WO2009042727A1 - Immediate early gene arc interacts with endocytic machinery and regulates the trafficking and function of presenilin - Google Patents

Abstract

In accordance with these discoveries, there are provided amino acid fragments of an Arc polypeptide, wherein the fragment binds presenilin 1. Also provided are methods of identifying a compound that inhibits binding of an Arc polypeptide to presenilin 1. Further provided are methods of inhibiting gamma-secretase activity in a cell, as well as inhibiting amyloid beta plaque formation. Finally, there are provided methods of inhibiting the formation of amyloid beta plaques in, on or around neurological tissue in subjects having or is at risk of having a neurodegenerative disease.

Description

IMMEDIATE EARLY GENE ARC INTERACTS WITH ENDOCYTIC MACHINERY AND REGULATES THE TRAFFICKING AND FUNCTION OF PRESENILIN

FIELD OF THE INVENTION

[0001] The present invention relates generally to the regulation of amyloid beta production and more specifically, to the interaction of Arc and presenilin 1 in the formation of a functional gamma secretase complex.

BACKGROUND OF THE INVENTION

[0002] Alzheimer's disease (AD) is one of the most common forms of dementia, and is one of the leading causes of death in the United States. Nearly 30% of all 85-year-olds have AD. Alzheimer's Disease is characterized by a progressive loss of cognitive function and histopathological changes that include amyloid plaque. The pathogenesis of AD is linked to the generation and deposition of the peptide Aβ. Recent studies indicate that cognitive defects are attributable to extracellular, soluble oligomeric Aβ that acts to alter forms of synaptic plasticity required for information storage, while insoluble Aβ is a major component of plaque but may be relatively inert.

[0003] The primary cause of the senile plaques is the amyloid-β peptide (Aβ), which is produced by proteolytic processing of amyloid precursor protein (APP). APP is a ubiquitously expressed integral membrane protein which is proteolytically processed by secretases in various pathways. The highly amyloidogenic Aβ is released from its precursor, the β-amyloid precursor protein (APP), by two sequential proteolytic cleavages mediated by β- and γ-secretase. β-Secretase (β- site APP cleaving enzyme, BACEl) removes the bulk of the ectodomain of APP and leaves behind a small membrane retained C-terminal stub. While β- secretase cleavage is mediated by a rather conventional aspartyl protease, the second cut is mediated by an unusual protease, γ-secretase.

[0004] Presenilins (PS) have been shown to form the catalytic subunit of the γ-secretase complex that produces the Aβ peptide. Most mutations in APP and PS increase the ratio of a 42-residue form of Aβ (Aβ42) versus 40-residue Aβ (Aβ40), thus defining a common AD phenotype caused by APP, PSl and PS2 mutations. Aβ peptides ending at residue 42 or 43 (long tailed Aβ) are thought to be more fibrillogenic and more neurotoxic than Aβ ending at residue 40, which is the predominant isoform produced during normal metabolism of βAPP. The Aβ42 peptide is thought to initiate the amyloid cascade, a pathological series of neurotoxic events, which eventually leads to neurodegeneration in Alzheimer's Disease.

[0005] Presenilins are known to be involved in the regulation of β-catenin stability, trafficking of membrane proteins, and γ-secretase cleavage of APP and other substrates. All PSl mutations associated with AD increase γ-secretase cleavage of βAPP and preferentially increase the production of long-tailed Aβ peptides ending at residue 42.

[0006] Most cells express both PSl -comprised γ-secretase and PS2-comprised γ-secretase, with PSl -comprised γ-secretase being primarily responsible for Aβ production and probably also Notch signaling. Thus, a major challenge in developing therapeutics for treating AD has been to identify inhibitors of γ-secretase that reduce the production of amyloid peptides from APP without significantly affecting the cleavage of other γ-secretase substrates such as Notch.

SUMMARY OF THE INVENTION

[0007] The present invention is based on the discovery that the novel immediate early gene, termed Arc, directly interacts with presenilin 1 (PSl) and induces PSl to accumulate in endosomes, where it co-localizes with Arc, other components of the gamma secretase complex, and APP. A peptide sequence in Arc has been identified that is necessary and suffcient to mediate the interaction between Arc and PSl. In accordance with these discoveries, there are provided amino acid fragments of an Arc polypeptide, wherein the fragment binds presenilin 1. In certain embodiments, the fragment induces PSl to accumulate in endosomes.

[0008] In another embodiment of the invention, there are provided methods of identifying a compound that inhibits binding of an Arc polypeptide to presenilin 1. The method includes contacting a test compound with a sample containing an Arc polypeptide or fragment thereof that binds PSl and a presenilin 1 polypeptide; and comparing Arc polypeptide and presenilin 1 (PSl) polypeptide binding in the presence and absence of the compound. When the binding in the presence of the compound is lower than in the absence of the compound, then the compound is an inhibitor of Arc polypeptide and a presenilin 1 (PSl) polypeptide binding. [0009] In another embodiment of the invention, there are provided methods of inhibiting gamma-secretase activity in a cell. The method includes contacting PSl with a compound that inhibits binding of Arc to PSl, thereby reducing or eliminating accumulation of PSl in endosomes and incorporation of PSl in gamma-secretase, thereby inhibiting gamma- secretase activity. In one aspect, the formation of amyloid beta is reduced.

[0010] In still another method of the invention, there are provided methods of inhibiting amyloid beta plaque formation in a subject, wherein the method includes inhibiting the interaction of Arc and presenilin 1 (PSl).

[0011] In yet another embodiment of the invention, there are provided methods of inhibiting the formation of amyloid beta plaques in, on or around neurological tissue. The method includes administering to a subject in need of such treatment an effective amount of a compound that blocks the interaction between Arc and PSl, thereby inhibiting formation of amyloid beta plaques in the subject. In certain embodiments, the subject has or is at risk of having a neurodegenerative disease. In particular embodiments, the subject has a mutation in the PSl gene.

BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1A shows a schematic structure of the Arc protein. Figure 1B shows the amino acid sequence of an exemplary human Arc (GenBank Accession No. AAF07185) and Figure 1C shows the corresponding nucleotide sequence (GenBank Accession No. AF193421).

[0013] Figure 2 A shows the amino acid sequence of an exemplary human presenilin 1 (GenBank Accession No. NP_000012) and Figure 2B shows the corresponding nucleotide sequence (GenBank Accession No. NM_000021).

[0014] Figure 3 shows plots Aβ secretion in primary cortical cultures from APPswe/PSlΔE9 transgenic mice in WT and Arc/Arg3.1 KO backgrounds (APP/WT and APP/ArcKO). Figures 3A and B show Aβ40 (A) and Aβ42 (B) secreted in conditioned medium. Figure 3C shows γ-secretase activity.

[0015] Figures 4A and B show plots depicting Aβ generation in vivo in12-month-old APP/ Arc KO as measured by ELISA (A) and filter trap (B) assays. Figure 4C shows a plot of an ELISA determination of PBS-soluble Aβ40/42 and formic acid-soluble Aβ40/42 in 6- month-old APP/WT and APP/Arc KO mice.

DETAILED DESCRIPTION OF THE INVENTION

[0016] As provided herein, the present invention is based on the discovery that the immediate early gene, termed Arc, directly interacts with PSl (and endocytosis proteins) and induces PSl to accumulate in endosomes where it co-localizes with Arc, other components of the gamma secretase complex, and APP. Thus, Arc functions to regulate the incorporation of PSl into endosomes. Agents that target this regulatory event offer a new and selective strategy for development of therapeutic compounds that modulate gamma secretase activity, because the gamma secretase activity is dependent on inclusion in endosomes. As reported herein, a peptide sequence in Arc that is necessary and suffcient to mediate the interaction has been identified. Accordingly, because the critical interactions between Arc and PSl are understood in molecular detail, high throughput assays for small molecule agents can be exploited for drug discovery. Thus, tools are provided for high throughput screens that can identify agents that selectively regulate the interaction. These discoveries promise a novel approach to develop agents to regulate PSl and treat neurological disorders such as Alzheimer's Disease and age-dependent memory decline.

[0017] Arc is an immediate early gene (IEG) that is dynamically regulated in brain neurons. The closest homology identified to date for Arc remains with α-spectrin. The region of homology spans 156 amino acids in the carboxyl terminal half of Arc (amino acids 155-316) where Arc is 20% identical to the 21st and 22nd repeats of α-spectrin. While this level of identity is generally too low to be predictive of function, it is noted that the degree of identity between "repeats" in α-spectrin is typically only 20%. Each α-spectrin repeat is believed to form a tri-α-helix bundle that is stabilized by hydrophobic, leucine-zipper like interactions with neighboring coils. Amino acids that are conserved between spectrin repeats correspond to sites of protein-protein interaction between coils and are generally conserved in Arc. Additionally, because many of the amino acid differences between α-spectrin and Arc represent conservative changes, the overall level of homology in this region is 77%. Accordingly, the deduced amino acid sequence of Arc suggests that the C-terminus possesses a spectrin-like domain that may be important in forming intermolecular interactions. [0018] An exemplary amino acid sequence for a human Arc protein is set forth in GenBank Accession No. AAF07185 (SEQ ID NO:1). The corresponding nucleotide sequence is set forth in GenBank Accession No. AF 193421 (SEQ ID NO: 2). An example of an Arc protein, termed Arg3.1, from rat has also been reported (GenBank Accession No. Z46925; Link et al, PNAS 92(12):5734-8, 1995).

[0019] Presenilin 1 (PSl) is a transmembrane protein that is endoproteolytically cleaved into C and N-terminal fragments and has been implicated in Alzheimer's disease (AD). The C-terminal fragment of PSl binds to endophilin 3 and is recruited to Arc-endophilin vesicles. Presenilins (PSl or PS2) have been shown to form the catalytic subunit of the γ-secretase complex that produces the Aβ peptide. Aβ peptides ending at residue 42 or 43 (long tailed Aβ) are thought to be more fibrillogenic and more neurotoxic than Aβ ending at residue 40, which is the predominant isoform produced during normal metabolism of βAPP. All PS 1 mutations associated with AD increase γ-secretase cleavage of βAPP and preferentially increase the production of long-tailed Aβ peptides ending at residue 42.

[0020] An exemplary amino acid sequence for a human presenilin 1 protein is set forth in GenBank Accession No. NP_000012 (SEQ ID NO:3). The corresponding nucleotide sequence is set forth in GenBank Accession No. NM 000021 (SEQ ID NO:4).

[0021] γ-Secretase is a multiprotein complex consisting of presenilin, nicastrin, Aph-1, and Pen-2. γ-Secretase cleaves the Amyloid Precursor Protein (APP) in its transmembrane domain, releasing the amyloid peptide Aβ. Aβ is the main constituent of the amyloid plaques in the brains of patients suffering from Alzheimer's disease. Several other type I integral membrane proteins are also cleaved by this protease. All four proteins are necessary for full proteolytic activity.

[0022] According to one embodiment of the invention, there are provided amino acid fragments of the Arc polypeptide as set forth in SEQ ID NO:1, wherein the fragment binds presenilin 1. In certain embodiments, the fragment induces PSl to accumulate in endosomes. In some embodiments, the fragment is about 2 to 20 amino acids in length. In certain embodiments, the fragment is 5 to 20 amino acids in length, or 5 to 15 amino acids, or 8 to 12 amino acids, or even 5 to 10 amino acids in length. In particular embodiments, the fragment comprises the sequence IKACLCRCQE (SEQ ID NO:5). In other embodiments, the fragment is the amino acid sequence as set forth in SEQ ID NO:5. In still other embodiments the amino acid fragment may be a peptidomimetic of the amino acid sequence set forth in SEQ ID NO:5.

[0023] In some embodiments, the amino acid fragment of Arc is modified. Such modifications include insertion, deletion, or substitution of one or more amino acids within the amino acid fragment. In certain embodiments, one amino acid is substituted with another amino acid, or two amino acids are substituted, or three amino acids are substituted with another amino acid. Amino acid substitutions may be conservative or non-conservative. Conservative amino acid substitutions generally refer to an exchange of a member of one class of amino acid with another member of the same class. There are four general classes of amino acids: nonpolar: Met, Gly, Pro, Ala, VaI, Leu, Cys, Ile, Trp, Phe; uncharged polar: Tyr, Ser, Thr, Asn, Gln; acidic: Asp, Glu; and basic, His, Lys, Arg. Non-conservative substitutions generally refer to an exchange of a member of one of these classes for a member from another class. Amino acid substitutions may encompass non-naturally occurring amino acid residues, which are typically incorporated by chemical peptide synthesis rather than by synthesis in biological systems. These include reversed or inverted forms of amino acid moieties.

[0024] In other embodiments, the peptide is modified by conjugation to a non-protein polymer. In some embodiments, the non-protein polymer is a polyethylene glycol, a polypropylene glycol or a polyoxyalkylene. In particular embodiments, the non-protein polymer is polyethylene glycol (PEG). In certain embodiments, the amino acid fragment may include one or more amino acid substitutions wherein the substitutions facilitate the site- specific coupling of at least one non-protein polymer, such as polypropylene glycol, polyoxyalkylene, or polyethylene glycol (PEiG) molecule to the fragment. Site-specific coupling of PEG, for example, allows the generation of a modified fragment which possesses the benefits of a polyethylene-glycosylated (PEGylated) molecule, namely increased plasma half life and decreased immunogenicity while maintaining greater potency over non-specific PEGylation strategies such as N-terminal and lysine side-chain PEGylation.

[0025] Methods of conjugating PEG (termed PEGylation) are well-known in the art. In general, the first step of the PEGylation is the suitable functionalization of the PEG polymer at one or both terminals. The choice of the suitable functional group for the PEG derivative is based on the type of available reactive group on the molecule that will be coupled to the PEG. For proteins, typical reactive amino acids include lysine, cysteine, histidine, arginine, aspartic acid, glutamic acid, serine, threonine, tyrosine. The N-terminal amino group and the C-terminal carboxylic acid can also be used. The techniques used to form first generation PEG derivatives are generally reacting the PEG polymer with a group that is reactive with hydroxyl groups, typically anhydrides, acid chlorides, chloroformates and carbonates. In the second generation PEGylation chemistry more efficient functional groups such as aldehyde, esters, amides etc made available for conjugation.

[0026] It should be understood that the structure of the attached PEG moieties is important in optimizing the PEGylation of the amino acid fragments of the invention. In one embodiment, the PEG moiety is linear. Linear PEG moieties are limited in size by the manufacturing process because the amount of PEG-diol increases as PEG molecular weight increases. In another embodiment, the PEG moiety is branched from a single attachment site. Branched PEG moieties have the advantage of increasing the size of the PEG molecule without increasing the number of site attachments.

[0027] In another embodiment of the invention, there are provided methods of identifying a compound that inhibits binding of an Arc polypeptide to presenilin 1. The method includes contacting a test compound with a sample containing an Arc polypeptide that binds PSl and a presenilin 1 polypeptide; and comparing the level of Arc polypeptide and the presenilin 1 (PSl) polypeptide binding in the presence and absence of the compound. When the binding in the presence of the compound is lower than in the absence of the compound, then the compound is an inhibitor of Arc polypeptide and a presenilin 1 (PSl) polypeptide binding.

[0028] The term "test compound" is used herein to mean any agent that is being examined for ability to inhibit binding of an Arc polypeptide to presenilin 1 in a method of the invention. Although the method generally is used as a screening assay to identify previously unknown molecules that can act as a therapeutic agent, a method of the invention also can be used to confirm that an agent known to have such activity, in fact has the activity, for example, in standardizing the activity of the therapeutic agent.

[0029] A candidate agent can be any type of molecule, including, for example, a peptide, a peptidomimetic, a polynucleotide, or a small organic molecule, that one wishes to examine for the ability to act as a therapeutic agent, which is an agent that provides a therapeutic advantage to a subject receiving it. It will be recognized that a method of the invention is readily adaptable to a high throughput format and, therefore, the method is convenient for screening a plurality of test compounds either serially or in parallel. The plurality of test compounds can be, for example, a library of test agents produced by a combinatorial method library of test agents. Methods for preparing a combinatorial library of molecules that can be tested for therapeutic activity are well known in the art and include, for example, methods of making a phage display library of peptides, which can be constrained peptides (see, for example, U.S. Pat. Nos. 5,622,699; 5,206,347; Scott and Smith, Science 249:386-390, 1992; Markland et al., Gene 109:1319, 1991; each of which is incorporated herein by reference); a peptide library (U.S. Pat. No. 5,264,563, which is incorporated herein by reference); a peptidomimetic library (Blondelle et al., Trends Anal. Chem. 14:8392, 1995; a nucleic acid library (O'Connell et al., supra, 1996; Tuerk and Gold, supra, 1990; Gold et al., slpra, 1995; each of which is incorporated herein by reference); an oligosaccharide library (York et al., Carb. Res., 285:99128, 1996; Liang et al., Science, 274: 1520-1522, 1996; Ding et al., Adv. Expt. Med. Biol., 376:261-269, 1995; each of which is incorporated herein by reference); a lipoprotein library (de Kruif et al., FEBS Lett., 399:232-236, 1996, which is incorporated herein by reference); a glycoprotein or glycolipid library (Karaoglu et al., J. Cell Biol., 130:567-577, 1995, which is incorporated herein by reference); or a chemical library containing, for example, drugs or other pharmaceutical agents (Gordon et al., J. Med. Chem., 37:1385-1401, 1994; Ecker and Crooke, Bio/Technology, 13:351-360, 1995; each of which is incorporated herein by reference). Accordingly, the present invention also provides a therapeutic agent identified by such a method, for example, a therapeutic agent useful in the treatment of a neurological disease or disorder.

[0030] In another embodiment of the invention, there are provided methods of inhibiting gamma-secretase activity in a cell. The method includes contacting PSl with a compound that inhibits binding of Arc to PSl, thereby reducing or eliminating accumulation of PSl in endosomes and incorporation of PSl in gamma-secretase, thereby inhibiting gamma- secretase activity. In certain embodiments, the formation of amyloid beta is reduced. In particular embodiments the formation of amyloid beta 42 is reduced. In other embodiments, the formation of amyloid beta 43 is reduced. [0031] In still another method of the invention, there are provided methods of inhibiting amyloid beta plaque formation in a mammal, wherein the method includes inhibiting the interaction of Arc and presenilin 1 (PSl).

[0032] In yet another embodiment of the invention, there are provided methods of inhibiting the formation of amyloid beta plaques in, on or around neurological tissue. The method includes administering to a subject in need of such treatment an effective amount of a compound that blocks the interaction between Arc and PSl, thereby inhibiting formation of amyloid beta plaques in the subject. In certain embodiments, the subject has or is at risk of having a neurodegenerative disease or disorder. In some embodiments, the neurodegenerative disease or disorder may be Alzheimer's Disease or age-dependent memory decline. In particular embodiments, the subject has a mutation in the PSl gene.

[0033] The route of administration of the compound will depend, in part, on the chemical structure of the compound and the target tissue. For example, the delivery of a compound to central nervous system (CNS) can be accomplished by administering the compound directly into the CNS or administering it systemically (e.g., by intravenous injection). Intravenous, intranasal, intracerebroventricular, intrathecal, intracranial intrapulmonary, or oral administration are commonly used to deliver compounds to the CNS.

[0034] The total amount of a compound to be administered in practicing a method of the invention can be administered to a subject as a single dose, either as a bolus or by infusion over a relatively short period of time, or can be administered using a fractionated treatment protocol, in which multiple doses are administered over a prolonged period of time. The compound can be formulated for oral formulation, such as a tablet, or a solution or suspension form; or can comprise an admixture with an organic or inorganic carrier or excipient suitable for enteral or parenteral applications, and can be compounded, for example, with the usual non-toxic, pharmaceutically acceptable carriers for tablets, pellets, capsules, suppositories, solutions, emulsions, suspensions, or other form suitable for use. The carriers, in addition to those disclosed above, can include glucose, lactose, mannose, gum acacia, gelatin, mannitol, starch paste, magnesium trisilicate, talc, corn starch, keratin, colloidal silica, potato starch, urea, medium chain length triglycerides, dextrans, and other carriers suitable for use in manufacturing preparations, in solid, semisolid, or liquid form. In addition auxiliary, stabilizing, thickening or coloring agents and perfumes can be used, for example a stabilizing dry agent such as triulose (see, for example, U.S. Pat. No. 5,314,695).

[0035] Biochemical data that suggest Arc functions as a multidomain scaffold. The current understanding of the domain structure of Arc is shown in Figure 1 (amino acid numbering below), with the names of proteins that interact with certain domains provided (above). A computer based structural analysis of the NH2 terminal 100 aa of Arc predicts coiled-coils (COILS software; available on the world wide web at ch.embnet.org/software/COILS_form). In support of this prediction, Genbank analysis indicates homology of the same region with the coiled-coil domain of troponin T. This region of Arc is of interest since current data indicates that it interacts with CaMKII and SH3 domains, as discussed below.

[0036] Arc protein interacts with calcium and calmodulin-dependent protein kinase type II (CaMKII), and this interaction has been confirmed by co-immunoprecipitation from brain. Arc also interacts with certain SH3 domain proteins including PSD95, SAP97 and endophilin 3. Regions of Arc essential for interaction with CaMKII and SH3 domain proteins have been mapped and are contiguous, yet distinct. Arc protein is palmitoylated, and this lipid modification may target Arc to membranes. Recent studies suggest a role for these biochemical properties in synaptic physiology. First, studies indicate that Arc expression results in selective down-regulation of synaptic AMPA receptor responses in hippocampal CAl neurons. Arc's effect is similar to classically defined long-term depression (LTD) in that it is blocked by calcineurin inhibitors. Second, Arc induces the formation of a novel endosome when co-expressed with endophilin 3. Arc-dependent endosome formation appears to be a consequence of the scaffold property of Arc as it requires the SH3 interaction domain, and Arc-endosomes are decorated by CaMKII. Interestingly, Arc-dependent LTD is also dependent on the SH3 -interaction domain of Arc. Finally, Arc-endophilin endosomes accumulate GluR2 when these proteins are co-expressed in neurons. These findings suggest that Arc plays a role in AMPAR trafficking, and this action may be mechanistically linked to Arc-induced endosomes.

[0037] Arc is an IEG that is enriched in brain and is rapidly regulated by neuronal activity. Like other IEGs, Arc mRNA is strongly induced in granule cell neurons of the hippocampus by maximal electroconvulsive seizure (MECS) and by non-epileptic, NMDA receptor- dependent synaptic stimuli in association with LTP. Among IEGs, Arc is unique in that the mRNA rapidly trafficks to dendrites. Arc mRNA trafficking is rapid and not dependent protein translation. The relationship between Arc mRNA localization and synaptic activity has been examined in vivo in the rat hippocampus. Arc expression was induced by activating the medial perforant pathway from the entorhinal cortex to the dentate gyrus. Synapses of the medial perforant path terminate in a precisely defined lamina in molecular layer and Arc mRNA was found to be targeted to these synapses. Whether mRNA targeting conferred co- localized protein expression at active synaptic regions was examined. Immunostaining of tissue sections from stimulated animals using an Arc-specific antibody revealed that the recently synthesized mRNA and protein were co-localized. There was a strong upregulation of Arc protein expression in granule cell bodies, and a discrete band of immunostaining in the molecular layer corresponding to the zone of termination of medial perforant path fibers. This selective distribution is quite different from the uniform distribution of immunostaining seen following MECS.

[0038] In another experiment, signals involved in induction of Arc mRNA were dissociated from those that target the mRNA. MECS was used to induce Arc mRNA and then the effect of medial perforant pathway stimulation was examined. The experimental preparation is particularly powerful since within animal comparisons of the effect of synaptic stimulation can be performed. Arc mRNA in the hippocampus that did not receive further stimulation was uniformly distributed through the molecular layer. By contrast, Arc mRNA showed focal accumulation in the medial molecular layer. Interestingly, the intensity of the in situ signal was identical to the average intensity of the contralateral side. This observation suggests that mRNA outside of the region of stimulation may be more rapidly degraded.

[0039] Arc mRNA is targeted to dendritic regions by NMDA receptor activation. MECS was administered to a rat, and in vivo LTP stimulation was administered via perforant pathway stimulation to one hemisphere. In this experiment the recording electrode included pharmacological antagonists. The most compelling observation was that inclusion of NMDA receptor antagonists resulted in Arc mRNA remaining uniformly distributed throughout the molecular layer. Regions remote from the recording electrode, but in the same plane of tissues showed the expected concentration of Arc mRNA in the medial molecular layer. This result suggests that NMDA receptor activation is required for focal enrichment of Arc mRNA. By contrast, AMPA receptor antagonists and mGluR antagonists do not appear to be involved in the process. These studies provide an initial sense for the complexity of cellular mechanisms that regulate Arc mRNA and protein targeting.

[0040] Arc mRNA expression identifies neural networks involved in learning. During the studies to examine the regulated expression of Arc mRNA, it was noted that following induction, mRNA first appears in the nucleus and later appears in the cytoplasm. The signal in the nucleus of CAl neurons appears within 3 min of placing rats in a novel environment and persists for only 20 min. By contrast, the signal in the cytoplasm begins after about 20 min and persists for ~45 min. The time-dependent appearance of the mRNA provided a means to assess when a neuron had been activated, and more importantly, it provided a means to assess its activity-history during 2 epochs. The essential technical innovation is that mRNA signal in the nucleus, or cytoplasm, can be distinguished using confocal microscopy. Thus, it is possible to identify patterns of neuronal activation in response to two, temporally discrete stimuli in the same animal. This method has been termed catFISH, for cellular analysis of temporal activity by fluorescent in situ hybridization.

[0041] catFISH was used to visualize a stable network of neurons that is active when an animal explores a novel environment. A ra1 was placed in environment "A" for 5 min., returned to its home cage for 20 min., then returned to "A" for 5 min and sacrificed. Consistent with electrophysiological recordings of CAl, it was observed that -40% of neurons showed FISH signal in the cytoplasm (first experience in "A") and -40% in the nucleus (second experience in "A"). Importantly, the signals were in the same neurons (95% overlap). By contrast, if the rat first experienced "A" and then a distinct environment, "B", again -40% of neurons showed FISH signal in nucleus and 40% in the cytoplasm, but only 16% showed signal in both cytoplasm and nucleus. This degree of overlap is consistent with electrophysiological recordings of place cells, which appear to be randomly "selected" for a particular environment. Moreover, the absolute percentage of neurons that show FISH signal is identical to that estimated by recordings for place cells. Thus, -40% of neurons are randomly selected for each environment A and B, with expected 16% overlap (0.4x0.4 = 0.16).

[0042] Induction of Arc mRNA is currently thought to be the cellular response that is most tightly linked to natural activity involved with information processing. This uniquely strong association is based on the demonstration that Arc is activated, and reactivated, in an identical population of neurons, if and only if, the environments are identical. From this, it can be inferred that Arc catFISH is detecting stable neural networks. Thus, catFISH is being used to study cellular responses during natural learning and memory. catFISH may therefore help understand learning deficits due to aging or gene mutation.

[0043] Arc is a synaptic protein that is enriched in the PSD. Arc transgene expressed in neurons localizes in dendrites and spines. Arc immunoreactivity in adult rat hippocampus is concentrated in dispersed granule cell and pyramidal neurons. The distributed cellular staining is consistent with its demonstrated transcriptional induction in discrete sets of neurons in awake behaving animals. Within these cells, Arc protein is enriched in cell bodies and dendrites. LM images illustrate Arc protein in fine, distal branches of the dendritic arbor. Additionally, fine punctae are resolved that have the characteristics of spines. Arc immunoreactivity is also present in the nucleus of granule cells and pyramidal neurons of the neocortex. Confocal images revealed its presence within the nucleus in a uniform pattern that was restricted from certain subregion (perhaps nucleolus). Nuclear localization of Arc appears to be specific since it is restricted to neurons that also express Arc in dendrites and is rapidly induced by neuronal activity. EM confirms the presence of Arc in spines, in which gold particles are present throughout the spine but are particularly enriched over the PSD where they appear at, or very near, the plasma membrane. Consistent with LM observations, no staining is detected in presynaptic elements or glial cells. Western blot analysis of biochemical fractionations of brain provides confirmation that Arc is a component of the PSD. Arc protein is enriched in the P2 fraction and is remains enriched in this fraction following Triton and Sarcosyl washes. These observations suggest that Arc is a "core PSD" protein and are consistent with a report of Arc in postsynaptic density (PSD) fractions.

[0044] Arc is palmitoylated and enriched in lipid raft preparations from brain. The EM localization of Arc very close to the plasma membrane was unexpected since Arc had been previously found to be enriched in the cell cortex, and biochemical studies indicated that Arc associates with F-actin preparations. While the actin network is enriched in the spine, it is not thought to extend to the plasma membrane of the PSD. Accordingly, the possibility that Arc might be targeted to the plasma membrane by addition of a lipid, as has been demonstrated for PSD-95 and GluR6 was examined. [0045] In one study, Arc was transiently expressed in HEK293 cells and 4 hrs prior to harvesting, the media was changed to DMEM with fatty acid free BSA containing 4-7 mCi of palmitate. Lysates were then immunoprecipitated, resolved by SDS-PAGE and assayed for incorporation of [³H] palmitate by autoradiography. A band of the predicted size indicated incorporation of [³H] into Arc. As a positive control, GluR6 showed prominent incorporation of [³H], while negative controls including Homer or empty vector did not. Palmitate is linked to proteins by thioester. Arc encodes 5 cysteines at positions 34, 94, 96, 98 and 159. To confirm that [³H] was added to Arc as a thioester, mutants that deleted cysteines or mutated cysteines to serine were tested. An Arc deletion mutant that lacks the N-terminal 40 aa (deletes the first cysteine) showed enhanced incorporation of [³H]. By contrast, the same deletion mutant in which cysteines 94, 96 and 98 were mutated to serine did not incorporate [³H]. These mutations in the full length protein markedly reduced palmitoylation. It is notable that cysteines 94, 96 and 98 are in a hydrophobic region that is similar to the site of palmitoylation in PSD-95 and other palmitoylated proteins.

[0046] Addition of palmitate has been shown to target proteins to specific lipid domains in the plasma membrane, termed lipid rafts, and thereby recruit proteins for signaling functions. Accordingly, the prediction that Arc may be enriched in lipid raft fractions was examined. Lipid rafts are characterized biochemically by insolubility in 1% Triton and relative bouncy in sucrose gradient centrifugation. Consistent with other reports, PSD-95, and Caveolin showed enrichment in the lipid raft fraction, while the transferrin receptor was enriched in more dense fractions. Arc showed prominent enrichment in the lipid raft fraction.

[0047] Arc interacts with the C-terminus of Endophilin 3. A yeast 2-hybrid screen of an adult rat forebrain cDNA library identified the C-terminus of a protein termed endophilin 3. There are three endophilin genes in the mammalian genome that share the same domain structure of an N-terminal coiled coil and a C-terminal SH3 domain. Endophilin 1 is known to play a role in endocytosis of clathrin coated vesicles and is predominantly present in presynaptic elements. Endophilin 3 is brain specific and is present both pre- and postsynaptically. The putative Arc-endophilin interaction was examined by first demonstrating that Arc Ab co-immunoprecipitates endophilin from a rat brain P2 fraction. An endophilin Ab was generated against a shared region of the protein family, but is relatively specific for endophilin 3. To further evaluate the interaction, a myc-tagged endophilin 3 C-terminus (endo3CT) was expressed in HEK293 cells and it was confirmed that it co-immunoprecipitates with wt Arc (CFP-Arc). By contrast, endo3CT did not co-IP with ArcΔ91-100 (CFP- ArcΔ91-100). The latter Arc deletion mutant was generated in studies of the interaction of Arc with the SH3-GK domain of PSD95. Its inability to interact with endo3CT suggests that Arc may interact with the SH3 domain of endo3CT.

[0048] Arc facilitates formation of endosomes with Endophilin 3. Interactions of Arc with endophilin 3 suggested that Arc may modulate endophilin function. Arc produced a marked redistribution of endo3CT. When individually expressed in HeLa cells, Arc protein localized to the cytosol with some enrichment at the plasma membrane. Endo3CT was also predominantly in the cytosol. However, when co-expressed, these proteins co-localized at the plasma membrane and in vesicle-like structures that appeared to be enriched near the cell membrane. The size of the vesicle increased with time after transfection and ranged from submicron vesicles close to the membrane to large, inclusion-like vesicles. The effect of Arc was dependent on its SH3 -interacting domain as it was not evident with ArcΔ91-100. Endogenous Transferrin receptor is recruited to these vesicles, which are also positive for the native endosomal marker EEAl. These observations support the hypothesis that the Arc- endophilin vesicle is an endosome.

[0049] Visualization of Arc-endophilin vesicle dynamics by total internal reflection fluorescence (TIRF) microscopy. The association of Arc-endo3CT vesicles with the plasma membrane was analyzed by TIRF. N-terminal GFP tagged Arc was transfected into HeLa cells and images of living cells were taken with widefield (Arc wf) and TIRF optics (Arc tirf). Arc in widefield (wf) showed uniform reticular pattern with additional enrichment in the nucleus. TIRF images of the same cells showed a weak, linear pattern but nuclear staining was absent. Absence of nuclear staining is consistent with TIRF optics, which detect proteins close to the membrane surface. As an additional negative control, GFP was strongly positive in wf (GFP wf) but absent in TIRF (GFP tirf). By contrast a GFP-tagged vinculin, which accumulates at focal adhesions showed reticular cytosolic distribution in wf (Vin wf) and bright presumptive focal adhesions in TlRF (Vin tirf). GFP- Arc was co-transfected with endo3CT and vesicles are evident in both wf and TIRF. Time lapse TIRF images revealed the dynamic nature of the vesicles at the membrane. Vesicles appeared to approach and recede from the membrane and two vesicles appeared to fuse. [0050] Arc-endosome recruits CaMKII. Arc also interacts with α and β isoforms of Ca2+/Calmodulin dependent protein kinase II (CaMKII). This interaction was also first detected in a yeast 2-hybrid screen. In vitro biochemical assays confirmed that Arc binds both α and β isoforms of CaMKII. Additionally, Arc and CaMKII co-immunoprecipitate from transfected HEK293 cells and from brain. Deletion analysis indicates that aa 50-80 of Arc are required for interaction with CaMKII and that the N-terminal 100 aa of Arc is sufficient to bind CaMKII. CaMKII requires a 20 aa sequence within the C-terminal association domain to interact with Arc. This region of CaMKII has been implicated in self- association of the multimeric holoenzyme. When βCaMKII is co-transfected with Arc and endo3CT, it is recruited to the vesicles. βCaMKII itself or cotransfection with endo3CT did not produce any vesicles. These data suggest that CaMKII may be recruited to the vesicle by Arc. Alternatively, Arc induces the endosome and another protein on the endosome recruits CaMKII.

[0051] Arc specifically interacts with Endophilin 3 CT, and not with Endophilin 1 CT. To assess the specificity of the interaction between Arc and endophilin family members, an equivalent deletion mutant of endophilin 1 (endolCT) was generated. Enophilin constructs were then transfected into HEK293 cells either alone or together with CFP-tagged Arc. Detergent lysates (l%Triton in PBS) were then precipitated with Arc Ab. Westerns were performed of lysates and precipitaes and probed with the indicated Ab. In contrast to endo3CT, which co-immunoprecipitates with Arc, endolCT did not. The same transfected cells were analyzed by immunohistochemistry. EndolCT did not produce vesicles when co- expressed with Arc. These observations indicate that the interaction of Arc and endophilin 3 is specific.

[0052] Arc expression causes a selective down regulation of AMPAR in CAl neurons. Organotypic hippocampal slices were prepared from P8 rats and cultured for 5 to 8 days. Slices were infected for 48 hrs with Sindbis virus GFP-Arc expression construct. Side-by- side control and infected cells are recorded simultaneously using paired patch clamp recordings in voltage clamp mode at holding potentials of -60 and +40mV, in order to differentiate AMPA and NMDA responses. Arc overexpression depresses AMPA transmission. [0053] Arc effects are blocked by inhibitors of Calcineurin. Control experiments were sought that would assess the specificity of Arc effects on AMPAR. Arc over-expression mimics LTD in that it produces a selective reduction of AMPAR responses. Accordingly, the hypothesis that effects of Arc could be blocked by agents that are known to block LTD was examined. Inhibitors of the phosphatase calcineurin block LTD, and also blocked Arc- induced depression.

[0054] Deletion of Arc domain required for Endophilin and SH3 interaction blocks its inhibitory activity on AMPAR responses. To provide further insight into the molecular mechanisms of Arc action, the effect of the Arc Δ91-100 mutant that does not interact with endophilin or the SH3-GK domain of PSD95/SAP97 was tested. The deletion mutant did not elicit AMPAR depression.

[0055] Arc induces co-clustering of Endophilin and AMPA receptors in vesicles. The preceding studies indicate that Arc induces down-regulation of AMPAR responses. This provides important insight into its possible function. However, direct interaction of Arc and AMPAR has not been detected, so the mechanism of action likely involves other proteins. The region of Arc required for AMPAR down regualtion is also required for formation of endosomes with endophilin 3. Since down regulation of AMPAR in LTD is mediated by clathrin-dependent endocytosis of the receptors, the possibility that Arc might interact with AMPAR indirectly by inclusion in the Arc-endophilin endosome was examined. Primary neuronal cultures were transfected with Arc, endo3CT and GluR2, either individually or together. When Arc is expressed alone, it accumulates in the dendrites and in the nucleus. Endo3CT is uniformly distributed in dendrites. When co-expressed, Arc and endo3CT co- localize in spherical structures that appear along the length of the dendrites. When GluR2 is co-expressed, it co-localizes with Arc and endo3CT.

[0056] Arc-Endophilin co-clusters the C-terminus of Presenilin 1. Presenilin 1 is cleaved into N- and C-terminal fragments and correct trafficking of the C-terminus is dependent on the C-terminal 4 amino acids. In a Y2H screen we identified a single clone (endophilin 3) that bound to the PSl C-terminus and required the last 4 amino acids. It was confirmed that the endophilin fragment co-IPs with native PSl when expressed in HEK293 cells. This finding was not evaluated further until recently as we discovered the. Based on the work identifying an Arc-endophilin interaction, the possibility that PSl might be recruited to Arc- endosomes was examined. PS-I was co-expressed in HeLa cells with Arc and endo3CT and cells stained for endo3CT and PSl. PSl was found to co-localize with endo3CT in these cells.

[0057] Arc knock-out mice show impaired LTP and long term memory. Homozygous Arc knockout mice are viable and appear healthy. This contrasts with a prior knockout that deleted part of the gene and reported non- viable embryos at Ell. The difference in phenotype is possibly due to generation of a dominant negative fragment of Arc that is lethal. Alternatively, differences in background could be important. The latter mouse shows expected Mendelian numbers and reproduces as homozygous knockout animal. In vivo extracellular recordings from the dentate gyras show strongly increased LTP. However, by 1-2 hrs there is a complete decay of LTP to baseline. LTP in littermate wild type mice was stable at the same time points. LTP was also examined in acute hippocampal slices. CAl LTP assayed by intracellular recordings in patch clamp by depolarization paired with presynaptic stimulation resulted in robust LTP for <30 min, then complete decay to baseline. Behavior analysis shows impaired fear conditioning (cued and context) as measured at 24 hours, whereas short term memory is not altered. Conditioned taste aversion is also markedly impaired in Arc knockout mice. In the water maze, Arc knockout mice show no deficits in acquisition but impairments are obvious at later trials. An analysis of the search strategy of the Arc knockouts revealed that they use all strategies except focal and spatial strategies. In summary, the Arc knockout mice show impaired LTP and long-term memory that affects multiple synapses and circuits.

[0058] Based on these studies, a model is proposed in which Arc functions as a multidomain scaffold. Arc mRNA is selectively expressed in neurons as part of information processing and storage, and Arc protein ma> be locally synthesized at active synapses. The interaction of Arc protein with certain SH3 domain proteins induces formation of early endosomes. It is hypothesized that recruitment of endophilin 3 is important in vesicle formation and perhaps in defining the composition of the vesicle. Arc also interacts with CaMKII and recruits the kinase to the endosome. Importantly, the Arc-endosome appears capable of trafficking AMPAR in neurons, and offers a cellular mechanism that is consistent with all available data for Arc function. The Arc-endosome hypothesis is also consistent with mechanisms known to be important in synaptic plasticity. For example, LTD is known to involve GluR2 trafficking from the synapse by a clathrin-dependent endocytosis mechanism. In a heterologous cell system, the endosomes do not co-localize with Endophilinl, PSD95, mGluR5, NMDARs NRl or NR2B, IP3 receptor or lamp. Thus, the Arc-endosome appears to be highly specific. And while the analysis emphasizes the endocytosis aspect of vesicular trafficking, it is hypothesized that the vesicular content also returns to the membrane surface. In this regard, the fact that Arc recruits CaMKII to the vesicle could be critical for function since activated CaMKII can drive GluRl into the synaptic membrane.

[0059] Arc is generally considered to possess three interaction domains. These include CaMKII (aa -50-80), SH3 (aa 91-100) and cytoskeletal proteins (C-terminal -2/3). Additionally, Arc is palmitoylated and this involves cysteines at positions 94, 96 and/or 98. By co-expressing Arc mutants with endophilin 3 SH3, the role of each domain in the physical interaction and endosome formation may be determined. Amino acids 91-100 of Arc are required for endosome formation. However, preliminary studies indicate that the N-terminal 130 aa is not sufficient. Accordingly, a role for the cytoskeletal domain in formation of the endosome is hypothesized.

[0060] The same Y2H screen that determined that Arc protein binds β-CaMKII and the SH3 domain of endophilin 3 identified a larger set of interacting proteins that appear especially interesting in the context of Arc-induced endosome formation. Candidate Arc interacting proteins derived from a comprehensive screen of full-length Arc include syntaxin B, dynamin 2, talin, alpha spectrin, 14-3-3 (beta, theta and zeta), kalirin and golgin. Each of these have been confirmed to interact with Arc in preliminary biochemical assays in which the cDNAs from the yeast screen were expressed in mammalian cells and assayed for binding to GST- Arc. To assess the hypothesis that Arc recruits these proteins to the Arc-endosome, the proteins can be co-expressed with Arc and endophilin 3 in heterologous cells and assess their co-localization with endosomes. This assay provides information beyond simple assays of protein-protein interaction. For example, Arc interacts with the SH3 domain of PSD-95 in co-IP assays, but PSD-95 does not co-localize in Arc-endophilin endosomes in HeLa cells. The initial survey will use the cDNAs derived from the Y2H screen since it is possible that their association is controlled by intramolecular effects similar to those of endophilin3. Those proteins that co-localize with the Arc-endosome can be further characterized by assessing whether full-length proteins co-localize with the endosome. Such assays may include parallel immunohistochemistry and co-IP. By this analysis, which proteins are most prominently associated with Arc endosomes in heterologous cells may be determined and characteristics that might distinguish it as a specific vesicle may be defined.

[0061] Presenilin 1 (PSl) is a transmembrane protein that is endoproteolytically cleaved into C and N-terminal fragments and has been implicated in Alzheimer's disease (AD). As provided herein, the C-terminal fragment of PSl binds to endophilin 3 and is recruited to Arc-endophilin vesicles. To further investigate this interaction N2a (mouse neuroblastoma) cell lines that stably overexpress PSl, as well as two lines that express familial AD mutant PSl (A256E, and ΔE9) were obtained. In ongoing studies, it was confirmed in these cells that PSl is recruited to Arc-endophilin vesicles, and preliminary results show that the mutant PSl ΔE9, which is not endoproteolytically cleaved, had higher levels of accumulated protein in vesicles than wild type PSl.

[0062] Studies have demonstrated that PSl affects the trafficking of several cell surface proteins and receptors including APP, Notch, TrkB, and APLPl. Furthermore PSl localizes to clathrin-coated vesicles as shown by electron microscopy studies (Efthimiopoulos, et al., J Neurochem 71:2365-72, 1998). While not v/ishing to be bound to any particular theory, the preliminary data mentioned above suggests that Arc may initiate the formation of a novel "signaling" endosome whereby PSl can cleave its substrates in response to activity. Alternatively Arc may regulate the surface levels of PSl and other proteins involved in substrate processing. It has recently been shown that inhibition of endocytosis can lead to nearly 90% reduction in beta-amyloid (Aβ) (the peptide found in AD plaques) production, implicating endosomes as the prime site of APP (amyloid precursor protein) cleavage (Ehehalt et al., J Cell Biol 160:113-23, 2003).

[0063] The presence of the components of the PSl/γ-secretase complex including Nicastrin, BACE, APH-I and PEN-2 in Arc endosomes may be examined using antiodies against these proteins to detect native or overexpressed protein that may be recruited to the Arc-endophilin vesicles. Control experiments examining mutants of these components that do not interact with PSl may be used to determine the effect of mutations of PSl on the recruitment of these components to vesicles. Substrates of secretase cleavage may be be examined to determine if they co-localize to Arc endosomes. In one approach, known substrates may be co-expressed with Arc/endophilin in N2a cell lines that express PS. [0064] The same fragment of endophilin that interacts with Arc also interacts with the C- terminus of PSl. This interaction requires the C-terminal 4 aa of PSl and appears essential for incorporation of PSl into Arc-endophilin endosomes.

[0065] The invention will now be described in greater detail by reference to the following non-limiting examples.

EXAMPLES

[0066] Animals and Reagents. Arc/Arg3.1 KO (Arc -/-) mice were generated as described by Plath et al. (Neuron 52:437, 2006). Arc/Arg3.1 heterozygous mice (Arc +/-) were crossed with APP_SWE/PS 1ΔE9 mice to generate APP/WT (APP_SWE/PS 1ΔE9; Arc +/+) and APP/Arc KO (APP_SWE/PS 1ΔE9; Arc -/-) mice.

[0067] Expression constructs were made by PCR and the PCR products were cloned into expression vectors ECFPCl (Clontech), pCIS (Genentech), pRK5 (Genentech) with myc or HA tags. All constructs were verified by sequencing.

[0068] Arc antibodies were generated against residues 155-396 of Arc; the other antibodies were purchased from commercial sources: Aβ (6E10, Signet), APP CTF

(Chemicon), flotillin2 (BD Transduction Laboratories), Endophilin2/3 (Santa Cruz), Dynamin (Upstate), Myc (9E10, Roche), HA-HRP (3F10,Roche), γ-secretase inhibitor JC-22 was described previously (Lewis et al., Bioorg Med Chem Lett 15:373, 2005).

[0069] Neuronal Culture, assays of Aβ secretion and surface protein trafficking. Primary neuronal cultures from embryonic day 18 (E 18) pups were prepared as reported previously (Shepherd et al., Neuron 52:445, 2006). DIV14 cortical cultures from APP/WT and APP/Arc KO embryos were assayed for Aβ40/Aβ42 peptides in medium measured 48hrs after addition of 40 μM bicuculline using a quantitative sandwich ELISA kit (Biosource International) that specifically detects human Aβ40/Aβ42. DIVl 4 high density cortical neurons were surface biotinylated. For internalization assay, surface biotinylated neurons were put back at 37 °C for 30 min. The remaining surface bound biotin was removed by glutathione-stripping buffer. The internalized surface proteins that were protected from surface stripping were isolated by Neutroavidin bead precipitation. [0070] In vitro γ-secretase activity assay. The γ-secretase activity of mouse brains were performed as described previously (Li et al., PNAS 97:6138, 2000).

[0071] Lipid raft isolation. Lipid rafts were isolated from Lubrol WXlysates of mouse brains by discontinuous flotation density gradients as described previously (Vetrivel et al., J Biol Chem 280:25892, 2005). Briefly, 200mg of tissue from brains of 4-month-old APP/WT and APP/ Arc KO mice were homogenized with a glass-Teflon homogenizer in buffer A. After passage through a 25-gauge needle three times, homogenates were spun at 960 x g for 10 min at 4 °C. The supernatant was collected, and the pellet was resuspended in buffer A, which was passed through a 25-gauge needle five times and spun as described above. The two supernatants were pooled and adjusted to 0.5% Lubrol WX, 25 niM Tris-HCl (pH 7.4), 150 mM NaCl, and 5 mM EDTA and mixed at 4 °C for 20 min. The lysate was then adjusted to 45% final concentration of sucrose and transferred to a ultracentrifuge tube. A discontinuous sucrose gradient was formed by sequentially layering 35% sucrose and 5% sucrose, and the tubes were subjected to ultracentrifugation at 39,000 rpm for 19 hrs at 4 °C. Twelve fractions of equal volume were collected from the top of the gradient and equal volume of each fraction was analyzed by immunoblotting.

[0072] Co-Immunoprecipitation. Mouse brain fractions were prepared as described previously (Chowdury et al., Neuron 52:445, 2006) with modifications. WT and Arc/Arg3.1 KO mice brains were homogenized in 10 volumes of buffered sucrose with a glass-Teflon homogenizer, centrifuged at 800 x g for 15 min. The supernatant was again centrifuged at 9000 x g for 15 min. The supernatant was mixed with Triton X-100 to make a final concentration of 0.5% and incubated at 4°C for 30 min. 25 μL mouse monoclonal Arc antibody was added to the detergent lysates and incubated at 4°C overnight. Protein G Sepharose slurry was then added and incubated for another 2 hours. The beads were washed with PBS+ 0.5% Triton X-100 three times and eluted with SDS loading buffer.

[0073] HEK293T cells were maintained in Dulbecco's modified Eagle's medium (DMEM) with GlutaMAX, containing 10% heat-inactivated fetal bovine serum. HEK293T cells grown in 6- well plates to 60% confluence were transfected with 0.5μg cDNA each per well, using the FUGENE 6 transfection reagent according to the manufacturer's protocol (Roche). HEK293 cells were transfected with Arc, presenilin or endophilin constructs, and cell lysates were prepared in PBS with 1% Triton X-100 and protease inhibitors (Roche). Presenilin N-terminal antibody or glutathione sepharose beads were incubated with cell lysates overnight. Protein A agarose beads were then added to those antibody incubation samples and rotated at 4°C for another 2 hours. The beads were washed twice with PBS+ 0.5-1% Triton X-100 and once with PBS. Proteins were eluted with SDS loading buffer.

[0074] Immunocytochemistry, APP internalization assay and Microscopy. Neuronal transfections were performed with LIPOFECT AMIN 2000 transfection reagent (Invitrogen) in DIV 14- 16 neurons and were analyzed 16-24 hrs after initial incubation. Neurons were fixed in 4% paraformaldehyde, 4% sucrose PBS solution for 20 min at 4°C and were subsequently permeabilized with 0.2% Triton X-100 in PBS for 10min. Cells were then blocked for 1 hr in 10% normal goat serum (NGS). Primary antibodies were diluted in 10% NGS and incubated with neurons for 1 hr at room temperature or overnight at 4°C. Alexa secondary antibodies (1:500; Molecular Probes) to the appropriate species were diluted in 10% NGS and incubated at room temperature for 1 hr. Coverslips were mounted on precleaned slides with ProLong Gold antifacle reagent with DAPI (Molecular Probes).

[0075] For studies comparing APP internalization in WT and Arc KO neurons, human APP695 was expressed in DIV 21 cortical neurons by Sindbis pseudovirus prepared according to the manufacturer's manual (Invitrogen). 8 hrs after infection, neurons were incubated with anti-Aβ (6E10) antibody at 10 °C for 30 min. The unbound excess antibody was washed with PBS. For internalization assays, neurons were returned to 37 °C for 15 min, and surface APP was removed by incubating neurons in stripping buffer (0.5 M NaCl and 0.2 M ascetic acid) live for 5 min and were then fixed. Neurons were randomly chosen from 3 to 5 coverslips. Immunofluorescence was viewed and captured with Zeiss LSM 510 confocal laser scanning microscope. Confocal images were obtained using a 63x objective at a resolution of 1024 x 1024 pixels. Total internalized APP particles per 50 μm of dendritic length were counted by thresholding the pariicle intensity in randomly selected dendrites using Image J software (N1H). For those selected 50 μm dendrites, co-localized particles between PS 1 and internalized APP were manually counted. All the analyses were done blinded with regard to the sample identity.

[0076] Aβ and Filter trap assays on mouse brain. Brains of APP/WT (APP_SWE/PS 1 ΔE9; Arc+/+) and APP/ Arc KO (APP_SWE/PS IΔE9; Arc-/-) mice were dissected on ice and homogenized in PBS buffer containing complete protease inhibitor cocktail. After the lysates were centrifuged at 100,000 x 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 storage on ice for 1 h, the formic acid lysates were centrifuged at 100,000 x g for 1 h, 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 sandwich ELISA kit (Biosource International) that specifically detects human Aβ40/Aβ42. BCA method was used to measure the total protein concentrations (Pierce).

[0077] Immunohistochemical analysis of amyloid plaques in mouse brains. Mouse brain hemispheres were immersed in 10% formalin/ PBS for histology. Brains were dehydrated in methanol, treated with xylenes, and embedded in paraffin. 4 μm sagittal sections ~800μm from bregma were cut and used for the plaque staining. Before immunostaining, slides were deparaffinized by xylenes. After rehydration through graded ethanols into water, they were incubated with 88% formic acid for 5 min. Endogenous peroxidase activity was quenched by incubation with 0.9% hydrogen peroxide in methanol. Slides were microwaved for 5 min in water, cooled gradually and washed in PBS. Nonspecific staining was blocked with 3% normal goat serum (NGS) in PBS for 1 hour. Slides were then incubated with anti-human Aβ antibody (6E10; 1:500 dilution) in PBS + 3% NGS overnight at RT. After washing with PBS, slides were incubated with biotinylated goat anti-mouse IgG antibody (VECTOR laboratories) in PBS + 2% NGS for one hour. Then ABC reagent (VECTOR laboratories) was applied and sections were developed with diaminobenzidine (VECTOR laboratories). Quantification of plaques was carried out using Image J software (N1H). Pictures of 4 individual parts of cortex in each section were taken at the same condition and saved as TIFF files. To measure plaques, background was subtracted and the same threshold was set, then plaque area was counted automatically. All the 4 areas were summed and the percentage of plaque area was calculated and divided by the total area. Statistic analysis was done by Mann-Whiteney U test.

[0078] Analysis of human brain tissue. Human brain tissue was obtained from the Johns Hopkins Brain Bank and used in accordance with Institutional guidelines. A summary of the clinical and histological data for AD and control brain samples is provided in Fig. S4A. For Western blots, samples of medial frontal gyius gray matter were obtained by punch biopsy and dissolved in 2% SDS lysis buffer. Immunohistochemical staining of human brain for Arc was performed on 5 μm formalin- fixed paraffin-embedded sections of the medial frontal gyrus. The sections were incubated with polyclonal Arc antibody for 24 hrs at room temperature after antigen retrieval and followed by incubation with biotinylated anti-rabbit IgG for 1 hour, and visualized using the ABC kit with diaminobenzidine as a substrate.

[0079] Statistical analysis. All data were analyzed statistically by Student's t test or Mann-Whiteney U test. In all tests, the level of significance was set at P < 0.05 or P < 0.01.

[0080] To examine the role of Arc in activity-dependent secretion of Aβ, primary neuronal cultures were prepared from cortex of APPswe/PSlΔE9 transgenic mice (a well-established model exhibiting accelerated Aβ amyloidosis attributable to coexpression of familial AD linked mutations of APP (APPswe) and PS1(PS1ΔE9) in CNS of mice (Jankowsky et al., J Neuropathol Exp Neurol 62: 1120, 2003), in either WT or Arc KO backgrounds and monitored Aβ in the medium. The levels of Aβ40 and Aβ 42 were significantly less in the medium of Arc KO cultures (Figure 3, A and B). Addition of bicuculline (40μM, 48 hrs), which inhibits GABA-A receptors and increases network activity and Arc expression, resulted in a 2.5-3 fold increase of Aβ40 and Aβ42 in medium of WT neurons. By contrast, bicuculline did not increase Aβ40/42 levels in medium of Arc KO neurons.

[0081] Since Arc increases AMPAR endocytosis, the question of whether Arc might enhance the rate of internalization of APP or its processing enzymes was addressed. Expression of APP, BACEl and components of γ-secretase, including presenilin 1 (PSl) and Nicastrin were not different in total cell lysates of WT and Arc KO neurons. Similarly, there was no change in their level of expression on the plasma membrane, or the amount of these proteins internalized during 30 min. Thus, in contrast to its effect on AMPAR, Arc did not alter the dynamics of APP/BACE1/PS1 trafficking from the membrane. Treatment of cultured neurons with the selective γ-secretase inhibitor JC-22 (Lewis et al., Bioorg Med Chem Lett 15:373, 2005) resulted in identical increases of C-terminal fragments of APP that result from cleavage by BACEl and/or α-secretase, suggesting that APP processing upstream of γ-cleavage esd similar in WT and Arc KO neurons, γ-secretase activity measured in vitro was not different in lysates of WT and Arc KO brain (Figure 3C). [0082] To assess molecular mechanisms that could mediate Arc-dependent Aβ secretion, sucrose gradient fractions of brain were prepared. The N and C-terminal fragments of PSl and Nicastrin were detected in buoyant fractions in association with flotillin-2 and lipid rafts. These fractions were also enriched for BACEl and APP C-terminal fragment, but not full length APP. Arc, together with proteins that physically interact with Arc and are essential for its endosomal pathway including endophilin 2, endophilin 3 and dynamin 2, were also enriched in the flotillin-2 fractions. To examine the possible physical association of Arc with APP processing enzymes, IP assays from brain were preformed. Using detergent conditions that dissociate γ-secretase, PS1-NTF co-immunoprecipitated with Arc, while PS1-CTF, Nicastrin and BACEl did not. These observations suggested that Arc selectively associated with the γ-secretase complex. This interaction was examined further in HEK293 cells that expressed PSl and Arc transgenes. PS1-NTF and full length PSl co-immunoprecipitated with full length Arc, as well as with an N-terminal fragment of Arc. Deletion of Arc aas 91- 100 or 101-130 disrupted its association with PSl. Since Arc aa 91-130 was also critical for interaction with endophilin, competitive binding in HEK293 cells expressing mammalian GST- Arc was assayed. Co-expression of endophilin 3 CT, which avidly binds Arc, did not alter the ability of PSl to co-immunoprecipi1ate with GST- Arc. These observations indicated that Arc can bind endophilin 3 and PSl simultaneously.

[0083] The Arc interaction with PSl was further examined histochemically in cultured hippocampal neurons. As reported previously by Chowdhury et al. (Neuron 52:445, 2006), Arc and endophilin 3 transgenes co-localized in dendrites as punctae. Both native PSl and PSl transgene localized to the somatodendritic compartment with a reticular and punctate pattern, consistent with its reported mixed ER and Golgi distribution. When co-expressed with Arc and endophilin 3, PSl transgene co-localized with Arc punctae within regions of distal dendrites.

[0084] The question of whether Arc-endophilin vesicles traffic APP was addressed. The anti-Aβ antibody 6E10 detects a region within human APP (hAPP), but not rodent APP, that is exposed when hAPP is on the plasma membrane (Cai et al., PNAS 103: 1936, 2006). Addition of 6E10 antibody to cultured rat or mouse neurons that express hAPP transgene detected specific surface expression. When applied to living neurons expressing hAPP, 6E10 antibody was rapidly internalized at 37°C, but not at 4°C. Internalized hAPP detected 15 min after addition of 6E10 co-localized with Arc-endophilin punctae in dendrites. Internalized hAPP also co-localized with internalized transferrin, a marker of recycling endosomes that co-localizes in Arc endosomes.

[0085] To examine the possibility that Arc might recruit PSl to APP containing recycling endosomes, hAPP was expressed in WT and Arc KO neurons, and assayed for co-localization of native PSl and internalized hAPP. Consistent with biochemical data, hAPP was identically internalized in WT and Arc KO neurons. In most dendrites the reticular pattern of PSl contrasted with the highly punctate pattern of internalized hAPP. Nevertheless, APP punctae frequently co-localized with punctae of PSl, and their co-localization was significantly greater in WT than Arc KO neurons. These data were consistent with the notion that PSl is present in multiple vesicular organelles, and that a portion of PSl associates with APP-containing recycling endosomes. Further, they indicated that Arc and endophilin recruit PSl to a recycling endosome that traffics APP, and support a current model of Aβ generation that proposes BACEl cleaves APP in early endosomes upon their acidification, and that γ- cleavage occurs subsequently with the fusion of vesicles that include PSl and mature γ- secretase.

[0086] Previous studies have suggested that presynaptic mechanisms contribute importantly to the deposition of insoluble Aβ in the pathogenesis of AD (Lazarov et al., J Neurosci 22:9785, 2002). Since Arc is expressed exclusively in postsynaptic neurons, the question of whether Arc-dependent mechanisms contribute to amyloid generation in the APP_SWE/PS 1ΔE9 transgenic mouse model of AD was addressed. Expression of hAPP and components of γ-secretase were identical in APP/WT and APP/ Arc KO forebrain assayed at 12 month of age. Aβ levels in PBS-soluble and formic acid-soluble fractions were reduced in forebrain of 12-month-old APP/ Arc KO mice (Figure 4A). A filter trap assay confirmed the reduction of Aβ deposits in APP/ Arc KO forebrain (Figure 4B). Immunohistochemical analysis using 6E10 antibody revealed reduced plaque area in APP/ Arc KO mice. Consistent with previous reports of this mouse model, Aβ levels were markedly age-dependent. Nevertheless, 6-month-old mice revealed reduced formic acid-soluble Aβ40 (Figure 4C)and plaque in APP/ Arc KO mice. These observations confirmed a role for Arc-dependent mechanisms in the generation of Aβ and formation of plaque. [0087] In healthy animals, Arc expression is linked to neuronal activity that underlies learning and memory (Guzowski el al., Ciiir Opin Neurohiυl 15:599, 2005). In view of the cognitive deficits of AD, it might be anticipated that Arc levels, and its contribution to Aβ generation, would be progressively reduced. Indeed, Arc protein was reduced by ~50% in APPswc/PSIΔE9 transgenic mice compared to age-matched WT mice (Dickey ci al., J Neurochem 88:434, 2004). This was examined in human AD brain samples (clinical and pathological summary in Table 1 below) and found Arc protein is increased compared to control brains. By contrast, another IEG Zif268 was not different in control and AD brains, lmmunohistology confirmed that Arc protein is present selectively in neurons of both control and AD subjects. While the basis of this increase is not known, it supports the hypothesis that Arc contributes to the generation of Aβ in human AD.

Table 1

F VlD liC.

^'. "Jf L-- .e-i -;•- PJCT C ^'.T .ι:r . 'I όi. -¹J >•'

I F ^":^■ :> Wl: ,1 Il -

_ F •-.; • I ItI- I Il -

Iv' i: O ItI-. ,1 Il - w ~: • •1 ItI- ,1 Il -

F •:■ Ur-. :l I I^'r/ ~: • •j Hi-: ^>l ll -

I/ M Ul-. ^•1 Il -

F -•; • I Il -

F .' I ^•:■■«!•: I Il - ι; F r : -C

I F :; r .- -C ι: F -- b IT : -C ι: I/ •:•: -C

I- Iv' ^"- I: : -C ι: F :- -C

I-: F ~: • -C

17 F -T > -C

I: F 4 -C ι- F >-: -C

[0088] The present studies idcntifed a postsynaptic, activity-dependent molecular pathway that modulates secretion of Aβ. These observations were consistent with reports that PSI is enriched in endocytic structures, and that Aβ42 accumulates in late endosomes that arc enriched in postsynaptic neurons. These data support a model wherein Arc and endophilin enhance the fusion of endosomes containing γ-secretase with recycling endosomes that contain APP. It was noted that endophilin binds synaptojanin, which can modify the protein and lipid composition of membranes to favor vesicular fusion. Endophilin was identified in a screen for genes that regulate Aβ formation. In addition to APP, γ-secretase cleaves several other membrane proteins, and is essential for signaling of Notch. Indeed, Notch signaling in the adult brain is dependent on Arc and synaptic activity.

[0089] Although the invention has been described with reference to the above examples, it will be understood that modifications and variations are encompassed within the spirit and scope of the invention. Accordingly, the invention is limited only by the following claims.

Claims

What is claimed is:

1. An amino acid fragment of the polypeptide set forth in SEQ ID NO: 1 , wherein the fragment binds presenilin 1 (PSl, SEQ ID NO:3).

2. The amino acid fragment of claim 1, wherein the fragment induces PSl to accumulate in endosomes.

3. The amino acid fragment of claim 1, wherein the fragment is about 2 to 20 amino acids in length.

4. The amino acid fragment of claim 1, wherein the polypeptide comprises the amino acid sequence set forth in SEQ ID NO:5.

5. The amino acid fragment of claim 1, wherein the wherein the amino acid fragment is modified at the N-terminus or the C-terminus.

6. The amino acid fragment of claim 5, wherein the wherein the amino acid fragment is modified to include a non-proteii polymer.

7. The amino acid fragment of claim 6, wherein the non-protein polymer is a polyethylene glycol, a polypropylene glycol or a polyoxyalkylene.

8. A method of identifying a compound that inhibits binding of an Arc polypeptide to presenilin 1, comprising:

(a) contacting a test compound with a sample containing an Arc polypeptide or fragment thereof that oinds PSl and a presenilin 1 polypeptide; and

(b) comparing the level of Arc polypeptide and the presenilin 1 (PSl) polypeptide binding in the presence and absence of the compound, wherein when the binding in the presence of the compound is lower than in the absence of the compound, then the compound is an inhibitor of Arc polypeptide and a presenilin 1 (PSl) polypeptide binding.

9. The method of claim 8, wherein the Arc polypeptide is about 2 to 20 amino acids in length.

10. The method of claim 8, wherein the Arc polypeptide comprises the amino acid sequence set forth in SEQ ID NO: 5.

11. A method of inhibiting gamma-secretase activity in a cell comprising: contacting PSl with a compound that inhibits binding of Arc to PSl, thereby reducing or eliminating accumulation of PSl in endosomes and incorporation of PSl in gamma-secretase, thereby inhibiting gamma-secretase activity.

12. The method of claim 11, wherein formation of amyloid beta is reduced.

13. A method of inhibiting amyloid beta plaque formation in a subject, comprising inhibiting the interaction of Arc and presenilin 1 (PSl).

14. A method of inhibiting the formation of amyloid beta plaques in, on or around neurological tissue comprising administering to a subject in need of such treatment an effective amount of a compound that blocks the interaction between SEQ ID NO:1 and PSl, thereby inhibiting formation of amyloid beta plaques in the subject.

15. The method of claim 14, wherein the subject has or is at risk of having a neurodegenerative disease.

16. The method of claim 15, wherein the neurodegenerative disease is Alzheimers's Disease or age-dependent memory decline.

17. The method of claim 14 wherein the subject has a mutation in the PSl gene.