which results in uncontrolled leakage and efflux of b-amyloid. There are no indications that b-amyloid undergoes a regulated release. It cannot be excluded that acidosis changes also the enzyme activity of secretases. An altered proteolytic processing of APP may lead to neurotoxic forms of b-amyloid, which then may play a pivotal role in triggering the death of neurons. Such a perturbed processing of APP due to ischemic events may result in increased production of b-amyloid at synapses and may be an early event in AD. Such changes in enzyme activity are indeed found in sporadic AD cases. It is well established that intraneuronal b-amyloid is a major risk factor and it serves as a source for extracellular amyloid aggregates and plaque deposition and may trigger the b-amyloid cascade.37,38 Extracellular vs intracellular b-amyloid This intracellular b-amyloid is secreted as a soluble peptide.37,38 Primary neurons of Tg2576 transgenic mice secrete approximately 500 pg/ml b-amyloid(1– Molecular Psychiatry


950

42) within 1 h, which is increased to up to 2600 pg/ml within 36 h.38 The secretion of b-amyloid(1–40) is much higher and up to 15 mg/ml of this peptide is secreted within 36 h.37,38 This ratio (20:1 for b-amyloid(1–40)/b-amyloid(1–42)) is of importance,37 because b-amyloid(1–40) is eliminated with a half-life of 1 h, whereas b-amyloid(1–42) is more resistant against degradation.37 The actual discussion has focused whether fibrillar or soluble oligomeric bamyloid is the active form of the peptide. The bamyloid cascade hypothesis14,37 favors the model that insoluble fibrillar b-amyloid triggers the neuronal degeneration. Evidence is now accumulating that soluble-activated monomers, soluble oligomers (dimer, trimer, tetramer) and protofibrils could be responsible for triggering the pathology in AD. The exact mechanism by which b-amyloid induces cell death is not known, but the ‘channel hypothesis’ suggests that certain fibrillar forms of the peptide cause neurodegeneration by forming ion channels that are generally large, voltage-independent and of relatively poor selectivity.29,39 It seems quite clear that neuronal axons contain a releasable pool of b-amyloid,40 but it is not clear if this synaptically released b-amyloid plays a role in the pathogenesis of AD. The data on release of b-amyloid in the rodent brain are very rare. Interestingly, C6 rat glioma cells produce and release a soluble 4 kDa peptide.41 Mouse microglia release soluble b-amyloid within a few hours, but release of fibrillar b-amyloid was less than half complete in 12 days.42 Our data suggest that under neutral pH b-amyloid(1–42) is physiologically secreted by controlled release into the medium of organotypic brain slices while under low pH an uncontrolled efflux of b-amyloid(1–42) from leaking neurons may occur, which can rapidly aggregate as extracellular depositions. Aggregation of rat b-amyloid Neuritic plaques and amyloid depositions have been detected in the humans, aged monkey, dog and polar bear and have rarely been found in aged rodents.43 It was suggested that the amino-acid sequence in humans, monkey, dog and polar bear are conserved and that the secondary structure of mouse and rat b-amyloid lacks a property to form a b-sheeted structure.43 b-amyloid can form insoluble assemblies of fibrills, which depends on the concentration, time, temperature, ionic strength and pH.7,44–46 It has been shown that low pH provides an ideal environment for b-amyloid aggregation.7 These conformational changes of b-amyloid are dependent on pH and absence of metal ions and soluble b-amyloid occurs when it is in an a-helical monomeric form between pH 1–4 and > 7. However, a b-sheet conformation occurs between pH 4–7 resulting in aggregation of bamyloid.7,47 This reduction in pH often mobilizes metal ions (such as Zn2 þ , Cu2 þ ), which are both found at high concentrations in the cortex and contribute to the promotion of fibrills and the aggregation of b-amyloid.7,29,45 It was suggested that

Molecular Psychiatry

histidine residues are essential for the metalmediated b-amyloid assembly—this may explain why b-amyloid deposition is not a feature of aged rats.43 A marked Cu2 þ -induced aggregation of bamyloid emerges when the pH is lowered to 6.8, indicating that H þ -induced conformational changes unmask a metal-binding site on b-amyloid that mediates reversible assembly of the peptide.45 Betaamyloid(1–42) has been shown to be more prone to amyloid fibril formation and appears to be the more toxic form of the peptide.29 The amino-acid sequence 15–22 of b-amyloid may control both aggregation at acidic pH but also its proteolytic activity at neutral pH.46 This direct interaction between low pH and bamyloid aggregation forced us to explore the role of acidosis in the brain slices. In fact, our aggregation experiments with Western blot show that rat bamyloid can undergo formation into small oligomeric soluble aggregates, which is markedly increased at pH 6.0. Furthermore, the immunohistochemical figures point to an aggregated form of extracellular bamyloid, a labeling pattern which was similar as seen in an AD postmortem human brain. Beta-amyloid and apolipoprotein E4 ApoE is an important transport molecule for lipids in the brain and triggers fast transport of b-amyloid within the brain and to the blood–brain barrier.29,47 Beta-amyloid binds to ApoE and such complexes are important for the clearance of b-amyloid in the brain. Dysfunction of ApoE may be of importance in AD, especially the ApoE4 allele is a major risk factor triggering late onset AD by increasing b-amyloid deposition.29,48–50 It has been found that the bamyloid–ApoE4 complex is not efficiently cleared from the brain, but more susceptible to amyloid aggregations and depositions. Indeed, several data show that ApoE4 markedly affects the deposition of bamyloid in a mouse model of AD.51 These potent effects of ApoE4 forced us to explore further the effects together with acidosis. Our data with Western blot now clearly show that ApoE4 plays an important role in the aggregation process of rat b-amyloid. It seems likely that the extracellular b-amyloid undergoes aggregation in the presence of ApoE4, whereas under neutral pH no such extracellular forms are seen when slices are incubated with ApoE4 at neutral pH. Taken together, this in vitro brain slice model provides a model for studying the regulation of physiological and pathological b-amyloid expression and secretion. This model has the advantage that cells express and regulate endogenous APP and b-amyloid, whereas in transgenic mice an exogenous recombinant protein is introduced. However, the brain slice model does not reflect the long-lasting chronic development of AD, because in our in vitro model a severe acute acidosis results in significant cell death of neurons accompanied by b-amyloid aggregation. Further experiments need to be done to balance the pH range and to determine how the slices react to very moderate pH changes (for example, 6.8–6.6–6.4–6.2)


over longer time periods. Perhaps it may be possible to tune the system to observe b-amyloid aggregation without such a significant neuronal cell death, which may allow us to explore more detailed mechanisms of cholinergic cell death and b-amyloid aggregation within a defined time frame. In conclusion, our findings show for the first time that rat APP/bamyloid is expressed in aging rat organotypic brain slices. We provide evidence that the survival of cholinergic neurons does not depend on the overexpression of APP and b-amyloid(1–42). Acidosis (pH 6.0) causes a degeneration of neurons, and leads to a marked loss of neurons. Under these conditions, bamyloid aggregates extracellularly, which is triggered by ApoE4. The aging organotypic brain slice model provides a potent tool to investigate the secretion and processing of APP/b-amyloid in a complex threedimensional physiological brain model. It will allow us to study putative therapeutics that may selectively suppress APP overexpression and b-amyloid accumulation and plaque formation.

Acknowledgments This study was supported by the Austrian Science Fund (P16130-B08 and P19122-B05). We thank Iris Berger, Michael Pirchl and Ursula KirzenbergerWinkler for excellent technical help. We also thank Dr Georg Wietzorrek for help with the confocal microscopy. Competing interests statement The authors declare that they have no competing financial interests.

References 1 Selkoe DJ. The cell biology of beta-amyloid precursor protein and presenilin in Alzheimer’s disease. Trends Cell Biol 1998; 8: 447–453. 2 Mattson MP, Barger SW, Cheng B, Lieberburg I, Smith-Swintosky VL, Rydel RE. Beta-amyloid precursor protein metabolites and loss of neuronal Ca2 þ homeostasis in Alzheimer’s disease. Trends Neurosci 1993; 16: 409–414. 3 Checler F. Processing of the beta-amyloid precursor protein and its regulation in Alzheimer’s disease. J Neurochem 1995; 65: 1431– 1444. 4 Masliah E. Role of amyloid precursor protein in the mechanisms of neurodegeneration in Alzheimer’s disease. Lab Invest 1997; 77: 197–209. 5 Gandy S, Greengard P. Processing of Alzheimer A-beta-amyloid precursor protein: cell biology, regulation, and role in Alzheimer disease. Int Rev Neurobiol 1994; 36: 29–50. 6 Yankner BA. Mechanisms of neuronal degeneration in Alzheimer’s disease. Neuron 1996; 16: 921–932. 7 Atwood CS, Obrenovich ME, Liu T, Chan H, Perry G, Smith MA et al. Amyloid-beta: a chameleon walking in two worlds: a review of the trophic and toxic properties of amyloid-beta. Brain Res Rev 2003; 43: 1–16. 8 Davies P, Maloney AJF. Selective loss of central cholinergic neurons in Alzheimer’s disease. Lancet 1976; 2: 1403. 9 Whitehouse PJ, Price DL, Struble RG, Clark AW, Coyle JT, Delon MR. Alzheimer’s disease and senile dementia: loss of neurons in basal forebrain. Science 1982; 215: 1237–1239. 10 Winkler J, Thal LJ, Gage FH, Fisher LJ. Cholinergic strategies for Alzheimer’s disease. J Mol Med 1998; 76: 555–567.

11 Selkoe DJ. Alzheimer’s disease is a synaptic failure. Science 2002; 298: 789–791. 12 Terry AV, Buccafusco JJ. The cholinergic hypothesis of age and Alzheimer’s disease-related cognitive deficits: recent challenges and their implications for novel drug development. J Pharmacol Exp Ther 2002; 306: 821–827. 13 Masliah E. Mechanisms of synaptic pathology in Alzheimer’s disease. J Neural Transm 1998; 53: 147–158. 14 Hardy J, Selkoe D. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 2002; 298: 353–356. 15 Humpel C, Weis C. Nerve growth factor and cholinergic CNS neurons studied in organotypic brain slices: Implications in Alzheimer’s disease. J Neural Transm 2002; 62: 253–263. 16 Weis C, Marksteiner J, Humpel C. Nerve growth factor and glial cell line-derived neurotrophic factor restore the cholinergic phenotype in organotypic brain slices of the basal nucleus of Meynert. Neuroscience 2001; 102: 129–138. 17 Lechner T, Adlassnig C, Humpel C, Kaufmann WA, Maier H, Reinstadler-Kramer K et al. Chromogranin peptides in Alzheimer’s disease. Exp Gerontol 2004; 39: 101–113. 18 Humpel C, Wetmore C, Olson L. Regulation of brain-derived neurotrophic factor messenger RNA and protein at the cellular level in pentylenetetrazol-induced epileptic seizures. Neuroscience 1993; 53: 909–918. 19 Brice A, Berrard S, Raynaud B, Ansieau S, Coppola T, Weber MJ et al. Complete sequence of a cDNA encoding an active rat choline acetyltransferase: a tool to investigate the plasticity of cholinergic phenotype expression. J Neurosci Res 1989; 23: 266–273. 20 Forloni G, Del Bo R, Angeretti N, Smiroldo S, Gabellini N, Vantini G. Nerve growth factor does not influence the expression of betaamyloid precursor protein mRNA in rat brain: in vivo and in vitro studies. Brain Res 1993; 620: 292–296. 21 Ga¨hwiler BH, Capogna M, Debanne D, McKinney RA, Thompson SM. Organotypic slice cultures: a technique has come of age. Trend Neurosci 1997; 20: 471–477. 22 Stoppini L, Buch PA, Muller D. A simple method for organotypic cultures of nervous tissue. J Neurosci Meth 1991; 37: 173–182. 23 Masliah E, Rockenstein E. Genetically altered transgenic models of Alzheimer’s disease. J Neural Transm 2000; 59: 175–183. 24 Duff K, Noble W, Gaynor K, Matsuoka Y. Organotypic slice cultures from transgenic mice as disease model systems. J Mol Neurosci 2002; 19: 317–320. 25 Kang J, Muller-Hill B. Differential splicing of Alzheimer’s disease amyloid A4 precursor RNA in rat tissues: PreA4(695) mRNA is predominantly produced in rat and human brain. Biochem Biophys Res Comm 1990; 166: 1192–1200. 26 Rohan de Silva HA, Jen A, Wickenden C, Jen LS, Wilkinson SL, Patel AJ. Cell specific expression of beta-amyloid precursor protein isoform mRNAs and proteins in neurons and astrocytes. Mol Brain Res 1997; 47: 147–156. 27 Moreno-Flores MT, Salinero O, Wandosell F. Beta-amyloid(25–35)induced APP expression in cultured astrocytes. J Neurosci Res 1998; 52: 661–671. 28 Mattson MP. Cellular actions of beta-amyloid precursor protein and its soluble and fibrillogenic derivatives. Physiol Rev 1997; 77: 1081–1132. 29 Tanzi RE, Bertram L. Twenty years of the Alzheimer’s disease amyloid hypothesis: a genetic perspective. Cell 2005; 120: 545–555. 30 Golde TE. The Abeta hypothesis: leading us to rationally-designed therapeutic strategies for the treatment or prevention of Alzheimer disease. Brain Pathol 2005; 15: 84–87. 31 Busciglio J, Gabuzda DH, Matsudaira P, Yankner BA. Generation of beta-amyloid in the secretory pathway in neuronal and nonneuronal cells. Proc Natl Acad Sci USA 1993; 90: 2092–2096. 32 Iadecola C. Neurovascular regulation in the normal brain and in Alzheimer’s disease. Nat Rev Neurosci 2004; 5: 347–360. 33 Humpel C, Marksteiner J. Cerebrovascular damage as a cause for Alzheimer’s disease? Curr Neurovasc Res 2005; 2: 341–347. 34 Siesjo BK. Acidosis and ischemic brain damage. Neurochem Pathol 1988; 9: 31–88. 35 Siesjo BK. Pathophysiology and treatment of focal cerebral ischemia. Part I: Pathophysiology. J Neurosurg 1992; 77: 169–184.

951



952

36 Pirchl M, Marksteiner J, Humpel C. Effects of acidosis on brain capillary endothelial cells and cholinergic neurons: relevance for vascular dementia and Alzheimer’s disease. NeurolRes 2006; 28: 657–664. 37 Wirths O, Multhaup G, Bayer TA. A modified beta-amyloid hypothesis: intrneuronal accumulation of beta-amyloid peptide—the first step of a fatal cascade. J Neurochem 2004; 91: 513–520. 38 Oddo S, Caccamo A, Smith IF, Green KN, LaFerla FM. A dynamic relationship between intracellular and extracellular pools of betaamyloid. Am J Pathol 2006; 168: 184–194. 39 Marchesi VT. An alternative interpretation of the amyloid Abeta hypothesis with regard to the pathogenesis of Alzheimer’s disease. Proc Natl Acad Sci USA 2005; 102: 9093–9098. 40 Schroeder BE, Koo EH. To think or not to think: synaptic activity and Abeta release. Neuron 2005; 48: 873–879. 41 Morato E, Mayor FJ. Production of the Alzheimer’s beta-amyloid peptide by C6 glioma cells. FEBS 1993; 336: 275–278. 42 Chung H, Brazil MI, Soe TT, Maxfield FR. Uptake, degradation, and release of fibrillar and solube forms of Alzheimer’s amyloid beta-peptide by microglial cells. J Biol Chem 1999; 274: 32301– 32308. 43 Johnstone EM, Chaney MO, Norris FH, Pascual R, Little SP. Conservation of the sequence of the Alzheimer’s disease amyloid peptide in dog, polar bear and five other mammals by crossspecies polymerase chain reaction analysis. Mol Brain Res 1991; 10: 299–305.


44 Carrotta R, Manno M, Bulone D, Martorana V, San Biagio PL. Protofibril formation of amyloid beta-protein at low pH via a noncooperative elongation mechanism. J Biol Chem 2005; 280: 30001– 30008. 45 Atwood CS, Moir RD, Huang X, Scarpa RC, Bacarra ME, Romano DM et al. Dramatic aggregation of Alzheimer Abeta by Cu(II) is induced by conditions representing physiological acidosis. J Biol Chem 1998; 273: 12817–12826. 46 Matsunaga Y, Fujii A, Awasthi A, Yokotani J, Takakura T, Yamada T. Eight residue amyloid beta peptides inhibit the aggregation and enzymatic activity of amyloid-beta42. Regul Peptides 1994; 120: 227–236. 47 Yip AG, McKee AC, Green RC, Wells J, Young H, Cupples LA et al. APOE, vascular pathology, and the AD brain. Neurology 2005; 65: 259–265. 48 Holtzman DM, Fagan AM, Mackey B, Tenkova T, Sartorius L, Paul SM et al. Hyman BT. Apolipoprotein E facilitates neuritic and cerebrovascular plaque formation in an Alzheimer’s disease model. Ann Neurol 2000; 47: 739–747. 49 Katzman R. Apolipoprotein E and Alzheimer’s disease. Curr Opin Neurobiol 1994; 4: 703–707. 50 Strittmatter WJ, Roses AD. Apolipoprotein E and Alzheimer disease. Proc Natl Acad Sci USA 1995; 92: 4725–4727. 51 Bales KR, Verina T, Cummins DJ, Du Y, Dodel RC, Saura J et al. Apolipoprotein E is essential for amyloid deposition in the APP(V717F) transgenic mouse model of Alzheimer’s disease. Proc Natl Acad Sci USA 1999; 96: 15233–15238.