EP1587820A2 - Isoform specific interactions of apolipoprotein e to an intermediate conformation of alzheimer a beta peptide - Google Patents

Abstract

The present invention provides assays useful in finding inhibitors of the interaction between apoliprotein E4 (apoE4) and an active form of Alzheimer Aβ peptide.

Description

Isoform Specific Interactions of Apolipoprotein E to an Intermediate Conformation of Alzheimer Aβ Peptide

FIELD ⁽XF THE INVENTION Trie present invention provides assays useful in fincling inhibitors of the interaction between apoliprotein E4 (apoE4) and an active form of Alzheimer Aβ peptide

BACKGROUND OF THE INVENTION Alzheimer's disease is characterized by progressive neurodegeneration associated with the deposition of extracellular proteinaceous amyloid-β peptide (Aβ) in trie form of senile plaques (Selkoe 1991). Plaque deposits are a result of self-aggregating monomeric Aβ peptides, by a process termed amyloidogenesis. Amyloidogenesis is initiated in vitro by micromolar amounts of monomeric Aβ (Harper 1997). In amyloidogenesis, the aggregation of monomeric subunits is consistent with a kinetic model where the peptide first slowly forms a conformer, which is the only species able to initiate the formation of linear aggregates and is the only neurotoxic form of Aβ.

Apolipoprotein E (apoE), a ligand that regulates lipid and cholesterol transport and clearance in brain through the LDL receptor related protein (LRP) (Boyles 1989, Corder 1993, Fagan 1996, Poirier 1994), has fceen identified to play a role in plaque deposition in AD t rain. ApoE has three predominant isoforms found in the human population, E2 (Cys¹¹², Cys¹⁵⁸), E3 (Cys¹¹², Arg¹⁵⁸), and E4 (Arg¹¹², Arg¹⁵⁸) (Weisgraber 1994). In individuals that had AD, apoE4 carriers have shown a greater number of senile plaques when compared to apoE3 carriers (Gearing 1996). Subjects with the ApoE4/ 4 genotype are as much as eight times as likely to be affected by Alzheimer's disease as subjects with the ApoE2/3 or ApoE3/3 genotypes. Further, the average age of onset of Alzheimer's disease and the average age of survival is lower for those having one ApoE4 allele, and lowest for those having two ApoE4 alleles (U.S. Pat. No. 5,508,167). Inheritance of the rare apoE2 allele is associated with a reduced risk of developing AD (Corder 1994).

ApoE is hypothesized to be involved in the deposition or clearance of Aβ peptide by direct protein-to-protein interactions. LaDu (1994, 1997) demonstrated apoE isoform-specific differences in the interaction with Aβ using native lipidated apoE molecules, whereas isoform specificity was abolished with purified apoE isoforms. The studies by LaDu did not investigate apoE effects on differing forms of Aβ evident throughout aggregation and therefore did not sliow a preferential association with apoE4. Tokuda et al reported an isoform specific interaction with apoE3 and Aβ. We however have established that various Aβ conformations confer differing degrees of neurotoxicity and that a toxic form of Aβ can be reproducibly produced. It was of great interest therefore to develop an isoform specific/toxic species specific assay procedure for the interaction of apoE4 with Aβ not only to validate the physiologic significance of the interaction but as a means of developing assays for finding modulators of the interaction. Such modulators would be useful in the treatment and prevention of amyloid associated diseases in patients with an apoE4 allele.

References Cited 1. Ainsztein, A.M. and D.L. Purich, Stimulation of tubulin polymerization by MA.P-2. Control by protein kinase C-mediated phosp τorylation at specific ites in the microtubule-binding region. J Biol Chem, 1994. 269(45): p. 28465-71.

2. Blow, D.M., et al., Control ofnucleation of protein crystals. Protein Sci, 1994. 3(10): p. 1638-43.

3. Boyles, J.K., et al:, A roleβor apolipoprotein E, apolipoprotein A-I, and low density lipoprotein receptors in cholesterol transjport during regeneration and remyelincition of the rat sciatic nerve. J Clin Invest, 1989. 83(3): p. 1015-31. 4. Cooper, J.A., et al., Kinetiσ evidence for a monomer activation step in actin polymerization. Biochemistry, 1983. 22(9): p. 2193-202. 5. Corder, E.H., et al., Gene dose of apolipoprotein E tyjye 4 allele and the risk of Alzheimer's disease in late onset families. Science, 1993. 261(5123): p. 921-3. 6. Corder, E.H., et al., Protective effect of apolipoprotein E type 2 allele for late onset Alzheimer disease. Nat Genet, 1994. 7(2): p. 180-4. 7. Fagan, A.M., et al., Apolipoprotein E-containing hign density lipoprotein promotes neurite outgrowth- and is a ligandfor the low density lipoprotein receptor-related protein. J Biol Chem, 1996. 271(47): p. 30121-5. 8. Findeis, M . and S.M. Molineaux, Design and testing of inhibitors of fibril formation. Methods Enzymol, 1999. 309: p.476-88. 9. Gallop, M.A., et al., Applications of combinatorial technologies to drug discovery. 1. Background and peptide combinatorial libraries. J Med Chem, 1994. 37(9): p. 1233-51. 10. Gearing, M., H. Mori, and S.S. Miτra., Abeta-peptide length and apolipoprotein E genotype in Alzheimer's disease. Ann Neurol, 1996. 39C3): p. 395-9. 11. Gordon, E.M., et al., Applications of combinatorial technologies to drug discovery. 2. Combinatorial organic synthesis, library screening strategies, and future directions. J Med Chem, 1994. 37(10): p. 1385-401.

12. Harper, J.D. and P.T. Lansbury, Jr., Models of amyloid seeding in Alzheimer's disease and scrapie: mechanistic truths and physiological consequences of the time-dependent solubility of amyloid proteins. Annu Rev Biochem, 1997. 66: p. 385-407.

13. Havel, R.J., N. Yamada, and D.M. Shames, Hole of apolipoprotein E im lipoprotein metabolism. Am Heart J, 1987. 113(2 Pt 2): p. 470-4. 14. Kadonaga, J.T. and R.. Tjian, Affinity purification of sequence-specific

DNA binding proteins. Proc Natl Acad Sci U S A, 1986. 83(16): p. 5889- 93. 15. LaDu, M.J., et al., Isoform-specific binding of apolipoprotein E to beta- amyloid. J Biol Chem, 1994. 269(38): p. 234-03-6. 16. LaDu, M.J., et al., Association of human, rat and rabbit apolipoprotein E with beta-amyloid. J Neurosci Res, 1997. 49<1): p. 9-18.

17. LaDu, M.J., et al., Nascent astrocyte particles differ from lipoproteins in CSF. J Neurochem, 1998. 70(5): p. 2070-81.

18. Mahley, R.W., Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science, 1988. 240(4852): p. 622-30.

19. Morrow, J.A., K.S. Arnold, and K.H. Weisgraber, Functional characterization of apolipoprotein E isoforms overexpressed in Escherichia coli. Protein Expr Purif, 1999. 1 (2): p. 224-30.

20. Poirier, J., Apolipoprotein E in animal models of CNS injury and in Alzheimer's disease. Trends Neurosci, 1994. 17(12): p. 525-30.

21. Selkoe, D.J., The molecular pathology of Alzheimer's disease. Neuron, 1991. 6(4): p. 487-9S.

22. Strittmatter, W.J., et al., Binding of human apolipoprotein E to synthetic amyloid beta peptide: isoform-specific effects and implications for late- onset Alzheimer disease. Proc Natl Acad Sci U S A, 1993. 90(17): p.

8098-102.

23. Tobacman, L.S. and E.D. Korn, The kinetics ofactin nucleation and polymerization. J Biol Chem, 1983. 258(5): p. 3207-14.

24. Voter, W.A. and H.P. Erickson, The kinetics of microtubule assembly. Evidence for a two-stage nucleation mechanism. J Biol Chem, 1984.

259(16): p. 10430-8.

25. Weisgraber, K.H., Apolipoprotein E: structwre-function relationships. Adv Protein Chem, 1994. 45: p. 249-302.

26. Tokuda, T. et al. Lipidation of apolipoprotein E influences its isoform— specific interaction with Alzheimer's amyloid β peptides Biochem J.

(2000) 348: p. 359-365.

27. Weisgraber, K. H., Apolipoprotein E Structi re-Function Relationships Advances in Protein Chemistry 45: p. 249-250

28- Jonas, A., Reconstit tion of High Density Lipoproteins, Methods in Enzymology, Plasma Lipoproteins Part A, Preparation, Structure, and

Molecular Biology Eds Segrest, J.P. and Albers, J.J. 128: p 553-575. SUMMARY OF THE INVENTION

The present invention addresses the need identified above in tliat it provides a method of identifying a lipidated ApoE4 derived polypeptide- activated Aβ complex formation inhibitor compound comprising the steps of contacting a lipidated apoE4 derived polypeptide with an activated Aβ polypeptide in presence of a test compound; and identifying test compounds that decrease the formation of the ApoE4 - activated Aβ complex wherein, a test compound which decreases the formation of the lipidated ApoE4 derived polypeptide- activated Aβ complex is complex formation inhibitor compounds. In addition to the foregoing, the invention includes, as an additional aspect, all embodiments of the invention narrower in scope in any way than the variations specifically mentioned above. Although the applicant(s) in. vented the full scope of the claims appended hereto, the claims appended hereto are not intended to encompass within their scope the prior art work of others. Therefore, in the event that statutory prior art within the scope of a claim is brought to the attention of the applicants by a Patent Office or other entity or individual, the applicant(s) reserve the right to exercise amendment rights under applicable patent laws to redefine the subject matter of such a claim to specifically exclude such statutory prior art or obvious variations of statutory prior art from the scope of such a claim. Variations of the invention defined by such amended claims also are intended as aspects of the invention.

Brief Description of the Figures

Figure 1- Typical Aβ^1"40 Shaken Aggregation Experiment by Turbidi"ty Detection: A 50 μ solution of Aβ^1"40 was induced to aggregate by sfcaking according to the details given in the Methods section. The ordinates are in units of A₄o₅ x 10³,with 138 units being equal to 50 μM. The solid lines are theoretical curves calculated using Equations 1-4 and the rate constan.ts given in Table 1. The left ordinate is for the starting peptide and the aggregated species while the right ordinate is for the activated monomer, the growing site, and the tetramer. The solid circles represent the individual experimental data points.

Figure 2- Time Dependencies of Turbidity and Peptide Concentration: The relationship between turbidity detection and peptide mass was determined in parallel experiments. The mass of the peptide remaining in solution after centrifugation was quantitated as described in Methods. The data wer-e analyzed in terms of Equations 1-4 and the solid lines represent the tbxeoretical fits to the experimental data points. Figure 3- Kinetics of Spontaneous Aggregation by CD Detection: The aggregation kinetics experiments were done using the shaking procedure with A₄₀₅ or CD ellipiticy at 200-205 nm as described in Methods. The solid lines represent best-fit theoretical curves. Figure 4- Changes in the CD Spectrum of Aβ^1"40 as a Function of Aggregation Time: CD spectra were recorded at different times during the aggregation process and the spectra analyzed as described in Methods. Figure 5- The Effect of Peptide Concentration on Aggregation; Detection by Turbidity and Fluorescence Polarization: Aggregation experiments were performed as a function of peptide concentration as described in IVlethods by turbidity and fluorescence polarization detection. Solid circles: 25 μM peptide; open circles: 50 μJM peptide. The solid lines are the theoretical fits to the experimental data points. Figure 6. Detection of lipidiated and unlipidated apoE binding to different forms of Aβ 1-40 by ELIS A utilizing anti-apoE N-terminal capture antibodies. Aβ 1-40 was obtained from 0, 60, 80, and 110 min that were shaken throughout spontaneous aggregation as described. Panels A and B = capture antibody 6C5, panels C and D = capture antibody 9H8.G5.F2, panels A and C = lipidated apoE, panels B and D = unlipidated apoE, • = apoE2, o = apoE3, ▲ = a oE4, * = p<0.05 versus apoE2, + = p<0.05 apoE4 versus apoE3. n = 5 for C5C5 and 9H8.G5.F2 lipidated apoE experiments, and n = 4 for 6C5 and9H8.G5.F2 unlipidated experiments.

Figure 7. Detection of lipidiated and unlipidated apoE binding to different forms of Aβ 1-40 byELISA utilizing anti-apoE C-terminal capture antibody 3H1. Aβ 1- 40 was obtained from 0, 60, 80, and 110 min that were shaken throughout spontaneous aggregation as described. Panel A = lipidated apoE, panel B = unlipidated apoE, • = apoE2, o = apoE3, ▲ = apoE4. n = 4 for lipidated apoE experiments and n = 3 for unlipidated apoE experiments. Brief Description of the Sequence Listings

SEQ ID NO:l arnino acid sequence of full length ApoE4 SEQ ID NO:2 amino acid sequence of Aβ 1-43 SEQ ID NO:3-10 Primer Sequences, Example 2

DETAILED DESCRIPTION OF THE INVENTION General Definitions

Apolipoprotein E (apoE), is a protein species which regulates lipid and cholesterol transport and clearance in brain through the LDL receptor related protein (LRP) (Boyles 1989, Corder 1993, Fagan 1996, Poirier 19S4), and which has been identified to play a role in plaque deposition in AD brain . There are three major isoforms of ApoE, referred to as ApoE2, ApoE3 and

ApoE4 which are products of three alleles at a single gene locus. Xhree homozygous phenotypes (Apo-E2/2, E3/3, and E4/4) and three heterozygous phenotypes (ApoE3/2, E4/3 and E4/2) arise from the expression o:f any two of the three alleles. The most common phenotype is ApoE3/3 and the most common allele is E3. See Mahley, R. W., Science 240:622-630 (1988). The amino acid sequences of the three types differ only slightly. ApoE4 differs from ApoE3 in that in ApoE4 arginine is substituted for the normally occurring cysteine at amino acid residue 112. The most common form of ApoE2 differs from ApoE3 at residue 158, where cysteine is substituted for the normally occurring arginine. See Mahley, Science, supra. The three isoforms therefore can be represented by the following informative shorthand representation of their sequence differences E2 (Cys¹¹², Cys¹⁵⁸), E3 (Cys¹¹², Arg¹⁵⁸), and E4 (Arg¹¹², Arg¹⁵⁸) ("Weisgraber 1994). Human apoE is a 34-kDa protein consisting of 299 amino acid residues that have three distinct functional domains (Weisgraber 1994) (Figure 1): (1) a N- terminal receptor binding region, (2) a random coil region susceptible to protease cleavage, and (3) the C-terminal lipid-binding region. Residues 136-158 in the N- terminal region interact with apoE associated receptors, residu.es 165-210 is highly susceptible to proteolysis, and residues 225-299 constitute the C-terminal portion with residues 268 to 289 involved in lipid binding. We have discovered that residues 243-272 are involved in binding to Aβ.

An "apoE4 derived polypeptide" as desecribed herein would include the mature human apoE4 polypeptide as described by Mahley arid as reproduced herein below (SEQ JD NO: 1):

KVEQAVETΞPEPELRQQTE QSGQR E AE-,GRF DYLR VQTLSEQVQEHI⁵⁰ LLSSQVTQΞLRALMDETMKE AYKSE EEQ TPVAΞΞTRAR SKELQA.^.¹⁰⁰ QAR GAD EDVJϋGR VQYRGEVQA GQSTEELRVRLASHLRKLRKRI- I ¹⁵⁰ DADDLQKRrjAVYQAGAREGAERGLSAIREl GPLVEQGRVRAATVGSLAG²⁰⁰ QPLQERAQA.WGERLRARMEEMGSRTRDRLT3EVKΞQVAEVRAKLEEQAQQ ²⁰⁰ RLQAEAFQA.RLKSWFEPLVEDMQRQ AGLVEKVQAAVGTSAAPVPSDNH² " as well as species homologues of the human apoE4. Many species homologues of apoE have been described by Weisgraber and this definition would include the baboon, cynomologous monkey, rat, mouse, guinea pig, rahbit, cow, dog, sea lion as well as other species homologues and polypeptide fragments of the above sequences .

The term " Aβ polypeptide" as used herein refers to a 38-43 amino acid peptide ha^lving a molecular weight of about 4.0 kD, which peptide is substantially homologous to the form of the protein described by Glenner, et al. (22) including mutations and post-translational modifications of the normal ^"beta -amyloid peptide. In whatever form, the beta -amyloid peptide is an approximately 38-41 amirio acid fragment (differing at the carboxy terminus of the fragment) of a laαrge membrane-spanning glycoprotein, referred to as the beta -amyloid precursor protein (APP). Its 43-amino acid sequence is: (SEQ ID NO: 2) 'DAEFRHDSGYEVHHQ LVF FAEDVGSNKGAI IGLMVGGWIAT" or a sequence which is homologous thereto. By way of example the corresponding rat and mouse sequences are about 95% homologous to the sequence above, are intended to be encompassed by this definition and can be deduced from Genbank accession number P08592 and AAB41502 respectively. Other homologues are easily deduced from public and private databases or via cloning and sequencing the corresponding APP from the appropriate species.

An Aβ polypeptide may exist in either soluble, insoluble, or intermediate forms. After a concentration-dependent lag period during in vitro incubations, soluble preparations of synthetic beta AP slowly form an activated beta amyloi species which then ultimately gives rise to fibrillar aggregates that resemble natural amyloid and are separable from the aqueous medium by sedimentation. For convenience, in this specification, an Aβ polypeptide is often referred to as Aβ 1-38, Aβ 1-39, Aβ 1-40, Aβ 1-41, Aβ 1-42, Aβ 1-43 or simply "Aβ"

The term " aggregated Aβ peptide" refers to Aβ peptide in an insoluble state.

The term "activated Aβ polypeptide" refers to a, polypeptide species generated during spontaneous aggregation which exhibits an isoform specific propensity to form a complex with an apoE4 polypeptide. Procedures for the generation of an activated A.β polypeptide is described in this specification in detail below.

A "lipidated apoE4 derived polypeptide" is an apoE4 derived polypeptide which comprises in addition to the apoE4 derived polypeptide polar or neutral lipid components and which is capable of specifically forming a complex with an activated beta -amyloid peptide where the interaction is sufficiently strong to permit measurement of their association. Methods of accessing the propensity of a polypeptide to associate with another polypeptide are well known in the art and illustrative examples are discussed Lipid componenets would include esterified or unesterified (free) cholesterol, triglycerides, and phospholipids. In one embodiment the lipid components in the aggregate are capable of forming a lipid bilayer. S ome emobodiments include an apoE4 complexed with phosphatidylcholin.es (PC) including dilauroyl PC (DLPC), dimyristoyl PC (DMPC), dipalmitoyl PC

(DPPC) and palmitoyl-2-oleoyl-sn-glycero-3-PC or phosphatidylehtanolamines, phosphatiydylserines, phosphatidylinositols, ceramides, and sphingornyelins, cerebrosides, sodium cholate, cholesterol, and cholesterol esters. Methods of creating reconstituted lipoprotein particles are well known in the art and can be made with various ratios of lipid mixtures and protein concentrations in the manner which has been described in detail by Jonas. Many methods exist for associating apolipoproteins with various lipids which include , cosoni ation of lipid and protein components, spontaneous interacion of protein with lipid vesicles and various methods of detergent-mediated reconstitution. As used herein, the term "contacting" means bringing together, either directly or indirectly, a compound into physical proximity to a polypeptide or of the invention. Additionally "contacting" may mean bringing a polypeptide of the invention into physical proximity with another polypeptide.

The term "complex" as used herein in intended to mean a combination of molecules bonded together and is not intended to convey any particular mode of bonding.

The term "homologous" is used here to illustrate the degree of identity between the amino acid sequence of a given polypeptide and another amino acid sequence. The amino acid sequence to be compared with the amino acid sequence of the given polypeptide may be deduced from a DNA sequence, e.g. obtained by hybridization as defined above, or may be o ^"btained by conventional amino acid sequencing methods. The degree of homology is preferably determined on the amino acid sequence of a mature polypeptide, It is preferred that the degree of homology is at least 80%, such as at least 90%, preferably at least 95^& or even 98% with the amino acid sequence between the amino acid sequences compared. Homologous amino acid sequences include those amino acid sequences which contain conservative amino acid substitutions. Percent homology can be determined by, for example, the Gap program (Wisconsin Sequence Analysis Package, Version 8 for Unix, Genetics Computer Group, University Research Park, Madison \NI), which uses the algorithm of Smith and Waterman (Adv. Appl. Math., 1981, 2, 482-489) using the default settings.

"Isolated" as used herein and as understood in the art, is taken to mean separated from the original cellular environment in which the polypeptide is normally found. As used herein therefore, by way of example onl^, a protein expressed by recombinant means in a cell type in which it does not naturally occur is "isolated". By way of example, a protein species, whether expressed in a naturally occurring cell or not, when purified to any extent is "isolated" As used hereinafter "polypeptide" refers to any peptide or protein comprising two or more amino acids joined to each other by peptide bonds or modified peptide bonds, i.e., peptide isosteres. "Polypeptide" refers to both short chains, commonly referred to as peptid s, oligopeptides or oligomers, and to longer chains, generally referred to as proteins. Polypeptides may contain amino acids other than the 20 gene-encoded amino acids. "Polypeptides" include amino acid sequences modified either by natural processes, such as post-translational processing, or by chemical modification techniques which are well known in the art. Glycosylated and non-glycosylated form of polypeptides are embraced by this definition. "Synthesized" as used herein and understood in the art, refers to polynucleotides produced by purely chemical, as opposed to enzymatic, methods. "Wholly" synthesized DΝA or polypeptide sequences are therefore produced entirely by chemical means, and "partially" synthesized DΝAs or polypeptides embrace those wherein only portions of the resulting DΝA or polypeptide were produced by chemical means.

The term "test compound" means any means identifiable natural or synthetic chemical or molecule, including, but not limited to a small molecule, peptide, protein, sugar, nucleotide, or nucleic acid which is assessed for its ability to modulate the propensity of a activated Aβ polypeptide to associate with a lipidated apoE4 derived polypeptide. Specifically included within this definition are antibodies (e.g., monoclonal and polyclonal antibodies, single chain antibodies, chimeric antibodies, bifunctional/bispecific antibodies, humanized antibodies, human antibodies, and complementary determining region (CDR)-grafted antibodies, including compounds which include CDR sequences which specifically bind apoE4 derived polypeptides or activated Aβ. This definition includes human antibodies that are produced and identified according to methods described in W093/11236, published June 20, 1993. This definition includes Antibody fragments, including Fab, Fab', F(ab')₂, and F_v, are also provided by the invention. The definition includes antisera isolated from an animal is an exemplary composition, as is a composition comprising an antibody fraction of an antisera that has been resuspended in water or in another diluent, excipient, or carrier. The term "lipidated apoE4 derived polypeptide-Aβ complex" or complex as used herein is intended to mean a complex between a apoE4 protein derived polypeptide and a beta -amyloid peptide.

Introduction A pathological hallmark: of Alzheimer's disease (-AD) is the presence of amyloid peptide (Aβ) in the form of extracellular plaques in brain. Prior to plaque formation, monomeric Aβ fragments undergo a transforroation from single units to aggregated products that lead to fibril formation and ultimately plaque deposits. Apolipoprotein E (apoE) is trrought to be involved in Aβ plaque formation. Individuals afflicted with AD carrying the apoE4 isoform have shown a greater number of Aβ plaques when compared to apoE3 carriers. Most notably, inheritance of an apoE4 allele increases the risk of AD when compared to apoE.2 and apoE3 carriers. The nature of the involvment of ApoDE with the pathogenesis of AD has been heretofore poorly understood. We have discovered a method of reproducibly generating an active species of Aβ, generated during spontaneous aggregation, which preferentially associates with the ApoE4 isoform. We describe below first how the active beta amyloid species can be generated and harvested. We then describe how lipidated apoE4 derived polypeptide can be generated. Lastly Λve describe how assays may be performed to assess the interaction between the two.

Generation of an Activated Aβ polypeptide exhibiting Isoform specific

Binding of ApoE

The assays described below make use of isolated A.β polypeptides either in one of it's molecular forms (typically the 38, 39, 40, 41„ 42 or 43 amino acid variants). Such polypeptides can be purchased ( For example from Sigma Biochemicals --Fragment 1-38 Cat # A0189, Fragment 1-40 Cat # A1075, Fragment 1-42 Cat # A9810, Fragment 1-43 Cat # AT712 ) Such polypeptides may also be synthetically produced by means well known in the art. Solid state peptide synthesis is well known to those of ordinary skill in the art, and is described generally by errifield, 1963, J. Amer. Chem. Soc. 85:2149-2156, Fields and Noble, 1990, Int. J. Pept. Protein Res. 35:161-214 and Solid Phase Peptide Synthesis: A practical approach" by Atherton and Sheppard (published by IRL press at Oxford University Press, 1989) The peptides and proteins disclosed herein may thus be prepared using these relatively routine techniques given the disclosure of the present invention. Solid state peptide synthesis is well known to those of ordinary skill in the art, as described by the references provided and can be performed manually or by an automated peptide synthesizer such as those sold by ABS. One method of synthesis is accomplished by solid phase peptide synthesis using the Fmoc strategy using an automated peptide synthesizer. This method involves building an amino acid chain from the -COOH terminus, which is attached to an insoluble polymeric support. The base-labile Fmoc group is used to protect the -amino group of each residue. Residues having potentially reactive side chains are protected with acid-labile groups such, as t-butyl. After removal of the Fmoc group during each cycle with piperidine, the next protected amino acid is added using either a coupling reagent or pre-activated amino acid derivative. At the end of the synthesis, the peptide is cleaved from the solid support to yield a peptide acid or amide, depending on the linking agent used, and the side-chain protecting groups are remo ed by treating the peptide-resin with a mixture of trifluoroacetic acid and various ion scavengers. Methyl t-butyl ether is added to precipitate the peptide out of the cleavage mixture. The crude peptide is dissolved and lyophilized, after which it was purified by high performance liqnid chromatography. The purified peptide is lyophilized and stored at - 80° C. Such polypeptides can also be produced as a consequence of the normal processing of the beta -amyloid precursor protein (APP) and subsequent purification. The invention includes several assay systems for identifying inhibitors of the apoE4- activated Aβ interaction. In order to understand the significance of such assays it is first necessary to understand that active forms of Aβ are the physiologically relevant species involved in neurotoxicity and also apoE4 binding. It is also necessary to appreciate that the generation of such an active species follows a fourth order kinetic model which we describe. Use of the model allows prediction of the time when the active species reaches a maximum and therefore affords a predictive tool for determining when to harvest the active species for use in an assay to access cell toxicity or apoE4 interaction. The kinetic study of the spontaneous aggregation of Aβ is complicated by the dependency of the nucleation rate on a high power of the peptide concentration and by the extreme sensitivity of the system to trace amounts fibrils, any particulate matter, seeds, and nonsedimenting oligomers often pres&nt in commercial samples. Also, the presence of trace amounts of moisture in the lyophilized peptide often lead to uncontrollable premature nucleation. One of the best ways to achieve purity and homogeneity of the starting material is b y pretreatment with acid or solvents, such as hex-afluoroisopropanol or acetic acid. (Findeis 1999) In addition to rigorous sample purification, shaking of thie aggregation mixture at a constant speed appeared to be necessary in order to obtain rates consistent from day to day.

We found that the aggregation data were most consistent with a three-step kinetic model in which the unactivated monomer (UM) is first slowly converted into an active species (AS). In a second slow step, several monomers cooperatively form an oligomeric nucleus that serves as the growing site (GS) for the fibril. Fibril growth occurs in a third step where successive addition of inactive monomer molecules elongates the aggregate, without destroying the growing site.

UM—^→AS

xAS—^→GS

Fast

UM + GS—^ ^k →GS + P The model is expressed in the following system of equations: dUM

-kJJM ^■ k_pGS ZJM dt Eql

= kΛJM -k„AS dt

Eq_2

dGS k

= ^-.AS^> dt Eq3

dF*

— = kΛJM . GS dt ^p

Eq4 where P is the concentration of aggregated monomers. Since these -rate equations are nonintegrable in the closed form, the P vs. time data were analyzed using a nonlinear least squares program combined with a fifth-order Runge— Kutta numerical integration of the above equations. The analysis yielded zx = 4 + 1 and the rate constants calculated. The analysis converged to about the same best fit parameters regardless of the values chosen for the initial guesses. Once rate constants are calculated the time point at which any species (including the active Aβ peptide species) can easily be calculated. We describe below several methods of assessing the aggregation of Aβ necessary to calculate the relevant rate constants.

Example 1 Measurement of Spontaneous Aβ Aggregation Aβ 1-40 was purchased from Polypeptide Laboratories (Torrance, CA) and from Bachem Biochemica, (Heidelberg Germany AG). Fluo-β-amyloid 1-40, containing a cysteine residue in the sequence to which a fluorescein moiety was covalently linked, was purchased from Advanced Bioconcept, Ltd. (Beckinghamshire, England). Molecular Sieve (4A), and 1,1,1,3,3,3-Hexafluoro- 2-Propanol (HFJP ) were from Sigma Chemical (St. Louis, MO). Preparation ofA.0^'40 in monomeric form. The Aβ 1-40 to be used as an HPLC standard was treated with HFJ A, which had been dried at 4°C over Molecular Sieve Type 4A and then centrifuged at 15,000 x g for 15 minutes to remove molecular sieve dust. For aggregation assays, the peptide was dissolved to 8 mg/mL in the HFTPA and held in the solvent for 18 to 24 hours to disaggregate any preformed particles. Following this step, 20 μl aliquots of ttiis solution were flash-frozen in liquid nitrogen and stored at -195°C. Before use, the HFIPA was removed by direct lyophilization — without thawing — under day ice. Removal of the HFIPA without thawing was essential to achieve non-aggregated starting material, since melting of the frozen pellets prior to lyophilization often produced some oligomeric peptides. The lyophilized Aβ 1-40 was dissolved in anhydrous DMSO (usually 20 μL) and bath-sonicated for 15 minutes to yie-ld a lOOx stock solution. The HPLC chromatogram of the peptide prepared in this manner showed only one peak — most likely the monomer — and the CD structure was consistent with a predominantly random coil peptide. We found that the procedure in which a 20x stock is prepared in 0.1% acetic acid (IFindeis, M.A. et al. (1989)) also yielded acceptable peptide solutions. Spontaneous aggregation ofAβl-40.

Shaken aggregation assays were performed essentially as described previously (Findeis, M.A. et al. (1989)). Assays were performed in 96-well plates with a final assay volume of 250 μL of Dulbecco's PBS (w/o Ca^H or Mg^44") containing 50 μ-M Aβl-40. The plates were shaken using a titer plate shaker (Lab-Line Model 4623) at a rate of 80O rpm. When antibodies were included in the reaction mixture, they were added from concentrated solutio ns. Antibody concentrations were calculated assuming a molecular weight of 150,000. Measurement of Aggregation by Turbidity

At given times the turbidities of the reaction mixtures were read at 405 nm in a Molecular Devices Nmax Microplate Reader (Sunnyvale CA). Aggregation of 50 μM Aβl-40 was also monitored continuously using the vertically polarized light scattering intensity on an ISS K2 spectrofluorimeter with e-xcitation at 490 nm and emission at 530 nm with both polarizers set in the vertical position. The cuvette contained peptide in two mL of buffer and the solution was stirred buy a stir bar at a rate of 80 rpm. Indirect measurement of Aggregation HPLC analysis of the solution concentration of low molecular weight Aβ^1'40.

For HPLC analysis 60 μl aliquots were centrifuged at 15,000 x g for 15 minutes at 4°C. Fifty μL supernatant was transferred to silanized mi ro-autosampler vials for determination of Aβl-40 monomer remaining in solution following spontaneous aggregation. Two HPLC methods were utilized. The first employed a Waters Delta-pak C18 column (150 x 4.6 mm, 5 μm stationary phase, 300 A pore size) with a 20 μL injection volume. This method used a Waters Alliance HPLC system with UN detection (Waters 996 photodiode array detector) at 214 nm. A gradient was applied at room temperature with a flow rate of 1 mL/min from 7O)% water, 30% acetonitrile (both containing 0.1% v/v trifluoroacetic acid) to 60% water, 40% acetonitrile, and 0.1 *7o trifluoroacetic acid v/vJv linearly over 10 minutes. The column was re-equilibrated to the original composition for 6 minutes before the next sample injection. The second method utilized an isocratic mobile phase (73% water, 27% acetonitrile, 0.1% trifluoroacetic acid v/v/v) and a Phenornenex C18 Luna column (250 x 2.1 mm, 5 μm). This method used a Perkin Elmer Series 200 pump and autosampler equipped, with a Waters 996 photodiode array detector at 214 nm. The column was at 45°C and the flow rate was 0.4 mL/min. Fifteen μL of sample was injected; retention times for Aβ 1-4-0 monomer approximately 5 minutes in the first method and 6 minutes in the second. Quantitation of Aβ^1"40 monomer was calibrated using a standard Aβ 1-40 solution, about 100 μM in DM O, whose concentration had been determined b-y amino acid analysis. The standard was stored under liquid nitrogen. Concentrations were calculated using a 5-point standard curve weighted 1/x. The lower limit of quantitation was approximately 1.7 μM.

Measurement of Aggregation by Circular Dichroism (CL>) Spectroscopy.

The kinetics of conformational changes occurring in solutions of Aβ 1-4-0 were monitored by CD spectroscopy at two different concentrations of peptide, 50 and 8O μM in Dulbecco's PBS. Aliquots of 120 μL were removed at timed intervals from sequential wells of a 96-well plate on an orbital shaker and transfened to a cylindrical quartz CD cell with a pathleng^th of 0.5 cm. Spectra were acquired using a Jasco (Easton, MD) J-715 spectrophotometer at 23 °C. A. CD spectrum of the buffer was also collected at 23 °C. T ie CD response was calibrated with ammonium -d-camphor-10-sulfonate. Far-UN CD spectra were collected from 193-260 nm with a response of 0.25 seconds, scan speed of 1 00 nm/rnin, resolution and bandwidth of 1.0 nm and 8 cumulative scans. A lower wavelength limit was imposed by the CI^" absorption below 195 nm but ΝaCl was required for consistency with other kinetic experiments.

Background-subtracted CD spectra were imported into Grams/32 for conformational analysis using principal component regression analysis (Blow, D.M., (1994)). Secondary structure estimates of Aβ 1-40 were determined using a basis set of CD spectra for 16 different proteins of known secondary structure obtained from public databases. The method has been adapted to run on a commercially supplied software package (PLSplus/IQ version 3.02) obtained from Galactic Industries (Salern, ΝH). The basis set consisted of mean centered CD data in units of Δε (liters/rnol cm) from 195 to 260 nm at 2 nm resolution along with the X-ray crystallographic secondary structure percentages for -helix, parallel and antiparallel β-sheet, β-turn and other or random structure. Basis set data were mean centered by subtracting the average spectrum from each of trie library entries. Five factors provided the optimum agreement between experimental and known secondary structure percentages. To estimate the conformational percentages of Aβ 1-40 , the experimentally obtained CD spectra were averaged to 2 nm resolution, converted to units of Δε (liters/mol cm) and analyzed using the five most significant principal components. Measurement of Aggregation of Aβ by fluorescence polarization.

Attachment of the peptide to a growing fibril should change the rotational mobility of any fluorophore on the peptide and, thus, the polarization of fluorescence of a labeled amyloid peptide should be a sensitive probe to detect and quantitate aggregation. Accordingly, we used an Aβ 1-40 with a covalently attached fluorescein molecule (Advanced Bioconcepts), mixed with unlabeled Aβ^1"40 for detection of aggregation.

Aggregation of fTuorescein-labeled Aβ 1-40 was examined using a BMG Fluostar Galaxy plate reader with the polarization optics head in place. A 480 nm filter (12 nm bandpass) was used in the excitation optics and a pair of matcbred 520 nm filters (35 nm bandpass) was used in the two emission optical paths- Five replicate solutions of Aβ 1-40 at three different concentrations (10, 25, an L 50 μM) in PBS were placed in wells of a Corning black opaque 96-well plate (part # 3915). An instrument procedure was established to read each well using 50 flashes of the xenon flash lamp averaged per read with a positioning delay of 1 second. Instrument gain was set according to the Fluostar software manual. A polarization intensity reading was measured at time zero and the plate was removed and stirred on an orbital shaker. A.n external orbital shaker was used instead of the instrument's shaker since precise conditions were previously established for the aggregation assay using the external device. The plate was stirred for 7 minutes, removed, placed back: in the plate reader and another polarization reading recorded for each well. In this manner, the kinetics of aggregation were recorded over a period of 200 minutes. Results

We first analyzed the time dependency of a typical Aβ aggre ation monitored by turbidity at 405 nm. The shape of the curve for Aβ spontaneous aggregation, shown in Figure 1, is reminiscent of that proposed for linear aggregation of proteins (Ainsztein, A. M. ( 1994), Tobacman, L.S, (1983), Voter W. A., (1984), Cooper, J.A., (1983)) but the induction period is muchi longer than that observed for either actin or tubulin. This indicated that the starting peptide species itself is unable to initiate fibril formation and that aggregation must be preceded by the slow formation of an active monomeric derivative. However, unlike the cases of actin and tubulin where the activation step is much faster than the aggregation and can be treated as an independent kinetic process, the Aβ aggregation appears to proceed through an extremely slow activation followed by a relatively rapid aggregation. After experimenting with a variety of lάnetic models we arrived at the model described above.

The agreement of the experimental points with the theoretical curve calculated using the best-fit constants (Figure 1) shows the congruity of the model. In the same figure we also show the calculated concentration s of the various kinetically competent species. Note that at any time AM represents only a small fraction of the total peptide concentration (2.6 μM versus 50 μl in this experiment) and that AM is the only species whose concentration passes through a maximum during the reaction. Although the proportionality between turbidity and concentration in linear polymerization was already established for actin (Tobacman, L-.S. (1983).), we ascertained its validity for our system by comparing, in parallel experiments, the A₄₀₅ measurements with those obtained by HPLC for the loss of the low molecular weight starting material from solution. The results are shown, in Figure 2. Both data sets were consistent with tlie model described by equations 1-4 and analysis yielded comparable rate constants, shown in Table 1. Table 1

The kinetics of shaken aggregation measured by turbidity were also compared to those measured by CD detection. The results of parallel experiments are shown in Figure 3. Again, both data sets were consistent with our kinetic model, and the analysis yielded the rate constants shown in Table 1.

The CD spectra of Aβ^1"40 were recorded as a function o>f the time of aggregation and the results are shown in Figure 4. Principal component analysis of the CD spectra collected during the aggregation of Aβ^1"40 repealed a slight loss of disordered coil structure with a concomitant increase in antiparallel β-sheet structure. Conformational analysis also suggested a slight increase in β-turn structure. The time of appearance of β-turn structure corresponded to the initiation of the aggregation process which we also observed using IR for detection (data not shown). Fluorescence Polarization Detection. The peptide concentration dependence of shaking-induced aggregation of Aβ 1-40 labeled with fluorescein was determined as shown in Figure 5, right panel. Aggregation resulted in a two-fold increase in. the steady state anisotropy at maximal change without a change in the intensity. These data were also consistent with the kinetic model evidenced by the agreement between the experimental data points and the theoretical curves calculated using Equations 1-4 and the best-fit parameters given in Table 1. The consistency of the kinetic behaviors observed by four different experimental detection methods indicates that the only major species observable during the whole time course of the reaction are the low molecular weight starting material and the aggregates. Any other intermediate species must ^"be at such a low concentration as to be undetectable within experimental error. The Effect of Peptide Concentration on Aggregation. The effect of the starting peptide concentration on the kinetics of aggregations was determined by measuring simultaneously the turbidity and fluorescence anisotropy. The results from the turbidity measurements at two concentrations of peptide are shown in Figure 5, left panel. Nonlinear least squares analysis of thie data according to Equations 1-4 yielded rate constants, shown in Table 1, ttaiat are most likely independent of the starting Aβ 1-40 concentration. The independence of the rate constants constitutes the best support for the validity of the kinetic model. Although it is possible to calculate the time point at which the active species of Aβ becomes maximal, it is also possible to simply take aliquots of the spontaneously aggregating A(3 solution and utilize those aliquots for binding to apoE4. In our hands this occurs at approximately 60 minntes when starting from completely dissagregated monomer under the buffer conditions described above . Production of :apoE4 and Other Isoforn s of ApoE In order to perform the assays of the invention it is necessary to isolate an apoE4 binding partner for the activated Aβ polypeptide. .Recombinant production of apoE isoforms has been described (Morrow JA et al. (1999). We describe here a procedure suitable as a means of producing apoE4 as a fcinding partner. It will be appreciated that other means are also suitable, albeit less convenient. For example apoE4 from human serum or CSF might be purified and isolated as source of material. While Morrow et al. utilize recombirtant material produced in E. coli it will be appreciated that virtually any recombinant protein production systems are equally suitable. We also describe the production of other isoforms of apoE. Such other isoforms are suitable as control polypeptides because the assays we describe show a preferential specificity for lipidated apoE4 in its complex formation with activated Aβ.

In Example 2 below we report a method for high level production of human apo E isoforms. The method is essentially that described recently b;y Vlorrow et al. the only modification being the ultracentrifugation step to reduce the amount of contaminant in the lipoprotein E fraction. Using this method we routinely obtain material that is greater than 90% pure. If ultrapure material is required, we find that re-passage of the apo E over the Sephacryl column wόll produce material that is approximately 99% pure, ^"but at slightly reduced recovery.

Example 2

MATERIALS AND METHODS Cloning of apo E isoforms and construction of expression vector

Complementary DNA corresponding to mature hvxman apo E was amplified from spleen cDNA (Clontech Laboratories, Palo Alto, CA). The amplification reaction contained 10 mM dNTP, 2 ng spleen cDNA, IX Q-solution (Qiagen, Inc., Valencia, CA), 0.2 μM each primer:

Forward, 5'-CATTGGATCCAAGGTGGTGGA<JCAAGCGGTGGAG-3' <SEQ 3D NO:3)

Reverse, 5'-CGACTCGAGTCAGTGATTGTCGCTGGGCAC-3', (SEQ ID

NO:4) 5 U TaqPlus Precision polymerase and IX buffer Stratagene, La Jolla, CA.) in a final volume of 50 μl. The PCR conditions for a plification were 30 cycles of 94 °C for 1 min, 55 °C for 1 min, 72 °C for 1 min, followed by a final extension of 72 °C for 10 min. The amplification product was ligated into pcDNA3.1 (Invitrogen, Carlsbad, CA). Subsequent plasmid rniniprep DNA samples were sequenced by the DNA Sequencing Core Laboratory. Sequencing results revealed full length, mature apoE 2 and apoE3 cDNA clones. To generate apoE4 cD NA, site-directed mutagenesis, using the QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA), was conducted on apoE3 cDNA, changing the codon corresponding to amino acid residue 112 from TGC (cys) to CGC (arg). The mutagenesis reaction contained 50 ng pcD NA3.1-apoE3, 10 mM dNTP, IX Q- solution (Qiagen, Inc., Valencia, CA), 125 ng each primer:

Forward, 5'-GACATGGAGGACGTGCGCGGCCGCCTCTGGT 3C-3' (SEQ ID NO:5)

Reverse, 5'-GCACCAGAGGCGGCCGCGCACGTCCTCCATGXC-3', (SEQ ID NO:6

2.5 U Pfu polymerase and IX buffer in a total volume of 50 μl. Cycling parameters were 95 °C for 30 sec, followed by 12 cycles of 95 °C for 30 sec, 55 °C for 1 min, and 68 °C for 12 min. The reaction was cooled to 31 °C, DPN I (10 U) was added, and digestion of parental DIN A was conducted for 1 h at 37 °C. The mutated DNA (1.5 μl) was transformed into DH5αFT cells (Life Technologies, Rockville, MD). Sequencing results of subsequent miniprep plasmid DNA showed several clones containing cDNA coding for mature apoE4. The cDN for mature apoE2, apoE3, and apoE4 was then cloned into the

E. coli expression vector pET32a (Novagen, Inc., Madison, WI), which was modified as previously described (17), such that a thrombin cleavage site was introduced immediately 5' to the BamHI restriction site. This modification was accomplished in two steps using the QuickChange site-directed nrutagenesis kit (Stratagene, La Jolla, CA). First, the nucleotides CCA CGC, encoding pro and arg respectively, were inserted immediately 5' to the BamHI site and the primers used for this insertion were:

S'-CATGGCTGATATCCCACGCGGATCCGAATTCG-S' (SEQ ID NO:7)

5'-CGAATTCGGATCCGCGTGGGATATCAGCCATG-3^/ (SEO> ID NO:8) Second, the nucleotides CTG GTA, encoding leu and arg respectively, were inserted immediately 5' to the first group of nucleotides inserted. The primers used for this insertion were:

5'-CATGGCTGATATCCTGGTACCACGCGGATCCG-3' (SEQ ID NO:9)

5'-CGGATCCGCGTGGTACCAGGATATCAGCCATG-3' (SEQ ID NO: 10) The mutagenesis reactions for each, of the above insertions contained 10 ng pET32a plasmid, 10 mM dNTP, 125 ng each primer, 2.5 U Pfu polymerase and IX buffer in a total volume of 50 μl. Cycling parameters were 95 °C for 30 sec, followed by 18 cycles of 95 °C for 30 sec, 55 °C for 1 min, and 68 °C for 12 min. Each reaction was cooled to 37 °C, DPN I (10 U) was added, and the reaction incubated for 1 h at 37 °C. The mutated DNA (1.5 μl) was transformed into DH5αFT cells. Core sequencing subsequently identified several clones containing the correctly inserted thrombin recognition sequence. The cDNA for apoE2, apoE3, and apoE4 was then ligated into the modified pET32a vector at the BamHI and Xhol restriction sites.

Expression of human recombinant ApoE isoforms

ApoE2, apo E3 and apo E4 DNAs -were transformed into BL-21(DE3) cells. For scale up expression and purification, 100 ml LB medium containing 100 μg/ml ampicillin was inoculated with each clone and incubated at 37 °C in a shaker until the OD₆₀o reached 0.6 at which point the cells were chilled on ice and kept at 4 °C overnight. The next morning the cells were centrifuged at 4O00 rpm using an SLA-150O rotor for 5 min and the pellet was resuspended in 100 ml of fresh LB media containing 100 μg/ml ampicillin. Two liters of fresh LB media containing 100 μg/ml ampicillin was inoculated with 35 ml of the resusroended cells and incubated at 37 °C in a shaker until the OD₆o₀ reached 0.6. The cells were then induced for 2 hrs with IPTG (100 mg/ml final concentration) . The cells were again centrifuged and the resulting pellet stored at -80 °C uirtil needed.

Isolation and Purification of the ApoE isoforms

The cell pellet obtained as described above was resuspended in what will hereafter be termed suspension buffer (50 mM Tris-HCl, pH 8.0 and 0.5 M NaCl containing 10 mg/ml each of Aprotinin, Bestatin, leupeptin a_nd 1 M Benzamidine) and then lysed using a French Press. The pH of the resulting lysate was adjusted to 8.0 using 2 M Tris and the sample was centrifuged at 11000 x g for 50 min. The supernatant was removed and loaded onto 20 ml Ni-IMAC column that had been pre-equilibrated with suspension buffer. The column was then washed with suspension buffer until the OD₂₈o of the effluent declined below 0.8. At that point, suspension buffer that contained 75 mM imidazole was used to wash the column and that washing continued until the OD₂₈₀ of the effluent declined below 0.7. The column was then washed with 2 column volumes of imidazole-free suspension buffer, after which the ApoE fusion protein was eluted using suspension buffer that contained 300 mM imidazole. Fifty-drop fractions were collected and fractions with an OD₂₈o value greater than 1.8 were pooled and dialyzed in 20 mM NH₄HCθ₃ The protein concentration of the pool was determined using BCA assay and the samples stored at -80 °C until ready for thrombin cleavage.

Cleavage of apo E isoforms by thrombin Before the fusion protein coυld be cleaved by thrornfcήn, the naturally occurring thrombin cleavage site present within the apo E sequence needed to be protected. This was accomplished by incubation of the fusion pjrotein with small, unilarnellar vesicles of dimyristoyl phosphatidylcholine (DMPC) (Sigma, St. Louis, MO}. DMPC (120 mg) was dissolved in 3 ml of chloroform in a 50 ml glass conical tube. The lipid was dried on the walls of the tube under a stream of nitrogen.

When the lipid appeared dry, the tube was alternately heated in a 50 °C water bath and flushed with N₂ until the solvent was completely evaporated. After solvent removal, 12 ml of 20 mM ammonium bicarbonate was added and the tube was incubated in a 50 °C water bath for 1 hour, then placed in a sonicating water bath for lO minutes. This dispersed the phospholipid from the walls of the tube to produce a milky suspension of very large multilamellar vesicles (MLV). Small unilarnellar vesicles (SUN) were prepared by the so ication of the MLV dispersions using a Heat Systems sonicator Model XX-2020 (Misonix, Farmingdale, ΝY) operating at a power of 90 W for 45 minutes, in 3 minute bursts at 1.5 minute intervals, with probe immersion of 4 mm. The sonication vessel was cooled with ice water to prevent the mate-rial from overheating. After sonication, metal fragments (from the probe) were removed by centrifugation at 40,000 rpm for 20 minutes at 15 °C. The DMPC-SL7N were immediately added to the fusion protein preparation described above at a ratio of approximately 3.8/1 (lipid/protein, w/w) and allowed to incubate overniglit at 24 °C. Thrombin was then added to the lipid protein complex at a ratio of 1/1000 (thrombin/fusioo. protein, w/w) and the cleavage reaction allowed to proceed at room temperature for 30 minutes. The reaction was stopped by the addition of β-mercaptoethanol (final concentration 0.1%), the mixture adjusted to a density of 1.225 with solid KBr and the lipoprotein fraction isolated by centrifuging for 44 h at 55,000 rpm using a TLS-55 rotor (Beckman Instruments Inc., Palo Alto, CA). The top 0.3 ml was removed by slicing the tube and the li oprotein fraction so obtained was dialyzed against 20 mM ammonium bicarbonate and lyophilized.

Purification of apo E by FPLC

The lyophilized lipoprotein was delipidated using 2:1 chloroform: methanol. The protein pellet was solubilized overnight in 6M guanidine-HCl, O.l Vl tris, pH 7.4 buffer containing 0.01% EDTA and 1% β-rnercaptoethanol, the protein solution was filtered through a 0.45 micron Millipore filter and applied to t>vo 2.6 x 60 cm HiPrep Sephacryl S -300 columns (Amersham Pharmacia Biotech, Piscataway, NJ) connected in tandem (bottom to bottom) wliich had been equilibrated with 4M guanidine- HC1, 0.1M tris, pH 7.4 buffer containing 0.01% EDTA and 0.1% β- mercaptoethanol. The column chromatography was managed by an FPLC system (Amersham Pharmacia Biotech). The sample was injected onto the column at 12 ml/hr and the column flow rate was 30 rnl/hir. The absorbance of the column effluent was monitored at 280 nm. The fractions making up a given absorbance peak were individually dialyzed against 10O mM ammonium bicarbonate, the protein concentration measured using the J3 CA method (Pierce, Rockford, IL) and the purity of the fractions determined by SDS-PAGE and Western analysis.

SDS-PAGE analysis was conducted using No vex NuPAGE 4-12 % Bis-Tris gels in MES running buffer. Gels were run at 2O0 volts for 35 min. Gels were fixed and stained using the Novex Colloidal Blue Staining Kit. For Western analysis, proteins were transferred to nitrocellulose (0.2 μm pore size; Novex, Carlsbad, CA) in NuPAGE transfer buffer. The transfer was run at 30 volts for 1 h. The membrane was placed in blocking buffer (50 mM Tris, pH 8.0, 2mM CaCl₂, 80 mM NaCl, 5 % nonfat dry milk, 0.2 % Nonidet P-40, and 0.01 % antifoam A) for 1 h at room temperature with shaking. The membrane was incubated with goat anti-cynomolgus apo E (12) (1:1000 dilution in blocking buffer) for 1 h at room temperature. The membrane was washed ttiree times with blocking buffer for 15, 5, and 5 min, consecutively and then incubated in the presence of ORP-labeled mouse anti-goat IgG (1 :5000 dilution in blocking buffer) for 50 min. The membrane was washed 4 times with blocking buffer for 15, 5, 5, and 5 min consecutively and then immersed in the presence of ECL detection reagent (Amersham Pharmacia Biotech) for 1 min before exposure to Hyperfilm ECL (Amersham Pharmacia Biotech).

Lipidation of Purified ApoE Purified apoE isoforms were obtained as described above. Lipidation of apoE was prepared by incubation of purified apoE isoforms with small, unilarnellar vesicles of l-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) (Avanti Polar Lipids, Catalog No. 850457). POPC (120 mg) was dissolved in 3 ml of chloroform in a 40 ml glass conical tube. The lipid was dried on the walls of the tube under a stream of nitrogen. When lipid appeared dry, the uxbe was heated in a 50°C water bath and flushed with N₂ until the solvent was completely evaporated. After solvent removal, 12 ml of 20 mM NFLΗCOs was added and the tube was incubated in a 50°C water bath for 1 hr, then placed in a sonicating water bath for 10-min. This incubation procedure was then repeated one time. This dispersed the phospholipid from the walls of the tube to produce a milky suspension of very large multilamellar vesicles (MLV). Small unilarnellar vesicles (SUV) were prepared by the sonication of the MLV dispersions using a Heat Systems sonicator Model XL-202O (Misonix), operating at a power of 90 W for 45 min, in 3 min bursts at 1.5 min intervals, with probe immersion of 4 mm. The sonication vessel was cooled with ice water to prevent the material from overheating. After sonication, metal fragments (from the probe) were removed by centrifugation at 40,000 rpm for 20 min at 15°C. The POPC-SIXN was immediately added to apoE or bovine serum albumin as a control (Sigma Catalog No. A7030) at a ratio of approximately 3.8/1 (lipid/protein, w ) and allowed to incubate for 17.5 hrs at 24°C. Samples w^{^}ere stored at 4°C and utilized 24 hrs later for interaction studies with Aβl-40.

Example 3

Isolation of ApoE from Conditioned Cell Media. Media containing human apoE was isolated from human astrocytoma cells that were stably transfected with human apoE2, apoE3, or apoE4 pcDNTA selected under neomycin resistance. Serum-free media was collected 3 days after medium replacement when cells were 50% and 80% confluent, and media was supplemented with protease inhibitors to prevent protein degradation (Complete Protease Inhibitor Cocktail Tablets, 1 tablet/50 ml media, Roche). Samples were s-tored at -20°C. Large-scale concentration of media was performed with Centricon Plus-80 centrifugal filters (Amicon Inc.) with a MW cutoff of 10-kDa. Due to low levels of apoE in conditioned media, apoE2 media was additionally concentrated with small-scale Centricon YM-10 centrifugal filters (Amicon Inc.) also with a MW cutoff of 10- kDa. ApoE3 and apoE4 media was concentrated between 30 to 50 fold and apoE-2 approximately 100 fold. Untransfected astrocytoma cells conditioned media was also concentrated in the same manner to be used for background determinations. Media was stored at -20°C until needed for Lipid Flotation Ultracentrifugation.

Conditioned Media Lipid Flotation by Ultracentrifugation. Ultracentrifugation procedures for obtaining lipidated apoE in conditioned media were performed using a modified protocol from Havel (1987). Concentrated conditioned media was adjusted to a density of 1.225 with solid KBr (0.7 6 g solid KBr was added to 2 ml of media, and 2.3 ml of .NaCl salt solution (1.225 g/ml) was added to advance height of the liquid). The lipoprotein fraction Λvas isolated by centrifuging for 24 hrs at 15°C at 55,000 rpm using a TI.S-55 rotor (Beckman Instruments). The top 0.3 ml was removed by slicing ttie tube with a Centritube sheer (Beckman Instruments) . The top layer containing lipid-associated apoE particles and the bottom layer containing unlipidated apoE were dialyzed agains-t DPBS using Slide- A-Lyzer Dialysis Cassettes with a MCW cutoff of 10-kDa (Pierce). We found that it was preferred that tissue culture derived apoE4 was used fresh and that activity was lost if the material was frozen. Assays Assessing the Association of activated beta amyloid peptide species with an lipidated apoE4 derived polypeptide

As noted above it is believed that apoE is invol"ved in deposition or clearance of Aβ by direct protein-to-protein interaction. A direct sandwich ELISA detected the interaction of the three predominant apoE isoforms (apoE2, apoE3, apoE4) with Aβ 1-40 obtained during different phases of spontaneous aggregation. However purified delipidated apoE had binding interactions only with O-min and an intermediate aggregated form of Aβ, and no isoform-specific differences were seen. However, when apoE was associated with biological lipid palmitoyl oleoyl phosphatidylcholine, (the biologically active form of apoE), apoE4 bound to the intermediate active Aβ with higher avidity than apoE-2 and apoE3. This interaction was detected with ELLS A capture apoE antibodies which have epitopes for the INT-terminus of apoE. An absence of detectable binding was demonstrated when a C-terminal apoE antibody was used for capturing. s noted below these (in light of the known binding specificities of the antibodies nsed) indicate that apoE4 interaction with the active form of Aβ resides at around amino acid residues 243 to 272 of apoE4.

The differential risk of AD related to a_poE genotype may be the result of enhanced capacity of apoE4 binding to an intermediate active form of A . As such, disruptions of this interaction may reduce brain Aβ plaque load in AD susceptible apoE4 carriers. Abolishment of the interaction between apoF_C4 and Aβ would provide an alternative therapeutic approach for the treatment of AD, and may be accomplished by drug intervention or immunization to apoE4. Example 4

ApoE Complex Formation with Aggregated Forms ofAβ 1-40. Binding reactions were carried out similar to concentration ratios of peptides performed by XaDu. Both purified and POPC lipidated isoforms of apoE were diluted to 10O nM with DPBS (Modified Dulbecco's Phosphate Buffered Saline (0.015M sodium phosphate, 0.003 M potassium phosphate, 0.29 M sodium chloride, 0.01 Tvl potassium chloride, pH 7.4), and 500 μl of eacli preparation were placed into a separate 14 ml sterile polystyrene round bottom tube (Falcon). Control conditions utilized purified apoE vehicle (20 mM NELLICO^ and BSA P^»OPC both diluted in DPBS . The contents from 14 to 18 wells of aggregating A ? were pooled each from the following intervals of aggregation: 0 min, 60 min, 80 min, and 110 min. The A ?(500 μl) was immediately added to apoE and controls giving a final concentration of 25 μM Aβ l-4O> and 50 nM apoE in the incubation mixture. Mixtures were gently rocked for 2 hrs at room temperature witti a Nutator (Clay Adams). Samples were immediately added to the direct sandwich ELISA for detection of complex formation.

Detection of ApoE Complex Fortnation with aggregated Forms ofAβ 1-40 Utilizing a Direct Sandwich ELISA. The interaction of lipidated apoE with A ? l- 40 was evaluated by a liquid-phase direct sandwich ELISA. The interactions were characterized with different anti-apoE antibodies. Antibodies recognizing the N- terminus of the 299 amino acid residues of apoE were mouse monoclonal antibody 6C5 (Ottawa Heart Institute Research Corp., Ottawa, Ontario), which recognizes residues 1-15, and mouse monoclonal antibody 9H8.G5.F2 (Biodesign International), which recognizes an unspecified epitope within the N-terminal domain. For C-terminal recognition, mouse monoclonal antibody 3H1 (Ottawa Heart Institute Research Corp., Ottawa Ontario) associates with, an epitope within residues 243-272. ELISA 1/2 well flat-Dottom 96-well plates (Corning) were coated with one of three mouse monoclonal antibodies (6C5 [2 μg/ml], 9H8.G5.F2 [2 μg/ml], and 3H1 [1 μg/ml]) to capture apoE; 50 μl/well. All antibodies were prepared in 100 mM NΣT HCO₃ and were incubated overnight at 4°C. Plates were blocked overnight at 4-°C with 1% BSA in DPJBS; 100 μl/well. ApoE and vehicle control reactions with, aggregated forms of Aβ 1-40 were added to anti-apoE antibody coated ELISA plates after complex reactions were completed, and the plates were incubated overnight at 4°C; 50 μil/well. Bound Aβ 1-40 was detected with biotinylated mouse monoclonal anti-A ? antibody 4G8 (0.5 μg/ml, recognizes residues 17-24 of A ? 1-40, Signet) diluted with 1% BSA in DPBS incubated for 2 hrs at room temperature; 50 μl well. Neutravidin conjugated to horseradish peroxidase (HRP, [0.1 μg/ml], Pierce) was reacted with biotin for 60 min; 50 μl/well. Addition of 50 μl/well TMB substrate (Kirkegaard & Perry) for 2 to 5 min produced a colored end-product when converted by HRP, and the reaction was stopped using 1 M H₃PO₄; 25 μl/well. The colored end-point was detected at an absorbance of 450 nnα using a Molecular Devices Nmax plate reader. Prior to addition of new layers in the ELISA, wells were washed with BupH Modified Dulbecco's PBS (Pierce) using a Skatron automated plate washer. Each treatment group was performed in triplicate wells and experiments were repeated three to four times for nonlipidated apoE and four to five times for lipid- associated apoE.

Binding Interactions of Nonlipidated and Lipid-Associated ApoE Isoforms with Different Aggregated Forms of Aβ 1-4O. Aliquots of Aβ 1-40 taken at different time points during the aggregation process were assessed for binding to the isoforms of apoE. The concentration of only one of these forms, AS, was shown to pass through a maximum during the full time course of the aggregation. Interactions of apoE's with Aβ 1-40 were observed with the O-min starting form and 60-min aggregated Aβ 1-40. ApoE N-terminal antibodies 6C5 ancl 9H8.G5.F2 detected isoform-specific differences of lipidated apoE binding to the 60-min aggregated Aβ aliquot (Figure 6A and 6C). These antibodies detected! high affinity apoE4 binding to the 60 min mixture that vere 2.2 to 3.3 fold greate-r than apoE2 and apoE3 (p<O.05). These isoform-specific differences were abolished when nonlipidated apoE isoforms were interacted with the 60-min aggregated Aβ (p>0.05) (Figure 6B and 6D). In contrast, apoE C-terminaT antibody 3H1 detected negligible interactions of lipid-associated apoE binding to the 60-min aggregated Aβ 1-40 (Figure 7A). These observations were also demonstrated with nonlipidated apoE (Figure 7B) in "which no isoform-specific interactions were observed. Interestingly, nonlipidated apoE binding to the 60- min aggregated form of Aβ 1-40 showed a trend of higher apoE2 binding than apoE3 and apoE4, utilizing 6C5, 9H8.G5.F2, and 3H1 antibodies (Figures 7B, 7D., and 7B). However, these differences were not significant (p>0.05). Also of note., 6C5 detected both lipid-associated and nonlipidated apoE interactions with the 0— min form of Aβ 1-40 (Figure 6 A and 6B), with no apoE isoform-specific differences observed. Incidentally, a similar magnitude of binding between., lipidated and nonlipidated apoE isoforms was demonstrated (p>0.05). lipidated apoE binding to the 0-min form of Aβ 1-40 was also detected with 3H1 (Figure 7A), however with a slightly lower magnitude than 6C5 detection. Little or no interactions of 0-min Aβ 1-40 were detected with nonlipidated apoE isoforms utilizing the 3H1 antibody (Figure 7B).

Identifying Inhibitors of the lipidated ApoE4-activated Aβ Interaction

Although the example above describes an ELISA. assay system a multitude of others systems are suitable for measuring this interaction. For this purpose an easy detection system should be available for at least one of the partners in the complex; such a detection system could be based on antibodies, or on labeling ones of the proteins with a marker molecule or radioactivity. The subsequent use of this assay would be to screen a compound collection for substances that would modulate the interaction between ApoE4-activated Aβ.

In solution assays, methods of the invention coroprise the steps of (a) contacting a ApoE4 or Aβ polypeptide with one or more candidate inhibitor compounds and (b) identifying the compounds that decrease the ApoE4 -Aβ complex. Agents that decrease the association. The selectivity of a corrrpound that modulates the propensity of ApoE4-activated Aβ to form a complex can be evaluated by comparing its effects on other protein protein interactions. Inhbitors of ApoE4-activated Aβ complex formation will be therapeutically useful in treatment of diseases and physiological conditions in which normal or aberrant amyloid deposition is involved.

Assays that may prove useful for measuring the ApoE4-activated - β interaction are well known in the art including, but not limited to: equilibrium or membrane flow dialysis, antibody binding assays, gel-shift assays, in vitro binding assays, filter binding assays, enzyme-linked iπumunoabsorbent assays (ELTSA), western blots, co-immunoprecipitation, immuriogold co- immunoprecipitation, coimmunolocalization, co-crystallization, fluorescence energy transfer, competition binding assays, chemical crosslink ng, and affinity purification. Compounds, including antibodies, that effect the ApoE4-activatecl Aβ interaction are useful therepeutics for the treatment of disease, including following pathologies: Alzheimer's disease, Parkinson's disease, multiple sclerosis, Huntington's disease, amylotrophic lateral sclerosis, head injury damage, Picks Disease, frontal lobe dementia, cerebellar degeneration, ischemia reperfusion injury, stroke, ischemic injury and schizophrenia EXAMPLE 5

In Vitro Binding Assays For Detecting Inhbitors Assays which monitor ApoE4-activated Aβ complex formation are of value in screening for inhibitors of the interaction. Binding assays theoretically ta ke one of two forms: Aβ polypeptide(s) can be used to bind ApoE4 polypeptide(s), or conversely, ApoE4 polypeptide(s) can be used to bind Aβ polypeptides. In each case, the other polypeptide is contacted with thie immobilized polypeptide under conditions that permit specific binding of the polypeptides(s) to form a complex in the absence of added agent. Particular aqueous conditions may be select ed by the practitioner according to conventional methods. For general guidance Modified Dulbecco's Phosphate Buffered Saline (0.015M sodium phosprrate, 0.003 M potassium phosphate, 0.29 M sodium, chloride, 0.01 M potassium chloride, pH 7.4)) with optional addition of div lent cation(s) and/or metal chelators and/or nonionic detergents and/or membrane fractions. Additions, deletions, modifications (such as pH) and substitutions (such as KC1 substituting for NaCl or buffer substitution) may b>e made to these basic conditions. Modifications can be made to the basic binding reaction conditions so long as specific binding of Aβ to ApoE4 occurs in the control reaction<s). In such reactions, at least one polypeptide species typically is labeled with a detectable marker. Suitable labeling includes, but is not limited to, radiolabeling by incorporation of a radiolabeled amino acid (e.g., ¹⁴C-labeled leucine, ³H-labeled glycine, ³⁵S-labeled methionine), radiolabeling by post-translational radioiodination with ^{1 5}I or ¹³¹I (e.g., Bolton-Hunter reaction and chloramine T), labeling by post-translational phosphorylation with ³²P (e.g., ptiosphorylase and inorganic radiolabeled phosphate) fluorescent labeling by incorporation of a fluorescent label (e.g., fluorescein or rhodamine), or labeling by other conventional methods known in the art. In embodiments where one of the polypeptide species is immobilized by linkage to a substrate, the other polypeptide is generally labeled with a detectable marker.

Labeled polypeptide(s) are contacted with immobilized polypeptide(s) under aqueous conditions as describee! herein. The time and temperature of incubation of a binding reaction is optionally varied, with the selected conditions permitting specific binding to occur in a control reaction where no agent is present. Preferable embodiments employ a reaction temperature of about at least 15 degrees Centigrade, more preferably 30 to 42 degrees Centigrade, and a time of incubation of approximately at least 1.5 seconds, although longer incubation periods, from 30 seconds to a minute to several minutes or more, are preferable so that, in some embodiments, a binding equilibrium is attained. Binding kinetics and the thermodynamic stability of bound ApoE4-activated Aβ complexes determine the latitude available for varying the time, temperature, salt, pH, and other reaction conditions. However, for any particular embodiment, desired binding reaction conditions can be calibrated readily by the practitioner using conventional methods in the art, whicli may include binding analysis using Scatchard analysis, Hill analysis, and other standard analytic methods (Proteins, Structures and Molecular Principles, 1984) Creighton (ed.), W . H. Freeman and Company, New York). Specific binding of labeled Aβ or ApoE4 polypeptide to immobilized Aβ or ApoE4, respectively, is determined by including unlabeled competitor protein(s) (e.g., albumin). Similarly, specific binding of labeled Aβ or ApoE4, respectively, is determined by including unlabeled competitor protein(s) (e.g., albumin). After a binding reaction is completed, labeled polypeptide(s) specifically bound to immobilized polypeptide is detected. For example and not by way of limitation, after a suitable incubation period for binding, the aqueous phase containing non-immobilized protein is removed and the substrate containing the immobilized polypeptide species and any labeled protein bound to it is washed with a suitable buffer, optionally containing unlabeled blocking agent (s), and the wash buffer(s) removed. After washing, the amount of detectable label remaining specifically bound to the immobilized polypeptide is determined (e.g., by optical, enzymatic, autoradiographic, or other radiochemical methods). In some embodiments, addition of unlabeled blocking agents that inhibit non-specific binding are included. Examples of such blocking agents include, but are not limited to, the following: calf thymus DNTA, salmon sperm DNA, yeast RNA, mixed sequence (random or pseudorandom sequence) oligonucleotides of various lengths, bovine serum albumin, nonionic detergents (NP-40, Tween, Triton X-1OO, etc.), nonfat dry milk; proteins, Denhardt's reagent, polyvinylpyrrolidone, Ficoll, and other blocldng agents. Practitioners may, in their discretion, select blocking agents at suitable concentrations to be included in binding assays; liowever, reaction conditions are selected so as to permit specific binding between a ApoE4- activated Aβ in a control binding reaction. Blocking agents are included to inhibit nonspecific binding of labeled protein to immobilized protein and/or to inhibit nonspecific binding of labeled polypeptide to the immobilization substrate. It will be appreciated that other isoforms of apoE can be used in control reactions to assess the specificity of the reaction. In embodiments where a polypeptide is immobilized, covalent or noncovalent linkage to a substrate may be used. Covalent linkage chemistries include, but are not limited to, well-characterized methods known in the art (Kadonaga and Tjian (1986) Proc. Natl. Acad. Sci. (U.S.A.) 83: 5889). One example, not for limitation, is covalent linkage to a substrate derivatized with cyanogen bromide (such as CNBr-derivatized Sepharose 4B). It ma;y be desirable to use a spacer to reduce potential steric hindrance from the substrate. Noncovalent bonding of proteins to a substrate include, but are not limited to, bonding of the protein to a charged surface (e.g., on a bead) and binding with specific antibodies.

In one class of embodiments, parallel binding reactions are conducted, wherein one set of reactions serves as control and at least one other set of reactions include various quantities of agents, mixtures of agents, of biological extracts, that are being tested for the capacity to inhibit binding or formation of an ApoE4-activated Aβ complex. Agents which, when added to a binding reaction, inhibit formation of ApoE4-activated Aβ complexes are thereby identified as inhibitors;

In a preferred embodiment, several binding reactions are monitored simultaneously, e.g., using a format whicb permits simultaneous analysis of several samples (rnicrotiter plates, etc.). In a preferred embodiment, the assays are automated, e.g., using robotics for pipetting samples into rnicrotiter plates.

One means for detecting binding of a ApoE4or Aβ to its partner is to immobilize either the lipidated ApoE4 derived polypeptide or the Avβ polypeptide, such as by covalent or noncovalent chemical linkage to a solid support, and to contact either the immobilized lipidated .ApoE4 or Aβ polypeptide with lipidated ApoE4 or Aβ (as appropriate) that has been labeled with a detectable marker (e.g., by incorporation of radiolabeled amino acid, by epitope tagging and reporting with a fluorescent-labelled anti-epitope tag antibody, and the like). Such contacting is typically performed in aqueous conditions which permit binding of a ApoE4 to Aβ. Binding of the labeled component to the immobilized component is measured by determining the extent to ^which the labeled component is immobilized as a result of a specific binding interaction. Such specific binding may be reversible, or may be optionally irreversible if a cross-linking agent is added in appropriate experimental conditions.

Agents that inhibit the formation of bound complexes as compared to a control binding reaction lacking agent are thereby identified as inhibitors and are candidate therapeutic agents. In one embodiment, the physical interaction of the bound labeled complex with the surface is reported, such as where the surface is a. fluor or scintillant and the label in the bound labeled complex emits radiation suitable for activating the fluor or scintillant of the surface; light emitted from the surface reports the relative amount of bound labeled complex. A suitable system is the scintillation proximity assay (Amersham), wherein the unlabeled component is bound to a fluor-containing bead. Alternative systems include the "Flash Plate" system (LKB).

In a variation, immobilization is not required; either the lipidated ApoE4 or the Aβ is labeled with a first fluor which absorbs radiation (particle or wave) and emits phosphorescent or fluorescent light at a first wavelength, the other component is labeled with a second fluor which absorbs radiation at said first wavelength and thereby emits fluorescent or phosphorescent radiation at a second wavelength. The labeled lipidated ApoE4 or Aβ and the appropriate binding partner are incubated under suitable binding conditions, and at suitable reactant concentrations whereby the amount of radiation of the second wavelength is approximately proportional to trie amount of ApoE4-activated Aβ complex, and excited with radiation of the firstwavelength (or particle type) and the amount of emitted radiation of the second wavelength is detected. Trie relative amount of " radiation of the second wavelength reports the relative amount of lipidated ApoE4 -activated Aβ complex. An example of suitable system is a dye-dye transfer system (Packard)

It should be recognized, of course, that these assays are mentioned by way of example only and that these methods might be modified or that other methods of assaying the ApoE4-activated Aβ interaction would be apparent to one skilled in the art.

Tbe invention also comprehends high throughput screening (HTS) assays to identify compounds inhibit the lipidated ApoE4 -activa-ted Aβ interaction. HTS assays permit screening of large numbers of compounds in an efficient manner. Cell-based HTS systems are contemplated to investigate apoE4-activated Aβ interactions. HTS assays are designed to identify "hits" or "lead compounds" having the desired property, from which modifications can be designed to improve the desired property. Typical examples of therapeutic agents discovered by the assays of the invention and described herein include, but are not limited to, defective (either engineered or naturally occurring) forms of the proteins that associate with the protein complexes, antibodies, inhibitory fragments of the proteins, wild type and altered genes that code for proteins that disrupt disrupt the association, small organic molecules, antisense nucleic acid sequences, oligonucleotides that inhibit expression or activity via a triplex mechanism, peptides, aptameric oligonucleotides, and the like. More particularly, examples of engineered proteins may include, but are not limited to, proteins that comprise inactivating mutations in conserved active sites (e.g., ATP binding motifs, DNA or protein binding domains, catalytic sites, etc.), fusion protei s that comprise at least one inhibitory domain, and the lik:e.

The above agents may be obtained from a wide variety of sources. For example, standard methods of organic syntriesis may be used to generate small organic molecules that specifically disrupt t ie relevant protein protein interactions. In addition, combinatorial libraries comprising a vast number of compounds (organic, peptide, or nucleic acid, reviewed in Gallop et al. 1994, J. Med. Chem. 37(9): 1233-1251; Gordon et al ., 1994, J. Med. Chem. 37(10):1385- 1401 ; and U.S. Pat. No. 5,424,186 all of wnich are herein incorporated by reference) may be screened for the ability to bind and inhibit protein protein interactions involved in ApoE4-activated Aβ association.

Additional features and variations of the invention will be apparent to those skilled in the art from the entirety of this application, including the detailed description, and all such features are intended as aspects of the invention. Likewise, features of the invention described herein can be re-combined into additional embodiments that also are intended as aspects of the invention, irrespective of whether the combination of features is specifically mentioned above as an aspect or embodiment of the invention. Also, only such limitations which are described herein as critical to the invention should be viewed as such; variations of the invention lacking limitations which have not been described herein as critical are intended as aspects of the invention. It will be clear that the invention may be practiced otherwise than as particularly described in the foregoing description and examples. Numerous modifications and variations of the present invention are possible in light of the above teachings and, therefore, are within the scope of tb_e invention.

Claims

What is claimed is:

1. A method of identifying a lipidated ApoE4 derived polypeptide-activated Aβ complex formation inhibitor compound comprising the steps of : (a) contacting a lipidated apoE4 derived polypeptide with an activated Aβ polypeptide in presence of a test compound; and

(b) identifying test compounds that decrease the formation of the ApoE4 - activated Aβ complex wherein a test compound which decreases the formation of the lipidated ApoE4 derived polypeptide- activated Aβ complex is complex formation inhibitor compounds.

2. A method according to claim 1, wherein the lipidated apoE4 derived polypeptide is a recombinant polypeptide isolated from a cell transformed or transfected with a polynucleotide comprising a nucleotide sequence that encodes the polypeptide.

3. A method according to claim 2 wherein the cell is E. coli.

4. A method according to claim 2 wtierein the cell is a human astrocytoma cell.

5. A method according to claim 1, whierein the activated Aβ polypeptide is a recombinant polypeptide purified and isolated from a cell transformed or transfected with a polynucleotide comprising a nucleotide sequence that encodes the polypeptide.

6. A method according to claim 1, wherein the Aβ polypeptide is synthesized.

7. A method according to claim 1 wherein the test compound is an antibody.

8. A method according to claim 7 wherein the test compound is a humanized antibody

. A method according to claim 1 the wherein the lipidated apoE4 derived polypeptide comprises residues 112-272 of SEQ ID NO:l

10. A method according to claim 1 the wherein the lipidated aρoE4 derived polypeptide comprises SEQ ID NO:l

11. A method of identifying ApoE4-activated Aβ complex formation inhibitor compounds comprising the steps of : (a) generating an activated Aβ polypeptide; and

(b) contacting said activated Aβ polypeptide witti an lipidated ApoE4 derived polypeptide in trie presence of a test co pound; and

(c) identifying compounds that decrease the for ation of an lipidated ApoE4 - activated Aβ complex wherein a test compound which decreases the formation of the lipidated ApoE4 derived polypeptide- activated Aβ complex is complex formation inhibitor compounds.

12. A method according to claim 11, wherein the lipidated apoE4 derived polypeptide is a recombinant polypeptide isolated from a cell transformed or transfected with a polynucleotide comprising a nucleotide sequence that encodes the polypeptide.

13. A method according to claim 12 wherein the cell is E. coli.

14. A method according to claim 12 wherein the cell is a human astrocytoma cell.

15. A method according to claim 11, wherein the activated Aβ polypeptide is a recombinant polypeptide purified and isolated from a cell transform or transfected with a polynucleotide comprising a nucleotide sequence that encodes the polypeptide.

16. A method according to claim 11, wherein the Aβ polypeptide is synthesized.

17. A method according to claim 1 1 wherein the test compound is an antibody.

18. A method according to claim 17 wherein the test compound is a humanized antibody

19. A method according to claim 11 tr e wherein the lipidated ap oE4 derived polypeptide comprises residues 112-272 of SEQ ID NO:l

20. A method according to claim 11 tbe wherein the lipidated ap»oE4 derived polypeptide comprises SEQ ID NO : 1

21. A method of identifying a lipidated ApoE4 derived polypeptide - activated Aβ complex formation inhibitor compound comprising the steps of :

(a) contacting a lipidated ApoE4 derived polypeptide with an activated Aβ polypeptide in presence of a test compound;

(i) in the presence of a test agent, and (ii) in the absence of said test agent,

(b) determining the amount of apoE-4-activated Aβ complex formed

(i) in the presence of a test agent, and (ii) in the absence of said test agent; and

(c) comparing the amount of lipidated ApoE4 derived polypeptide - activated Aβ complex formed at corresponding times, with, and without the test agent, and selecting a test agent that decreases the amount of apoE4-activated Aβ complex formed.

22. A method according to claim 21, wherein the lipidated apoE4 derived polypeptide is a recombinant polypeptide isolated from a cell transformed or transfected with a polynucleotide comprising a nucleotide sequence triat encodes the polypeptide.

23. A method according to claim 22 wherein the cell is E. coli.

24. A method according to claim 22 wherein the cell is a human astrocytoma cell.

25. A method according to claim 21, wherein the activated Aβ polypeptide is a recombinant polypeptide purified and isolated from a cell transformed or transfected with a polynucleotide comprising a nucleotide sequence tt at encodes the polypeptide.

26. A method according to claim 21, wherein the Aβ polypeptide is synthesized.

27. A method' according to claim 21 wherein the test compound is an antibody.

28. A method according to claim 27 wherein the test compound is a humanized antibody

29. A method according to claim 21 the wherein the lipidated apoE4 derived polypeptide comprises residues 112-272 of SEQ 3D NO:l

30. A method according to claim 21 the wherein the lipidated apoE4 derived polypeptide comprises SEQ ID NO:l