Arc/Arg3.1 Regulates an Endosomal Pathway Essential for Activity-Dependent β-Amyloid Generation

doi:10.1016/j.cell.2011.09.036

Cell. Author manuscript; available in PMC 2012 October 28.

Published in final edited form as:

Cell. 2011 October 28; 147(3): 615–628.

doi: 10.1016/j.cell.2011.09.036

PMCID: PMC3207263

NIHMSID: NIHMS331002

Arc/Arg3.1 Regulates an Endosomal Pathway Essential for Activity-Dependent β-Amyloid Generation

Jing Wu,¹ Ronald S. Petralia,⁵ Hideaki Kurushima,¹ Hiral Patel,⁶ Mi-young Jung,⁶ Lenora Volk,^1,⁴ Shoaib Chowdhury,¹ Jason D. Shepherd,^1,⁹ Marlin Dehoff,¹ Yueming Li,⁷ Dietmar Kuhl,⁸ Richard L. Huganir,^1,⁴ Donald L. Price,^1,^2,³ Robert Scannevin,⁶ Juan C. Troncoso,^2,³ Philip C. Wong,^1,³ and Paul F. Worley^1,²^*

¹Solomon H. Snyder Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA

²Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA

³Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA

⁴Howard Hughes Medical Institute, Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA

⁵Laboratory of Neurochemistry, National Institute on Deafness and Other Communication Disorders, National Institutes of Health, Bethesda, MD 20892, USA

⁶Biogen Idec, 14 Cambridge Center, Cambridge, MA 02142, USA

⁷Molecular Pharmacology and Chemistry Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA

⁸Molecular Neurobiology, Freie Universität of Berlin, Berlin 14195, Germany

^* Correspondence: Email: pworley/at/jhmi.edu

⁹Present address: The Picower Institute for Learning and Memory, Massachusetts Institute of Technology, Cambridge, MA, USA.

The publisher's final edited version of this article is available at Cell

See other articles in PMC that cite the published article.

Publisher's Disclaimer

Summary

Assemblies of β-amyloid (Aβ) peptides are pathological mediators of Alzheimer's Disease (AD) and are produced by the sequential cleavages of amyloid precursor protein (APP) by β-secretase (BACE1) and γ-secretase. The generation of Aβ is coupled to neuronal activity, however the molecular basis is unknown. Here, we report that the immediate early gene Arc is required for activity-dependent generation of Aβ. Arc is a postsynaptic protein that recruits endophilin2/3 and dynamin to early/recycling endosomes that traffic AMPA receptors to reduce synaptic strength in both Hebbian and non-Hebbian forms of plasticity. The Arc-endosome also traffics APP and BACE1, and Arc physically associates with presenilin1 (PS1) to regulate γ-secretase trafficking and confer activity-dependence. Genetic deletion of Arc reduces Aβ load in a transgenic mouse model of AD. In concert with the finding that patients with AD can express anomalously high levels of Arc, we hypothesize that Arc participates in the pathogenesis of AD.

Introduction

Alzheimer's Disease (AD) is the most common cause of dementia in the elderly, and is characterized by a progressive loss of cognitive functions. The histopathology of AD includes accumulations of extracellular Aβ peptides (neuritic plaques) and intra-neuronal hyperphosphorylated tau (neurofibrillary tangles) (Hardy and Selkoe, 2002). Several lines of evidence suggest that cognitive failure is linked to the generation and deposition of neurotoxic species of Aβ derived by secretase cleavage of the amyloid precursor protein (APP). APP is an integral type I membrane protein that is trafficked through a constitutive secretary pathway and processed at the cell surface, trans-Golgi network (TGN) and endocytic organelles (Thinakaran and Koo, 2008). β APP cleavage enzyme1 (BACE1) and the γ-secretase complex, which includes presenilin1 (PS1), Nicastrin, Aph-1 and Pen2, are also present in ER, TGN, endosomes, and on the cell surface (De Strooper and Annaert, 2010; Small and Gandy, 2006). Amyloidogenic processing of APP is thought to occur in endosomes, but the precise trafficking events remain unclear.

Aβ generation can be modulated by neural activity in the interstitial fluid (ISF) in vivo (Cirrito et al., 2005) or in hippocampal slices (Kamenetz et al., 2003), suggesting that a substantial fraction of Aβ generation is dependent on activity; however, the contribution of activity-dependent Aβ generation to amyloid deposition in vivo is as yet unclear. Aberrant activity in the hippocampus “default pathway” is linked to cognitive decline in patients with AD (Buckner et al., 2005). By contrast, behavioral activation is reported to reduce plaque load in a mouse model of AD (Lazarov et al., 2005), and epidemiological studies suggest that cognitive activity is protective for AD (Cracchiolo et al., 2007). The absence of a molecular understanding of how activity increases Aβ accumulation has limited deeper appreciation of the contribution of activity to AD pathogenesis.

Our previous studies have focused on cellular immediate early genes (IEG) as potential mediators of protein synthesis dependent memory. Among them, Arc is a neuron-specific, postsynaptic protein that interacts with endophilin and dynamin, and contributes to an endocytic pathway that accelerates removal of AMPA receptors from excitatory synapses (Chowdhury et al., 2006). Arc is required for multiple forms of synaptic plasticity that depend on its endocytic function including homeostatic scaling (Shepherd et al., 2006) and mGluR-LTD (Park et al., 2008). Arc is an intriguing candidate that could participate in activity-dependent Aβ generation since endocytic pathways are important for regulation of BACE1 activity (Huse et al., 2000). Moreover, dominant negative dynamin, which blocks endocytosis, reduces Aβ levels in ISF by as much as 70%, and prevents activity-dependent increases in Aβ (Cirrito et al., 2008).

Here, we report that Arc is required for activity-dependent increases of Aβ generation, and reveal a role for Arc in postsynaptic trafficking and processing of APP that appears relevant to the pathogenesis of AD. Arc directly binds the N-terminus of PS1, the catalytic subunit of the γ-secretase complex, and these proteins co-localize in early/recycling endosomes within dendrites. Interruption of the Arc-PS1 interaction prevents activity-dependent increases of Aβ, and cell biological studies support a role for Arc in trafficking γ-secretase to vesicles that process APP. Arc-dependent Aβ generation contributes to plaque deposition in a mouse model of AD, and Arc expression is significantly elevated in the medial prefrontal cortex of patients with pathologically confirmed AD. Together, these findings demonstrate that Arc-dependent mechanisms, which are known to control synaptic strength, also control activity-dependent generation of Aβ that is relevant in the pathogenesis of AD.

Results

Genetic Deletion of Arc Prevents Activity-Dependent Generation of Aβ

We crossed the amyloidogenic AD mouse model (APPswe;PS1ΔE9 transgenic mice) to Arc KO (Arc^-/-) mice. The APPswe;PS1ΔE9 mouse is an established model exhibiting accelerated Aβ amyloidosis attributable to co-expression of two familial AD linked mutations (hereafter referred to as AD mice) (Jankowsky et al., 2001). We first examined soluble hAβ levels in 5-month-old AD;Arc^+/+ and AD;Arc^-/- mice. hAβ40 was not significantly different in hippocampus of AD;Arc^+/+ versus AD;Arc^-/- mice sacrificed immediately upon removal from their home cage (Figure 1A). We used maximal electroconvulsive seizure (MECS) to acutely increase neural activity, and confirmed a ~2 fold increase of soluble hAβ40 5 hr after a single MECS in hippocampus from AD;Arc^+/+ mice (Figure 1A). By contrast, MECS did not increase soluble hAβ40 in hippocampus from AD;Arc^-/- mice despite typical MECS behavioral responses and MECS-induced increase of IEG Homer1a (Figure 1A).

Figure 1

Genetic Deletion of Arc Abolishes Activity-dependent Increases of hAβ40

We examined acute brain slices prepared from 8 to 9 week-old AD;Arc^+/+ and AD;Arc^-/-mice. Hippocampal slices were treated with picrotoxin (100μM), which induces epileptiform activity by blocking GABAergic inhibition. Soluble hAβ40 increased during in vitro incubation, and basal levels were reliably detected after 10 hr incubation (Figure 1B). As reported (Kamenetz et al., 2003), picrotoxin treatment increased soluble hAβ40 in slices from AD mice (Figure 1B). By contrast, picrotoxin failed to increase soluble hAβ40 in slices from AD;Arc^-/-mice.

We next examined DIV10 primary neuronal cultures prepared from AD;Arc^+/+ and AD;Arc^-/- embryos. Soluble hAβ40 was assayed in whole cell lysates of neurons, or in the medium, 4 hr after a complete replacement with fresh glial conditioned medium (which was confirmed to have undetectable levels of Aβ, data not shown). >95% of soluble hAβ40 was present in the medium and only low levels were detected in neuronal lysates (Figure 1C); consistent with reports that Aβ is rapidly secreted after it is generated. In basal conditions, the level of hAβ40 was significantly less in the medium of AD;Arc^-/- neuronal cultures compared to AD;Arc^+/+ cultures. Addition of picrotoxin (100 μM) for 4 hr, resulted in a significant increase of hAβ40 in medium of AD;Arc^+/+ neuronal cultures but not in medium of AD;Arc^-/- cultures. BDNF (100 ng/ml for 4 hr) produced a prominent upregulation of Arc expression and resulted in an increase of hAβ40 in medium of AD;Arc^+/+ but not AD;Arc^-/- cultures (Figure 1C). Lentiviral expression of Arc in AD;Arc^-/- cultures resulted in an increase of soluble Aβ40 in the media compared to either uninfected or lentiviral-GFP infected AD;Arc^-/- cultures. Thus, Arc transgene expression is sufficient to increase soluble Aβ40 generation in AD;Arc^-/- neurons.

Arc Binds the N-terminus of PS1

We performed immunoprecipitation (IP) assays from WT and Arc KO brains using detergent conditions (1% Triton) that dissociate proteins in the γ-secretase complex. PS1-NTF selectively co-immunoprecipitated with Arc, while PS1-CTF, BACE1 and Nicastrin did not (Figure 2A). PS1-CTF and additional components of the γ-secretase, e.g. Nicastrin, co-IPed with Arc when the IP was performed using detergent conditions that retain interactions between components of γ-secretase (see Figure 6). In a reciprocal assay, Arc and PS1-CTF co-immunoprecipitated with PS1-NTF (Figure 2B). The physical interaction of Arc and PS1 was reconstituted in HEK-293 cells using PS1 and Arc transgenes (Figure 2C). PS1-NTF and full length PS1 co-immunoprecipitated with full length Arc, or an N-terminal fragment of Arc (Arc 1-154). Deletion of Arc amino acids 91-100 or 101-130 disrupted Arc's association with PS1. Binding between Arc and PS1 was also reconstituted using recombinant Arc protein purified from E. coli together with Flag-PS1-NTF affinity purified from HEK293 cells (Figure 2D). This in vitro binding was selectively inhibited by a synthetic peptide mimic of the N-terminal 40 amino acids of PS1-NTF, but not a peptide mimic of amino acids of 41-81 (Figure 2E). These observations suggest that Arc and PS1 directly and specifically interact.

Figure 2

Arc Binds to Presenilin 1 N-terminus and Arc Does Not Alter APP or its Processing Enzymes Expression, Trafficking from the Plasma Membrane or Activity

Figure 6

Interruption of Arc Binding to γ-secretase Prevents Activity-dependent Aβ Generation

Arc KO Does Not Alter Surface Expression of APP, BACE1 or γ-secretase, or their Trafficking from the Plasma Membrane

Current models of Aβ generation suggest that APP is processed by BACE1 and γ-secretase in early and recycling endosomes (Small and Gandy, 2006; Vetrivel and Thinakaran, 2006). We examined the possibility that Arc might increase Aβ generation by increasing the rate of endocytosis of APP, BACE1 or γ-secretase. DIV 14 cortical cultures were prepared from AD;Arc^+/+ and AD;Arc^-/- embryos, and surface proteins were labeled by incubating cultures at 4°C with membrane-impermeable biotinylating reagent. Expression of APP, BACE1 and components of γ-secretase, including PS1 and Nicastrin was not different between genotypes in either the total cell lysates or on the plasma membrane (Figure 2F). PS1 NTF antibody detected both PS1ΔE9 transgene and a lower amount of processed native PS1 NTF, and this was not different in AD;Arc^-/- neurons. The rate of endocytosis of surface labeled APP and presenilin assayed 10, 30, and 60 min after return to 37°C was not different between genotypes (Figure S1A and S1B). Treatment of cultured neurons with the selective γ-secretase inhibitor JC-22 (Lewis et al., 2005) resulted in identical increases of C-terminal fragments of APP that result from cleavage by BACE1 and/or α-secretase (Figure 2G), suggesting that APP processing upstream of γ-cleavage is not different in neurons derived from AD;Arc^+/+ and AD;Arc^-/- mice. An in vitro assay of γ-secretase activity measured from forebrain lysates using biotinylated recombinant APP substrate Sb4 (Shelton et al., 2009) in the presence of 0.25% CHAPSO, did not detect an effect of Arc in comparisons of Arc^+/+ versus Arc^-/- or AD;Arc^+/+ versus AD;Arc^-/- mice in terms of production of Aβ40 peptides using a G2-10 antibody (Tian et al., 2010) (Figure 2H). Thus, the effect of Arc to increase Aβ40 generation in cultures could not be related to altered dynamics of APP, BACE1 or γ-secretase trafficking from the plasma membrane, or to prominent changes in either BACE1 or γ-secretase expression or activity.

Arc and PS1 Co-localize in Postsynaptic Dendrites in Association with Endosome Membranes

Biochemical and electron microscopic (EM) methods indicate that Arc protein is present exclusively in the postsynaptic compartment (Chowdhury et al., 2006). EM studies have reported APP and PS1 in both pre and post-synaptic compartments (Ribaut-Barassin et al., 2000). We detected native PS1 in dendrites (Figure 3A) using an antibody with confirmed specificity for immunostaining (Figure S1C). Co-expression of Arc and PS1 transgenes in DIV14 neurons resulted in co-localizing punctae in dendrites (Figure 3B), especially in fine processes where punctae are discrete (technical note: low power images that indicate the position of the processes typically appear overexposed at regions of the neuron including the soma and proximal dendritic branches as a consequence of their greater thickness). Co-localization required co-expression of endophilin 3, which enhances Arc association with endosomes (Chowdhury et al., 2006). To determine whether the Arc-PS1 punctae are endosomes, we examined the colocalization of Arc-PS1 punctae with endosomal markers. In neurons expressing GFP-Rab11, a recycling endosome marker, we could detect the colocalization between Arc, PS1 and Rab11 (Figure 3C). Costes' method (Costes et al., 2004) was applied to evaluate the probability of colocalization between Arc and PS1, or PS1 and Rab11 (Figure 3C, probability graph). The Pearson's coefficient of double-channel images (PS1 and Arc, or PS1 and Rab11) confirms a high probability of colocalization (Figure3D, scatter plot). Quantitative analysis revealed prominent colocalization of Arc with PS1(45.0±3.0%), Arc with EEA1(26.3±3.4%) and Rab11(50.0±4.4%), and PS1 with EEA1(23.5±3.6%) and Rab11 (43.3±4.5%) (Figure 3D, table). EM analysis of normal rat hippocampus revealed that native PS1 and native Arc co-localize in postsynaptic spines of CA1 neurons (Figure 3E and 3F). Gold particles associated with vesicles immediately beneath the postsynaptic density, as well as with structures identified as early/recycling endosomes based on their tubulovesicular structures (Figure 3G-P). Several profiles of endosomal complex structures extending from the sorting/early endosome (oval, spindly vesiculate shape) labeled with Arc and PS1 (Figure 3G-L, 3N and 3O). Double EM also revealed co-localization of PS1 with Rab11 (Figure 3M and 3P).

Figure 3

Arc and Presenilin1 Co-localize in Endosomes in Neuronal Dendrites

APP is Rapidly Internalized in Postsynaptic Early/Recycling Endosomes that Associate with Arc

We examined trafficking of APP in dendrites of DIV14 neurons prepared from AD;Arc^+/+ mice. Neurons were fixed and stained with 6E10 monoclonal antibody, which reacts with the Aβ sequence present in hAPP. Staining was present along dendrites that co-stained with MAP2 (Figure 4A). A similar pattern of staining was detected with two additional APP antibodies: 22C11, which reacts with the N-terminus of human or mouse APP (Figure S1D), and 4G8, which reacts with the Aβ epitope in both human and mouse APP (data not shown). By incubating live neurons from AD;Arc^+/+ with 6E10 at 10°C, and then returning the neurons to 37°C for 10 min, we could detect hAPP internalized into dendrites (Figure 4B). The same result was obtained with 22C11 (Figure S1E). hAPP labeled on the surface by 6E10 was internalized into punctae that co-localized with Rab11 (Figure 4C), suggesting APP is trafficked in recycling endosomes. Consistent with dendritic localization, APP, PS1ΔE9 and Arc are enriched in biochemical fractions of the postsynaptic density (Figure 4D). EM analysis of normal rat hippocampus with two different APP antibodies verified the presence of APP in postsynaptic vesicles near the PSD, and in endosomal complexes composed of different tubulovesicular structures (Figure 4E-N, 4G8 antibody; Figure S1F-K, 22C11 antibody). Double EM revealed co-localization of APP with Arc (Figure 4E-G, 4I-K, 4M and 4N; Figure S1G-J), and Arc with Rab11 (Figure S2A-G) at endosomes. Quantitative analysis of immunogold labeling of dendrites demonstrated a substantial amount of Arc, PS1 and APP are associated with tubulovesicular structures (Figure 4O). Additionally, Arc, endophilin3 and dynamin co-fractionate with PS1 and APP-CTF in detergent-resistant, buoyant fractions (lipid raft) prepared from AD;Arc^+/+ mice that are enriched for proteins that process APP (Vetrivel et al., 2005) (Figure S2H). This distribution was not altered in AD;Arc^-/- mice (Figure S2I).

Figure 4

APP is Rapidly Internalized into Endosomes in Neuronal Dendrites and Associates with Arc

Arc Increases Association of γ-secretase with APP Containing Endosomes

Transiently expressed hAPP was labeled on the dendritic surface with 6E10 monoclonal antibody (Figure S3A), and endocytosed hAPP was detected as bright, discrete punctae in both dendrites and axons (Figure S3C and S3D), consistent with endogenous APP. Identical results were observed with wild-type hAPP and hAPPswe (data not shown). In comparison to natively expressed hAPP, hAPP transgene could be detected with direct conjugated secondary antibody, and could be easily assessed for co-localization with other proteins. When hAPP trangene was co-expressed with Arc, endophilin3 and GFP-Rab11, surface labeled hAPP rapidly internalized into dendrites and co-localized with Arc and Rab 11 (Figure 5A). Evaluation of colocalization by Costes' approach revealed a high probability of colocalization between internalized APP and Arc, or internalized APP and Rab11 (Figure 5A, graph). Quantitative analysis further demonstrated the positive correlation (Figure 5B, scatter plot) and a significant population of colocalized APP (internalized) and Arc is present in endosomes and with higher percentage in recycling endosomes [Figure 5B and S4A, 33.6±4.3%, 28.4±1.7% and 45.8±3.7% respectively for early (Rab5⁺), late (Rab7⁺) and recycling (Rab11⁺) endosomes]. ~50% of internalized hAPP colocalized with Arc (Figure S3E and S3F). Another early/recycling endosome marker, transferrin receptor, also colocalized with Arc and internalized hAPP (Figure S4B). hAPPswe was similarly internalized and co-localized with transferrin receptor (Figure S4C). Association between internalized APP and early or recycling endosomes was further confirmed by live confocal imaging using GFP-Rab5 or Rab11 and Alexa 568 labeled APP antibody (Figure S4D). These observations support the hypothesis that APP is trafficked in dendritic early/recycling endosomes that associate with Arc.

Figure 5

Trafficking Assays; Native Arc Enhances Association of Native γ-secretase with hAPP Transgene Containing Endosomes

A similar approach was used to monitor trafficking of BACE1. HA-BACE1 was detected on the neuronal surface by live labeling with HA antibody (Figure S3B), and was rapidly internalized into dendrites where it co-localized with Arc-endophilin punctae (Figure5C).

To examine the possibility that Arc might regulate the association of APP with γ-secretase, we expressed hAPP transgene in Arc^+/+ and Arc^-/- neurons and assessed co-localization with native PS1. hAPP was labeled on the membrane with 6E10 and internalized following a 10 min return to 37°C. hAPP was identically internalized in Arc^+/+ and Arc^-/- neurons (Figure 5E), consistent with biochemical assays of native APP (Figure 2F and S1A). PS1 particle number in randomly selected 50 μm dendritic segments was not different between Arc^+/+ and Arc^-/- neurons (Figure S5). Internalized hAPP punctae frequently co-localized with PS1 punctae (Figure 5D). Costes' method revealed a high probability of co-localization (not shown). We quantitatively compared the degree of co-localization of internalized hAPP with native PS1 in randomly selected 50 μm dendritic segments. Colocalization of PS1 with APP containing vesicles was significantly greater in Arc^+/+ than Arc^-/- neurons (Figure 5F). This finding supports a model in which Arc functions to increase γ-secretase in dendritic trafficking endosomes that process APP.

Interruption of Arc Association with γ-secretase Prevents Activity-Dependent Increase of Aβ Generation

To further examine the role of Arc binding to PS1 in the generation of Aβ, we generated lentivirus expressing Flag tagged PS1 NTF and confirmed that Flag-PS1 NTF co-immunoprecipitates with Arc from neurons (Figure 6A). A previous study reported that PS1-NTF does not integrate into γ-secretase (Levitan et al., 2001), and we confirmed that Flag-PS1 NTF does not associate with components of the native γ-secretase complex in neurons (Figure 6B). Using lysates prepared from neurons and detergent conditions that do not disrupt the γ-secretase complex (Capell et al., 1998), Arc co-immunoprecipitated PS1-CTF and Nicastrin (indicators of its association with the γ-secretase complex), and this association was interrupted by Flag-PS1 NTF (Figure 6C). Importantly, expression of Flag-PS1 NTF in AD;Arc^+/+ neurons blocked the activity-dependent increase of Aβ40 (Figure 6D). Flag-PS1 NTF also produced a modest reduction of basal Aβ in AD;Arc^+/+ neurons. By contrast, Flag-PS1 NTF did not alter basal or induced expression of Aβ40 in AD;Arc^-/- cultures indicating that Flag-PS1 NTF does not alter γ-secretase activity in neurons that lack Arc.

Aβ and Plaque Load are Reduced in AD;Arc^-/- Mice

We next determined if Arc contributes to amyloid generation and deposition in vivo. The APP_SWE;PS1ΔE9 model shows an age-dependent increase of PBS-soluble and PBS-insoluble (formic acid-soluble) hAβ together with demonstrable plaque in cortex by 6 months of age. Genetic crosses of AD;Arc^+/- with Arc^+/- assured that all mice carried a single copy of the APPswe and PS1ΔE9 transgene, and provided littermate controls. Western blots confirmed that expression of native mouse APP and human APPswe was identical in forebrain of AD;Arc^+/+ versus AD;Arc^-/- mice (Figure S6A). Similarly, expression of BACE1 and components of γ-secretase complex were not different between genotypes (Figure S6A). However, in 6-month-old male mice, the amount of Aβ40 in a formic acid soluble fraction was reduced in AD;Arc^-/- mice. Aβ42 showed the same trend but was not statistically different between genotypes (Figure 7A). In 12-month-old mice, both Aβ40 and Aβ42 levels were reduced in AD;Arc^-/- versus AD;Arc^+/+ male (Figure 7B) and female mice (Figure S6B). Immunohistochemistry in cohorts of 6-month-old male (Figure 7C and 7D) and 12-month-old female mice (Figure S6C and S6D) confirmed that areas occupied by Aβ plaque were reduced in AD;Arc^-/-. These results suggest that Arc contributes significantly to amyloid burden in this AD model.

Figure 7

Arc Contributes to Aβ levels and Plaque Load in AD Mice, and is Elevated in Medial Frontal Cortex (MFC) of Patients with Alzheimer's Disease

Arc Expression is Increased in Medial Frontal Cortex of Patients with Alzheimer's Disease

To further explore the role of Arc in AD, we compared expression of Arc protein in gray matter punch samples from medial frontal cortex (MFC) or occipital cortex (OCC) of autopsies of patients with dementia that was highly likely due to AD based on the presence of both neuritic plaque and neurofibrillary tangles in neocortex (CERAD C and Braak Stage V and VI) and patients with intermediate likelihood that dementia was due to AD based on neuritic plaque in neocortex and neurofibrillary tangles in limbic regions (CERAD B and Braak Stage III and IV) versus samples from age matched controls (patient data is summarized in Table S1). MFC and OCC were selected because MFC is commonly impacted in AD and used in both CERAD and Braak Staging as representative neocortex, while OCC (primary visual area) is not a major brain region impacted in AD. Levels of Arc protein in MFC were increased in the combined AD group compared to the control group (Figure 7E, 7F and S7A) by ~2 fold. Arc levels were not significantly different between the high likelihood versus intermediate likelihood AD cases. Zif268, an IEG that shows similar response as Arc to activity, was not different between AD and control brains. Levels of Arc or Zif268 protein in OCC were not different between AD and control brains. While there are important limitations of post mortem tissue, the present data suggest that Arc expression is maintained at normal or supranormal levels in a brain area impacted by AD, and thus could contribute to Aβ generation and pathology.

Discussion

The present study demonstrates that Arc is required for activity-dependent increases of Aβ, APP is trafficked in the dendritic compartment, and associates with Arc in early and recycling endosomes that presumably return processed APP as Aβ to the extracellular space where it accumulates. Arc also associates with presenilin in dendritic vesicular structures that include early endosomes and recycling endosomes. Arc directly binds the N-terminus of PS1, and interruption of the interaction blocks activity-dependent increases of Aβ generation. Importantly, inhibitory effects of Flag-PS1 NTF on Aβ generation are evident only in neurons that express Arc, and γ-secretase activity is not different in Arc KO brain, suggesting that Arc does not directly modify the activity of γ-secretase. Rather, Arc increases the association of PS1/γ-secretase with endosomes that traffic APP. The catalytic site of the γ-secretase complex is present within the transmembrane domain, and this mandates that its substrate be present in the same membrane. Since APP is present in the plasma membrane, along with BACE1 and low levels of components of γ-secretase, Arc could function to increase γ-secretase recruitment from the plasma membrane to the same endosome as APP and BACE1. However, we do not detect an Arc-dependent change in the rate of internalization of γ-secretase, as might be expected in such a model. Consequently, we favor a model in which Arc that associates with recycling endosomes assists in sorting γ-secretase from an intracellular source, (where it is enriched in tubulovesicular structures including presumptive sorting endosomes in dendrites) to early and recycling endosomes that process APP-CTF.

Notch is another well-characterized substrate of γ-secretase that is processed to generate transcription regulating peptide NICD. While Notch signaling is best characterized during development, Notch also plays a role in adult brain and is implicated in synaptic plasticity. A recent study demonstrates that Notch is processed to NICD in dendrites of mature neurons, and this is regulated by activity, and is dependent on Arc (Alberi et al., 2011). Notch processing is associated with endocytic trafficking, and the Arc/γ-secretase mechanism in the present study could underlie activity-dependent increases of γ-secretase processing of Notch.

Arc is best known for its role in trafficking of AMPA receptors and synaptic plasticity. Persistent activity that increases Arc expression in neuronal culture models or in visual cortex results in homeostatic down-regulation of synaptic strength by increasing the rate of AMPA receptor endocytosis (Gao et al., 2010; Shepherd et al., 2006). The present study indicates that Arc induction also increases the generation of Aβ. Persistent aberrant activity as in the “default network” (Buckner et al., 2005) may underlie the increase of Arc that we detect in MFC in AD brains, and the effect of Arc to increase Aβ generation could contribute to the amyloid deposition associated with this activity (Sperling et al., 2009). Our studies validate the association between Arc-dependent Aβ generation and plaque deposition in an in vivo model. This link between Arc and AMPA receptor down regulation suggests that the persistence of aberrant activity in the default pathway in the AD brain may be indicative of failure of homeostatic scaling, which in a healthy brain would reestablish AMPA receptor dependent excitability and restore normal levels of activity. Arc could also be up regulated as part of a neuroinflammatory process (Rosi et al., 2005) or as a consequence of accumulation of Aβ oligomers (Lacor et al., 2004). We have confirmed that synthetic hAβ40 dimer can increase Arc expression in cultured neurons (Figure S7B-D), which suggests that Arc and Aβ oligomer could act in a positive feedback loop. The dual functions of Arc in the neural plasticity related to cognitive function, and Aβ generation, suggest a fundamental link between these processes that is disrupted in AD.

Experimental Procedures

Mouse Strains

Both Arc^+/+ and Arc^−/− mice are in congenic C57BL/6J background. Arc/Arg3.1 heterozygous mice (Arc^+/−) were crossed with AD mouse model (APP_SWE;PS1ΔE9) (C3H/HeJ * C57BL/6J, obtained from Dr. Philip Wong's lab) to generate AD;Arc^+/- mice, which were backcrossed to Arc^+/− to generate AD;Arc^+/+ and AD;Arc^−/− mice. All procedures involving animals were under the guidelines of JHU Institutional Animal Care and Use Committee.

Co-Immunoprecipitation, In vitro Binding, Peptide Blocking Assay, and Surface Protein Trafficking Assay

Detailed protocols are provided in the Extended Experimental Procedures.

Immunoelectron Microscopy

Immuno-electron microscopy was performed by post-fixation immunogold labeling as described in detail in the Extended Experimental Procedures.

Aβ Assays on Mouse Brain

Brains of AD;Arc^+/+ and AD;Arc^-/- mice were dissected on ice and homogenized in PBS buffer containing complete protease inhibitor cocktail (Roche). After the lysates were centrifuged at 100,000 × g for 30 min, the supernatants containing soluble Aβ peptides were collected for assay, and the pellets were homogenized in 70% formic acid solution. After incubation on ice for 1 hr, the formic acid lysates were centrifuged at 100,000 × g for 1 hr, and the supernatants were collected and neutralized by 1 M Tris-base solution. The concentrations of Aβ40/Aβ42 peptides in PBS-soluble and formic acid-soluble fractions were measured using a quantitative ELISA kit (Invitrogen) that specifically detects human Aβ40/Aβ42. BCA method was used to measure the total protein concentrations (Pierce).

Immunohistochemical Analysis of Amyloid Plaques in Mouse Brains

Mouse brains were fixed in 10% formaldehyde, dehydrated in methanol, treated with xylenes, and embedded in paraffin. The brain was cut parasagitally into 4 μm sections. Aβ load was measured in a parasagittal section immunostained with 6E10 antibody and counterstained with hematoxylin. Using the Stereo Investigator Optical Disector software (Version 7) from MBF Biosciences, we defined a contour that encompassed the whole cerebral cortex (excluding basal ganglia and hippocampus) with a grid size of 250×250μm. The distance between the fractionator sampling sites was 10 μm. The fractional area occupied by Aβ immunoreactivity in the demarcated area was measured using the area fraction fractionator probe (AFF) with a 40× objective. Statistical differences between the fractional areas of Aβ in AD;Arc^+/+ and AD;Arc^-/-mice were evaluated using the Mann-Whiteney U test.

Statistical Analysis

All data were analyzed statistically by Student's t test, two-tailed Mann-Whiteney U test and ANOVA with Turkey post hoc test using GraphPad Prism. All data are presented as mean ± SEM.

Supplementary Material

Click here to view.^{(16M, pdf)}

Acknowledgments

We would like to thank Gopal Thinakaran and Sangram Sisodia (University of Chicago), Tong Li for critical reagents and helpful discussions. Thanks to Gay Rudow for helping with unbiased stereology. Thanks to Desheng Xu and Shi Yang for help with lentivirus packaging. Many thanks to Robert Ardiuni, Melissa Levesque, Darren baker, and Stephan Miller from Biogen Idec for their excellent technical support on in vitro binding assay. We thank Ya-Xian Wang for help with the EM studies. We also thank Robert Malinow (University of California at San Diego) and Mike Ehlers (Duke University) for APP Sindbis virus constructs and GFP-Rab constructs. This work was supported by NIMH grant RO1 MH053608 (P.F.W.), the Johns Hopkins Alzheimer's Disease Research Center (NIH grant P50AG05146) and in part by the NIDCD Intramural Program (R. S. P.) and NIH grant MH084020.

Footnotes

Supplemental Information: Supplemental information includes Extended Experimental Procedures, seven figures, and Table S1.

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

Alberi L, Liu S, Wang Y, Badie R, Smith-Hicks C, Wu J, Pierfelice TJ, Abazyan B, Mattson MP, Kuhl D, et al. Activity-induced Notch signaling in neurons requires Arc/Arg3.1 and is essential for synaptic plasticity in hippocampal networks. Neuron. 2011;69:437–444. [PMC free article] [PubMed]
Buckner RL, Snyder AZ, Shannon BJ, LaRossa G, Sachs R, Fotenos AF, Sheline YI, Klunk WE, Mathis CA, Morris JC, Mintun MA. Molecular, structural, and functional characterization of Alzheimer's disease: evidence for a relationship between default activity, amyloid, and memory. J Neurosci. 2005;25:7709–7717. [PubMed]
Capell A, Grunberg J, Pesold B, Diehlmann A, Citron M, Nixon R, Beyreuther K, Selkoe DJ, Haass C. The proteolytic fragments of the Alzheimer's disease-associated presenilin-1 form heterodimers and occur as a 100-150-kDa molecular mass complex. J Biol Chem. 1998;273:3205–3211. [PubMed]
Chowdhury S, Shepherd JD, Okuno H, Lyford G, Petralia RS, Plath N, Kuhl D, Huganir RL, Worley PF. Arc/Arg3.1 interacts with the endocytic machinery to regulate AMPA receptor trafficking. Neuron. 2006;52:445–459. [PMC free article] [PubMed]
Cirrito JR, Kang JE, Lee J, Stewart FR, Verges DK, Silverio LM, Bu G, Mennerick S, Holtzman DM. Endocytosis is required for synaptic activity-dependent release of amyloid-beta in vivo. Neuron. 2008;58:42–51. [PMC free article] [PubMed]
Cirrito JR, Yamada KA, Finn MB, Sloviter RS, Bales KR, May PC, Schoepp DD, Paul SM, Mennerick S, Holtzman DM. Synaptic activity regulates interstitial fluid amyloid-beta levels in vivo. Neuron. 2005;48:913–922. [PubMed]
Costes SV, Daelemans D, Cho EH, Dobbin Z, Pavlakis G, Lockett S. Automatic and quantitative measurement of protein-protein colocalization in live cells. Biophys J. 2004;86:3993–4003. [PubMed]
Cracchiolo JR, Mori T, Nazian SJ, Tan J, Potter H, Arendash GW. Enhanced cognitive activity--over and above social or physical activity--is required to protect Alzheimer's mice against cognitive impairment, reduce Abeta deposition, and increase synaptic immunoreactivity. Neurobiol Learn Mem. 2007;88:277–294. [PMC free article] [PubMed]
De Strooper B, Annaert W. Novel Research Horizons for Presenilins and gamma-Secretases in Cell Biology and Disease. Annu Rev Cell Dev Biol 2010 [PubMed]
Gao M, Sossa K, Song L, Errington L, Cummings L, Hwang H, Kuhl D, Worley P, Lee HK. A specific requirement of Arc/Arg3.1 for visual experience-induced homeostatic synaptic plasticity in mouse primary visual cortex. J Neurosci. 2010;30:7168–7178. [PMC free article] [PubMed]
Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science. 2002;297:353–356. [PubMed]
Huse JT, Pijak DS, Leslie GJ, Lee VM, Doms RW. Maturation and endosomal targeting of beta-site amyloid precursor protein-cleaving enzyme. The Alzheimer's disease beta-secretase. J Biol Chem. 2000;275:33729–33737. [PubMed]
Jankowsky JL, Slunt HH, Ratovitski T, Jenkins NA, Copeland NG, Borchelt DR. Co-expression of multiple transgenes in mouse CNS: a comparison of strategies. Biomol Eng. 2001;17:157–165. [PubMed]
Kamenetz F, Tomita T, Hsieh H, Seabrook G, Borchelt D, Iwatsubo T, Sisodia S, Malinow R. APP processing and synaptic function. Neuron. 2003;37:925–937. [PubMed]
Lacor PN, Buniel MC, Chang L, Fernandez SJ, Gong Y, Viola KL, Lambert MP, Velasco PT, Bigio EH, Finch CE, et al. Synaptic targeting by Alzheimer's-related amyloid beta oligomers. J Neurosci. 2004;24:10191–10200. [PubMed]
Lazarov O, Robinson J, Tang YP, Hairston IS, Korade-Mirnics Z, Lee VM, Hersh LB, Sapolsky RM, Mirnics K, Sisodia SS. Environmental enrichment reduces Abeta levels and amyloid deposition in transgenic mice. Cell. 2005;120:701–713. [PubMed]
Levitan D, Lee J, Song L, Manning R, Wong G, Parker E, Zhang L. PS1 N-and C-terminal fragments form a complex that functions in APP processing and Notch signaling. Proc Natl Acad Sci U S A. 2001;98:12186–12190. [PubMed]
Lewis SJ, Smith AL, Neduvelil JG, Stevenson GI, Lindon MJ, Jones AB, Shearman MS, Beher D, Clarke E, Best JD, et al. A novel series of potent gamma-secretase inhibitors based on a benzobicyclo[4.2.1]nonane core. Bioorg Med Chem Lett. 2005;15:373–378. [PubMed]
Ribaut-Barassin C, Moussaoui S, Brugg B, Haeberle AM, Huber G, Imperato A, Delhaye-Bouchaud N, Mariani J, Bailly YJ. Hemisynaptic distribution patterns of presenilins and beta-APP isoforms in the rodent cerebellum and hippocampus. Synapse. 2000;35:96–110. [PubMed]
Rosi S, Ramirez-Amaya V, Vazdarjanova A, Worley PF, Barnes CA, Wenk GL. Neuroinflammation alters the hippocampal pattern of behaviorally induced Arc expression. J Neurosci. 2005;25:723–731. [PubMed]
Shelton CC, Tian Y, Frattini MG, Li YM. An exo-cell assay for examining real-time gamma-secretase activity and inhibition. Mol Neurodegener. 2009;4:22. [PMC free article] [PubMed]
Shepherd JD, Rumbaugh G, Wu J, Chowdhury S, Plath N, Kuhl D, Huganir RL, Worley PF. Arc/Arg3.1 mediates homeostatic synaptic scaling of AMPA receptors. Neuron. 2006;52:475–484. [PMC free article] [PubMed]
Small SA, Gandy S. Sorting through the cell biology of Alzheimer's disease: intracellular pathways to pathogenesis. Neuron. 2006;52:15–31. [PubMed]
Sperling RA, Laviolette PS, O'Keefe K, O'Brien J, Rentz DM, Pihlajamaki M, Marshall G, Hyman BT, Selkoe DJ, Hedden T, et al. Amyloid deposition is associated with impaired default network function in older persons without dementia. Neuron. 2009;63:178–188. [PMC free article] [PubMed]
Thinakaran G, Koo EH. Amyloid precursor protein trafficking, processing, and function. J Biol Chem. 2008;283:29615–29619. [PubMed]
Tian Y, Bassit B, Chau D, Li YM. An APP inhibitory domain containing the Flemish mutation residue modulates gamma-secretase activity for Abeta production. Nat Struct Mol Biol. 2010;17:151–158. [PubMed]
Vetrivel KS, Cheng H, Kim SH, Chen Y, Barnes NY, Parent AT, Sisodia SS, Thinakaran G. Spatial segregation of gamma-secretase and substrates in distinct membrane domains. J Biol Chem. 2005;280:25892–25900. [PMC free article] [PubMed]
Vetrivel KS, Thinakaran G. Amyloidogenic processing of beta-amyloid precursor protein in intracellular compartments. Neurology. 2006;66:S69–73. [PubMed]