Background
Alzheimer's disease (AD) represents the so-called "storage disorder" of amyloid β (Aβ). The AD brain contains soluble and insoluble Aβ, both of which have been hypothesized to underlie the development of cognitive deficits or dementia [
1‐
3]. The steady-state level of Aβ is controlled by the generation of Aβ from its precursor, the degradation of Aβ within the brain, and transport of Aβ out of the brain. The imbalance among three metabolic pathways results in excessive accumulation and deposition of Aβ in the brain, which may trigger a complex downstream cascade (e.g., primary amyloid plaque formation or secondary tauopathy and neurodegeneration) leading to memory loss or dementia in AD. Accumulated lines of evidence indicate that such a memory loss represents a synaptic failure caused directly by soluble Aβ oligomers (AβOs) [
4‐
6], whereas amyloid fibrils may cause neuronal injury indirectly via microglial activation [
7]. Thus, the classical amyloid cascade hypothesis [
8] underwent a modification in which the emphasis is switched to the intermediate form of Aβ such as AβOs [
9‐
12], rather than fibrillar Aβ [
7]. If this were the case, therapeutic intervention targeting AβOs may be effective in blocking this pathogenic cascade. The outcome of a recent human AN-1791 trial confirmed that plaque removal did not prevent the progression of neuronal degeneration [
13], supporting this hypothesis.
However, the distinct assembly states of AβOs remain to be elucidated. Several forms of AβOs have been found to be neurotoxic, from LMW-oligomers (dimers, trimers, and tetramers) disrupting memory function [
14,
15], synaptic function [
15,
16] and long-term potentiation (LTP) [
14,
17], to dodecamers affecting memory [
18]. In addition, Aβ-derived diffusible ligands (or ADDLs) [
9,
19], globulomers [
11], fibrillar Aβ oligomers [
20,
21], and toxic soluble Aβ assembly (TAβ) [
22] have been shown to be highly synaptotoxic or neurotoxic. Recently, a particular form of AβO, named the native amylospheroids [
23], has been isolated from AD brains and found to induce neuronal loss through its binding to synaptic targets [
24].
In this study, we chose a prophylactic passive immunization as a tool to define not only the pathological relevance of AβOs as the trigger of synaptic or neuronal degeneration, but also the possible mechanism underlying the neurotoxic action of endogenous AβOs. To address this issue, we successfully generated monoclonal 1A9 and 2C3 antibodies using a novel design method. When extracellular high-molecular-weight (HMW)-AβOs were controlled by 1A9 or 2C3 in Swedish-type amyloid precursor protein (APP) transgenic mice (Tg2576), we demonstrated that synaptic/neuronal degeneration or accumulation of intraneuronal AβOs was effectively prevented. These results argue for a role of both extracellular and intracellular HMW-AβOs in the induction and progression of synaptic or neuronal degeneration and provide a potential explanation for the extracellular one as the primary molecular basis for a toxic process.
Discussion
The development of AβOs-selective antibodies has greatly facilitated the understanding of
in vivo relevance of endogenous AβOs-mediated synaptic failure or neuronal degeneration. To prove this issue, a prophylactic study to control endogenous AβOs using AβOs-selective antibodies is required. Several conformation-dependent antibodies such as oligomer- or fibril-specific antibodies were reported previously [
12,
20,
21,
27,
30,
31], but none of them was examined for this purpose. In the current study, we successfully generated monoclonal oligomer-specific 1A9 and 2C3 using a novel design concept. Monoclonal 1A9 recognizes HMW-oligomers (100~230-mers), whereas 2C3 recognizes LMW- and HMW-oligomers ranging in sizes larger than pentamers (5~230-mers). In support of a previous report [
12], prefibrillar oligomer-selective A11 appears to be specific to HMW-oligomers. Under conditions of SDS-PAGE, 1A9-, 2C3-, or A11-oligomeric conformers were consistently detected at 70-180 kDa corresponding to 15~40-mers. Note that neither 1A9 nor 2C3 reacted with monomers and fibrils. In spite of heterogeneity in size, AFM clearly demonstrated that the toxic 1A9-, 2C3-, or A11-oligomeric conformers display relatively compact spherical particles, not fibrilar structure.
Using these oligomer-selective 1A9 and 2C3, we found that the majority of AβOs exclusively accumulated in neurons, whereas the degree of staining of diffuse plaques varied among antibodies tested (1A9>2C3 = A11). Under our conditions tested, A11-immunoreactivity is occasionally found in small diffuse plaques, in contrast with the finding of a previous report [
12]. These findings are direct evidence that heterogeneous oligomeric conformers exist as a distinct entity in both extracellular and intraneuronal deposits in human brains. Furthermore, the change of nuclear appearance in AβOs-burned neurons is highly indicative of impending neuronal degeneration: 1A9-positive AβOs may be associated with the most severe neuronal degeneration among the three anti-oligomer-specific antibodies tested. FJB binding assay, which specifically detects degernerating neurons [
29], also confirmed that 1A9- or 2C3-burned neurons in AD brains were unequivocally FJB-positive, indicating that intraneuronal accumulation of AβOs is closely associated with neuronal degeneration.
Using above-mentioned 1A9 and 2C3, we evaluated whether a specific control of endogenous, extracellular AβOs
in vivo would be sufficient to prevent the disruption of neuronal function leading to memory loss. Our
in vivo investigation demonstrated that immunized subjects have less intraneuronal accumulation of AβOs and fewer degenerating neurons than untreated controls. In addition, 2C3 and 1A9 protected Tg2576 mice from postsynaptic degeneration and from the degeneration of dendritic spines, respectively. These results place both extracellular [
22,
32] and intraneuronal AβOs [
33,
34] centrally within the mechanisms mediating AβO-induced neuronal dysfunction leading to memory loss [
26,
27]. In support of these data, our in vivo investigations clearly demonstrated that 1A9- and 2C3-treated Tg2576 mice aged 13 months showed cognitive performance superior to that of untreated Tg2576 mice, and, ultimately, performed better than and/or as well as untreated wild-type mice. It is unlikely that the impaired performance of Tg2576 mice in learning and memory tests is due to changes in motivation or sensory motor function, because the purpose of each behavioral test is different, and different skills are required for a good performance in each test. There were no differences in locomotor activity and the total time spent exploring objects in the novel object test between the wild-type and Tg2576 mice. Thus, it is likely that endogenous oligomeric 1A9- or 2C3-conformer is not only generated in Tg2576 mice, but is also actually a bioactive molecule
in vivo, and its selective immunoneutralization by systemic administration of 1A9 or 2C3 is sufficient to prevent either short-term or long-term memory loss. This
in vivo neuroprotective activity of 1A9 appeared to be superior to that of 2C3, supporting our current finding that neuronal degeneration in AD brains is more severe in 1A9-burned neurons than in 2C3-burned neurons. Recently, the generation of a new mouse model expressing only AβOs in neurons has demonstrated that endogenous AβOs are neurotoxic
in vivo inducing synaptic alteration, abnormal tau phosphorylation, glial activation, and neuronal loss [
35], supporting the pathological relevance of AβOs as shown herein. Taken together, the results from this study indicate that both extracellular and intraneuronal AβOs represent a molecular basis of memory loss
in vivo. Additional studies that attempt to identify the cellular and molecular targets on the cellular surface with which 1A9 or 2C3 interacts may yield insights into the mechanisms underlying the synaptotoxic or neurotoxic effects of AβOs or synaptoprotective or neuroprotective effect of 1A9 or 2C3.
Methods
Generation of monoclonal 1A9 and 2C3
Synthetic Aβ1-42 (r-peptide, Osaka, Japan) was dissolved in distilled, deionized H
2O, or 10 mM phosphate buffer at 250 μM, and allowed to incubate at 37°C for 18 h. Twenty microgram of preincubated Aβ1-42 was separated by NuPAGE 4-12% Bis-tris glycine gels, followed by CBB staining. Balb-c mice were immunized by injection with 2.5 μg of Aβ1-42 tetramer alone, which was excised from the gel and emulsified with complete Freund's adjuvant, into foot pads, followed by six additional injections. Inguinal lymphonode was used to generate hybridomas by fusion with Sp2/O-Ag14 myeloma cells with polyethylene glycol 1500. Initial screening was performed by dot blot analysis, applying 2.5 μl of seed-free fresh or 18-h preincubated Aβ1-42 (2.5 ng/dot) to a nitrocellulose membrane [
22]. The blots were then allowed to dry and blocked with 5% low-fat milk and 1% BSA in PBS containing 0.05% Tween-20 (PBST) and incubated with culture medium supernatant, followed by horseradish peroxidase (HRP)-labeled goat anti-mouse or anti-rabbit F(ab')
2 antibody (1:3000; Amersham). Dot immunoblots were visualized with an ECL kit using LAS3000 mini (Fujitsu, Tokyo, Japan).
Preparation of Aβ1-42 peptide
Synthetic Aβ1-42 was dissolved in 0.02% ammonia solution at 250 μM. To obtain seed-free Aβ42 solutions (540,000 g sup), the prepared solutions were centrifuged at 540,000
g for 3 h using an Optima TL ultracentrifuge (Beckman, USA) to remove undissolved peptides, which can act as preexisting seeds. The supernatant was collected and stored in aliquots at -80°C until use. Immediately before use, the aliquots were thawed and diluted with TBS (150 mM NaCl and 10 mM Tris-HCl, pH 7.4). For time- course experiments, 540,000 g sup was incubated for an indicated time (0-72 h), and soluble AβOs were retained after 100,000 g ultracentrifugation for 1 h, followed by dot immunoblot analysis (2.5 ng/dot) [
22]. Because the 4-h-incubated mixture is suitable for the characterization of soluble AβO standards, we used it in further experiment instead of the 18-h-incubated mixture used in the first screening. To further assess whether monoclonal 1A9 or 2C3 recognizes Aβ fibrils, seed-free Aβ42 solutions (25 μM) were incubated for 5 days at room temperature. Electron-microscopy-confirmed Aβ fibrils were subjected to dot immunoblot analysis (2.5 ng/dot) [
22].
Electron microscopy (EM)
For electron microscopy, samples were diluted with distilled water and spread on carbon-coated grids. The grids were negatively stained with 1% phosphotungstic acid and examined under a Hitachi H-7000 electron microscope (Tokyo, Japan) with an acceleration voltage of 77 kV.
BN-PAGE and two-dimensional Natie/SDS-PAGE
BN-PAGE analysis was performed following the manufacture's instruction (Invitrogen, Carlsbad, CA): 4%, 6%, 8%, 10%, 12%, 14%, 16%, 18%, 20%, and 4-16% Novex
® Bis-Tris gel was used. The apparent molecular masses of LMW-oligomers were calculated from the Ferguson plots with known molecular mass standards (α-lactalbumin, 14.2 kDa; carbonic anhydrase, 29 kDa; chickin egg albumin, 45 kDa; bovine serum albumin, 66 kDa monomer, and 132 kDa dimer; urease, 272 kDa monomer and 545 kDa dimer) (Sigma). For two-dimensional native/SDS-PAGE, one lane was excised from the gel and applied directly for SDS-PAGE. Immunoblot analysis was performed as described previously [
36].
Aβ-induced toxicity assay
We conducted the Aβ-induced toxicity assay according to previously published methods [
22]. Briefly, rat pheochromocytoma 12 (PC12) cells were cultured in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% heat-inactivated horse serum (Invitrogen) and 5% FBS (Invitrogen). For their differentiation, PC12 cells were plated on 10-cm
2 poly-L-lysine-coated (10 mg/ml) dishes at a density of 20,000 cells/cm
2 and cultured for 6 days in DMEM supplemented with 100 ng/ml nerve growth factor (NGF; Alomone Labs, Jerusalem, Israel) (PC12N). Basically, toxicity was assessed using different Aβ conformers: seed-free Aβ1-42 vs. 2-h- or 4-h-preincubated Aβ1-42 with or without 540,000 g ultracentrifugation for 1 h. PC12N was exposed to seed-free or preincubated Aβ1-42 at 25 μM for 48 h at 4°C. Toxicity was assessed by Live/Dead two-color fluorescence assay (Molecular Probes, Eugene, OR) or CytoTox 96 Non-Radioactive Cytotoxicity Assay Kit (Promega, Madison, DI) in accordance with the manufacturer's instructions as described previously [
22].
Ultrafiltration and molecular sieving (UF/MS)
To characterize the size of de novo formed toxic AβOs at 25 μM, serial ultrafiltration using Microcon® 3-, 10-, 30-, and 10-kDa cut-off membranes (Millipore Corp. Billerica, MA) was performed to prepare four filtrates (12,000 g-centrifuge for 90 min, Fr. 1, >3 kDa; 12,000 g-centrifuge for 60 min, Fr. 2, 3-10 kDa; 12,000 g-centrifuge for 30 min, Fr. 3, 10-30 kDa; 12,000 g-centrifuge for 15 min, Fr. 4, 30-100 kDa) and one retentate (Fr. 5, >100 kDa). Each fraction at 25 μM was subjected to Aβ-induced toxicity assay as described above. PC12N cells were exposed to each fraction and the toxic fraction was assessed as described above. The distribution of the A11, 1A9, 2C3, or 4G8 conformer was characterized by dot blot analysis as described above. To determine the morphology of toxic oligomers, each fraction was also subjected to atomic force microscopy (AFM).
Atomic force microscopy (AFM)
AFM assessment was performed as reported previously [
25]. Briefly, samples were dropped onto freshly cleaved mica. After allowing them to stand for 1 h, following by washing with water, the samples were assessed in a solution using a Nanoscope IIIa (Digital Instruments, Santa Barbara, CA, USA) set at tapping mode. OMCL-TR400PSA (Olympus, Japan) was used as a cantilever. Consecutive scans were monitored until distortion due to creep or shifts in the slow scan direction were negligible before collecting scans at sizes of 2 μm with the maximum 512 × 512 pixel resolution.
Immunohistochemistry
The left hemispheres of the brains of Tg2576 mice were sagitally cut into 30-μm-thick sections using a freezing microtome (RM 2145; Leika, Wetzlar, Germany). The tissue blocks from human subjects (4 AD subjects and 3 normal controls) or mice were fixed in 4% paraformaldehyde with 0.1 M phosphate-buffered saline (pH 7.6) and embedded in paraffin wax. After deparaffinization, heat-induced antigen retrieval was achieved by boiling sections for 10 min in a microwave oven in 0.01 M citrate buffer pH6.0, followed by 3 min incubation in 99% formic acid and then blocking of endogenous peroxidase. Then sections were subsequently incubated for 1 hour with primary antibody diluted in blocking buffer with normal goat or horse serum (2%), and after washing for 1 hour, with a secondary antibody in the same buffer. All incubations were done in parallel and photograph exposures were equal for sections in human and mice.
The following primary antibodies were used: monoclonal antibodies 6E10 and 4G8 against human Aβ sequence corresponding to residues 1-16 and 17-24, respectively, (Covance Immuno-Technologies, Dedham, MA), Polyclonal A11 specific to AβOs (Biosource, Camarillo, CA), anti-SYP (D4) antibody, monoclonal antibody against β-actin (C4) (Santa Cruz, Santa Cruz, CA), monoclonal anti-drebrin antibody (MBL, Nagoya, Japan), polyclonal anti-PSD-95 (CT) antibody (Invitrogen, Camarillo, CA), and our monoclonal 1A9 and 2C3 antibodies specific to AβOs.
The following secondary antibodies were used (1:1000): Goat anti-rabbit or anti-mouse IgG conjugated with horse-radish peroxidase (HRP) (Invitorogen, Carlsbad, CA), and Goat anti-mouse IgG conjugated with Alex Fluor (AF) 488 or 594 and goat anti-rat IgG conjugated with AF 488 (Molecular Probes, Eugene, OR). Immunopositive signals were visualized using an observation system (Compix Imaging System, Lake Oswego, OR) linked to an Olympus microscope BX50 through a highly sensitive CCD camera or a confocal laser scanning microscope (Carl Zeiss LSM510).
Fluoro-Jade B was purchased from Chemicon (now part of Millipore, Schwalbach, Germany). 7-AAD was purchased from Invitrogen (Carlsbad, CA).
Histochemistry
For Fluoro-Jade B histochemistry (1:50000), 5-μm-thick paraffin-embedded sections were deparafinized and stained following the manufacturer's instruction (Chemicon, now part of Millipore, Schwalbach, Germany). The Fluoro-Jade B-stained product fluoresces when exited at 488 nm and staining was imaged using a confocal laser scanning microscope (Carl Zeiss LSM510). The same procedure was applied for 30-μm-thick cryostat sections.
Cell culture and cellular uptake
Human neuroblastoma SH-SY5Y (SY5Y) cells were cultured in Dulbecco's modified Eagle's medium/Ham's F-12 medium supplemented with 10% fetal bovine serum. To investigate the fate of extracellular Aβ, SY5Y cells were exposed to HiLyte Fluor™488-labeled AβMs (10 kDa-filtrate), HiLyte Fluor™488-labeled AβOs (30 kDa-retentate) at 5 μM (AnaSpec, San Jose, CA), or Fluor™488 alone for 10 min, 30 min, and 180 min. In a separate set of experiment, cultures were treated at 37°C for 0, 3, 6, and 24 h with 5 μM Fluor™488-labeled AβOs, and for 24 h with 5 μM Fluor™488-labeled AβMs and synthetic Aβ42-1 (AnaSpec, San Jose, CA). Toxicity was assessed by CytoTox 96 Non-Radioactive Cytotoxicity Assay Kit in accordance with the manufacturer's instructions (Promega, Madison, DI) as described previously [
22].
Protein extraction and immunoblotting
Saline-soluble, saline-insoluble, SDS-soluble fractions, or SDS-insoluble, formic acid (FA)-extractable fractions were prepared from the Tg2576 mouse brains as described previously [
32]. Briefly, frozen brain samples were homogenized with a motor-driven Teflon/glass homogenizer (20 strokes) in TBS containing a cocktail of protease inhibitors (150 mg/ml), followed by centrifugation at 100,000
g for 1 h. The resultant supernatant (soluble fraction) was subjected to dot blot immunoanalysis or western blotting employing the same antibodies as those used for immunohistochemical staining. The pellet was further extracted with 2% sodium dodecylsulphate (SDS), followed by 70% FA, and the homogenate was ultracentrifuged as described above. The resultant supernatant (insoluble fraction) was also subjected to dot blot immunoanalysis and western blot analysis. For western blotting, aliquots of isolated fractions were separated using NuPAGE 4-12% bis-tris-glycine gels and transblotted to nitrocellulose membrane or Immobilon P (Millipore) for 1 h at 400 mA using 10 mM 3-cyclohexylamino-1-propancsulphonic acid (pH 11) containing 10% methanol. Membranes were blocked for 3 h at room temperature with 5% low-fat milk and 1% BSA in PBST and incubated with either the polyclonal anti-A11 (1:1000) or anti-PSD95 antibody (1:250), and monoclonal anti-drebrin (1:100), anti-SYP (1:2000), and anti-β actin antibodies (1:1000), followed by HRP-labeled goat anti-rabbit or anti-mouse F(ab')
2 antibody (1:3000; Amersham). Immunoblots were visualized with an ECL kit using LAS3000 mini (Fujitsu, Tokyo, Japan). Densitometric analysis of immunoblot was performed using Multi Gauge v3.0 software (Fuji Film, Tokyo), and bands of interest were normalized to the corresponding actin bands indicated.
Immunization and behavioral analyses
All animal procedures were performed in accordance with a protocol approved by the Animal Care Committee of the National Institute for Longevity Sciences. Several 3-month-old female nontransgenic (non-Tg) and Tg2576 mice that carry and overexpress the human APP gene with the Swedish double mutation (K670N; M671L) of familial AD were purchased from Taconics (Germantown, NY, USA) and maintained at our animal care facility until 13 months of age. To determine whether immunization prevents the development of Alzheimer-like phenotype, 4-month-old Tg2576 mice were administered 1A9 or 2C3 (0.4 mg/kg/week), or PBS intravenously via the tail vein until 13 months of age. Memory functions were measured at 13 months of age in the following four behavioral paradigms, as described previously [
28]: (1) spontaneous alternation in the Y-maze test; (2) novel object recognition test; (3) Morris water maze test; and (4) cued and fear conditioning tests. Mice were sacrificed 3 days after the termination of the behavioral tests for biochemical and histological assessments. Experimental results were analyzed by one-way ANOVA and two-way ANOVA, with Fisher's test for post hoc analysis.
Spontaneous alternation in Y-maze test
The maze was made of black painted wood; each arm was 40 cm long, 12 cm high, 3 cm wide at the bottom, and 10 cm wide at the top. The arms converged at an equilateral triangular central area that was 4 cm long at its longest axis. Each mouse was placed at the center of the apparatus, the sequence and number of arm entries were recorded for each mouse over an 8 min period. Alternation was defined as successive entry into the three arms on overlapping triplet sets. Alternation behavior (%) was calculated as the ratio of actual alternations to possible alternations (defined as the number of arm entries minus two) multiplied by 100.
Novel object recognition test
The test procedure consisted of three sessions: habituation, training, and retention. Each mouse was habituated to the box (30 × 30 × 35 cm), with 10 min of exploration in the absence of objects for 3 days (habituation session). During the training session, two objects were placed at the back corner of the box. A mouse was then placed midway at the front of the box and the total time spent exploring the two objects was recorded for 10 min. During the retention session, the mice were placed back in the same box 24 hr after the training session, in which one of the familiar objects used during the training was replaced with a novel object. The animals were then allowed to explore freely for 10 min, and the time spent exploring each object was recorded. Throughout the experiments, the objects used were counterbalanced in terms of their physical complexity and emotional neutrality. A preference index, a ratio of the amount of time spent exploring any one of the two objects (training session) or the novel object (retention session) over the total time spent exploring both objects, was used to measure cognitive function. To eliminate the influence of the last behavioral test, the objects were changed each time.
Morris water maze test
The Morris water maze test was conducted in a circular pool (120 cm in diameter) filled with water at a temperature of 22 ± 1°C. A hidden platform (one block) (7 cm in diameter) was used. The mice underwent two trials (one block) for 10 consecutive days, during which the platform was left in the same position. The distance taken to locate the escape platform was determined in each trial using the Etho Vision system (Neuroscience Idea Co., Ltd., Osaka, Japan).
Cued and contextual fear conditioning tests
For measuring basal levels of freezing response (preconditioning phase), mice were individually placed in the conditioning cage (17 × 27 × 12.5 cm) for 1 min, then in the conditioning cage (25 × 31 × 11 cm) for 2 min. For training (conditioning phase), mice were placed in the conditioning cage, then a 15 sec tone (80 dB) was delivered as a conditioned stimulus. During the last 5 sec, an unconditioned stimulus was applied through a shock generator (Neuroscience Idea Co., Ltd.). This procedure was repeated 4 times at 15 sec intervals. Cued and contextual tests were carried out 1 day after the fear conditioning. For the cued test, the freezing response was measured in the neutral cage for 1 min in the presence of a continuous-tone stimulus identical to the conditioned stimulus. For the contextual test, mice were placed in the conditioning cage and the freezing response was measured for 2 min in the absence of the conditioned stimulus.
Ethics Statement
The research protocol was approved by the local animal esthetics committees at Research Institute, National Center for Geriatrics and Gerontology (Animal Care Committee) prior to initiation of the study. The research project was approved by the local ethics committee of Hirosaki University Graduate School of Medicine, and Research Institute, National Center for Geriatrics and Gerontology prior to initiation of the study.
Statistical analyses
We used factorial design analysis of variaence (ANOVA) or Mann-Whitney test to analyze data as appropriate. Significant ANOVA values were subsequently subjected to simple main effects analyses or post hoc comparisons of individual means using the Tukey's or Dannett's method as appropriate. We considered p values of 0.05 as significant for all studies. Some of the data obtained from animal experiments were analyzed by two-way ANOVA, with Fisher's test for post hoc analysis.
Competing interests
YO, TY, and MS are the employees in Immunas Pharma Incorporation. TY and MS hold stock options in Immunas Pharma Incorporation. EM, TY, and MS are co-inventors of two filed provisional patent applications titled "Antibody Specific Binding to a Beta Oligomer and The Use" and "Antibodies That Specifically Bind to Aβ Oligomers and Uses Thereof" that cover the antibodies described in this paper, but this does not alter the adherence to all the Molecular Neurodegeneration policies on sharing data and materials. This study has in some parts been funded by a commercial funder, but that does not alter the author's adherence to all the Molecular Neurodegeneration policies on sharing data and materials.
Authors' contributions
Conceived and designed the experiments: MM EM. Performed the experiments: AT YO TK AM TN HS TU NY EM. Analyzed the data: MS MS KY MM. Contributed reagents/materisals/analysis tools: YO TY MS KA. Wrote the paper: AT EM. All authors read and approved the final manuscript.