Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Jan 1.
Published in final edited form as: Neurobiol Dis. 2011 Sep 3;45(1):381–387. doi: 10.1016/j.nbd.2011.08.027

Preferential accumulation of amyloid-beta in presynaptic glutamatergic terminals (VGluT1 and VGluT2) in Alzheimer’s disease cortex

Sophie Sokolow 1,2,3, Sanh H Luu 1, Karabi Nandy 1,4, Carol A Miller 5, Harry V Vinters 6, Wayne W Poon 7, Karen H Gylys 1,2,3
PMCID: PMC3339276  NIHMSID: NIHMS333721  PMID: 21914482

Abstract

Amyloid-beta (Aβ) is thought to play a central role in synaptic dysfunction (e.g. neurotransmitter release) and synapse loss. Glutamatergic dysfunction is involved in the pathology of Alzheimer’s disease (AD) and perhaps plays a central role in age-related cognitive impairment. Yet, it is largely unknown whether Aβ accumulates in excitatory boutons. To assess the possibility that glutamatergic terminals are lost in AD patients, control and AD synaptosomes were immunolabeled for the most abundant vesicular glutamate transporters (VGluT1 and VGluT2) and quantified by flow cytometry and immunoblot methods. In post-mortem parietal cortex from aged control subjects, glutamatergic boutons are fairly abundant as approximately 40% were immunoreactive for VGluT1 (37%) and VGluT2 (39%). However, the levels of these specific markers of glutamatergic synapses were not significantly different among control and AD cases. To test the hypothesis that Aβ is associated with excitatory terminals, AD synaptosomes were double-labeled for Aβ and for VGluT1 and VGluT2, and analyzed by flow cytometry and confocal microscopy. Our study demonstrated that Aβ immunoreactivity (IR) was present in glutamatergic terminals of AD patients. Quantification of Aβ and VGluT1 in a large population of glutamatergic nerve terminals was performed by flow cytometry, showing that 42% of VGluT1 synaptosomes were immunoreactive for Aβ compared to 9% of VGluT1 synaptosomes lacking Aβ-IR. Percentage of VGluT2 synaptosomes immunoreactive for Aβ (21%) was significantly higher than VGluT2 synaptosomes lacking Aβ-IR (9%). Moreover, Aβ preferentially affects VGluT1 (42% positive) compared to VGluT2 terminals (21%). These data represent the first evidence of high levels of Aβ in excitatory boutons in AD cortex and support the hypothesis that Aβ may play a role in modulating glutamate transmission in AD terminals.

Introduction

Alzheimer’s disease (AD) is the most common form of dementia, affecting > 5 million Americans (Hebert et al., 2003). Abnormal amyloid-beta (Aβ) metabolism, tau hyperphosphorylation and synapse loss are characteristic features of AD. Although the molecular triggers of AD are unknown, extensive studies indicate that Aβ plays a principal role in AD pathogenesis (Kuo et al., 1996; McLean et al., 1999; Naslund et al., 2000).

Presynaptic buildup of soluble Aβ is suggested as a proximal cause of synapse dysfunction in AD (Fein et al., 2008; Shankar et al., 2007; Shankar et al., 2008). Natural human Aβ dimers are able to (i) mediate loss of dendritic spines and excitatory synapses; (ii) inhibit long-term potentiation (LTP) and facilitate long-term depression (LTD) in normal rodent hippocampus, and (iii) impede learned behavior in normal rats (Shankar et al., 2007; Shankar et al., 2008). The degree of clinical impairment strongly correlates to the decrease of electrophysiologically active synapses especially glutamatergic pyramidal synapses (Shankar et al., 2007; Shankar et al., 2008; Terry et al., 1991). Nonetheless, a direct link between presynaptic Aβ and the loss of excitatory function has yet to be established.

Glutamate is a key neurotransmitter in primary perception and cognition. It is the principal excitatory neurotransmitter in the neocortex and hippocampus, major brain areas affected in AD. Furthermore, glutamate is associated with excitotoxicity and LTP. Despite the central role of glutamatergic synapses in learning and memory, their involvement in AD pathology remains unclear. Many neurochemical studies have associated a dysfunction of the glutamatergic system with AD pathology, including reduction of glutamic acid content in AD brain (Procter et al., 1988), reduced receptor binding (Cross et al., 1987) and decreased cortical and hippocampal glutamate uptake which was interpreted as a result of glutamatergic synapse loss (Cowburn et al., 1988; Hardy et al., 1987). Recently, synthetic Aβ peptide species were shown to potentiate K+-induced glutamate release from normal rodent hippocampus (Kabogo et al., 2010).

With the availability of VGluT specific antibodies, immunohistochemical investigation of excitatory boutons has become possible. VGluT1 and VGluT2 are presynaptic proteins responsible for glutamate transport into synaptic vesicles and are found in functionally different subtypes of neurons (Fremeau et al., 2004; Herzog et al., 2001; Takamori et al., 2000; Varoqui et al., 2002). Several immunohistochemical studies yielded conflicting results in VGluT1 and VGluT2 expression in AD post-mortem brain tissue. Glutamatergic presynaptic bouton density has been reported as elevated in midfrontal gyrus of patients with mild cognitive impairment (MCI) and depleted in mild- and severe stage of AD (Bell et al., 2007). Another group observed an important reduction in VGluT1 and VGluT2 expression in AD prefrontal cortex; VGluT1 reduction was correlated to the clinical dementia rating score (Kashani et al., 2008). Others found reduced VGluT1 levels in parietal and occipital cortex but not in frontal cortex of AD patients; VGluT2 expression was not modified in these brain areas (Kirvell et al., 2006). In a more recent investigation, VGluT1 expression and cognitive scores were correlated but no difference was observed in VGluT1 expression in prefrontal and temporal cortex of control and AD patients (Kirvell et al., 2010). Because of the discrepancy between studies, the involvement of excitatory synapses in AD remains unclear, despite their central role in clinical symptoms that characterize the disease.

In this study, we aimed to (i) detect alterations in glutamate transporter levels by examining large population of AD and control synaptosomes using flow cytometry analysis, and (ii) determine the degree to which Aβ is present in excitatory terminals. In the present experiments, Aβ, VGluT1 and VGluT2 were quantified by flow cytometry and confocal microscopy in surviving terminals of AD parietal cortex (Brodmann area A7). This area of the CNS was previously reported to be severely affected in AD and abundantly expressing Aβ (Fein et al., 2008; Kirvell et al., 2006).

Methods

Human brain specimens

A total of 15 individuals were studied including 8 with AD (7 females, 1 males, age 86.0 ± 3.3 yr; mean postmortem delay: 6.9 ± 0.8 h) and 7 aged cognitively normal controls (4 females, 3 males, age 89.4 ± 3.3 yr, mean postmortem delay: 7.4 ± 1.4 h). Brain specimens from parietal cortex (Brodmann area A7) were obtained at autopsy from the Alzheimer’s disease Research Centers at the University of Southern California, the University of California at Los Angeles and the University of California at Irvine. Table 1 shows the demographic characteristics of the participants, postmortem interval (PMI), Braak stage and frequency of plaques, neutropil threads, gliosis and neuron loss.

Table 1. Sokolow et al.

Clinical and neuropathological characteristics of patients and controls Neuropil threads and plaques are reported for parietal cortex and refer to silver (Gallyas) stain. The plaque number includes plaques with and without cores; sparse (< 5/field), mod (6 to 20/field), freq (21 to 30/field).

Cases Sex Age
(yr)		 Ethnicity PMI
(h)		 MMSE
(1-30)		 CDR
(1-3)		 Parietal
cortex
atrophy		 Neuritic
plaques		 Neurofib.
changes
(neuropil
threads)		 Neuronal
loss		 Gliosis
Normal
726# F 97 Caucasian 5.5 ND ND Mild None None None None
779*,# F 89 Caucasian 8.4 29 ND Moderate None None None None
789*,# F 105 Caucasian 6.0 30 0.5 Mild None None None None
810* M 81 Hispanic 3.0 27 ND Mild None None None None
824*,# F 87 Caucasian 7.0 29 0 None None None None None
1509# M 82 ND 14.9 ND ND None None None None None
1209# M 85 ND 6.8 ND ND None None None None None
AD
825# III F 68 Caucasian 9.75 ND ND None Moderate Moderate None None
782*,# IV-V F 89 Caucasian 9.0 ND ND Mild Moderate Sparse None None
787*,# IV F 87 Hispanic 6.0 ND ND Mild Moderate Sparse None Sparse
745*,# V F 92 Caucasian 4.8 ND ND Mild Moderate Sparse None None
776*,# V F 99 Caucasian 7.5 ND ND Mild Sparse None None None
1009# VI F 82 ND 4.5 ND ND Frequent Frequent Frequent Sparse Sparse
102*,# VI F 80 ND ND ND ND Moderate Moderate Frequent None None
028*,# VI M 91 ND ND ND ND Moderate Frequent Frequent None None
Open in a new tab

Abbreviations: CDR, cognitive dementia rating; F, female; M, male; ND, not documented; PMI, post-mortem interval.

*

Samples used for flow cytometry analysis.

#

Samples used for Western Blot analysis.

Synaptosome preparation

Synaptosomes were prepared from cryopreserved human brain tissue. Samples (1 to 3 g) were minced and slowly frozen on the day of autopsy in 10% dimethyl sulfoxide and 0.32 M sucrose and stored at −80°C until homogenization. The crude synaptosome fraction was prepared as previously described (Fein et al., 2008). Briefly, the minced tissue was homogenized in 10 volume of 0.32 M sucrose containing protease and phosphatase inhibitors. The homogenate was first centrifuged at 1,000 g for 10 min. The supernatant was centrifuged at 10,000 g for 20 min to obtain the crude synaptosomal pellet (P-2) which contains synaptic terminals that have resealed into functional spheres during homogenization in sucrose. Aliquots of P-2 are routinely cryopreserved in 0.32 M sucrose and stored at −80°C, according to protocols previously described for optimal preservation (Dodd et al., 1986).

Western blotting analysis

Protein concentrations in synaptosome-enriched fractions (SEF) were determined using BCA assays (Thermo Scientific, Waltham, MA). Samples were not boiled, as VGluT proteins aggregate after boiling. Non-boiled samples were electrophoresed on 8% SDS-page gels (Expedeon, San Diego, CA). Membranes were blocked for 1 h at room temperature in 5% BSA, followed by incubation overnight at 4°C with the primary antibodies in PBS containing 0.01% Tween 20 (PBS-T) and 1.5% (W/V) BSA and 0.05% sodium azide: VGluT1 1:1000 (Abcam, Cambridge, MA) and VGluT2 1:1000 (Synaptic Systems, Germany). Membranes were then incubated for 1 h with an anti-rabbit IgG horseradish peroxidase (HRP)-conjugated (1:30,000; Jackson ImmunoResearch, West Grove, PA). Membranes were incubated with SuperSignal West Dura substrate (Thermo Scientific, Waltham, MA) and exposed to an OptiChemi HR Camera 600 (UVP Imaging, Upland, CA). Quantification of proteins was performed following optimal exposure time using the VisionWorksLS Image Acquisition and Analysis software (UVP Imaging, Upland, CA). To strip the immunoblots, membranes were incubated at room temperature in 0.1 M glycine at pH 2.5. At the end of each experiment, membranes were rinsed in large volumes of PBS-T and equal loading of protein was verified by staining PVDF membranes with Coomassie Blue. Control and AD samples were analyzed simultaneously within the same blot. All experiments were performed in duplicate.

Immunolabeling of synaptosomal fraction

P-2 aliquots were immunolabeled for flow cytometry analysis according to an intracellular antigen staining method (Gylys et al., 2000). Pellets were fixed in 0.25% buffered paraformaldehyde (PAF) and permeabilized in 0.2% Tween20/PBS (15 min., 37°C). VGluT1-2 antibodies were labeled directly with Zenon kit Alexa Fluor 488 or 647 reagents according to manufacturer (Invitrogen, Carlsbad, CA). This mixture was added to P-2 aliquots and incubated at RT for 30 min. Pellets were washed 2 times with 1 ml 0.2% Tween20/PBS, then resuspended in PBS buffer for flow cytometry analysis. The synaptosomal pellet was dispersed for all washes and for incubations with fixative, detergent, antibodies and then collected by centrifugation (1310 × g at 4°C). Control and AD synaptosomes were immunolabeled in the same experiment.

Fluo-4NW

Synaptosomal preparation was resuspended in a physiologic salt solution (PSS) consisting of 145 mM NaCl, 10 mM HEPES, 5 mM KCl, 1.2 mM MgCl2, 1.2 mM CaCaCl2 and 10 mM glucose. Synaptosomes were incubated for 30 min at 37°C, with the loading solution containing the calcium indicator dye fluo-4NW (Molecular Probes, Eugene, OR). Synaptosomes were then washed for 5 min and the synaptosomal pellet was incubated for an additional 15 min with the PSS solution in order to allow complete esterase cleavage of the fluo-4 dye. After a final wash, the synaptosomal pellet was resuspended in PSS and analyzed by flow cytometry analysis.

Flow cytometry

For flow cytometry analysis, immunolabeled synaptosomes were transferred to a flow tube and analyzed as previously described (Gylys et al., 2000). Data was acquired using a BD-Calibur I analytic flow cytometer (Becton-Dickinson, San Jose, CA) equipped with argon 488 nm and 635 nm diode lasers. Ten thousand particles were collected and analyzed for each sample. Debris was excluded by establishing a size threshold set on forward light scatter. Alexa-488 and fluo-4NW were detected by the FL1 photomultiplier tube detector, whereas Alexa 647 was assessed by the FL4 photomultiplier tube detector. Analysis was performed using FCS Express software (DeNovo Software, Los Angeles, CA).

Confocal microscopy

Synaptosome-enriched fractions were immunolabeled as described above and washed, then dispersed with a pipette and spread on slides. Slides were dried, coverslipped with Prolong Antifade (Molecular Probes, Eugene, OR), and stored at 4°C. VGluT1-2 antibodies were coupled with Alexa 488 and 10G4, were labeled with Alexa 568. Confocal fluorescence and differential interference contrast images of synaptosomes were taken using a 100X 1.4 Planapo objective lens on a Leica SP2 Confocal Inverted Microscope (Heidelberg, Germany) equipped with argon laser (488 nm excitation) and diode pump (561 nm excitation).

Statistical data analysis

For simple statistical analysis, means ± SEM were calculated using SigmaPlot (Systat Software, San Jose, CA) and One-way ANOVA was used for multiple comparisons of means. For comparisons of VGluT1 and VGluT2 between normal and AD cases, a mixed model ANOVA was fit to the data, treating VGluT as a within-subject factor and the two groups as a between-subject factor. Groups and VGluT were considered as fixed effects while subjects were considered as random effects. Treating subjects as random allowed for the covariances in the model to vary due to differences among subjects, although subjects were still assumed to be independent. Interaction between factors was also included in the model. In the event where there were significant interaction between group and VGluT, we investigated further into their simple effects and differences in the least square means. A similar mixed model ANOVA was fit to compare Aβ in VGluT1 and VGluT2 among AD cases. VGluT and Aβ were both considered to be within-subject fixed effects while subjects were treated as random. Descriptive statistics were obtained and mixed models were built using SAS 9.2. (SAS Institute Inc., Cary, NC).

Results

Quantification of VGluT1 and VGluT2 terminals in human parietal cortices

As synaptic terminals are not well resolved by light microscopy, even at the highest magnification, we analyzed immunolabeled synaptosomes by flow cytometry (FACS). This method quantifies multiple parameters on each nerve terminal in a sample (Fein et al., 2008), including fluorescence and forward scatter (FSC), which is proportional to the size of a particle. We have previously reported that non-synaptosomal elements are excluded by drawing a size-based gate that includes only particles that are ~ 0.75 to 1.5 microns (Gylys et al., 2004a; Gylys et al., 2007). In the present study, ten thousand size-gated particles were acquired using an acquisition gate based on size standard beads (Gylys et al., 2004a; Gylys et al., 2007). Only specific immunolabeled particles were included in a gate drawn on background labeling in the presence of an isotype-specific control antibody (Fig.1A). The difference in protein level was expressed as a change in the number of positive particles (Gylys et al., 2004a). Background labeling is illustrated for a representative sample in Fig. 1A. The degree of synaptosomal purity is illustrated in Fig. 1B: synaptosomes were acquired based on size and analyzed as percentage of immunopositive particles for the synaptosomal-associated protein SNAP-25 (Fein et al., 2008). Fig. 1C shows a representative dot plot of a total synaptosome preparation and size standards used to determine the size exclusion. Synaptosome integrity was evaluated using fluo-4NW, a Ca2+ sensitive fluorescent dye. Fluo-4NW is an acetoxymethyl (AM) ester derivative capable of permeating cell as an uncharged molecule. Once inside the cell, cytoplasmic esterases cleave the ester group, resulting in a charged form of the dye that cannot leak out of intact cells. Among 10,000 synaptosomes labeled with the fluorescent dye, 99% displayed an intense fluorescent signal (Fig. 1D). This indicates that synaptosomal membranes were intact in our preparation.

Fig. 1.

Fig. 1

Open in a new tab

Flow cytometry analysis of VGluT1 and VGluT2 in control and AD synaptosomes. Human cortical synaptosomal preparation immunolabeled with (A) non-immune IgG and (B) SNAP-25. Positive control sample labeled for SNAP-25 with side scatter dot plot showing polystyrene size standards (C). Synaptosome integrity assessed with the fluorescent dye, Fluo-4NW (D). Representative density plots of normal (E and F) and AD (G and H) synaptosomes labeled for VGluT1 (E-G) and VGluT2 (F-H). The forward scatter parameter on the X-axis is proportional to size. Synaptic distribution of VGluT1 and VGluT2 in control and AD parietal cortex are shown (I). Mixed model ANOVA, VGluT * group interaction [F =66.90; P < 0.001], values represent means ± SEM. *, P < 0.05 between VGluT1-IR in AD compared to controls.

Synaptosomes from aged control and AD were immunolabeled with VGluT isotype-specific antibodies and analyzed by flow cytometry. Representative density plots of VGluT1 and VGluT2 immunofluorescence in one normal and one AD patient are shown in Fig. 1E-H. The number of immunofluorescent particles included in the rectangular gate represents the synaptosomes specifically immunolabeled with isotype-specific antibodies for VGluT1 (Fig. 1E and G) and VGluT2 (Fig. F and H). Glutamatergic terminals were fairly abundant in human parietal cortex of aged control and AD cases and there was a significant VGluT by group interaction (Fig. 1I). The mean number of VGluT1 synaptosomes was significantly higher in AD (49%, N = 6) as compared to normals (37%, N = 4, p < 0.05). A reduced number of VGluT2 terminals was noted for the AD cases but the difference was not significant. Samples were also dual labeled for VGluT1 and VGluT2 to determine the degree of VGlut1 and VGluT2 co-localization, which revealed that ~18% of glutamatergic terminals contained both VGluT1 and VGluT2 (data not shown). Taken together with the data in Fig. 1 I, these results indicate that approximately 60-70% of synaptic terminals in parietal cortex are glutamatergic.

Our Western blot analysis showed that the concentration of the VGluT1 and VGluT2 in SEF were subject to intergroup variation with a trend for reduced VGluT1 and VGluT2 levels. However these differences did not reach significance for the region studied with this method (Fig. 2). In contrast to flow data, Western blot analysis showed an apparent decrease in VGluT1 immunoreactivity in AD versus controls. This difference was not significant and can be justified by the difference in methods: indeed Western blot is a semi-quantitative technique and less accurate than flow cytometry. Cases used for flow cytometry and Western Blot analyses are shown in Table 1.

Fig. 2.

Fig. 2

Open in a new tab

Representative immunoblots of VGluT 1 (A) and VGluT2 (B) expression in SEF isolated from Brodmann area A7 in 3 aged normal controls and 3 Alzheimer’s disease (AD) patients. (C) Densitometric analyses of VGluT1 and VGluT2 proteins. Data were expressed as percentage of the control group, taken as 100%. Values are means ± SEM. N = 6 normal and 8 AD cases for VGluT1 (100.0 ± 6.9 vs. 83.6 ± 6.5, P > 0.05). N = 5 normal and 5 AD cases for VGluT2 (100.0 ± 15.4 vs. 67.7 ± 19.0, P > 0.05).

Co-localization of Aβ in glutamatergic terminals

Confocal analysis reveals that Aβ is co-localized in glutamatergic boutons immunoreactive for VGluT1 (Fig. 3A-B) and VGluT2 (Fig. 3E-F). Arrowheads showed co-localization of Aβ and VGluT1-2 in a subset of synaptosomes (Fig.3C and 3G); differential contrast images (Fig. 3D and 3H) revealed the spherical structure of the synaptosome particles.

Fig. 3.

Fig. 3

Open in a new tab

Confocal analysis of VGluT1 and Aβ co-localization (A-C). Washed synaptosomes isolated from a 95 y/o AD case (parietal cortex) were dual labeled for VGluT1 (A) and Aβ (B). Merge image, arrowheads showing the co-localization of VGluT1 and Aβ in a subset of synaptosomes (C). VGluT2 and Aβ co-localization (E-G). Synaptosomal preparation dual labeled for VGluT2 (E) and Aβ (F). Merge image, arrowheads showing the co-localization of VGluT2 and Aβ in a subset of synaptosomes (G). Differential interference contrast images (D and H).

The degree of synaptic co-localization was then quantified by flow cytometry. Samples were double labeled for Aβ and VGluT1-2 specific antibodies and only size gated particles were examined in order to ensure ~ 90% synaptosomal purity (Gylys et al., 2004a). Ten thousand synaptosomes were collected. The number of single-positive (Aβ−/VGluT1-2+) and double-positive (Aβ+/VGluT1-2+) synaptosomes was analyzed using quadrant analysis (Fein et al., 2008). Positive immunofluorescent synaptosomes were identified by drawing a quadrant that excludes background labeling in the lower right, upper left and upper right quadrants (Fig. 4A). This method allowed us to quantify (i) glutamatergic synaptosomes which are immunoreactive only for VGluTs (Aβ−/VGluT1-2+); (ii) non glutamatergic synaptosomes positive for Aβ only (Aβ+/VGluT1-2−) and (iii) glutamatergic synaptosomes immunoreactive for Aβ (Aβ+/VGluT1-2+) (Fig. 4B-E). The degree of Aβ co-localization in glutamatergic terminals of a representative AD case is illustrated in Fig. 4B (Aβ/VGluT1) and Fig. 4C (Aβ/VGluT2), and for a typical normal control sample in Fig. 4D and 4E.

Fig. 4.

Fig. 4

Open in a new tab

Co-localization of VGluT1-2 and Aβ. (A) Quadrant analysis of human cortical synaptosomal preparation immunolabeled with non-immune IgG. A representative sample of AD parietal cortex double labeled for VGluT1 and Aβ (B) and VGluT2 and Aβ (C).Quadrant analysis for one representative aged normal control sample double labeled for VGluT1 and Aβ (D) and VGluT2 and Aβ (E). Glutamatergic synaptosomes containing Aβ (Abeta+/VGluT+) are in the upper right quadrant, non-glutamatergic synaptosomes positive for Aβ (Abeta+/VGluT−) are in the lower right quadrant and glutamatergic synaptosomes lacking Aβ (Abeta−/VGluT+) are in the upper left quadrant. Numbers represent the percentage of synaptosomes in each quadrant. (F) Aβ variation in single-labeled (Abeta+/VGluT− only, silver bars; Abeta−/VGluT+ only, grey bars) and dual-labeled (Abeta+/VGluT+, black bars) fractions of six AD parietal cortices. Due to the low Aβ levels in aged normal controls, this analysis was only performed in AD cases. Mixed model ANOVA, VGluT * Aβ interaction [F = 7.27; P < 0.01]. Values represent means ± SEM. *P<0.05 and **P < 0.01, dual-labeled compared to single-labeled fractions; ##P < 0.0001 Abeta+/VGluT1+ dual-labeled fractions compared to Abeta+/VGluT2+ dual-labeled fractions.

In AD parietal cortex, Aβ preferentially accumulates in VGluT1 synaptosomes (42% of Aβ+/VGluT1+ N = 6) as compared to non-VGluT1 synaptosomes (27% of Aβ+/VGluT1−, N = 6). For both transporters, the number of glutamatergic terminals accumulating Aβ (42% of Aβ+/VGluT1+ and 21% Aβ+/VGluT2+) was significantly higher than the number of glutamatergic terminals unaffected by Aβ 9% of Aβ−/VGluT1+ and 9% of Aβ−/VGluT2+). Lastly, using a mixed model ANOVA, we found a significant interaction between VGluT and Aβ signifying that Aβs accumulates preferentially in VGluT1 surviving terminals (42% of Aβ+/VGluT1+) compared to VGluT2 (21% of Aβ+/VGluT2+)(Fig. 4F).

Discussion

Glutamate dysfunction mediated by Aβ oligomeric species is thought to be an important aspect of AD pathogenesis but the underlying mechanisms remain intangible. Abnormal synaptic plasticity induced by Aβ is also attributed to early memory loss that precedes neuronal degeneration (Selkoe, 2002). Several studies reported the co-localization of Aβ with postsynaptic NMDA-receptors (Dewachter et al., 2009; Lacor et al., 2007) but it remains unclear whether Aβ is present within presynaptic glutamatergic terminals. The present results highlight colocalization of Aβ and vesicular glutamate transporters in the synaptic compartment of human diseased brain.

The major finding of our study is that the percentage of VGluT1 synaptosomes containing Aβ (42% of Aβ+/VGluT1+) was almost twice higher than the percentage of non-glutamatergic synaptosomes affected by Aβ (27% of Aβ+/VGluT1−); i.e., indicating that a significant pool of surviving glutamatergic terminals have accumulated Aβ. Moreover, for both transporters, the number of glutamatergic terminals accumulating Aβ was two to four times higher than the number of glutamatergic terminals devoid of Aβ (Aβ−/VGluT1-2+); i.e., most surviving excitatory terminals in AD neocortex are likely to be affected by Aβ. We also showed up regulation of VGluT1 but not VGluT2 in AD cortical synapses and a preferential Aβ buildup in VGluT1 surviving terminals. The elevation in VGluT1 positive terminals in AD may be a feedback response to the higher Aβ levels in this population, which would be expected to differentially affect glutamatergic function. VGluT1 and VGluT2 have been considered as separate neuronal populations but our data suggest that there may be expression of both transporters and some overlap of function in a minor subpopulation of terminals. Consistent with this observation, VGluT1 mRNA is expressed in pyramidal neurons in all layers of the human cortex whereas VGluT2 mRNA was only found in the superficial layers (II and III). This expression pattern suggests that VGluT1 and VGluT2 are associated with specific efferent circuits (i.e. cognition, emotion, sensation) (McCullumsmith and Meador-Woodruff, 2003).

Increased interstitial Aβ has been associated with synaptic activity in vivo in mouse models (Cirrito et al., 2008; Cirrito et al., 2005) and human patients (Brody et al., 2008). Therefore the presynaptic accumulations observed in the present experiments may contribute to production and synaptic release of Aβ by glutamatergic neurons. Release of Aβ by glutamatergic synapses is also indicated by the finding that activation of group II metabotrophic glutamate receptors (mGluRs) triggered secretase activity and synaptic Aβ release in TgCRND8 mice (Kim et al., 2010). Glutamate as well as Aβ release may be affected; for example, age-related reduction in stimulated glutamate release has been observed in APP/PS1 transgenic mice (Minkeviciene et al., 2008); moreover, spontaneous excitatory release may also be affected in AD (Pratt et al., 2011). Glutamatergic signaling and function is also regulated by interactions of receptors like NR2B with post-synaptic scaffolding proteins, which may ultimately enhance presynaptic glutamate release (Proctor et al., 2011). Another mechanism by which the presence of Aβ in glutamatergic terminals may induce synaptotoxicity is through the inhibition of glutamate reuptake. The resulting elevation of extracellular glutamate followed by desensitization would be expected to ultimately activate perisynaptic NMDA-receptors and mGluRs (Li et al., 2009).

In this report, the Aβ assembly states associated with VGluT1-2 are more likely to be a 56 kD oligomeric species and small oligomers < 20 kD, which are prominent in synaptic terminals (Fein et al., 2008; Lesné et al., 2006; Sokolow et al., 2011). Indeed, our group previously reported a strong correlation between presynaptic Aβ and several oligomeric species of Aβ peptide in AD synaptosomes (Fein et al., 2008; Sokolow et al., 2011). Selkoe and his collaborators have suggested that functional synaptic changes in AD lead to synaptic failure and are responsible for the cognitive deficits prior to neurodegeneration (Selkoe, 2002). In rat hippocampus, secreted Aβ dimers were capable of selectively blocked LTP of excitatory synaptic transmission, an electrophysiological correlate of learning and memory (Shankar et al., 2008; Walsh et al., 2002). Inhibition of LTP by Aβ oligomers has also been attributed to NMDA-receptor-dependent pathways (Shankar et al., 2007). Hence, analogous effects caused by oligomeric Aβ buildup in glutamatergic boutons might contribute to AD synaptotoxicity and subsequently cognitive impairment in AD patients. More recently, Puzzo et al. have suggested that at picomolar concentrations Aβ can actually play a positive role on neurotransmission and synaptic plasticity (Puzzo et al., 2008). Since the pathological concentration of Aβ oligomers is unknown, Aβ peptides may trigger synaptic potentiation in AD glutamatergic synapses, perhaps at the early stages of the disease. This is consistent with the work of Abramov et al. who demonstrated that release of endogenous Aβ peptides play a critical role in regulating synaptic function and mediate presynaptic potentiation (Abramov et al., 2009). In their report, excitatory synapses were shown to be highly sensitive to changes in Aβ levels (Abramov et al., 2009). Moreover, GluR2 subunit of the α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR) increased as MMSE declined in incipient AD synaptoneurosomes (Williams et al., 2009). Within this context, it will be important to examine glutamatergic synaptic Aβ levels in a larger set of early AD cases (Braak and Braak I-III) as well as comparing glutamatergic synaptic Aβ levels in brain regions differently affected by the disease progression (e.g. entorhinal cortex vs. motor cortex).

Analyzing intact nerve terminals, we did not observed loss of glutamatergic endings in AD compared to aged controls. Although other investigators have observed reduced levels of VGluT1 in AD parietal cortex (Kirvell et al., 2006), the discrepancy could be explain by the difference in methods. Integrity of AD synaptosomes was demonstrated by the size of the acquired particles and the conservation of presynaptic markers in synaptosomes affected by Aβ (Gylys et al., 2004a; Gylys et al., 2004b) as well as by the retention of the viability dye, fluo-4AM, in this report. The fact that in AD cortex we did not observe any loss of VGluT1-2 boutons does not rule out the possibility that Aβ-bearing terminals are abnormally active compared to Aβ-free synapses. Indeed, unusual patterns of neuronal activity, such as reduced glutamatergic excitatory currents and increased proportion of hyperactive and hypoactive neurons were reported in mouse models of AD (Busche et al., 2008; Palop et al., 2007; Palop and Mucke, 2010). In AD patients, aberrant neural network activity was also documented (Buckner et al., 2005).

Taken together, the present results support the hypothesis that accumulation of Aβ in glutamatergic terminals may have a direct effect on surviving glutamatergic terminals and glutamate function in AD cortex. Complimentary studies using animal models of AD are needed to determine the functional effects of Aβ accumulation in glutamatergic boutons over the course of AD pathology.

Conclusion

Our report demonstrates the presence of Aβ in cortical glutamatergic terminals and underlies their important contribution in AD pathology. These findings emphasize the importance of elucidating the synaptic cascade triggered by Aβ in order to prevent glutamate dysfunction and cognitive loss. This study supports the valuable therapeutic benefits of targeting glutamatergic neurons to treat Alzheimer’s disease.

Highlights.

>We quantify glutamatergic terminals in Alzheimer’s disease and aged control brain.>Glutamatergic boutons are fairly abundant in pathologic and aged control cortex. >We examine levels of Abeta in glutamatergic boutons in Alzheimer’s disease cortex. > High levels of Abeta are found in excitatory boutons in Alzheimer’s disease cortex.

Acknowledgement

We are thankful Mr. David Moom for technical assistance. This work was supported by NIH AG27465 to KHG, by NIA AG18879 to CAM. HVV is supported by the Daljit S. and Elaine Sarkaria Chair in Diagnostic Medicine. Tissue was obtained from the Alzheimer’s Disease Research Center Neuropathology Cores of USC (NIA 050 AG05142), UCLA (NIA P50 AG 16970), and UC Irvine (NIA P50 AG016573). Flow cytometry was performed in the UCLA Jonsson Comprehensive Cancer Center (JCCC) and Center for AIDS Research Flow Cytometry Core Facility supported by NIH CA16042 and AI 28697, and by the JCCC, the UCLA AIDS Institute, the David Geffen School of Medicine and the Chancellor’s Office at UCLA. The authors wish to thank the families who generously donated the brain samples for the present research program.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Abramov E, et al. Amyloid-beta as a positive endogenous regulator of release probability at hippocampal synapses. Nat Neurosci. 2009;12:1567–76. doi: 10.1038/nn.2433. [DOI] [PubMed] [Google Scholar]
  2. Bell KFS, et al. Paradoxical Upregulation of Glutamatergic Presynaptic Boutons during Mild Cognitive Impairment. J. Neurosci. 2007;27:10810–10817. doi: 10.1523/JNEUROSCI.3269-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Brody DL, et al. Amyloid-beta dynamics correlate with neurological status in the injured human brain. Science. 2008;321:1221–4. doi: 10.1126/science.1161591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Buckner RL, et al. Molecular, structural, and functional characterization of Alzheimer’s disease: evidence for a relationship between default activity, amyloid, and memory. J Neurosci. 2005;25:7709–17. doi: 10.1523/JNEUROSCI.2177-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Busche MA, et al. Clusters of hyperactive neurons near amyloid plaques in a mouse model of Alzheimer’s disease. Science. 2008;321:1686–9. doi: 10.1126/science.1162844. [DOI] [PubMed] [Google Scholar]
  6. Cirrito JR, et al. Endocytosis Is Required for Synaptic Activity-Dependent Release of Amyloid-[beta] In Vivo. Neuron. 2008;58:42–51. doi: 10.1016/j.neuron.2008.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cirrito JR, et al. Synaptic activity regulates interstitial fluid amyloid-beta levels in vivo. Neuron. 2005;48:913–22. doi: 10.1016/j.neuron.2005.10.028. [DOI] [PubMed] [Google Scholar]
  8. Cowburn R, et al. Presynaptic and postsynaptic glutamatergic function in Alzheimer’s disease. Neurosci Lett. 1988;86:109–113. doi: 10.1016/0304-3940(88)90192-9. [DOI] [PubMed] [Google Scholar]
  9. Cross AJ, et al. Sodium dependent d-[3H]aspartate binding in cerebral cortex in patients with Alzheimer’s and Parkinson’s diseases. Neurosci Lett. 1987;79:213–217. doi: 10.1016/0304-3940(87)90699-9. [DOI] [PubMed] [Google Scholar]
  10. Dewachter I, et al. Deregulation of NMDA-receptor function and down-stream signaling in APP[V717I] transgenic mice. Neurobiol Aging. 2009;30:241–56. doi: 10.1016/j.neurobiolaging.2007.06.011. [DOI] [PubMed] [Google Scholar]
  11. Dodd P, et al. Optimization of freezing, storage, and thawing conditions for the preparation of metabolically active synaptosomes from frozen rat and human brain. Molecular and Chemical Neuropathology. 1986;4:177–198. doi: 10.1007/BF02834357. [DOI] [PubMed] [Google Scholar]
  12. Fein JA, et al. Co-localization of amyloid beta and tau pathology in Alzheimer’s disease synaptosomes. Am J Pathol. 2008;172:1683–92. doi: 10.2353/ajpath.2008.070829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fremeau RT, Jr., et al. Vesicular glutamate transporters 1 and 2 target to functionally distinct synaptic release sites. Science. 2004;304:1815–9. doi: 10.1126/science.1097468. [DOI] [PubMed] [Google Scholar]
  14. Gylys KH, et al. Quantitative characterization of crude synaptosomal fraction (P-2) components by flow cytometry. J Neurosci Res. 2000;61:186–92. doi: 10.1002/1097-4547(20000715)61:2<186::AID-JNR9>3.0.CO;2-X. [DOI] [PubMed] [Google Scholar]
  15. Gylys KH, et al. Enrichment of presynaptic and postsynaptic markers by size-based gating analysis of synaptosome preparations from rat and human cortex. Cytometry A. 2004a;60:90–6. doi: 10.1002/cyto.a.20031. [DOI] [PubMed] [Google Scholar]
  16. Gylys KH, et al. Increased cholesterol in Abeta-positive nerve terminals from Alzheimer’s disease cortex. Neurobiol Aging. 2007;28:8–17. doi: 10.1016/j.neurobiolaging.2005.10.018. [DOI] [PubMed] [Google Scholar]
  17. Gylys KH, et al. Synaptic changes in Alzheimer’s disease: increased amyloid-beta and gliosis in surviving terminals is accompanied by decreased PSD-95 fluorescence. Am J Pathol. 2004b;165:1809–17. doi: 10.1016/s0002-9440(10)63436-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hardy J, et al. Region-specific loss of glutamate innervation in Alzheimer’s disease. Neurosci Lett. 1987;73:77–80. doi: 10.1016/0304-3940(87)90034-6. [DOI] [PubMed] [Google Scholar]
  19. Hebert LE, et al. Alzheimer Disease in the US Population: Prevalence Estimates Using the 2000 Census. Arch Neurol. 2003;60:1119–1122. doi: 10.1001/archneur.60.8.1119. [DOI] [PubMed] [Google Scholar]
  20. Herzog E, et al. The Existence of a Second Vesicular Glutamate Transporter Specifies Subpopulations of Glutamatergic Neurons. J. Neurosci. 2001;21:181RC. doi: 10.1523/JNEUROSCI.21-22-j0001.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kabogo D, et al. [beta]-Amyloid-related peptides potentiate K+-evoked glutamate release from adult rat hippocampal slices. Neurobiol Aging. 2010;31:1164–1172. doi: 10.1016/j.neurobiolaging.2008.08.009. [DOI] [PubMed] [Google Scholar]
  22. Kashani A, et al. Loss of VGLUT1 and VGLUT2 in the prefrontal cortex is correlated with cognitive decline in Alzheimer disease. Neurobiol Aging. 2008;29:1619–30. doi: 10.1016/j.neurobiolaging.2007.04.010. [DOI] [PubMed] [Google Scholar]
  23. Kim SH, et al. Group II metabotropic glutamate receptor stimulation triggers production and release of Alzheimer’s amyloid(beta)42 from isolated intact nerve terminals. J Neurosci. 2010;30:3870–5. doi: 10.1523/JNEUROSCI.4717-09.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kirvell SL, et al. Vesicular glutamate transporter and cognition in stroke: a case-control autopsy study. Neurology. 2010;75:1803–9. doi: 10.1212/WNL.0b013e3181fd6328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kirvell SL, et al. Down-regulation of vesicular glutamate transporters precedes cell loss and pathology in Alzheimer’s disease. J Neurochem. 2006;98:939–50. doi: 10.1111/j.1471-4159.2006.03935.x. [DOI] [PubMed] [Google Scholar]
  26. Kuo YM, et al. Water-soluble Abeta (N-40, N-42) oligomers in normal and Alzheimer disease brains. J Biol Chem. 1996;271:4077–81. doi: 10.1074/jbc.271.8.4077. [DOI] [PubMed] [Google Scholar]
  27. Lacor PN, et al. Abeta oligomer-induced aberrations in synapse composition, shape, and density provide a molecular basis for loss of connectivity in Alzheimer’s disease. J Neurosci. 2007;27:796–807. doi: 10.1523/JNEUROSCI.3501-06.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lesné S, et al. A specific amyloid-[beta] protein assembly in the brain impairs memory. Nature. 2006;440:352–357. doi: 10.1038/nature04533. [DOI] [PubMed] [Google Scholar]
  29. Li S, et al. Soluble Oligomers of Amyloid [beta] Protein Facilitate Hippocampal Long-Term Depression by Disrupting Neuronal Glutamate Uptake. Neuron. 2009;62:788–801. doi: 10.1016/j.neuron.2009.05.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. McCullumsmith RE, Meador-Woodruff JH. Expression of transcripts for the vesicular glutamate transporters in the human medial temporal lobe. Ann N Y Acad Sci. 2003;1003:438–42. doi: 10.1196/annals.1300.046. [DOI] [PubMed] [Google Scholar]
  31. McLean CA, et al. Soluble pool of Aβ amyloid as a determinant of severity of neurodegeneration in Alzheimer’s disease. Ann Neurol. 1999;46:860–866. doi: 10.1002/1531-8249(199912)46:6<860::aid-ana8>3.0.co;2-m. [DOI] [PubMed] [Google Scholar]
  32. Minkeviciene R, et al. Age-related decrease in stimulated glutamate release and vesicular glutamate transporters in APP/PS1 transgenic and wild-type mice. J Neurochem. 2008;105:584–94. doi: 10.1111/j.1471-4159.2007.05147.x. [DOI] [PubMed] [Google Scholar]
  33. Naslund J, et al. Correlation between elevated levels of amyloid beta-peptide in the brain and cognitive decline. Jama. 2000;283:1571–7. doi: 10.1001/jama.283.12.1571. [DOI] [PubMed] [Google Scholar]
  34. Palop JJ, et al. Aberrant excitatory neuronal activity and compensatory remodeling of inhibitory hippocampal circuits in mouse models of Alzheimer’s disease. Neuron. 2007;55:697–711. doi: 10.1016/j.neuron.2007.07.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Palop JJ, Mucke L. Amyloid-[beta]-induced neuronal dysfunction in Alzheimer’s disease: from synapses toward neural networks. Nat Neurosci. 2010;13:812–818. doi: 10.1038/nn.2583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Pratt KG, et al. A novel role for {gamma}-secretase: selective regulation of spontaneous neurotransmitter release from hippocampal neurons. J Neurosci. 2011;31:899–906. doi: 10.1523/JNEUROSCI.4625-10.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Procter AW, et al. Evidence of Glutamatergic Denervation and Possible Abnormal Metabolism in Alzheimer’s Disease. J Neurochem. 1988;50:790–802. doi: 10.1111/j.1471-4159.1988.tb02983.x. [DOI] [PubMed] [Google Scholar]
  38. Proctor DT, et al. Post-synaptic scaffolding protein interactions with glutamate receptors in synaptic dysfunction and Alzheimer’s disease. Prog Neurobiol. 2011;93:509–21. doi: 10.1016/j.pneurobio.2011.02.002. [DOI] [PubMed] [Google Scholar]
  39. Puzzo D, et al. Picomolar amyloid-beta positively modulates synaptic plasticity and memory in hippocampus. J Neurosci. 2008;28:14537–45. doi: 10.1523/JNEUROSCI.2692-08.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Selkoe DJ. Alzheimer’s disease is a synaptic failure. Science. 2002;298:789–91. doi: 10.1126/science.1074069. [DOI] [PubMed] [Google Scholar]
  41. Shankar GM, et al. Natural oligomers of the Alzheimer amyloid-beta protein induce reversible synapse loss by modulating an NMDA-type glutamate receptor-dependent signaling pathway. J Neurosci. 2007;27:2866–75. doi: 10.1523/JNEUROSCI.4970-06.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Shankar GM, et al. Amyloid-beta protein dimers isolated directly from Alzheimer’s brains impair synaptic plasticity and memory. Nat Med. 2008;14:837–42. doi: 10.1038/nm1782. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sokolow S, et al. AD synapses contain abundant Abeta monomer and multiple soluble oligomers, including a 56-kDa assembly. Neurobiol Aging. 2011 doi: 10.1016/j.neurobiolaging.2011.05.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Takamori S, et al. Identification of a vesicular glutamate transporter that defines a glutamatergic phenotype in neurons. Nature. 2000;407:189–94. doi: 10.1038/35025070. [DOI] [PubMed] [Google Scholar]
  45. Terry RD, et al. Physical basis of cognitive alterations in Alzheimer’s disease: synapse loss is the major correlate of cognitive impairment. Ann Neurol. 1991;30:572–80. doi: 10.1002/ana.410300410. [DOI] [PubMed] [Google Scholar]
  46. Varoqui H, et al. Identification of the Differentiation-Associated Na+/PI Transporter as a Novel Vesicular Glutamate Transporter Expressed in a Distinct Set of Glutamatergic Synapses. J. Neurosci. 2002;22:142–155. doi: 10.1523/JNEUROSCI.22-01-00142.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wallén-Mackenzie Å, et al. Restricted Cortical and Amygdaloid Removal of Vesicular Glutamate Transporter 2 in Preadolescent Mice Impacts Dopaminergic Activity and Neuronal Circuitry of Higher Brain Function. The Journal of Neuroscience. 2009;29:2238–2251. doi: 10.1523/JNEUROSCI.5851-08.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wallén-Mackenzie Å, et al. Genetic inactivation of the vesicular glutamate transporter 2 (VGLUT2) in the mouse: What have we learnt about functional glutamatergic neurotransmission? Upsala Journal of Medical Sciences. 2010;115:11–20. doi: 10.3109/03009730903572073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Walsh DM, et al. Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature. 2002;416:535–9. doi: 10.1038/416535a. [DOI] [PubMed] [Google Scholar]
  50. Williams C, et al. Transcriptome Analysis of Synaptoneurosomes Identifies Neuroplasticity Genes Overexpressed in Incipient Alzheimer’s Disease. PLoS ONE. 2009;4:e4936. doi: 10.1371/journal.pone.0004936. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Wojcik SM, et al. An essential role for vesicular glutamate transporter 1 (VGLUT1) in postnatal development and control of quantal size. Proc Natl Acad Sci U S A. 2004;101:7158–63. doi: 10.1073/pnas.0401764101. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES