WO2000028811A1

WO2000028811A1 - Animal model for neurological disorders

Info

Publication number: WO2000028811A1
Application number: PCT/US1998/024341
Authority: WO
Inventors: Beverly H. Koller; Amy R. Mohn
Original assignee: The University Of North Carolina At Chapel Hill
Priority date: 1998-11-13
Filing date: 1998-11-13
Publication date: 2000-05-25
Also published as: AU1409999A

Abstract

Genetically modified non-human animals are described which serve as a model for neurological disorders such as schizophrenia; the animals express a reduced level of NMDA receptor. Methods for making and using the genetically modified animals are also described.

Description

ANIMAL MODEL FOR NEUROLOGICAL DISORDERS

GOVERNMENT SUPPORT

The invention was supported, in whole or in part, by a grant T32-GM07092 from the National Institutes of Health. The Government has certain rights in the invention.

BACKGROUND OF THE INVENTION

Schizophrenia is arguably one of the most devastating mental illnesses when one considers the emotional and financial burden that it creates for victims of the disease and their families. Schizophrenia affects 1% of the population worldwide. The symptoms of schizophrenia include hallucinations, paranoia, delusions and formal thought disorders (the "positive" symptoms), and blunted affect, loss of motivation, social withdrawal and cognitive impairments (the "negative" symptoms). Although current drug therapies have been beneficial in the treatment of schizophrenia, even the most recently developed drugs fail to completely relieve the symptoms of the disease. Additionally, these drags may have undesirable side effects such as agranulocytosis, orthostatic hypotension sedation, convulsions and weight gain, all of which decrease the quality of life for the individual and often lead to non-compliance, resulting in relapse and hospitalization.

Current animal models for disorders in thinking and cognition (e.g., schizophrenia and delirium) are generated by treatment of animals with psychotic drugs such as phencyclidine (PCP). However, drug-induced models have a number of limitations. First, in addition to induction of the trait to be studied, the drug treatment can induce other changes in the animals that can affect assessment of the traits of interest. Additionally, use of one or more drugs to induce the disease prevents the use of the animal model to assess the efficacy of drugs that target the same molecule as the inducing drugs. Interactions between the test drugs and the drug used to induce the condition may result in unexpected changes in the physiology of the animal, making the results difficult to interpret. Finally, the acute disease brought about by brief pharmacological treatment may not be an accurate model of chronic disease states and the potentially associated secondary changes, particularly in the nervous system.

SUMMARY OF THE INVENTION Work described herein relates to the development of an animal model for neurological disorders using genetic alteration of expression of cell surface molecules in the mouse central nervous system. This model is useful in studying neurological disorders, learning, cognition, memory, dementia, addictive behavior and neuronal death. In a particular embodiment the invention is an animal model for schizophrenia. In one embodiment the invention relates to alteration of expression of glutamate receptors, e.g., NMDA receptors, in an animal such as a mouse.

As described herein, a mouse line has been generated in which the expression of a family of glutamate neurotransmitter receptors, NMDA receptors, is reduced to about 10% of normal levels. This was done by the manipulation of the structure of the Nrl gene in embryonic stem (ES) cells and the generation of a mouse line carrying the altered Nrl gene from these ES cells. Mice homozygous for this alteration were characterized and shown to display altered behavior that in rodents is believed to correspond to the neurological changes that occur in schizophrenia. Nrl message levels were reduced to less than 10% of normal levels in the brains of animals homozygous for this mutation.

The behavior of these mice is similar to that induced in normal animals with PCP, a drug that blocks ΝMDA receptors and is used as the best pharmacological model of schizophrenia in humans and in animals. Specifically, these animals exhibit hyperlocomotion that is both qualitatively and quantitatively similar to that seen in wild type mice that are treated with PCP or dizocilpine; this behavior is believed to be analogous to the positive symptoms of schizophrenia. In addition, the animals display altered social behavior that is analogous to the negative symptoms of schizophrenia. Finally, male mice fail to mate, and female mice do not care for their young. These behaviors appear to be manifestations of schizophrenia in animals. The behavior of mutant mice becomes indistinguishable from that of normal mice following treatment of the mutant mice with two different classes of antipsychotic drugs, haloperidol and clozapine. Thus, the correction of this altered behavior, e.g., altered mating and nurturing, provides a novel and rapid means by which to assess the efficacy of anti-psychotics. Other behavioral abnormalities include spasticity, increased sensitivity to ethanol, and an increased susceptibility to spontaneous and chemically-induced seizures.

One benefit of this model is that genetic alterations leading to a functional decrease in a molecule may more closely mimic the development of the human disease phenotype, particularly in that the target for potential therapeutic agents is still present in the animal. This allows the identification of drugs that can increase the activity of the receptor. Moreover, because this model addresses the role of glutamatergic neurotransmission in the pathology of schizophrenia, there is the potential for discovering drugs that are mechanistically different from current therapies and may be more effective in the normalization of schizophrenic symptoms. For example, NMDA receptor agonists or a combination of dopamine/serotonin antagonists and NMDA receptor agonists can be assessed in these model animals.

Thus, the invention relates to a genetically modified non-human animal expressing a reduced level of functional NMDA receptor, wherein said reduced level is not pharmacologically induced. In a particular embodiment the animal is a mouse. In a preferred embodiment, the genetically modified non-human animal exhibits behavior similar to the behavior of a wild type animal treated with an NMDA receptor antagonist. The described animals can be used as models of neurological disorders, such as schizophrenia.

The invention also relates to a neuronal cell line cultured in vitro from a genetically modified animal described herein, or from non-human embryonic stem cells that have the same genetic modification; such cultured cells will express a reduced level of functional NMDA receptor. In a particular embodiment, both copies of the gene for the NMDA receptor are modified as described herein to express a reduced level of functional NMDA receptor. Embryonic stem cells described herein can also be induced to differentiate into neurons, e.g., with retinoic acid, and the invention pertains to these differentiated cells having or expressing a reduced level of functional NMDA receptor. The invention also pertains to a method of making a genetically modified non-human animal, comprising the steps of introducing into a living, non-human embryo a nucleic acid molecule encoding all or part of an NMDA receptor subunit, wherein the nucleic acid molecule has been altered to result in the reduced expression of NMDA receptors or subunits thereof. The resulting embryo is allowed to complete development and then bred to generate animals heterozygous for said nucleic acid; further breeding of the heterozygous animals to produce a genetically modified non-human animal which is homozygous for said nucleic acid molecule can also be carried out. In a particular embodiment, the nucleic acid molecule is contained in an embryonic stem cell and the embryo is a blastocyst.

In one embodiment of the invention, the alteration comprises alteration, e.g., deletion or substitution of a nucleotide sequence, or insertion of one or more foreign sequences nucleotide sequences, in one or more non-coding regions of the nucleic acid molecule encoding the NMDA receptor. The foreign sequence can be any nucleotide sequence, including nucleotide sequences which do not encode a particular gene product. In a particular embodiment, the foreign sequence is selected from the group consisting of selectable markers. For example, the selectable marker can include, but is not limited to, the neo gene, the Hart gene, the hygromycin (Hyg) gene, the blasticidin (Bsr) gene, the zeocin gene, the thymidine kinase gene (e.g., the heφes simplex virus thymidine kinase gene) and the puromycin resistance gene. In one embodiment of the invention, the altered nucleic acid molecule encodes the NR1 subunit of an NMDA receptor. In another embodiment, the non- coding region of an NMDA receptor gene is selected from one or more introns, the promoter of an NMDA receptor gene, the 5' or 3' untranslated regions, and combinations thereof.

The invention also pertains to a method of assessing a compound for the ability to normalize the behavior of a genetically modified non-human animal described herein, comprising administering the compound to be assessed to the genetically modified animal and assessing the behavior of the animal relative to an untreated wild type animal. In another embodiment, the invention relates to a method of assessing a compound for the ability to function as an agonist of NMDA receptor function, comprising contacting a cultured neuron described herein with the compound to be assessed and assessing the effect on NMDA receptor function. The invention also relates to a method of assessing a compound for the ability to function as an agonist of NMDA receptor function, comprising administering the compound, with an optional physiologically acceptable vehicle, to a genetically modified non- human animal described herein and assessing the effect on NMDA receptor function. Compounds identified as having the desired activity can be used in the therapeutic treatment of neurological disorders such as schizophrenia.

BRIEF DESCRIPTION OF THE DRAWINGS

Figures 1A-1C show the introduction of a hypomoφhic mutation at the Nrl locus. The targeting strategy depicted in Figure 1A demonstrates the insertion of the 1.8 kb neo gene into intron 20 of Nrl. Homologous recombination results in a change of Hindlll restriction fragment length from 11 kb to 9 kb. Southern blot of genomic DNA was used to genotype drug-resistant cell lines and animals generated from ES cell transgenesis; the results are depicted in Figure IB. This Southern blot demonstrates the Hindlll restriction pattern for animals with a genotype of Nrl^neo +/+, Nr/"^eo +/-, and Nrl"^eo -/-. Figure 1C shows a Northern analysis demonstrating that Nrl ^e° is a hypomoφhic allele. Total RNA probed for Nrl message shows significant reductions of full-length Nrl message in animals homozygous for the Nri"^eoallele. Equivalent loading was verified by subsequent hybridization with a β- actin probe.

Figures 2A-F show that locomotor behavior of Nrl"^e° -I- mice is similar to that of wild type mice treated with MK-801 or PCP. Locomotion, measured as the total distance of horizontal movement traveled, and stereotypy, measured as the number of stereotypic behaviors, are increased in Nrl"^eo -I- mice (Figures 2A and 2B, respectively). Locomotion is 3.5 fold higher in mutant mice than in wild type mice (P-0.000095), and stereotypy is doubled (P=0.005). Increases in locomotion and stereotypy are quantitatively similar to the increases caused by administration of MK-801 (0.2 mg/kg) (Figures 2C and 2D, respectively) or PCP (3 mg/kg) (Figures 2E and 2F, respectively) in wild type mice. Locomotor activity for the experiments described in (2C-2F) are represented as the averaged sum of activity over a two-hour period ±S.E.M. Figures 3A-D show that MK-801 and PCP fail to elicit motor stimulatory effects in Nrl^neo -I- mice. The motor stimulatory effects of MK-801 are demonstrated in Figures 3C and 3D. MK-801 administration (0.2 mg/kg) causes marked increase of locomotion and stereotypy in wild type mice; MK-801 does not increase motor activity in Nrl^ne° -I- mice. In fact, stereotypy is actually reduced. Similar results are seen when PCP (3 mg kg) is administered (Figures 3A and 3B).

Figures 4A-4D show that haloperidol and clozapine ameliorate the increased motor activity oϊNrl^neo -I- mice. Inteφeritoneal administration of the anti-psychotic drug haloperidol (0.5 mg/kg) results in decreases in motor activity in wild type mice (black bar). The elevated motor activity of Nrl^neo -I- mice (white bar) is completely reduced to levels similar to haloperidol treated wild type mice. Inteφeritoneal administration of clozapine (0.5 mg/kg) normalizes both the locomoter activity (Figure 4C) and stereotype (Figure 4D) of Nrl^neo-/- mice (white bar) to wild type levels of activity (black bar). At this dose clozapine does not affect wild type motor activity. Motor activity is measured as the averaged sum of the total horizontal motion ±S.E.M (Figure 4 A) and as the averaged sum of the total number of stereotypic movements±S.E.M. (Figure 4B) over a two-hour observation period. Figures 5 A-B show that Nrl"^eo -I- mice display social withdrawal in the resident-intruder behavioral assay. Figure 5A shows that Nrl^ne° -I- mice (white bar, MT) spend less time over a 6-minute period actively pursuing investigation of the intruder mouse than wild type mice (black bar, WT). The social withdrawal of Nrl^neo -I- mice is improved by prior administration of clozapine ((0.5 mg/kg) striped bar, MT+CLOZ). Figure 5B shows that Nrl^neo -I- mice also exhibit escape behaviors, where they actively avoid social interaction with the intruder. These behaviors are rarely seen in wild type mice, and are diminished after treatment with clozapine. (**=statistically significant, P<0.05, two-tailed T-test; ***=highly statistically significant, P<0.005, two-tailed T-test.)

DETAILED DESCRIPTION OF THE INVENTION

The etiology of schizophrenia is complex, having both environmental and genetic components that contribute to the expression of the disease. The first theory that approached schizophrenia as a biochemical imbalance was the dopaminergic -7-

hypothesis of schizophrenia which suggested that the symptoms of schizophrenia were due to a hyperactivity of the dopaminergic neuro transmitter systems. This hypothesis was supported by convincing evidence that efficacy of classical neuroleptics strongly correlated with D2 dopamine receptor antagonism (Seeman et 5 al, Nature, 261:111-119 (1976)). However, this result does not prove that a hyperactive dopaminergic system is the primary defect in schizophrenia, nor does it exclude the possibility that other neurotransmitter systems are also affected in schizophrenia. In fact, antagonism of dopamine receptors is only effective in the amelioration of the positive symptoms of schizophrenia, while the negative

10 symptoms of the disease persist (Kane and Meyerhoff Br J Psychiatry, 755:115-118 (1989)). Furthermore, several weeks of treatment are required before classical neuroleptics gain therapeutic potency (Freed, Schizophrenia Bull, 14:269-211 (1988)). Thus, it appears that the dopaminergic hypothesis alone is insufficient to explain some aspects of the disease.

15 Another candidate system that may be altered in schizophrenics is the glutamatergic system (Carlsson, et al, Life Sciences, 61:75-94 (1997); Coyle, Harv Rev Psychiatry, 3:241-253 (1996); Tamminga, Crit Rev Neurobiol, 12:21-36 (1998)). There are several lines of evidence that glutamatergic neurotransmission is reduced in schizophrenic brains, and that this represents either a primary or

20 secondary defect that results in schizophrenic symptoms. There are four classes of glutamate receptors, one of which is the class of N-methyl-D-aspartate (NMDA)- type glutamate receptors.

The NMDA receptor is a critical component of neuron development and biology (Hahm et al, Nature, 357:568-570 (1991); Li et al, Cell, 76:427-437

25 (1994); Mitrovic et al, Eur J Neurosci, 5:1793-1802 (1996); Collingridge et al, J Physiol (London), 334:33-46 (1983); Stringer and Guyenet, Brain Res, 255:159-164 (1983); Morris et al, Nature, 319:114-116 (1986); Tsein et al, Cell 57:1327-1338 (1996)) and is implicated in a number of human disease states including epilepsy, neuronal death following excitotoxic insult, schizophrenia, and neurodegenerative

30 diseases such as Parkinson's, Huntington's and Alzheimer's disease. The NMDA receptor controls the flow of both divalent (Ca++) and monovalent (Na+, K+) ions into the postsynaptic neural cell. The relatively recent cloning of the genes that encode subunits of the NMDA receptor has broadened this base of knowledge and expanded the possibilities for further study at the molecular level (Moriyoshi et al, Nature, 354:31-31 (1991); Ishii et al. Biol Chem, 265:2836-2843 (1993); Yamazaki et al, FEBS Lett, 300:39-45 (1992); Meguro et al, Nature, 357:70-74 (1992); Kutsuwada et al, Nature, 355:36-41 (1992); Ikeda et al, FEBS Lett, 373:34- 38 (1992); Moyner et al, Science, 256:1217-1221 (1992); Hollman and Heinemann, Annu Rev Neurosci 77:31-108 (1994)).

All NMDA receptors characterized to date are composed of an NRl subunit and one of four different NR2 subunits. Because of this stoichiometry, alteration of the expression of NRl is expected to affect all NMDA receptor subtypes. Although null mutations of the Nr7 gene have been introduced by gene targeting (Li et al, Cell, 76:427-437 (1994); Forrest et al, Neuron, 73:325-338 (1994)), mice homozygous for null 7Vr7 mutations die perinatally. As a result these animals do not provide a means of addressing the role of these receptors in the adult animal. The most compelling evidence to support a role for ΝMDA receptors in the pathogenesis of schizophrenia is the observation that non-competitive antagonists of this receptor (phencyclidine, ketamine and dizocilpine) are psychotogenic. In fact, intoxication with PCP results in a psychosis that is indistinguishable from schizophrenia and provides a better model of the disease than amphetamine, as it reproduces both positive and negative symptoms (Javitt and Zukin, Am J Psychiatry, 745:1301-1308 (1991)). Administration of ΝMDA receptor antagonists also exacerbates symptoms in diagnosed schizophrenics (Itil et al, Can Psychiatr Assoc J, 72:209-212 (1967); Malhotra et al, Neuropsychopharmacology, 77:141-150 (1997)). Because these drugs act specifically to block ΝMDA receptors, it is reasonable to suggest that decreases in ΝMDA receptor function underlie the symptoms of schizophrenia.

From this initial hypothesis has come research that supports the idea that ΝMDA receptor function may be reduced in schizophrenics (Sherman et al, Biol Psychiatry, 30:1191-1198 (1991); Akbarian et al, J Neurosci, 76:19-30 (1996); Riva et al, Brain Res Mol Brain Res, 50:136-142 (1997); Humphries et al,

Neuroreport, 7:2051-2055 (1996); Farber et al, Schizophr Res, 21:33-31 (1996); Wang and Liang, Neuropsychopharmacology, 79:74-85 (1998); Moore et al, J Clin Psychiatry 59 Suppl, 70:37-44 (1997); Corbett et al, Psychopharmacology, 120:61- 74 (1996); Carlsson and Carlsson, Trends Neural Sci, 73:272-276 (1990)). While these studies do not prove that schizophrenia results from a defect in NMDA- mediated transmission, they show that suppression of NMDA receptors induces a state that is similar to schizophrenia.

Observation of the psychomimetic effects of NMDA receptor blockade in humans, and its similarity to schizophrenia, has been the basis for the use of PCP in animal models of schizophrenia. In rodent models of schizophrenia, PCP is used to induce psychosis as measured by increases in locomotion and stereotyped behavior (Deutsch, et al, Clin Neuropharmacol, 20:315 (1997); Corbett et al, Psychopharmacology, 120:61-14 (1995); Gleason and Shannon, Psychopharmacology, 129:19 (1997)) as well as negative symptoms measured by social interactions (Corbett et al, Psychopharmacology, 120:61-14 (1995)). The efficacy of antipsychotic drugs can be reliably predicted by the ability to attenuate these PCP-induced behaviors (Corbett et al, Psychopharmacology, 120:61-14 (1995); Gleason and Shannon, Psychopharmacology, 129:19 (1997)).

Unfortunately, current models to generate hypoactive NMDA receptors rely on the use of pharmacological agents. There are several disadvantages to drug- induced models of schizophrenia. In addition to issues of receptor specificity and drug interactions, there is also the concern that acute administration of drugs cannot mimic a chronic suppression of NMDA receptors. Because schizophrenia has been postulated to develop from abnormalities in the "wiring" of the developing brain, chronic suppression through all stages of development may provide a better model of schizophrenia than current drug-induced psychoses. As described herein, an animal model has been produced in which the level of functional glutamate receptors, e.g., NMDA receptors, is reduced. Despite the low levels of NRl expression, and in contrast to NRl null animals, these mice survive into adulthood to allow the study of NMDA receptors in vivo.

Genetic manipulation is carried out in order to reduce the functional level of NMDA receptor; that is, the nucleotide sequence encoding one or more subunits of the NMDA receptor is altered (i.e., contains an insertion, deletion or substitution of one or more nucleotides). The genes that encode subunits of the NMDA receptor have previously been cloned and their nucleic acid sequences published (Moriyoshi et al, Nature, 354:31-31 (1991); Ishii et al, JBiol Chem, 265:2836-2843 (1993); Yamazaki et al, FEBS Lett, 300:39-45 (1992); Meguro et al, Nature, 357:70-74 (1992); Kutsuwada et al, Nature, 355:36-41 (1992); Ikeda et al, FEBS Lett, 313:34- 38 (1992); Moyner et al, Science, 256:1217-1221 (1992); Hollmann and

Heinemann, Annu Rev Neurosci 17:31-108 (1994)); many of the sequences have also been deposited in Genbank.

One method of reducing the functional level of receptor is, as described herein, the insertion of a foreign nucleotide sequence into an intron of an NMDA receptor gene, e.g., the Nr7 gene; this results in a decreased level of ΝR1 mRΝA and concomitant decrease in functional ΝMDA receptor protein. Without wishing to be bound by theory, it is likely that alterations in non-coding regions of the gene result in decreased stability of the primary transcript.

As used herein, a "foreign" nucleotide sequence is intended to mean any nucleotide sequence which does not normally (in nature) occur at the insertion location. For example, foreign nucleotide sequences can include, but are not limited to, selectable markers (positive and negative selectable markers) such as antibiotic resistance genes (e.g., Hprt, neo, hygromycin (Hyg), blasticidin (Bsr), puromycin resistance gene) and the heφes simplex virus thymidine kinase gene (HSV-TK). Typically the positive selectable marker enables the scoring of recombination and confers a phenotype not normally exhibited by the wild type animal. The construct can also comprise one or more negative selection markers to facilitate identification of proper homologous recombination; for example, the marker can confer sensitivity to a substance not normally toxic to the cell in a wild type animal. In one embodiment the targeting construct contains a neo positive selection marker (encoding resistance to the neomycin analog G418) and a negative selection marker which is the heφes simplex virus thymidine kinase gene (HSV-TK) (encoding susceptibility to ganciclovir). Upon successful gene targeting and homologous recombination, the positive selection marker is incoφorated into the genome, making the cells resistant to G418, while the negative selection marker is excluded, maintaining the cells' resistance to ganciclovir. To enrich for homologous recombinants, gene targeted cells are grown in culture medium containing G418 and ganciclovir. However, the invention is not intended to be limited to the use of positive and/or negative selection criteria; other methods for identifying homologous recombination events are known in the art.

The site of alteration, e.g., insertion, will typically be a non-coding region of the gene (Kingsmore et al, Nature Genet .7:136 (1994); McDevitt et al, Proc. Natl Acad. Sci. USA 94:6781 (1997); Meyers et al, Nature Genet 75:136 (1998); Moens et al, Genes Dev 6:691 (1992); van Deursen et al, Proc. Natl. Acad Sci USA 97:9091 (1994); Wassarman et al, Development 124:2923 (1997); Wilson et al, J Immunol 757:1571 (1993)). For example, the insertion site can be an intron, such as intron 10 or intron 20 of the Nr7 gene. Alternatively or additionally, the insertion can be into the promoter or 3' untranslated region of the gene to achieve altered levels of mRΝA. Insertion of an appropriate foreign nucleotide sequence can be carried out using a variety of gene targeting constructs known in the art. One example of an appropriate targeting construct is shown in Figure 1A. An alternative method of achieving reduced functional levels of ΝMDA receptors is the deletion or substitution of one or more nucleotides in the nucleotide sequence encoding one or more of the ΝMDA receptor subunits. In one embodiment, one or more complete introns can be deleted in their entirety. Again, the deletion or substitution will usually take place in a non-coding region of the gene (e.g., an intron or the promoter) so as to alter the level of receptor produced but not alter the structure or function of the receptor itself. For example, an ΝMDA promoter (e.g., the Nr7 or Nr2B promoter (Sasner and Buonanno, J Biol Chem 277:21316 (1996); Bai and Kusiak, J Biol Chem 272:5936 (1997); Bai et al, J Biol Chem 273:1086 (1998); Klein et al, Gene 208:259 (1998)) can be replaced with a heterologous promoter (i.e., a promoter from another gene) to reduce the functional level of receptor expressed by the cells. Suitable promoters can include chick β- actin promoter, human cytomegalovirus promoter, and mouse phosphoglycerylkinase promoter; other suitable promoters can be determined using methods known in the art. Alterations in the promoter of a gene encoding an ΝMDA receptor subunit (e.g., the Nr7 or Nr2B promoter) can alter the level of expression of the gene without affecting the tissue distribution of the gene. -12-

Alternatively, the nucleotide sequence encoding an NMDA receptor gene can be truncated at the 3' or 5' untranslated regions with the goal of destabilizing the primary transcript, altering RNA processing and transport from the nucleus or producing inefficient splicing. Alterations (e.g., point mutations, substitutions, insertions or deletions) in the splice acceptor or splice donor sites are also suitable methods for altering the level of functional receptor produced.

Alterations in the coding sequence of an NMDA receptor gene that alter the primary sequence can also result in lower NMDA receptor function. For example, a mutation can alter the rate at which the molecule is transported to the cell surface through the endoplasmic reticulum and golgi apparatus. Preferred alterations in the coding sequence do not affect receptor binding or signal transduction. Additionally, although work described in the examples below was carried out by homologous recombination in embryonic stem cells, the invention is not limited to this particular method of genetic modification. For example, embryonic stem cells or embryos can be exposed to viruses and other sequences which integrate into the mammalian genome according to methods known in the art. These cells or embryos can then be screened to identify cells or embryos in which the insertion is at or flanking the NMDA locus using art-recognized methods.

Although work described herein was carried out utilizing the Nr7 gene of the ΝMDA receptor family (as it is the subunit which is common to all of the ΝMDA receptors), this embodiment should not be construed to limit the invention to alterations in the Nr7 gene. Additional glutamate receptors are also suitable for use in the invention. Furthermore, alteration of the Nr2 genes of the ΝMDA receptor family (e.g., Nr2A, e\, Nr2B, e2, Nr2C, e3, Nr2D, e4) to produce a decrease in functional receptor is also the subject of this invention. Alteration of a particular ΝR2 subunit would be expected to affect only NMDA receptors which are composed of that particular subunit and not NMDA receptors which are comprised of a different NR2 subunit. Thus, selection of an NR2 subunit for alteration provides the ability to target the effects of the alteration to a particular subclass of NMDA receptor.

Alternatively, a particular population of neurons can also be targeted by altering expression of all NMDA receptors and then returning expression to normal in some tissues using a ere lox or similar approach. For example, lox sites (or other sequences recognized by a particular recombinase) can be located around a foreign nucleotide sequence that has been inserted into e.g., an intron. Animals can be mated to generate animals that, in addition to being homozygous for the mutant NRl gene, carry a transgene encoding the recombinase. The recombinase expression can be regulated by a promoter capable of driving expression of the recombinase in neurons in which it is desirable to re-establish normal levels of expression of the NMDA receptor (e.g., the brain stem or hippocampus). The resultant animal has reduced NMDA expression only in neurons in which the recombinase is not expressed.

Alternatively, NMDA receptor gene expression can be eliminated in a subset of neurons. This can be achieved by placing the sequences recognized by a recombinase flanking exons essential for the expression of the gene, with expression of the recombinase being driven by a promoter that drives expression in those neurons in which loss of normal NMDA receptor expression results in the schizophrenic phenotype.

Although work herein is exemplified specifically with respect to use of a particular gene targeting construct to direct the altered sequence into an embryonic stem cell, other targeting vectors and strategies will be readily apparent to the skilled artisan. The invention is not intended to be limited to this particular method of reducing the functional level of receptor. Any non-pharmacological method which reduces the functional level of NMDA receptors is suitable to create the animal model of this invention.

For example, an animal with a null mutation in the NRl or NR2 gene can be bred to an animal carrying a NRl or NR2 transgene, respectively, driven by any of a variety of suitable promoters. The promoter driving expression of the transgene would allow restoration of expression in those regions of the brain, or to appropriate levels, required for survival. By selection of the promoter and expression pattern, particular animals can be selected and assessed for the schizophrenic phenotype. For example, transgene expression can be selected such that the transgene promoter drives expression in the brain stem, where complete loss of expression is believed to lead to perinatal lethality in the NMDA null animals. A dominant-negative, antisense or ribozyme strategy can also be used to produce animal models having a reduction in functional NMDA receptors. For example, a genetically modified animal can be made in which an antisense RNA is expressed under the control of a neuron specific promoter. This antisense mRNA will bind to the normal NMDA transcript, forming double stranded DNA, thereby preventing expression of at least some of the NMDA transcripts.

As used herein, a reduction in functional NMDA receptor level is intended to include any inhibition in, reduction or decrease of functional receptor level. Reduction in the functional level of receptor is intended to include both a decrease in the overall amount of receptor protein expressed, as well as a decrease in function, i.e., reduced efficacy, of the receptor protein itself. However, in a preferred embodiment functional receptor is not completely absent; that is, some level of functional receptor is still present to facilitate assessment of therapeutic agents which target the receptor. It is also preferred that the expression level is sufficient to provide the critical function of NMDA receptors postnatally, allowing the animal to survive to adulthood. For example, preferred animals express less than about 50% of normal functional receptor level, particularly preferred animals express less than about 25% of normal functional receptor level, and most particularly preferred animals express less than about 10% of normal functional receptor level. Methods for measurement of gene expression and protein expressed by a gene are varied and well-known, including protein antibody measurements of tissue samples, screening with nucleic acid probes, northern blot analysis, Rnase protection assay, and radioligand binding.

Animals useful in the present invention are typically those which are easily bred, raised and readily manipulated genetically. Additionally, the animal should preferably naturally express the NMDA receptor(s). For example, suitable animals include the mouse, rat, gerbil, rabbit, hamster and guinea pig, as well as non-human primates.

Genetically modified animals comprising the cells described herein can be generated by standard methods in which a genetic construct is introduced by pronuclear injection into a fertilized egg. Additionally, the introduction of gene sequences can be achieved by microinjection of embryos to inject genetically modified embryonic stem cells into blastocysts. The procedures for manipulation of the rodent embryo and for microinjection of DNA are described in detail in the art, for example in Hogan et al., Manipulating the Mouse Embryo, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1986)). Embryonic stem cells can also be manipulated using the method of Gossler et al, Proc. Nat Acad. Sci. USA 53:9065- 9069 (1986).

Generally, the alteration of the selected nucleic acid sequence is carried out in embryonic stem cells using homologous recombination (see, for example, Capecchi, Trends Genet 5:70-76 (1989); Koller and Smithies, Ann Rev Immunol 70:705-730 (1992); U.S. Patent 5,614,396 to Bradley et al.). The embryonic stem cells containing the altered sequence are then introduced into a developing embryo or blastocyst and thereby incoφorated into an animal. Breeding the resulting mice and their progeny result in some animals which are heterozygous for the altered sequence and some mice which are homozygous for the altered sequence. However, other methods of mutagenesis can also be used, such as other methods of gene targeting, chemical mutagenesis (Rinchik, Trends Genet 7:15-21 (1991)) and random integration of vectors such as retroviruses (Robertson et al, Ciba Found Symp 765:237-250 (1992)), and these methods can be carried out at the level of the embryo and whole animal. For example, early stage embryos (preimplantation) can be exposed to retrovirus and then transferred into a foster mother. When born, the location of the inserted retrovirus into the genome of the embryo can be determined. One advantage of carrying out these procedures with ES cells is that you can determine the mutation made before continuing to produce the animal.

It is intended that the animal model described herein can be used as a tool for screening and assessing compounds for potential diagnostic or therapeutic efficacy for neurological disorders. In addition to schizophrenia, suitable neurological disorders include disorders associated with learning, memory, excitatory toxicity, cognition, dementia and addictive behavior. The animal model described herein can also be used to elucidate the mechanisms by which NMDA receptor activity modulates the activity of other neurotransmitter systems.

In addition, genetically modified animals described herein expressing functionally reduced levels of NMDA receptors can be used as a source of neurons which express reduced levels of NMDA receptors and which can be cultured in vitro and, optionally, immortalized. The invention also relates to an embryonic stem cell line genetically modified to express a reduced level of functional NMDA receptor. In a particular embodiment, both copies of the gene for the NMDA receptor are modified as described herein to express a reduced level of functional NMDA receptor. Embryonic stem cells described herein can also be induced to differentiate into neurons, e.g., with retinoic acid, and the invention pertains to these differentiated cells having or expressing a reduced level of functional NMDA receptor. These cells can be used as an in vitro system for assessing the effects of various compounds on NMDA receptor activity.

For example, compounds can be tested for their ability to act as agonists of NMDA receptor activity. Compounds that are initially identified as agonists-of NMDA receptor function in vitro can then be further assessed and studied for their in vivo efficacy, e.g., as NMDA receptor agonists and/or anti-psychotics, in a genetically modified animal as described herein.

The neurons obtained from genetically modified animals described herein, or from embryonic stem cells described herein, as well as the genetically modified animals described herein, can be used for the identification of other proteins whose expression is altered in the disease model as a result of alterations in the NMDA receptor levels. These proteins can in turn represent new targets for intervention in the neurological disorders. Such technologies are known in the art and include transcriptional profiling, differential display and DNA chip technology.

A wide variety of compounds can be assessed using the methods described herein. For example, the compound to be tested can be a known agonist of the NMDA receptor, including, but not limited to, glutamate, glycine, and polyamines (e.g., diamines, triamines, and tetraamine) such as spermine and spermidine. Additionally, libraries of chemical compounds can be assessed using the methods described herein. For example, compounds can be assessed for their ability in vivo to correct, partially or fully, the hyperlocomotion, altered social behavior, altered mating and/or altered nurturing exhibited by the genetically modified animal models. Alternatively, compounds can be assessed for their ability to normalize (i.e., bring closer to a control) expression of proteins identified as being altered by alteration in functional NMDA receptor levels.

Administration of the compound to be tested can be carried out by any appropriate means. For example, administration to the genetically modified animals described herein can be carried out intracerebrally, intravascularly, parenterally, subcutaneously, intramuscularly, intraperitoneally, orally, interstitially, hyperbarically, intraocularly, sublingually, intravenously and the like. Other suitable methods of introduction can also include gene therapy, rechargeable or biodegradable devices and slow release polymeric devices. The pharmaceutical compositions of this invention can also be administered as part of a combinatorial therapy with other agents. For example, NMDA receptor agonists or a combination of dopamine/serotonin antagonists and NMDA receptor agonists can be assessed as described herein. In the case of in vitro genetically modified cells, administration of the compound can be carried out by contacting the cells with the compound, e.g., by placing it in the culture medium.

The present invention also pertains to pharmaceutical compositions comprising compounds described herein. For instance, a compound of the present invention can be formulated with a physiologically acceptable medium to prepare a pharmaceutical composition. The particular physiological medium may include, but is not limited to, water, buffered saline, polyols (e.g., glycerol, propylene glycol, liquid polyethylene glycol) and dextrose solutions. The optimum concentration of the active ingredient(s) in the chosen medium can be determined empirically, according to well known procedures, and will depend on the ultimate pharmaceutical formulation desired. The generation of mice expressing reduced levels of NMDA receptors by mutation of Nrl provides the opportunity to study the biology of NMDA receptors in behaving, adult mice. Genetic suppression of Nrl mimics the use of PCP and MK-801 as determined by analysis of habituation motor behavior. These drugs fail to elicit further motor stimulatory effects in 7Vr7"^eo -/- mutant animals, providing strong evidence that PCP and MK-801 evoke their motor stimulatory effects through interaction with NMDA receptors. The Nr7 mutant animals demonstrate that motor stimulatory effects are seen without the use of PCP, and result from down-regulation of NMDA receptor-mediated neurotransmission.

Hyperlocomotion is the result of increased dopaminergic activity in the striatum. The hyperlocomotion oϊNrl"^eo -I- animals provides genetic evidence that suppression of NMDA-mediated neurotransmission augments dopaminergic neurotransmission. The action of NMDA receptor antagonists on the dopaminergic system has been a point of debate, however several studies have demonstrated the direct or indirect action of NMDA receptor antagonists in the upregulation of dopaminergic systems (Hertel et al, Behav Brain Res, 72: 103-114 (1995), Yan et al, Brain Res, 765:149-158 (1997); Lapin and Rogawski, Behav Brain Res, 70: 145- 151 (1995); Nabeshima et al, Ann N Y Acad Sci, 507:29-38 (1996); Steinpress, Behav Brain Res, 74:45-55 (1996). The action of haloperidol on the behavior of Nrl^ne°-I- animals also demonstrates the requirement for an intact dopaminergic system in the expression of these behaviors. Moreover, animal models described herein can assist in the diagnosis of schizophrenia. The described animal models can be studied and compared with normal animals to identify differences, e.g., a change in levels of a particular blood protein, which are associated with schizophrenia. The identification of such differences alone or in combination can then be used to diagnose or assist in the diagnosis of schizophrenia in an individual. For example, a blood sample can be taken from an individual, assessed for the presence, absence or amount of a particular blood protein associated with schizophrenia, and compared with levels in a normal individual. Such diagnostic tests can be combined with psychological and/or behavioral tests to determine the proper diagnosis. The present invention will now be illustrated by the following Examples, which are not intended to be limiting in any way. The teachings of all references cited herein are incoφorated herein by reference in their entirety. EXAMPLES Materials and Methods Generation of the Nrl mutation

Three overlapping genomic clones spanning the Nrl locus were isolated from a 129/SvEv lambda bacteriophage library (Stratagene, La Jolla) using Nr7 cDΝA exons 11-20 as a probe. These genomic clones were subcloned into pBluescriptIISK+ (Stratagene, La Jolla) and mapped for restriction endonuclease sites. Restriction fragments from these clones were used for the construction of the targeting construct, Νrl^neo, which, upon integration into the locus by homologous recombination, results in the insertion of a neomycin resistance gene into intron 20 of the Nr7 gene. An 8.0 kb DΝA fragment extending from the Eco RI site of intron 10 to the Sma I site of intron 20 was inserted 5' of the neo gene in the Not I site of the targeting vector JNS2 (Dombrowitz et al, J Immunol 757:1645 (1996)). A 2.5 kb Smal-BamHI fragment corresponding to intron 20-intron 21 was cloned into the Xbal and BamHI sites of JNS2 with a Nhe I linker (New England Biolabs, Beverly, MA) to complete the targeting construct Nrl^neo.

The targeting construct was linearized by Pvul digestion and electroporated into the embryonic stem cell line E14Tg2a (Hooper et al, Nature, 326:292-295 (1987)) as described previously (Mohn and Koller, DNA Cloning 4: Mammalian Systems, D.M. Glover and B.D. Hames, eds. (New York: Oxford University Press), pp. 143-184 (1995). Colonies resistant to G418 and gancyclovir were expanded and genotyped by Southern analysis. ES cell lines 364-1 1, 364-71, and 364-77 were karyotyped and micro injected into C57BL/6J blastocysts to generate germline- transmitting chimeras. Chimeras were mated to B6D2 animals to generate animals heterozygous for the mutant allele.

Genotype analysis of ES cell lines and animals

Genomic DNA was prepared from ES cell pellets or tail biopsy as previously described (Mohn et al, Mol Cell Neurosci, 9:63-76 (1997)). Hindlll restriction digests of approximately 7 ug of genomic DNA were electrophoresed and transferred to Hybond nylon membrane (Amersham, Arlington Heights, IL) by Southern blot. Membranes were hybridized in Rapid Hybe hybridization solution (Amersham) with a 2.0 kb genomic probe which includes exon 22 of Nrl. Membranes were subsequently washed at 59°C with 2X SSC, 0.1% SDS and with 0.2X SSC, 0.1% SDS.

Sequence analysis of Nrl cDNA generated from Nr "^eo homozygotes Total RΝA from whole brain homogenates of Nrl"^eo homozygotes and their wild type littermates was isolated with RΝAzolB (Tel-Test, Friendswood, TX). 3 μg of total RΝA was used as a template for reverse transcription using the cDΝA cycle kit (Invitrogen, San Diego). The cDΝA generated was used as a template for PCR amplification of Nrl exons 10-22. Three primer sets were used to amplify overlapping regions of /Vr7 cDΝA; they were: 5'ACAGAGAAGCCTCGAGGATA (SEQ ID ΝO:l) with 5'AGGAAAACCACATGGCAGAG (SEQ ID NO: 2), 5'TTCAGTCCCTTTGGCCGATT (SEQ ID NO: 3) with 5'GCGGGAGTCACATTCTTGAT (SEQ ID NO: 4), and 5'GGGTACTCTTACCGAAGTAC (SEQ ID NO: 5) with 5'AGAAATACACAGACAAGGCG (SEQ ID NO: 6). Each PCR product was approximately 600 bp, and was gel purified and cloned into pCR2.1 with the TA cloning kit (Invitrogen). For each PCR product, two independent clones were sequenced by Taq cycle sequencing using an Applied BioSystems 373 A Automated Sequencer (UNC-CH Automated Sequencing Facility). This was done to ensure that basepair mutations had not been introduced in the Nrl"^e° allele.

Northern analysis

Northern blot of 20 ug of total RNA was performed by the method of Kroczek and Siebert, Anal Biochem, 754:90-95 (1990) using Immobilon-NC membranes as the transfer support membrane (Millipore, Bedford, MA). The northern blot was hybridized with a 1.35 kb cDNA probe specific for Nrl exons 11- 20 at 68 °C using QuickHybe hybridization solution (Stratagene, La Jolla). Following hybridization the blot was washed at 65 °C once in 2X SSC, 0.1% SDS and then in 0.2X SSC, 0.1% SDS. Northern blots were subsequently stripped and hybridized with a 1.3 kb mouse β-actin probe (Stratagene) using the same hybridization conditions to ensure equal loading of RNA for the three genotypes. NIH Image 1.62 Beta was used to digitally capture autoradiograpic images from the northern blots. Densitometry was performed with NIH Image 1.62 Beta by measuring the area of the plot of each band.

Examination of motor behaviors Eight Nrl"^eo-/- mice and eight of their wild type littermates were tested at 8-

10 weeks of age for their habituation to novel environments by measuring spontaneous open field locomotion and rearing behaviors. Motor activity studies were performed with an Omnitech digiscan activity monitor, which measures several parameters of vertical and horizontal motion within a 42 cm² chamber. Horizontal distance and stereotyped motions were recorded in 5 minute intervals over a 4 hour period, and the raw numbers were averaged to give values for 20 minute intervals. Activity tests were performed between the hours of 10:00 am and 6:00 pm.

Behavioral responses to pharmacological agents

Inteφeritoneal injections of MK-801 (0.2 mg/kg, Research Biochemicals International, Natick, MA), PCP (3 mg/kg, Research Biochemicals International), haloperidol ( 0.5 mg/kg, Research Biochemicals International), or clozapine (0.5 mg/kg, Research Biochemicals International) were administered to mice immediately prior to assessment of motor behaviors. Motor activity was recorded with the Omintech digital activity monitor as described, except that observation was limited to 2 hours. Activity tests were performed with 8 Nrl^ne0-I- mice and 8 wild type littermates aged 8-10 weeks old, and were performed between the hours of 10:00 am and 6:00 pm.

Results

Generation of an Nrl mutant mouse strain Homologous recombination replaces the genomic sequence of Nrl with a targeting construct in the embryonic stem cell line E14TG2a (Fig. 1A). The incoφoration of the targeting construct into the genome does not delete Nr7 sequences, but introduces a selectable marker gene (neo) within intron 20. The coding sequence of Nrl remains intact as determined by sequence analysis of cDΝA from exons 10-22. Southern analysis determines the genotype of drug-resistant ES cell lines as well as the animals that are generated by ES cell transgenesis (Fig. IB). Intercross of animals heterozygous for the targeted allele (Nrl"^e° +/-) results in a ratio of 1 (Nr7"^eo +/+) : 2 (Nr7"^eo +/-) : 1 (Nr7"^eo -/-). Northern analysis demonstrates that Nrl^ne° is a hypomoφhic allele; Nrl^neo -I- animals express at least 10-fold less full-length message than wild type controls (Fig. 1C). Repeated densitometry analysis, normalized with actin controls, estimates Nr7 message levels of mutant animals at 9-13% of wild type. This drastic reduction in Nr7 message is most likely due to the insertion of the 1.8 kb neo gene; the insertion may prevent splicing of intron 20 and result in message degradation. Instances of hypomoφhic mutation by intronic insertion have been reported in naturally occurring mouse mutants and as a result of gene targeting (Seperack et al, 1995; Kingsmore et al, Nat Genet, 7:136-141 (1994); Moens et al, Genes Dev, 6:691-704 (1992); van Duersen et al, Proc Natl Acad Sci USA, 97:9091-9095 (1994); Wilson et al, J Immunol, 757:1571-1578 (1993); McDevitt et al, Proc Natl Acad Sci USA, 94:6781-6785 (1997); Wassarman et al, Development, 124:2923- 2934 (1997); Meyers et al, Nature Genet, 75:136-141 (1998). A common feature of these reported intronic insertion events is a decrease in functional message levels, as we have found in the hypomoφhic allele Nrl"^e0.

Nrl mutant mice weigh significantly less than littermates as juveniles and adults.

Nrl^ne° -I- mice can be identified from their wild type and heterozygote littermates by their reduced size. Diminished body weight is first apparent at postnatal day 10 (P10). The runted appearance of mutants is exacerbated in large litters (10+ pups), and occasionally leads to death from malnutrition in these litters. The large majority of mutants survive into adulthood; however, they remain significantly smaller than their wild type counteφarts. Heterozygotes are not significantly smaller than wild type controls.

Reductions in body weight persist through adulthood, and mutant mice maintain body weights that are 25-35% of their wild type counteφarts. Mutant mice have difficulty obtaining chow pellets from the wire racks above the cage; these mice will eat ravenously if chow is placed in the bedding of the cage. Placing chow in the bedding improves the luster of the mutant's coat, and they appear healthier, but diminished body weight persists.

Crosses between Nrl"^eo -I- mice and wild type animals or between Nrl^ne0 -I- mice and heterozygotes have not resulted in viable litters. Nrl""⁰ -I- males are infertile, and while a few Nrl"^eo -I- females have become pregnant, their litters are small and delivered pups die perinatally. It seems unlikely that the reduced fertility or infertility is due to chronic malnutrition, since mutant animals can weigh as much as 30 grams and develop fat pads around reproductive organs and along the back.

Nrl mutant mice display behavioral phenotypes similar to high doses of non- competitive NMDA receptor antagonists.

Pharmacological suppression of the NMDA receptors leads to varied behavioral effects in rodents, depending on the dose and nature of the receptor antagonist. Moderate doses of competitive antagonists such as AP5 or CPP can cause animals to display ataxia, muscle weakness, and splayed hind limbs (Loscher and Schmidt, Epilepsy Res, 2:145-181 (1988). Higher doses cause sedation. Moderate doses of noncompetitive antagonists such as phencyclidine(PCP) and dizocilpine (MK-801) have the opposite effect of increasing motor activity and stereotypic movements such as grooming, sniffing, and head nodding (Tricklebank et al, Eur J Pharmacol, 767:127-135 (1989)). Nrl"^eo -I- mice do display some behaviors that may be related to application of competitive antagonists. Mutant animals tend to splay their hind limbs when held suspended by the tail. They also have difficulty grasping a wire rack, or obtaining food stored above them; these observations may indicate muscle weakness similar to that seen after application of competitive NMDA receptor antagonists.

In general however, Nrl"^eo -I- mice display behavioral abnormalities that are more consistent with the actions of non-competitive rather than competitive antagonists. Mutant mice display significant increases in motor activity and stereotyped motion when introduced to novel environments (Figs. 2A-B). Horizontal activity, described as the total distance traveled by mice within a four hour time period, is increased 3.5 times above wild type controls (P=0.000095, two- tailed T-test). Stereotypic movements, measured as the total number of stereotypy counts within the same time period, are double that of wild type controls (P=0.005, two-tailed T-test). Increased locomotion and stereotypy are obvious during the process of habituation to novel environments; mutant mice habituate more slowly to new cages, but after 4 hours their motor activity is similar to wild type.

Increases in locomotor behavior and stereotypy are quantitatively similar to acute administration of the non-competitive NMDA receptor antagonists MK-801 and PCP. Wild type mice given a 0.2 mg/kg dose of MK-801 have 4-fold increases in horizontal activity (Figs. 2E-F). Stereotypic movements increase 2-3 fold with the application of MK-801. PCP administration results in a different kinetic profile, but motor activity is similarly increased (Figs. 2C-D). When measured quantitatively, the motor behaviors of wild type mice treated with MK-801 or PCP are not significantly different from untreated Nr7"^e0 -/- animals.

These results can be considered in the context of a potential mouse model of schizophrenia. Current mouse models of schizophrenia are drug-induced psychosis elicited by PCP or amphetamine (Alexander, et al, Neuropsychopharmacology, 75:484-490 (1996); Ellison, Brain Res Rev, 79:223-239 (1994); Corbett et al, Psychopharmacology, 120:61-14 (1995); Iversen, J Psychopharmacol, 7:154-176 (1987). These models demonstrate hyperlocomotion similar to that seen following PCP administration (as we have demonstrated) and also to amphetamine administration (Alexander, et al, Neuropsychopharmacology, 75:484-490 (1996)).

Nrl mutant mice are unaffected by ΝMDA receptor antagonists.

PCP and MK-801 fail to elicit motor stimulatory responses in Nr7^neo -/- mice (Figs. 3A-D). Profiles of locomotor activity are remarkably similar between drug- treated and untreated mutant animals. Stereotypic movements are actually decreased with the administration of both MK-801 and PCP, although the differences are not highly significant (P=0.043). This provides genetic evidence that PCP and MK-801 are specific antagonists of the ΝMDA receptor . These results also demonstrate that the motor stimulatory effects generated by PCP and MK-801 are produced solely through the blockade of ΝMDA receptors. Although MK-801 has been classified as a specific antagonist of the ΝMDA receptor complex, PCP has been shown to interact with other neurotransmitter systems, including cholinergic receptors, potassium channels, sigma binding sites, and biogenic amine reuptake systems (DAT) ( Rothman, Neurotoxicol Teratol, 76:343-353 (1994)). Genetic blockade of DAT through the creation of a DAT null mutation in mice results in significant increases in motor activity (Giros et al, Nature, 379:606-612 (1996)). If PCP mediates part of its motor stimulatory effect through the pharmacological blockade of DAT, it would be expected that PCP would provoke a motor stimulant effect above baseline levels seen in untreated Nrl"^eo -I- animals. However, since this is not the case; our results suggest that the motor stimulatory effects of PCP are primarily due to its action at the NMDA receptor complex.

Haloperidol and clozapine ameliorate behavioral effects of genetic Nrl suppression

Haloperidol is a potent neuroleptic agent that is used to treat schizophrenia and is thought to mediate its antipsychotic effects by antagonism of the D₂ dopamine receptor (Carlsson, J Psychiatr Res, 77:57-64 (1974); Kebabian and Calne, Nature, 277:93-96 (1979)). Its action on the nigrostriatal dopaminergic system leads to suppression of motor activity, while suppression of the mesolimbic dopaminergic system results in amelioration of the "positive symptoms" of schizophrenia (formal thought disorder, delusions, hallucinations). Haloperidol is effective in antagonizing the motor behaviors characteristic of apomoφhine, amphetamine or phencyclidine intoxication in animal models of schizophrenia (Freed et al, Psychopharmacology, 77:291-297 (1980); Iversen, J Psychopharmacol, 7:154-176 (1987)). The attenuation of motor behaviors produced by PCP and MK-801 has been proposed as an indicator of antipsychotic action for new pharmaceuticals (Carlsson and Carlsson, Trends Neural Sci, 73:272-276 (1990); Corbett et al, Psychopharmacology, 120:61- 74 (1995)).

Since genetic suppression of NMDA receptors results in an altered motor behavior similar to PCP or MK-801 administration, the ability to attenuate these behaviors by haloperidol administration was assessed. Attenuation of hyperlocomotion would further support the notion that Nrl mutant mice represent a potential model for schizophrenia. Indeed, haloperidol administration results in significant suppression of the locomotor and stereotyped behaviors of Nmdarl"^eo homozygotes (Figs. 4A-B). Quantitative analysis of motor behavior demonstrates that haloperidol normalizes Nrl^neo -I- animals to wild type levels of activity.

Clozapine and other atypical antipsychotics have been demonstrated to suppress psychotic symptoms without significant extra-pyramidal side effects (Gerlach, Schizophr Bull, 77:289 (1991); Ereshefsky et al, Clin Pharm, 5:691 (1989); Van Tol et al, Nature, 350:610 (1991); Gingrich and Caron, Annu Rev Neurosci, 16:299 (1993)). Clozapine's pharmacological profile includes D2/D4 dopamine receptor antagonism and 5HTA2 receptor antagonism, but also interacts with a-adrenergic receptors and HI histamine receptors. Although clozapine does not directly act on glutamate receptors, it has been suggested that part of clozapine's therapeutic effect is due to the augmentation of glutamatergic neurotransmission (Goff and Wine, Schizophrenia Res, 27:157-168 (1997); Daly and Moghaddam, Neurosci Lett, 752:61 (1993)). Even at the relatively low dose of 0.5 mg/kg, clozapine is effective at attenuating the abnormal motor behavior of Nrl^ne° -I- mice; at this dose clozapine does not affect the motor behaviors of wild type littermates (Figs. 4C-D). A comparison can be drawn between clozapine and haloperidol in the treatment of schizophrenia in humans and in the efficacy of these two drugs in the attenuation of hyperlocomotion and stereotypy of Nrl^neo -I- mice. In humans, clozapine differs from haloperidol in that it has fewer EPS and is more effective in the treatment of negative symptoms of schizophrenia. In Nrl^neo -/- mice, clozapine is more effective than haloperidol in suppressing abnormal behaviors with a dose that lacks effects on control mice.

Negative symptoms of schizophrenia An additional measure of the Nrl"^e0 -I- mice as a model of schizophrenia is the demonstration of "negative symptoms of schizophrenia", specifically social withdrawal. Patients with schizophrenia exhibit social withdrawal as a persistent symptom of the disease. This behavior is also observed in the phencyclidine- induced rodent models of schizophrenia. Social withdrawal was apparent in Nrl"^e0- /- mice that were housed with wild type littermates; the mutant mice often do not sleep in a nest with cagemates, and preferred to be physically distant. The extent of social withdrawal was quantified with a resident-intruder behavioral assay (Fig. 5). In this assay, a resident male mouse is housed alone for at least a week and then a group-housed intruder male is added to the resident cage. Typically the resident male actively initiates social investigation of the intruder, rarely avoids social interaction, and may initiate fights. The amount of time that wild type or Nrl^ne0 -I- residents spent engaging in these behaviors was monitored over a 6-minute period (Fig. 5). Nr7"^eo -/- mice displayed significant reductions in their amount of social investigation when compared to wild type mice. Furthermore, Nrl"⁶⁰ -I- mice exhibited "escape behaviors", where they actively avoided interaction with the intruder male. In contrast, escape behavior was only rarely displayed by wild type residents.

The administration of clozapine to Nrl"^e° -I- mice one hour prior to testing resulted in increases in social investigation and decreases in escape behavior (Fig. 5). Thus clozapine is effective in normalizing not only the hyperlocomotion of these animals (related to positive symptoms), but also their social withdrawal (related to negative symptoms).

The social withdrawal apparent in Nrl^neo -I- mice extended to their mating behavior. Nrl^neo -I- males were infertile due to a failure to mate with ovulating females. Hypothesizing that this failure to mate was due to a behavioral defect, not a physical one, clozapine (0.5 mg/kg) was administered to Nrl"^eo -I- males prior to mating. Clozapine was, in fact, effective in restoring or improving the fertility of Mrl^neo -I- males (Table 1). Following clozapine administration the mutant males mated with ovulating females approximately 30% of the time, as evidenced by the production of copulation plugs and fertilized embryos in the females. The infertility of Nrl"^eo -I- males may related directly to the human condition of schizophrenia. Schizophrenic males have a significant reduction in fertility. This observation has been attributed to reduced marriage rates and reduced numbers of conjugal relationships. The explanation for reduced fertility in schizophrenic males is a "reduced ability to obtain sexual partners", and is in a sense a behavioral, and not a physical type of infertility. These results with the Nr7"^eo -/- mice suggest that the infertility of these mice is also behavioral and not physical, and can be improved with drugs that improve social withdrawal symptoms. Table 1. Clozapine improves the fertility of Nrl"^e0 -I- males.

Mating frequency determined as the number of times that copulation plugs were identified from super-ovulated females mated with the individually-housed males. Clozapine was administered at a dose of 0.5% mg/kg inteφeritoneally to males just prior to the addition of the female.

While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

CLAIMS What is claimed is:

1. A genetically modified non-human animal expressing a reduced level of NMDA receptor, wherein said reduced level is not pharmacologically induced.

2. A genetically modified non-human animal according to Claim 1 which is a mouse.

3. A genetically modified non-human animal according to Claim 1, wherein the genetically modified non-human animal exhibits behavior similar to the behavior of a wild type animal treated with an NMDA receptor antagonist.

4. A genetically modified non-human animal according to Claim 1, wherein the animal is a model of a neurological disorder.

5. A genetically modified non-human animal according to Claim 4, wherein the neurological disorder is schizophrenia.

6. A neuron cultured in vitro from a genetically modified animal according to

Claim 1.

7. A neuron obtained from a genetically modified embryonic stem cell expressing a reduced level of NMDA receptor, wherein said reduced level is not pharmacologically induced.

8. A method of making a genetically modified non-human animal expressing a reduced level of NMDA receptor, wherein said reduced level is not pharmacologically induced, comprising the steps of: a) introducing into a living non-human embryo a nucleic acid molecule encoding all or part of an NMDA receptor subunit, wherein said nucleic acid molecule has been altered to result in expression of a reduced level of functional NMDA receptor or subunit thereof, thereby producing an animal; and b) breeding the animal to generate animals heterozygous for said nucleic acid molecule.

9. A method according to Claim 8, further comprising breeding said animals heterozygous for said nucleic acid molecule to produce a genetically modified non-human animal which is homozygous for said nucleic acid molecule.

10. A method according to Claim 8, wherein the nucleic acid molecule is contained in an embryonic stem cell and wherein the embryo is a blastocyst.

11. A method according to Claim 8, wherein the alteration comprises insertion of one or more foreign nucleotide sequences in one or more non-coding regions of said nucleic acid molecule.

12. A method according to Claim 11, wherein the foreign nucleotide sequence is selected from the group consisting of the genes encoding neo, Hprt, hygromycin, blasticidin, puromycin resistance, zeocin, and thymidine kinase.

13. A method according to Claim 11, wherein the non-coding region is selected from one or more introns of said nucleic acid molecule, the promoter of said nucleic acid molecule, the 3' or 5' untranslated region of said nucleic acid molecule, and combinations thereof.

14. A method according to Claim 8, wherein said nucleic acid molecule encodes the NMDA NRl receptor subunit.

15. A method according to Claim 8, wherein the genetically modified non- human animal exhibits behavior similar to the behavior of a wild type animal treated with an NMDA receptor antagonist.

16. A method of assessing a compound for the ability to normalize the behavior of a genetically modified non-human animal according to Claim 3, comprising administering the compound to be assessed to the genetically modified animal and assessing the behavior of the animal relative to an untreated wild type animal.

17. A method of assessing a compound for the ability to function as an agonist of NMDA receptor function, comprising contacting a neuron according to

Claim 6 with the compound to be assessed and assessing the effect on NMDA receptor function.

18. A method of assessing a compound for the ability to function as an agonist of NMDA receptor function, comprising contacting a neuron according to Claim 7 with the compound to be assessed and assessing the effect on

NMDA receptor function.

19. A method of assessing a compound for the ability to function as an agonist of NMDA receptor function, comprising administering the compound, with an optional physiologically acceptable vehicle, to a genetically modified non- human animal according to Claim 3, and assessing the effect on NMDA receptor function.

20. A method of normalizing the behavior of a schizophrenic individual comprising administering to the individual a compound identified according to the method of any one of Claims 16 to 19.