WO1994012627A1

WO1994012627A1 - Transgenic animal model for alzheimer's disease

Info

Publication number: WO1994012627A1
Application number: PCT/US1993/011480
Authority: WO
Inventors: Richard W. Scott; David S. Howland
Original assignee: Cephalon, Inc.
Priority date: 1992-11-25
Filing date: 1993-11-24
Publication date: 1994-06-09

Abstract

Disclosed is a method for producing animal models that display aspects of human neuropathology found in Alzheimer's disease. Specifically, the neuropathology is duplicated in transgenic non-human mammals which have incorporated genetic material in their genomes that encodes amyloid proteins under the transcriptional control of heterologous gene promoters. These animals are useful in evaluating cognitive behavior during the progression of a neuropathological condition and for testing various therapeutic procedures for treatment of the disease.

Description

TRANSGENIC ANIMAL MODEL FOR ALZHEIMER'S DISEASE Background of the Invention

Three hallmark neuropathological features of Alzheimer's disease (AD) are extracellular amyloid plaques, intracellular neurofibrillary tangles, and neuronal loss. These changes occur predominantly in the cortex and hippocampus of the human brain and a variety of neuronal populations are affected in these areas (Hamos et al.. Neurology 39:355-361, 1989; Struble et al., J. Neuropathol . Exp. Neurol . 46:567-584, 1987; Selkoe, Neuron 6:487-498, 1991). One particularly sensitive class is the group of cholinergic projection neurons of the basal forebrain whose dysfunction may contribute to the memory loss associated with AD (Davies et al.. Lancet 3:1403, 1976; Whitehouse et al., Ann . Neurol . 10:122-126, 1981). The molecular mechanisms responsible for the neuronal loss are not understood. However, the frequent appearance of dystrophic neurites surrounding extracellular plaques implicates the amyloid deposits in mediating or contributing to neuronal dysfunction. Additional evidence discussed below also supports a direct role for the amyloid plaque in neurodegeneration. There are two general types of deposits; diffuse noncompacted deposits termed preamyloid plaques and the dense neuritic deposit containing a core of amyloid fibrils surrounded by dystrophic neurites, activated microglia, and fibrillary astrocytes. The dense plaques occur mainly in the brain parenchyma and cerebral vasculature within affected regions of the brain in AD patients (Terry et al., Ann. Neurol. 10:184-192, 1981; Price et al., Drug Development Research 5:59-68, 1985). The cores of these plaques are comprised of unpaired amyloid fibrils 8 nanometers in diameter that have a cross-beta sheet structure (Glenner, N. Engl . J. Med . 302:1283-1291, 1980). The preamyloid deposits are widely distributed, often found in unaffected brain regions (such as the cerebellum) , and are rarely associated with dystrophic neurites or glia (Tagliavini et al., Neurosσ. Letters 93:191-196, 1988; Joachim et al.. Am . J. Pathol . 135:309-319, 1989). A temporal connection between preamyloid and dense amyloid deposits is indicated in Down's syndrome (trisomy of chromosome 21). Individuals with Down's syndrome that reach 40 or 50 years of age develop neuritic amyloid plaques and neurofibrillary tangles identical to those in AD. Preamyloid deposits are found in Down's patients at earlier ages (mid-teens through twenties) but dense neuritic plaques are sparse or absent (Giaconni et al., Neurosci . Lett . 97:232-238, 1989; Motte et al., Acta Neuropathol . 77:535-546, 1989). This suggests that the preamyloid deposits precede and develop into neuritic plaques over time in Down's syndrome and, most likely, AD.

The primary component of the amyloid plaque is a 28 to 43 amino acid peptide called the beta/A4 peptide or 3/A4 (Glenner et al., Biochem . Biophys . Res . Commun . 120:885-890, 1984; Masters et al., Proc. Natl . Acad. Sci . USA 82:4245-4249, 1985). The peptide is derived from a larger protein termed the amyloid precursor protein (APP) by proteolytic cleavage. APP is a membrane-associated glycoprotein encoded by a single copy gene (Kang et al.. Nature 325:733-736, 1987). There are three predominant forms, or isoforms, of the protein, 695 (APP695) , 751 (APP751) , and 770 (APP770) amino acids in length, that are generated by alternative splicing of a common RNA precursor (see Figure 1 for a schematic diagram of APP) . The APP isoforms are inserted in the membrane and are subject to at least two post-translational processing pathways. A fraction of the APP is cleaved adjacent to the membrane-spanning domain (see Figure 1 for the exact cleavage site) resulting in the release or secretion of the N-terminal portion of APP (Esch et al.. Science 248:1122-1124, 1990). The C-terminal cell-associated fragment and a large fraction of the uncleaved APP appear to undergo endocytosis and are degraded in the lysosomes (Caporaso et al., Proc. Natl . Acad. Sci . USA 89:2252- 2256, 1992; Estus et al. , Science 255:726-728, 1992). The 751 and 770 species differ from APP695 by the inclusion of a serine-protease inhibitory domain found in the family of Kunitz protease inhibitors (Ponte et al.. Nature 331:525-527, 1988; Tanzi et al.. Nature 331:528- 532, 1988). The secreted form of APP751 has been identified as a protease inhibitor called protease nexin II (Van Nostrand et al.. Nature 341:546-549, 1989). APP751 and APP770 are synthesized by a variety of neuronal and non-neuronal cell types in the brain and in peripheral tissue while APP695 is produced predominantly in neurons (Selkoe et al., Proc. Natl . Acad. Sci . USA 85:7341-7345, 1988; Golde et al.. Neuron 4:253-267, 1990) . The isoform from which 3/A4 is produced is unknown. Although APP695 is neuron-specific, levels of APP751 and APP770 relative to APP695 appear to be enriched in AD brain (Johnson et al., Science 248:854- 857, 1990). Accordingly, arguments for all three isoforms as being the 3/A4 progenitor have been presented.

The cleavage that occurs during secretion is within the 0/A4 domain between amino acid residues 16 and 17 of the 0/A4 sequence (Figure 1, Wang et al., J. Biol . Chem . 266:16960-16964, 1991). As a result, /J/A4 or so- called amyloidogenic fragments capable of producing /3/A4 must be formed by alternate processing events outside of the secretory pathway. The cellulai? protease or proteases that generate amyloidogenic fragments or the j8/A4 peptide in AD brain have not been identified. Candidate proteases include lysosomal proteases (Cataldo et al., Proc. Natl . Acad. Sci . USA 87:3861-3865, 1990; Golde et al., Science

255:728-730, 1992), the multicatalytic protease (Ishiura et al. , Neuroεc. Let . 115:329-334, 1990; Siman et al., Abs . Soc. Neurosci . 17:1070, 1991), and a calcium activated protease isolated from AD brain (Abraham et al., Biochem . Biophys . Res . Commun . 174:790-796, 1991). Also, the cellular origin of the amyloidogenic fragments or the 3/A4 peptide is not known. The wide tissue distribution of APP synthesis opens several possibilities. The amyloidogenic fragments may be produced locally in the brain by neurons or glia, in the vasculature by blood and/or endothelial cells, or in peripheral tissues and transported to the brain by the bloodstream. A neuronal origin is supported by several criteria. APP is expressed at high levels in neurons and amyloid plaques are frequently found associated with neurites and synaptic termini in the brain parenchyma. It is argued that the beta-peptide component of the plaque originates from aberrant processing of APP that accumulates in the neuronal synapse (Koo et al., Proc. Natl . Acad. Sci . USA 87i1561-1565, 1990). Alternatively, a glial role in J/A4 deposition has been proposed since plaques are also sites of active gliosis which has been shown to induce APP expression and may promote aberrant synthesis or processing of β/2.4 (Siman et al.. Neuron 3:275-285, 1989; Delacourte, Neurology 40:33-37, 1990; Itagaki et al., J. Neuroimmun . 24:173-182, 1989). The deposition of amyloid plaques in the cerebral vasculature and possibly in peripheral tissues, such as skin, supports a circulatory source for /3/A4 in AD (Joachim et al.. Nature 341:226-230, 1989). Also, individuals with the disease hereditary cerebral hemorrhage with amyloidosis-Dutch type (HCHWA-DT) develop vascular plaques in the brain (van Duinen et al., Proc. Natl . Acad. Sci . USA 84:5991-5994, 1987). These different locations of amyloid deposits in AD and HCHWA-DT suggest there may be multiple cellular origins of β/2.. .

Several findings support a direct role for 0/A4 in the neuropathology of Alzheimer's disease. As described above, dense neuritic plaques are almost always associated with dystrophic neurites. These degenerating neurites are enriched in neurofibrillary tangles, and tangle-bearing neurons often project to regions rich in amyloid plaques. A direct neurotoxic activity of the 3/A4 peptide has been described for neurons in primary cultures or in the brain following injection of purified 3/A4 preparations (Yankner et al.. Science 250:279-281, 1990; Kowall et al., Proc. Natl . Acad. Sci . USA 88:7247- 7251, 1991). A neurotrophic effect has also been described for β/2-A in primary cultures of immature neurons and this may explain the abnormal neuritic sprouting that surrounds many amyloid plaques (Whitson et al., Neuroscience Letters 110:319-324, 1990). Similar neurotoxic and neurotrophic activities have also been reported for a 100 amino acid carboxy-terminal fragment of APP (CIOO, see Figure 1) that includes the full 0/A4 sequence (Yankner et al., Science 245:417-420, 1989). These differential trophic and toxic activities attributed to 3/A4 and β/A4 fragments may be related to local concentrations of the peptides and the differentiation status of the neurons.

The genetic linkages of mutations in APP to diseases that exhibit jS/A4 pathology provide the most compelling evidence for a causative role of β/A4 in the disease process. In a subset of individuals with AD, designated as familial Alzheimer's disease (FAD) , the disease is inherited in an autosomal dominant fashion and is characterized by an early onset of symptoms. In at least four separate families with FAD, a mutation in APP at an amino acid 4 residues down from the carboxy- terminal border of the beta-peptide has been identified and linked to the disease (Goate et al., Nature 349:704- 706, 1991; Yoshioka et al., Biochem . Biophys . Res . Commun . 178:1141-1146, 1991; Chartier-Harlin et al. , Nature 353:844-846, 1991; Hardy et al.. Science 256: 184- 185, 1992) . More recently, a double mutation at two codons that immediately precede the N-terminal border of the β peptide have also been linked to FAD (Mullen et al.. New Genetics 1:345-347, 1992). In HCHWA-DT described above, all afflicted individuals carry a specific mutation in APP within the beta-peptide sequence (Levy et al.. Science 248:1124-1126, 1990). The effects of the mutations in the diseases remain unclear but a subtle alteration in the processing of APP has been postulated and this may promote production and deposition of the 0/A4 peptide.

Summary of the Invention In general, the invention features a transgenic non-human animal harboring a transgene coding for an amyloid protein, the transgene being under the transcriptional control of a synapsin gene promoter. In preferred embodiments the transgene may also contain DNA coding for one or more RNA processing signals; for example, the SV40 small t intron sequences or the SV40 polyadenylation signal. The signals may be positioned proximally and/or distally to DNA coding for an amyloid protein; for example, synapsin promoter:SV40 RNA t intron:amyloid protein DNA:SV40 polyadenylation signal or synapsin promoter:amyloid protein DNA:SV40 t intron:SV40 polyadenylation signal. In preferred embodiments the amyloid protein DNA is a precursor protein that includes sequences coding for human APP695, APP751, APP770, or fragments thereof, or sequences coding for the fragments CIOO or 3/A4. The sequences can further contain a mutation; for example, the hereditary cerebral hemorrhage with amyloidosis - Dutch Type (HCHWA-DT) mutation or the familial Alzheimer's disease (FAD) mutations.

The transgene, under the transcriptional control of the synapsin gene promoter, is expressed preferentially in the transgenic animals in neuronal cells of brain tissue, resulting in the progressive development of a neuropathological condition. The deposition of amyloid plaques, formation of intracellular neurofibrillary tangles, neuritic dystrophy, or neuronal loss in brain tissue are distinguishing features of the pathology consistent with Alzheimer's disease, an Alzheimer's-like disease, a hereditary cerebral hemorrhage with amyloidosis - Dutch Type disease or a familial Alzheimer's disease.

In preferred embodiments the transgenic non-human animal of the present invention is a mammal, and is preferably a rodent such as a mouse or a rat. The animals can serve as experimental systems for testing behavioral function and the development of a neuropathological condition. Analyses would include testing for deficits in cognitive ability during the progression of a neuropathological condition. Furthermore, the transgenic animals of the present invention are useful as experimental systems which enable an evaluation of the efficacy of treatments to attenuate or ameliorate the advancement of a neuropathological condition.

In a second aspect, the invention features a transgenic non-human animal harboring a transgene coding for an amyloid protein, the transgene being under the transcriptional control of a human cytomegalovirus gene promoter. In preferred embodiments the transgene may also contain DNA coding for one or more RNA processing signals; for example, the SV40 small t intron sequences or the SV40 polyadenylation signal. The signals may be positioned proximally and/or distally to DNA coding for an amyloid protein; for example, human cytomegalovirus promoter:SV40 RNA t intron: amyloid protein DNA: SV40 polyadenylation signal or human cytomegalovirus promoter:amyloid protein DNA:SV40 t intron:SV40 polyadenylation signal.

In preferred embodiments the transgene, under control of the cytomegalovirus promoter, is expressed preferentially in neuronal and non-neuronal cells of brain tissue and peripheral tissue, in non-neuronal cells of brain tissue and peripheral tissue, or exclusively in peripheral tissue and not in cells of brain tissue, resulting in the progressive development of a neuropathological condition. The deposition of amyloid plaques, formation of intracellular neurofibrillary tangles, neuritic dystrophy or neuronal loss in brain tissue are distinguishing features of the pathology consistent with Alzheimer's disease, an Alzheimer's-like disease, a hereditary cerebral hemorrhage with amyloidosis - Dutch Type disease or a familial Alzheimer's disease.

In preferred embodiments the transgenic non-human animal of the present invention is a mammal, and is preferably a rodent such as a mouse or a rat. The animals can serve as experimental systems for testing behavioral function and the development of a neuropathological condition. Analyses would include testing for deficits in cognitive ability during the progression of a neuropathological condition. Furthermore, the transgenic animals, expressing amyloid protein under the transcriptional control of the human cytomegalovirus gene promoter, are useful as experimental systems which enable an evaluation of the efficacy of treatments to attenuate or ameliorate the advancement of a neuropathological condition.

In a third aspect, the invention features a recombinant transgene coding for an amyloid protein under the transcriptional control of a synapsin gene promoter. In preferred embodiments the transgene may also contain DNA coding for one or more RNA processing signals; for example, the SV40 small t intron sequences or the SV40 polyadenylation signal. The signals may be positioned proximally and/or distally to DNA coding for an amyloid protein; for example, synapsin promoter:SV40 RNA t intron: amyloid protein DNA: SV40 polyadenylation signal or synapsin promoter:amyloid protein DNA:SV40 t intron:SV40 polyadenylation signal.

In a final aspect, the invention features a recombinant transgene coding for an amyloid protein under the transcriptional control of a human cytomegalovirus gene promoter. In preferred embodiments the transgene may also contain DNA coding for one or more RNA processing signals, for example, the SV40 small t intron sequences or the SV40 polyadenylation signal. The signals may be positioned proximally and/or distally to DNA coding for an amyloid protein; for example, human cytomegalovirus promoter:SV40 RNA t intron: amyloid protein DNA: SV40 polyadenylation signal or human cytomegalovirus promoter:amyloid protein DNA:SV40 t intron:SV40 polyadenylation signal.

"Transgenic" as used herein means a mammal which includes a DNA sequence which is inserted by artifice into a cell and becomes part of the genome of the animal which develops from that cell. In the transgenic animals described herein, the DNA sequence encodes amyloid protein. The transgenic animals thus produced, and their descendants, especially their homozygous descendants, are destined to develop a neuropathological condition. Any non-human mammal which might be produced by transgenic technology is included in the invention; preferred mammals include, in addition to mice and rats, cows, pigs, sheep, goats, rabbits, guinea pigs, hamsters and horses. By "transgene" is meant DNA which is partly or entirely heterologous (i.e., foreign) to the transgenic animal, or DNA homologous to an endogenous gene of the transgenic animal, but which is inserted into the animal's genome at a location which differs from that of the natural gene.

By "transgene encoding an amyloid protein" is meant DNA encoding a protein which is at least 85% homologous in amino acid sequence with a region of APP695, APP751, APP770; contains at least 42 amino acids; and causes at least one Alzheimer-like symptom in a transgeneic animal harboring the transgene. Preferred DNA includes sequences that code for human precursor proteins APP695, APP751, APP770, or fragments thereof; or sequences coding for the fragments CIOO or S/A4. The sequence can contain a mutation; for example, the hereditary cerebral hemorrhage with amyloidosis - Dutch Type (HCHWA-DT) mutation or the familial Alzheimer's disease (FAD) mutations. As is described further below, a preferred amyloid protein is a precursor protein of at least 600 amino acids, which is processed in vivo to yield a toxic amyloid fragment. Alternatively, DNA encoding a fragment can be used.

By "promoter" is meant a segment of DNA to which the transcriptional enzyme complex binds prior to initiating transcription of the gene. By "synapsin gene promoter" is meant a DNA sequence gene that controls the transcription of the synapsin gene, which is at least 85% homologous to a vertebrate synapsin gene, and is sufficient to direct transcription of an amyloid protein encoding DNA to a greater extent in neuronal tissue than in non-neuronal tissue.

By "cytomegalovirus gene promoter" is meant a promoter existing naturally in a CMV strain. In preferred embodiments the CMV promoter is a DNA sequence controlling transcription of the immediate early (IE) gene of CMV. Preferably, the promoter can direct transcription of an amyloid protein-encoding DNA in a non-tissue specific manner, e.g., at high levels in a plurality of cell types.

By "RNA processing signal" is meant a segment of DNA that when transcribed to RNA provides cues for modifying the RNA molecule. As used herein, it includes signals coding for RNA splicing (e.g., removal of intron sequences) and for polyadenylation (e.g., the addition of a poly A tail to the 3' end of a mRNA transcript) .

By "mutation" is meant an alteration of a DNA sequence which results in a change in the natural wild type amino acid sequence. These changes can arise, e.g., spontaneously by chemical energy, or by other forms of mutagenesis, by genetic information. Mutations include, e.g. base changes, deletions, insertions, inversions, translocations, or duplications.

The invention provides a method for producing a transgenic non-human animal model of Alzheimer's disease. The model is based on the over-expression of native or mutant forms of human APP to mimic two hallmark features of AD brain: extracellular deposition of _/S/A4 and neurodegeneration. Human APP is expressed specifically in neurons or constitutively in brain and peripheral tissues using chimeric transgenes encoding full-length APP protein. The use of full-length protein has the advantage that issues associated with protein processing and neuronal degeneration can be addressed and correlated in one model system. Also, APP isoforms encoding native proteins or proteins carrying the genetic HCHWA-DT and FAD mutations can be used to determine the effects of these mutations on APP processing and neuropathology in this animal system. Alternatively, expression of specific APP sub-fragments, such as the 42 amino acid beta-peptide or the related 100 amino acid C-terminal peptide (CIOO) , can be done to bypass protein processing issues and concentrate exclusively on aspects of neuronal degeneration and aberrant APP expression. The invention also features transgenic rodent lines expressing human APP, or APP derivatives, in a cell-specific or constitutive manner that will serve as a model system for Alzheimer's disease. These animals can be used to study the mechanisms responsible for the onset of AD and test the efficacy of compounds in treating the disease. One advantage of overexpressing full-length APP is that it provides a model system for testing therapeutic compounds directed at proteases involved in the production of ,5/A4 from the precursor protein for their effects on the progression of the AD-like pathology. The demonstration of efficacy is strong proof that the proteases targeted by the therapeutic compounds are involved in modulating AD pathology.

Transgenic animals overexpressing APP carrying the genetic HCHWA-DT and FAD mutations will provide animal systems to examine the effects of these mutations on AD pathology. If the mutations enhance pathology, these animals will also supply a means to identify the mechanisms which regulate development of AD neuropathology. Finally, transgenic animals overexpressing specific APP sub-fragments, including the 3/A4 or CIOO peptides, would elucidate the role of these peptides in AD neuropathology and furnish an animal system to test therapeutic compounds directed at mechanisms subsequent to protein processing that also modulate progression of AD pathology.

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

Detailed Description The drawings will first briefly be described. Drawings

Fig. 1 is a diagram of APP cDNA and protein structures with amino acid sequence of the 0/A4 domain from native and mutant genes.

Fig. 2 is the cDNA map of a clone for APP751. Fig. 3 is the construction of a full-length APP751 cDNA clone. Fig. 4A is a diagram of the PCR screen for identification of cDNA clones encoding either APP695 or APP770.

Fig. 4B is the construction of a full-length CDNA clone for APP695. Fig. 4C is the construction of a full-length cDNA clone for APP770.

Fig. 5 is the insertion of APP cDNA clones into a cytomegalovirus (CMV) mammalian expression vector.

Fig. 6 is a Western (i munoblot) analysis of APP695, APP751, and APP770 expression in transfected human 293 cells.

Fig. 7 is the construction of APP cDNA clones that carry the FAD and HCHWA-DT mutations and insertion into the CMV expression vector. Fig. 8A is of the preparation of transgene vector backbones.

Fig. 8B illustrates the insertion of human APP751 into the transgene vectors. Fig. 9 is the construction of transgene plasmids Ceph 1 and Ceph 2.

Fig. 10 is the construction of transgene plasmids Ceph 3 and Ceph 4.

Fig. 11 is the cloning of intermediate vectors for transgene synthesis containing the FAD and HCHWA-DT mutations.

Fig. 12 is the construction of transgene plasmids Ceph 5 and Ceph 6.

Fig. 13 is the construction of transgene plasmids Ceph 7 and Ceph 8.

Fig. 14 shows the structures of transforming DNA Ceph 1 through Ceph 8.

Fig. 15A is the construction of a transgene plasmid encoding the 0/A4 domain of APP linked to the 350 bp synapsin promoter.

Fig. 15B is the construction of a transgene plasmid encoding the β/ A domain of APP linked to the 4.4 kb synapsin promoter.

Fig. 15C is the construction of an APP transgene plasmid encoding the CIOO domain of APP linked to the 350 bp synapsin promoter.

Fig. 15D is the construction of an APP transgene plasmid encoding the CIOO domain of APP linked to the 4.4 kb bp synapsin promoter. Fig. 16 is the Southern blot analysis of genomic DNA samples from Ceph 1 Fl mice.

Fig. 17A is the diagram for the RNA PCR screen to detect transgene expression. Fig. 17B is the Southern blot analysis of the PCR RNA screen on brain and-heart samples from selected Ceph 1 Fl mice.

Fig. 18 is the RNA protection probe to detect transgene mRNA.

Fig. 19A is the RNA protection results with brain RNA samples from selected transgenic Fl animals.

Fig. 19B shows the quantitation of the RNA protection results by densito etric scans. Fig. 20 is a protein blot showing APP751 protein expression in transgenic mouse brain.

Fig. 21A shows the results of a DNA slot blot hybridization analysis of the Ceph 8 transgenic rat line (founder (Fo) animals) . Fig. 2IB shows the results of a DNA slot blot hybridization analysis of the Ceph 8 transgenic rat line (progeny (Fl) animals) .

Fig. 22 shows the results of a PCR screen of each Ceph 8 transgenic rat line disclosed in Figures 21A and 21B.

Fig. 23 shows the results of a RNA expression assay using RNAse protection. Transσenes

Neuron-specific expression is directed preferably by the rat synapsin promoter, and constitutive expression is directed by the major early promoter of cytomegalovirus (CMV) . Neuron-specific expression is preferred in order to localize APP over-expression to pertinent cell types and minimize potential toxicity related to widespread and unregulated expression of APP that would preclude the establishment of transgenic animals. However, since the neuron may not be the cellular origin of ,9/A4, a more generalized pattern of APP expression would be required for someone to develop an animal model of amyloid deposition. The synapsin promoters are described in Howland et al., Mol . Brain Res . 11:345-353, 1991 and Thiel et al., Proc. Natl . Acad. Sci . USA 88:3431-3435, 1991. Proper control of transcription by the synapsin promoter is reflected by preferential RNA synthesis in neurons, or neuronal-like cells, relative to non-neuronal cell types. There appear to be several DNA elements involved in the proper regulation of eukaryotic gene expression and these can be divided into two categories: basal promoter elements responsible for efficient initiation of RNA synthesis located immediately upstream of the synapsin coding sequences and tissue-specific regulatory elements responsible for directing RNA synthesis in a subset of cell types. For the synapsin gene, these latter elements can be positive elements that promote efficient RNA synthesis in neurons or negative elements that prevent synapsin RNA synthesis in non-neuronal tissue. Inclusion or exclusion of these regulatory sequences in transgene constructions would be expected to dramatically influence the level and tissue distribution of transgene expression. Both positive and negative regulatory elements have been defined in the rat synapsin gene using DNA transfection assays in cell culture systems and are located within a 4.5 kilobase segment upstream of the synapsin coding sequences. The preferred embodiment of this invention utilizes two rat synapsin promoter fragments, 350 bp and 4.4 kb in size, that flank and span the 5'-terminus of the synapsin gene and direct preferentially RNA synthesis in neuronal-like cells in cell culture expression systems (Howland, et al., supra) . It is expected that analogous synapsin promoter fragments from other species or differing in size but containing essential transcriptional regulatory sequences found in the preferred fragments will also direct proper cell- specific expression. The major immediate early promoter of CMV is well characterized and directs gene expression at high levels in many cell types (Boshart et al., Cell 41:521-530, 1985) . Efficient and widespread expression of several chimeric genes in transgenic mice was also successfully obtained with the CMV promoter (Schmidt et al., Mol . Cell . Biol . 10:4406-4411, 1990); Furth et al. , Nuσ. Acids Res . 19:6205-6208, 1991). The CMV promoter has been linked to APP coding sequences and it efficiently directs APP expression in several animal cell lines in culture (Weidemann et al. Cell 57:115-126, 1989; Oltersdorf et al., J. Biol . Chem . 265:4492-4497, 1990). Efficient and widespread expression of several chimeric genes in transgenic mice was also successfully obtained with the CMV promoter (Schmidt et al., Mol . Cell . Biol . 10:4406- 4411, 1990; Furth et al.. Nucleic Acids Res . 19:6205- 6208, 1991) but the expression of CMV/APP chimeric genes in transgenic mice has not been reported. The preferred embodiment of the present invention utilizes a 637 bp fragment that flanks and spans the 5' terminus of the IE gene. It is expected that similar CMV IE promoters differing in size but containing essential transcriptional regulatory sequences found within the preferred 637 bp fragment would have analogous activities in transgenic animals.

APP coding sequences are derived from full-length cDNA clones isolated from a cDNA library synthesized using mRNA from human temporal cortex. Full-length clones for APP695, 751 and 770 have been prepared and utilized in transgene constructions. Coding sequences for APP subfragments, such as the 3/A4 peptide and the CIOO fragment, are derived from the full-length cDNA clones by standard molecular biology techniques (see, e.g., Ausubel et al.. Current Protocols in Molecular Biology, John Wiley & Sons, New York, 1989; herein incorporated by reference) . In addition, incorporation of specific mutations into APP has been accomplished by site-directed mutagenesis to determine the effect of these alterations on 3/A4 processing and neuropathology in transgenic animals. The mutations include the FAD and HCHWA-DT genetic mutations that have been linked to AD or AD-like diseases in man.

All APP coding sequences in the transgenes are fused to RNA processing signals derived from the SV40 virus. It has been shown in numerous systems that the SV40 signals can supply the information required for RNA splicing (t splice) and polyadenylation (pA) events which are essential for efficient RNA synthesis and accumulation. The location of the RNA splice sequences in some unrelated transgenes dramatically influences the level of transgene expression in the animals (Huang et al., Mol . Cell . Biol . 10:1805-1810, 1990; Palmiter et al., Proc. Natl . Acad. Sci . USA 88:478-482, 1991). In view of this information, the SV40 splice sequences have been inserted either upstream or downstream of the APP coding sequences in several transgene templates.

Transgenic mice

Mouse lines over-expressing the APP transgenes will be selected for the occurrence of AD-like pathologies. The transgenic mice are established by injection of transgene DNA into the male pronucleus of a single cell embryo (Wagner, T.E. and Hoppe, P.C. U.S. Patent 4,873,191). The injected embryos are implanted into pseudo-pregnant females and allowed to develop to term. At three to four weeks of age the animals are weaned and analyzed for transgene DNA by standard immunoblot and DNA detection techniques (see, e.g., Ausubel et al., supra) . Transgene-positive animals (Fo or founder animals) are mated with non-transgenic animals and the progeny (Fl animals) are screened for transmission of the transgene DNA. The Fl transgene-positive animals are hemizygous for the transgene DNA and homozygous transgenic animals can be generated by parent-progeny or sibling matings. Each founder animal and its transgenic progeny are unique in comparison to other transgenic mice established with the same transgene. Integration of the transgene DNA into the mouse genomic DNA is random and the site of integration can profoundly effect the levels, and the tissue and developmental patterns of transgene expression. Consequently, a number of transgenic lines are usually screened for each transgene to identify and select animals with the most appropriate expression profiles.

Transgenic lines are evaluated on levels of transgene expression and neuropathology. Expression at the RNA level is determined initially to identify and quantitate expression-positive animals. Standard techniques for RNA analysis are employed and include PCR amplification assays using oligonucleotide primers designed to amplify only transgene RNA templates and solution hybridization assays using transgene-specific probes (see, e.g., Ausubel et al., supra) . The RNA- positive animals are then analyzed for protein expression by Western immunoblot analysis using APP-specific antibodies (see, e.g., Ausubel et al., supra) . In addition, in situ hybridization and immunocytochemistry can be done using transgene-specific nucleotide probes and antibodies, respectively, to localize sites of expression within transgenic tissue and determine the correlation between expression and neuropathology. Since the conditions for transgene expression that lead to neuropathology are unknown, selection of transgenic lines expressing human APP over a range of levels is necessary.

Neuropathological analyses of transgenic mouse tissue is done using procedures that have been successfully used on human AD brain tissue (Papasozmenos, Laboratory Investigation 60:123-137, 1989; Kitamoto et al. Laboratory Investigation 57:230-236, 1987; Ikeda et al. Acta Neurol . Pathol . 78:137-142, 1989). Mouse lines are analyzed at various ages for parenchyma amyloid plaques, vascular amyloid deposits, neuritic dystrophy, and neuronal loss. Transgenic lines exhibiting significant amyloid pathology will be characterized for behavioral function to determine if the appearance of AD- like pathology is correlated with deficits in cognitive abilities; for example, by using tests such as the

Morris water maze (Dekker et al. Neuroscience 48:111-119, 1992) and radial arm maze (Murray et al., Psychopharmacology 89: 378-381, 1986).

The following examples are intended to illustrate but not to limit the invention.

Example 1. Cloning and Characterization of Human APP695. 751 and 770 cDNAs.

A cDNA clone containing the entire coding region of APP751 was constructed from two partial but overlapping cDNA clones isolated from a human temporal cortex library (catalogue number 935205, Stratagene, La Jolla, CA) . This is a λZAP bacteriophage library prepared from polyA⁺ RNA isolated from brain tissue of a two year old female. Using standard molecular biology techniques (see, e.g., Ausubel et al., supra) , approximately lxlO⁶ recombinant bacteriophage were screened for the presence of APP sequences by hybridization with a small, radiolabeled APP cDNA fragment termed amy-7, the cDNA clone FB68L (Tanzi et al. Science 235:880-884, 1987) that contains carboxy-terminal coding and 3' untranslated sequences from APP mRNA (Figure 2) . The clone, 1-3-1, was identified, isolated, and verified as a partial cDNA for APP751 by standard restriction enzyme digestion assays and direct nucleotide sequencing (see, e.g., Sanger et al., Proc,. Natl . Acad. Sci . USA 74:5463-5467, 1977). To identify larger clones that cover the N-terminal coding regions of APP, the 5' fragment of 1-3-1 that extends between the EcoRI restriction site of 1-3-1 at position +988 to the EcoRI restriction site at position 1,963 was isolated (amy

13IN) and used as a hybridization probe in a re-screen of 10⁶ bacteriophage from the Stratagene (La Jolla, CA) temporal cortex library. A duplicate screen of the same bacteriophage was also done using a small oligonucleotide probe (RS-1; 5'dCTGCCCGGTTTGGCACTGCT3'; SEQ ID NO: 1) homologous to the extreme N-terminal coding region of APP beginning at nucleotide +4. This combined screen identified two additional partial cDNA clones, C5-1 and C6-1, that overlap and span the C-terminal and N-terminal coding regions of APP, respectively. The identity of these clones was confirmed by restriction enzyme analysis and nucleotide sequencing according to standard methods (see, e.g., Ausubel et al, supra) . The cDNA clones were sub-cloned from the λZAP bacteriophage according to a standard recombination procedure (Stratagene, La Jolla, CA) . This manipulation yields cDNA clones inserted in the SK (Stratagene, La Jolla, CA) bacterial plasmid vector (C5-1/SK and C6-1/SK) . A full-length APP cDNA clone was constructed by joining C5-1/SK and C6-1/SK at the Xhol site within APP at position 1078 that they both share (Figure 3) . This clone, designated flAPP751, encodes human APP751 (contains the Kunitz-inhibitor domain) and has 7 bp of 5' untranslated sequences and 675 bp of 3' untranslated sequences. Full-length APP695 and APP770 cDNA clones were constructed by subcloning the alternately spliced regions specific for each subtype into the flAPP751 backbone. In the screen of the human temporal cortex cDNA library described above a total of 36 clones were identified with the amy 13IN probe that overlaps the KPI domain. To identify clones that encode APP695 and APP770 a polymerase chain reaction (PCR) protocol was designed and is illustrated in Figure 4A. Oligonucleotide fragments that flank the KPI domain at nucleotide +865 but anneal to all APP isoforms were used as primers for the PCR reaction that synthesizes (or amplifies) the DNA sequences between the primers. The PCR analysis was done using bacteriophage DNA preparations from the 36 positive clones (see, e.g., Gussow et al., Nucleic Acids Res . 17:4000, 1989). The expected size fragments of 376 bp for APP695 that lacks the KPI domain and 601 bp for APP770 that has a 57 bp extension of the KPI domain were obtained and the bacteriophage harboring the positive clones were plaque purified. Clones of the bacteriophage inserts in the SK plasmid vector were obtained by the recombination procedure and are designated as C6-4.1 for the APP695 clone and C8-1.1 for the APP770 clone (Figures 4B and 4C, respectively) . Restriction enzyme analysis demonstrated that neither clone encoded full-length APP. Therefore, a strategy of "cut and paste" was designed to assemble full-length APP695 and APP7770 cDNAs using the C6-4.1 and C8-1.1 partial clones and the C6-1 and C5-1 partial clones of APP751. Briefly, to obtain full-length APP695 (Figure 4B) the 860 bp EcoRI insert of C6-4.1 was isolated and partially digested with Taql under conditions that promoted cleavage at the Taql site at position +861 (APP695 coordinates) but not at the Taql sites at positions +911 and +1310. This is accomplished by using limiting amounts of enzyme (1 unit/1 ug DNA) for brief periods of time (1 to 5 minutes) at 65°C prior to stopping the reaction with 20 mM EDTA. All DNA fragments are isolated by resolving the digestion products according to size on agarose gels, identifying the DNA fragments by ethidium bromide staining, and solubilizing the pertinent gel fragments in Nal using GeneClean® (Bio 101, Inc., La Jolla, CA) . The 660 bp Taql/EcoRl was isolated after preparative gel electrophoresis and ligated with the 1007 bp Sacl/Taql fragment from C6-1 and the 3 kb Sad/EcoRI SK- plasmid (Stratagene) fragment. Following DNA transformation of competent MC1061 E. coli (see, e.g.. Dower et al.. Nucleic Acids Res . 16:6127- 6145, 1988) , plasmid DNA was isolated from ampicillin- resistant bacteria (see, e.g.. Holmes et al.. Anal.

Biochem . 114:193-197, 1981) and analyzed by restriction enzyme analysis. The resultant plasmid (termed C6-1/695) is analogous to C6-1 except it lacks the KPI domain and encodes instead APP695-specific amino acids. Full-length APP695 (flAPP695) in the SK plasmid is assembled by joining C6-695 and C5-1 as described for the construction of APP751 (Figure 3).

APP770 was constructed by joining fragments from C8-1.1 and C6-1 (Figure 4C) . The 3962 bp Hindlll/Scal fragment from C6-1 was obtained by first isolating the linear fragment from a partial Seal digest of C6-1 and then treating with Hindlll. This fragment was then ligated with the 537 bp Scal/Hindlll fragment from C8-1.1 to form C6-770 which is analogous to C6-1 except it encodes APP770-specific sequences. Full-length APP770

(flAPP770) was assembled in the SK plasmid by joining C6- 770 with C5-1 as described for APP751 (Figure 3) . Example 2. Insertion of APP Into A Constitutive Mammalian Expression Vector.

To verify that the full-length APP clones encode the intact protein, the cDNA fragments were inserted in a eukaryotic expression vector that directs and promotes efficient RNA synthesis when introduced into mammalian cells. A number of vectors suitable for stable transfection of mammalian cells are available to the public, e.g., see Pouwels et al. Cloning Vectors: A Laboratory Manual , 1985; methods for constructing such cell lines are also publicly available, e.g., in Ausubel et al., (supra) . In the present example, the expression vector, termed pcDNAI/neo (Figure 5) , is utilized (Invitrogen, San Diego, CA) . It contains several features required for growth and selection in bacteria and the selection of stable and transient transfectants in mammalian cells. The APP cDNA sequences are inserted in the multiple cloning site, downstream of the CMV promoter and upstream of SV40 RNA processing signals that include an RNA intron and polyadenylation sequences (pA) . When introduced into mammalian cells, the CMV promoter directs efficient transcription of the APP cDNA sequences and the RNA processing signals stabilize the RNA transcripts. The recombinant plasmid containing full-length APP751 in the pcDNAI/neo expression vector (Figure 5) is assembled using three fragments; a BamHl/BamHl fragment from flAPP751 containing residues -7 to +1497 of APP751, a BamHl/XhoI fragment from flAPP751 containing residues +1497 to +2928, and the expression plasmid fragment linearized by digestion with Xhol and BamHl. Insertion of the full-length cDNA fragments for APP695 and APP770 are done similarly by substituting the BamHl/BamHl fragments containing N-terminal APP sequences from flAPP695 and flAPP770, respectively. Example 3. Expression of APP Clones in Mammalian Cell Culture.

To detect protein expression, the plasmid DNAs from CMV/APP695, CMV/APP751, and CMV/APP770 were introduced into human 293 cells by calcium phosphate co- precipitation (see, e.g., Chen et al., Mol . Cell . Biol . 7:2745-2752, 1987). Typically, lxlO⁶ human 293 cells are transfected with 10 ug plasmid DNA for 12 hours, washed to remove excess DNA, and cultured for an additional 36 hours to allow for protein synthesis. At the conclusion of the incubation period the transfected cells are harvested and analyzed for APP production by immunoblot analysis of cell lysates (see, e.g., Towbin et al., Proc. Natl . Acad. Sci . USA 76:4350-4354, 1979; Burnette et al., Anal. Biochem. 112:195-203, 1981). Also at this time, conditioned medium samples in which the cells have been cultured for one hour prior to harvest are collected for analysis on immunoblots to detect APP species secreted from the transfected cells. Crude protein preparations from cell lysates (L) prepared from lxlO⁶ cells and conditioned medium (M) samples from 2.5xl0⁶ cells are resolved according to size on 6% polyacrylamide gels, transferred to nitrocellulose support membranes, and incubated with a commercially available monoclonal antibody (22C11, Boehringer Mannheim, Indianapolis, IN) that specifically recognizes an epitope within amino acid residues 60 to 100 of APP. In duplicate experiments shown in Figure 6, all three APP isoforms are efficiently expressed in the transfected 293 cells. Several APP species are detected in the cell lysate with the 22C11 antibody and most likely represent full-length proteins that differ in the extent of glycosylation. This conclusion is supported by the apparent molecular weight of the proteins, their sensitivity to pharmacological agents that disrupt glycosylation, and the finding that all species detected in the cell lysate in Figure 6 are recognized by N- and C-terminal-specific APP antibodies. All three isoforms are also normally processed as indicated by the accumulation of a truncated APP species in the conditioned medium samples that is recognized by the 22C11 antibody (Figure 6) but not by a C-terminal- specific antibody for APP.

Example 4. Construction of APP Clones Containing Site- Specific Mutations Linked to Human Diseases That Exhibit Beta-Amyloid Pathology.

The FAD and HCHWA-DT genetic mutations illustrated in Figure 1 were introduced into APP751 by PCR site- directed mutagenesis using an overlap extension method (see, e.g., Ho et al., Gene 77:51-59, 1989). This method uses the original plasmid DNA as a template and mutations are introduced into a pair of overlapping amplified sequences that are subsequently fused by recombining them in a second PCR reaction (Figure 5A) . Construction of the FAD mutation is initiated in separate amplification reactions where two fragments of the target APP751 sequence are amplified. Reaction 1 uses flanking primer a (5'dCGTAGCCGTTCTGCTGCATC3'; SEQ ID NO: 2) and internal primer b (5'dCGACAGTGATCATCATCACC 3'; SEQ ID NO: 3) that hybridizes at the site of the mutation and contains the mismatched base (underlined) A for G. Reaction 2 uses flanking primer d (5'dTAATACGACTCACTATAGGGAGA 3'; SEQ ID NO: 4) and internal primer c (5'dGGTGATGATGATCACTGTCG3' ; SEQ ID NO: 5) that contains the complementary base mismatch (underlined) . The mismatched bases are introduced into the newly synthesized fragments since the primers are incorporated into the amplification products. Because the internal primers are complementary, the two fragments ab and cd generated in the first amplification reactions are fused by combining and annealing them in the second amplification reaction. The flanking primers were designed so that the resultant amplified fragment ad contains restriction recognition sites that flank the mutation site. To introduce the mutation into full- length APP751, the amplified fragment ad is cleaved with EcoRI and Haell and assembled into the APP subclone,amy- 7, by ligation with the 420 bp Haell/Pstl fragment of amy-7 and the EcoRl/Pstl Gem3 plasmid fragment to form amy-7/FAD. The 563 bp EcoRI/Spel fragment from amy-7/FAD is then inserted into a larger APP subclone that serves as a shuttle vector for assembly of the mutation into full-length APP751 in the CMV expression vector. The shuttle vector APP751sh/SP72 contains APP sequences from the Xho I restriction site at +1078 to the Clal restriction site at +2641 cloned in the SP72 bacterial vector (Promega, Madison, WI) . The FAD mutation is introduced into the shuttle vector by replacing the 563 bp EcoRl/Spel fragment from APP751sh/SP72 with the analogous fragment from amy-7/FAD creating APP751sh/FAD. The Xhol/Xbal APP fragment from APP751sh/FAD is joined to the Xhol/Xbal APP/plasmid fragment from APP751/CMV to generate a full-length APP751 cDNA clone that carries the FAD mutation (CMVAPP751/FAD) .

The construction of a full-length APP751 cDNA clone containing the HCHWA-DT mutation was done in an identical manner as that described for the FAD mutation except internal primer b in reaction 1 has the sequence 5'dGTGTTCTTTGCACAAGATGTGGG3' (SEQ ID NO: 6) containing the mismatched base C for G (underlined) , and primer c has the sequence 5'dCCCACATCTTGTGCAAAGAACAC3' (SEQ ID NO: 7) containing the complementary base mismatch (underlined) . The introduction of both mutations in amy- 7 and subsequently in CMVAPP751/FAD was confirmed by nucleotide sequence analysis. In addition, the entire sequence of the amplified portion of the mutant amy-7 clones was verified by nucleotide sequence analysis to ensure that additional mutations through PCR-based errors were not introduced into the APP751 clones.

Example 5. Preparation of Transgene Vector Backbones and Insertion of Human APP751.

The multiple cloning region of the eukaryotic expression vector plasmid pcDNAI (Invitrogen, San Diego, CA) was modified by the addition of a Sail site (Figure 8A) . Plasmid pcDNAI was cut with Hind III and blunt ends were created by filling in with Klenow polymerase in the presence of 0.5 mM deoxyribonucleotide triphosphates (dNTPs) . The plasmid was then cut with Bam HI. A synthetic double stranded oligonucleotide (DH3/4) , containing internal Sail and Hindlll restriction site recognition sequences and flanking 5' blunt and 3' Bam HI sequences, was ligated into the blunt and BamHI sites of the pcDNAI vector to create plasmid pcDNAS.

DH3/4: 5' CGTCGACGCAAGCTTG 3' (SEQ ID NO: 8)

3' GCAGCTGCGTTCGAACCTAG 5' (SEQ ID NO: 9) Plasmid pcDNAS was then further modified to delete the SV40 small t intron located 3' to the multiple cloning region (Figure 8B) . Briefly, plasmid pcDNAS was cut with Xbal followed by treatment with Klenow polymerase in the presence of 0.5 mM dNTPs to create a blunt end. The plasmid was then cut with Hpal (blunt) and was recircularized by ligation. The resulting plasmid (pcDNAS(-ti) lacks 754 bp of SV40 sequence encoding the small t intron.

Plasmid flAPP751/pcDNAIneo was cut with Hindlll and Nsi I to liberate a 2.7 kb fragment encompassing the human APP751 cDNA (-7 to +2716) . Plasmids pcDNAS and pcDNAS(-ti) were cut with Hindlll and Nsil and were ligated to the 2.7 kb Hindlll/Nsil APP751 cDNA. These ligations resulted in the generation of plasmids pcDNAS- 751 and pcDNAS(-ti)-751.

Example 6. Cloning of Transgene Plasmids Ceph l and Ceph 2. Plasmids pSyn4400 and pSyn349 (Howland, D. et al., supra) provide a source of rat synapsin I promoter sequences. Plasmid pSyn4400 (Figure 9) contains a 4.5 kb fragment encompassing -4.4 kb to +110 bp of the 5' flanking region of the rat synapsin I gene cloned into a pUCCAT expression vector (Kislauskis et al. Neuron 4:783- 795, 1990) Plasmid pSyn349 contains a 460 bp fragment encompassing -349 bp to +110 bp of rat synapsin I 5' flanking DNA. Each promoter fragment can be excised with Sail and Ncol restriction enzymes. Plasmid pcDNAS-751 was cut with Hindlll and blunt ended with Klenow polymerase in the presence of 0.5 mM dNTPs. The plasmid was then cut with Sail. For preparation of the rat synapsin I inserts, plasmids pSyn4400 and pSyn349 were cut with Sail and Smal to liberate a 4.5 kb and a 450 bp fragment, respectively. These fragments were each ligated to the Sail/ blunt ended pcDNAS-751 vector resulting in the generation of the Ceph 1 and Ceph 2 transgene plasmids. Hence, the Ceph 1 and Ceph 2 plasmids^* contain the large (4.4 kb) and small (349 bp) synapsin I promoter fragments, respectively, immediately 5' to the native human APP751 cDNA. Each plasmid contains SV40 processing signals including the small t intron and polyadenylation signals located 3' to the APP751 cDNA.

Example 7. Cloning of Transgene Plasmids Ceph 3 and Ceph 4.

To enable cloning of Ceph plasmids 3 and 4, a PCR strategy was employed to generate the SV40 small t intron sequences with appropriate 5' and 3' end restriction sites for insertion into transgene vectors at a 5' loci.

Oligonucleotides DH5 (5' TATCCCGGGCTGTGGTGTGACATAATTGG 3'; SEQ ID NO: 10) and DH6 (5' GCCAAGCTTAGGTTGGAATCTAAAATACAC 3';. SEQ ID NO: 11) were synthesized and used as primers to generate a 127 bp fragment encoding the SV40 small t intron. Briefly, 1 ng of p776RVA template DNA, containing the entire SV40 sequence cloned into pGEM3, was mixed with 1 μM of DH5 and DH6, 200 μV of each dATP, dCTP, dTTP, and dGTP, 2.5 units of Vent DNA polymerase (New England Biolabs, Beverly, MA) in a 100 μl reaction containing 10 mM Tris, pH 8.3, 50 mM KCL, 1.5 mM MgCl2, and 0.001% gelatin. The reaction was placed in a thermocycler at 94°C, for 5 min. for l cycle followed by 94°C, l min.; 48°, 2 min.; 72°C, 3 min. for 30 cycles and 72°C for 7 minutes for 1 cycle. The product of the PCR was electrophoresed on 2% NuSieve/1% agarose gel, (FMC, Rockland, ME) excised and eluted with MerMaid reagent (Bio-101 Inc., La Jolla, CA) . The DNA was then cut with Sma I and Hind III, electrophoresed on 2% NuSieve/1% agarose gel, and eluted with MerMaid reagent.

Plasmids pSyn4400 and pSyn349 (Figure 10) were cut with Sal I and Sma I to liberate the large (4.4 kb) and small (-349 bp) synapsin I promoter fragments. Plasmid pcDNAS (ti-)751 was cut with Sail and Hindlll. Three-way ligations were set up between the Sall/S a I (4.4 kb) synapsin I fragment or the Sall/Smal (349 bp) synapsin I fragment and, the Smal/Hindlll 127 bp small t intron DNA, and the Sall/Hindlll cut pcDNAS (ti)-751 vector. The resulting Ceph 3 and Ceph 4 transgene plasmids contain the large (4.4 kb) or small (349 bp) synapsin I promoter fragments, respectively, cloned immediately upstream of the SV40 small t intron. The human APP751 cDNA and SV40 polyadenylation signals are located immediately downstream of the synapsin I/t intron fusion.

Example 8. Cloning of Intermediate Vectors Containing FAD and HCHWA-DT APP Seguences. To enable construction of Ceph 5, 6, 7, and 8 transgene plasmids, intermediate vectors pmt693-pre and pmt717-pre were cloned (Figure 11) . The pmt693-pre contains APP751 harboring the HCHWA-DT mutation at amino acid 693 cloned into pCDNAS. Plasmid pmt7l7-pre contains APP751 harboring the FAD mutation at amino acid 717 cloned into pCDNAS.

Briefly, plasmid flAPP75l(HCHWA-DT)/pcDNAlneo was cut with Hindlll and Hgal to liberate a 2.2 kb fragment (-7 to +2158) harboring the HCHWA-DT mutation at amino acid 693. Plasmid f1APP751(FAD)/pcDNAlneo was cut with Hindlll and Hgal to liberate a 2.2 kb fragment (-7 to +2158) harboring the FAD mutation at amino acid 717. Plasmid flAPP751/pcDNAIneo was cut with Hgal and Nsil to liberate a 558 bp fragment containing the 3' end of the APP751 cDNA. Two-way ligations between the 558 bp Hgal/Nsil APP fragment and the 2.2 kb Hindlll/Hgal fragments were done. The resulting full length APP751 cDNA fragments (-7 to +2158) containing either the FAD or the HCHWA-DT mutations were ligated into pcDNAS after cutting with Hindlll and Nsil resulting in pmt693-pre and pmt717-pre plasmids, respectively.

Example 9. Cloning of Transgene Plasmids Ceph 5 and Ceph 6.

Plasmids pmt693-pre and pmt717-pre were cut with Hindlll followed by treatment with Klenow polymerase in the presence of 0.5 mM dNTPs to create blunt ends (Figure 12). The plasmids were then cut with Sal I. The large (4.4 kb) synapsin I promoter fragment was excised from plasmid pSyn4400 using Sail and Smal and was ligated into the blunt/Sal I sites of pmt693-pre and pmt7l7-pre to create the Ceph 5 and Ceph 6 transgene plasmids. Ceph 5 and Ceph 6 plasmids contain the large (4.4 kb) synapsin I promoter immediately upstream of the APP751 cDNA harboring the HCHWA-DT or FAD mutations, respectively, followed by SV40 processing sites (t intron and poly adenylation signal) .

Example 10. Cloning of Transgene Plasmids Ceph 7 and Ceph 8.

To enable construction of transgene plasmids Ceph 7 and Ceph 8, it was necessary to delete SV40 small t intron sequences from the 3' end of the plasmids pmt693- pre and pmt7l7-pre vectors (Figure 13) . The pmt693-pre and pmt717-pre DNA were cut with Xbal followed by treatment with Klenow polymerase in the presence of 0.5 mM dNTPs to create blunt ends. The plasmids were then cut with Hpal (blunt end) and recircularized by ligation. The resulting plasmids pmt693pre(-ti) and pmt7l7pre(-ti) were cut with Hindlll and Sal I to enable insertion of the large (4.4 kb) Sall/Sma I synapsin I promoter fragment and the 127 bp Smal/Hindlll SV40 small t intron, generated by PCR as previously described. The resulting 3-way ligation yielded plasmids Ceph 7 and Ceph 8. The Ceph 7 and Ceph 8 transgene plasmids contain the large (4.4 kb) synapsin I promoter immediately upstream of the SV40 small t intron splice. The APP751 cDNA harboring either the HCHWA-DT or FAD mutations and SV40 polyadenylation signal are located immediately downstream of the synapsin I/SV40 t intron fusion.

The general structural features of transgenes Ceph 1 through Ceph 8 are summarized in Figure 14. The transgene DNA in each construct can be isolated from plasmid sequences by cleavage with Sail and Ncol. Example 11. Construction of Synapsin Transgenes Encoding the β/A4 and CIOO Domains of APP.

Recombinant transgene templates have also been constructed with the synapsin promoter and DNA fragments encoding 3/A4 and CIOO. Both APP coding segments have been linked to 350 bp or 4.4 kb synapsin promoter fragments. To construct the 3/A4 recombinants (Figure 15A and B) , two PCR primers were prepared to amplify the DNA domain encoding 40 amino acids of ,9/A4, supply protein initiation and termination signals, and provide convenient restriction enzyme recognition sites for subcloning. The forward a primer,

5'dAATTGGTACCGTGAAGATGGATGCAGAATTCCGAC3' (SEQ ID NO: 12) , includes a Kpnl restriction site at the 5' terminus and overlaps nucleotides +1954 to +1972 of APP751 that begins with the methionine residue at amino acid position 652 immediately upstream to the ?/A4 domain. The reverse b primer, 5'dCATTATGCATCTAGACAACACCGCCCACCATGA3' (SEQ ID NO: 13) , includes an Nsil restriction site at the 5' terminus and a CTA complementary sequence for the TAG termination codon followed by the last 19 nucleotides of the 40 amino acid form of 0/A4. The PCR amplification (see, e.g., Saiki et al.. Science 230:1350-1354, 1985) is done using 10 ng of flAPP751 template, 1 mM oligonucleotide primers, 200 mM nucleotide triphosphates, and 2 units of Vent DNA polymerase (New England Biolabs, Beverly, MA) in a 100 μl reaction volume. The PCR products are isolated after agarose gel electrophoresis (described above) and treated with Kpnl and Nsil to create compatible ends for subcloning. To link 3/A4 with the short (Figure 15A) and long (Figure 15B) synapsin promoters the 3/A4 PCR fragment is ligated with the Sall/Nsil pcDNAI plasmid fragment and either the 350 bp Sall/Kpnl or 4.4 kb Sall/Kpnl synapsin promoter fragments. The resultant plasmids are designated Syn350/?A4 and Syn4.4/jøA4, respectively, and have the SV40 RNA splice signals in the downstream configuration. The CIOO recombinants (Figures 15C and 15D) were constructed in a similar fashion except that the reverse primer for the PCR reaction to amplify the CIOO fragment has the sequence 5'dATTATGCATCTAGTTCTGCATCTGCTCAAAG3' (SEQ ID NO: 14) and includes an Nsil restriction site at the 5' terminus and a CTA complementary sequence for the TAG termination codon followed by the last 19 nucleotides of APP.

Example 12. Preparation of Linearized Ceph 1 Through 8 Transgene Inserts for Microiniection.

Each transgene insert, containing rat synapsin I promoter, APP751 cDNA, and SV40 processing signals, was excised as full length from plasmids Ceph 1 through Ceph 8 by a Sall/Ncol digest. The resulting DNA was electrophoresed on a 1% agarose gel. The transgene inserts were excised and electroeluted in O.lx TBE, at 200 volts for 2 hours. The eluates were passed over a Qiagen tip-100 column (QIAGEN, Chatsworth, California) , followed by isopropanol precipitation. The pellets were washed in 80% ethanol and were resuspended to 20 ng/μl in 10 mM Tris, pH 8, 0.25 mM Na₂EDTA.

Example 13. Establishment of Transgenic Mice with the APP transgenes.

Introduction of the transgenes into the mouse genome was accomplished by injection of approximately 10 picoliters (pi) of the DNA solution into the male pronucleus of a C57B16/SJL zygote. Methods for collection of the fertilized zygotes, pronuclear injection, and implantation into pseudopregnant females are described in U.S. Patent 4,873,191 which is incorporated herein by reference. Identification of DNA-positive animals (Fo or founder animals) in the litters that developed from the injected zygotes was determined by Southern blot analysis of genomic DNA samples prepared from tail sections. At the time of weaning (4 to 6 weeks of age) , a 1 cm portion of the tail was taken from each mouse and processed for isolation of high molecular weight genomic DNA. The DNA samples (10 ug) were resolved on 1% agarose gels, transferred to a nylon support membrane and hybridized with a radiolabeled Bglll - Spel DNA fragment from Cephl that is present, at least in part, in all synapsin/APP transgenes used in this invention (Figure 16) . Hybridization was at 42°C for 15 to 18 hours in 50% formamide, 20 mM sodium phosphate, pH 6.8, 6X SSC (IX SSC is 150 mM NaCl, 15 mM sodium citrate), 0.1% SDS, IX Denhardt's solution (0.02% Ficoll, 0.02% polyvinylpyrrolidone, and 0.02% bovine serum albumin), and 100 ug/ml heat denatured salmon sperm DNA. The filters were then washed in a O.lx SSC, 0.1% SDS solution at 65°C and exposed to autoradiographic film to identify DNA samples that contain DNA sequences complementary to the transgene probe. Under these hybridization and wash conditions, the human APP probe anneals preferentially to human versus mouse APP sequences. An example of the Southern blot results appears in Figure 16 where genomic samples prepared from mice injected with Ceph 1 transgene DNA are presented. Integration of transgene DNA after microinjection commonly occurs as an array of multiple copies of the DNA template inserted as "head to tail" concatemers at a single site in the mouse genome. The samples in Figure 16 were digested with BamHl which cleaves the transgene DNA at two sites at the 5' terminus of the synapsin promoter and within the APP coding sequences at +1497. Multiple copies of transgene DNA commonly integrate into mouse genomic DNA in a head to tail arrangement and under these conditions, detection of a 2 kb fragment is expected. Integration of a single transgene copy would yield a fragment of unpredictable size following BamHl digestion since the location of the second BamHl recognition sequence would be in the mouse genomic DNA and the site of transgene integration is largely random. The Southern blot results show the presence of transgene DNA in Fo mice designated 6161, 6164, 6166, 6173, 6175, 6181, 6182, 6192, and 6193. The expected 2 kb transgene fragment (arrow) , diagnostic of multiple copy integration, is detected in all samples except 6166 which contains a higher molecular weight fragment that presumably represents integration of a single transgene copy. The additional fragments detected by the APP-specific probe in the multiple copy integrants most likely correspond to junction fragments at either ends of the transgene arrays that border the mouse genomic DNA. Results on the identification of founder animals for transgenes Ceph 1 through Ceph 8 are summarized in Table 1. Copy number approximations were made by comparison of the hybridization signals to plasmid DNA standards in which known amounts of APP transgene DNA were added in parallel on each Southern blot. In certain samples, there were difficulties in determining the amounts of genomic DNA resolved on the agarose gels so copy number values were not calculated (NC) .

Table 1: Founder Animals

Transgene Founder DOB Copy# Ceph 1

NC 5 1

20 3

NC 3

10

NC

Ceph 2

3-5 3-5 3-5

1 1

Ceph 3

3 1 5

Ceph 4

5-10 5-10

1 5-10

5

Ceph 5

NC

1 1 5 5 1 1 1 1 2-3 10+ 5

Ceph 6

2

5 3 3 2 2 2

Ceph 7

3 2 5 5 5 2

Ceph 8

3

1 5 2 1

3

NC - not calculated Transgenic lines were established by breeding the founder mice with non-transgenic C57B16/SJL mice and analyzing the Fl litters for transmission of the transgene. This was done using genomic DNA samples prepared from tail sections of the Fl mice taken at the time of weaning. Presence of the transgene was determined either by a PCR strategy that specifically amplifies transgene DNA which is a modification of the PCR method presented in Figure 17 or a slot blot method in which enhancement of the hybridization signal over the endogenous background was used to score positive animals. Positive signals for both methods of detection were confirmed by Southern blot analysis. Results on the transmission frequencies obtained for each founder animal appear in Table 2. The frequencies are presented as the number of DNA-positive Fl animals/total number of Fl animals analyzed. Copy number estimates for the Fl mice were made from slot blots. Many of the Fl estimates are different from the corresponding founders but this is not unusual since the founders are chimeric and the copy number present in their germline which is inherited by the Fl progeny may differ from the somatic tissue. For two of the founders (Ceph 1, #6181; Ceph 5, #6847) the Fl progeny inherited two different transgene copy numbers. This is due to integration of the transgene DNA at two different loci after microinjection in the mouse zygote. Subsequent breeding of these animals will result in segregation, or differential inheritance, of the two loci in the progeny. Table 2: Fl Transmission and RNA Expression

Transgene Founder Fl Ceph 1

6161

6164

6166

6173

6175

6181

6182 6192 6193

Ceph 2

6341 7340 7346 7347 7348

Ceph 3

6431 6433 6435

Ceph 4

6611 6621 6628 6637 6640

Ceph 5

6817 6822 6832 6833 6834 6837 6839 6842 6845 6847

6848 6855

Ceph 6

6891 6893

6902 6906 6911 6916 6917

Ceph 7

6931 6933 6934 6935 6936 6939

Ceph 8

6948 6951 6956 6966 6973 6980

NC - not calculated Example 14. Tissue-specific Expression of the APP Transgenes.

Once the transgenic lines are established, expression of the transgenes can be analyzed in some of the Fl animals for each line. Expression is characterized at the RNA and protein level prior to histological examination. To rapidly and reliably identify expression positive animals a RNA PCR strategy was developed and is outlined in Figure 17A. Forward and reverse PCR primers corresponding to synapsin 5' untranslated and APP coding sequences, respectively, were synthesized for the amplification reaction. The primer design exploits the unique synapsin/APP junction of the transgenes so that only transgene RNA, and not APP or synapsin RNA, can serve as the template for amplification. Synthesis of a 405 bp fragment is diagnostic for transgene RNA from templates containing a 3' intron (Ceph 1, 2, 5, and 6) and a 465 bp (spliced) or 532 bp (unspliced) fragment is diagnostic for transgene RNA from templates with a 5' intron (Ceph 3, 4, 7, and 8) . An additional component of the RNA amplification reaction is rTth thermostable polymerase (Promega, Madison, WI) that is capable of using RNA or DNA as a template for transcription depending on the Mn²⁺ and Mg²⁺ concentrations.

Total RNA was prepared from brain and heart tissue by the RNAzol B method using conditions suggested by the supplier (Cinna/Biotecx Laboratories, Houston, Texas) . Frozen tissue samples were resuspended in 4 ml of the RNAzol B solution and homogenized by mechanical disruption. The samples were extracted with an equal volume of chloroform and centrifuged at 12,000xg at 4°C for 15 minutes to separate the aqueous and organic phases. The aqueous phase was collected and treated with 2 ml of isopropanol to precipitate the RNA. The precipitate was collected by centrifugation at 12,000xg for 20 minutes at 4°C, air dried, and resuspended in 0.4 ml water. The samples were extracted with an equal mixture of phenol and chloroform and centrifuged to separate the phases. The aqueous phase was collected, brought to 0.3 M sodium acetate, and treated with 1 ml ethanol to precipitate the RNA. The RNA was recovered by centrifugation, rinsed two times in 80% ethanol, dried under vacuum, and resuspended in 75 μl water. The RNA was quantitated by UV absorbance and stored at -70°C.

Prior to PCR amplification, the RNA samples were treated with DNasel to remove any residual DNA. This was done by resuspending 10 ug of RNA in 100 mM sodium acetate, 5 mM MgCl₂, 100 units RNasin, and 1 unit RQ-1 DNasel (Promega, Madison, WI) and incubating at 37°C for 30 minutes. The DNasel was inactivated by heating to 65°C for 5 minutes and the samples were extracted with an equal mixture of phenol and chloroform. The RNA was collected from the aqueous phase by ethanol precipitation, washed in 80% ethanol, dried under vacuum, and resuspended in 80 μl water.

The PCR reactions with each RNA sample were done in duplicate using either primers specific for the transgene RNA or, as a control for RNA integrity, primers specific for endogenous mouse APP RNA (Figure 17A) . The forward transgene primer (5'dACCGACCCACTGCCCCTTGGATCC3'; SEQ ID NO: 15) extends from +87 to +110 of the synapsin I sequence and the reverse transgene primer (5'dGGATGGGTCTTGCACTGCTTGCGG3'; SEQ ID NO: 16) extends from +303 to +326 of the human APP sequence. The forward control primer (5'dACTCGCACACGGAGCACTCGGTGG3'; SEQ ID NO: 17) extends from -17 to -40 of the mouse APP sequence and the reverse control primer(5'dGCTTGCGGCCCCGCTTGCACCAGT3'; SEQ ID NO: 18) extends from +287 to +311 of the mouse APP sequence. To begin the amplification 250 ngs of RNA was incubated with 0.75 mM of the reverse primer, 200 mM dNTPs, 10 mM Tris.HCl, pH 8.3, 90 mM KC1, 1 mM MnCl₂, and 5 units rTth polymerase at 70°C for 15 minutes. Under these conditions, the reverse transcriptase activity of the polymerase is stimulated and an antisense DNA strand is synthesized from the RNA template. The reaction was then brought to 0.75 mM EGTA, 0.05% Tween 20, 2.5 mM MgCl₂, and 0.75 mM of the forward primer to stimulate the DNA polymerase activity. The amplification reaction was continued for 35 cycles of 95°C for 1 minute and 60°C for 1 minute incubations. The amplified products were resolved by agarose gel electrophoresis, transferred to a nylon support membrane, and hybridized with an oligonucleotide (5'dGGCAGCATCGCGACCTGGCGGGAATTCCTG3'; SEQ ID NO: 19) end-labeled with γ-³²P-ATP and polynucleotide kinase. This oligonucleotide spans the synapsin/APP junction and is specific for amplification products from the synapsin/APP transgenes. Hybridization was done at 65°C for 18 hours in 0.5x SSPE (lx SSPE is 180 mM NaCl, 10 mM sodium phosphate, pH 7.4, 1 mM EDTA), 2x Denhardts, 50 mg/ml denatured salmon sperm DNA, 50 mg/ml E. coli tRNA, and 0.1% SDS. The filters were washed in 2x SSPE, 0.5% SDS at room temperature for 1 hour and in 5x SSPE, 0.1% SDS at 65°C for 20 minutes. Results of the PCR screen on Fl animals derived from some Ceph 1 founders appear in Figure 17B. Synthesis of transgene RNA is apparent in brain samples from Fl animals 5-1-3, 5-1-15, 6-7-1, and 6-7-7 as indicated by the 405 bp hybridization signal. The 5-1-3 and 5-1-15 Fl mice are progeny from founder 6173 and the 6-7-1 and 6-7-7 Fl mice are progeny from founder 6181. Brain-specific synthesis of transgene RNA is suggested by the lack of amplification products in the heart RNA samples from these same animals (Figure 17B) and confirmed by the lack of amplification products in liver, kidney, and spleen RNA samples. As expected. transgene-specific amplification products are not detected in DNA-negative control mice (6-7-4, 6-7-6, and 5-1-16) . The integrity of the RNA samples was verified in parallel PCR reactions using the mouse APP primers and all samples supported synthesis of the diagnostic 351 bp fragment.

Example 15. Ouantitation of Relative Levels of Transgene RNA Expression.

The complete PCR data demonstrate that the transgenes are transcriptionally competent and both the 350 bp and 4.4 kb synapsin promoters are able to direct brain-specific expression of the linked APP sequences. However, the PCR assay is not quantitative and an accurate determination of the relative levels of transgene expression between the various transgenic lines can not be made. To establish a quantitative assay, an RNA:RNA solution hybridization strategy was designed (see, e.g., Zinn, K.D., et al., Cell 34:865-879, 1983). An antisense RNA strand uniformly labeled with α-³ P-UTP was synthesized from a subclone of the synapsin/APP transgenes (Figure 18) . The R4T72 subclone contains the 5' terminus of human APP extending from -7 to +226. Antisense RNA was synthesized in vitro by using the R4T72 DNA, linearized by Bglll digestion, as a template, α-³²P- UTP, and SP6 RNA polymerase according to the procedures of the vendor (Promega, Madison, WI) . This results in the synthesis of a 344 nucleotide RNA transcript (probe) containing a 275 nucleotide segment that is complementary to transgene mRNA. The additional sequences in the full- length probe are from flanking plasmid sequences.

Hybridization reactions were done under conditions of probe excess with 30 ug of tissue RNA and 1.5 x 10⁶ cpm of the radiolabeled RNA. The tissue and probe RNA samples were mixed, dried under vacuum, resuspended in 80% formamide, 1OO mM sodium citrate, pH 6.4, 300 mM sodium acetate, pH 6.4, and 1 mM EDTA, heat-shocked at 85°C for 4 minutes to denature intramolecular hybrids, and incubated overnight at 45°C. The samples were treated with 20 mg/ml RNase A and 400 units/ml RNase TI at 37°C for 30 minutes to remove single-stranded RNA species. The RNase-resistant species were collected by ethanol precipitation following inactivation of the RNases with guanidinium (Ambion, Austin, Texas) . The RNA pellets were resuspended in an 80% formamide sample buffer and resolved on a 5% polyacrylamide, 8 M urea gel. Hybridization of the probe to transgene RNA will result in the protection of a 275 base radio-labeled fragment from RNase digestion. Results of this RNA protection assay on brain samples from Fl animals identified as RNA- positive and RNA-negative by PCR analysis appear in Figure 19A. The status of each sample with regard to transgene DNA (+ or -) or RNA expression (+ or -) determined by RNA PCR analysis is indicated above each sample number. As controls, samples from DNA-negative (7-2-6, 6-1-1, and 4-3-14) and DNA-positive but RNA- negative (4-1-6) are also included. The closed arrow indicates the position of the 275 base RNA fragment protected from RNase digestion by transgene mRNA. This fragment is detected in all samples identified as RNA positive by PCR analysis but is not found in the DNA- or RNA- minus control samples. The open arrow marks the position of undigested, full-length probe RNA. A measurement of the relative levels of transgene expression between the various transgenic lines can also be made. RNA samples from Fl animals 7-2-8, 7-2-7, and 7-2-1 show the highest intensity protected fragment indicating that these animals are relatively high expressors of transgene mRNA. All three of these Fl animals are progeny of the same founder mouse, 6935. Densitometric scans of the RNA protection results (Figure 19B) show that expression levels in the various lines assayed in this experiment cover a 200-fold range. The results of the RNA protection assay on the relative levels of transgene RNA expression for many of the transgenic lines are summarized on Table 2. The lines have been categorized as non-expressors (-) or expressors ranging from low (+) to high (+++++) and expression levels between all lines tested cover a 500-fold range. The RNA expression data confirms the utility of the synapsin/APP transgenes in directing efficient brain- specific expression of APP coding sequences. The ability of the synapsin promoter to direct brain-specific expression is not limited to APP coding sequences and would be applicable to any linked gene sequence containing the proper RNA processing signals.

Example 16. Ouantitation of Transgene Protein Expression.

Expression of human APP751 protein was analyzed on immunoblots using an antiserum prepared against the

Kunitz protease inhibitor (KPI) domain found in APP751 and APP770, but absent in APP695. This was expected to reduce the signal from endogenous APP695 and, as a result, improve detection and quantitation of the human APP751 transgene protein. The R7 antiserum (Anderson et al., EMBO J. 8:3627-3632, 1989) was prepared against a portion of the KPI-domain in App751 (amino acids 296 to 315) and recognizes both human and mouse APP isoforms. Crude protein samples of tissue lysates were prepared from individual brain regions, including temporal cortex, frontal cortex, hippocampus, striatum, and cerebellum, from a Ceph 8; 4-9 transgenic mouse and a littermate non- transgenic control mouse. The samples were resolved according to size on a 6% polyacrylamide gel, transferred to nitrocellulose, and incubated with the R7 antibody. Specific antibody binding was visualized by enhanced chemiluminescence (ECL, Amersham Corporation, Arlington Heights, Illinois) using an HRP (horse radish peroxidase)-conjugated goat anti-rabbit IgG secondary antibody (BioRad Laboratories, Richmond, California) . The expression results presented in Figure 20 indicate that APP751 is overexpressed, relative to non-transgenic levels, in all brain regions of the Ceph 8 transgenic mouse. Overexpression varied between regions, being highest (10-fold) in the cerebellum and cortices, intermediate in the hippocampus, and lowest in striatum. The regional variation in APP751 overexpression correlates well with the neuronal densities of each region and supports a neuron-specific expression pattern of the transgene. This establishes the human APP751 transgene protein is being expressed in brain regions that are highly susceptible to _/9-amyloid pathology in the human disease.

Example 17. Establishment and Analysis of Transgenic APP Rats.

A transgenic rat line has also been established with the synapsin/APP751 transgene carrying the FAD V to I mutation (Ceph 8 transgene) . Introduction of the transgene DNA into the rat genome was accomplished by the injection of approximately 10 pi of a 20 ng/ul DNA solution into the male pronucleus of a Sprague-Dawley zygote. Methods for collection of the fertilized zygotes, pronuclear injection, and implantation into pseudopregnant females are similar to the procedures described for the establishment of transgenic mice (US patent 4,873,191). DNA-positive founder rats are identified in the litters that developed from the injected zygotes by a PCR screen and slot blot hybridization analysis of genomic DNA samples prepared from tail portions taken at the time of weaning.

The PCR analysis is done using a modification of the procedure described for the detection of transgene RNA in mice (Example 14) . The forward (SEQ ID NO: 15) and reverse (SEQ ID NO: 16) transgene-specific primers are incubated at 0.4 uM concentrations with 500 ng genomic DNA, 5 units Taq DNA polymerase (Promega Biotec, Madison, WI) , 150 uM dNTPs, lOmM Tris.HCL, pH 8.3, 1.5 mM MgCL2₁90 mM KCI, in a 50 ul reaction volume. The amplification reactions continued for 35 cycles of 94°C for 20 seconds, 60°C for 30 seconds, and 72°C for 30 seconds. The slot blot hybridization assays were done by immobilizing the genomic DNA samples on nylon membranes and incubating overnight at 65°C with a radio-labeled 589 bp Bgl II (nucleotide 1937) - Spe I (nucleotide 2526, App751 coordinates) APP-specific fragment that is present in all APP transgenes and was isolated from flAPP751. The hybridization buffer is 0.25 M NaHP04, pH 7.0, 7% SDS, 1% BSA, and 1 mM EDTA.

Results of the slot blot analysis for identification of the DNA-positive Ceph 8 founder rat and several of the Fl progeny are shown in Figures 21A and B. DNA-positive animals are indicated by an enhanced hybridization signal over the endogenous background. Figure 21A contains DNA samples from litters that developed from injected zygotes and only animal 303 was DNA-positive. Based on comparisons to DNA standards containing known amounts of DNA, founder (Fo) animal 303 was estimated to contain 30 copies of the transgene.

Transmission of the transgene DNA to Fl progeny of Fo 303 is shown in Figure 21B. Fl progeny 419, 556, 557, and 558 are DNA-positive and animal 559 is a non-transgenic littermate. The DNA results were confirmed by PCR analysis of the same and additional DNA samples (Figure 22) . Presence of the transgene DNA in Fo 303 and Fl progeny 410, 413, 417, 419, 556, 557, and 558 is indicated by the appearance of the 532 bp amplification product (arrow) . As expected the 532 bp fragment is absent in the non-transgenic littermate (animal 559) . These data demonstrate the establishment of a transgenic rat line carrying the Ceph 8 transgene.

Expression of the transgene at the RNA level in the Ceph 8 rat line was quantitated by the RNase protection assay described in Example 15. RNA expression results with samples prepared from brain, heart, liver, kidney, and spleen from a line 303 Fl animal (#413) appear in Figure 23. The closed arrow marks the position of the 275 base RNA fragment diagnostic for transgene mRNA. These fragment is detected only in brain RNA from animal #413 and not in other peripheral tissue RNA samples, indicating brain-specific expression. As expected, there is no evidence of transgene mRNA in brain RNA prepared from a non-transgenic littermate (animal #414) . Comparison of the hybridization signal with the internal positive-control signal indicates that expression levels are moderate (+++) in the Ceph 8 transgenic rat line.

SEQUENCE LISTING

(1) GENERAL INFORMATION: (i) APPLICANT: Scott, Richard W. Howland, David S.

(ii) TITLE OF INVENTION: TRANSGENIC ANIMAL MODEL FOR ALZHEIMER'S DISEASE

(iii) NUMBER OF SEQUENCES: 22 (iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: Fish & Richardson

(B) STREET: 225 Franklin Street

(C) CITY: Boston

(D) STATE: Massachusetts

(E) COUNTRY: U.S.A.

(F) ZIP: 02110-2804

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: 3.5" Diskette, 1.44 Mb

(B) COMPUTER: IBM PS/2 Model 50Z or 55SX

(C) OPERATING SYSTEM: MS-DOS (Version 5.0)

(D) SOFTWARE: WordPerfect (Version 5.1)

(vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER:

(B) FILING DATE:

(C) CLASSIFICATION:

(▼ii) PRIOR APPLICATION DATA:

(A) APPLICATION NUMBER: 07/989,850

(B) FILING DATE: 25 November 1992

(viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: Clark, Paul T.

(B) REGISTRATION NUMBER: 30,162

(C) REFERENCE/DOCKET NUMBER: 02655/035001

(ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: (617) 542-5070

(B) TELEFAX: (617) 542-8906

(C) TELEX: 200154

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1:

CTG CCC GGT TTG GCA CTG CT 20

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 2: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2:

CGT AGC CGT TCT GCT GCA TC 20

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 3: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 3:

CGA CAG TGA TCA TCA TCA CC 20

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 4: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 23

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 4:

TAA TAC GAC TCA CTA TAG GGA GA 23

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 5: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 5:

GGT GAT GAT GAT CAC TGT CG 20

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 6: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 23

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 6:

GTG TTC TTT GCA CAA GAT GTG GG 23

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 7: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 23

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 7:

CCC ACA TCT TGT GCA AAG AAC AC 23

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 8: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 16

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 8:

CGT CGA CGC AAG CTT G 16

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 9: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 9:

GAT CCA AGC TTG CGT CGA CG 20 (2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 10: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 29

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 10:

TAT CCC GGG CTG TGG TGT GAC ATA ATT GG 29

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 11: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 30

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 11:

GCC AAG CTT AGG TTG GAA TCT AAA ATA CAC 30

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 12: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 35

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 12:

AAT TGG TAC CGT GAA GAT GGA TGC AGA ATT CCG AC 35

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 13: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 33

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 13:

CAT TAT GCA TCT AGA CAA CAC CGC CCA CCA TGA 33

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 14: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: ' 31

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 14:

ATT ATG CAT CTA GTT CTG CAT CTG CTC AAA G 31

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 15: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 24

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 15:

ACC GAC CCA CTG CCC CTT GGA TCC 24

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 16: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 24

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 16:

GGA TGG GTC TTG CAC TGC TTG CGG 24

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 17: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 24

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 17:

ACT CGC ACA CGG AGC ACT CGG TGG 24

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 18: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 24 (B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 18:

GCT TGC GGC CCC GCT TGC ACC AGT 24

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 19: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 30

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 19:

GGC AGC ATC GCG ACC TGG CGG GAA TTC CTG 30

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 20: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 53

(B) TYPE: amino acid

(C) STRANDEDNESS:

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 20:

Glu Val Lys Met Asp Ala Glu Phe Arg His Asp Ser Gly Tyr Glu Val 1 5 10 15

His His Gin Lys Leu Val Phe Phe Ala Glu Asp Val Gly Ser Asn Lys 20 25 30

Gly Ala He He Gly Leu Met Val Gly Gly Val Val He Ala Thr Val 35 40 45

He Val He Thr Leu 50

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 21:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 53

(B) TYPE: amino acid

(C) STRANDEDNESS:

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 21:

Glu Val Lys Met Asp Ala Glu Phe Arg His Asp Ser Gly Tyr Glu Val 1 5 10 15 His His Gin Lys Leu Val Phe Phe Ala Glu Asp Val Glu Ser Asn Lys 20 25 30

Gly Ala He He Gly Leu Met Val Gly Gly Val Val He Ala Thr Val 35 40 45

He He He Thr Leu 50

(2) INFORMATION FOR SEQUENCE IDENTIFICATION NUMBER: 22: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 53

(B) TYPE: amino acid

(C) STRANDEDNESS:

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 22:

Glu Val Lys Met Asp Ala Glu Phe Arg His Asp Ser Gly Tyr Glu Val 1 5 10 15

His His Gin Lys Leu Val Phe Phe Ala Gin Asp Val Gly Ser Asn Lys 20 25 30

Gly Ala He He Gly Leu Met Val Gly Gly Val Val He Ala Thr Val 35 40 45

He Val He Thr Leu 50

Claims

We claim:

1. A transgenic non-human animal harboring a transgene coding for an amyloid protein, wherein said transgene is under the transcriptional control of a synapsin gene promoter.

2. A transgenic non-human animal harboring a transgene coding for an amyloid protein under the transcriptional control of a human cytomegalovirus gene promoter.

3. The animal of claim 1 or claim 2, wherein said transgene contains DNA coding for a RNA processing signal.

4. The animal of claim 3, wherein said DNA comprises the SV40 small t intron sequences.

5. The animal of claim 4, wherein said SV40 small t intron is proximal to DNA coding for an amyloid protein.

6. The animal of claim 4, wherein said SV40 small t intron is distal to DNA coding for an amyloid protein.

7. The animal of claim 3, wherein said DNA comprises a SV40 polyadenylation signal.

8. The animal of claim 7, wherein said SV40 polyadenylation signal is distal to DNA coding for an amyloid protein.

9. The animal of claim 1 or claim 2, wherein said amyloid protein DNA comprises sequences coding for a precursor protein.

10. The animal of claim 9, wherein said precursor protein DNA comprises sequences coding for a human APP695, APP751, or APP770.

11. The animal of claim 10, wherein said DNA comprises fragments of sequences encoding human APP695, APP751, or APP770.

12. The animal of claim 11, wherein said fragments are CIOO or β/K .

13. The animal of claim 10 or claim 11, wherein said amyloid protein or precursor protein DNA comprises a coding sequence containing a mutation.

14. The animal of claim 13, wherein said mutation is the hereditary cerebral hemorrhage with amyloidosis - Dutch Type (HCHWA-DT) .

15. The animal of claim 13, wherein said mutation is the familial Alzheimer's disease (FAD).

16. The animal of claim l, wherein the transgene is expressed preferentially in neuronal cells of brain tissue of said animal.

17. The animal of claim 2, wherein the transgene is expressed preferentially in all cells of brain tissue and peripheral tissue of said animal.

18. The animal of claim 17, wherein the transgene is expressed preferentially in non-neuronal cells of brain tissue and peripheral tissue of said animal.

19. The animal of claim 17, wherein the transgene is expressed preferentially in peripheral tissue and not in cells of brain tissue of said animal.

20. The animal of claim 16 or claim 17, wherein expression of said transgene in said animal results in a neuropathological condition.

21. The animal of claim 20, wherein said neuropathological condition comprises deposition of amyloid plaques, formation of intracellular neurofibrillary tangles, neuritic dystrophy or neuronal loss in brain tissue of said animal.

22. The animal of claim 20, wherein said neuropathological condition is selected from the group consisting of an Alzheimer's disease, an Alzheimer's-like disease, a familial Alzheimer's disease, and a hereditary cerebral hemorrhage with amyloidosis - Dutch Type.

23. The animal of claim 1 or claim 2, wherein said non-human animal is a mammal.

24. The animal of claim 23, wherein said mammal is a rodent.

25. The animal of claim 24, wherein said rodent is a mouse.

26. The animal of claim 24, wherein said rodent is a rat.

27. A method of testing behavioral function associated with a neuropathological condition, said method comprising providing a transgenic animal of claim 23 and testing said transgenic animal for deficits in cognitive ability during the development of a neuropathological condition in said transgenic animal.

28. A method of testing treatment for a neuropathological condition said method comprising providing a transgenic animal of claim 23, exposing said transgenic animal to said treatment and evaluating the effect of said treatment on the development of a neuropathological condition of said transgenic animal.

29. A recombinant transgene coding for an amyloid protein under the transcriptional control of a synapsin gene promoter.

30. A recombinant transgene coding for an amyloid protein under the transcriptional control of a human cytomegalovirus gene promoter.

31. The recombinant transgene of claim 29 or claim 30, wherein said recombinant transgene contains DNA coding for a RNA processing signal.

32. The recombinant transgene of claim 31, wherein said DNA comprises the SV40 small t intron sequences.

33. The recombinant transgene of claim 32, wherein said SV40 small t intron is proximal to DNA coding for an amyloid protein.

34. The recombinant transgene of claim 32, wherein said SV40 small t intron is distal to DNA coding for an amyloid protein.

35. The recombinant transgene of claim 31, wherein said DNA comprises a SV40 polyadenylation signal.

36. The recombinant transgene of claim 35, wherein said SV40 polyadenylation signal is distal to DNA coding for an amyloid protein.

37. The recombinant transgene of claim 31, 32, or claim 35, wherein said amyloid protein DNA comprises sequences coding for a precursor protein.

38. The recombinant transgene of claim 37, wherein said precursor protein DNA comprises sequences coding for a human APP695, APP751, or APP770.

39. The recombinant transgene of claim 38, wherein said DNA comprises fragments of sequences encoding human APP695, APP751, or APP770.

40. The recombinant transgene of claim 39, wherein said fragments are CIOO or 3/A4.

41. The recombinant transgene of claim 37 or claim 38, wherein said amyloid protein or precursor protein DNA comprises a coding sequence containing a mutation.

42. The recombinant transgene of claim 41, wherein said mutation is the hereditary cerebral hemorrhage with amyloidosis - Dutch Type (HCHWA-DT) .

43. The recombinant transgene of claim 41, wherein said mutation is the familial Alzheimer's disease (FAD) .