WO1992006187A1 - Transgenic animals with alzheimer's amyloid precursor gene - Google Patents

Abstract

A transgenic rodent useful for studying Alzheimer's disease having a transgene comprising a mammalian metallothionein I (MtI) promoter operably linked to a nucleotide sequence encoding Alzheimer amyloid precursor protein (AAP protein) operably linked to a mammalian growth (GH) hormone 3'-untranslated region is disclosed.

Description

TRANSGENIC ANIMALS WITH ALZHEIMER'S AMYLOID PRECURSOR GENE

FIELD OF THE INVENTION

The present invention relates to transgenic animals useful as models for studying Alzheimer's disease and useful for identifying compounds for treating Alzheimer's disease.

BACKGROUND OF THE INVENTION

Alzheimer's disease (AD) is the most common cause of dementia in late life. AD results in a progressive loss of intellectual function characterized by progressive impairments in memory, language, visuospatial skills and behavior. Those afflicted with AD eventually become unable to speak or think or take care of themselves. AD is a terminal disorder, but patients generally die of some complication that afflicts bedridden patients. It is estimated that in the United States, from 1.5 to two million people suffer from this degenerative disorder of the central nervous system.

What causes AD and how its characteristic changes are brought about are not known.

There is no known treatment or cure for AD. The diagnosis of AD can only be inferred during the patient's lifetime since no unique pattern of behavioral abnormalities has been established and there is no satisfactory laboratory test short of a brain biopsy. An autopsy, however, shows highly characteristic pathologic changes in the brain.

The clinical manifestations of AD are the result of a degeneration of neurons, particularly in regions essential for memory and cognition, or thought processes. There is a loss of neurons located in the basal forebrain cholinergic complex, several monoaminergic brainstem nuclei, amygdala, hippocampus and neocortex. There is a significant loss of neurons in certain more primitive regions at the base of the brain, with consequent reduction in the amount of the neurotransmitters, notably acetylcholine, normally released from the terminals of those neurons in higher brain centers.

AD is associated with abnormal protein structures. The three major pathologic signs of AD are neurofibrillary tangles within neurons, amyloid surrounding and invading cerebral blood vessels and amyloid-rich plaques proximal degenerating nerve terminals. Each of these signs reflects an accumulation of proteinaceous structures not normally found in the brain.

Neurofibrillary tangles result from accumulation of proteinaceous deposits which form abnormal fibers within the perikaryon of neurons. These accumulations of twisted filaments and other abnormal structures are found within neuronai cell bodies and contribute to the degeneration of nerve cell processes.

In addition to neurofibrillary tangles, a central feature of the pathology of AD is the presence of deposits of amyloid within plaques and around blood vessels. The major diagnostic lesion of AD is the deposits of abnormal amyloid proteins in intracellular and extracellular locations. The cellular dysfunction and death that eventually result from these deposits are common consequences of diseases termed "amyloidosis", which are characterized by the deposition of abnormal fibrillar proteins in these extracellular and intracellular spaces.

The term "amyloid" is applied to pathological accumulations within tissues of a protein-rich mass notable mainly for its staining properties: when amyloid is stained with a dye called Congo red and viewed under polarized light, it emits a greenish yellow glow, and under polarized light, a red/green birefringence. Some amyloid is seen in the brain of most old people and in other organs, such as the liver and kidney, of people with certain chronic diseases. Abundant cerebral amyloid is, however, always associated with AD, where it is seen as deposits in and adjacent to blood vessels and as a components of neuritic plaques. The abnormal proteins of the neurofibrillary tangles also can exhibit the staining properties of amyloid.

The neuritic (or senile) plaque is the pathological structure whose presence signals AD to the neuropathologist. Plaques are usually most abundant in die cerebral cortex and hippocampus and in the amygdala, a nucleus of cells near the hippocampus that seems to be particularly damaged in the disease. Within each region the plaques are localized in areas containing the axonal terminals of neurons rather than their cell bodies. The consistent evidence that the fibrillar deposits in plaques and cerebral vessels are amyloid fibers and that the paired helical filaments in tangles are twisted, β -pleated sheet fibrils, have led to die conclusion that AD is a form of cerebral amyloidosis. This signifies that the above lesions may be directly or indirectly responsible for neuronai cell death and represent an important stage of the pathogenetic process leading to AD.

Biochemical studies have revealed that the plaque core protein in AD is formed from a 4500-dalton protein. The protein is referred to as either amyloid A4, or as me β-protein. The full- length protein consists of only 42 to 43 residues. The discovery of β -protein from amyloid-laden cerebral vessels of patients with AD has provided a means to begin deciphering the pathogenesis of AD.

Considerable evidence has accumulated that most amyloid fibril proteins are formed from precursor proteins by proteolytic cleavage to produce β-pleated sheet fibrils and that the precursor proteins have an abnormal sequence or amino acid substitution. Based on these precedents, one would expect the amyloid fibril β-protein of cerebrovascular amyloid, having a maximum of 43 amino acids, to be formed by proteolytic cleavage of a putative abnormal β-protein precursor. Proteolysis of the precursor to form β-protein is accepted; however, despite precedent, no evidence for an abnormal β-protein precursor in AD has thus far been demonstrated. Cloning and cDNA sequencing have indicated that the self-aggregating amyloid protein of AD is encoded as part of one of three larger precursor protein genes. Each protein is referred to as the Alzheimer's Amyloid Precursor Protein (AAP Protein). The respective proteins have 695 residues (AAP₆₉₅), 751 residues (AAP₇₅₁), and 770 residues (AAP₇₇₀). The AAP proteins are encoded by a unique gene on chromosome 21. The various mRNAs are generated by alternative splicing of this gene's primary transcript. An interesting observation is that the brains of Down's patients who grow to adulthood degenerate in much the same way as those of Alzheimer's patients. Bio- chemical studies have revealed that the plaque core protein in both Alzheimer's disease and Down's syndrome is the identical β -protein. Since the gene encoding AAP protein resides on chromosome 21, overexpression of all AAP may affect the associated amyloidosis. Thus, any treatment to slow or prevent the progression of AD may be useful in the treatment of adult Down's patients.

Presently, the only animal models available to study AD and screen compounds which may be useful for treatment of AD are aged primates which exhibit age-associated memory deficits.

These animals display structural/chemical changes in the brain similar to those found in aged humans, particularly those suffering AD. However, the usefulness of these animals is limited and a better animal model is desired.

Among the uses foreseen for a better AD animal model is the ability to use such a model to screen compounds useful in prevent, slow or reverse the accumulation of amyloid in the brain.

Identification of such compounds could provide potential therapeutics for AD.

The present invention provides a transgenic animal useful as a model to study the accumulation of amyloid in brain tissue. Furthermore, the present invention relates to a transgenic animal useful in the identification of compounds which can prevent, slow or reverse the accumulation of amyloid in the brain. The present invention provides a transgenic animal useful in the discovery of drugs for the treatment AD and for the prevention of brain tissue degeneration in adults with Down's syndrome. According to the present invention, a transgenic animal is provided which displays tissue-specific overexpression of a gene encoding AAP protein in the regions of the brain where amyloid deposits are commonly found in patients with AD. Thus, amyloid deposits are produced in the transgenic animal models of die present invention in the same pattern as those occurring in AD patients. The transgenic animals of the present invention therefore provide an in vivo model which possesses a physical condition that closely resembles a pathological condition of patients afflicted with AD.

INFORMATION DISCLOSURE

Swanson, et al., "Novel developmental specificity in the nervous system of transgenic animals expressing growth hormone fusion genes", Nature, Vol. 317, 26 September 1985, pp. 363- 366, report that transgenic animals expressing rat-growth hormone (rGH) under ; control of the mouse metallothionein I (mMtl) promoter express such proteins in a tissue specific pattern in neuronai cells. Similar experiments were performed using transgenes containing mMtl promoter controlling expression of a human growth hormone gene. Localized expression in neuronai tissue of transgenic mice was observed. It is noted that neither metallothionein nor growth hormone are locally expressed in the neuronai cells which express these fusion genes. Fusion proteins containing other structural genes under the control of the mMtl promoter did not exhibit similar pattern of expression.

Russo, A. F., et al., "Neuronal Expression of Chimeric Genes in Transgenic Mice", Neuron, Vol. 1, June, 1988, pp. 311-320, reports chimeric genes containing the mMtl promoter linked to either rGH or hGH genes or the calcitonin/CGRP gene are expressed in very similar patterns of neuronai regions. It is suggested that the ectopic expression which is unexpected is due to regulatory signals from multiple DNA elements; that is, the interplay between the mMtl promoter and the 3' region of growth hormone gene bring about expression.

Evans, R.M. et al., "Inducible and Developmental Control of Neuroendocrine Genes", Cold Springs Harbor Symp. Quant. Biol., Vol. 50, pp. 389-397, report that the localized pattern of expression of fusion genes containing the mMtl promoter and 3' untranslated flanking regions of growth hormone genes results from the combination of such gene elements in fusion genes.

Kang, J., et al., "The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor", Nature, Vol. 325, 19 February 1987, pp. 733-736, report the isolation and sequence of a full length cDNA clone encoding a 695-residue precursor of the amyloid proteins subunit A4.

Pome, P., et al., "A new A4 amyloid mRNA contains a domain homologous to serine proteinase inhibitors", Nature, Vol. 331, 11 February 1988, pp. 525-527, disclose a novel gene encoding AD protein. The novel precursor is longer than the AAP₆₉₅. It contains an additional 168 base-pair insert, encoding a 56 amino acid domain within the so-called extracellular region of the protein.

Kitaguchi, N., et al., "Novel precursor of Alzheimer's disease amyloid protein shows protease inhibitory activity", Nature, Vol. 331, 11 February 1988, pp. 530-532, report a novel precursor of the amyloid protein A4. This novel precursor is longer than AAP₇₅₁. It contains an additional 57 base pairs encoding a 19 amino acid domain of unknown function, inserted immediately C-terminal to the insert in AAP₇₅₁.

Selkoe, D. J., "Deciphering Alzheimer's Disease: The Amyloid Precursor Protein Yields New Clues", Science, Vol. 248, pp. 1058-1060, provides a review of AAP genes and proteins. It is reported that the gene occurs in three forms, AAP₆₉₅ AAP₇₅₁, and AAP₇₇₀ and a discussion of the conversion from precursor to the amyloid β-protein is included.

Wurtman, R.J., "Alzheimer's Disease", Scientific American 252:62-74, provides a review of six hypotheses which underlie the current focus on research on AD. The abnormal protein model that is reported in the reference discusses the presence of amyloid deposits in the brains of patients afflicted with Alzheimer's disease.

Glenner. G.G., "The Pathobiology of Alzheimer's Disease", Ann. Rev. Med. 40:45-51 (1989), provides a review of the pathology of AD. The role of the β -protein as the major component of amyloid fibrils of plaques and cerebral vessels and the paired helical filaments of neurofibrillary tangles is discussed.

Muller-Hill, B. et al, "Molecular Biology of Alzheimer's Disease", Annu. Rev. Biochem.,

58:287-307 (1989), provide a review of the molecular biology of Alzheimer's disease. A discussion of the genes encoding die β-protein, referred to as A4 amyloid, is included. Additionally, the cDNAs of AAP protein, the genes encoding AAP protein, and the link of the

AAP protein with AD are discussed.

Price, D.L. et al., "Cellular and Molecular Biology of Alzheimer's Disease", BioEssays, Vol. 10, Nos. 2 & 3, February-March 1989, pp. 69-74, provide a review of the cellular and molecular biology of Alzheimer's disease. Included is a discussion of me animal models presently being used. The section entitled "Animal Models" reports the use of nonhuman primates, specifically aged Rhesus monkeys. The usefulness and shortcomings of these models are reported. In addition, the use of transgenic mice as a potential animal model for AD is suggested. The advantages of such transgenic mice models are outlined and research strategies using these mice are proposed. However, at page 72, column 3, line 39, it is noted that a crucial problem exists in designing a transgene which provides tissue specific expression. The present invention overcomes this obstacle.

U.S. Patent Number 4,736,866 issued April 12, 1988 to Leder et al discloses a transgenic non-human animal having a transgene comprising an activated oncogene sequence which increases the probability of development of neoplasms in the animal.

Strojek R.M., et al, The Use of Transgenic Animal Techniques for Livestock

Improvement, Genetic Engineering: Principles and Methods, J.K. Setlow, Ed. Vol. 10 New York (1988) reviews work in the area of transgenic mice. Methods are disclosed and various transgenic lines are described and discussed.

Skangos and Bieberich, Gene transfer into mice, Advances in Genetics, 24:285-322 (1987), provide a review of work in the area of transgenic mice. A list of reported transgenic mice species is included, listing various transgene constructs introduced into mice.

Palmiter, R.D. et al., Nature (London) 300:611-615 (1982) refers to a transgenic mouse containing a recombinant gene comprising mMtl promoter and rGH sequences. The mMtl promoter is inducible by the presence of heavy metal. Thus, expression of the rat growth hormone may be controlled.

SUMMARY OF THE INVENTION

The present invention provides a transgenic rodent having a transgene comprising a mouse metallothionein I (mMtl) promoter operably linked to a nucleotide sequence encoding Alzheimer amyloid precursor protein (AAP protein) operably linked to a mammalian growth (GH) hormone 3 '-untranslated region. The present invention also provides a recombinant DNA molecule comprising a mammalian Mtl promoter operably linked to a nucleotide sequence encoding AAP protein operably linked to a mammalian GH 3 '-untranslated region.

DETAILED DESCRIPTION

The major impediment to etiological studies of Alzheimer's Disease (AD) and related drug development is die lack of any suitable animal model. With the increasing evidence in the literature that amyloid deposition is an early, if not primary event in AD pathogenesis, the present invention relates to transgenic animals which will develop Alzheimer-type amyloid deposits in brain regions corresponding to those effected in AD. These animals can be used as a basis for studies of AD etiology and as a screening system for novel compounds designed to interfere with the process of amyloid deposition.

To develop an animal model according to the present invention, transgenic animals are produced which carry a transgene whose expression results in tissue-specific amyloid deposition.

Expression of the transgene must occur at a high level and in specific regions of the brain in order for the transgenic animal to provide a suitable AD model.

As used herein, the term "ectopic expression" means expression of a transgene in neurons within regions of the brain which do not correspond to regions normally directed by the control sequences, i.e. the promoter and me 3'-untranslated sequence.

As used herein, the term "ectopic regulatory sequences" means those genetic regulatory sequences which when operably linked to a gene, facilitate the ectopic expression of the gene.

Transgenes according to the present invention referred to herein as "AAP transgenes" are constructed to contain ectopic regulatory sequences operably linked to an Alzheimer's Amyloid

Precursor gene (AAP gene).

As used herein, the term "Alzheimer's Amyloid Precursor gene" or "AAP gene" means a nucleotide sequence which encodes an Alzheimer's Amyloid Precursor protein (AAP protein), a protein that can be processed into Amyloid β-protein, or the amyloid β-protein itself. AAP genes include gnomic clones, cDNAs, synthetically produced nucleotide sequences and combinations thereof. Conventions used to represent plasmids and fragments in Charts 7-13, though unique to this application, are meant to be synonymous with conventional representations of plasmids and their fragments. Unlike the conventional circular figures, the single line figures on the charts represent both circular and linear double-stranded DNA with initiation or transcription occurring from left to right (5' to 3'). Asterisks (*) represent the bridging of nucleotides to complete the circular form of the plasmids. Endonuclease restriction sites are indicated above die line. Gene markers are indicated below the line.

Transgenic animals carrying AAP transgenes may be produced using techniques well known by those having ordinary skill in the art. Transgenic animals carrying AAP transgenes will be genetically programmed to overexpress AAP genes in neurons of regions of the brain corresponding to those regions in humans which are effected in AD. in order to facilitate the development of amyloid deposits. Accordingly, such transgenic animals are useful in studying AD and in drug discovery efforts. As a screening tool, the transgenic animals according to the present invention can be used to identify compounds which are useful to prevent, impede or reverse the progression of AD and the accompanying brain function loss and dementia brought upon by amyloid deposition.

The ectopic regulatory sequences are modeled after the chimeric promoter system originally described by Swanson et al. Swanson placed the structural genes for both rat and human growth hormone (GH) under the control of the mouse metallothionein-I (mMtl) promoter in transgenic mice. For unknown reasons, ectopic expression of GH occurred in the brains of these mice; specifically, in neurons within regions later discovered to correspond to many of those primarily effected in AD. The observation was made in constructions containing rat growth hormones (rGH) sequences and in constructions containing human growth hormone sequences (hGH). This was an unexpected observation since neither mMtl nor GH are normally expressed in these neurons. It was subsequently found that when GH coding sequences were replaced by those for calcitonin/CGRP (calcitonin gene-related peptide), a similar pattern of ectopic expression was obtained as long as the mMtl promoter and the GH 3'-untranslated region (3'-UTR) flanked the cDNA. It therefore appears that some undefined interaction between these sequences directs die expression of inserted cDNAs in the neurons of brain regions of the brain that degenerate in AD.

The GH 3'-UTR from several species has been shown to provide similar results. Accordingly, transgenes according to the present invention may comprise any mammalian GH 3'-

UTR sequences.

Essential to the present invention is the ectopic expression of the gene introduced in the transgenic animal. This ectopic expression is accomplished by the unexpected interaction of the promoter and the 3'-UTR of the ectopic regulatory sequences.

A transgenic animal according to the present invention will have the predisposition to develop Alzheimer's-related brain amyloidosis. Thus, an essential feature of the present invention is a transgene which contains a gene that encodes a protein or preprotein which, when expressed ectopically, results in the brain amyloidosis condition.

The amyloid protein, also referred to as the β-protein, is a 42-43 amino acid protein that is originally expressed as a precursor protein. Three different forms of precursor proteins have been identified. The dominant form in brain tissue is produced by translation of mRNA encoding a 695 amino acid polypeptide. Two other forms have also been described: one contains 751 amino acids, the other contains 770. The present invention uses any of the three precursor forms in the transgene. When expressed, each precursor form is subsequently processed to generate the amyloid deposit.

Each of the three precursors used contain a transmembrane domain. When the native AAP protein is produced, it is thought to be partially secreted out of the cell. Three contiguous lysine residues, c-terminal to the single domain effectively serves as a cytoplasmic anchor, preventing full secretion of the molecule. In addition to transgenes made using the native AAP coding sequences, transgenes were also made using modified AAP coding sequences. Each of the three AAP coding nucleotide sequences were subjected to mutagenesis to convert a codon in the transmembrane domain into a stop codon. The modified AAP coding sequences when expressed produce truncated proteins that no longer contain the cytoplasmic anchor. These truncated AAP proteins are secreted.

The starting materials used to produce transgenes and transgenic animals according to die present invention are readily available to one having ordinary skill in die art. Metallothionen-I promoters are well known in the art. The mMtl promoter is well known in the art and can be purchased (Nichols Institute) or readily obtained from natural sources by those having ordinary skill in the art using well known techniques. Similarly, mammalian GH gene 3 '-untranslated region sequences are readily available. Such sequences are well known and can be purchased (Nichols Institute) or readily obtained from natural sources by those having ordinary skill in the art using well known techniques. Any of the three forms of the AAP gene are also readily obtained from natural sources by those having ordinary skill in die art using well known techniques. Chart 1 shows the amino acid sequence of AAP₆₉₅. Chart 2 shows the cDNA nucleotide sequence encoding AAP₆₉₅. Chart 3 shows the amino acid sequence of AAP₇₅₁. Chart 4 shows the cDNA nucleotide sequence encoding AAP₇₅₁. Chart 5 shows the amino acid sequence of AAP₇₇₀. Chart 6 shows the cDNA nucleotide sequence encoding AAP₇₇₀. This sequence can be used by one having ordinary skill in the art to obtain a copy of the gene. Alternatively, one having ordinary skill in the art can produce a transgene according to die present invention or one or more components of the transgene by synthesizing the nucleotide sequences using well known nucleotide sequence synthesizer technology.

Transgenic animals are animals which have integrated foreign DNA in their somatic cells and germ cells. The most common way of introducing the foreign DNA into the animal is by either microinjection or retroviral infection of the animal when it is in an embryonic state. The foreign DNA then integrates itself into the genetic material of the animal after which it is replicated along with the native genetic material of the animal during the development and life of the animal. Additionally, because the foreign DNA is integrated into the germ cell DNA, the offspring of such an animal will contain copies of the foreign DNA. Transgenic animals according to the present invention can be made following the procedure described in U.S. Patent No. 4,873,191 issued October 10, 1989 to Wagner et al., which is incorporated herein by reference.

The present invention provides an AAP operably linked to ectopic regulatory sequences. Constructs according to the present invention contain mammalian Mtl promoters operably linked to AAP genes operably linked to mammalian GH3'-UTR sequences. Optionally, nucleotide sequences encloded mammalian GH signal sequences, including the intron contained therein, operably linked upstream of the AAP gene are included in the present invention. Rodent species, especially rats, are particularly useful, since rats provide a wider array of behavioral and physiological paradigms than mice. Contemplated equivalents include transgenes that contain ectopic regulatory sequences operably linked to incomplete fragments of the AAP gene such that expression of the transgene results in formation of amyloidosis conditions. Contemplated equivalents of animal models according to the present invention include other non-human mammals which comprise the ectopic AAP transgene and equivalents thereof.

Example 1 Production of transgenic mice with transgene pNAN

Construction of pNAN

The first transgene construct described herein is referred to as pNAN. The transgene contains coding sequences from AAP₆₉₅ operably linked to and between the mMtl promoter linked to the sequence encoding the bovine growth hormone signal sequence, including the intron contained therein, and the 3' flanking regions of the bovine growth hormone (bGH) gene. The transgene was constructed by inserting a fragment of AAP₆₉₅ consisting bases 1923-2233 into a plasmid, pBGH-10, which contains the appropriate ectopic regulatory sequences.

The vector pBGH-10 is described in Kelder, B. et al. Gene 76:75-80 (1989) which is incorporated herein by reference. pBGH-10 contains the bGH structural gene placed under the control of the mMtl promoter.

AAP sequences used were subclones of AAP₆₉₅ cDNA obtained from a human brain cDNA library. The human brain cDNA library, and appropriate host cells were from Clontech (#HI1003, lot #2002). Clone Lambda SADE-1 was obtained from human brain cDNA library by hybridization with oligonucleotides BDG-1, BDG-2, BDG-4 and BDG-5.

BDG-1 5'- ccaatttttgatgatgaacttcatatcctgagtcatgtcg -3'

BDG-2 5'-gttctgcatctgctcaaagaacttgtaggttggattttcg-3'

BDG-4 5'-ctcggtcggcagcagggcgggcatcaacaggctcaacttc-3'

BDG-5 5'-cagagatctcctccgtcttgatatttgtcaacccagaacc-3'

A subclone, pSADE-1B, was constructed by inserting into pUC13 an EcoRI fragment from Lambda SADE-1 approximately 780 bp's extending from AAP₆₉₅ bases 1941 to about 2700. Plasmids described as shown in Chart 7.

Subclone AAP sequences from pSADE-1B were inserted within the bGH gene of pBGH-10. The coding sequence for the entire bGH signal sequence was maintained, including the intron contained within this region of the gene. The AAP sequences replaced codons #1-188 of mature bGH, maintaining the last three bGH codons plus its termination codon and 3'-UTR.

The inserted AAP sequence corresponded to AAP₆₉₅ bases 1923-2233. This was accomplished by appropriately adapting an EcoRI-Mael fragment (bases 1941-2233) from Lambda pSADE-1B, replacing the bGH segment in pBGH-10 from the Narl site at position 648 to the PvuII site at position 1942. The remaining AAP bases 1923-1940 were provided by the adaptors.

To form the 5'-insertion site, the bGH gene in pBGH-10 was cleaved with restriction enzyme Narl. The AAP gene fragment from pSADE-1B was cut with Mael, flush ended, and cleaved with restriction enzyme EcoRl . The 3' end of the pBGH-10 Narl fragment was linked to die 5' end of the pSADE-1B EcoRl fragment by inserting previously annealed oligos BDG-41 and

BDG-42.

BDG-41 5'-cgaagtgaagatggatgcag-3'

BDG-42 5'-aattctgcatccatcttcactt-3 '

The ligation of the fragments and the oligos resulted in the 5' insertion of AAP fragment into bGH. To form die 3 '-insertion site the bGH gene was cleaved with the restriction enzyme PvuII. The 5' pBGH-10 PvuII fragment was ligated to the 3' flush ended Mael fragment of pSADE-1B to form the 3' insertion of AAP into bGH.

The methods performed to generate completed construct are well known. The order of steps followed can be summarized as:

AAP:

1. Cut pSADE-1B DNA with Mael.

2. Flush-end pSADE-1B Mael fragments.

3. Cut fragments from step 2 with EcoRl.

Adaptors:

4. Phosphorylate BDG-42 with T4 polynucieotide kinase.

5. Anneal phosphorylated BDG-42 with BDG-41.

AAP + adaptors:

6. Ligate annealed oligos to DNA fragments from step 3.

7. Gel-purify appropriate 292 base-pair fragment.

pBGH-10:

8. Cut with PvuII + Nad.

9. Gel -purify appropriate fragment.

Final construction:

10. Ligate fragments from steps 7 and 9.

11. Transform into E. coli.

12. Sequence junctions by the standard techniques. The bGH-AAP 5'-junction was sequenced from die EcoRl site at AAP position 1941. The AAP-bGH 3'-junction was sequenced from the Asp718 site in die bGH 3'-UT.

The usefulness of pNAN construct was determined by in vitro transcription/translation of the pNAN sequence. For this purpose, the intron interrupting the bGH signal sequence had to be removed, and the sequence to be expressed was placed within a vector, pSP72 (Promega), that is suitable for transcriptional analyses.

The plasmid pNAN was cut and ligated to annealed oligos BDG-78 and BDG-79.

BDG-78 5'-agcttaccagctatgatggctgcaggcccccg-3'

BDG-79 5'-gtccgggggcctgcagccatcatagctggta-3'

To summarize the construction for transcription/translation experiments:

1. Isolate the 956 bp pNAN/Avall-Clal fragment.

2. Anneal oligos BDG-78 and BDG-79 create a Hindlll site at dieir 5' end.

In a 3-way reaction, ligate these annealed oligos, the 956 bp fragment and pSP72 cut with Hindlll and Clal.

3. Transform into E. coli and confirm by sequence analysis using the universal SP6 primer. This clone was called pSPNAN2.

Translation of capped in v/Vro-generated transcripts in the presence of ³⁵S-Met yielded radiolabeled protein which migrated at approximately 18-19 Kd on SDS-polyacrylamide gels. When translated in the presence of microsomal membranes, the band shifted slightly but perceptibly toward a lower molecular weight. This indicates appropriate initiation and cleavage of the signal sequence in the presence of the microsomal membranes.

To generate transgenic animals, the transgene segment was generated as follows. Asp718 sites exist within the mMtl promoter and the bGH 3'-UTR. The entire transgene was liberated from pNAN as an Asp718 fragment containing approximately 700 bp mMtl promoter, the 5'- flanking sequence of the bGH gene, the bGH-AAP segments described above, plus approximately 260 base pairs bGH 3'-UTR. This fragment is introduced into mouse embryos using the methods described in Wagner, T.E. et al, Microinjection of a rabbit β-globin gene into zygotes and its subsequent expression in adult mice and their offspring. Proc. Natl. Acad. Sci. USA Vol. 78, No. 10 pp.6376-6380, (Oct. 1981), and U.S. Patent Number 4,873, 191 issued Oct. 10, 1989 to Wagner, both incorporated herein by reference.

Two hundred thirty oocytes were microinjected with the construct and transferred into nine pseudo-pregnant female recipients. DNA was collected from the tails of 60 offspring and four were shown to contain the transgene (three males, one female). This was determined by Southern blot analysis of 10 μg DNA samples cut with Asp718, probed with radiolabeled pSADE-1B insert. These four F_o mice were bred with non-transgenic cohorts and a similar analysis was done on tail DNA obtained from the resulting F₁ offspring to determine whether the transgene was transmitted through the germ line. Two of these founders did transmit the transgene and the resulting lines were bred to homozygosity for the transgene array. Selected homozygous and heterozygous mice were placed on 76 mM ZnSO₄, while others were maintained on water without zinc. Following anesthetization by inhalation with metofane, brains were removed and RNA extracted. Densitometric analysis of Northern blots probed with the pSADE-1B insert revealed that homozygotes expressed 2-3 fold more transgene-coded RNA man heterozygotes. Moreover, zinc intake resulted in a 1-3 fold increase in transgene-coded RNA relative to littermates containing me same transgene copy number which were maintained without zinc.

Example 2 AAP cDNAs and Modifications

Each of the three known AAP cDN As were used as starting materials for constructions of transgenes. In addition to using subclones of each form of precursor, modifications were made to each form to insert a stop codon in the AAP coding sequences upstream from the region of the gene which encodes the cytoplasmic carboxyl terminus. When expressed, these modified AAP subclones produce molecules lacking the cytoplasmic anchor normally found in die AAP protein. Thus, the modified genes will produce modified proteins that are secreted.

For each AAP gene, cDNA was first isolated from a human brain cDNA lambda phage library using oligonucleotides BDG-1, BDG-2. BDG-4 and BDG-5. The AAP-encoding cDNA was then subcloned into pUC13 plasmids to facilitate further manipulations.

AAP₆₉₅ Plasmids:

To generate a full length clone of AAP₆₉₅, the N-terminal portion was recovered by amplifying cDNA from Alzheimer brain RNA using PCR. The C-terminal portion was recovered by subcloning a cDNA obtained from a human brain cDNA library.

Lambda SADE-1 which contains the AAP₆₉₅ cDNA was obtained from human brain cDNA library by hybridization with oligonucleotides BDG-1, BDG-2, BDG-4 and BDG-5. Lambda SADE-1 extend from AAP₆₉₅ bases 996 to approximately 2700. The 3'-terminus was not accurately established but this was unnecessary for further work.

Two subclones were obtained from Lambda S ADE-1 : pSADE-1 A and pSADE-1B. In both cases, the AAP₆₉₅ sequence from Lambda SADE-1 was subcloned into the EcoRl site of pUC13. Lambda SADE-1 was cut with EcoRl and the 947 bp EcoRl fragment extending from AAP₆₉₅ bases 996-1942 was inserted into pUC13, generating plasmid pSADE-1A. Plasmid pSADE-1B is the subclone of Lambda SADE-1 EcoRl fragment into pUC13 which contains approximately 780 bp EcoRl fragment extending from AAP₆₉₅ bases 1943 to about 2700.

Plasmid pSADE-3 which contains AAP₆₉₅ bases 131-1243 was derived from single stranded cDNA that was generated from Alzheimer brain RNA using BDG-75 as a primer, and double-stranded cDNA generated by PCR using BDG-74 and BDG-75 on cDNA template. The AAP sequences were subcloned as an EcoRl fragment into pUC13.

BDG-74 5'-gggaattcccccgcgcagggtcgcg-3'

BDG-75 5'-gggaattcgattccactttctcctg-3 '

Plasmid pSADE-4 contains AAP₆₉₅ bases 131-1942. The subcloned EcoRl inserts from pSADE-3 and pSADE-1A were inserted into pBR322 cut with EcoRl in 3-way ligation to generate the insert that includes AAP₆₉₅ bases 131-1942.

Plasmid pSADE-695 was constructed next. pSADE-695 contains AAP₆₉₅ bases 131 to about 2700 which constitutes essentially the full length coding sequence. To construct pSADE-695, EcoRl inserts from pSADE4 and pSADElB were subcloned into pBR322 cut with EcoRl. This was performed as a 3-way ligation.

Plasmid pSP695F contains the same AAP₆₉₅ insert as pSADE-695 subcloned into pSP73. Sense strand orientation reads 5' to 3' from the SP6 promoter. Plasmid pSP695R contains the same AAP₆₉5 insert as pSP695F except reverse orientation, i.e. sense strand orientation reads 5' to 3' from the T7 promoter.

Plasmid pSP695R-TL-f was derived from pSP695R. In order to remove the ATG codon between the T7 promoter and the AAP initiation codon, the plasmid was cut with Sail and Hindlll, flush-ended and religated. The Hindlll site was regenerated and the ATG codon was deleted.

Plasmid pSP₆₉5R-TL-s was engineered to encode a secreted form of AAP₆₉5 by replacing the valine codon diat is two positions downstream of the amyloidogenic domain (AAPg₉₅ amino acid 640) with a termination codon. This functionally deletes the C-terminal 56 amino acids including nine amino acids of the transmembrane domain, the cytoplasmic anchor and the entire cytoplasmic domain. To construct pSP695R-TL-s, the 565 bp EcoRI-Spel fragment of pSADE-1B (AAP₆₉₅ bases 1941-2504) was subcloned into M13mpl8 and mutagenized by site-directed mutagenesis using oligo BDG-80.

BDG-80 5'-catagcgacatagatcgtcatcacc-3'

The corresponding fragment was removed from pSP₆₉₅R-TL-f by limit digestion with Spel plus partial digestion with EcoRl, and replaced by this mutagenized fragment. In addition, sites for SphI, PstI, AccI and Sail were also deleted.

Plasmids pSP695R-TL

B-f and pSP695R-Tl B-s are clones that contain a BamHI site

deletion at AAP₆₉₅ position 1475. These were generated for use in the pSAR constructions. The site is deleted without altering the coding sequence. A 1600 bp Sad fragment was subcloned from pSP695R-TL-f into M13mpl9 (fragment extends from Sad site in vector polylinker through AAP₆₉₅ bases 131-1738), then mutagenized by site-directed mutagenesis with oligo DEL-2.

DEL-2 5'-gcatggtggaccccaagaaa-3'

The AccI-SacI fragment (AAP₆₉₅ bases 73-1738) in pSP695R-TL-f and pSP695R-TL-s were replaced with the corresponding mutagenized fragment.

Plasmid pAAP-695AB-f was constructed by subcloning the Nrul-Spel fragment of pSP695R-TL

B-f (AAP₆₉₅ bases 144-2504) into pGEM-5Zf( + )/EcoRV-SpeI.

Plasmid pAAP-695 B-s was constructed by replacing the EcoRI-Spel fragment of pAAP-

695AB-f (AAP₆₉₅ bases 1941-2504) with EcoRI-Spel fragment of pSP695-RTL B-s. This was then moved into the vector pGEM-5Zf( + ) cleaved with EcoRV and Spel as an Nrul-Spel fragment, deleting the EcoRV and Nrul sites, but maintaining the Spel site and placing the entire construct immediately downstream of an Ncol site necessary for further steps in the construction.

AAP₇₅₁ Plasmids

Plasmid pSADE-5 which contains AAP₇₅₁ bases 131-1411 was derived from single stranded cDNA that was generated from Alzheimer brain RNA using BDG-75 as a primer, and double-stranded cDNA generated by PCR using BDG-74 and BDG-75 on cDNA template.

BDG-74 5'-gggaattcccccgcgcagggtcgcg-3 '

BDG-75 5 '-gggaattcgattccactttctcctg-3 '

The AAP sequences were subcloned as an EcoRl fragment into pUC13.

Plasmid pSADE-7 contains AAP_{75 j} bases 131-2110. The subcloned EcoRl inserts from pSADE-5 and pSADE-1 A were inserted into pBR322 cut with EcoRl in 3-way ligation to generate the insert that includes AAP₇₅ι bases 131-2110.

Plasmid pSP751R-TL-f was constructed to replace a portion of AAP₆₉₅ in pSP695R-TL with a corresponding portion from AAP₇₅₁. The Accl-Xhol fragment of pSP695R-TL (AAP₆₉₅ bases 373-1056) was removed by limit digestion with AccI plus partial digestion with Xhol due to die presence of another Xhol site in the vector polyl inker. This fragment was replaced by the

Accl-Xhol fragment of clone pSADE-5 (AAP751 bases 373-1224) to generate pSP751R-TL.

Plasmid pSP751R-TL-s was engineered to encode a secreted form of AAP751 by replacing the valine codon that is two positions downstream of the amyloidogenic domain (AAP₆₉₅ amino acid 640) with a termination codon. This functionally deletes the C-terminal 56 amino acids including nine amino acids of the transmembrane domain, the cytoplasmic anchor and the entire cytoplasmic domain. To construct pSP751R-TL-s the 565 bp EcoRI-Spel fragment of pSADE-1B (AAP₆₉₅ bases 1941-2504) was subcloned into M13mpl8 and mutagenized by site-directed mutagenesis using oligo BDG-80.

BDG-80 catagcgacatagatcgtcatcacc

The corresponding fragment was removed from pSP751R-TL by limit digestion with Spel plus partial digestion with EcoRl, and replaced by this mutagenized fragment.

Plasmid pAAP-751 AB-f was constructed by replacing the Asp718-XhoI fragment of pAAP- 695 B-f (AAP₆₉₅ bases 203-1056) with Asp718-XhoI fragment of pSP751R-TL-f (AAP751 bases 203-1225).

Plasmid pAAP-751AB-s was constructed by replacing the Asp718-XhoI fragment of pAAP- 695 B-S (AAP₆₉₅ bases 203-1056) with Asp718-XhoI fragment of pSP751R-TL-f (AAP₇₅₁ bases 203-1225).

AAP₇₇₀ Plasmids

Plasmid pSADE-6 which contains AAP₇₇₀ bases 131-1468 was derived from single stranded cDNA that was generated from Alzheimer brain RNA using BDG-75 as a primer, and double-stranded cDNA generated by PCR using BDG-74 and BDG-75 on cDNA template. The AAP sequences were subcloned as an EcoRl fragment into pUC13.

BDG-74 5 '-gggaattcccccgcgcagggtcgcg-3 '

BDG-75 5'-gggaattcgattccactttctcctg-3'

Plasmid pSADE-8 contains AAP₇₇₀ bases 131-2167. The subcloned EcoRl inserts from pSADE-6 and pSADE-1A were inserted into pBR322 cut with EcoRl in 3-way ligation to generate die insert that includes AAP₇₇₀ bases 131-2167.

Plasmid pSP770R-TL-f was constructed to replace a portion of AAP₆₉₅ in pSP695R-TL with a corresponding portion from AAP₇₇₀. The Accl-Xhol fragment of pSP695R-TL (AAP₆₉₅ bases 373-1056) was removed by limit digestion with AccI plus partial digestion with Xhol due to the presence of another Xhol site in the vector polylinker. This fragment was replaced by the

Accl-Xhol fragment of clone pSADE-6 (AAP₇₇₀ bases 373-1224) to generate pSP770R-TL.

Plasmid pSP770R-TL-s was engineered to encode a secreted form of AAP₇₇₀ by replacing the valine codon diat is two positions downstream of the amyloidogenic domain (AAP₆₉₅ amino acid 640) with a termination codon. This functionally deletes the C-terminal 56 amino acids including nine amino acids of the transmembrane domain, the cytoplasmic anchor and the entire cytoplasmic domain. To construct pSP770R-TL-s, the 565 bp EcoRI-Spel fragment of pSADE-1B (AAP₆₉₅ bases 1941-2504) was subcloned into M13mp18 and . mutagenized by site-directed mutagenesis using oligo BDG-80.

BDG-80 5'-catagcgacatagatcgtcatcacc-3'

The corresponding fragment was removed from pSP770R-TL by limt digestion with Spel plus partial digestion with EcoRl. and replaced by this mutagenized fragment.

Plasmids pAAP-770AB-f were constructed by replacing the Asp718-XhoI fragment of pAAP-695 B-f (AAP₆₉₅ bases 203- 1056) with Asp718-XhoI fragment of pSP770R-TL-f (AAP₇₇₀

bases 203-1281).

Plasmids pAAP-770AB-s were constructed by replacing the Asp718-XhoI fragment of pAAP-695 B-s (AAP₆₉₅ bases 203-1056) with Asp718-XhoI fragment of pSP751R-TL-f (AAP₇₇₀

bases 203-1281).

Example 3 pSAR

Several transgenes were constructed containing nucleotide sequences from rat growth hormone (rGH). A vector, pSAR. was constructed which contains the mMtl promoter, the rGH signal sequence including the intron contained therein, and the rGH3'-UTR. Plasmid pSAR contains cloning sites which allow for insertion of AAP coding sequences which can then be expressed when the transgene constructed is liberated and used to generate a transgenic animal.

In order to construct a transgene according to the present invention using the rGH 3'- untranslated sequences, the growth hormone sequences must be modified. Thus, 5 segments of the rGH gene were subcloned into five different plasmids to facilitate manipulations. The five subclones were modified and ligated back together to produce a modified rGH sequence. A mouse mMtl promoter was then inserted upstream of the rGH material. The mMtl promoter was recovered from starting material and amplified using PCR technology which allowed for the generation of a Smal site at the 3' end which is not naturally present. This Smal site was useful in the ligation of the mMtl promoter to the rGH sequence. To complete the transgene construction, an internal portion of the rGH sequence was deleted and one of die six versions of the AAP sequence was inserted in its place. The inserted AAP sequence was then modified to place it in proper reading frame for expression in transgenic animals. Charts 8-12 illustrate plasmids constructed to make pSAR.

The starting material for the rGH 3'-UT was a rat growth hormone structural gene clone in bacteriophage Lambda-Charon 4A described in Chien, Y.-H. & E.B. Thompson, Proc. Natl. Acad. Sci. USA 77:4583-4587 (1980). Aliquots of this DNA were packaged using standard techniques, amplified, and DNA was extracted from the resulting bacteriophage preparations. The DNA was digested with both BamHI and Xhol. Fragments which migrated on agarose gels at about 5 Kb were purified. These fragments were subcloned into pSP73 (Promega) cut with the BamHI and Xhol (see Chart 8). Appropriate clones, designated pRGH, were identified by hybridization with BDG-86.

BDG-86 5'-caagaggctggtgctttccctgccatgccc-3'

The pRGH clone was divided into five fragments of workable size and complexity to enable appropriate modifications. Numbering was according to the rGH sequence coordinates:

#1 XhoI-PvuII fragment (407-789)

#2 PvuII-PstI fragment (789-1714)

#3 Pstl-PstI fragment (1714-2564)

#4 Pstl-PstI fragment (2564-3764)

#5 Pstl-BamHI fragment (3764-5644)

Modifications were performed on Pstl-PstI fragment #3 to enable insertion of the 3'-AAP sequences, including die AAP stop codon, immediately upstream of the rGH 3'-UT. These modifications were termed Step 1. The Pvull site within the 5th codon upstream of the rGH stop was selected as an insertion site. Since several PvuII sites exist within the rGH gene, it was necessary to mutate this one to enable insertion of the AAP cDNAs without further fragmentation of rGH sequences in the cloning vector. Hpal was chosen to replace PvuII. It was selected because die enzyme used needed to cut uniquely within rGH, and generate a flush-ended terminus to enable proper insertion of the AAP cDNAs. Replacement of the PvuII site with a Hpal site was accomplished using a PCR-based protocol.

New cloning vectors were generated for this section. pSP72 K was generated by cutting pSP72 (Promega) with Asp718. flush-ending and recircularizing. pSP72AKH was generated by cutting pSP72

K with Hpal + EcoRV and recircularizing.

The Step 1 segment was initially subcloned in Bluescript (Stratagene) as a PstI fragment referred to as pStep-PPP (see Chart 8). The strategy used to replace the PvuII site at rGH 2373 employed PCR on two segments of this clone.

The 5' Pstl-PvuII segment was mutated by amplifying the insert of pStepl in pStep-PPP using oligos BDG-158 and BDG-156.

BDG- 156 5' -ggggaattcgttaactgctttccgcaaagcggcg-3 '

BDG-158 5'-cagccctaactgcagtctaggcca-3'

BDG-158 corresponds to the rGH sequence surrounding the PstI site at position 1714. BDG-156 contains the rGH antisense sequence surrounding position 2373 (downstream of the PvuII site), but replaces the PvuII site with a Hpal site so that the amplified products contain die Hpal site in place of the PvuII site. BDG-156 also contains an EcoRI site downstream of the Hpal site to facilitate cloning. This PCR-generated fragment was subcloned into pUC13-SmaI as a blunt-ended fragment to generate pStepl-5'. It was inserted in such an orientation that there was an EcoRI site in die vector upstream of the 5'-end of this fragment, i.e. this fragment was now flanked by EcoRI sites.

Insertion was random, i.e. it was in both orientations. This orientation was selected since it was the one useful for the construct. The EcoRI site at the 5' end was supplied by die vector, since it exists upstream of the Smal site used for insertion. The EcoRI site at the 3'-end was created with the PCR primer, adjacent to the Hpal site on that primer.

The 3'-segment was mutated by amplifying the insert of pStepl-3' using oligos BDG-157 and BDG-159. BDG-159 corresponds to a cloning vector sequence 3' to the insert. BDG-157 contains the rGH sense strand sequence surrounding position 2373 (upstream of the Pvu site), but replaces the PvuII site with a Hpal site so that the amplified products contain the Hpal site in place of the PvuII site. BDG-157 also generates an EcoRI site upstream of the Hpal site to facilitate cloning. This PCR-generated fragment was cut with PstI + EcoRI and cloned into pSP72

KH cut with PstI + EcoRI.

BDG-157 5'-cccgaattcgttaacgctgtgctttctaggcacacac-3'

BDG- 159 5 ' -gacgttgtaaaacgacggccagt-3 '

Oligos BDG-156 and BDG-157 were designed so that the two PCR-generated Step 1 segments could be cut with Hpal and ligated together to yield the appropriate PvuII to Hpal modification at rGH position 2373. This was accomplished by cutting pStep 1-5' and pStep 1-3' with EcoRl and Hpal. The 5' portion of fragment #3 from pStep1-5' was subcloned into the 3' portion contained in pStep1-3' to generate the final Stepl plasmid. pStep1-PHP.

Modifications on the Xhol-PvuII fragment (407-789), termed Step 2 modifications were performed to permit insertion of the 5' AAP terminus near the first rGH codon downstream of the rGH signal peptidase cleavage site. This was accomplished by generating an Ncol site in rGH. The AAP sequences were then cloned into this site by using the immediately upstream Ncol site in vector pGEM-5Zf(+), in the pAAP series of constructs. These constructs were further modified by inserting appropriate oligonucleotide adaptors between the engineered rGH Ncol site and die natural AAP Asp718 site, so that the AAP sequence will begin with the first codon of the mature protein, expressed as a fusion with the first 5 rGH residues. This is designed so that the rGH signal sequence should be clipped within an rGH milieu.

pStep2 is the XhoI-PvuII rGH segment (coordinates 407-789) containing the engineered

Ncol site at coordinate 736, cloned as a flush-ended PCR-generated fragment into pUC13-SmaI (see Chart 9). To construct pStep2, the 382 bp XhoI-PvuII fragment was cloned into M13 cut with Xhol and Smal. The Ncol site was engineered by site-directed mutagenesis using oligo BDG-

112.

BDG-112 5'-ccctgccatggccttgtccag-3'

The mutated insert was then excised from the M13 clones by PCR from using primers BDG-122 & 123 to preserve Xhol site and regenerate the PvuII site.

BDG-122 5'-cagcagccagctggtgcaggtgctgggctc-3'

BDG-123 5'-tccagcaccctcgagcccagattccaaact-3;

The PCR-generated segment was sequenced to ensure its integrity. The XhoI-PvuII fragment was subcloned into M13/SmaI-XhoI, and the correct sequence confirmed tiirough the region going into the final construct, including the presence of the engineered Ncol site.

No modifications to the nucleotide sequence were required for the PvuII-PstI fragment (789-1714) and die Pstl-PstI fragment (2564-3764). However, both fragments were inserted into different vectors.

In step 3, plasmid pStep3 which contains the rGH PvuII-PstI fragment (coordinates 789- 1714) was cloned into similarly-cut pSP72 (see Chart 9).

In step 4, plasmid pStep4 which contains the rGH PstI fragment (coordinates 2564-3764) cloned into Pstl-cut Bluescript M13+-SK (see Chart 9).

Modifications were performed on the Pstl-BamHI fragment (3764-5644) to construct pStep5. pStep5 was produced when the rGH Pstl-BamHI fragment (coordinates 3764-5644) was cloned into similarly-cut pSP72

KH. Ncol and Asp718 sites were mapped within this rGH segment at positions 4760 and 5470. respectively. These sites must be deleted to enable appropriate cloning of AAP sequences. Since these are approximately 3000 bases away from the termination codon, it was felt that some modifications at these sites within the 3'-UTR are unlikely to be detrimental to the expressional specificity of the transgenic constructs. These sites were individually eliminated by cutting with the enzyme, flush-ending and recircularizing (see Chart 9).

Once modification were performed on the five subclones. the fragments were relinked to produce a modified version of the original rGH sequence. The attachment and modifications of plasmids from steps 1-5 was performed sequentially and produced a series of plasmids.

Plasmid pStep23 results from ligation of subclone XhoI-PvuII insert from pStep2 into similarly-cut pStep3, generating a plasmid containing rGH coordinates 407-1714 (see Chart 10).

Fragment 1 is then combined with pStep23 and a portion of the plasmid is deleted. Then fragments 4 and 5 are added. In order to subclone the rGH segments from steps 4 and 5 into the Step231 construct, the PstI site at position 1714 must be deleted. This is not problematic since the rGH region between coordinates 736-2373 is deleted in all pSAR-AAP constructs. However, the PstI site at position 469 within the pStep23 segment must be maintained. Therefore, the following steps were done in order:

1. Generate two subclones which separate the PstI site at position 469 in pStep23 from the remainder of the pStep23 insert at a site unique within pRGH (Bglll-PstI 1271-1714).

2. Subclone the pStepl insert into this plasmid and do the necessary deletion.

3. Subclone die pStep45 insert into this plasmid.

4. Add back the missing pStep23 sequences.

To accomplish this plasmid pStep23-XB is constructed. PStep23-XB is a subclone of pStep23 which contains the XhoI-BgUI fragment of pStep23 (rGH coordinates 407-1271) into similarly-cut ρSP72 (see Chart 10).

The remaining fragment of pStep23 is pStep23-BP, the subclone containing the Bglll-PstI fragment of pStep23 (rGH coordinates 1271-1714) inserted into similarly-cut pSP72 (see Chart 10).

The step 1 fragment is ligated to the pStep23-BP to form pStep23-BP-1. The step 1 fragment is the PstI fragment of pStep1-PHP (rGH coordinates 1714-2564). It is inserted into similarly-cut pStep23-BP. Orientation confirmed by analytical restriction digestions. Clone contains rGH coordinates 1271-2564 (see Chart 10).

Plasmid pStep23-BP-1

is pStep23-BP-1 with the necessary deletion. In order to delete the rGH region containing PstI-1714, pStep23-BP-1 was cut with Styl and Asp718, flush-ended and recircularized. This deleted the region between rGH coordinates 1396-1907 and recreated an

Asp718 site at the recircularized junction (see Chart 10).

pStep45 is the combination of fragments from steps 4 and 5. To form pStep45, the PstI fragment from pStep4 was subcloned into similarly-cut pStep5. Orientation confirmed by analytical restriction digestions. pStep45 contains rGH coordinates 2564-5644 (see Chart 11).

pStep23-BP-1

-45 resulted from a 3-way ligation of: 1 ) the Bglll-PstI insert from pStep23-

BP-1A (rGH coordinates 1271-2564): 2) the BamHI-PstKpartial) insert from pStep45 (rGH coordinates 2564-5604); and 3) phosphatased BamHI-Bglll-cut pSP72. Orientation confirmed by analytical restriction digestions. Clone contains rGH coordinates 1271-5604 with 1396-1907 deleted (see Chart 11). pStep231A45 was generated by the following steps:

1. Cut pStep23-XB and pStep23-BP-1A-45 with Bglll.

2. Ligate these two Bglll-cut plasmids.

3. Cut ligated DNAs with Xhol + BamHI and purify the 4726 bp fragment (rGH coordinates 407-5644 with 1396-1907 deleted).

4. Clone this XhoI-BamHI fragment into similarly-cut pGEM-11Zf(-), generally pStep231A45 (see Chart 11).

This completed the rGH segment of the pSAR construction. The rGH sequence could now appropriately be combined with the mMtl promoter to provide a vector in which AAP sequences can be inserted to form a working transgene.

The next step was to combine the mouse metallothionein-I promoter with the rGH segments. The starting plasmid for the metallothionein component was pXGH5 (Nichols Institute) (see Chart 12).

The desired junction between mMtl and rGH was generated by flush-ending die rGH Xhol terminus and recreating the remainder of the junction using PCR on the mMtl clone.

PCR utilizing primers BDG-68 and BDG-213 amplified an approximately 2100 bp fragment from pXGH5 containing pUC12 polylinker sequence (from pXGH5) at the 5'-end and the appropriate junction sequence at the 3 '-end.

BDG-68 5'-gttttcccagtcacgac-3'

BDG-213 5'-gggatctggtgaagctggag-3'

The fragment was cut with EcoRI and cloned into pSP73/EcoRI-SmaI generating plasmid pSP73mMtI; the EcoRI site was supplied by the amplified polylinker sequence from pXGH5, and the Smal site is supplied by flush-end ligation of the amplified sequence (contains half an Smal site) to the Smal site of pSP73. The Asp718 site must then be deleted from this mMtl segment since it will interfere with the cloning of AAP sequences into pSAR, but the change is irrelevant to the transgenic constructs since it will not be included within the transgenes (see Chart 12).

Thus,. to make a construct from the plasmids described that contains the mMtl promoter sequence upstream from the 3'UTR sequence of rGH, the following plasmids were constructed.

To construct plasmid pmMtl

K, mMtl sequence amplified from pXGH5 with primers BDG-68 and BDG-213 was cut with EcoRI and cloned into EcoRI-Smal-cut pSP73. The Asp718 site at mMtl position 1100 was deleted by cutting with Asp718, flush-ending and recircularizing

(see Chart 12).

Plasmid pSAR was then constructed. The EcoRI-Smal insert from pmMtl

K was subcloned into pStep231

45 cut with Xhol. flush-ended, then recut with EcoRI. This generated the completed pSAR cloning vector (see Chart 12). The sequence of the PCR-generated stretch of mMtl promoter was confirmed. Insertion of AAP sequence into pSAR

To insert AAP fragments into pSAR. AAP clones pAAP-(695.751,770)AB/(f,s) were cut with Mael. This cut the AAP clones at AAP termination codon and within pGEM5-Zf(+). Mael fragments were flush ended, recut with Ncol. and subcloned into pSAR cut with Ncol and Hpal.

The following plasmids were constructed:

pSAR-695NM/f,s: AAP-695 full length and secreted constructs from pAAP-695

B-s in pSAR, not yet adapted for expression.

pSAR-751NM/f,s: AAP-751 full length and secreted constructs from pAAP-751

B-s in pSAR, not yet adapted for expression.

pSAR-770NM/f,s: AAP-770 full length and secreted constructs from pAAP-770

B-s in pSAR, not yet adapted for expression. Collectively these plasmids are referred to herein as pSAR- NM.

The region between the Ncol site in the 5' rGH segment and the Asp718 site in me AAP segment had to be removed and replaced by appropriate adaptors to place me AAP sequences under proper control of the mMtl promoter and rGH signal sequence. This was accomplished as follows: pSAR-NM was cut with Ncol and Asp718 and NcoI-Asp718 annealed adaptors BDG-173

& BDG-174 were subcloned therein.

BDG 173 5'-catgctggaa-3'

BDG 174 5'-gtacttccag-3'

The following plasmids were constructed:

pSAR-695/(f,s): Final AAP-₆₉₅ full length and secreted constructs from pSAR-695NM/f,s, properly adapted for expression.

pSAR-751/(f,s): Final AAP-₇₅₁ full length and secreted constructs from pSAR-751NM/f,s, properly adapted for expression.

pSAR-770/(f,s): Final AAP-₇₇₀ full length and secreted constructs from pSAR-770NM/f,s, properly adapted for expression.

To generate transgenic animals, the transgene segment was generated as follows: pSAR- 695/(f,s) was cut with Bgll and BamHI. The entire transgene was liberated. The pSAR-695/(f,s) Bgll-BamHI transgenes were microinjected fertilized mouse ali following the directions in U.S. Patent 4,873,191. pSAR-695/f was injected into 668 eggs yielding 101 pups, and pSAR-695/s into 570 eggs yielding 94 pups. Of the 101 pSAR-695/f pups, 41 were shown to be potential founders by Southern blot analysis of DNA extracted from tails. Of the 94 pSAR-695/s

, 29 were similarly shown to potential founders.

Example 4 Mouse Brain Analyses

The brains of sacrificed transgenic mice are each analyzed as follows. Northern blot and ribonuclease protection assays on RNAs extracted from brain tissue are performed to evaluate gross transgene expression at the RNA level. Western blot analyses using several antisera that recognize the AAP region encoded by the pNAN transgene are perform ed on protein extracted from brain tissue to evaluate gross transgene expression at the protein level. In situ hybridizations are performed on sections of brain tissue to evaluate regional and cellular-specificity transgene expression at the RNA level. Immunocytochemistry studies on sections of brain tissue are performed to evaluate regional and cellular-specificity of transgene expression at the protein level. By histological methods including but not restricted to Congo red, Thioflavin T, Thioflavin S, silver staining methods are performed to evaluate neuronai and other pathological abnormalities.

Claims

1. A transgenic rodent comprising a mammalian metallothionein I (Mtl) promoter operably linked to an Alzheimer amyloid precursor gene (AAP gene) operably linked to a mammalian growth hormone 3' untranslated region (GH 3-'UTR).

2. A transgenic rodent according to Claim 1 wherein said Mtl promoter is a mouse metallothionein promoter (mMtl).

3. A transgenic rodent according to Claim 2 wherein said mammalian GH 3'-UTR is selected from the group comprised of mouse GH 3'-UTR (mGH 3'-UTR), rat GH 3'-UTR (rGH 3'-UTR), bovine GH3'-UTR (bGH3'-UTR) and human GH 3'-UTR (hGH 3'-UTR).

4. A transgenic rodent according to Claim 3 wherein said mammalian GH 3'-UTR is rGH 3'-UTR.

5. A transgenic rodent according to Claim 3 wherein said mammalian GH 3'-UTR is bGH 3'-UTR.

6. A transgenic rodent according to Claim 1 wherein said AAP cDNA is selected from the group consisting of AAP ₆₉₅, AAP₇₅₁ , and AAP₇₇₀.

7. A transgenic rodent according to Claim 3 wherein said AAP cDNA is selected from the group consisting of AAP _695, AAP₇₅₁ , and AAP₇₇₀.

8. A transgenic rodent according to Claim 7 wherein said AAP cDNA is AAP₆₉₅.

9. A transgenic rodent according to Claim 8 wherein said mammalian GH3'-UTR is bGH 3'-UTR.

10. A transgenic rodent according to Claim 8 wherein said mammalian GH-3'-UTR is rGH 3'-UTR.

11. A transgenic rodent according to Claim 1 further comprising DNA encoding of mammalian GH signal sequence.

12. A transgenic rodent according to Claim 11 wherein said AAP gene is AAP₆₉₅.

13. A transgenic rodent according to Claim 12 wherein Mtl promoter in mMtl, said mammalian GH-3'-UTR is bGH-3'UTR and said DNA encoding a mammalian signal sequence is DNA encoding bGH signal sequence.

14. A transgenic rodent according to Claim 13 wherein said Mtl promoter is mMtl, said mammalian GH-3'-UTR is rGH-3'UTR and said DNA encoding a mammalian signal sequence is DNA encoding rGH signal sequence.

15. A transgenic rodent according to Claim 1 wherein said rodent is a mouse.

16. A recombinant DNA molecule comprising a metallothionein I (Mtl) promoter operably linked to an Alzheimer amyloid precursor gene (AAP gene) operably linked to a mammalian growth hormone 3' untranslated region (GH 3-'UTR).

17. A recombinant DNA molecule according to Claim 16 wherein said Mtl promoter is a mouse metallothionein I (mMtl) promoter.

18. A recombinant DNA molecule according to Claim 16 wherein said mammalian GH 3'-UTR is selected from the group comprised of mouse GH 3'-UTR (mGH 3'-UTR), rat GH 3'-UTR (rGH 3'-UTR) bovine GH3'-UTR (bGH3'-UTR) and human GH 3'-UTR (hGH 3'-UTR).

19. A recombinant DNA molecule according to Claim 16 wherein said AAP gene is selected from the group consisting of AAP₆₉₅, AAP₇₅₁ , and AAP₇₇₀.

20. A recombinant DNA molecule according to Claim 16 further comprising DNA encoding of mammalian signal sequence.