CA2174632C - Methods of screening for beta-amyloid peptide production inhibitors - Google Patents

Abstract

Processing of .beta.-amyloid precursor protein (.beta.APP) is monitored by d etecting the secretion of a soluble amino-terminal fragm ent or .beta.APP (ATF-.beta.APP) resulting from cleavage of .beta.APP at the amino- terminus of .beta.-amyloid peptide. Secretion of ATF-.be ta.APP in animal models may be monitored to identify inhibitors of .beta.-amyloid production. The ATF-.beta.APP may be detected using antibodies an d other specific binding substances which recognize a carboxy terminal residue on th e fragment. Animals expressing the Swedish mutation of .beta.APP are described which produce abundant amounts of ATF-.beta.APP.</SD OAB>

Description

METHODS OF SCREENING FOR
BETA-AMYLOID PEPTIDE PRODUCTION INHIBITORS
BACKGROUND OF THE INVENTION
1. Field of the Invention The present invention relates generally to methods and compositions for monitoring the processing of (3-amyloid precursor protein. More particularly, the present invention relates to the use of such methods and compositions for the diagnosis, prognosis, and monitoring response to therapy of Alzheimer's disease, and for screening and evaluation of potential drugs for the treatment of Alzheimer's disease.
Alzheimer's disease is characterized by the presence of numerous amyloid plaques and neurofibrillary tangles (highly insoluble protein aggregates) present in the brains of Alzheimer's disease patients, particularly in those regions involved with memory and cognition. While in the past there was significant scientific debate over whether the plaques and tangles are a cause or are merely the result of Alzheimer's disease, recent discoveries indicate that amyloid plaque is a causative precursor or factor. In particular, it has been discovered that the production of /3-amyloid peptide, a major constituent of the amyloid plaque, can result from mutations in the gene encoding amyloid precursor protein, a protein which when normally processed will not produce the ~3-amyloid peptide.
It is presently believed that a normal (non-pathogenic) processing of the ~3-amyloid precursor protein occurs via cleavage by a putative "a-secretase" which cleaves between amino acids 16 and 17 of the protein. It is further believed that pathogenic processing occurs via a putative "/3-secretase"
at the amino-terminus of the /3-amyloid peptide within the precursor protein.
The identification of mutations in the amyloid a 35 precursor protein gene which cause familial, early onset Alzheimer's disease is the strongest evidence that amyloid metabolism is the central event in the pathogenic process underlying the disease. Four reported disease-causing 2 i l4.<~32 2 mutations include with respect to the 770 isoform, valine717 to isoleucine (Goate et al. (1991) Nature 349:704-706), valine717 to glycine (Chartier Harlan et al. (1991) Nature 353:844-846, valine717 to phenylalanine (Murrell et al. (1991) Science 254:97-99) and with respect to the 695 isoform, a double mutation changing lysine595-methionine596 to asparagine595-leucine596 (Mullan et al. (1992) Nature Genet 1:345-347) referred to as the Swedish mutation. Moreover, (3-amyloid peptide appears to be toxic to brain neurons, and neuronal cell death is associated with the disease.
Thus, the ability to monitor cellular processing of the amyloid precursor protein would be of significant value in the diagnosis, prognosis, and therapeutic supervision of Alzheimer's disease. In particular, it would be desirable to identify minimally invasive procedures for screening and evaluating detectable diagnostic markers in readily obtainable patient samples, such as serum, cerebrospinal fluid (CSF), and the like.
A number of potential diagnostic markers for Alzheimer's disease have been proposed. Of particular interest to the present invention are certain fragments of the amyloid precursor protein, including carboxy terminal fragments (such as the (3-amyloid peptide itself and fragments thereof), and amino-terminal fragments (such as certain 25 kD, 105 kD, and 125 kD fragments). As yet, none of the proposed markers has proved to be definitive for the antemortem diagnosis or monitoring of Alzheimer's disease.
Thus, it would be desirable to identify additional and alternative diagnostic markers for Alzheimer's disease.
Such markers should be useful by themselves and/or in combination with other diagnostic markers and procedures.
Preferably, the diagnostic markers would be detectable in body fluids, such as CSF, blood, plasma, serum, urine, tissue, and the like, so that minimally invasive diagnostic procedures can be utilized.
Of further interest to the present invention are in vitro and in vivo systems and methods for screening candidate drugs for the ability to inhibit or prevent the production of WO 95/1199=1 217 ~ 6 3 2 pCT~S94/11782 /3-amyloid plaque. It would be desirable to provide methods and systems for screening test compounds for the ability to inhibit or prevent the conversion of amyloid precursor protein to /3-amyloid peptide. In particular, it would be desirable to base such methods and systems on metabolic pathways which have been found to be involved in such conversion, where the test ' compound would be able to interrupt or interfere with the metabolic pathway which leads to conversion. Such methods and systems should be rapid, economical, and suitable for screening large numbers of test compounds.
2. Description of the Background Art ~i-amyloid peptide (also referred to as A4, ~iAP, A~3, or A(3P; see, U.S. Patent No. 4,666,829 and Glenner and Wong (1984) Biochem. Biophys. Res. Commun. 120:1131-1135) is derived from /3-amyloid precursor protein (aAPP), which is expressed in differently spliced forms of 695, 751, and 770 amino acids.
See, Kang et al. (1987) Nature 325:773-776; Ponte et al. (1988) Nature 331:525-527; and Kitaguchi et al. (1988) Nature 331:530-532. Normal processing of amyloid precursor protein involves proteolytic cleavage at a site between residues Lysl6 and Leul7 (as numbered for the vAP region where Asp597 is residue 1 in Kang et al. (1987)), supra, near the transmembrane domain, resulting in the constitutive secretion of an extracellular domain which retains the remaining portion of the /3-amyloid peptide sequence (Esch et al. (1990) Science 248:1122-1124).
This pathway appears to be widely conserved among species and present in many cell types. See, Weidemann et al. (1989) Cell 57:115-126 and Oltersdorf et al. (1990) J. Biol. Chem.
265:4492-4497. This normal pathway cleaves within the region of the precursor protein which corresponds to the /3-amyloid peptide, thus apparently precluding its formation. Another T constitutively secreted form of ~3APP has been noted (Robakis et al. Soc. Neurosci. October 26, 1993, Abstract No. 15.4, Anaheim, CA.) which contains more of the aAP sequence carboxy terminal to that form described by Esch et al. supra.
Golde et al. (1992) Science 255:728-730, prepared a series of deletion mutants of amyloid precursor protein and observed a single cleavage site within the /3-amyloid peptide region. Based on this observation, it was postulated that /3-amyloid peptide formation does not involve a secretory pathway.
Estus et al. (1992) Science 255:726-728, teaches that the two largest carboxy terminal proteolytic fragments of amyloid precursor protein found in brain cells contain the entire /3-amyloid peptide region. ' PCT application WO 92/00521 describes methods for evaluating Alzheimer's disease based on measuring the amounts of certain 25 kD, 105 kD, and 125 kD soluble derivatives of amyloid precursor protein in a patient's cerebrospinal fluid.
Fig. 3 of WO 92/00521 suggests that cleavage of amyloid precursor protein may occur adjacent to the amino-terminus of ~-amyloid peptide to produce a soluble amino-terminal fragment, but no evidence or discussion of such cleavage is presented in the application. Kennedy et al. (1992) Neurodegeneration 1:59-64, present data for a form of secreted (3APP, which was characterized by its reactivity with antibodies to residues 527-540 of /3APP and the lack of reactivity with antibodies to the first fifteen residues of /3AP. No direct evidence is provided to suggest the cleavage site or identity of the carboxy terminus of the (3APP form. PCT application WO 91/16628 describes methods for diagnosing disease based on detection of amyloid precursor proteins and fragments thereof utilizing antibodies to protease nexin-2 or amyloid precursor protein.
Recent reports show that soluble /3-amyloid peptide is produced by healthy cells into culture media (Haass et al.
(1992) Nature 359:322-325) and in human and animal CSF (Seubert et al. (1992) Nature 359:325-327).
Palmert et al. (1989) Biochm. Biophys. Res. Comm.
165:182-188, describes three possible cleavage mechanisms for aAPP and presents evidence that ~3APP cleavage does not occur at methionine596 in the production of soluble derivatives of /3APP.
U.S. Patent No. 5,200,339, discusses the existence of certain proteolytic factors) which are putatively capable of cleaving aAPP at a site near the ~3APP amino-terminus.
SUMMARY OF THE INVENTION

t~ PCT/US94/11782 Methods and compositions are provided for detecting and monitoring an amino-terminal fragment of (3-amyloid precursor protein (~3APP) resulting from ~i-secretase cleavage.
The resulting fragment, referred to hereinafter as ATF-(3APP, ' 5 may be detected in biological samples and is useful for monitoring the processing of /3APP in animal models. In particular, the present invention provides for monitoring of QAPP processing in vivo where the presence of the ATF-~3APP is detected in a specimen from an animal transformed to express the Swedish mutation of human f3APP and where the ATF-(3APP has been cleaved from (3APP between LEU596 and ASP597, based on the numbering of Kang, et al., supra, in the 695 amino acid isoform. The Swedish mutation results from a substitution of ASN595-LEU596 for LYS595-MET596 which are present in the wild type 695 isoform of /3APP.
It has been found that animal models expressing the human Swedish mutation of the ~iAPP gene are particularly prolific producers of the amino-terminal fragment thereof.
That is, non-human animal models are able to cleave the human Swedish mutation at a greater frequency than cleavage of either the endogenous /3APP or the wild type human ~3APP. It is further believed that intracellular processing of the Swedish mutation of human LAPP results in greater production of the ATF-/3APP
than is produced by other human mutations of the /3APP gene.
Thus, transgenic animal systems, such as transgenic mice, expressing the Swedish mutation of ~3APP are particularly suitable as models for monitoring intracellular processing of LAPP as well as for screening test compounds for the ability to inhibit or modulate cleavage of ~iAPP as a result of /3-secretase activity, the apparently pathogenic form of aAPP processing.
In a preferred aspect of the present invention, ATF-LAPP produced by (3-secretase cleavage of the Swedish mutation of /3APP is detected by a binding substance which specifically binds to an epitope present at or near the carboxy-terminus of ATF-(3APP, where the epitope comprises the carboxy terminal residue (LEU596), and preferably the five amino acid residues which are present at the carboxy terminus of ATF-LAPP. It has been found that binding between binding 2~7~632 6 substances, such as antibodies, against such a C-terminal epitope provides for high specificity detection of the Swedish form of ATF-aAPP. Preferred antibodies are raised against a peptide comprising the five-residue carboxy terminus which is exposed by /3-secretase cleavage of the Swedish (3APP.
In another preferred aspect of the present invention, /3-amyloid production inhibitors may be identified in transgenic animal models as described above. The animals are exposed to test compounds) and those compounds which affect, usually by diminishing, the production of ATF-~3APP are considered candidates for further testing as drugs for the treatment of ~iAP-related conditions.
In yet another aspect of the present invention, antibodies are provided which are specific for the amino-terminal fragment of Swedish aAPP. Preferably, the antibodies are raised against a peptide comprising the amino-terminal residues of the Swedish ATF-~iAPP, and the antibodies may be polyclonal or monoclonal.
BRIEF DESCRIPTION OF THE DRAWINGS
Figs. lA and iB illustrate the various isoforms of normal (3APP and the corresponding isoforms of ATF-aAPP, respectively.
Figs. 2A, 2B, and 2C, are chemiluminescent gel patterns of material derived from conditioned medium of a human fetal brain cell culture. Lanes 1, 2, and 3 of each gel represent the untreated conditioned medium, the conditioned medium depleted by reaction with antibody which recognizes an epitope within the ~i-amyloid peptide residues 1-16, and the material removed by this antibody, respectively. Panels A, B, and C represent the following probes: anti-5 antibody (which recognizes locations on ~iAPP amino-terminal to (3-amyloid peptide region), antibody 92 (which was raised against a synthetic peptide terminating in the C-terminal methionine exposed by cleavage of ~3AP from aAPP), and antibody lOD5 (a monoclonal antibody which recognizes an epitope of (3AP within residues 1-16), respectively.

WO 95111994 2 ~ 7 4 6 ~ 2 PCTIUS94/11782 Fig. 3 is a chemiluminescent gel pattern obtained by examining human lumbar CSF. The CSF was probed with 92 antibody either alone (lane 1) or preincubated with various peptides representing variations of the C-terminus of ATF-~3APP.
' S A significant competition (reduction in binding) was observed with peptides terminating in the C-terminal methionine (lanes 3, 4, 6, and 7). The peptides were as follows: Lane 1, no competing peptide added; Lane 2, GSGLTNIKTEEISEVK; Lane 3, YSGLTNIKTEEISEVKM; Lane 4, ISEVKI~i; Lane 5, EISEVKMD; Lane 6, CISEVKM; Lane 7, YISEVKM. MW = molecular mass markers (indicated in kilodaltons).
Fig. 4 is an autoradiogram representing electrophoretic gel patterns obtained by immunoprecipitation of conditioned medium from various cell lines. The material secreted by human fetal brain cultures and immunoprecipitated by antibody 92 (lane il at the arrow) is apparently smaller than the material precipitated by antibody 6C6 (lane 10 at the arrow). Antibody 6C6 recognizes an epitope within residues 1-16 of ~iAP.
Fig. 5 is an autoradiogram representing electrophoretic gel patterns obtained by immunoprecipitation of conditioned medium from human 293 cell lines transfected with cDNA encoding both normal and Swedish aAPP. The amount of AFT-/3APP material secreted by the Swedish transfected cells (lanes 11 and 12) is qualitatively greater than that produced by normal (3APP transfectants (lanes 9 and 10).
Fig. 6 is an immunoblot demonstrating specificity of a monoclonal antibody raised against the Swedish mutation of ATF-f3APP (referred to hereinafter as the Swedish 192 antibody).
Fig. 7A is a map of plasmid pNSEAPPswn3'.
Fig. 7B is a map of plasmid pNSEAPPsw.
Fig. 8 is a Western blot of soluble fractions of transgenic and non-transgenic (control) mice brain homogenates probed for the presence of (3APP fragments.
Fig. 9 are Western blots of brain homogenates from transgenic and non-transgenic (control) mice demonstrating that the Swedish 192 antibody does not cross-react with fragments of SAPP cleaved at the a-secretase site.

2 ~ 7~-X32 DESCRIPTION OF THE SPECIFIC EMBODIMENTS
The present invention results from the identification of a novel secreted fragment of /3-amyloid precursor protein (/3APP) which results from cleavage of an intact ~i-amyloid peptide (/3AP) region from the precursor protein. The novel secreted fragments comprise the amino-terminal portion of aAPP
which remains after such cleavage and will be referred to hereinafter as the amino-terminal fragment form of aAPP (ATF-/3APP). ATF-~3APP is believed to be the product of an l0 alternative secretory processing pathway for /3APP, which pathway is present even in normal (non-diseased) cells. It is further believed, however, that the alternate secretory pathway may be responsible for an essential event in the production of (3AP in diseased cells in patients, and that abnormal production of ATF-/3APP may be involved in diseases related to aAP plaque, particularly Alzheimer's disease and Down's syndrome. Thus, the present invention provides methods and compositions for monitoring the cellular processing of (3APP based on the detection and measurement of ATF-aAPP in biological samples.
ATF-(3AP is identified and recognized by specific binding to antibodies raised against peptides comprising certain residues of ~iAPP which lie immediately adjacent to the (3AP region and for normal /3APP include the carboxy terminal methionine (numbered as methionine596 in the 695 isoform, as set forth below). The peptides will usually include at least five contiguous residues up to and including residue596, and specific methods for producing such antibodies are set forth below.
Referring now to Figs. lA and 1B, aAPP is found in three isoforms comprising 695, 751, and 770 amino acids, respectively. The 695 isoform is the most common in neuronal cells, with the 751 and 770 isoforms resulting from insertions at residue 289 on the 695 isoform (all numbering of the 695 isoform will be in accordance with Kang et al. (1987) Nature 325:733-736). ATF-aAPP apparently results from proteolytic T
cleavage of the various ~3APP isoforms at or within the five residues on the amino-terminal side of the amino-terminus of the ~i-amyloid peptide (aAP) region, which is located between WO 95/11994 217 4 6 J ~ pCT~S94/11782 residues 596 and 597 of the 695 isoform. Such cleavage results in the exposure of a C-terminal residue, which will usually be methionine596, lysine595 or leucine596, more usually being methionine596 or leucine596, shown as MET596 and LEU596 in Fig. 1B. Exposure of LEU596 occurs with ~3-secretase cleavage of the Swedish mutation of human f3APP in both humans and animal ' models, as described in more detail hereinbelow. It will be appreciated, of course, that the C-terminal residues would have a different numbering when the ATF-(3APP is derived from a different (3APP isoform. In particular, the C-terminal methionine would be MET652 and MET6~1 and the C-terminal leucine would be LEU651 and LEU6~~ in the 751 and 770 aAPP isoforms, respectively. As used hereinafter and in the claims, methionine596, lysine595, and leucine596 will refer generally to corresponding residues in all other isoforms or variants of aAPP. Presently, it is believed that the N-terminal residue of ATF-/3APP is LEU18 in all isoforms (based on processing of the amino-terminal end of ,LAPP in secreted forms which are cleaved within the (3AP region).
According to the present invention, ATF-aAPP may be detected and/or measured in a variety of biological samples, including in vitro samples, such as conditioned medium from cultured cells, and patient and animal model samples, typically CSF, blood, serum, plasma, urine, tissue, and the like.
Detection and measurement may be accomplished by any technique capable of distinguishing ATF-aAPP from other ~i-APP fragments which might be found in the sample. Conveniently, immunological detection techniques may be employed which utilize antibodies, antibody fragments, or other equivalent specific binding substances, which bind to an epitope characteristic of the ATF-~3APP and which are substantially free of cross-reactivity with other /3APP fragments. Of particular use are antibodies and other binding substances which bind to an epitope which comprises the C-terminal residue of ATF-~iAPP
which is exposed upon (3-secretase cleavage of the ~3AP region, e.g., methionine596, leucine596 or lysine595. It has been found that such C-terminal-specific antibodies are able to discriminate between the ATF-RAPP and related aAPP fragments.

21?4632 Alternatively, immunological detection techniques may be based on isolated and purified ATF-~iAPP using conventional techniques. The preparation of both C-terminal residue-specific antibodies and purified and isolated ATF-~3APP are 5 described hereinafter. Particularly suitable detection techniques include ELISA, Western blotting, radioimmunoassay, and the like.
Other techniques for detecting ATF-~iAPP which do not require the use of ATF-~3APP specific antibodies and/or 10 competing antigen may also be employed. For example, two-dimensional gel electrophoresis may be employed to separate closely related soluble fragments of ~iAPP. Antibodies cross-reactive with many or all of the fragments may then be used to probe the gels, with the presence of ATF-aAPP being identified based on its precise position on the gel. Other techniques for detection of ATF-aAPP are also well within the skill in the art. For example, the secreted /3APP species which contain the amino-terminal region of ~3AP can be immunologically removed from a sample to isolate ATF-(3APP (see Fig. 2A, lane 2 and Fig. 5, lanes 11 and 12), which can then be detected by any of several methods as discussed above.
Antibodies specific for an epitope characteristic of the wild type ATF-/3APP may be prepared against a suitable antigen or hapten comprising the C-terminal ATF-~3APP sequence including the methionine residue. Antibodies specific for an epitope characteristic of the Swedish mutation of ATF-~iAPP may be prepared against a suitable antigen or hapten comprising the C-terminal ATF-(3APP sequence including the leucine residue at position 596. Conveniently, synthetic peptides may be prepared by conventional solid phase techniques, coupled to a suitable immunogen, and used to prepare antisera or monoclonal antibodies by conventional techniques. One suitable synthetic peptide consists of six residues of ATF-~3APP (ISEVHI~I) which are located on the immediate amino-terminal side of ~iAP and which may be coupled to an immunogen and used to prepare specific antibodies as described in detail in the Experimental section.
Other suitable peptide haptens will usually comprise at least five contiguous residues within aAPP on the immediate amino-WO 95/11994 PC'T/US9=i/11783 terminal side of SAP, and may include more than six residues (although a peptide including sixteen amino-terminal residues was found to yield antisera which was less specific). The carboxy terminal 25 residues of the normal ATF-j3APP areas follows (using the single letter amino acid designations).
DRGLT TRPGSGLTNI KTEEISEVIQ~i The carboxy terminal 25 residues of the Swedish mutation of ~iAPP are as follows:
DRGLT TRPGSGLTNI KTEEISEVNL

Synthetic polypeptide haptens may be produced by the well-known Merrifield solid-phase synthesis technique where amino acids are sequentially added to a growing chain (Merrifield (1963) J. Am. Chem. Soc. 85:2149-2156). The amino acid sequences may be based on the sequence of ATF-~iAPP set forth above or may utilize naturally occurring or engineered mutant sequences. For example, the peptides mimicking Swedish mutant would have asparagine595-leucine596 substituted for lysine59s-methionine59s and another substitution might include only the leucine596 substitution for methionine596. A preferred peptide comprises the carboxy terminal five amino acids of Swedish ATF-~iAPP: SEVNL.
Once a sufficient quantity of polypeptide hapten has been obtained, it may be conjugated to a suitable immunogenic carrier, such as serum albumin, keyhole limpet hemocyanin, or other suitable protein carriers, as generally described in Hudson and Hay, Practical Immunology, Blackwell Scientific Publications, Oxford, Chapter 1.3, 1980.
Once a sufficient quantity of the immunogen has been obtained, antibodies specific for the C-terminal residue exposed upon cleavage of ~3AP from ATF'-~iAPP may be produced by in vitro or in vivo techniques. In vitro techniques involve exposure of lymphocytes to the immunogens, while in vivo techniques require the injection of the immunogens into a PCTIUS9a/11782 suitable vertebrate host. Suitable vertebrate hosts are non-human, including mice, rats, rabbits, sheep, goats, and the like. Immunogens are injected into the animal according to a predetermined schedule, and the animals are periodically bled with successive bleeds having improved titer and specificity.
The injections may be made intramuscularly, intraperitoneally, subcutaneausly, or the like, and an adjuvant, such as incomplete Freund's adjuvant, will be employed.
If desired, monoclonal antibodies can be obtained by to preparing immortalized cell lines capable of producing antibodies having desired specificity. Such immortalized cell lines may be produced in a variety of ways. Conveniently, a small vertebrate, such as a mouse is hyperimmunized with the desired immunogen by the method just described. The vertebrate is then killed, usually several days after the final immunization, the spleen cells removed, and the spleen cells immortalized. The manner of immortalization is not critical.
Presently, the most common technique is fusion with a myeloma cell fusion partner, as first described by Kohler and Milstein (1975) Nature 256:495-497. Other techniques including EBV
transformation, transformation with bare DNA, e.g., oncogenes, retroviruses, etc., or any other method which provides for stable maintenance of the cell line and production of monoclonal antibodies. Specific techniques for preparing monoclonal antibodies are described in Antibodies: A
Laboratory Manual, Harlow and Lane, eds., Cold Spring Harbor Laboratory, 1988.
In addition to monoclonal antibodies and polyclonal antibodies (antisera), the detection techniques of the present invention will also be able to use antibody fragments, such as F(ab), Fv, VL, VH, and other fragments. It will also be possible to employ recombinantly produced antibodies (immunoglobulins) and variations thereof as now well described in the patent and scientific literature. See, for example, EPO 8430268.0; EPO 85102665.8; EPO 85305604.2; PCT/GB 85/00392;
EPO 85115311.4; PCT/US86/002269; and Japanese application 85239543.

a WO 9511199a ~ ~ PCTlUS9:1111782 It would also be possible to prepare other recombinant proteins which would mimic the binding specificity of antibodies prepared as just described.
The present invention further comprises isolated and purified ATF-(3APP, usually obtained in substantially pure form.
"Substantially pure" means at least about 50o w/w (weight/weight) or more purity with substantial freedom from interfering proteins and contaminants. Preferably, the ATF-~APP will be isolated or synthesized in a purity greater than 50% w/w, preferably being 80% w/w or higher. The ATF-~3APP may be purified from a natural source by conventional protein purification techniques, with homogeneous compositions of at least about 50% w/w purity being purified by use of antibodies prepared as described above using conventional immunoaffinity separation techniques. Suitable natural starting materials include conditioned medium from ATF-(3APP-producing cell lines, such as fetal brain cell cultures; and the like.
Alternatively, the ATF-QAPP may be isolated from biological samples obtained from a human host, such as CSF, serum, and the like. Suitable protein purification techniques are described in Methods in Enzymology, Vol. 182, Deutcher, ed., Academic Press, Inc., San Diego, 1990.
Antibodies and purified ATF-aAPP prepared as described above can be used in various conventional immunological techniques for detecting ATF-~3APP in biological samples, particularly patient samples and animal specimens for the monitoring of ~3-amyloid-related diseases and drug screening, and in conditioned media from cell culture for monitoring the intracellular processing of aAPP. Suitable immunological techniques include immunoassays, such as ELISA, Western blot analyses, and the like. Numerous specific immunological detection techniques are described in Harlow and Lane, supra.
Detection of ATF-aAPP in patient samples can be used for diagnosing and monitoring of Alzheimer's disease and other diseases related to (3-amyloid plaque deposition, such as Down's syndrome. Suitable patient samples include CSF, blood, serum, ~WO 9511199:1 ~ ) PCTIUS9a111782 plasma, urine, tissue, and the like. Presence of the disease will generally be associated with elevated levels of ATF-J3APP, or elevated ratios of the amount of ATF-j3APP to the amounts of other secreted ~3APP fragments (i.e., those (3APP fragments cleaved within or carboxy terminal to the ~iAP region) when compared to those values in normal individuals, i.e., individuals not suffering from Alzheimer's disease or other Q-amyloid-related disease. The amount of ATF-~iAPP may be compared to the amount of another species of APP, either an isoform (e. g., 695, 751 or 770) and/or a form further defined by its carboxy terminus (e. g., forms cut at and/or carboxy terminal to that site described by Esch et al.). In addition to initial diagnostic procedures, levels of ATF-~3APP may be monitored in order to follow the progress of the disease, and potentially follow the effectiveness of treatment. It would be expected that levels of ATF-~iAPP would decrease with an effective treatment regimen.
In vitro monitoring of ATF-aAPP levels in cultured medium from a suitable cell culture may be used for drug screening. By growing cells under conditions which result in the secretion of ATF-~3APP into the culture medium, and exposing the cells to test compounds, the effect of these test compounds on ATF-~iAPP secretion may be observed. It would be expected that test compounds which are able to diminish the amount of ATF-~iAPP would be candidates for testing as inhibitors of (3AP
formation. Suitable cell lines include human and animal cell lines, such as 293 human kidney cell line, human neuroglioma cell lines, human HeLa cells, primary endothelial cells (e. g., HWEC cells), primary human fibroblasts or lymphoblasts (including endogenous cells derived from patients with ~3APP
mutations), primary human mixed brain cells (including neurons, astrocytes and neuroglia), Chinese hamster ovary (CHO) cells, and the like. Cell lines which preferentially increase the levels or ratios of ATF-aAPP would be particularly useful in the methods of invention.
Similarly, in vitro monitoring of ATF-(3APP in animal models of Alzheimer's disease, such as the mouse model disclosed in WO 91/19810 WO 95/1199:1 PCT/US94/11782 may also be used to screen test compounds for therapeutic effectiveness (usually for testing of compounds which have previously been identified by an in vitro screen). The test compounds) are administered to 5 the animal and the level of ATF-QAPP or ratio of ATF-~3APP to other aAPP fragments observed. Those compounds which reduce the level of ATF-~iAPP, or decrease the ratio of ATF-~3APP to other QAPP fragments, will be considered to be candidates for further evaluation.
10 Particularly preferred animal models for ~3-secretase cleavage of QAPP are transgenic animals which express the Swedish mutation of the /3APP gene, as described above. It has been found that such transgenic animals, particularly transgenic mice, produce high quantities of ATF-~3APP which may 15 detected according to the methods of the present invention. In particular, it has been found that Swedish mutation of aAPP
produces quantities of the ATF-aAPP which will usually be at least two-fold higher than wild type human (3APP expressed in animals. Usually, production will be significantly higher, typically being at least two-fold higher. With such elevated levels of ATF-(3APP production, monitoring ~i-secretase activity under different conditions is greatly facilitated. In particular, screening for drugs and other therapies for inhibiting j3-secretase activity (and thus inhibiting (APP
production) are greatly simplified in animals models expressing the Swedish mutation of human ~3APP.
The use of animal models for screening drugs for ~3--secretase inhibition activity will often be performed after in' vitro cell culture screening techniques, as described above, have been performed. Drugs which appear promising in in vitro screening may then be administered to test animals, such as test mice, which are transgenic and which express the Swedish mutation of human ~iAPP. Particular techniques for producing transgenic mice which express the Swedish form of aAPP are described in the Experimental section hereinafter. It will be appreciated that the preparation of other transgenic animals expressing the Swedish human aAPP may easily be accomplished, including rats, hamsters, guinea pigs, rabbits, and the like.

'~ i ~ ~-632 16 The effect of test compounds on ATF-/3APP production in test animals may be measured in various specimens from the test animals. In the Experimental section, the detection of ATF-~iAPP in brain homogenates is described in detail.
Detection of ATF-/3APP in brain homogenates is exemplary, but not necessarily preferred. In some cases, it will be advantageous to measure the ATF-aAPP in other specimens, such as cerebrospinal fluid, blood, and the like, which may be obtained from the test animal without sacrifice of the animal.
In all cases, it will be necessary to obtain a control value which is characteristic of the level of ATF-aAPP
production in the test animal in the absence of test compound(s). In cases where the animal is sacrificed, it will be necessary to base such control values on an average or a typical value from other test animals which have been transgenically modified to express the Swedish mutant of human ~iAPP but which have not received the administration of any test compounds or any other substances expected to affect the level of production of ATF-aAPP. Once such control level is determined, test compounds can be administered to additional test animals, where deviation from the average control value indicates that the test compound had an effect on the /3-secretase activity in the animal. Test substances which are considered positive, i.e., likely to be beneficial in the treatment of Alzheimer's disease or other (3-amyloid-related conditions, will be those which are able to reduce the level of ATF-aAPP production, preferably by at least 20%, more preferably by at least 50%, and most preferably by at least 80%.
The test compounds can be any molecule, compound, or other substance which can be added to the cell culture or administered to the test animal without substantially interfering with cell or animal viability. Suitable test compounds may be small molecules, biological polymers, such as polypeptides, polysaccharides, polynucleotides, and the like.
The test compounds will typically be administered to the culture medium at a concentration in the range from about 1 nM
to 1 mM, usually from about 10 ~.M to 1 mM. The test compounds '1 ) WO 95!11994 PCT/US9a111782 will typically be administered at a dosage of from 1 ng/kg to mg/kg, usually from l0 ug/kg to 1 mg/kg.
Test compounds which are able to inhibit secretion or animal production of ATF-/3APP are considered as candidates for 5 further determinations of the ability to block ~3-amyloid production in animals and humans. Inhibition of secretion ar production indicates that cleavage of ~3APP at the amino-terminus of ~3AP has likely been at least partly blocked, reducing the amount of a processing intermediate available for 10 conversion to ~-amyhoid peptide.
The present invention further comprises. methods for inhibiting (3-amyloid production in cells, where the method includes administering to the cells compounds selected by the method described above. The compounds may be added to cell culture in order to inhibit ~3AP production by the cultured cells. The compounds may also be administered to a patient in order to inhibit the deposition of amyloid plaque associated with Alzheimer's and other (3AP-related diseases.
The present. invention further comprises pharmaceutical compositions incorporating a compound selected by the above-described method and including in a pharmaceutically acceptable carrier. Such pharmaceutical compositions should contain a therapeutic or prophylactic amount of at least one compound identified by the method of the present invention. The pharmaceutically acceptable carrier can be any compatible, non-toxic substance suitable to deliver the compounds to an intended host. Sterile water, alcohol, fats, waxes, and inert solids may be used as the carrier.
Pharmaceutically acceptable adjuvants, buffering agents, dispersing agents, and the like may also be incorporated into the pharmaceutical compositions. Preparation of pharmaceutical conditions incorporating active agents is well described in the medical and scientific literature. See, for example, Remington's Pharmaceutical Sciences, Mack Publishing Company, Easton, Pennsylvania, 16th Ed., 1982.
The pharmaceutical compositions just described are suitable for systemic administration to the host, including both parenteral, topical, and oral administration. The pharmaceutical compositions may be administered parenterally, i.e. subcutaneously, intramuscularly, or intravenously. Thus, the present invention provides compositions for administration to a host, where the compositions comprise a pharmaceutically acceptable solution of the identified compound in an acceptable carrier, as described above.
Frequently, it will be desirable or necessary to introduce the pharmaceutical compositions directly or indirectly to the brain. Direct techniques usually involve placement of a drug delivery catheter into the host's ventricular system to bypass the blood-brain barrier. Indirect techniques, which are generally preferred, involve formulating the compositions to provide for drug latentiation by the conversion of hydrophilic drugs into lipid-soluble drugs.
Latentiation is generally achieved through blocking of the hydroxyl, carboxyl, and primary amine groups present on the drug to render the drug more lipid-soluble and amenable to transportation across the blood-brain barrier. Alternatively, the delivery of hydrophilic drugs can be enhanced by intra-arterial infusion of hypertonic solutions which can transiently open the blood-brain barrier.
The concentration of the compound in the pharmaceutical carrier may vary widely, i.e. from less than about 0.1 ~ by weight of the pharmaceutical composition to about 20% by weight, or greater. Typical pharmaceutical composition for intramuscular injection would be made up to contain, for example, one to four ml of sterile buffered water and one ~,g to one mg of the compound identified by the method of the present invention. The typical composition for intravenous infusion could be made up to contain 100 to 500 ml of sterile Ringer's solution and about 1 to 100 mg of the compound.
The pharmaceutical compositions of the present invention can be administered for prophylactic and/or therapeutic treatment of diseases related to the deposition of SAP, such as Alzheimer's disease and Down's syndrome. In therapeutic applications, the pharmaceutical compositions are 1~4~~~
WO 9511199.1 PCT/US94/11782 administered to a host already suffering from the disease. The pharmaceutical compositions will be administered in an amount sufficient to inhibit further deposition of aAP plaque. An amount adequate to accomplish this is defined as a "therapeutically effective dose." Such effective dose will depend on the extent of the disease, the size of the host, and the like, but will generally range from about 0.01 ~g to 10 mg of the compound per kilogram of body weight of the host, with dosages of 0.1 ~g to 1 mg/kg being more commonly employed.
For prophylactic applications, the pharmaceutical compositions of the present invention are administered to a host susceptible to the (3AP disease, but not already suffering from such disease. Such hosts may be identified by genetic screening and clinical analysis, as described in the medical literature. The pharmaceutical compositions will be able to inhibit or prevent deposition of the j3AP plaque at a very early stage, preferably preventing even the initial stages of the (3-amyloid disease. The amount of the compound required for such prophylactic treatment, referred to as a prophylactically-effective dosage, are generally the same as described above for therapeutic treatment.
The following examples are offered by way of illustration, not by way of limitation.
ERPERIMENTAL
Materials and Methods 1. Antibody and Affinity Matrix Preparation Monoclonal antibody 6C6 was raised and screened in the same manner as antibody lODS (Hyman et al. (1992) J.
Neuropath. Exp. Neurol. 51:76) using a synthetic peptide containing (3AP residues 1-28 conjugated to rabbit serum albumin as the immunogen. Both lOD5 and 6C6 recognize an epitope within the first 16 amino acids of the aAP sequence. 6C6 was more efficient than lOD5 in immunoprecipitation and was used as a capture antibody. To prepare 6C6 resin, 4 mls of Affigel~10 (Bio-Rad Laboratories, Hercules, CA) was washed with cold water and combined with 3 mls of 6C6 (12.5mg/ml in PBS (2.7 mM KC1, 1.5 mM KH2P04, 8.1 mM Na2HP04, 137 mM NaCl, pH 7.5) 0.5 M NaCl.

PCT/US9a/11782 WO 95/1199.) The coupling proceeded overnight at 4°C with gentle shaking.
4001 of 1M Tris, pH 8.0, was then added, and shaking was continued for 40 minutes. The resin was then washed with TTBS
(137 mM NaCl, 5 mM KC1, 25 mM Tris, 0.5o Tween~20; pH ?:5) 5 exhaustively before use. Antibody 7H5 is also described in Hyman et al. (1992), supra. Anti-5 antibodies were raised against aAPP 4~4-592 (Oldersdorf et al. (1989) supra, and (1990) supra.
Antibodies (designated antibody 92) were raised 10 against a synthetic peptide including residues 591-596 of aAPP
(as numbered in Kang et al. (1987), supra). The peptide (N-acetyl-CISEVKM) was conjugated to rabbit serum albumin which had been activated with sulfo-maleimido benzoyl-N-hydroxysuccinimide ester to form an immunogen. Antisera were 15 raised against the immunogen in rabbits by standard methodologies. During each inoculation, rabbits received 5~sg of immunogen in 0.1 ml injections subcutaneously at approximately 10 sites (50~g/boost). The same peptide was coupled to Sulfo-linkT" gel (Pierce Chemical Co., Rockford, I:L) 20 for the affinity purification of antibodies from the IgG
fraction.
A more detailed description of the antibody 92 preparation is as follows. Rabbit serum albumin (12.3 mg) was incubated with 13 mg of sulfo-maleimido benzoyl-N-hydroxysuccinimide ester in 1.25 mls of 0.05 M KHZP04, pH 7.0 for 20 minutes at 0°C. The mixture was then immediately subjected to gel filtration on a 1 x 75 cm column of Sephadex G-10 equilibrated with the phosphate buffer. The protein eluant in the excluded volume was pooled and immediately combined with 30 mg of N-acetyl-CISEVKM peptide which was synthesized by standard automated solid phase methodologies.
The coupling reaction (20 ml volume) was allowed to proceed overnight and was then sent to a commercial facility for antibody generation. The injection protocol was to emulsify the antigen in an equal volume of Freund's complete adjuvant and subcutaneously inject a total of 50 ~cg of antigen in o.l ml aliquots in approximately 10 sites. Every three weeks thereafter, a booster injection was given by an identical * Trademark WO 95111994 ? PCT/US9a/11782 protocol except Freund's incomplete adjuvant was used as the emulsifier. Rabbits were bled one week following each injection and the serum examined for titer by reaction to peptide in ELISA. The IgG was purified from the positive reacting sera by precipitation with 50% (NH~)2504, (2 x's) and dialyzed against PBS. The N-acetyl-CISEVKM peptide was conjugated to Sulfo-link"' gel (Pierce Chemical Co., Rockford, IL) using the manufacturer's recommendations to generate an affinity resin to purify the peptide specific antibodies. The IgG fraction was applied to the column and, after washing through non-specifically bound material with PBS, the antibodies were eluted with 0.1 M glycine pH 2.5 0.5 M NaCl and then dialyzed vs PBS before freezing.
Swedish 192 antibody was raised against a synthetic peptide composed of residues 590-596 of the Swedish ~iAPP
sequence. In addition to the aAPP sequence, two glycines and a cysteine were added as a spacer and a linker giving the following sequence: CGGEISEVNL. The peptide was conjugated to a commercially available maleimide activated cationized Bovine Serum Albumin (Pierce Imject Supercarrier Immune Modulator, hereafter referred to as cBSA.) Antiserum was raised by following the injection schedule described above for antibody 92.
In general, cBSA was resuspended in deionized water to a concentration of 10 mg/ml. An equal milligram amount of peptide was added to the carrier and mixed for four hours at room temperature. The conjugate was then dialyzed extensive7.y against Dulbecco's phosphate buffered saline without calcium and magnesium.
The conjugate was compared to the starting cBSA on a 6% Novex~pre-poured Tris-glycine gel. Successful conjugation was indicated by a visible shift to a higher molecular weight.
2. Human Fetal Brain Cell Culture Fetal neural tissue specimens were obtained from 1.2-14 week old fetal cadavers. Samples of cerebral cortex were.
rinsed twice with Hank's Balanced Saline Solution (HBSS).
Cortical tissue (2-3 grams) was placed in 10 mls of cold HBSS
to which l mg of DNase (Sigma Chemical Co., St. Louis, MO
* Trademark x WO 9S/1199a l j ) PC?lUS9:111178?

D3427) was added. The triturated suspension was filtered through Nite nylon screens of 210 um then 130 ~.m, as described by Pulliam et al. (1984) J. Virol. Met. 9:301.
Cells were harvested by centrifugation and resuspended in neuronal medium (MEM fortified with 10% fetal bovine serum, 1% glucose, 1 mM NaPyruate, 1 mM glutamine, 20 mM
KC1). Polyethyleneimine coated 100 mm dishes were seeded with 1.5 x 10~ cells in 8 mls of neuronal medium. The medium was exchanged twice weekly. All cultures in this study were grown in vitro at least 30 days. For serum-free growth condition , cultures were shifted into defined medium (DMEM supplemented with 5 ~cg/ml bovine insulin; 0.1 mg/ml human transferrin;
0.1 mg/ml BSA fraction V; 0.062 ug/ml progesterone, 1.6 ~cg/ml putrescine; 0.039 ~cg/ml sodium selenite, 0.042 ug/ml thyroxine;
and 0.033 ~cg/ml triiodo-L-thyronine), and after 3 days the supernatant was harvested.
Conditioned medium from the cells (10 ml) was harvested. EDTA (5 mM), leupeptin (10 ~g/ml), and Tris-HC1 (20 mM, pH 8.0) were added to each l0 ml sample at the indicated final concentration, and the sample spun 30,000 xg for 20 minutes at 4°C. The resulting supernatant was divided into two equal aliquots, 6C6 resin was added to one of the aliquots (200 ~,1 of resin with approximately 5 mg/ml 6C6 bound). Both aliquots were gently mixed for 6 hours at 4°C, the resin was pelleted, and a second 200 ~,1 aliquot of resin was added. The samples were further mixed overnight at 4°C.
The combined resins were washed twice with TTBS, then briefly extracted twice with one ml aliquots of 0.1 M glycine, 0.1 M
NaCl, pH 2.8.
The material extracted from the resin, the medium depleted by the resin, and the starting medium~were individually precipitated with 10% TCA (trichloro-acetic acid) at 0°C for one hour, the pellets washed with acetone and then resuspended in 150 ~C1 of SDS-PAGE sample buffer under reducing conditions and boiled. Each sample (25 ~cl) was subjected to SDS-PAGE using 10-20% tricine gels (Novex). The proteins were transferred to ProBlot PVDF membranes overnight at 40V.
Visualization of immunoreactive proteins employed the TROPIX
* Trademark WO 95/11994 ~ ~ ~ PCT/US94/11782 chemiluminescence system according to the manufacturer's directions for the AMPPD substrate. Primary antibody concentrations used were: anti-5, 0.1 ~,g/ml; 92, 2 ~,g/ml; lOD5, 2 ~,g/ml.
3. Culture of Human 293 Cells Human 293 cells (ATCC No. CRL-1573) were modified to overexpress APP (Selkoe et al. (1988) Proc. Natl. Acad. Sci.
USA 85:7341). Cells were grown in 10 cm dishes to subconfluency prior to use. Metabolic labelling and immunoprecipitation were performed essentially as previously described in Oltersdorf et al. (1989) Nature 341:144 and (1990) J. Biol. Chem. 265:4492. In brief, labelling was performed in 10 cm dishes. Cells were washed in methionine-free medium, incubated for 20 minutes in 2 ml methionine-free medium supplemented with 0.5 mCi 35S-methionine, washed in full medium, and chased for 2 hours in 3 ml of full medium.
Conditioned medium was collected and cleared at 3000 xg for 10 minutes followed by preabsorption with protein A Sepharose~
(Pharmacia, Piscataway, NJ). Immunoprecipitation was performed with 1.5 mg protein A Sepharose~ per sample. Antibody anti-5 was used at 2 ~,g per sample; 6C6, 7H5 and 92 were used at 10 ~Cg per sample. 5 mg of rabbit anti mouse IgG were used with 6C6 and 7H5 as well as in the control samples. Precipitates were washed four times in TBS (137 mM NaCl, 5 mM KCl, 25 mM Tris, pH

7.5), 0.1% NP40, 5 mM EDTA, 1 mM PMSF, 10 ~,g/ml leupeptin.
SDS-PAGE was performed on 5% Laemmli gels.
4. Culture of Human 293 Cells Transfected with Swedish Mutation.
Duplicate,wells in a 6 well tray of human kidney 293 cells were transiently transfected with plasmid vectors expressing either normal human f3APP or the Swedish mutation variant /3APP using DOTAP mediated transfection as described by the manufacturer (Boehringer Mannheim). 40 hours later the cells were placed into methionine free DME containing 10% fetal calf serum and 20 minutes later they were labeled for 35 minutes with 200 ~,Ci/ml 35S-methionine. The cells were then placed back into normal DME medium containing 10% fetal calf serum and incubated another 2.5 hours. The medium was 2; 74E~~ 24 collected from the cells, and spun at 1000 xg for 15 minutes to remove all the cells. The supernatants were split in half and half was immunoprecipitated with anti-5 antibody by standard methods. The other half was incubated overnight with agarose-coupled 6C6 antibody, and the material bound to the 6C6 agarose was separated by centrifugation. The remaining material was then immunoprecipitated with anti-5 antibody. The total anti-5 immunoprecipitates (a5+), 6C6 bound precipitates (6C6+) and 6C6 non-reactive, anti-5 reactive immunoprecipitates (6C6-,a5+) were run on a 5% Laemmli gel and immunoreactive proteins were visualized by autoradiography.
5. Specificity of the Swedish 192 antibody.
Conditioned medium from 293 kidney cells, which have been stably transfected to overexpress the Swedish /3APP
protein, was collected. One milliliter aliquots were added to either 100 ~,1 of immobilized 6C6-affinity resin (see above) or 100 ~1 of heparin agarose (Sigma). The reaction with the 6C6 resin was for 5 hours at 4°C; the heparin-agarose was reacted for 30 minutes at 4°C. After incubation, the resins were washed with TTBS and then 100 ~,1 of 2 X SDS-PAGE sample buffer were added to each sample, the samples were boiled (5 minutes) and briefly centrifuged. Twenty ~,1 of the samples were loaded onto 6% SDS-polyacrylamide gels and electrophoresed. The proteins were transferred to ProBlot~ membranes as described above. The samples were probed with the following antibodies:
6C6 (described above), Swedish 192 (see above), or 8E5 (a monoclonal antibody which recognizes an epitope of /3APP in the region of amino acids 444-592, using the numbering of the 695 form.) All antibodies were used at 2 ~,g/ml during the probing of the immunoblot. The visualization of immunoreactive material was achieved using the Amersham ECL~ system according to the manufacturer's recommendations. Blocking and antibody dilutions were made using 5% non-fat dry milk (Carnation) in TTBS.
6. Transgenic Mice.
Transgenic mice were generated using the plasmids shown in Fig. 7 (NSEAPPsw and NSEAPPswn3'). These plasmids contain the 751 form of fiAPP containing the Swedish mutation . a WO 95/11994 PCT/US9~/11782 (KM to NL at position 595 and 596 of the 695 form). The neural specific enolase promoter drives expression and provides a splice sequence. The rat NSE promoter and splice sequences were derived from pNSE6 (Okayama and Berg (1982) Mol. Cell.
5 Biol. 2:161-170. This vector contains the 4.2 kb BglII
fragment of the rat NSE promoter region (starting from the upstream BglII_site and continuing to the BglII site in the second intron) cloned into the BamHI site of the vector pSP65 (Promega). The vector-derived XbaI site at the 5'end of the 10 promoter defined the 5' end of the promoter used and the NSE
translation initiating ATG, contained within the second intron, was fused to the J3APP-initiating ATG.
NSEAPPsw also contains a.splice sequence from SV40 in the 3' region of the gene. This splice sequence was derived 15 from the Okayama/Berg vector pLl, and is a fusion of the late 16s and 19s message splice sequences (Forss-Petter et al.
(1990) Neuron 5:187-197). Polyadenylation is provided by SV40 sequences.
Transgenic mice incorporating these plasmid sequences 20 were generated using standard techniques. The NotI fragment containing the above described expression cassette was purified and injected into eggs obtained from a C57B1/DBA hybrid mouse.
The eggs were implanted into pseudopregnant mice and the offspring were screened for presence of plasmid sequences in 25 tail DNA by PCR, slot blot and Southern analysis. The-founder mice thus generated were screened for expression of human (3APP
by analysis of their F1 transgenic offspring. Brains from the F1 animals were homogenized with a hand held homogenizer (polytron PT122B, Kinematica~AG) either in SDS buffer (2% SDS~, 20 mM Tris pH 8.0, 150 mM NaCl, 10 mM EDTA) or homogenized in NP-40 buffer (1% NP40, 50 mM Tris pH 7.5, 10 mM EDTA, and a cocktail of protease inhibitors containing 5-10 ~cg/ml leupeptin, 2-4 ~g/ml Pepstatin A, 5-lO~Cg/ml Aprotinin, and 1-2mM PMSF). The SDS lysates were loaded directly onto gels for Western analysis. The NP40 homogenates were spun at 55,000 rpm for 10 minutes in a Beckman ultracentrifuge (T1100.3 rotor) and the supernatants were loaded onto gels for Western analysis. The Western analysis was done by standard procedures * Trademark WO 95111994 ~ C' S PCT/US94/11782 t 26 utilizing either anti-5 (0.4 ~.g/ml) or 8E5 (5~g/ml) antibodies to detect the human specific l3APP. Those lines expressing relatively high levels of f3APP were chosen for further analysis. This included the lines Hillary 14, Chelsea 32 and Chelsea 58. The experiments described here were done on heterozygote animals of these lines derived by breeding transgene containing animals with wild type animals and screening the offspring for presence of the transgene.
Similarly, homozygous animals from a selected number of lines can be used.
Description of the Experimental Figures Fig. 2: Demonstration of truncated LAPP in conditioned medium from human mixed-brain cell cultures.
Sample 1 is the conditioned-medium from culture;
sample 2 is the medium depleted of 6C6-reactive aAPP; and sample 3 is the material extracted from the 6C6 resin. Panel A
was probed with anti-5 antibodies which were raised against the aAPP sequence 444-592 (Oltersdorf et al. (1989) supra, and (1990) supra. Panel B was probed with antibody 92, described in the Materials and Methods section. Panel C was probed with lODS, a monoclonal antibody which recognizes an epitope within ~3AP residues 1-16, as described in the Materials and Methods section. The lower molecular weight bands observed in C2 and C3 were not seen in C1 and are derived from the 6C6 resin and are recognized by goat-anti-mouse IgG alkaline phosphatase conjugate independent of a primary antibody (data not shown).
Fig. 3: Specificity of 92 Antibodies.
One milliliter of a human lumbar CSF specimen obtained from a 75 year old male was precipitated with loo TCA
to effect a ten-fold concentration and processed as described in Fig. 2, except that the gel well was a 4 cm slot. The 92 antibody was diluted to 6.7 ~,g/ml in 0.5 mls of TTBS in the presence of various potentially competing peptides, each at an approximate concentration of 60 ~,M, for 10 hours at 4°C with gentle mixing. The antibody was then diluted eight-fold in 1%
gelatin/TTBS before incubation with strips of the blot of CSF-derived material and processed as described in Fig. 2. The competing peptides were as follows: Lane 1, no competing ";7 WO 95111994 ~ ~ PCT/US94/11782 peptide added; Lane 2, GSGLTNIKTEEISEVK; Lane 3, YSGLTNIKTEEISEVKM; Lane 4, ISEVKM; Lane 5, EISEVKMD; Lane 6, CISEVKM; Lane 7, YISEVKM. MW = molecular mass markers (indicated in kilodaltons).
Fig. 4: Molecular mass heterogeneity of secreted forms of APP
in immunoprecipitation detected by antibodies against different C-termini in cell Lines and primary human fetal brain cultures.
Antibodies: anti-5: Lanes 3, 6, 9; 6C6 (directed against ~iAP peptide residues 1-16): Lanes 4, 7, 10; antibody 92 (against APP amino acids 591 to 596): Lanes 5, 8, 11; 7H5 (against APP-KPI): Lane 12. Cells: left panel (lanes 1 and 3-5): 293 cells stably transfected with APP 695; middle panel (lanes 2 and 6-8): 293 cells stably transfected with APP 751;
right panel (lanes 9-13): human fetal brain cultures.
Controls: lanes 1, 2, and 13: rabbit anti mouse IgG antibody.
Arrows: an example of molecular mass difference between secreted forms of APP recognized by antibodies 6C6 and 92.
SDS-PAGE was performed on a 5% Laemmli gel. MW = molecular mass markers (indicated in kilodaltons).
Fig. 5: Demonstration of truncated /3APP in conditioned medium from human 293 cells transfected with Swedish mutation.
Fig. 5 shows results from duplicate transfections for both normal and Swedish forms. Lanes 1-4 are a5+; lanes 5-8 are 6C6+; and lanes 9-12 are 6C6-, a5+ samples. Lanes 1, 2, 5, 6, 9 and 10 are from normal ~3APP, lanes 3, 4, 7, 8, 11 and 12 are from Swedish ~3APP. The Swedish mutation results in the production of increased AFT-aAPP as lanes 11 and 12 contain more ATF-~iAPP material than lanes 9 and 10.
Fig. 6: ~ Immunoblot demonstrating specificity of the Swedish 192 antibody.
Figure 6 shows an immunoblot demonstrating specificity of the Swedish 192 antibody. Lanes 1, 3, 5 contain material eluted from heparin agarose. Lanes 2, 4, 6 contain material eluted, from the 6C6 resin. Lanes 1 and 2 were probed with antibody 8E5; Lanes 3 and 4 were probed with the Swedish 192 antibody; Lanes 5 and 6 were probed with antibody 6C6.

Figs. 7A and 7B: Plasmid Maps.
Figs. 7A and 7B are plasmid maps of pNSEAPPswn3' and pNSEAPPsw, respectively, which are used to produce transgenic mice as described above.
Fig. 8: Western Blot Fig. 8 is a Western blot of soluble fractions of transgenic and control animal brains probed for the presence of secreted ~iAPP fragments reactive with the Swedish 192 antibody.
Lane 1: molecular weight markers; lane 2: non-transgenic line;
lane 3: transgenic line.
Fig. 9: Western Blot.
Fig. 9 are Western blots of brain homogenates from transgenic (+) and non-transgenic (-) animals depleted of 6C6 antibody-reactive /3APP forms probed with antibody 8E5 (panel A) and Swedish 192 antibody (panel B).
Results Monoclonal antibody 6C6 which recognizes an epitope of (3AP within residues 1-16 was used to immuno-deplete certain /3APP fragments from various samples. The monoclonal antibody 6C6 was coupled to resin (as described above) and incubated with the conditioned medium from human fetal brain cell cultures as described above. As can be seen (Fig. 2, lane C2), this resin effectively removes the (3APP containing QAP 1-6 from the conditioned medium of the cell culture. Substantial aAPP
immunoreactivity, however, is not captured by the resin as detected by anti-5 antibody directed against an epitope N-terminal to the aAP region (Fig. 2, lane A2).
In order to characterize this apparently novel form of'(3APP, we raised antibodies against a synthetic peptide which included (3APP residues 591-596 (as described above). This antibody (designated 92) was found to recognize the species of (3APP not captured by the resin, (Fig. 2, lane B2) but surprisingly did not react with the secreted form of ~3APP
containing the ~3AP 1-16 sequence (lane B3).
The explanation for this lack of cross reactivity appears to be that the 92 antibody recognizes an epitope in aAPP including the carboxy terminal methionine, corresponding to residue 596. Accordingly, we examined the ability of WO 95111994 L ~ ~ ~- 6 3 PCT/US94/11782 various synthetic peptides to block the immunoreactivity generated with 92. As can be seen in Fig. 3, ~iAPP sequence-based peptides ending with the equivalent of methionine 596 substantially block the reaction of 92, while peptides one amino acid longer or shorter at their carboxy termini are comparatively ineffective in competition. The same pattern of peptide competition was observed in cell culture supernatants (data not shown) and in CSF. A series of pulse-chase experiments revealed that detectable amounts of antibody 92 immunoprecipitable material are produced by 293 cells overexpressing either the 695 or 751 isoforms of ~iAPP (Fig. 4, lanes 5 and 8). Similar experiments on human fetal brain cell cultuYes show that 92 immunoprecipitable material can be resolved from 6C6 reactive (3APP by low percentage (5%) SDS-PAGE
(Fig. 4, lanes 9-11). In the fetal brain cell cultures, the alternative processing of Kunitz protease inhibitory domain (KPI)-containing aAPP forms is less apparent, although faint co-migrating hands are observed with antibody 92 and anti-KPI
antibody 7H5 immunoprecipitations (lanes 11, 12).
The ability to resolve the antibody 92 and 6C6 precipitable materials in mixed brain cultures is due, at least in part, to the nearly equal amounts of the respective forms produced as compared to the situation in 293 cells. Estus et al. (1992), supra, observed that compared with other tissues, human brain contained a relatively higher amount of the potentially amyloidogenic carboxy terminal fragment that, based upon size, appears to begin at or near the amino-terminus of (3AP .
The temporal coincidence of the appearance of antibody 92 and 6C6 precipitable aAPP materials argues against the likelihood of a second proteolytic event occurring post-secretion, particularly since longer chase times do not result in a noticeable alteration in the ratio of the 92 and 6C6 reactive species (data not shown). The resolution of the secreted forms by SDS-PAGE, coupled with the complete lack of immunological cross-reactivity of these species, further demonstrate the existence of an alternative secretory pathway.
The alternative cleavage site was designated as the (3-secretase WO 95/1199-!
site to emphasize that cleavage occurs amino-terminal to the ~iAP as distinct from the cleavage described by Esch et al.
(1990) supra which occurs within the ~iAP.
As can be seen in Figure 6, lane 4, the Swedish 192 5 antibody does not appreciably recognize the 6C6 reactive form of ~3APP despite the fact that more total (3APP is present in lane 4 compared to lane 3 (compare lanes 1 and 2.) The lack of reactivity with ~3APP forms containing the partial (3AP sequence (6CG-reactive) suggests the Swedish 192 antibody recognizes 10 j3APP cleaved at or near the amino-terminus of RAP, but not when (3APP extends past this region.
In order to verify the utility this approach as an animal model, soluble fractions of transgenic animal brains were probed for the presence of the "92" form of the secreted 15 ~3APP (Fig 8). This form is produced as a byproduct of the production of f3AP and inhibition of the production of this form in cultured cells accompanies inhibition of the cleavage of the N-terminal end of 13AP, the site cleaved by l3-secretase.
Brains from transgenic (Swedish Hillary 14) or non-20 transgenic mice were homogenized in 50 mM Tris 10 mM EDTA
together with the above described protease inhibitor cocktail and centrifuged at 55K rpm for l0 min as described above. The supernatant was analyzed by Western blot utilizing the Swedish "192" antibody that reacts only with the secreted form of ~3APP
25 produced by !i-secretase. For Western analysis proteins were separated on a 6 % SDS PAGE gel (from Novex) and then transferred to immobilonP by standard techniques. The filter was incubated with 2 ~g/ml of the Swedish "192" antibody using standard techniques, and the bound~antibody visualized using 30 the AmershamECL kit. As shown in Fig 8, lane 3, there was robustly detectable "192" reactive material in the supernatant from the transgenic animal. The non-transgenic animal brain homogenate contained a low amount immunoreactive material that.
is slightly faster in mobility on the gel than the material specific to the transgenic animal (.lane 1). This material is probably not related to BAPP since it does not hybridize with other I3APP antibodies (e.g. anti-5) .
* Trademark WO 9511199:1 In tissue culture systems, the Swedish 192 antibody does not cross-react with secreted f3APP that is cleaved at the a-secretase site at position 17 in the middle of the f3AP
sequence. To prove that this is also true in the mouse models, brain homogenates were depleted of the a-secretase-cleaved LAPP
forms using resin bound to 6C6 antibody, which is specific for the first 16 amino acids of the I3AP, and therefore reacts with amino-terminal fragment of the a-secretase cleaved secreted BAPP fragment but not with the shorter f3-secretase cleaved secreted fiAPP fragment. Resin was produced by using Actigel--ALS coupled in suspension as described by the manufacturer (Sterogene). An excess of resin-antibody was incubated with the brain homogenates from animals either containing or not containing the transgene for an initial incubation of 3 hours at 4°C with shaking, and bound and unbound material was separated by centrifugation at 14,000 rpm for 1 min. The supernatant was again incubated with an excess of 6C6 coupled resin for 16 hours at 4°C, and again centrifuged to separate the unbound material. Material that bound during the first incubation and material that did not bind to the 6C6 coupled resin were analyzed by Western blot utilizing 8E5 and Swedish 192 antibodies (Fig. 9).
Homogenates from transgenic (+) or non-transgenic (-) mice were probed with 8E5 (panel A) or Swedish 192 (panel B).
Lanes 1 refer to total homogenate, lanes 2 refer to the fraction that did not bind to the 6C6 resin (.i.e., which is depleted of the a-secretase cleaved fragment of ~3APP), and lanes 3 refer to the fraction that bound to the 6C6 coupled resin (i.e., which retains the,a-secretase cleaved fragment o.f QAPP). None of the bound f3APP, identified by its reactivity to anti-5 antibody, cross-reacted with the Swedish 192 antibody.
Unbound material, identified by reactivity to anti-5, reacted with the Swedish 192 antibody.
Thus, the present invention provides a viable animal model for elucidating the processing of ~3APP into ~3AP and related fragments and further provides a convenient system for.
screening for inhibitors of B-secretase activity and/or for drugs that modulate f3-secretase activity.
* Trademark WO 95/11994 217 4 6 3 ~ PCT/US94/11782 Although the foregoing invention has been described in detail for purposes of clarity of understanding, it will be obvious that certain modifications may be practiced within the scope of the appended claims.

WO 95/11994 ~ PCT/US94/11782 SEQUENCE LISTING
(1) GENERAL INFORMATION:
(i) APPLICANT: ATHENA NEUROSCIENCES, INC.
(ii) TITLE OF INVENTION: Methods of Screening for Beta-Amyloid Peptide Production Inhibitors (iii) NUMBER OF SEQUENCES: 12 (iv) CORRESPONDENCE ADDRESS:
(A) ADDRESSEE: Townsend and Townsend Khourie and Crew (B) STREET: One Market Plaza, Steuart Tower, Ste. 2000 (C) CITY: San Francisco (D) STATE: California (E) COUNTRY: USA
(F) ZIP: 94105 (v) COMPUTER READABLE FORM:
(A) MEDIUM TYPE: Floppy disk (B) COMPUTER: IBM PC compatible (C) OPERATING SYSTEM: PC-DOS/MS-DOS
(D) SOFTWARE: PatentIn Release #1.0, Version #1.25 (vi) PRIOR APPLICATION DATA:
(A) APPLICATION NUMBER: US 08/143,697 (B) FILING DATE: 27-OCT-1993 (C) CLASSIFICATION:
(vii) PRIOR APPLICATION DATA:
(A) APPLICATION NUMBER: US 07/965,971 (B) FILING DATE: 26-OCT-1992 (viii) PRIOR APPLICATION DATA:
(A) APPLICATION NUMBER: US 07/868,949 (B) FILING DATE: 15-APR-1992 (ix) ATTORNEY/AGENT INFORMATION:
(A) NAME: Heslin, James M.
(B) REGISTRATION NUMBER: 29,541 (C) REFERENCE/DOCKET NUMBER: 15270-4-2PC
(x) TELECOMMUNICATION INFORMATION:
(A) TELEPHONE: 415-326-2400 (B) TELEFAX: 415-326-2422 WO 95!11994 ~ ~ PCT/US94/11782 (2) INFORMATION FOR SEQ ID NO:1:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 16 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID NO:1:
Gly Ser Gly Leu Thr Asn Ile Lys Thr Glu Glu Ile Ser Glu Val Lys 217~~02 WO 95/11994 PCT/US94/i 1782 (2) INFORMATION FOR SEQ ID N0:2:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 17 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID N0:2:
Tyr Ser Gly Leu Thr Asn Ile Lys Thr Glu Glu Ile Ser Glu Val Lys Met ~~~4~32 (2) INFORMATION FOR SEQ ID N0:3:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 6 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID N0:3:
Ile Ser Glu Val Lys Met WO 95/11994 L ~ ~ 4 E 3 2 PCT/US94111782 (2) INFORMATION FOR SEQ ID N0:4:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 8 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID N0:4:
Glu Ile Ser Glu Val Lys Met Asp 146?
(2) INFORMATION FOR SEQ ID N0:5:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 7 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID N0:5:
Cys Ile Ser Glu Val Lys Met WO 95111994 ~ PCT/US94/11782 (2) INFORMATION FOR SEQ ID N0:6:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 7 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID N0:6:
Tyr Ile Ser Glu Val Lys Met WO 95111994 2 ~ 7 4 6 3 2 PCTIUS94/11782 (2) INFORMATION FOR SEQ ID N0:7:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 5 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID N0:7:
Asp Arg Gly Leu Thr WO 95111994 ? ~ ~ PCT/US9=1111782 (2) INFORMATION FOR SEQ ID N0:8:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 10 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID N0:8:
Thr Arg Pro Gly Ser Gly Leu Thr Asn Ile WO 95111994 2 ~ 7 ~ ~ ~ ~ PCT/US9:l/11782 (2) INFORMATION FOR SEQ ID N0:9:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 10 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID N0:9:
Lys Thr Glu Glu Ile Ser Glu Val Lys Met WO 95/11994 ? 1 7 a ~ ~ ~ PCT/US94/11782 (2) INFORMATION FOR SEQ ID NO:10:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 10 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID NO:10:
Lys Thr Glu Glu Ile Ser Glu Val Asn Leu 21 l ~ 632 (2) INFORMATION FOR SEQ ID NO:11:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 5 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID NO:11:
Ser Glu Val Asn Leu WO 95/11994 L~ ~ ~ 4. ~~ ~ ~ PCT/US94/11782 (2) INFORMATION FOR SEQ ID N0:12:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 9 amino acids (B) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID id0:12:
Cys Gly Gly Glu Ile Ser Val Asn Leu

Claims (20)

WHAT IS CLAIMED IS:

1. A method for monitoring B-amyloid precursor protein (.beta.APP) processing in vivo, said method comprising detecting the presence of an amino terminal fragment of .beta.APP (ATF-.beta.APP) in a specimen from a non-human animal transformed to express the Swedish mutation of human .beta.APP, wherein the amino terminal fragment has been cleaved between Leu S96 and Asp597.

2. A method for identifying .beta.-amyloid production inhibitors, said method comprising detecting the amount of an amino terminal fragment .beta.APP (ATF-.beta.APP) cleaved between Leu596 and Asp597 in a specimen from a non-human animal transformed to express the Swedish mutation of human .beta.-amyloid precursor protein (.beta.APP) and to which a diagnostically effective amount of a test compound has been administered; and comparing the detected amount of ATF-.beta.APP with a control amount of ATF-.beta.APP produced in the absence of said test compound.

3. A method as in claim 2, wherein the control amount of ATF-.beta.APP is measured in the same non-human animal which receives the test compound or wherein the control amount of ATF-.beta.APP is an average value determined from a plurality of test non-human animals expressing the Swedish mutation of human .beta.APP.

4. A method as in any one of claims 1 to 3, wherein the presence of the ATF-.beta.APP is detected by reaction of the specimen with a diagnostically effective amount of a binding substance specific for an epitope at the carboxy terminus or ATF-.beta.APP comprising the amino acid sequence Ser-Glu-Val-Asn-Leu, optionally wherein the binding substance is a monoclonal antibody raised against a peptide comprising the amino acids.

5. A method as in any one of claims 1 to 4, wherein the specimen is cerebrospinal fluid or brain homogenate.

6. A method for detecting an amino-terminal fragment of the Swedish mutation of .beta.-amyloid precursor protein (.beta.APP) in a biological specimen, said method comprising:

exposing the specimen to a diagnostically effective amount of a binding substance which binds specifically to an epitope which comprises a C- terminal leucine596 which is exposed by cleavage of .beta.-amyloid peptide (.beta.AP) from the Swedish mutation of .beta.APP: and detecting binding between the substance and the amino-terminal fragment.

7. A method as in claim 6, wherein the binding substance is an antibody.

8. Antibodies specific for the C-terminus of the amino terminal fragment of Swedish .beta.-amyloid precursor protein.

9. Antibodies as in claim 8, raised against a peptide comprising SEVNL.

10. A method of making antibodies comprising:
raising antibodies specific for an immunogen comprising the C-terminus of the amino-terminal fragment of Swedish .beta.-amyloid precursor protein, after having exposed lymphocytes to or after having injected a non-human vertebrate with said immunogen.

11. A method as in any one of claims 1 to 3, wherein the presence of the ATF-.beta.APP is detected by reaction of the specimen with a binding substance able to discriminate between the ATF-.beta.APP and related .beta.APP fragments.

12. A method as in any one of claims 1 to 5 or 11, wherein said non-human-animal is a mouse.

13. Antibodies as in claim 8, raised against a peptide comprising residues 590-596 of the Swedish .beta.APP amino acid sequence.

14. Antibodies as in claims 8 or 13, wherein said antibodies are polyclonal antibodies.

15. A method as in claim 10, wherein said antibodies are raised against a peptide comprising SEVNL.

16. A method as in claim 15, wherein said peptide comprises residues 590-596 of the Swedish .beta.APP amino acid sequence.

17. A method as in any of claims 10, 15 or 16, wherein said antibodies are polyclonal antibodies.

18. A method as in claim 7, wherein said antibody has been raised against a peptide comprising SEVNL.

19. A method for screening test compounds for the ability to inhibit or modulate cleavage of .beta.APP, said method comprising:
administering a diagnostically effective amount of a test compound to a mouse transformed to express the Swedish mutation of human (3APP, wherein said human .beta.APP is processed to ATF-.beta.APP in a sufficient amount to be detectable in a brain homogenate of said transformed mouse, and detecting a change in .beta.APP processing in said transformed mouse as compared to said processing in the absence of the test compound.

20. A method as in claim 19, wherein said change in .beta.APP processing is detected by measuring a change in the amount of ATF-.beta.APP.