WO1990005138A1

WO1990005138A1 - Cytotoxic amyloid precursors and screening assays using them

Info

Publication number: WO1990005138A1
Application number: PCT/US1989/005041
Authority: WO
Inventors: Rachael L. Neve; Bruce A. Yankner
Original assignee: The Children's Medical Center Corporation
Priority date: 1988-11-08
Filing date: 1989-11-08
Publication date: 1990-05-17

Abstract

Neurotoxic amyloid precursor proteins (NAPP's) are produced, e.g. using recombinant DNA, and used in an assay to screen candidate compounds for their ability to antagonize neuronal toxicity. Specifically, neurons are cultured in the presence of an NAPP that has been treated with the candidate compound. The assay is useful to screen candidate therapeutics for Alzheimer's Disease. Assays for the presence of NAPP are useful for identifying and monitoring the progression of Alzheimer's Disease.

Description

CYTOTOXIC AMYLOID PRECURSORS AND SCREENING ASSAYS USING THEM

Background of the Invention

A 4.2 kilodalton (kD) polypeptide (called amyloid, the beta protein or the A4 peptide) is deposited in neuritic plaques and along the walls of the cerebral vasculature in Alzheimer's disease, in Down Syndrome, and to a lesser extent in normal aging. Apparently, amyloid is produced from a precursor (amyloid precursor protein or APP). Based on cDNA clones that have been isolated and sequenced, it appears that APP is considerably larger than the amyloid polypeptide. Also, based on the cDNA clones, various investigators have reported that APP is the product of a single gene on chromosome 21 expressing at least three RNA species.

For convenience, we have included in Figs. 1A and IB of this document the APP sequence reported by Kang et al. Nature 325:733-736 (1987). See also, Tanzi et al.. Science 235:880-884 (1987); Goldgaber et al. ,

Science 235:887-880 (1987); Robakis et al., Proc. Nat' 1. Acad. Sci. USA 84:4190-4194 (1987). Nucleotide residues are numbered in the 5' to 3' direction, beginning with the first base of the initiation codon AUG. The untranslated sequence following the poly (G) tail is indicated by negative numbers. The sequence shows a 695-residue open reading frame. The deduced amino-acid sequence is numbered, and the amino-terminal methionine included. The amino-acid sequence of the A4 polypeptide is boxed. Zigzag underline is a probable membrane-spanning sequence of the A4 polypeptide. The synthetic oligonucleotide mixture which Kang et al. used as probe is indicated as a line above the corresponding cDNA. Polyadenylation signals are underlined. Nucleotide 3207 is followed by a poly(dA) tail linked to vector DNA.

Summary of the Invention One aspect of the invention features screening candidate compounds for the ability to antagonize APP-related neuronal toxicity. Specifically, we have discovered amyloid precursor proteins that are toxic to neuronal cells, and those proteins can be used to evaluate candidates which antagonize that toxicity. We use the term neurotoxic amyloid precursor proteins (or "NAPP's") to include any APP and any APP fragment which kills cultured neuronal cells, as described in greater detail below. Specific NAPP's are described below. NAPP's also include conservative substitutions, deletions or additions to the APP sequences described below. Neuronal cells include both primary neurons and cells which can be made to differentiate into a neuronal cell type under differentiating conditions. The assay specifically features culturing a neuronal cell in the presence of an NAPP which has been treated with the candidate compound, and evaluating the toxicity of the treated NAP? to the neuronal cell culture. The neuronal cell may be cultured in the presence of both the NAPP and the candidate compound. For example, the neuronal cell can be a cell that has been genetically altered by transformation with an expression vector encoding the NAPP, so that the neuronal cell, which produces NAPP in vivo, is treated with the candidate compound. Alternatively, the NAPP (or a mixture containing the NAPP) can be treated with the candidate compound, after which the candidate compound is separated from the NAPP or the treated mixture, and the neuronal cell is cultured in the presence of the treated NAPP.

While we do not wish to bind ourselves to a particular theory, it appears that NAPP's generally include the amyloid polypeptide sequence, and they lack the serine protease inhibitor domain shown in Fig. 3. Specifically, Fig. 3 is a restriction map of the APP coding region taken from commonly owned Neve, USSN 154,236, cited above. The vectors depicted in Fig. 3 ^' are described in greater detail in that application. The sequence of the serine protease inhibitor domain is shown as an insert. The relationship between the insert sequence in Fig. 3 and the APP sequence of Figs. 1A and IB is as follows. The VAL-VAL-ARG sequence beginning the insert of Fig. 3 constitutes residues 286,287, and 288 in Fig. 1A. Residue 289 (VA in Fig. 1A) is changed to GLU by the protease inhibitor insert. The last residues of the Fig. 3 insert are ILE-PRO-THR, the last two residues of which are residues 290 and 291, respectively of Fig. 1A. The APP-l described in Tanzi et al., Nature

33_1: 528-530 (1988) and in commonly owned USSN 154,236, referenced above, is an NAPP that can be used in the above assay. Moreover, applicants have discovered other NAPP's, e.g. those which consist of APP that has been truncated at the amino terminus. Specifically, the truncation leaves the amyloid polypeptide sequence and at least 10 residues of the APP to the amino terminus side of the amyloid polypeptide sequence. For example, the truncation can be at ARG578 or at GLU590. While the carboxy terminus of APP also can be truncated, preferably the NAPP includes at least 10 residues of the APP sequence to the carboxy terminal side of the amyloid polypeptide. It is not necessary to truncate the APP at all at the carboxy terminus—i.e., the NAPP includes the APP sequence extending from the amyloid peptide sequence all the way to the carboxy terminus of APP.

A third aspect of the invention features substantially isolated DNA encoding the above-described NAPP's, and a fourth aspect of the invention features host cells transformed with that DNA which, when _^expressed, produces those NAPP's.

Finally, the invention features assaying samples for the presence of NAPP's, e.g. as part of a diagnostic assay of a bodily fluid extract.

The above described assays also can be used to evaluate or screen the efficacy of potential Alzheimer of Downs Syndrome therapeutics, or to evaluate the mechanism responsible for neurotoxicity. Description of the Preferred Embodiment

I. Brief Description of the Figures

Figs. 1A-1D are a nucleotide sequence of APP-encoding DNA.

Fig. 2 is a diagram showing DNA encoding APP in relation to various segments of APP, including ABl and ADl.

Fig. 3 is a diagram showing a serine protease inhibitor domain present in one version of APP. II. Obtaining Neurotoxic Amyloid

Precursor Proteins; Neurotoxicity

The preferred NAPP's described above generally can be produced by known techniques of recombinant DNA construction and expression. See, generally, Maniatis et al., Molecular Cloning, Cold Spring Harbor Lab. 1982. As described above, various groups have cloned DNA expressing amyloid precursor proteins. See Kang et al., Tanzi et al . (1987), Goldbaber et al., and Robakis et al. cited above. Preferred toxic APP fragments can be obtained by truncating the N-terminal-encoding end of the APP gene. As shown in Figs. 1A-1D, the APP-encoding region of the human genome is less than 3.0 kb. , and DNA encoding toxic APP fragments can be obtained from about l .0 kb of the C-terminal encoding portion of that APP-encoding DNA.

Fig. 2 shows two specific neurotoxic APP-encoding fragments, ABl and ADl, which are described in greater detail below. The precise terminal points of the APP fragments can be varied, however, without substantially altering the toxic effect of the fragment. One skilled in the field therefore will readily understand that our discovery applies to APP fragments other than those described below. For example other restriction sites can be selected which will preserve the amyloid-encoding sequence (see Fig. 1C) without including the serine protease inhibitor domain described by Tanzi et al . (1988) and USSN 154,236 cited above. Having obtained the desired NAPP-encoding fragment, standard techniques can be used to construct an expression vehicle for transforming a selected host cell which can then be used to produce the NAPP, Suitable host cells include bacteria, yeast, mammalian and insect cells. One particular system which is provided by way of example only and not as a limitation are pEx, sold by Genofit, Geneva, Switzerland, or the vectors described by Stanley and Luzio (1984) EMBO J. 3(6) :1429-1434. NAPP's can be produced using vaccinia expression systems. NIH 3T3 fibroblast cells also can be used.

The transformed cells can be cultured by known techniques and the NAPP can be purified by known techniques, including affinity purification using antibodies to NAPP- raised by standard techniques, e.g. those described in USSN 154,236 or using immobilized proteins to which the NAPP can bind, e.g., serine protease inhibitor linked to agarose from Sigma, St. Louis, Missouri.

The following Examples describe cloning of DNA encoding APP. The assays described in parts III and IV, below, illustrate expression of APP in various cells. They also illustrate methods that can be used to screen candidate APP fragments for neurotoxicity.

Example 1 ABl initiates at nucleotide 1769 of the APP sequence of Fig. 2, 17 basepairs upstream of the MET codon immediately preceding the amyloid* protein-encoding portion of APP. ABl extends to bp 2373, 286 base pairs beyond the APP stop codon. Thus ABl encodes an APP fragment which has the entire amyloid protein sequence at its N terminus, and which includes the remaining C-terminal portion of APP. Throughout this document, we use the numbering system of Figs. 1A and IB. By that system, the APP-l full-length clone of Fig. 2 extends from -95 to 2953.

ABl was constructed from a full-length APP-l cDNA which was: a) cleaved with Bgl II and Xmn I; b) ligated to Xho I linkers to convert the blunt Xmn I end to to an Xho I cohesive end; and c) inserted into the BamHI/Sal I-digested of a varient of the vector DO, as disclosed in Korman et al. , Proc. Nat' 1. Acad. Sci. USA 84:2150 (1987).

Example 2 ADl also begins at nucleotide 1769, and it extends to base pair 1959, 45 base pairs downstream of the termination of the amyloid peptide encoding region. To construct ADl, the full-length APP1 cDNA was cleaved with Bglll and Hindi, and ligated to DOJ that had been cleaved with BamHI . III. Screening For NAPP Antagonists

The neurotoxic amyloid precursor proteins described above generally can be used in an assay that screens for antagonists to neurotoxicity. As summarized above, neuronal cells are exposed to NAPP that has been treated with the potential antagonist, and the toxicity of the treated NAPP to the neuronal cells is evaluated. It is possible with some antagonists to expose cultured neurons simultaneously to the NAPP and the antagonist. Ordinarily, however the NAPP will b treated with the antagonist separately* e.g. by passing the NAPP over antagonist immobilized to a separate phase. The resulting treated NAPP can their be added to the medium used to culture the neurons.

The NAPP can be provided as a purified protein or as conditioned medium from a transfected cell as described in Part II, above. The antagonist candidates can be any candidate for treating Alzheimer's Disease. One category of drugs to be assayed includes serine protease inhibitors, which generally have been well studied. See, e.g., Billings et al., Proc. Nat' 1. Acad. Sci. 84:4801-4805. Specifically, cleavage involving SER 604 (within the amyloid peptide sequence) of APP is implicated as an active site in the improper APP processing.

Preferred primary neuronal cells include those which are most affected by plaques in Alzheimer's Disease, such as hippocampal cells and neocortical cells. Other primary neurons (e.g. dorsal root ganglial cells) can be used.

Immortalized cells can be used, e.g. those which can be exposed to conditions which cause them to differentiate into neuronal cell types. Among the suitable immortalized cells are fibroblast tumors, neuroblastoma and embryonial carcinoma cells. Differentiating agents can be used such as nerve growth factor (NGF) (Collaberative Research); fibroblast growth factor (FGF) (Collaberative Research); cyclic AMP; and phorbol esters. Serum deprivation will provoke some cells such as neuroblastomas to differentiate into neuronal cell types. The neuronal cells can be cultured by known techniques generally as described below.

Examples of neuronal cell cultures that can form the basis of the- assay are given below.

Example 3 Neurotoxicity of ABl in PC12 Cells

The construction of Example 1 (AB1/DOJ) was transfected into PC12 cells by standard techniques, e.g. the calcium phosphate method of Chen et al. Mol. Cell Biol. 1:2745-2752 (1987) and Graham et al. Virology 52:256-467 (1973). Transfected cells were selected with neomycin analogue G418 (Sigma, St. Louis, Missouri) and resistant colonies were subcloned and expanded.

Cells expressing the transfected contructs synthesize, under the control of the 5' Moloney Murine Leukemia Virus Long Terminal Repeat (MoMuLV LTR) , a fusion mRNA transcribed from the cDNA insert and the neomycin resistance gene (4.1 kb) which are then translated as separate polypeptides by the general technique of Korman et al. , Proc. Nat' 1■ Acad. Sci. USA 8_4:2150 (1987). Internal MET 596 initiates transcription. PC12 cells were cultured by the general technique of Greene et al. , Proc. Natl. Acad. Sci ■ USA 73:2424-2428 (1976) in Dulbecco's Modified Eagle Medium (DMEM) containing 10% fetal calf serum (FCS) and 5% fetal horse serum (HS) . βNGF was added at a concentration of 50 ng/ml with a change of medium and 3NGF every two days.

Transfected cells initially were similar in morphology and viability to non-transfected cells. The neurotoxicity of ABl was observed when the cells were treated with NGF, which causes PC12 cells to differentiate into a neuronal cell type. Graham et al., cited above. The general technique of Greene et al . , Proc. Nat'1. Acad. Sci. (USA) 73:2424-2428 (1976) was used, as described below.

Addition of NGF to the ABl-transfected clones Bll. and B17 resulted in the initial outgrowth of small neuritic processes. However, after 3 to 4 days there was a clear deviation from the normal course of differentiation with a slowing of neurite outgrowth and the development of vacuolar inclusions. This was followed by swelling and granularity of the soma and subsequent cell death. After 7 days in NGF, 50-75% of the cells had died (Table 1).

The degeneration of ABl-expressing cells was NGF-dependent and was greatest during the period in which maximal process outgrowth normally occurs, i.e. 5-8 days. Controls were: PC12 cells transfected with AS1 (Fig. 1) encoding APP; PC12 cells transfected with DOJ alone; and normal PC12 cells.

Specifically, normal PC12 cells and PC12 cells transfected with ABl (clones Bll and B17) were plated on

₄ polylysine-coated wells at a density of 2x10 cells/well. Cells were incubated in the absence (a) or presence (b) of NGF for 8 days and then dissociated in trypsin and incubated with trypan blue for 5 min. Viable cells which excluded trypan blue were counted in a hemocytometer. Each value represents the mean of 8 determinations. Although there may have been a small difference in the growth rate of Bll and B17 relative to PC12 in the absence of NGF, there was not apparent evidence of cell death or degeneration. In the presence of NGF, the Bll and B17 cultures showed clear evidence of cellular degeneration and death whereas PC12 and S16 did not.

Table I - Viability of Transfected PC12 Cell Lines After Treatment With NGF

# Viable Cells/Well (x!0^~3)

Cell Line - NGF<^a> . + NGF<^b>

PC12 145 58

S16 128 65 Bll 110 30

B17 108 15

Example 4 Neurotoxic Conditioned Medium/PC12 Cells The ABl-transfeeted PC12 cells of Example # were grown as described in Example 3. Then normal untransfected PC12 cells (after 8 days in NGF) were treated as described below with conditioned medium from the ABl-transfected PC12 cells. Conditioned medium was prepared by incubating subconfluent undifferentiated cultures with DMEM supplemented with 2 mM glutamine and 5 μg/ml insulin for 48 hours and then withdrawing the medium and filtering it.

For 4-7 days after treatment with conditioned medium, the normal PC12 cells showed gradual retraction of neuritic processes and cellular degeneration. Controls were undifferentiated PC12 (i.e. without NGF) and the use of conditioned medium from ASl-transfected

PC12 cells on NGF-treated PC12 cells.

Example 5

Neurotoxic Conditioned Medium/

Primary Neuronal Cultures

Rat hippocampal cultures were treated with conditioned medium prepared as described in Example 4.

Specifically, primary hippocampal cultures were established by dissecting hippocampi from E18 rat fetuses and incubating them for 30 min. with 0.03% trypsin in Ca/Mg-free Earles salt solution and then for 20 min. in Earles salts. After trituration, cells were plated in 24 well plated on polylysine coated glass coverslips in DMEM supplemented with 2 mM gluatimine, 20 mM KC1, 1 mM pyruvate, 0.5% glucose and 10% FCS. After 24 hr. , media was replaced with DMEM containing the N2 supplements (see Boltenstein et al. , Proc. Natl ■ Acad. Sci. USA J_6:514-517 (1979)) with or without 50% of conditioned medium. In some experiments with conditioned medium, cells were maintained in the FCS-containing medium. Experiments were terminated after 72 hr. in culture (48 hr. with conditioned medium) by fixation of the cells in buffered 4% paraformaldehyde. Neuronal viability, as assessed by trypan blue exclusion, phase brightness and intactness of neurites, was determined in 10 fields of 4 sister cultures for each condition. Prior to these experiments, rat hippocampal cultures were analyzed immunocytochemically with an antibody to the neuron-specific 68 kD neurofilament protein to verify the morphological identification of neuronal cells. After 66 hours virtually all neurons treated with B17 conditioned medium were dead.

Example 6 Antagonists of Neurotoxicity

The procedure of Example 5 was repeated, using conditioned medium that was first immunoabsorbed with antibody M4, an antibody to murine APP-l. Specifically, antibody M4 was prepared by injecting 350μg of an uncoupled peptide representing amino acids 597-611 of the murine APP1 (i.e., the first 15 amino acids of the murine homologue of the amyloid polypeptide) into rabbits. Second and third injections of 200 μg each were done at 4 and 6 weeks. The antibody was affinity purified on a peptide column. Immunoblot analysis with M4 antibody revealed a prominent, positively reacting band at 135 kD in membrane fractions of murine and human brain homogenates; preabsorption of the antibody with the peptide to which it was made greatly reduced this immunoreactivity.

Immunoabsorption of the conditioned medium with M4 was performed as follows. M4 was applied to polystyrene immunoassay plates at a concentration of 10 μg/ml for 2 hrs. Following blockage of nonspecific sites with 5% BSA in PBS for 2 hrs., the wells were washed with PBS and the conditioned medium was applied for 3 hrs., and then to a second set of antibody-treated wells for an additional 3 hrs. As a control, conditioned medium was treated in a parallel fashion in wells blocked with BSA but not pretreated with antibody M4. All treatments were carried out at room temperature, after which the medium was filter sterilized and applied to hippocampal cultures. The immunoabsorption was confirmed by immunoblot analysis. ABl-conditioned medium, treated with M4 antibody was added to hippocampal neurons as described in Example 5. After 66 hours cultures treated with immunoabsorbed conditioned medium were similar in appearance to untreated cultures.

Example 7 Conditioned Medium From 3T3-AB1 Cells

NIH 3T3 cells were transfected with AB1/DOL, a DO vector varient, as by the techniques described above. While expression of ABl was not lethal to 3T3 cells, the conditioned medium of such cells was lethal to rat hippocampal cultures, using the technique of Example 5.

Other Embodiments IV. Assay For Presence Of NAPP

As outlined above, a sample of bodily fluid (e.g. blood, urine, cerebrospinal fluid obtained by lumbar puncture, or a tissue sample,^" such as a brain biopsy) can be assayed for the presence of NAPP, e.g. to determine proclivity for plaque development or to evaluate progress of a disease state.

Ordinarily, a sample will be treated first to remove unwanted contaminants. For example, blood could be passed over an affinity medium such as serine protease inhibitor coupled to agarose (Sigma) or other affinity material including standard immunoaffinity material using anti-APP antibody. The fraction enhanced for NAPP can then be exposed to cultured neuronal cells and the neurotoxicity compared to standards.

Other embodiments are within the following claims.

Claims

1. An assay to screen a candidate compound for its ability to antagonize neuronal toxicity, said assay comprising, a) culturing a neuronal cell in the presence of a neurotoxic amyloid precursor protein (NAPP) which has been exposed to said candidate compound; and b) evaluating toxicity to said neuronal cell to said treated NAPP.

2. An assay according to claim 1 in which said neuronal cells are cultured in the presence of said NAPP and of said candidate compound.

3. An assay according to claim 1 in which first a mixture comprising said NAPP is treated with said candidate compound, then said candidate compound is separated from said mixture, and finally said neuronal cell is cultured in the presence of said treated mixture.

. An assay according to claim 1 in which said neuronal cell is a primary neuronal cell.

5. An assay according to claim 4 in which said primary neuronal cell is selected from the group consisting of hippocampal, neocortical and dorsal root ganglial cells.

6. An assay according to claim 1 in which said neuronal cell is a undifferentiated cell treated with a neuronal differentiating conditions.

7. An assay according to claim 6 in which said undifferentiated cell is selected from the group consisting of neuroblastoma, embryonial carcinoma and fibroblast tumor cells.

8. An assay according to claim 6 or 7 in which said differentiating conditions are selected from the group consisting of: a) NGF; b) FGF; c) cyclic AMP; d) phorbol esters; and e) serum deprivation.

9. An assay according to claim 1 in which said neurotoxic amyloid precursor protein includes the amyloid sequence and lacks serine protease inhibitory activity.

10. The assay of claim 9 in which said neurotoxic amyloid precursor protein comprises at least 10 residues of the APP sequence at the carboxy terminus of the amyloid sequence.

11. The assay of claim 9 or claim 10 in which the neurotoxic amyloid precursor protein comprises at least 10 residues of the APP sequence at the amino terminus of the amyloid sequence.

12. The assay of claim 11 in which the neurotoxic amyloid precursor protein comprises APP whose amino terminus is truncated at a residue from ARG578 to GLU590, inclusive.

13. The assay of claim 12 in which the neurotoxic amyloid precurosr protein comprises an APP sequence extending from the amyloid peptide to the carboxy terminus of APP.

14. Substantially isolated DNA encoding a neurotoxic amyloid precursor protein which consists of APP truncated at the amino terminus.

15. Substantially isolated DNA according to claim 14 in which said neurotoxic amyloid precursor protein: a) includes the amyloid sequence; and b) lacks serine protease inhibitory activity.

16. Substantially isolated DNA according to claim 15 in which said neurotoxic amyloid precursor protein comprises at least 10 residues of the APP sequence at the carboxy terminus of the amyloid sequence.

17. Substantially isolated DNA according to claim 15 or 16 in which the neurotoxic amyloid precursor pxotein comprises at least 10 residues of the APP sequence at the amino terminus of the amyloid sequence.

18. Substantially isolated DNA according to claim 17 in which the neurotoxic amyloid precursor protein comprises APP whose amino terminus is truncated at a residue from ARG578 to GLU590, inclusive.

19. Substantially isolated DNA according to claim 18 in which the neurotoxic amyloid precurosr protein comprises an APP sequence extending from the amyloid peptide to the carboxy terminus of APP.

20. A host cell transformed with the substantially isolated DNA of claim 14, claim 15, claim 16 or claim 18.

21. A host cell transformed with the substantially isolated DNA of claim 17.

22. A host cell transformed with the substantially isolated DNA of claim 19

23. A neurotoxic amyloid precursor protein which consists of APP truncated at the amino terminus.

24. A neurotoxic amyloid precursor protein according to claim 23 which: a) includes the amyloid sequence; and b) lacks serine protease inhibitory activity.

25. A neurotoxic amyloid precursor protein according to claim 24 which comprises at least 10 residues of. the APP sequence at the carboxy terminus of the amyloid sequence at the caroboxy terminus of the amyloid sequence.

26. A neurotoxic amyloid precursor protein according to claim 24 or 25 which comprises the neurotoxic amyloid precursor protein comprises at least 10 residues of the APP sequence at the amino terminus of the amyloid sequence.

27. A neurotoxic amyloid precursor protein according to claim 24 which comprises the neurotoxic amyloid precursor protein comprises APP whose amino terminus is truncated at a residue between ARG578 and GLU590, inclusive.

28. A neurotoxic amyloid precursor protein according to claim 27 which comprises the neurotoxic amyloid precurosr protein comprises an APP sequence extending from the amyloid peptide to the carboxy terminus of APP.

29. An assay for the presence of NAPP in a sample comprising subjecting said sample to antibody to NAPP and detecting binding of said antibody to said NAPP.

30. A therapeutic composition comprising an antagonist to NAPP in a pharmaceutically acceptable vehicle.

31. The therapeutic composition of claim 30 wherein said antagonist to NAPP is a serine protease inhibitor.

32. A method for counteracting neuronal toxicity comprising administering the therapeutic composition of claims 30 or 31 to a patient at risk for development of APP-related neuronal plaque.