WO1999018124A1

WO1999018124A1 - Assays for nuclear receptor ligands using fret

Info

Publication number: WO1999018124A1
Application number: PCT/US1998/021049
Authority: WO
Inventors: Richard T. Cummings; Jeffrey D. Hermes; David E. Moller; Gaochao Zhou
Original assignee: Merck & Co., Inc.
Priority date: 1997-10-07
Filing date: 1998-10-06
Publication date: 1999-04-15
Also published as: EP1021462A4; CA2305711A1; EP1021462A1; JP2001519525A

Abstract

Provided is a method of identifying agonists and antagonists of nuclear receptors that comprises measuring agonist-dependent fluorescence resonance energy transfer (FRET) between a fluorescently labeled nuclear receptor or ligand binding domain and fluorescently labeled CREB-binding protein (CBP), p300, other nuclear co-activator, or binding portion thereof. The method is simple, rapid, and inexpensive. Nuclear receptors and nuclear receptor co-activators labeled with fluorescent reagents for use in the above-described method are also provided.

Description

TITLE OF THE INVENTION

ASSAYS FOR NUCLEAR RECEPTOR LIGANDS USING FRET

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of U.S. Provisional Application No. 60/061,385, filed 10/7/97, the contents of which are incorporated herein by reference in their entirety.

STATEMENT REGARDING FEDERALLY-SPONSORED R&D Not applicable.

REFERENCE TO MICROFICHE APPENDIX Not applicable.

FIELD OF THE INVENTION

This invention relates to methods of identifying novel agonists and antagonists of nuclear receptors utilizing the agonist- dependent interaction of such receptors with CREB-binding protein

(CBP) or other nuclear receptor co-activators in which this interaction is detected by fluorescence resonance energy transfer.

BACKGROUND OF THE INVENTION Nuclear receptors are a superfamily of ligand-activated transcription factors that bind as homodimers or heterodimers to their cognate DNA elements in gene promoters. The superfamily, with more than 150 members, can be divided into subfamilies (e.g. the steroid, retinoid, thyroid hormone, and peroxisome proliferator-activated [PPAR] subfamilies). Each subfamily may consist of several members which are encoded by individual genes (e.g. PPARα, PPARγ, and PPARδ). In addition, alternative mRNA splicing can result in more than one isoform of these genes as in the case of specific PPARs (e.g. PPARγl and PPARγ2). The nuclear receptor superfamily is involved in a wide variety of physiological functions in mammalian cells: e.g., differentiation, proliferation, and metabolic homeostasis. Dysfunction or altered expression of specific nuclear receptors has been found to be involved in disease pathogenesis.

The PPAR subfamily of nuclear receptors consists of three members: PPARα, PPARγ, and PPARδ. PP.ARα is highly expressed in liver and kidney. Activation of PPARα by peroxisome proliferators (including hypolipidimic reagents such as fibrates) or medium and long-chain fatty acids is responsible for the induction of acyl-CoA oxidase and hydratase-dehydrogenase (enzymes required for peroxisomal β-oxidation), as well as cytochrome P450 4A6 (an enzyme required for fatty acid ω-hydroxylase). Thus, PPARα has an important role in the regulation of lipid metabolism and is part of the mechanism through which hypolipidimic compounds such as fibrates exert their effects. PPARγ is predominantly expressed in adipose tissue. Recently, a prostaglandin J2 metabolite, 15-Deoxy-D12,14-prostaglandin J2, has been identified as a potential physiological ligand of PPARγ. Both 15- Deoxy-D12,14-prostaglandin J2 treatment of preadipocytes or retroviral expression of PPARγ2 in fibroblasts induced adipocyte differentiation, demonstrating the role of PPARγ in adipocyte differentiation and lipid storage. The demonstration that anti-diabetic and lipid-lowering insulin sensitizing compounds known as thiazolidinediones are high affinity ligands for PPARγ suggests a broad therapeutic role for PPARγ ligands in the treatment of diabetes and disorders associated with insulin resistance (e.g. obesity and cardiovascular disease).

Nuclear receptor proteins contain a central DNA binding domain (DBD) and a COOH-terminal ligand binding domain (LBD). The DBD is composed of two highly conserved zinc fingers that target the receptor to specific promoter/enhancer DNA sequences known as hormone response elements (HREs). The LBD is about 200-300 amino acids in length and is less well conserved than the DBD. There are at least three functions for the LBD: dimerization, ligand binding, and transactivation. The transactivation function can be viewed as a molecular switch between a transcriptionally inactive and a transcriptionally active state of the receptor. Binding of a ligand which is an agonist flips the switch from the inactive state to the active state. The COOH-terminal portion of the LBD contains an activation function domain (AF2) that is required for the switch. The ligand-induced nuclear receptor molecular switch is mediated through interactions with members of a family of nuclear receptor co-activators (e.g., CBP/p300, SRC-l/NcoA-1, TIF2/GRIP- l/NcoA-2, and p/CIP). Upon binding of agonist to its cognate receptor LBD, a conformational change in the receptor protein creates a co- activator binding surface and results in recruitment of co-activator(s) to the receptor and subsequent transcriptional activation. The binding of antagonist ligands to nuclear receptors will not induce the required conformational change and prevents recruitment of co-activator and subsequent induction of transcription. The co-activators CREB-binding protein (CBP) and p300 are two closely related proteins that were originally discovered by virtue of their ability to interact with the transcription factor CREB. These two proteins share extensive amino acid sequence homology. CBP can form a bridge between nuclear receptors and the basic transcriptional machinery (Kamei et al., 1996, Cell 85:403-414; Chakravarti et al., 1996, Nature 383:99-103; Hanstein et al., 1996, Proc. Natl. Acad. Sci. USA 93:11540-11545; Heery et al., 1997, Nature 387:733-736). CBP also contains intrinsic histone acetyltransferase activity which could result in local chromatin rearrangement and further activation of transcription. Ligand- and AF2-dependent interaction between certain nuclear receptors and CBP has been demonstrated in in vitro pull down assays and far- western assays. This interaction is both necessary and sufficient for the transcriptional activation that is mediated by these nuclear receptors. Thus, an AF2 mutant of the estrogen receptor (ER) which abolishes the transcriptonal function of the receptor is incapable of interacting with CBP.

The N-termini of CBP and p300 have been shown to interact with the ligand-binding domains of some nuclear receptors (Kamei et al., 1996, Cell 85:403-414, hereinafter "Kamei"). Kamei was able to demonstrate direct interaction of CBP and p300 with nuclear receptors by several different methods:

(1) Kamei produced GST fusion proteins of the first 100 amino acids of the N-terminus of CBP. These fusion proteins were run out on a polyacrylamide gel, transferred to a membrane, and the membrane was exposed to 32p.l_abeled ligand-binding domains of nuclear receptors. In the presence of ligand, a specific binding interaction between the CBP and nuclear receptor fragments was detected in that the 32p-l_abeled ligand-binding domains were observed to bind to the bands on the membrane containing the GST-CBP fusion proteins.

(2) Kamei also utilized the yeast two-hybrid system. The ligand-binding domain of the nuclear receptor fused to the DNA-binding domain of the LexA protein was used as bait. The amino terminal domain of CBP fused to the gal4 transactivation domain was used as prey. In the presence of ligand, a specific binding interaction (occurring in vivo, i.e., within the yeast) was observed between the CBP and nuclear receptor fragments.

(3) Kamei observed ligand-induced binding between CBP and nuclear receptors via a gel-shift assay. This assay is based on the observation that, in the presence of ligand, nuclear receptors will bind to oligonucleotides containing their target recognition sequence. Such binding results in the formation of a nuclear receptor-ligand- oligonucleotide complex having a higher molecular weight than the oligonucleotide alone. This difference in molecular weight is detected via a shift in position of the 32p-l_abeled oligonucleotide when it is run out on a polyacrylamide gel. Kamei found that a fragment of CBP (the N- terminal 100 amino acids) was capable of binding to the nuclear receptor-ligand-oligonucleotide complex and shifting the complex's position on the gel to an even higher molecular weight. (4) Kamei was able to co-immunoprecipitate CBP using antibodies to nuclear receptors in extracts from a variety of cells in the presence of ligand.

(5) By the use of transcriptional activation assays, Kamei was able to demonstrate that nuclear receptors and CBP interact in a functional manner. Such transcriptional activation assays can indicate that two proteins are involved in a pathway that results in transcriptional activation but these assays do not prove that the interaction between the proteins is one of direct binding.

By the above-described methods, Kamei was able to demonstrate specific binding interactions between CBP and the retinoic acid receptor (RAR), glucocorticoid receptor (GR), thyroid hormone receptor (T3R), and retinoid X receptor (RXR). Kamei also demonstrated specific binding between the N-terminus of p300 and RAR. However, Kamei did not demonstrate specific binding between CBP, p300, or any other nuclear receptor co-activators and PPARs. What is striking about the methods used by Kamei is their extremely laborious and time consuming nature. Such methods involve, among other things, the construction of fusion proteins, the preparation of 32p-l_abeled proteins, the construction of specialized expression vectors for the yeast two-hybrid assay and the transcriptional activation assays, the running of many gels, and the raising of antibodies. Most of these assays take days to carry out and preparing the reagents needed to carry them out may take weeks. Because of the complicated reagents that are involved in these assays and the time needed to prepare and run the assays, these assays tend to be costly. Investigators other than Kamei who have studied the interaction between nuclear receptors and CBP have also been forced to rely on such cumbersome methods (see, e.g., Chakravarti et al., 1996, Nature 383:99- 103; Hanstein et al., 1996, Proc. Natl. Acad. Sci. USA 93:11540-11545; Heery et al., 1997, Nature 387:733-736). Kamei did not use the above-described methods to identify novel agonists or antagonists of nuclear receptors. The focus of Kamei was not on agonists or antagonists, but rather on the interaction between nuclear receptors and CBP. Although modifying the methods of Kamei to identify agonists or antagonists might be possible, such methods would suffer from serious disadvantages. This is because, as discussed above, all of the assays employed by Kamei to study the interaction of CBP and p300 with nuclear receptors are very laborious, slow, and costly. Given the therapeutic importance of steroid hormones such as estrogen, cortisol, progesterone, and other nuclear receptor agonists such as thyroid hormone and antidiabetic thiazolidinedione compounds, the need for improved high-throughput screening assays to identify potential pharmaceutical compounds affecting nuclear receptors is clear. Historically, therapeutically useful nuclear receptor ligand compounds were identified by screening animal models, an approach which is even more labor intensive and time consuming than the methods used by Kamei. Also, approaches such as those used by Kamei are ill-suited for the identification of antagonists of nuclear receptors. It is now widely appreciated that antagonists of nuclear receptors can be valuable therapeutic agents. Examples of such therapeutically useful antagonists are tamoxifene, raloxifene, and RU- 486.

What is needed is a high throughput, time and labor- saving, non-radioactive, inexpensive, and very reliable assay for the identification and characterization of both agonists and antagonists of nuclear receptors. Such an assay is provided by the present invention.

SUMMARY OF THE INVENTION

The present invention provides novel methods of identifying agonists and antagonists of nuclear receptors. The methods take advantage of the agonist-dependent binding of nuclear receptors and CBP, p300, or other nuclear receptor co-activators. In the absence of agonist, binding between the nuclear receptor and CBP, p300, or other nuclear receptor co-activators does not occur. If agonist is present, however, such binding occurs and can be detected by fluorescence resonance energy transfer (FRET) between a fluorescently-labeled nuclear receptor and fluorescently-labeled CBP, p300, or other nuclear receptor co-activator. Antagonists can be identified by virtue of their ability to prevent or disrupt the agonist-induced interaction of nuclear receptors and CBP, p300, or other nuclear receptor co-activators. In contrast to prior art methods of identifying agonists and antagonists of nuclear receptors, the methods of the present invention, are simple, rapid, and less costly.

The present invention provides a nuclear receptor or ligand binding domain thereof labeled with a fluorescent reagent for use in the above-described methods of identifying agonists and antagonists of nuclear receptors. The present invention also provides CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a fluorescent reagent. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates a method of fluorescently labelling a protein or polypeptide with Europium cryptate (Eu3+K).

Figure 2 illustrates the format for experiments 1 and 2 of

Table 1.

Figure 3 illustrates the format for experiment 3 of

Table 1. Figure 4 illustrates the format for experiment 4 of Table 1.

Figure 5 shows the results of studies using the methods of the present invention with four known PPARγ agonists. -o~ = AD5075; -D-- = Pioglitazone; -X- = Troglitazone; -0-- = BRL49653.

Figure 6 shows a measurement of the binding constant for the interaction between hCBP and PPARγlLBD.

Figure 7A shows the amino acid sequence of human CBP (SEQ.ID.NO.:l).

Figure 7B shows the nucleotide sequence of a cDNA encoding human CBP (SEQ.ID.NO.:2). The open reading frame is at positions 76-1290.

Figure 8A shows the amino acid sequence of human PPARα (SEQ.ID.NO.:3).

Figure 8B shows the nucleotide sequence of a cDNA encoding human PPARα (SEQ.ID.NO.:4). The open reading frame is at positions 217-1623.

Figure 9A shows the amino acid sequence of human PPARγl (SEQ.ID.NO.:5).

Figure 9B shows the nucleotide sequence of a cDNA encoding human PPARγl (SEQ.ID.NO.:6). The open reading frame is at positions 173-1609.

Figure 10A shows the amino acid sequence of human PPARδ (SEQ.ID.NO.:7).

Figure 10B-C shows the nucleotide sequence of a cDNA encoding human PPARδ (SEQ.ID.NO.:8). The open reading frame is at positions 338-1663. DETAILED DESCRIPTION OF THE INVENTION For the purposes of this invention:

- an "agonist" is a substance that binds to nuclear receptors in such a way that a specific binding interaction between the nuclear receptor and CBP or other nuclear receptor co-activator can occur.

- an "antagonist" is a substance that is capable of preventing or disrupting the agonist-induced specific binding interaction between a nuclear receptor and CBP, p300, or another nuclear receptor co- activator. - a "ligand" of a nuclear receptor is an agonist or an antagonist of the nuclear receptor.

- a "specific binding interaction," "specific binding," and the like, refers to binding between a nuclear receptor and CBP, p300, or other nuclear receptor co-activator which results in the occurrence of fluorescence resonance energy transfer between a fluorescent reagent bound to the nuclear receptor and a fluorescent reagent bound to CBP, p300, or other nuclear receptor co-activator.

With respect to CBP, p300, or other nuclear receptor co- activators, a "binding portion" is that portion of CBP, p300, or other nuclear receptor co-activators that is sufficient for specific binding interactions with nuclear receptors.

With respect to nuclear receptors, a "ligand binding domain" is that portion of a nuclear receptor that is sufficient to bind an agonist or antagonist of the nuclear receptor. The present invention provides a high throughput, time and labor-saving, non-radioactive, inexpensive, and very reliable assay for the identification and characterization of both agonists and antagonists of nuclear receptors. In a general embodiment, the present invention provides methods of identifying agonists and antagonists for any nuclear receptor for which CBP, p300, or another nuclear receptor binding protein is a co-activator. Such agonists and antagonists are identified by virtue of their ability to induce or prevent binding between the ligand binding domain of a nuclear receptor and CBP, p300, or other nuclear receptor co-activator. The interaction between the nuclear receptor and CBP, p300, or other nuclear receptor co-activator is monitored by observing the occurrence of fluorescence resonance energy transfer (FRET) between two fluorescent reagents. One fluorescent reagent is bound to the nuclear receptor; the other fluorescent reagent is bound to CBP, p300, or other nuclear receptor co-activator. The binding of fluorescent reagent to nuclear receptor, CBP, p300, or other nuclear receptor co-activator can be by a covalent linkage or a non-covalent linkage.

The present invention makes use of fluorescence resonance energy transfer (FRET). FRET is a process in which energy is transferred from an excited donor fluorescent reagent to an acceptor fluorescent reagent by means of intermolecular long-range dipole-dipole coupling. FRET typically occurs over distances of about lOe to lOOe and requires that the emission spectrum of the donor reagent and the absorbance spectrum of the acceptor reagent overlap adequately and that the quantum yield of the donor and the absorption coefficient of the acceptor be sufficiently high. In addition, the transition dipoles of the donor and acceptor fluorescent reagents must be properly oriented relative to one another. For a review of FRET and its applications to biological systems, see Clegg, 1995, Current Opinions in Biotechnology 6:103-110. The present invention makes use of a nuclear receptor or ligand binding domain thereof labeled with a first fluorescent reagent and CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a second fluorescent reagent. The second fluorescent reagent comprises a fluorophore capable of undergoing energy transfer by either (a) donating excited state energy to the first fluorescent reagent, or (b) accepting excited state energy from the first fluorescent reagent. In other words, according to the present invention, either the first or the second fluorescent reagents can be the donor or the acceptor during FRET. The first and second fluorescent reagents are spectropscopically complementary to each other. This means that their spectral characteristics are such that excited state energy transfer can occur between them. FRET is highly sensitive to the distance between the first and second fluorescent reagents. For example, FRET varies inversely with the sixth power of the d stance between the first and second fluorescent reagents. In the absence of agonist, the first fluorescent reagent, bound to the nuclear receptor or ligand binding domain thereof, will not be near the second fluorescent reagent, bound to CBP, p300, or other nuclear receptor co-activator, or binding portion thereof. Thus, no FRET, or very little FRET, will be observed. In the presence of agonist, however, interaction between the nuclear receptor and CBP, p300, or other nuclear receptor co-activator will occur, thus bringing close together the first and the second fluorescent reagents, allowing FRET to occur and be observed.

Accordingly, the present invention provides a method of identifying an agonist of a nuclear receptor that comprises providing:

(a) a nuclear receptor or ligand binding domain thereof labeled with a first fluorescent reagent;

(b) CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a second fluorescent reagent; and (c) a substance suspected of being an agonist of the nuclear receptor; under conditions such that, if the substance is an agonist of the nuclear receptor, binding between the nuclear receptor or ligand binding domain thereof and CBP, p300, or other nuclear receptor co- activator, or a binding portion thereof, will occur; and

(d) measuring fluorescence resonance energy transfer (FRET) between the first and second fluorescent reagents; where the occurrence of FRET indicates that the substance is an agonist of the nuclear receptor. In particular embodiments, the nuclear receptor is selected from the group consisting of steroid receptors, thyroid hormone receptors, retinoic acid receptors, peroxisome proliferator-activated receptors, retinoid X receptors, glucocorticoid receptors, vitamin D receptors, and "orphan nuclear receptors" such as LXR, FXR, etc. In a particular embodiment, the nuclear receptor or ligand binding domain thereof is a full-length nuclear receptor. In another embodiment, the nuclear receptor or ligand binding domain thereof is a ligand binding domain of a nuclear receptor. In another embodiment, the nuclear receptor or ligand binding domain thereof comprises an AF- 2 site of a nuclear receptor. In a particular embodiment, the nuclear receptor or ligand binding domain thereof is a full-length PPAR. In another embodiment, the nuclear receptor or ligand binding domain thereof is the ligand binding domain of a PPAR. In a further embodiment, the PPAR is selected from the group consisting of PPARα, PPARγl, PPARγ2, and PPARδ. In a further embodiment, the ligand binding domain of the PPAR contains amino acid residues 176-478 of human PPARγl.

In a particular embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 143-462 of human RARα. In another embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 122-410 of rat T/3Rαl. In another embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 227-463 of mouse RXRγ. In another embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 251-595 of human ER.

In a particluar embodiment, the above-described methods utilize full-length CBP, either mouse or human. In other embodiments, the methods utilize amino acid residues 1-113 of human CBP. In another embodiment, the methods utilize amino acid residues 1-453 of human CBP.

The conditions under which the methods described above are carried out are conditions that are typically used in the art for the study of protein-protein interactions: e.g., physiological pH; salt conditions such as those represented by such commonly used buffers as PBS; a temperature of about 4°C to about 55°C. The presence of commonly used non-ionic detergents, e.g., NP-40®, sarcosyl, Triton X- 100®, is optional. When europium cryptates are used as fluorescent reagents, reactions should contain KF at a concentration of at least 200 mM. Heery et al., 1997, Nature 387:733-736 showed that interactions between nuclear receptors and a variety of nuclear receptor co-activators are mediated by a short amino acid sequence in the nuclear receptor co-activators having the amino acid sequence LXXLL, where L is leucine and X represents any amino acid. Accordingly, the present invention can be practiced with a binding portion of a nuclear receptor co-activator, provided that the binding portion contains the amino acid sequence LXXLL. Therefore, the present invention includes a method of identifying an agonist of a nuclear receptor that comprises providing:

(a) a nuclear receptor or ligand binding domain thereof labeled with a first fluorescent reagent; (b) a binding portion of a nuclear receptor co-activator, where the binding portion contains the amino acid sequence LXXLL, and where the binding portion is labeled with a second fluorescent reagent; and

(c) a substance suspected of being an agonist of the nuclear receptor; under conditions such that, if the substance is an agonist of the nuclear receptor, binding between the nuclear receptor or ligand binding domain thereof and the binding portion of the nuclear receptor co-activator will take place; and (d) measuring fluorescence resonance energy transfer

(FRET) between the first and second fluorescent reagents; where the occurrence of FRET indicates that the substance is an agonist of the nuclear receptor.

In a particular embodiment, the nuclear receptor co- activator is selected from the group consisting of: human RIP- 140, human SRC-1, mouse TIF-2, human or mouse CBP, human or mouse p300, mouse TIF-1, and human TRIP proteins.

In a particular embodiment, the nuclear receptor co- activator is human RIP- 140 and the binding portion includes a contiguous stretch of amino acids of human RIP- 140 selected from the group consisting of: positions 20-29, 132-139, 184-192, 266-273, 379-387, 496-506, 712-719, 818-825, 935-944, and 935-942.

In another embodiment, the nuclear receptor co-activator is human SRC-1 and the binding portion includes a contiguous stretch of amino acids of human SRC-1 selected from the group consisting of: positions 45-53, 632-640, 689-696, 748-755, and 1434-1441.

In another embodiment, the nuclear receptor co-activator is mouse TIF-2 and the binding portion includes a contiguous stretch of amino acids of mouse TIF-2 selected from the group consisting of: positions 640-650, 689-699, and 744-754. In another embodiment, the nuclear receptor co-activator is human or mouse CBP and the binding portion includes a contiguous stretch of amino acids of human or mouse CBP selected from the group consisting of: positions 68-78 and 356-366. In another embodiment, the nuclear receptor co-activator is human or mouse p300 and the binding portion includes a contiguous stretch of amino acids of human or mouse p300 selected from the group consisting of: positions 80-90 and 341-351.

In another embodiment, the nuclear receptor co-activator is mouse TIF-1 and the binding portion includes a contiguous stretch of amino acids of mouse TIF-1 containing positions 722-732.

In another embodiment, the nuclear receptor co-activator is human TRIP2 and the binding portion includes a contiguous stretch of amino acids of human TRIP2 containing positions 23-33. In another embodiment, the nuclear receptor co-activator is human TRIP3 and the binding portion includes a contiguous stretch of amino acids of human TRIP3 containing positions 97-107.

In another embodiment, the nuclear receptor co-activator is human TRIP4 and the binding portion includes a contiguous stretch of amino acids of human TRIP4 containing positions 36-46.

In another embodiment, the nuclear receptor co-activator is human TRIP5 and the binding portion includes a contiguous stretch of amino acids of human TRIP5 containing positions 26-36.

In another embodiment, the nuclear receptor co-activator is human TRIP8 and the binding portion includes a contiguous stretch of amino acids of human TRIP8 containing positions 36-46.

In another embodiment, the nuclear receptor co-activator is human TRIP9 and the binding portion includes a contiguous stretch of amino acids of human TRIP9 selected from the group consisting of: positions 73-83, 256-266 and 288-298.

For amino acid sequences of nuclear receptor co-activators, see Yao et al., 1996, Proc. Natl. Acad. Sci. USA 93:10626-10631 (SRC-1); 0§ate et al., 1995, Science 270:1354-1357 (SRC-1); Cavailles et al., 1995, EMBO J. 14:3741-3751 (RIP-140); Voegel et al., 1996, EMBO J. 15:101-108 (TIF-2); Kwok et al., 1994, Nature 370:223-226 (CBP); Arias et al., 1994, Nature 370:226-229 (CBP); Eckner et al., 1994, Genes Dev. 8:869-884 (p300); Le Douarin et al., 1995, EMBO J. 14:2020-2033 (TIF-1); Lee et al., 1995, Nature 374:91-94 (TRIP proteins).

The particular embodiments of the present invention described above are all particular embodiments of a more general method that is also part of the present invention. That general method is a method of identifying an agonist of a nuclear receptor that comprises providing:

(a) a nuclear receptor or ligand binding domain thereof labeled with a first fluorescent reagent; Ob) a polypeptide containing the amino acid sequence

LXXLL where the polypeptide is labeled with a second fluorescent reagent; and

(c) a substance suspected of being an agonist of the nuclear receptor; under conditions such that, if the substance is an agonist of the nuclear receptor, binding between the nuclear receptor or ligand binding domain thereof and the polypeptide will take place; and

(d) measuring fluorescence resonance energy transfer (FRET) between the first and second fluorescent reagents; where the occurrence of FRET indicates that the substance is an agonist of the nuclear receptor.

In a particular embodiment, the amino acid sequence LXXLL is present in an α helical portion of the polypeptide. In another embodiment, the amino acid sequence LXXLL is present in an α helical portion of the polypeptide and the leucines form a hydrophobic face. The present invention provides methods for identifying antagonists of a nuclear receptor. Such methods are based on the ability of the antagonist to prevent the occurrence of agonist-induced binding between a nuclear receptor and CBP, p300, or other nuclear receptor co- activator, or to disrupt such binding after it has occurred. Thus, the present invention provides a method for identifying antagonists of nuclear receptors that comprises providing:

(a) a nuclear receptor or ligand binding domain thereof labeled with a first fluorescent reagent; (b) CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a second fluorescent reagent; (c) an agonist of the nuclear receptor; and

(d) a substance suspected of being an antagonist of the nuclear receptor; under conditions such that, in the absence of the substance, binding between the nuclear receptor or ligand binding domain thereof and CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof will occur; and

(e) measuring fluorescence resonance energy transfer (FRET) between the first and second fluorescent reagents when the substance is present and measuring FRET between the first and second fluorescent reagents when the substance is absent; where the a decrease in FRET when the substance is present indicates that the substance is an antagonist of the nuclear receptor. In particular embodiments, the nuclear receptor is selected from the group consisting of steroid receptors, thyroid hormone receptors, retinoic acid receptors, peroxisome proliferator-activated receptors, retinoid X receptors, glucocorticoid receptors, vitamin D receptors, and "orphan nuclear receptors" such as LXR, FXR, etc. In a particular embodiment, the nuclear receptor or ligand binding domain thereof is a full-length nuclear receptor. In another embodiment, the nuclear receptor or ligand binding domain thereof is a ligand binding domain of a nuclear receptor. In another embodiment, the nuclear receptor or ligand binding domain thereof is an AF-2 site of a nuclear receptor.

In a particular embodiment, the nuclear receptor or ligand binding domain thereof is a full-length PPAR. In another embodiment, the nuclear receptor or ligand binding domain thereof is the ligand binding domain of a PPAR. In a further embodiment, the PPAR is selected from the group consisting of PPARα, PPARγ, and PPARδ. In a further embodiment, the ligand binding domain of the PPAR contains amino acid residues 176-478 of human PPARγl.

In a particular embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 143-462 of human RARα. In another embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 122-410 of rat T₃Rαl. In another embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 227-463 of mouse RXRγ. In another embodiment, the nuclear receptor or ligand binding domain thereof contains amino acids 251-595 of human ER. In a particular embodiment, the above-described methods utilize full-length CBP, either mouse or human. In other embodiments, the methods utilize amino acid residues 1-113 of human CBP. In another embodiment, the methods utilize amino acid residues 1-453 of human CBP. The conditions under which the methods described above are carried out are conditions that are typically used in the art for the study of protein-protein interactions: e.g., physiological pH; salt conditions such as those represented by such commonly used buffers as PBS; a temperature of about 4°C to about 55°C. The presence of commonly used non-ionic detergents, e.g., NP-40®, sarcosyl, Triton X- 100®, is optional. When europium cryptates are used as fluorescent reagents, reactions should contain KF at a concentration of at least 200 mM.

In principle, one could measure FRET by monitoring either (a) a decrease in the emission of the donor fluorescent reagent following stimulation at the donor's absorption wavelength and/or (b) an increase in the emission of the acceptor reagent following stimulation at the donor's absorption wavelength. In practice, FRET is most effectively measured by emission ratioing. Emission ratioing monitors the change in the ratio of emission by the acceptor over emission by the donor. An increase in this ratio signifies that energy is being transferred from donor to acceptor and thus that FRET is occurring. Emission ratioing can be measured by employing a laser-scanning confocal microscope. Emission ratioing is preferably done by splitting the emitted light from a sample with a dichroic mirror and measuring two wavelength bands (corresponding to the donor and the acceptor emission wavelengths) simultaneously with two detectors. Alternatively, the emitted light can be sampled consecutively at each wavelength (by using appropriate filters) with a single detector. In any case, these and other methods of measuring FRET are well known in the art. Although a variety of donor and acceptor fluorescent reagents can be used in the practice of the present invention, preferred embodiments of the present invention make use of cryptates of fluorescent reagents as donor reagents. Inclusion of a substrate into the intramolecular cavity of a macropolycyclic ligand results in the formation of a cryptate. The macropolycyclic ligand shields the substrate from interaction with solvent and other solute molecules. If the substrate is a fluororescent reagent, formation of a cryptate may result in markedly different spectroscopic characteristics for the reagent as compared to the spectroscopic characteristics of the free reagent.

The present invention includes the use of europium (EuHI) or terbium (TbM) cryptates as donor fluorescent reagents. Such EuIH or TblH cryptates, as well as methods for their formation, are well known in the art. For example, see Alpha et al., 1987, Angew. Chem. Int. Ed. Engl. 26:266-267; Mathis, 1995, CHn. Chem. 41:1391-1397. A europium cryptate is formed by the inclusion of a europium ion into the intramolecular cavity of a macropolycyclic ligand which contains bipyridine groups as light absorbers. When europium cryptates are present in solution together with fluoride ions, a total shielding of the europium cryptate fluorescence is occurs. The molecular structure of a europium cryptate is shown below.

NH₂ NH₂

Europium cryptates can be conjugated to proteins by the use of well-known heterobifunctional reagents (see, e.g. , International Patent Application WO 89/05813; Prat et al., 1991, Anal. Biochem. 195:283-289; Lopez et al., 1993, Clin. Chem. 39:196-201). The present invention includes the use of XL665 as the acceptor fluorescent reagent. XL665 is a crosslinked derivative of allophycocyanin (APC). APC is a porphyrin containing protein which is derived from the light harvesting system of algae (Kronick, 1986, M. Immunol. Meth. 92:1-13). XL665 has an absorption maximum at =620 nm and an emission maximum at 665 nm. In some embodiments of the invention, XL665 is labeled with streptavidin in order to effect the binding of the streptavidin-labled XL665 to a biotin-labeled substance, e.g. , CBP or the ligand binding domain of a nuclear receptor. Streptavidin labeling of XL655 and biotin labeling of CBP, or the ligand binding domain of a nuclear receptor, can be performed by well known methods.

In a preferred embodiment of the invention, XL665 as the acceptor fluorescent reagent is combined with Europium cryptate (Eu3+K) as the donor fluorescent reagent. Europium cryptate (Eu3+K) has a large Stokes shift, absorbing light at 337 nm and emitting at 620 nm. Thus, the emission maximum of Europium cryptate (Eu3+K) overlaps the absorption maximum of XL665. Europium cryptate (Eu3+K) has a large temporal shift; the time between absorption and emission of a photon is about 1 millisecond. This is advantageous because most background fluorescence signals in biological samples are short-lived. Thus the use of a fluorescent reagent such as europium cryptate, with a long fluorescent lifetime, permits time-resolved detection resulting in the reduction of background interference.

The spectral and temporal properties of europium cryptate (Eu3+K) result in essentially no fluorescence background and thus assays using this fluorescent reagent can be carried out in a "mix and read" mode, greatly facilitating its use as a high throughput screening tool. For the embodiment using Europium cryptate (Eu3+K) and XL665, the measuring instrument irradiates the sample at 337 nm and measures the fluorescence output at two wavelengths, 620 nm (B counts, europium fluorescence) and 665 nm (A counts, XL665 fluorescence). The extent of flurorescent resonance energy transfer is measured as the ratio between these two values. Typically this ratio is multiplied by 10,000 to give whole numbers.

Other FRET donor-acceptor pairs are suitable for the practice of the present invention. For example, the following donor- acceptor pairs can be used: dansyl/fluorescein; fluorescein/rhodamine; tryptophan/aminocoumarin.

The present invention provides a nuclear receptor or ligand binding domain thereof labeled with a fluorescent reagent for use in the above-described methods of identifying agonists and antagonists of nuclear receptors. The present invention also provides CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a fluorescent reagent.

In a particular embodiment, the nuclear receptor or ligand binding domain thereof is selected from the group consisting of PPARα, PPARγ, PPARδ, a ligand binding domain of PPARα, PPARγ, or PPARδ, and amino acid residues 176-478 of human PPARγl and the fluorescent reagent is selected from the group consisting of XL665 and Europium cryptate (Eu3+K). In a particular embodiment, CBP, p300, or other nuclear receptor co-activator is labeled with a fluorescent reagent selected from the group consisting of XL665 and Europium cryptate (Eu3+K).

The following non-limiting examples are presented to better illustrate the invention.

EXAMPLE 1

Cloning, expression, and purification of human CBP and PPAR proteins

To test whether human CBP can interact with PPARs in an agonist-dependent manner, we cloned the human cDNA fragments encoding the NH2-terminal 1-113 amino acids (hCBPl-113) and 1-453 amino acids (hCBPl-453) of human CBP by the polymerase chain reaction (PCR). The DNA and amino acid sequences of human CBP are disclosed in Borrow et al., 1996, Nature Genet. 14:33-41 and in GenBank, accession no. U47741.

The primers used for hCBPl-113 were:

5'-ACTCGGATCCAAGCCATGGCTGAGAACTTGCTGGACGG-3' (SEQ.ID.NO.:9) and

5'-CACAAAGCTTAGGCCATGTTAGCACTGTTCGG-3' (SEQ.ID.NO.:

10).

These primers were expected to amplify a 0.9 kb DNA fragment. The primers for hCBPl-453 were: 5'-ACTCGGATCCAAGCCATGGCTGAGAACTTGCTGGACGG-3'

(SEQ.ID.NO.:9) and

5'CTCAGTCGACTTATTGAATTCCACTAGCTGGAGATCC-3'

(SEQ.ID.NO.:ll).

These primers were expected to amplify a 1.5 kb DNA fragment.. The template for the PCR reaction was a human fetal brain cDNA library (Stratagene, Catalogue #IS 937227). Of course, any human cDNA library from a tissue expressing CBP could have been used. The PCR amplified 0.9 kb and 1.5 kp DNA fragments which were digested with restriction endonucleases and ligated into pBluescript II vector. DNA sequencing analysis confirmed that the amplified fragments were identical to the corresponding published nucleic acid sequences of human CBP.

Based on the publicly available sequences for human CBP cited above, other primers could be readily identified and prepared by those skilled in the art in order to amplify and clone other portions of cDNA encoding human CBP from appropriate cDNA libraries. Once such portions of human CBP are produced, they could be used in the methods of the present invention in a manner similar to that described herein for hCBPl-113 and hCBPl-453. The amino acid sequence of human CBP is shown in Figure 7 A; the nucleic acid sequence of the cDNA encoding human CBP is shown in Figure 7B.

To express the polypeptides encoded by the PCR fragments, vectors encoding fusion proteins of the polypeptides and glutathione S- transferase (GST) were constructed and expressed in E. coli. The PCR fragments were subcloned into the expression vector pGEX (Pharmacia

Biotech) to generate pGEXhCBPl-113 and pGEXhCBP 1-453. pGEXhCBPl-113 and pGEXhCBPl-453 were transfected into the DH5α strain of E. coli (GIBCO BRL) and the bacteria hosting either pGEXhCBPl-113 or pGEXhCBPl-453 were cultured in LB medium (GIBCO BRL) to a density of OD600 = 0.7-1.0 and induced for overexpression of the GST-CBP fusion proteins by addition of IPTG (isopropylthio-β-galactoside) to a final concentration of 0.2 mM. The IPTG induced cultures were further grown at room temperature for 2-5 hrs. The cells were harvested by centrifugation for 10 min at 5000g. The cell pellet was used for GST-CBP fusion protein purification by following the procedure from Pharmacia Biotech using Glutathione Sepharose beads. hCBPl-113 and hCBPl-453 proteins were generated by cleaving the corresponding GST fusion proteins with thrombin. SDS- polyacrylamide gel electrophoresis analysis showed that the preparation from pGEXhCBPl-113 gave two polypeptide bands, with apparent molecular weight of 12 kd and 10 kd. The 12 kd band is the expected size of hCBPl-113 and the 10 kd band is most likely a premature translational termination product. The preparation from pGEXhCBP 1-450 gave a single band with the expected size, 50 kd. cDNAs encoding full-length PPARα and PPARγl were subcloned into pGEX vectors for the production of GST-PPARα and GST- PPARγl fusion proteins in E.coli. PPARγl was cloned from a human fat cell cDNA library (see Elbrecht et al., 1996, Biochem. Biophys. Res. Comm. 224:431-437). A cDNA encoding the human PPARγl ligand binding domain (PPARγlLBD; amino acids 176-478 of PPARγl) was subcloned from a modified pSG5 vector as a Xho I (site located in the N- terminus of the LBD)/ Xba I (site located in the pSG5 vector) fragment. The Xba I site was blunt-ended with T4 DNA polymerase. The 1.1 kb fragment containing the LBD was purified from an agarose gel and ligated into pGEX-KG (see Guan & Dixon, 1991, Anal. Biochem. 192:262- 267) that had been digested with Xho I and Hind III (the Hind III site had been blunt-ended with T4 DNA polymerase). This construct was used for the production of GST-hPPARγlLBD and hPPARγlLBD (the ligand binding domain cleaved free of GST). The overexpression and purification of PPARα, PPARγl, and PPARγlLBD were as described above for CBP. The DNA and amino acid sequences of human PPARα are disclosed in Schmidt et al., 1992, Mol. Endocrinol. 6:1634-1641 and in GenBank, accession no. L07592. See Figure 8A and 8B.

The DNA and amino acid sequences of human PPARγl are disclosed in Greene et al., 1995, Gene Expr. 4:281-299; Qi et al., 1995, Mol. Cell. Biol. 15:1817-1825; Elbrecht et al., 1996, Biochem. Biophys. Res. Comm. 224:431-437; and in GenBank, accession no. L40904. See Figure 9A and 9B. Human PPARγ2 contains the same amino acid sequence as human PPARγl except for an amino terminal addition of 24 amino acids (see Elbrecht et al., 1996, Biochem. Biophys. Res. Comm. 224:431-437). Thus, the amino acid sequence of the ligand binding domain of human PPARγ2 is the same as the amino acid sequence of the ligand binding domain of human PPARγl, although the numbering of the amino acids differs (176-478 for human PPARγl and 200-502 for human PPARγ2). The DNA and amino acid sequences of human PPARδ are disclosed in Sher et al., 1993, Biochemistry 32:5598-5604 and in GenBank, accession no. L02932. See Figure lOA-C.

EXAMPLE 2

Interaction between PPARs and hCBP fragments

Experiments were first conducted using hCBPl-113 and hPPARγlLBD. Purified hPPARγlLBD was biotinylated with Sulfo-NHS- LC-Biotin (PIERCE) to a biotin:hPPARγlLBD ratio of 3:1 according to the procedure provided by PIERCE. Purified hCBPl-113 was directly labeled with europium cryptate (Eu3+K) by the method illustrated in Figure 1. Biotin-labeled hPPARγlLBD, Eu3+K-labeled hCBPl-113, and streptavidin-labeled XL665 (SA-XL665; from PACKARD) were incubated together in the presence or absence of 1 μM of known PPARγ agonist (BRL49653 or AD5075). Thus, this experimental format made use of the fluorescent reagent pair europium cryptate (Eu3+K), which acted as donor, and

XL665, which acted as acceptor. hCBPl-113 was directly labeled with europium cryptate (Eu3+K); hPPARγlLBD was indirectly labeled with

XL665 by means of a biotin- streptavidin link. The emission maximum of europium cryptate (Eu3+K) overlaps with the absorption maximum of XL665. Therefore, when europium cryptate (Eu3+K) and XL665 are in close proximity, and the sample is illuminated with light at 337 nm (the absorption maximum of europium cryptate (Eu3+K)), FRET can occur between europium cryptate (Eu3+K) and XL665. This FRET manifests itself as increased emission at 665 nm by XL665. Figure 2 shows a schematic of the format used in this experiment (experiment 1 of Table 1). When agonist is bound to hPPARγlLBD, a specific interaction occurs between hPPARγlLBD and hCBPl-113, thus bringing europium cryptate (Eu3+K) and XL665 into close enough proximity for FRET to occur. In the absence of agonist, no interaction occurs between hPPiVRγlLBD and hCBPl-113 and thus europium cryptate (Eu3+K) and XL665 are not brought into close proximity and no FRET occurs. When FRET occurs, the amount of light given off by the sample at the emission maximum of XL665 (665 nm) is increased relative to the amount of fight given off by the sample at the emission maximum of europium cryptate (Eu3+K) (620 nm). Therefore, measuring the ratio of emission at 665 nm to 620 nm in the presence and the absence of a substance suspected of being an agonist allows for the determination of whether that substance actually is an agonist. If the substance is an agonist, an increase in the ratio of emission at 665 nm to 620 nm in the presence of the substance will be observed.

Reactions were carried out in microtiter plates. Reaction conditions were: appropriate volume (total 250 μl) of the reaction buffer (either PBS or HEPES, see below, containing 500 mM KF, 0.1% bovine serum albumin, BSA) was added to each well, followed by addition of ligands (BRL49653 or AD5075 at a final concentration of 1 μM and 0.1% dimethylsulfoxide (DMSO) or vehicle control (0.1% DMSO), Eu3+K labeled hCBP (100 nM), biotin-hPPARγlLBD (100 nM), and streptavidin- labeled XL665 (100 nM) to appropriate wells. After mixing, 200 μl of reaction mixture was transferred to a new well. The plate was either directly measured for fluorescence resonance energy transfer (FRET) or covered with sealing tape (PACKARD) to avoid evaporation and incubated at room temperature for up to 24 hrs before measuring FRET. The results of this experiment and others described below yielded ratio values as follows: Table 1

Experiment 1 of Table 1 was carried out using PBS (137 mM NaCl, 2.7 mM KC1, 4.3 mM Na2HPθ4, 1.4 mM KH2PO4, pH 7.4). The greater emission ratio observed in the presence of .AD5075 demonstrated that a specific interaction between hCBPl-113 and hPPARγlLBD occurred in the presence of the agonist AD5075. Although it was clear that FRET was occurring, the signal-noise ratio was small. In experiment 2 of Table 1, HEPES buffer (N-2-hydroxyethylpiperazine-N'- 2-ethane sulfonic acid, 100 mM, pH 7.0) containing 0.05% NP40 (Nonidet P-40) was used instead of PBS and an improved signal-noise ratio was obtained.

In order to get an even better signal-noise ratio, the above- described format was modified slightly for experiment 3. In experiment 3, SA-XL665 (500 nM), biotin-labeled hPPARγlLBD (100 nM), GST- hCBPl-113, and Eu3+K labeled anti-GST antibody (2.5 μl) were incubated in the presence or absence of AD5075 (1 μM) in HEPES buffer containing 0.05% NP40. A two-fold signal- noise ratio was obtained. Figure 3 shows a schematic of the format used in experiment 3.

The anti-GST antibody was a goat antibody to GST from Pharmacia (catalogue number 27-4577-01) that was labeled with Eu3+K according to the procedure summarized below.

- Make up @ 10 mg/mL in H2O. Need 42.2 μg (4.2 μL, 96.6 nmol) for 49.0 μg Eu3+ Reagent

-Resuspended @ -FW = 1465 ²-^{9 E(}l^uiv SULFO-SMCC,

2.5 mg/mL in 10% Use 49.0 μg 20 mM Pi buffer, 10% DMF

DMF/PBS (19.6 μL, 33.4 nmol) RT, 30 minutes

From Pharmacia, 5.0 mg/mL, 5.0 Equiv SPDP, FW =312, Dissolve FW = 150 kD Use 200 μL (1 mg, RT, 5 hours @ 1.00 mg/mL in EtOH. 6.66 nmol) exchange into 10 mM Add 10.4 μL (5 equiv., Borate, 350 mM NaCl, 10% Gly, 10.4 μg, 33.4 nmol) to pH 8.5 with BioSpin-30 protein.

To further improve the signal to noise ratio, a series of experiments were conducted. Experiment 4 of Table 1 exemplifies results obtained from those efforts. cDNA encoding a longer fragment of hCBP was cloned and expressed to get hCBPl-453. hCBPl-453 was biotinylated. Biotin-labeled hCBPl-453 (25 nM), SA-XL665 (100 nM), GST-hPPARγlLBD (1 nM), and Eu3+K-labeled anti-GST antibody (2 nM) were mixed together in the presence or absence of 1 μM AD5075. The detergent was changed from 0.05% NP40 to 0.5% CHAPS (3-{[3- cholamidopropyl]dimethyl-ammoniol}-l-propanesulfonate). A three- to four-fold signal-noise ratio was obtained. Figure 4 shows the strategy used for experiment 4 and similar experiments.

The correlation between results from the above-described assays and previously reported results from in vitro binding and transcriptional activation assays of selected antidiabetic insulin sensitizers that are known to be PPARγ agonists (Elbrecht et al., 1996, Biochem Biophys Res Comm 224:431-437) was analyzed by titrating those known PPARγ agonists in the assays described above and comparing EC50S so obtained with previously described values for potency in binding or transcriptional activation assays for the known agonists. The results are shown in Figure 5. From Figure 5, the following EC50s can be derived:

AD5075 = 8 nM BRL49653 = 53 nM Troglitazone = 646 nM

Pioglitazone = 890 nM. These EC50s generated in the above-described assays are in close agreement with those generated by in vitro binding and transcriptional activation studies (Elbrecht et al., 1996, Biochem Biophys Res Comm 224:431-437).

The above-described assay can also be used to characterize the interaction between nuclear receptors with co-activators as, e.g. , by determining the binding constant for that interaction. Figure 6 shows an example of such an application. Saturating amounts of PPARγ agonist (10 μM BRL49653) were used. Increasing concentrations of non- biotinylated hCBPl-453 were used to titrate away biotin-hCBP- PPARγlLBD complex and decrease the fluorescence energy transfer. A Kd of 300 nM for the interaction between hCBPl-453 and PPARγlLBD can be derived from the results illustrated in Figure 6 and this Kd (300 nM) is a measurement ofthe affinity between CBP and PPARγ.

The present invention is not to be limited in scope by the specific embodiments described herein. Indeed, various modifications of the invention in addition to those described herein will become apparent to those skilled in the art from the foregoing description. Such modifications are intended to fall within the scope of the appended claims. Various publications are cited herein, the disclosures of which are incorporated by reference in their entireties.

Claims

WHAT IS CLAIMED:

1. A method of identifying an agonist of a nuclear receptor that comprises providing:

(b) CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a second fluorescent reagent; and

(c) a substance suspected of being an agonist of the nuclear receptor; under conditions such that, if the substance is an agonist of the nuclear receptor, binding between the nuclear receptor or ligand binding domain thereof and CBP, p300, or other nuclear receptor co- activator, or a binding portion thereof, will occur; and

2. The method of claim 1 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of steroid receptors, thyroid hormone receptors, retinoic acid receptors, peroxisome proliferator-activated receptors, retinoid X receptors, glucocorticoid receptors, vitamin D receptors, LXR, and FXR.

3. The method of claim 1 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of a full-length nuclear receptor, a ligand binding domain of a nuclear receptor, and an AF-2 site of a nuclear receptor.

4. The method of claim 1 where the nuclear receptor or ligand binding domain thereof comprises an AF-2 site of a nuclear receptor.

5. The method of claim 1 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of a full-length PPAR, a ligand binding domain of a PP.AR, and amino acid residues 176-478 of human PPAR╬│l.

6. The method of claim 1 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of

PPAR╬▒, PPAR╬│l, PPAR╬│2, and PPAR╬┤.

7. The method of claim 1 where the nuclear receptor or ligand binding domain thereof comprises a ligand binding domain selected from the group consisting of amino acids 143-462 of human RAR╬▒, amino acids 122-410 of rat T3R╬▒l, amino acids 227-463 of mouse

RXR╬│, and amino acids 251-595 of human ER.

8. The method of claim 1 where CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof is selected from the group consisting of full-length human CBP, full-length mouse CBP, amino acid residues 1-113 of human CBP, and amino acid residues 1-453 of human CBP.

9. The method of claim 1 where the first fluorescent reagent is selected from the group consisting of XL665 and Europium cryptate (Eu3+K).

10. The method of claim 1 where the second fluorescent reagent is selected from the group consisting of XL665 and Europium cryptate (Eu3+K).

11. A method of identifying an agonist of a nuclear receptor that comprises providing: (a) a nuclear receptor or ligand binding domain thereof labeled with a first fluorescent reagent;

(b) a binding portion of a nuclear receptor co-activator, where the binding portion contains the amino acid sequence LXXLL, and where the binding portion is labeled with a second fluorescent reagent; and (c) a substance suspected of being an agonist of the nuclear receptor; under conditions such that, if the substance is an agonist of the nuclear receptor, binding between the nuclear receptor or ligand binding domain thereof and the binding portion of the nuclear receptor co-activator will take place; and

12. The method of claim 11 where the binding portion of a nuclear receptor co-activator is selected from the group consisting of human RIP-140, human SRC-1, mouse TIF-2, human or mouse CBP, human or mouse p300, mouse TIF-1, and human TRIP proteins.

13. A method of identifying an agonist of a nuclear receptor that comprises providing:

(b) a polypeptide containing the amino acid sequence LXXLL where the polypeptide is labeled with a second fluorescent reagent; and

(d) measuring fluorescent resonance energy transfer (FRET) between the first and second fluorescent reagents; where the occurrence of FRET indicates that the substance is an agonist of the nuclear receptor.

14. A method for identifying an antagonist of a nuclear receptor that comprises providing: (a) a nuclear receptor or ligand binding domain thereof labeled with a first fluorescent reagent;

(b) CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a second fluorescent reagent; (c) an agonist of the nuclear receptor; and

(e) measuring fluorescence resonance energy transfer (FRET) between the first and second fluorescent reagents when the substance is present and measuring FRET between the first and second fluorescent reagents when the substance is absent; where the a decrease in FRET when the substance is present indicates that the substance is an antagonist of the nuclear receptor.

15. The method of claim 14 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of steroid receptors, thyroid hormone receptors, retinoic acid receptors, peroxisome proliferator-activated receptors, retinoid X receptors, glucocorticoid receptors, vitamin D receptors, LXR, and FXR.

16. The method of claim 14 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of a full-length nuclear receptor, a ligand binding domain of a nuclear receptor, and an AF-2 site of a nuclear receptor.

17. The method of claim 14 where the nuclear receptor or ligand binding domain thereof comprises an AF-2 site of a nuclear receptor.

18. The method of claim 14 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of a full-length PPAR, a ligand binding domain of a PPAR, and amino acid residues 176-478 of human PPAR╬│l.

19. The method of claim 14 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of

PPAR╬▒, PPAR╬│l, PPAR╬│2, and PPAR╬┤.

20. The method of claim 14 where the nuclear receptor or ligand binding domain thereof comprises a ligand binding domain selected from the group consisting of amino acids 143-462 of human

RAR╬▒, amino acids 122-410 of rat T/3R╬▒l, amino acids 227-463 of mouse

RXR╬│, and amino acids 251-595 of human ER.

21. The method of claim 14 where CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof is selected from the group consisting of full-length CBP, amino acid residues 1-113 of human CBP, and amino acid residues 1-453 of human CBP.

22. The method of claim 14 where the first fluorescent reagent is selected from the group consisting of XL665 and Europium cryptate (Eu3+K).

23. The method of claim 14 where the second fluorescent reagent is selected from the group consisting of XL665 and Europium cryptate (Eu3+K).

24. A nuclear receptor or ligand binding domain thereof labeled with a fluorescent reagent.

25. The nuclear receptor or ligand binding domain thereof of claim 24 where the nuclear receptor or ligand binding domain thereof is selected from the group consisting of PPAR╬▒, PPAR╬│l, PPAR╬│2, PPAR╬┤, a ligand binding domain of PPAR╬▒, PPAR╬│l, PPAR╬│2, or PPAR╬┤, and amino acid residues 176-478 of human PP.AR╬│l and the fluorescent reagent is selected from the group consisting of XL665 and Europium cryptate (Eu3+K).

26. CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, labeled with a fluorescent reagent.

27. The CBP, p300, or other nuclear receptor co-activator, or a binding portion thereof, of claim 26 where the fluorescent reagent is selected from the group consisting of XL665 and Europium cryptate (Eu3+K).