WO2000017334A2 - Analysis of ligand activated nuclear receptors i(in vivo) - Google Patents

Abstract

The present invention relates to detection of ligands for nuclear receptors in vivo. In particular, the present invention provides transgenic constructs and transgenic animals, as well as assays using the same to detect ligands for nuclear receptors in transgenic animals. The transgenic constructs, transgenic animals and assays of the present invention are useful for identification and isolation of ligands for orphan receptors. In addition, the invention is useful for analyzing pharmacological properties of natural and synthetic ligands for nuclear receptors.

Description

ANALYSIS OF LIGAND ACTIVATED NUCLEAR RECEPTORS XiV VIVO

FIELD OF THE INVENTION The present invention relates to detection of ligands for nuclear receptors in vivo . In particular, the present invention provides transgenic constructs as well as assays using such constructs to detect ligands for nuclear receptors in transgenic mice. The transgenic constructs and assays of the present invention are useful for identification and isolation of ligands for orphan receptors. In addition, the invention is useful for analyzing pharmacological properties of natural and synthetic ligands for nuclear receptors.

BACKGROUND OF THE INVENTION

Nuclear receptors (NR) are intracellular proteins that include receptors for steroid hormones, retinoids, thyroid hormone and vitamin D. Some of the NRs act as ligand-inducible transcription factors, while a large number of them have no defined ligand and are hence described as orphan receptors (En ark et al . , Mol . Endocrinol . 10: 1293-1307, 1996). Virtually all studied physiological processes have been demonstrated to be affected by NRs. Consequently, it is not surprising that NRs, including several of the orphan receptors, have been implicated in pathological conditions such as atherosclerosis, diabetes, obesity, Parkinson's disease and various types of cancer. See, for example, Y. Labelle et al . (1995) Hum . Mol . Genet . 4 : 2219-2226; J. M. Lehmann et al . (1995), J. Biol . Chem . , 270: 12953-6; P. Sarraf et al . (1998) Nature Medicine 4 : 1046-1052; P. Tontonoz et al . (1998) Cell 93 : 241-252; and R.H. Zetterstrδm et al . (1997) Science 276: 248-250. Thus, it is of particular importance to explore the involvement of NRs in the above mentioned disorders and in normal physiology.

It is believed that the ligands of NRs are generated in vivo in at least three different ways: at a source other than the target cell in classic endocrine fashion (e.g., thyroid hormone); within the target cell from an apohormone (e.g., 9-cis retinoic acid); or within the target cell and is not secreted (e.g., prostaglandin) . See review by Mangelsdorf et al . , Cell 83: 841-850, 1995, and Mangelsdorf et al . , Cell 83: 839- 840, 1995. Upon ligand binding, nuclear receptors mediate transcriptional responses by controlling the activity of the target genes .

Despite the recent progress in this field, our knowledge concerning NR ligands and signaling in vivo is still limited. NR ligands are small nonprotein lipophilic molecules that are not themselves encoded in the genome. As a result, although some novel ligands and activators of NRs have been identified, the identification of ligands has remained a great challenge for most of the recently identified orphan receptors. The present invention provides ligand trap assays which detect NR ligand distribution in vivo . Most NRs signal in a para- or autocrine fashion and their ligands are synthesized locally within tissues where NRs are of functional importance. Thus, defining the sites of NR ligand synthesis facilitates the understanding of NR signaling and provides opportunities for identifying ligands for orphan receptors. Additionally, the assays of the present invention can be used in assessing receptor activation in vivo following pharmacological administration of nuclear receptor ligands.

SUMMARY OF THE INVENTION The present invention relates to vectors and methods for detecting ligands for nuclear receptors in vivo .

In one embodiment, the present invention provides a first assay that detects ligands of nuclear receptors in vivo by using a pair of transgenic vectors: an effector vector and a corresponding reporter vector.

The effector vector for use in such first assay of the present invention codes for an effector protein which includes the DNA binding domain (DBD) of a transcription factor and the ligand binding domain (LBD) of a nuclear receptor. The effector-coding sequence is operably linked to a first promoter. The reporter vector codes for a reporter protein. Such reporter-coding sequence is operably linked to a promoter, which promoter is placed after a DNA binding sequence to which the DBD of the effector protein binds. A double transgenic animal can be established by using an effector vector and a corresponding reporter vector of the present invention. The ligands of a nuclear receptor can be detected by determining the expression of the reporter gene in the transgenic animal .

In another embodiment, the present invention provides expression vectors which include coding sequences for both an effector protein and a reporter protein. Such vectors are also referred to herein as

"effector-reporter vectors". The effector protein is a fusion between the DBD of a transcription factor and the LBD of a nuclear receptor. The effector-coding sequence is operably linked to a first promoter which, in turn, is placed downstream of a first DBD-binding sequence. The reporter-coding sequence is linked to a second promoter which, in turn, is placed downstream of a second DBD- binding sequence. The DBD of the effector protein can bind to both the first and the second DBD-binding sequence. Such transgenic vector can include other sequences that may be appropriate, such as a polyadenylation signal and a gene coding for selection marker.

In another embodiment of the present invention, cells transformed with an effector-reporter vector of the present invention are provided. Another embodiment of the present invention provides a transgenic animal containing an effector- reporter vector of the present invention. A preferred transgenic animal is a transgenic mouse.

In still another embodiment, the present invention provides a second assay for detecting ligands of an NR in vivo . Such second assay of the present invention employs transgenic animals which contain an effector-reporter vector of the present invention. The NR ligands can be detected by examining the expression of the reporter gene from the transgenic vector in the transgenic animal .

In a further embodiment, the present invention provides methods for identifying ligands of a nuclear receptor. Another embodiment of the present invention provides methods of making immortal ligand-producing cell lines . A further aspect of the present invention is directed to methods of treating a subject in need of ligands of an NR by administering to such subject, ligands identified or purified by the foregoing methods of the present invention, or synthetic analogs thereof.

In a further aspect of the invention, methods of assessing the pharmacological properties of an NR ligand, an NR ligand analog or an antagonist are provided.

BRIEF DESCRIPTION OF DRAWINGS

Figure 1 depicts the effector-reporter transgenic assay. a. The effector and reporter constructs. b. The induction of lacZ reporter expression by gRAR and gRXR. β-galactosidase activity was detected when CMV-gRXR and lacZr were cotransfected in the presence of all- trans RA (RA; 1 μM) , or the RXR-selective ligand SR11237 (SR; 1 μM) , but not the RAR-specific ligand TTNPB (TT; 0.1 μM) . In contrast, gRAR was efficiently activated by all- trans RA and TTNPB but not by SR11237. Transfections were performed in JEG-3 choriocarcinoma cells and β-galactosidase activity was measured by a luminometric assay. Bars represent mean of quadruples +/- standard deviations. c. and d. In si tu hybridization analysis of gRXR mRNA distribution in gRXR-positive embryos at stage Ell.5. The expression signal was observed along the entire spinal cord and developing brain (c) and in transverse sections of the spinal cord (d) . e. Binding of the transgenic effector protein to GAL4 binding site (UAS) DNA in gel-shift analysis. Extracts from E10.5 gRXR-transgenic embryos were incubated with ³²P-labeled UAS probe. A complex was formed (lane 1; lower arrow) migrating with similar mobility as in vitro-synthesized DNA-bound gRXR protein. Formation of the protein-DNA complex was inhibited in the presence of excess (50-fold) unlabeled specific UAS competitor DNA (s; lane 4), but not by non-specific DNA (ns; lane 3). The complex was supershifted when anti-HA antiserum was included in the binding reaction (lane 2; ab, upper arrow) . A more slowly migrating complex was not supershifted and apparently corresponded to a non-specific protein-DNA complex. Scale bar in c=lmm and in d=100μm.

Figure 2 depicts the induction of β-galactosidase expression by transgenic effector genes in vivo . β-galactosidase was specifically induced in double-positive gRXR/lacZr (a, b, c, d) and gRAR/lacZr (e, f) embryos at E10.5 (a, e) , Ell .0 (b, f) Ell.5(c) and E12.5 (d) , as visualized by whole-mount X-gal staining. In single-positive lacZr embryos (2g) , no staining was detected. In a, b, c and d, staining appeared in the spinal cord restricted to the fore- and hindlimb levels of gRXR/lacZr embryos. The staining patterns of gRAR/lacZr-positive embryos were similar but temporally distinct (e and f) . In h, the induction of the reporter gene was observed to be expanded by exogenously administered all- rans RA. β-galactosidase was expressed in a continuous pattern along the spinal cord at 12 hours post fetal treatment.

Figure 3 depicts cervical transsections of X-gal stained double-positive gRAR/LACZr (a, b) and gRXR/lacZr (d, e, f) embryos, which revealed a ventral-to-dorsal spatiotemporal shift of β-galactosidase-positive cells in the developing spinal cord. "As comparison, an Ell.5 control section of a -/lacZr embryo was shown (c) . At E10.5, strongly blue-stained cells of gRAR/lacZr embryos were detected in the ventral part of the developing spinal cord (a) , which were replaced by dorsal staining at Ell.0(b) . No blue cells were detected in the -/lacZr embryo (c) . A similar shift, but clearly delayed, appears in gRXR/lacZr-embryos . Only some few weakly positive cells were present in the ventral E10.5 spinal cord (d) . These increased in number and intensity at Ell.0(b), but have faded at Ell.5, when strong staining was observed dorsally (d) . Scale bar = lOOμm.

Figure 4 illustrates induction of β-galactosidase expression in E10.5 spinal cord explant cultures of gRXR/lacZr transgenic embryos. Explants derived from thoracic (a, b, c) segments were cultured in the presence of SR 11237 (1 μM; n=4) (a), TTNPB (0.1 μM; n=4) (b) , or vehicle (n=9) (c) . a, b, c, β-galactosidase positive cells, normally absent from thoracic spinal cord (c) , were detected in SRll237-treated (a) , but not in TTNPB-treated (b) cultures. Bars represent mean of quadruplets ± standard deviations. Scale bar in c = 100 mm.

Figure 5 depicts a single plasmid construct in effector-reporter transgenic assay. The figure graphically depicts the combined effector and reporter plasmid construct (gRAR-lacZ) and the generation of transgenic mice carrying the plasmid transgene. The figure further illustrates the use of X-gal staining for identifying tissues expressing the reporter lacZ gene.

Figure 6 depicts the induction of lacZ reporter expression by plasmid constructs g-lacZ and gRAR-lacZ in the presence or absence of TTNPB, which is a RAR selective agonist. g-lacZ is similar to the construct in Figure 5 with the only difference that it lacks the RAR ligand binding domain (indicated as "hRARα" in Figure 5) . β-galactosidase activity was detected when gRAR-lacZ was transfected in the presence of the synthetic RAR ligand TTNPB. The induction was not observed when the ligand binding domain of RAR was not included in the construct (g-lacZ) . Transfections in JEG-3 cells and β-galactosidase activity was measured as described in Figure 1.

Figure 7 depicts the induction of β-galactosidase expression by the combined transgenic construct gRAR-lacZ in vivo . β-galactosidase was detected in the Ell.5 embryo as visualized by whole-mount X-gal staining. Staining appeared in the spinal cord restricted to the fore- and hindlimb regions. Additional staining was observed, e.g., in the junction between the mid- and hindbrain.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to detection of ligands for nuclear hormone receptors in vivo .

The term "nuclear hormone receptors" or "nuclear receptors" as used herein refers to a family of intracellular protein receptors for steroids, vitamin D, thyroid hormone, retinoids, prostanoids and the like. A typical nuclear receptor can be divided into functional domains which include a DNA binding domain (or DBD) , a ligand binding domain (or LBD) and transactivating domains. There also exists a dimerization domain within LBD (Mangelsdorf et al . , 1995). Those skilled in the art also recognize unusual variants of NRs containing a LBD but lacking a DBD. An example of such NR is the orphan receptbr Dax-1 (Zanaria et al . (1994) Nature 372 : 635- 641. Examples of the nuclear receptors of the present invention include, but are not limited to, RXR (for 9-c retinoic acid), PPARα,β,γ (for fatty acids, prostanoids) , RAR (for retinoic acid), VDR (for 1, 25- (OH) 2-VD3 ) , TRα, β (for thyroid hormone) , FXR (for farnesoids) and other NRs described by Laudet et al . , Cell 97: 161-163 (1999). Proteins that are substantially homologous to members of the NR family are also included. By "substantially homologous" is meant that the degree of identity of a protein with any member of the NR family, or with the DBD or LBD of any member of the NR family, is at least about 35%, preferably at least about 40%. The degree is the similarity index calculated using the Lipman-Pearson Protein Alignment program with the following choice of parameters: Ktuple = 2, Gap Penalty = 4, and Gap Length Penalty = 12.

Nuclear receptors of the present invention include orphan receptors, proteins that by homology belong to the nuclear receptor superfamily, but for which ligands have not been identified (Mangelsdorf et al . (1995) ) . Orphan receptors contemplated by the present invention include, but are not limited to, LXRα,β, XONR,α,β,γ, HNF-4, GCNF (Germ Cell Nuclear Factor), Rev Erbα,β, SF-1, ERR1,2, RORα,β, TLX, LRH (Liver Receptor Homologue) , NGFl-B (Nerve Growth Factor Inducible-B) , NOR1, NURR1, MB67 and Dax-1.

In one embodiment, the present invention provides a first assay that detects ligands of nuclear receptors in vivo by using a pair of transgenic vectors: an effector vector and a corresponding reporter vector.

The effector vector for use in the first assay of the present invention includes a DNA sequence coding for an^' effector protein. As used herein, an effector protein includes a ligand binding domain (LBD) and a DNA binding domain (DBD) .

DNA-binding domains suitable for use in the effector protein are typically obtained from transcription factors. The term "DNA-binding domain" is understood in the art to refer to a polypeptide that is able to bind to DNA. As used herein, the term "DNA- binding domain" encompasses a minimal peptide sequence of a transcriptional factor up to the entire length of a transcriptional factor, so long as the DNA-binding domain functions to bind a specific DNA sequence. DNA-binding domains are known to function heterologously in combination with other functional domains. In other words, a DNA binding domain, when linked with another polypeptide to form a chimeric fusion protein, is able to maintain the ability to bind the specific DNA sequence (see, e.g., Brent and Ptashne, Cell 43: 729-736, 1985). Transcription factors suitable for use herein as a source of such DNA binding domains include, e.g., homeobox proteins (such as HOX, STF-1, Antp, Mat A-2 and INV) , zinc finger proteins (Zif268, GLI and XFin) , hormone receptors, helix-turn-helix proteins, helix-loop- helix proteins, basic-Zip proteins (bzip) , and the like. See, for example, Harrison (1991) , "A Structural Taxonomy of DNA-binding Domains", Nature 353: 715-719; Klug and Rhodes (1987) Trends Biochem . Sci . 12: 464; Jacobs and Michaels (1990) New Biol . 2: 583; and Jacobs (1992), EMBO J. 11: 4507-4517. DNA-binding domain(s) from members of the nuclear receptor superfamily can be used as well.

Additional DNA binding domains contemplated for use in the present invention include DNA-binding domains of the yeast transcription factor GAL4 and the bacterial transcription factor LexA. The DNA binding domain of the yeast GAL4 protein comprises at least the first 74 amino terminal amino acids thereof (see, for example, Keegan et al., Science 231: 699-704, 1986). Preferably, the first 90 or more, or more preferably, the first 147 amino terminal amino acids of the GAL4 protein can be used.

Ligand-binding domains (LBDs) suitable for use IN the effector protein are obtained from NRs, e.g., those described by Mangelsdorf et al . (1995) and by Laudet et al . (1999). As used herein, the term "ligand- binding domain" encompasses a continuous peptide sequence of a nuclear receptor that binds to a particular ligand(s) and that does not have an ability to bind DNA. In general, a potent ligand dependent activation domain is embedded within the LBD of NRs. This activation domain is usually refered to as activation domain 2 (AF2) (see e.g. Mangelsdorf et al . (1995)).

According to the present invention, an LBD can be positioned at either the amino or carboxy terminus of a DBD in the effector protein. Additional amino acids can be placed as a linker sequence between the LBD and the DBD of the effector protein, and/or at the N- or C- terminus of the effector protein outside the LBD and the DBD. The DNA sequence coding for the effector protein (also referred to herein as the "chimeric effector protein" or "effector fusion protein"), is placed on the effector vector in an operable linkage to a promoter. As used herein, the term "promoter" refers to a specific nucleotide sequence recognized by RNA polymerase, the enzyme that initiates RNA synthesis. The promoter sequence is the site at which transcription can be specifically initiated under proper conditions. Promoters contemplated for use in the practice of the present invention include inducible (e.g., minimal CMB promoter, minimal TK promoter, modified MMLV LTR and heatshock promoter), constitutive (e.g., β-actin promoter, MMLV LTR (non-modified) and DHFR promoter) , or tissue specific promoters (e.g., a nestin promoter, Zimmerman et al . , Neuron 12:11-24 (1994)). The nestin promoter is a tissue-specific promoter that directs expression in developing central nervous system (CNS) . An effector protein under the control of a nestin promoter is particularly useful for detection and localization of NR ligands in CNS.

In accordance with the present invention, a reporter vector includes a DNA sequence coding for a reporter protein. As used herein, a reporter protein can be any protein that provides a detectable signal, such as lacZ, GFP (green fluorescence protein) , BFP (blue fluorescence protein) , luciferase, alkaline phosphatase, and chloramphenicol acetyl transferase. The use of GFP allows detection in live cells. Coding sequences for these reporter proteins are available to those skilled in the art. A reporter-coding sequence, or "a reporter gene", is placed on the reporter vector in an operable linkage to a promoter. Any of the above-described promoters for use in an effector vector can be used in the reporter vector.

According to the present invention, the reporter vector also includes a DNA sequence to which a DBD binds, also referred to herein as "a DBD-binding sequence" . Such DBD-binding sequence is preferably positioned upstream of the promoter on the reporter vector. A DBD-binding sequence can include, for example, one or more copies in tandem of a DNA motif known to be recognized and bound by the DBD of a transcription factor.

Given an effector vector, "the corresponding reporter vector" is a vector that includes a DNA sequence to which the DBD of the effector protein binds. For example, when the DBD of the effector protein is the DBD of the yeast transcriptional factor GAL4, one or more copies of the GAL binding sequence CGGAGTACTGTCCTCCG (SEQ ID NO: 1) (Kang et al . , J^". Biol . Chem . 268:9629-9635 (1993)) is placed, in tandem, upstream of the promoter on the reporter vector.

Additional components which can also be incorporated into a reporter vector or an effector vector include polyadenylation signal sequences, or sequences coding for selectable markers, such as genes conferring antibiotic resistance, genes which enable cells to process metabolic intermediaries, and the like. Exemplary antibiotic resistance genes include genes which impart tetracycline resistance, genes which impart ampicillin resistance, zeomycin resistance, neomycin resistance, hygromycin resistance, puromycin resistance, and the like. Genes which enable cells to process metabolic intermediaries include genes which permit cells to incorporate L-histidinol, genes encoding thymidine kinase, genes encoding xanthine-guanine phosphoribosyl transferase (gpt) , genes encoding dihydrofolate reductase, genes encoding asparagine synthetase, and the like.

According to the present invention, the vectors into which reporter genes or effector genes are inserted can be any of the expression or gene-transfer vectors known in the art that can effect the transport of the reporter or the effector gene into desired host cells for expres^'sion and/or replication thereof. These vectors can be either circular or linear. Suitable vectors for use herein include plasmids, phage, recombinant virus, and other vectors that, upon introduction into an appropriate host cell, result in expression of the inserted DNA.

Appropriate expression vectors are well known to those of skill in the art and include those that are replicable in eukaryotic cells and/or prokaryotic cells, and those that remain episomal or integrate into the host cell genome. Examples of such vectors include pBluescript (Stratagen) and pUC18 (Clontech) .

To practice the first assay of the present invention, an effector vector and a corresponding reporter vector are introduced into an animal host to produce a double transgenic animal. Preferably, such animal host is a mouse.

As used herein, the phrase "transgenic animal" refers to an animal that contains one or more inheritable expression sequences containing one or more exogenous nucleic acid(s) . For purposes of the present invention, "a single transgenic animal" refers to an animal that contains either an effector vector or an reporter vector. "A double transgenic animal" contains both vectors. The vectors can remain episomal or integrate into the genome of the host animal. A vector which can be used to create a trangenic animal is also referred to herein as a transgenic vector.

A double transgenic mouse can be produced by first establishing separate single transgenic mouse lines and intercrossing single transgenic mouse lines.

Alternatively, double transgenic mouse embryos can be produced by injecting into fertilized eggs of a single transgenic mouse, the second vector. Both procedures are well known in the art. See, B. Hogan et al . (1994) Manipulating the Mouse Embryo (Cold Spring Harbor Press: Cold Spring Harbor, New York) . The introduction of the effector and the reporter gene into mouse tissues or embryos can be confirmed by a variety of routine methods, such as PCR, in si tu hybridization or immunohistochemistry.

In an alternative approach, transgenic mice can be produced by introducing an effector and/or the corresponding reporter vector into the mouse genome of embryonic stem (ES) cells via homologous recombination. ES cells having the effector and/or the reporter gene integrated in their genome can be used to generate mouse strains containing the desired effector and/or reporter gene. This approach allows targeting ("knock in") of the transgenes into a preselected genetic locus. Thus, the effector gene can be placed under the control of a particular promoter ensuring the desired expression of the transgene. The procedure of gene targeting by homologous recombination in ES cells is well known in the art.

Once a double transgenic mouse or embryo is established, the in vivo distribution of the ligands for the nuclear receptor can be analyzed. Activated by ligand binding, the effector protein binds to the DBD- binding sequence in the reporter vector and induces transcription of the reporter gene. To detect the expression of the reporter gene, a tissue sample from the transgenic animal, or the whole transgenic embryo, can be taken for analysis in a variety of assays . The assays can be based on the nucleotide sequence of the reporter gene, e.g., PCR, Northern Blot analysis, in si tu hybridization. The assays can also be based on the protein product of the reporter gene, e.g., immunostaining, Western Blot analysis or biochemical assays such as an enzymatic activity assay. For immunohistological staining, embryo cells can be fixed by formalin or other standard histological preservatives, dehydrated and embedded in paraffin as routinely done. Sections can be cut from the paraffin and mounted on glass slides, or the sections may be prepared from cryo-preserved tissue. Alternatively, cytological preparations can be made, e.g., by fixing cells on a slide. This can be achieved by exposing the cells to formalin in a buffer of a physiological pH, followed by suspension in acetone and pelleting onto gelatin-coated slides by centrifugation. The expressed reporter protein can be localized and quantitated, either by exposure to labeled antibody or by exposure to unlabeled antibody and a labeled secondary antibody. See, e.g., Current Protocols in Molecular Cloning (Ausubel et al . , John Wiley & Sons, New York) . When the lacZ gene is used as a reporter gene, the expression of the lacZ gene can be detected by X-gal staining, a procedure well known in the art and also described in Example 1. Transgenic embryos at different stages can be analyzed for the expression of the reporter gene. The pattern of the reporter expression reflects the locations at which the ligands are present.

In another embodiment, the present invention provides expression vectors which include coding sequences for both an effector protein and a reporter protein. Such vectors are also referred to herein as "effector-reporter vectors".

An effector-reporter vector of the present invention includes a first sequence encoding an effector proteih. Such first sequence is linked operably to a first promoter. Upstream of such first promoter is placed a first DBD-binding sequence, to which the DBD of the effector protein encoded by such vector binds. Such effector-reporter vector of the present invention also includes a second sequence coding for a reporter protein. Such second sequence is linked operably to a second promoter. Upstream of such second promoter is placed a second DBD-binding sequence, to which the DBD of the effector protein encoded by such vector binds.

Effector proteins and reporter proteins have been defined hereinabove. The choices for promoters, LBDs, DBDs, and DBD-binding sequences and the types of vectors are also as described hereinabove. The first promoter can be different from the second promoter. The first and the second DBD-binding sequence can also be different, e.g., by including different number of copies of the same minimal DBD-binding motif.

An example of such effector-reporter is gRAR- lacZ. The effector protein encoded by gRAR-lacZ is a fusion between the LBD of RAR (receptor for retinoic acid) and the DBD of GAL4. The reporter gene of gRAR- lacZ is the lacZ gene. The first promoter and the second promoter are identical in this case: both are the heat- shock promoter (hsp) . The first and the second DBD- binding sequences are identical as well: both sequences contain four tandem repeats of the GAL4 binding motif CGGAGTACTGTCCTCCG (SEQ ID NO: 1) . Intermediate vectors which are useful for making an effector-reporter vector of the present invention are also provided by the present invention. For example, g-lacZ is a vector which is the same as gRAR-lacZ except that g-lacZ lacks the LBD of RAR. Thus, g-lacZ can be used conveniently as an intermediate vector into which a sequence coding for the LBD of an NR can be inserted to generate a complete effector-reporter vector.

In another embodiment of the present invention, cells transformed with an effector-reporter vector of the present invention are provided. Such cells can be used for the purpose of maintaining or propagating a vector, testing the functionality of a vector, assessing the properties of a known or newly-identified NR ligand, or screening for a potential NR ligand.

Both prokaryotic and eukaryotic cells are contemplated by the present invention. Exemplary eukaryotic cells suitable for use in the present invention include, e.g., JEG-3 cells, CV-1 cells, P19 cells, ES cells (embryonic stem cells), COS cells, mouse L cells, Chinese hamster ovary (CHO) cells, primary human fibroblast cells, human embryonic kidney cells, African green monkey cells, cultured primary tissues, cultured tumor cells, neuronal progenitor or precursor cells, neuronal cells lines such as cerebellum derived neuronal precursors and PC12 cells, neurons, primary astrocytes, oligodendrocytes, and the like. Another embodiment of the present invention provides transgenic animals containing an effector-reporter vector of the present invention. Preferably, such animal is mouse. The methods for generating transgenic animals have been described hereinabove.

In still another embodiment of the present invention, a second assay is provided for detecting ligands of an NR receptor in vivo .

The second assay of the present invention employs a transgenic animal the present invention which contains an effector-reporter vector of the present invention. According to the present invention, the expression of the effector and the reporter genes in the transgenic animal is controlled in an autoregulatory manner. Ligands for the nuclear receptor at issue bind to the LBD of the effector proteins which are expressed at basal levels in the transgenic animal. The ligand- effector complexes thus formed, in turn, bind to the first and second DBD-binding sequence on the vector and upregulate the transcription of both the effector and the reporter gene. The in vivo distribution of the ligands for a particular nuclear receptor can be determined from detecting the expression of the reporter gene in various tissues and at various developmental stages of the transgenic animal. Detection methods which are described for use in the first assay of the present invention can be applied here as well, such as PCR, Northern Blot analysis, in si tu hybridization, immunostaining, Western Blot analysis or biochemical assays such as an enzymatic activity assay. To illustrate the efficiency of the second assay, transgenic mouse embryos containing the vector gRAR-lacZ have been examined by the present inventors. The vector gRAR-lacZ was injected into the fertilized eggs of mice. Transgenic embryos at various stages were taken and the pattern of the lacZ expression was analyzed by histological X-gal staining. As described in Example 7, the X-gal staining pattern is consistent with the previous knowledge concerning the distribution of endogenous retinoids . In addition, new regions not previously known for having retinoids were also identified.

In a further embodiment, the present invention provides methods for identifying ligands of a nuclear receptor. Such ligands can be novel ligands of an NR, or ligands of an orphan receptor.

According to this embodiment, the in vivo sites of ligands of a nuclear receptor can be determined by using the assays of the present invention described hereinabove. Cells at such sites can be isolated from the transgenic animal or a normal non-transgenic animal host, and ligands can be extracted and purified by routine procedures for purifying small lipophilic molecules, e.g., HPLC . The structure of the ligands can be determined by procedures known to those skilled in the art, e.g., mass-spectrometry.

According to the present invention, most NRs signal in a para- or autocrine fashion and their ligands are synthesized locally within tissues where such ligands are of functional importance. Thus, once the in vivo tissue sites of ligands for a nuclear receptor are determined by the foregoing assays of the present invention, immortal cell lines can be established from cells of such tissues. Immortalization can be achieved by a variety of methods . One such method is to generate a transgenic animal containing an effector-reporter vector of the present invention, wherein the reporter gene can be the gene coding for a temperature sensitive large T antigen from SV 40, polyoma virus, adenovirus

ElA, or E7 of HPV16 or HPV17, or other oncogenes, such as vMyc. Relevant teachings can be found in, e.g., Whitehead et al . (1993) Proc . Natl . Acad . Sci . USA 90 : 587-591 and Morgan et al . (1994) Dev. Biol . 162 : 486-498. In this manner, an oncogene is expressed in the transgenic animal in those tissues where ligands are present. Cells which synthesize NR ligands can then be made immortal for use as an unlimited source of NR ligands. Accordingly, such methods of making immortal ligand-producing cell lines are also features of the present invention.

The established ligand-producing cell lines also express the enzymatic machinery required for biosynthesis of a particular NR ligand. Defining the enzymatic mechanisms can provide additional opportunities for pharmacological manipulation of NR signaling pathways . A further aspect of the present invention is directed to methods of treating a subject in need of ligands of an NR receptor by administering to such subject, ligands identified or purified by the foregoing methods of the present invention. The term "subject" used herein refers to any mammalian subject. Preferably, the subject is a human subject. According to the present invention, subjects in need of ligands of an NR include patients suffering a disorder characterized by an abnormal NR signaling, e.g., atherosclerosis, diabetes, obesity, Parkinson's disease and various types of cancers, Alzheimer's disease, inflammation, and disorders of the immune system. Additionally, disorders not associated with abnormal NR signaling can also be treated. For example, glucocorticoids can be used to treat disorders characterized by inflammation, such as asthma, because of the potent anti-inflammatory effects mediated by the active glucocorticoid receptor. Similarly, ligands for the nuclear receptor PPARgamma (thiazolidinedioneε) have been shown to be beneficial in patients suffering from non-insulin dependent diabetes, even though such disorders are not associated with abnormal PPARgamma signaling. The term "treatment" refers to modulating the NR signaling in an effective manner so as to prevent or delay the onset, retard the progression or ameliorate the symptoms of the disorder. In accordance with the present invention, the purified NR ligands can be easily delivered to subjects in need thereof due to the small size and lipophilic nature of these ligands. It is known that NR ligands in vertebrates are highly conserved. Thus, NR ligands identified from animals, e.g., mice, are likely to be identical or closely related to the human counterparts, and thus, suitable for therapeutic use in humans.

In a further aspect of the invention, methods of assessing the pharmacological properties of an NR ligand, an NR ligand analog or an antagonist are provided.

Pharmacological properties which can be assessed by such methods of the present invention include, for example, the efficacy in activating a nuclear receptor, the bioavailability to a particular tissue, the stability, the effective dosage for treating a disorder. The types of substances which can be used in making formulations containing NR ligands, as well as the properties of such formulations can be assessed using the methods of the present invention as well. In addition, other drugs which may modulate NR signaling, e.g., by affecting metabolism synthesis of NR ligands can be studied using the assays of the present invention.

Such methods of the present invention employ an effector-reporter vector of the present invention.

Depending on the nature of the property to be assessed, the assays can be carried out either in vi tro or in vivo, or both. For in vi tro assays, cells transformed with an effector-reporter vector of the present invention can be used. Ligands of interest can be added to cultures of such cells and the pharmacological properties can be analyzed. For in vivo studies, transgenic animals established according to the present invention can be used. For example, specific ligands can be administered to a transgenic mouse established by using an transgenic vector of the present invention. The pharmacological properties of the ligands in the mouse can be examined by using approaches, e.g., those described in Example 4.

The present invention is further illustrated by the following examples.

All the publications mentioned in the present disclosure are incorporated herein by reference. The terms and expressions which have been employed in the present disclosure are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, recognizing that various modifications are possible within the scope of the invention.

EXAMPLE 1 GENERAL METHODS Plasmid constructions

The transgenic reporter construct, lacZr, contained four repeats of GAL4-specific binding sites (UAS) followed by a minimal thymidine kinase (tk) promoter, which was further linked to the lacZ gene. Plasmid lacZr was generated by subcloning a PvuII-XhoI fragment from pMHlOO-tk-luc (Perlmann, T., & Jansson, L. Genes Dev. 9, 769-782 (1995)) into the EcoRI sites of ptkβ (Clontech) .

To generate the effector constructs gRXR and gRAR, the nine amino acid long hemagglutinin epitope of Influenza virus (HA: nucleotide sequence CTACCCCTACGACGTCCCCGACTACGC (SEQ ID NO: 2) , was introduced at the NH₂-terminus of the yeast GAL4 DNA binding domain of pCMX-GAL4 (amino acids 1-147) (Sadowski, I., & Ptashne, M. Nucl . Acids Res . 17, 7539

(1989)) to generate pCMX-HA-GAL4. GAL4 was exchanged for HA-GAL4 in pCMX-GAL4-RXR and pCMX-GAL4-RAR, that contained GAL4-DBD fused to the ligand binding domains of the human RXR and the RAR gene. Resulting constructs were named CMV-gRXR and CMV-gRAR. The HA-GAL4-RXR and HA-GAL4-RAR coding sequences were subcloned into pNes-SV40pA containing the rat nestin promoter and enhancer (Zimmerman, L., et al . Neuron 12, 11-24 (1994)). For microinjections, cesium-chloride purified plasmid DNA was digested and vector sequences were removed in low-melting gel electrophoreses . The insert was purified and dissolved in injection buffer (10 mM Tris, 0.1 mM EDTA, pH 7.4). Transfections

Transfections of human JEG-3 choriocarcinoma cells were performed in quadruplets in 24-well plates by the calcium phosphate method as described by Perlmann, T., & Jansson, L. Genes Dev. 9, 769-782 (1995). At 15 hrs post-transfection, all-trans RA (1 μM) , TTNPB ( (E)-4-[2-(5,5,8,8-tetramethyl-5, 6,7,8 tetrahydro -2-naphtalenyl) -1-propenyl] benzoic acid; 0.1 μM) , SR11237 (1 μM) or ethanol was added with fresh medium. At 36 hours, cell lysates were luminometrically assayed for β-galactosidase (Galacto-Light™, Tropix, Bedford, MA, USA) in a microplate luminometer/photometer (Lucy-1, Anthos, Salzburg, Austria).

Transgenic animals

For prenatal retinoid treatment, pregnant mice were gavage-fed 11 days post coitum with all- trans RA dissolved in 300 μl corn oil at a concentration of 20 mg/kg x weight^"1. After 12 hours, treated embryos were analyzed by X-gal histochemistry.

Gel-shift analyses

Total embryonic extracts were prepared from E10.5 gRXR transgenic embryos by sonicating in binding buffer containing 10 mM Tris, (pH 8.0), 40 mM KCl, 0.05% NP-40, 6% glycerol with 1 mM phenylmethylsufonylfluoride, and centrifugation. After analysis for total protein content, extracts were incubated in binding buffer with 1 mM DTT and 0.2 μg poly[d(IC)], mixed with 0.2-0.5 ng ³²P-labeled UAS oligonucleotide probe

(AGCTCGACGGAGTACTGTCCTCCGTCGA (SEQ ID NO: 3) ) , alone, in presence of 50-fold excess of unlabeled UAS-or non-specific double stranded DNA oligonucleotide, or with monoclonal anti-HA antibodies. Samples were run in a 4% non-denaturing polyacrylamide gel and analyzed by autoradiography.

Histochemical analyses

Embryos were analyzed at indicated stages from E10.5 to E13.5 where the morning of plugging was counted as 0.5 days post coitum (E0.5). Stages were confirmed based on morphology. For X-gal (5-bromo-4-chloro-3-indoyl-b- D-galactopyranoside) whole-mount staining, embryos were fixed by immersion in 0.2% glutaraldehyde 7-30 minutes and processed as described by Nilsson, E. & Lenhahl, U. Mol . Repr. Dev. 34, 149-157 (1993) . Embryos were postfixed in 2-4% paraformaldehyde for 1-2 days and stored in 30% sucrose. For indirect immunohistochemistry and in si tu hybridization, explants and embryos were fixed by immersion in 2% paraformaldehyde for 10-30 minutes, stored in 30% sucrose before cryosectioning at 14 μm. Incubations with primary antibodies were carried out overnight with indicated dilutions: Tuj-1 (anti-γ-tubulin) at 1:250; 4D5 (anti-Islet 1/2) at 1:50.

For in si tu hybridizations, sections were incubated overnight with ³⁵S-labeled oligonucleotide probes and processed as described by Dagerlind, A. ,

Friberg, K. & Hokfelt, T. Histochem . 98: 39-49 (1992). Two GAL-4 specific probes with the following sequences were used: AAAACCAAAAGGTCTCCGCTGACTAGGGCACATCTGACAGAAGTGGAA (SEQ ID NO: 4) , and

ATGCGATATTTGCCGACTTAAAAAGCTCAAGTGCTCCAAAGAAAAACC (SEQ ID NO: 5) . Both probes resulted in identical hybridization patterns .

Sections were analyzed under florescence-, bright-, and darkfield microscopy (Nikon, Tokyo, Japan). Photographed sections were scanned and processed in Photoshop™ (Adobe, Mountain View, CA) .

Explant Cultures

Explants were cultured for 1-5 days, as described previously (A.G.S. Lumsden et al . (1983) Nature

306: 786-788), in Dulbecco's Modified Eagle's Medium containing 0.2% ethanol, 1 μM SR11237, or 0.1 μM TTNPB.

Cultures were subsequently fixed and X-gel stained or analyzed by immunohistochemistry.

EXAMPLE 2 EXPRESSION OF gRAR/gRXR AND lacZr IN EG-3 CELLS

The transgenic effector and reporter constructs, CMV-gRAR, CMV-gRXR and lacZr, were first tested in a transient transfection assay in human chorion carcinoma JEG-3 cells. JEG-3 cells were transfected as described in Example 1. β-galactosidase activity was measured by a luminometric assay. (See, Example 1, General Methods . ) As shown in Figure lb, β-galactosidase activity was detected when CMV-gRXR and lacZr were cotransfected in the presence of all- trans RA (RA; 1 μM) , or the RXR-selective ligand SR11237 (SR; 1 μM) , but not the RAR-specific ligand TTNPB (TT; 0.1 μM) . In contrast, gRAR was efficiently activated by all- trans RA and TTNPB but not by SR11237. Activation of both receptors by all-trans-RA was due to partial isomerization of all-trans RA into 9-cis RA (Allenby, G., et al . Proc . Natl . Acad. Sci . USA 90, 30-34 (1993)). Thus, transfection and expression of CMV-gRXR and CMV-gRAR efficiently induced lacZ expression in response to high doses of all-trans RA. In contrast, RAR- and RXR-selective synthetic retinoids specifically activated gRAR and gRXR, respectively. These results demonstrated that the ligand responses of the hybrid receptors are highly selective. EXAMPLE 3 EXPRESSION OF gRAR/gRXR AND lacZr in vivo

For ligand detection in vivo, gRXR and gRAR effector proteins were expressed in transgenic animals. To analyze the function of retinoid signaling during CNS development, gRAR and gRXR were cloned downstream of the rat nestin promoter and enhancer as described in Example 1. The nestin promoter was previously demonstrated to promote highly efficient expression in the developing CNS (Lendahl U. , Zimmerman, L.B. & McKay, R.D. Cell 60,

585-95 (1990) and Zimmerman, L., et al . Neuron 12, 11-24 (1994)). The resulting effector plasmids, nes-gRXR and nes-gRAR, and the reporter plasmid lacZr, were used to generate transgenic mice as described by Nilsson, E. & Lenhahl, U. Mol . Repr. Dev. 34, 149-157 (1993). Founder lines were established for nes-gRXR and lacZr, whereas nes-gRAR embryos of different stages were obtained by transient injection experiments since it was found that the nes-gRAR transgene was not transmitted through the germline. Transgenic embryos and mice were genotyped by PCR using primers specific for the lacZ and GAL4-DBD sequences, respectively, and using DNA extracted from amnion sac or tail as templates. Three out of four nes-gRXR effector founder lines showed high and reproducible transgenic expression as judged from in si tu hybridization analysis. One reporter line showing reproducible lacZ staining when crossed with gRXR-effector lines was chosen and used in all subsequent experiments. gRAR/lacZr double-transgenic embryos were generated by injecting nes-gRAR DNA into fertilized eggs from mice heterozygous for the lacZr transgene.

The expression of transgenic gRXR protein is illustrated in Figure lc and Id, where in si tu hybridization, using a GAL4-specific probe, detected high levels of gRXR mRNA expression in the developing CNS at embryonic day 11.5 (Ell.5). Immunostaining with antibodies specific for the HA-epitope correlated well with the distribution of gRXR mRNA. The DNA binding capacity of the transgenic gRXR protein was analyzed in a gel retardation assay using a ³²P-radiolabeled GAL4 DNA binding site as probe (Figure le) . In gRXR whole-embryo extracts, a shifted complex was formed (lane 1; lower arrow) that was efficiently inhibited by the addition of excess non-radioactive specific competitor DNA in the binding reaction (lane 4) , but not by non-specific DNA (lane 3) . Furthermore, the complex was supershifted when an anti-HA antibody was included in the binding reaction (lane 2; upper arrow). This supershift confirmed the presence of gRXR in the protein-DNA complex.

EXAMPLE 4

DETECTION OF LIGAND-ACTIVATED gRAR/gRXR Transgenic embryos carrying effector and reporter genes either individually or in combination, were collected and examined for expression of β-galactosidase by whole-mount X-gal histochemistry. The analysis focused on stages from E10.5 and onwards when the nestin promoter was highly active. As shown in Figure 2, β-galactosidase expression was detected in gRAR/lacZr and gRXR/lacZr embryos, but not in embryos carrying lacZr, gRAR or gRXR transgenes alone. A specific X-gal staining was detected in the spinal cord of gRXR/lacZr embryos. At E10.5, gRXR-dependent reporter induction was barely detectable, while at Ell clear staining was observed at the forelimb level of the spinal cord (Figures 2a and 2b) . Staining became intensified at Ell.5 when a weak signal was also detected at the hindlimb level of the spinal cord (Figure 2c) . At subsequent stages, β-galactosidase induction declined in the forelimb region while staining in the hindlimb region peaked at E12.5 (Figure 2d). This pattern is consistent with the delayed development of hindlimbs relative to forelimbs. A distinct but related pattern was detected in gRAR/lacZr double-positive embryos. Reporter activity significantly preceded that induced by gRXR (Figures 2e and 2f) . Thus, gRAR-dependent induction was detected already at E10.5 at the forelimb level of the spinal cord (Figure 2e) , while gRXR-induced reporter expression, as shown in Figure 2b, appeared at E11.0. To verify the reporter gene could be induced also at other levels of the spinal cord, transgenic embryos were treated in utero with high doses of all-trans RA. In gRXR/lacZr double-positive embryos, this treatment resulted in strong β-galactosidase staining evenly distributed throughout the spinal cord, while staining in the developing brain was weak despite expression of the effector gene in this region (Figure 2h) .

These results indicated that the β-galactosidase expression was concentrated to cervical and lumbal segments of the spinal cord. These regions have been demonstrated to be "hot spots" for retinoid synthesis in other related studies. A retinaldehyde dehydrogenase (RALDH-2), proposed to be critical for RA synthesis, is expressed in the embryonic ventral spinal cord specifically at these levels (McCaffery, P. & Drager, U.C. Proc . Natl . Acad. Sci . USA 91, 7194-7197 (1994); Zhao, D., et al . Eur. J. Biochem . 240, 15-22

(1996) and Niederreither, K. , McCafferty, P., Drager, U. , C, Chambon, P. & Dolle, P. Mech . Dev. 62, 67-78 (1997)). In addition, previously described analysis of embryos containing RA-responsive lacZ reporter genes have indicated the presence of activating retinoids in a similar pattern (Roosant, J., Zirngibl, R. , Cado, D., Shago, M. & Giguere, V. Genes Dev. 5, 1333-1344 (1991); Reynolds, K. , Mezey, E. & Zimmer, A. Mech . Dev. , 36, 15-29 (1991); Mendelsohn, C, Ruberte, E., LeMeur, M. , Morriss-Kay, G. & Chambon, P. Development 113, 723-734

(1991) and Balkan, W. , Colbert, M. , Bock, C. & Linney, E. Proc . Natl . Acad. Sci . USA 89, 3347-3351 (1992)). However, these studies did not distinguish between RAR and RXR activating ligands whereas the effector-reporter strategy described in the present application was designed to assess the specific activation by RXR and RAR, respectively. To further compare the locations of ligand-activated RXRs and RARs, sections from double-positive X-gal stained embryos were examined. Patterns of β-galactosidase expression revealed a ventral-to-dorsal spatiotemporal shift in cervical sections of stained embryos. In gRAR-positive embryos, staining was most intense in the ventral spinal cord at E10.5 (Figure 3a) and later stained strongly in the dorsal region (Figure 3b) . A similar but delayed shift was observed in gRXR embryos. At E10.5, only scattered cells positive for β-galactosidase could be detected in the ventral spinal cord (Figure 3d) . In contrast, lacZr was highly induced ventrally at Ell.O (Figure 3e) while staining increased in dorsal regions at later stages (Ell.5, Figure 3f; and E12.5). In sections from the reporter mice, no reporter induction was detected (-/lacZr; Figure 3c) . Taken together, the data showed that gRAR and gRXR were both active in spatially overlapping but temporarily unique patterns in the ventral and dorsal spinal cord. Such selective activation of gRXR and gRAR was consistent with the specific activation by synthetic ligands in transfected tissue cultured cells (Figure 1) .

EXAMPLE 5 VERIFYING SPECIFICITY OF EFFECTOR FUSION PROTEIN in vivo

At stages when both gRAR and gRXR induce β-galactosidase expression (see Ell.O; Figures 2b and 2f) , the dorsoventral distribution of positive cells is distinct, with gRAR-signaling being most prominent dorsally and that of gRXR ventrally (Figure 3) supporting the selectivity in activation of gRXR and gRAR. These results indicate that reporter induction by gRXR is unlikely a consequence of heterodimerization with endogenous RAR .

The specificity of reporter gene induction in gRXR/lacZr transgenic mice was further confirmed in the following way:

Transgenic spinal cord explants from thoracic segments of E10.5 embryos, which do not contain X-gal positive cells in gRXR/lacZr transgenic embryos, were treated with RXR- or RAR selective retinoids, SR11237 and TTNPB, respectively (see also Figure lb) . After 24 hours, only explants cultured in the presence of SR11237 showed strong X-gal staining (Figures 4, a-c) .

EXAMPLE 6 TRANSGENIC CONSTRUCT gRAR-lacZ

Transgenic expression vector (gRAR-lacZ including both effector (GAL4-RAR) and reporter (lacZ) genes (Figure 5) was generated as follows:

1. A reporter plasmid (uas-hsp-lacZ) was generated and included the following sequences: 4 GAL4 binding sites followed by the hsp 68 promoter, a lacZ gene and a polyadenylation site. To generate such reporter construct (uas-hsp-lacZ) , a Sall-Pvull fragment from the UASx4-tk-luc reporter (Forman et al . , Cell 81,541-550, 1995) was sublconed into the Smal site of pKS-hsp-lacZ (Kothary et al . , Development 105: 707-714, 1989) . The new plasmid contains four GAL4-specific binding sites (CGGAGTACTGTCCTCCG (SEQ ID NO: 1) ; Kang et al., J. Biol . Chem . 268: 9629-9635, 1993) followed by the hsp 68 promoter linked to a bacterial lacZ gene and polyadenylation site of pKS-hsp-lacZ (Kothary et al . , Development 105: 707-714, 1989).

2. The effector construct uas-hsp-gRAR included the following sequences: 4 GAL4 binding sites followed by the hsp 68 promoter, a hybrid gene encoding the GAL4 DNA binding domain linked in frame to the ligand binding domain of the human retinoic acid receptor α (RAR) and a polyadenylation site. To generate such effector construct, the pCMX-GAL4-RAR vector (Forman et al . , Cell 81, 541-550, 1995) was used as a template. Primers were designed to amplify most of the GAL4 DNA-binding domain

(aa 1-84), including the Xhol site at amino acid 74. The 5' primer (5'-C AAA ACC ATG GCG AAG CTA CTG TCT TCT ATC GAA C-3 ' (SEQ ID NO: 6) ) was designed to incorporate a Ncol restriction site upstream of the GAL4 DBD sequence, and the 3' primer (5'-CCC GGC GGC CGC GCT AGC CCA TTT TCA AAA TCA TGT CAA G-3 ' (SEQ ID NO: 7)) incorporated a Nhel site and a Notl site downstream of the amplified GAL4 DBD sequence. The resulting PCR fragment was digested with Ncol and Notl, and cloned into Ncol/Notl digested pKS- hsp-lacZ (Kothary et al . , Development 105: 707-714, 1989) . This generated a plasmid in which the lacZ coding region and downstream polyadenylation signal from pKS- hsp-lacZ was replaced by the sequence encoding the a fragment of the GAL 4 DBD (amino acid 1-74) . This vector was digested with Xhol and Nhel, and ligated with the Xhol and Nhel fragment from pCMX-GAL4-RAR (Forman et al . , Cell 81, 541-550, 1995), incorporating the rest of the GAL4 DBD (amino acids 75-147) and the ligand binding domain of the human RARα LBD (from Glul56 to Pro462) . A polyadenylation signal sequence from pCMX (Umesono et al . Cell 65, 1255-1266, 1991) was PCR amplified using the primers 5 ' -GAG AAC CCA CTG CTT AAC TGG C-3' (5') (SEQ ID NO: 8)) and 5 ' -CCA CCC GCG GCC GCT CCA GAC ATG ATA AGA TAC ATT-3' (3') (SEQ ID NO: 9)) and was inserted ino the Nhel and Notl sites downstream of the GAL4-RAR sequence to generate the effector plasmid uas-hsp-gRAR. 3. The combined effector and reporter construct was generated by ligating a fragment including the entire insert from uas-hsp-lacZ downstream of the entire insert of the effector vector uas-hsp-gRAR to generate the combined vector gRAR-lacZ (Figure 5) . In this plasmid, all vector sequences can be excised using Notl to generate a fragment containing the elements indicated in Figure 5. Transgenic construct gRAR-lacZ was first tested in a transient transfection system. JEG-3 cells were transfected with plasmid gRAR-lacZ, and β-galactosidase activity was measured posttransfection following the procedure as described in Example 1. As a control, a similar plasmid carrying a non-functional effector, lacking the RAR LBD, was used in the same experiment. As shown in Figure 6, the GAL4-RAR effector fusion protein was activated by the RAR-specific ligand TTNPB (0.1 μM) indicated by significant β-galactosidase activity detected. In contrast, when the non-functional effector was used, ligand addition did not induce the activation of lacZ.

Expression of the gRAR-lacZ transgene in vivo was examined by X-gal staining. Transgenic embryos were generated by injecting the transgenic construct into fertilized eggs. Analysis of X-gal stained embryos staged between 10.5 and 12.5 days post coitum (dpc) , revealed a specific pattern of β-gal expression, confined to the developing CNS and limb buds, an example of which is shown in Figure 7. At 11.5 dpc, strong staining was observed in the spinal cord, with intensified staining at limb levels (Figure 7A) . LacZ expression was also detected in the developing fore- and hindbrain (Figure 7B) . Furthermore, in several embryos, blue cells were present at two distinct regions in the proximo-anterior and -posterior part of the forelimb buds (Figure 7C) . Importantly, the pattern of blue cells was consistent with previous knowledge concerning the distribution of endogenous retinoids. In addition, new regions not previously identified as a source of retinoids were also identified. EXAMPLE 7 GENERATION OF gPPARγ-lacZ MICE

The ligand binding domain of human RARα in the vector gRAR-lacZ was replaced by the ligand binding domain of mouse PPARγ to generate gPPARg-lacZ. The construct was generated by digesting gRAR-lacZ with Xhol and Nhel (excises a fragment encoding amino acids 75-147 of the GAL4 DBD and the entire LBD of human RARα. The digested vector was ligated to a fragment containing the GAL4 DBD (from amino acid 75-147) followed by the ligand binding domain of mouse PPARy (amino acid 174-475) . The transgenic vector was used to generate transgenic mice.

Claims

WE CLAIM :

1. An expression vector, which comprises: a first DNA sequence encoding an effector protein comprising a DNA-binding domain (DBD) and a ligand-binding domain (LBD) ; a first promoter; a first DBD-binding sequence; wherein said first DNA sequence is operably linked to said first promoter, said first promoter is operably linked to said first DBD-binding sequence and said effector protein binds to said first DBD-binding sequence via said DBD; a second DNA sequence encoding a reporter protein; a second promoter; and a second DBD-binding sequence, wherein said second DNA sequence is operably linked to said second promoter, said second promoter is operably linked to said second DBD-binding sequence and said effector protein binds to said second DBD-binding sequence via said DBD;

2. The vector of claim 1, wherein said LBD is the LBD of a nuclear receptor.

3. The vector of claim 2, wherein said nuclear receptor is selected from the group consisting of RXR, PPARα,β,γ, RAR, VDR, TRα,β, FXR, LXRα,β, XONR, COUPα,β,γ,

HNF-4, GCNF, Rev Erbα,β, SF-1, ERR1,2, RORα,β, TLX, LRH, NGFI-B, NOR1, NURRl , MB67 and Dax-1.

4. The vector of claim 1, wherein said DBD is the DBD of a transcription factor selected from the group consisting of a homeobox protein, a zinc finger protein, a hormone receptor, a helix-turn-helix protein, a helix- loop-helix protein, a basic-Zip protein, LexA and GAL-4.

5. The vector of claim 4, wherein said transcription factor is the yeast GAL4.

6. The vector of claim 5, wherein said first

DBD-binding sequence and said second DBD-binding sequence comprise at least one copy of the DNA-binding motif for GAL4, CGGAGTACTGTCCTCCG (SEQ ID NO: 1) .

7. The vector of claim 1, wherein said reporter protein is selected from the group consisting of lacZ, GFP, BFP, luciferase, alkaline phosphatase and chloramphenicol acetyl transferase.

8. The vector of claim 1, wherein said first promoter and said second promoter are selected from the group consisting of a minimal CMB promoter, a minimal TK promoter, an MMLV LTR, a heatshock promoter, a β-actin promoter, a DHFR promoter and a nestin promoter.

9. An expression vector, having the designation gRAR-lacZ.

10. An expression vector, having the designation g-lacZ.

11. A cell transformed with the vector of claim 1.

12. A transgenic animal, comprising the vector of claim 1.

13. The transgenic animal of claim 12, wherein the animal is mouse.

14. A method for detecting ligands of a nuclear receptor in vivo, comprising:

(1) producing a transgenic animal containing an expression vector, wherein said vector comprises: a first DNA sequence encoding an effector protein comprising a DNA-binding domain (DBD) and the ligand-binding domain (LBD) of said nuclear receptor; a first promoter; a first DBD-binding sequence; wherein said first DNA sequence is operably linked to said first promoter, said first promoter is operably linked to said first DBD-binding sequence and said effector protein binds to said first DBD-binding sequence via said DBD; a second DNA sequence encoding a reporter protein; a second promoter; and a second DBD-binding sequence, wherein said second DNA sequence is operably linked to said second promoter, said second promoter is operably linked to said second DBD-binding sequence and said effector protein binds to said second DBD-binding sequence via said DBD; and (2) detecting the expression of said reporter in said transgenic animal thereby detecting ligands of said nuclear receptor.

15. The method of claim 14, wherein said transgenic animal is mouse.

16. The method of claim 15, wherein the transgenic mouse is produced from fertilized eggs injected with said vector.

17. The method of claim 15, wherein the transgenic mouse is produced from embryonic stem cells comprising said vector and wherein said vector has been integrated into the genome of said stem cells.

18. The method of claim 14, wherein the expression of said reporter is detected by in situ hybridization, immunostaining, histochemical analysis or an biochemical assay.

19. The method of claim 14, wherein said nuclear receptor is selected from the group consisting of RXR, PPARα,β,γ, RAR, VDR, TRα,β, FXR, LXRα,β, XONR,

COUPα,β,γ, HNF-4, GCNF, Rev Erbα,β, SF-1, ERR1,2, RORα,β, TLX, LRH, NGFI-B, NORl , NURRl , MB67 and Dax-1.

20. The method of claim 14, wherein said DBD is the DBD of a transcription factor selected from the group consisting of a homeobox protein, a zinc finger protein, a hormone receptor, a helix-turn-helix protein, a helix-loop-helix protein, a basic-Zip protein, LexA and GAL-4.

21. The method of claim 20, wherein said trascription factor is GAL-4, and wherein said first and said second DBD-binding sequence comprise at least one copy^'of the DNA-binding motif for GAL4 , CGGAGTACTGTCCTCCG (SEQ ID NO: 1) .

22. The method of claim 14, wherein said first promoter and second promoter are selected from the group consisting of a minimal CMB promoter, a minimal TK promoter, an MMLV LTR, a heatshock promoter, a β-actin promoter, a DHFR promoter and a nestin promoter.

23. The method of claim 14, wherein said vector is gRAR-lacZ.

24. The method of claim 14, wherein said reporter is selected from the group consisting of lacZ, GFP, BFP, luciferase, alkaline phosphatase and chloramphenicol acetyl transferase.

25. A method for identifying ligands of a nuclear receptor, comprising (1) producing a transgenic animal containing an expression vector, wherein said vector comprises: a first DNA sequence encoding an effector protein comprising a DNA-binding domain (DBD) and the ligand-binding domain (LBD) of said nuclear receptor; a first promoter; a first DBD-binding sequence; wherein said first DNA sequence is operably linked to said first promoter, said first promoter is operably linked to said first DBD-binding sequence and said effector protein binds to said first DBD-binding sequence via said DBD; a second DNA sequence encoding a reporter protein; a second promoter; and a second DBD-binding sequence, wherein said second DNA sequence is operably linked to said second promoter, said second promoter is operably linked to said second DBD-binding sequence and said effector protein binds to said second DBD-binding sequence via said DBD;

(2) determining the tissues where the ligands for said nuclear receptor are present by detecting the expression of said reporter in said transgenic animal; and

(3) isolating the ligands from the tissues identified in (2), thereby identifying ligands of said nuclear receptor.

26. An isolated ligand of a nuclear receptor, identified by the method of claim 25.

27. A method of making immortal cell lines which produces ligands of a nuclear receptor, comprising:

(1) producing a transgenic animal containing an expression vector, wherein said vector comprises: a first DNA sequence encoding an effector protein comprising a DNA-binding domain (DBD) and the ligand-binding domain (LBD) of said nuclear receptor; a first promoter; a first DBD-binding sequence; wherein said first DNA sequence is operably linked to said first promoter, said first promoter is operably linked to said first DBD-binding sequence and said effector protein binds to said first DBD-binding sequence via said DBD; a second DNA sequence encoding an oncogenic protein; a second promoter; and a second DBD-binding sequence, wherein said second DNA sequence is operably linked to said second promoter, said second promoter is operably linked to said second DBD-binding sequence and said effector protein binds to said second DBD-binding sequence via said DBD; (2) isolating the tissues where the ligands for said nuclear receptor are present; and

(3) establishing immortal cell lines from the tissues isolated in (2) .

28. The method of claim 27, wherein said oncogenic protein is vMyc, a temperature sensitive large T antigen from SV40, polyoma virus, adenovirus ElA, E7 of HPV16 or E7 of HPV17.

29. A method of treating a subject in need of a ligand of an NR receptor of claim 26, comprising administering said ligand to said subject.

30. The method of claim 29, wherein said subject is a patient suffering a disorder selected from the group consisting of atherosclerosis, diabetes, obesity, Parkinson's disease, cancer Alzheimer's disease, inflammation and an immunological disorder.

31. A method for assessing a pharmacological property of a nuclear receptor ligand in a transgenic animal, comprising: (1) producing a transgenic animal containing an expression vector, wherein said vector comprises: a first DNA sequence encoding an effector protein comprising a DNA-binding domain (DBD) and the ligand-binding domain (LBD) of said nuclear receptor; a first promoter; a first DBD-binding sequence; wherein said first DNA sequence is operably linked to said first promoter, said first promoter is operably linked to said first DBD-binding sequence and said effector protein binds to said first DBD-binding sequence via said DBD; a second DNA sequence encoding a reporter protein; a second promoter; and a second DBD-binding sequence, wherein said second DNA sequence is operably linked to said second promoter, said second promoter is operably linked to said second DBD-binding sequence and said effector protein binds to said second DBD-binding sequence via said DBD;

(c) administering said ligand to the transgenic animal generated in step (b) ; and

(d) analyzing the pharmacological property of said ligand by detecting the expression of said reporter gene in said transgenic animal .