WO2004034288A1

WO2004034288A1 - The three-dimensional crystal structure of bcatc complexes and methods of use thereof

Info

Publication number: WO2004034288A1
Application number: PCT/IB2003/004301
Authority: WO
Inventors: John Ronald Rubin; Lain-Yen Hu; Anil Mistry; Jeffrey David Scholten; David Winslow Moreland; William Thomas Mueller; Patrick Charles Mcconnell
Original assignee: Warner-Lambert Company Llc
Priority date: 2002-10-09
Filing date: 2003-09-29
Publication date: 2004-04-22
Also published as: AU2003263540A1

Abstract

The three-dimensional crystal structures of BCATc peptide: co-factor and BCATc peptide: co-factor: ligand complexes, and methods for obtaining the crystals, are described. The structures provide three-dimensional descriptions of BCATc and its binding sites and can be used, for example, in drug discovery and design.

Description

THE THREE-DIMENSIONAL CRYSTAL STRUCTURE OF BCATc COMPLEXES AND METHODS OF USE THEREOF TECHNICAL FIELD AND INDUSTRIAL APPLICABILITY OF THE INVENTION

The present invention relates to the crystallization and resultant crystal structure, of mammalian, e.g., human, cytosolic branched-chain amino acid aminotransferase (BCATc) in complex with a cofactor, such as covalently linked pyridoxal-5'-phosphate co-factor (PLP), or with a cofactor, such as PLP, and a ligand. This invention further relates to the three- dimensional structural information derived from the BCATc complexes and the use of that three-dimensional structural information, for example, to screen for, identify, design, modify,. and/or evaluate potential BCATc peptide activity inhibitors and/or enhancers, i.e., ligands, for therapeutic use in treating, for example, neurological diseases.

REFERENCE TO TABLES 1-3 SUBMITTED ON COMPACT DISC The atomic coordinates in Tables 1-3 are submitted herewith on duplicate compact discs. The material on the compact discs is incorporated by reference herein. The files on the compact discs containing the atomic coordinates of Tables 1-3 are labeled Table 1, Table 2 and Table 3, respectively.

BACKGROUND OF THE INVENTION

The neurotransmitter glutamate is produced by two pathways: (1) reversible transamination; and (2) oxidative deamination. In general, during reversible transamination, the main pathway, branched-chain amino acid aminotransferases (BCATs) catalyze the transfer of an amino group from the branched-chain amino acids (BCAAs), i.e., isoleucine, leucine and valine, to the substrate alpha-ketoglutarate (AKG), resulting in the production of glutamate. More specifically, the proposed enzyme mechanism involves the formation of a pyridoxamine intermediate from a BCAA, followed by the transfer of an amino group from the pyridoxamine intermediate to AKG resulting in glutamic acid formation. The BCAA leucine is the source of approximately 25% of the nitrogen used in glutamate synthesis in the central nervous system (CNS).

The BCAT employed in reversible transamination is either BCATm (mitochondrial) or BCATc (cytosolic or cytoplasmic) depending upon the location of glutamate synthesis as the two BCAT isoforms work in tandem in different cell types to generate glutamate. For example, the major source of glutamate in glutamatergic neurons is derived from glutamate biosynthesized in astrocytes wherein BCATm is located. However, BCATc is responsible for the amino group transfer in neurons. BCATm is a ubiquitous monomeric enzyme, found in nearly all tissues, whereas dimeric BCATc is found predominantly in the brain (Hutson, S.M. et al., Journal of Biological Chemistry 270: 30344-30352 (1995)).

BCATc expression is highly regulated such that inhibition of BCATc activity may slow glutamate synthesis which, in turn, would decrease the amount of glutamate released during the excitation of neurons. The ability to manipulate the activity of BCATc and thus, glutamate synthesis, is believed to be of therapeutic benefit in neurological disorders such as diabetic retinopathy and Alzheimer's Disease (AD), as well as in certain behavioral' disorders which rely on glutamatergic mechanisms (Beal, M.F., Current Opinions in Neurobiology 2: 657 (1992)). This is illustrated by the competitive inhibition of BCATc by the anti-convulsant drug gabapentin (Hutson, S.M. et al., J. Neurochem. 71: 863-874 (1998)).

BCATs are found in nearly all organisms including bacteria. For example, the crystal structure of E. coli BCAT (eBCAT) is described by Okada et al. in the Journal of Biochemistry 121: 637-641 (1997). In addition, the eBCAT three-dimensional structure is described by Okada, K. et al. in Biochemistry 40: 7453-7463 (2001). Okada et al. set forth the three- dimensional structure of three forms of eBCAT and indicate that eBCAT is a hexamer containing three dimer units surrounding a 3-fold axis, wherein the dimer units have a molecular weight of approximately 31 ,500 Da and contain 308 amino acid residues. Similarly, Hutson et al. report that the S. typhimurium BCAT is also a hexamer and contains identical 33.9 kDa subunits (Hutson, S.M. et al., Journal of Biological Chemistry 270: 30344-30352 (1995)). The structure of the H. influenzae BCAT is not known, but it is reported to share various properties of the other isolated bacterial BCATs, e.g., single isoenzyme (Hutson, S.M. et al., Journal of Biological Chemistry 270: 30344-30352 (1995)).

In the eukaryotic group, BCATs have been cloned from humans, rats, mice and sheep. Such cloning has confirmed the placement of mammalian BCATs in the same folding group as bacterial BCATs (Hutson, S.M. et al., Journal of Biological Chemistry 270: 30344- 30352 (1995)). However, mammalian BCATs have acquired additional amino acids during evolution which is a structural change that has not been observed in other aminotransferase subclasses (Hutson, S.M. et al., Journal of Biological Chemistry 270: 30344-30352 (1995)). Due to mammalian BCAT cloning, it is now known that human BCATc is a dimeric enzyme, wherein each monomer contains 384 amino acid residues (Davoodi, J. et al., Journal of Biological Chemistry 273: 4982-4989 (1998)). The dimer molecular weight is approximately 86,000 Da since the monomers have a molecular weight of approximately 42,806 Da each. It is further known that rat BCATc, described by Hutson, S.M. et al. in the Journal of Biological Chemistry 270(51): 30344-30352 (1995), is a 410-amino acid protein with a molecular weight of 46,045 Da; and, similarly, murine BCATc, described by Niwa, O. et al., Nucleic Acids Research 18: 6709 (1990), has a molecular weight of 43.6 Da.

Bonfils, J. et al., Biochimca et Biαphysica Ada 1494: 129-136 (2000), have predicted the structure of sheep BCATc using a three-dimensional computational model. According to Bonfils et al., sheep BCATc cDNA encodes 385 amino acids. Bonfils et al. further report that sheep BCATc is a dimeric enzyme having a molecular weight of approximately 86 kDa (each monomer has a molecular weight of approximately 43 kDa). Bonfils et al. note that sheep BCATc cDNA exhibits 87% identity to human BCATc cDNA and the two amino acid sequences are 90% identical. However, the two BCAT species are distinct. For example, sheep BCATc likely has a different function than human BCATc due to its ubiquitous property. Further, Bonfils et al. state that sheep BCATc is found in muscle which suggests that it possesses a unique role in sheep BCAA metabolism, whereas human BCATc, expressed at high levels in the CNS, has been implicated in neurological diseases and associated with neurological pain. Bonfils et al. have predicted the BCATc three-dimensional structure, but have not produced the structure. In fact, there is no known actual mammalian BCATc three- dimensional structure. The crystal structures of human BCATm and human D-amino acid aminotransferase are known (Yennawar et al., Ada Cryst. D57: 506-515 (2001) and Sugio et al., Biochemistry 34: 9661-9669 (1995), respectively), but these two proteins, along with eBCAT, although structurally related to human BCATc, possess different amino acid sequences, different cellular origins and different functions.

Thus, although various mammalian BCATc DNA and amino acid sequences are known, one mammalian three-dimensional model has been predicted and related protein crystal structures are known, this knowledge is not sufficient to screen for, identify, design, modify and/or evaluate inhibitors and/or enhancers of BCATc activity. It is the three- dimensional crystal structure (as defined by atomic coordinates) obtained from an electron density map produced by X-ray crystallography data) which provides a basis for the identification, design, modification and, ultimately, evaluation of BCATc activity inhibitors and enhancers via computer modeling by illustrating the BCATc binding sites and catalytic regions. Such inhibitors and enhancers could inhibit or enhance, respectively, glutamatergic mechanisms, and thus, could be used as drugs to treat conditions involving such mechanisms. Therefore, there is presently a need for obtaining a form of BCATc that can be crystallized with or without a ligand (such as a BCATc activity inhibitor or enhancer) to form crystals of sufficient quality to allow X-ray diffraction data and thus, atomic coordinates, to be obtained. There is also a need for determining the three-dimensional structures of such crystals to use in, for example, structural-based drug screening, identification, design and/or modification and/or drug assessment. See, e.g., Bugg et al., Scientific American Dec: 92-98 (1993); West et al., TiPS 16: 67-74 (1995); Dunbrack et al., Folding & Design 2: 27-42 (1997).

SUMMARY OF THE INVENTION The present invention relates to crystalline structures of a mammalian, e.g., human, BCATc peptide and peptides that are structurally related to the BCATc peptide, including crystalline structures of a human BCATc peptide in complex with its covalently linked co- factor, PLP, with and without a ligand, and methods for crystallizing the crystalline structures. It has been discovered that human BCATc comprises a pyridoxal-5'-phosphate (PLP) binding pocket, a catalytic binding pocket, an alpha-ketoglutarate (AKG) binding pocket, and a channel between the catalytic binding pocket and the AKG binding pocket, as more fully defined below. Thus, the invention also relates to these binding pockets and to crystal structures of BCATc peptide or a structurally related peptide (defined below) comprising these binding pockets. Additionally, the invention includes binding pockets that are defined by the atoms found in the atomic coordinates as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A (most preferably 1.25 A) from the binding pocket C alpha atoms of any one of the BCATc binding pockets defined below, or a conservatively substituted variant thereof, and crystal structures of BCATc peptide or a structurally related peptide comprising these binding pockets.

Specifically, it has been discovered that BCATc comprises a PLP binding pocket that has a rectangular shape, is at the interface of N-terminal and C-terminal domains of the BCATc peptide, and has the dimensions of about 14 A x about 10 A x about 6 A. It has also been discovered that the PLP binding pocket is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co-factor located in the PLP binding pocket: thr258, met259, asn260, gly330, Iys220, arg117, val287, thr288, thr331 , tyr225, glu255, and ser329. Thus, the invention relates to crystal structures of a BCATc peptide or structurally related peptide comprising the PLP binding pocket or a PLP binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A (most preferably 1.25 A) from the C alpha atoms of the PLP binding pocket. It has also been discovered that BCATc comprises a catalytic binding pocket having dimensions of about 14 A x about 10 A x about 19 A and that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co-factor located in the catalytic binding pocket: (a) amino acid residues Ieu45, val46, phe47, gly48, thr49, phe93, glu94, gly95, Ieu96, Iys97, tyr159, arg161 , vahδδ, gly189, pro190, tyr191 , phe192, ser193, val200, ser201 , Ieu202, arg210, asn224, ser228, Ieu229, gln232, val236, cys240, gln241 , gln242, val243, trp245, val256, gly257, ser329, ala332, cys333 and cys336 from a first BCATc peptide monomer; and (b) amino acid residues ser84, his87, tyrδδ, glu168, ser170, Ieu171 , gly172, val173, Iys174 and Iys175 from a second BCATc peptide monomer. Thus, the invention relates to crystal structures of a BCATc peptide or structurally related peptide comprising the catalytic binding pocket or a catalytic binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A (most preferably 1.25 A) from the C alpha atoms of the catalytic binding pocket.

Additionally, it has been discovered that BCATc comprises an AKG binding pocket that is rectangular in shape, has dimensions of about 18 A x about 7 A, and is defined by the atomic coordinates of the following amino acid residues within about 5 A of a substrate located in the AKG binding pocket: Ieu96, phe99, phe109, val287, thr288 and val334 of the peptide. Thus, the invention relates to crystal structures of a BCATc peptide or structurally related peptide comprising the AKG binding pocket or an AKG binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A (most preferably 1.25 A) from the C alpha atoms of the AKG binding pocket.

Additionally, it has been discovered that BCATc includes an AKG binding pocket having a channel that is defined by the atomic coordinates of the following amino acid residues within about 5 A of the channel: phe47, gly95, gln110, pro111, asn112, Ieu113, ser155, ser157, val188, pro190, tyr191 , phe192, cys291 , gly330, ala332, cys333, Ieu364, Ieu367, thr368 and tyr372 of the BCATc peptide. Thus, the invention also relates to crystal structures of a BCATc peptide or structurally related peptide comprising the AKG binding pocket channel or an AKG binding pocket channel that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A (most preferably 1.25 A) from the C alpha atoms of the AKG binding pocket channel.

It has also been discovered that the BCATc peptide comprises a channel between the catalytic binding pocket and the AKG binding pocket, and that the channel is defined by the atomic coordinates of the following amino acid residues within about 5 A of the channel: (a) arg210, asn224, tyr225, ser228, Ieu229, gln232, trp245, gly257 and thr258 from a first peptide monomer; and (b) val173 from a second peptide monomer. Thus, the invention relates to crystal structures of a BCATc peptide or structurally related peptide comprising the channel between the catalytic binding pocket and the AKG binding pocket or a channel that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A (most preferably 1.25 A) from the C alpha atoms of the channel. The invention further relates to the three-dimensional structural coordinates, e.g, atomic coordinates, obtained from an electron density map produced by X-ray diffraction data of the BCATc crystalline structures. These atomic coordinates are set forth in Table 1 , Table 2, and Table 3. Related sets of structural coordinates having a root mean square deviation from the core C alpha atoms of the atomic coordinates set forth in Table 1 , Table 2, or Table 3 of from not more than about 1.5 A to not more than about 0.50 A are also a part of the invention. Preferably the root mean square deviation of the structural coordinates is not more than about 0.50A, more preferably not more than about 0.75 A, even more preferably not more than about 1.00 A, and most preferably not more than about 1.25 A. The structural coordinates reflect the three-dimensional structure of the BCATc peptide complex and illustrate, to atomic resolution, the chemical environment around the BCATc binding sites.

The invention also relates to machine-readable medium having stored thereon (1) data comprising the atomic coordinates as set forth in Table 1 , Table 2, or Table 3, or a related set of atomic coordinates; or (2) data comprising a BCATc or BCATc-like binding pocket (defined below).

Additionally, the invention relates to the use of the coordinates set forth in Table 1 , Table 2, and Table 3 in methods for generating a three-dimensional representation of a human BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket. The three-dimensional representations are generated by applying the atomic coordinates set forth in Table 1 , Table 2, or Table 3, or a related set of atomic coordinates (defined below), to a computer algorithm to generate a three-dimensional representation of the peptide or peptide binding pocket. The invention also relates to the use of these three-dimensional representations in, for example, drug discovery and design.

Specifically, the invention provides a method for modifying a chemical entity that includes: (a) generating the three-dimensional computer representation of BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket; (b) modeling the chemical entity based on the three-dimensional representation; and (c) modifying the chemical entity to improve its ability to associate with the peptide or peptide binding pocket.

The invention also includes a method for designing a chemical entity comprising: (a) generating the three-dimensional computer representation of BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket; (b) generating a chemical entity that spatially conforms to the three-dimensional representation of the peptide or the peptide binding pocket; and (c) evaluating whether the chemical entity has the potential to associate with the peptide or peptide binding pocket.

The invention also relates to the use of the three-dimensional representations of BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket in a method for screening and identifying a potential inhibitor or enhancer of the activity of a BCATc peptide or a structurally related peptide. One such method for screening and identifying a potential inhibitor or enhancer of the activity of BCATc peptide or a structurally related peptide comprises: (a) generating the three-dimensional representation of BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket; (b) applying an iterative process whereby a chemical entity is applied to the three- dimensional representation to determine whether the chemical entity associates with the peptide; and (c) evaluating the effect(s) of the chemical entity on peptide activity to determine whether the chemical entity functions as an activity inhibitor or enhancer.

Another such method for screening and identifying a potential inhibitor or enhancer of the activity of BCATc peptide or a structurally related peptide comprises: (a) generating a three-dimensional representation of a BCATc or BCATc-like PLP, catalytic, or AKG binding pocket, (b) generating a potential inhibitor or enhancer by (i) assembling molecular fragments into a chemical entity; (ii) de novo design of a chemical entity; (iii) selecting a chemical entity from a small molecule database; or (iv) modifying a known chemical entity; and (c) evaluating by computer modeling whether the potential inhibitor or enhancer associates with the binding pocket. Alternatively, the invention relates to a method for screening and identifying a potential inhibitor or enhancer of the activity of a human BCATc peptide or a structurally related peptide comprising: (a) generating a three-dimensional representation of a BCATc or BCATc-like PLP, catalytic, or AKG binding pocket; (b) generating a chemical entity that spatially conforms to the binding cavity, wherein the chemical entity is generated by (i) assembling molecular fragments into the chemical entity; (ii) de novo design of the chemical entity; (iii) selecting the chemical entity from a small molecule database; or (iv) modifying a known inhibitor or enhancer, or portion thereof, of BCATc activity; (c) synthesizing the chemical entity or analogs thereof; and (d) evaluating whether the chemical entity associates with the binding pocket. The method may further comprise (e) growing a crystal comprising the peptide and the chemical entity; and (f) determining the three-dimensional structure of the crystal using molecular replacement.

The invention further includes the use of the three-dimensional representations of BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket in methods for evaluating the potential of a chemical entity to associate with a human BCATc peptide or structurally related peptide. One such method comprises: (a) generating the three dimensional representation of BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket; (b) applying a chemical entity (e.g. a three- dimensional representation of the chemical entity) to the three-dimensional representation; and (c) quantifying the association between the chemical entity and the peptide or peptide binding pocket.

The invention also relates to use of the atomic coordinates set forth in Table 1 , Table 2, and Table 3, or a related set of atomic coordinates, in a method of utilizing molecular replacement to obtain structural information about a molecule or a molecular complex of unknown structure. The method includes: (a) crystallizing said molecule or molecular complex; and (b) generating an X-ray diffraction pattern from said crystallized molecule or molecular complex; and applying at least a portion of the atomic coordinates set forth in Table 1 , Table 2, or Table 3, or a related set of atomic coordinates, to the X-ray diffraction pattern to generate a three-dimensional electron density map of at least a portion of the molecule or molecular complex whose structure is unknown.

Additionally, the invention is related to methods of growing crystals of a peptide o- factor complex comprising a human BCATc peptide. The method comprises:(a) providing a BCATc peptide solution comprising a human BCATc peptide, co-factor, buffering agent, reducing agent, chelating agent, and ionic source; (b) providing a precipitant solution comprising an ionic source, precipitating agent, buffering agent, chelating agent, and reducing agent;(c) mixing the BCATc peptide solution with the precipitant solution to form a third solution;(d) suspending the third solution over a container housing additional precipitant solution, wherein the vapor pressure of the additional precipitant solution is lower than the vapor pressure of the third solution; (e)allowing the suspended third solution to stand at a temperature of from about 4°C to about 20°C for a period of time until BCATc peptide:co- factor complex crystals grow to a predetermined size; (f) dipping the complex crystals into a cryoprotective solution; and (g) freezing the dipped complex crystals. Likewise, the invention is related to methods of growing crystals of a peptide:co- factor:ligand complex comprising a human BCAT peptide. The method comprises: (a) providing a BCATc solution comprising a human BCATc peptide, co-factor, ligand, buffering agent, reducing agent, chelating agent, and ionic source; (b) providing a precipitant solution comprising an ionic source, precipitating agent, buffering agent, chelating agent, and reducing agent; (c) mixing the BCATc peptide solution with the precipitant solution to form a third solution; (d) suspending the third solution over a container housing additional precipitant solution, wherein the vapor pressure of the additional precipitant solution is lower than the vapor pressure of the third solution; (e) allowing the suspended third solution to stand at a temperature of from about 4°C to about 20°C for a period of time until BCATc peptide:co- factor complex crystals grow to a predetermined size; (f) dipping the complex crystals into a cryoprotective solution; and (g) freezing the dipped complex crystals.

The invention also includes a method of making crystallizable BCATc peptide, comprising: (a)isolating a BCATc peptide; (b) charging the isolated BCATc peptide with a molar excess of pyridoxal-5'-phosphate (PLP); (c) subjecting the charged BCATc peptide to anion-exchange chromatography; and (d) recovering the charged BCATc peptide. BRIEF DESCRIPTION OF THE FIGURES

The application file contains at least one drawing executed in color. Copies of the patent appliction publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee. Figure 1 is a ribbons diagram of the BCATc structure.

Figure 2 is a representation of the PLP binding site of BCATc.

Figure 3 is a ribbons diagram of a representation of the catalytic site of BCATc.

Figure 4 is a representation of the AKG binding site of BCATc.

Figure 5 is a representation of the BCATc activity inhibitor 2-benzofurancarboxylic acid, 5-chloro-, 2-[[2-(trifluoromethyl)phenyl]sulfonyl] hydrazide (compound 1 ) bound to BCATc.

Figure 6 is a representation of the BCATc activity inhibitor [2-(2,6- dichlorophenylamino)-4-oxo-4H-thiazol-5-ylidine]acetic acid methyl ester (compound 2) bound to BCATc. DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

The present invention provides crystals of BCATc peptide:co-factor complexes comprising BCATc, or a protein structurally related to BCATc, and a co-factor, e.g., PLP. The invention further provides crystals of BCATc peptide:co-factor:ligand complexes comprising BCATc, or a protein structurally related to BCATc, a co-factor, e.g., PLP, and a ligand. The invention also provides the structural coordinates (e.g. atomic coordinates) derived from crystals of BCATc peptide:co-factor complexes and crystals of BCATc:co- factoπligand complexes. Particularly, the present invention provides the atomic coordinates of a BCATc:PLP crystal as set forth in Table 1 , and the atomic coordinates of BCATc:PLP:ligand crystals as set forth in Tables 2 and 3, and related set(s) of structural coordinates.

The atomic coordinates reflect the three dimensional structure of the BCATc peptide:co-factor complexes and BCATc peptide:co-factor:ligand complexes and illustrate, to atomic resolution, details regarding the BCATc peptide crystal structure and the chemical environment around BCATc and BCATc-like binding sites. As used herein, the term "site" is also referred to as "pocket."

The present invention also includes the use of the atomic coordinates set forth in Tables 1-3, or a related set of structural coordinates, and other structural information derived therefrom, including, but not limited to, the characterization of the BCATc and BCATc-like PLP binding site, catalytic binding site and AKG binding site in, for example, computer modeling to screen and identify, design and/or modify compounds that associate with BCATc or structurally related peptides and thus, may inhibit or enhance BCATc activity or the activity of a peptide that is structurally related to BCATc. Identification of possible inhibitors or enhancers may be accomplished via screening known candidate compounds or synthesizing candidate compounds de novo.

The present invention also provides methods of obtaining BCATc peptides and proteins structurally related to BCATc (wherein all BCATc forms may be modified as described herein), and the peptides so obtained. Definitions

As used herein, the terms "comprising" and "including" are used in the conventional, non-limiting sense.

As used herein, the term "ligand" means a molecule that binds to or associates with an enzyme and can be used to mean a BCATc activity inhibitor or enhancer.

As used herein, the term "co-factor" means an inorganic molecule, an organic molecule or a co-enzyme that is required for enzymatic activity of a protein associated with the co-factor. For example, PLP (pyridoxal-5'-phosphate) is a co-factor associated with BCATc. As used herein, the phrase "BCATc peptide" means native or modified BCATc peptides or proteins structurally related to native BCATc or modified BCATc. The BCATc peptide may be modified as described herein. Modified versions of BCATc possess the same activities as native BCATc. The term "BCATc" and the phrase "BCATc peptide" are used interchangeably herein. As used herein, the phrase "root mean square deviation" ("RMS") denotes a measure of the structural relationship between two or more species of proteins. It may be determined by, for example, superimposing one three-dimensional structure onto another, which may be solved by using, for example, X-ray crystallography or nuclear magnetic resonance (NMR), and then, calculating the difference in the RMS of the distance from the C-q and/or backbone (N, C, O, and C alpha) trace (or atoms) of one protein to another protein in units of Angstroms (A). The superimposition of three-dimensional structures may be performed using a molecular modeling program such as, for example, the Superimpose command in Insight II (Accelrys Inc., San Diego, CA), CNX (Accelrys Inc., San Diego, CA), XtalViewTM (Scripps Research Institute, La Jolla, CA), SYBYL® (Tripos, Inc., St. Louis, MO), or O (Aarhus Univ., Denmark (Jones, T.A. et al., Acta Cryst. A47: 110-119 (1991)), or other related computer modeling programs or scripts, alone or in combination. For example, the Superimpose command in Insight II, performs a minimum RMS alignment of two molecules on selected sets of atoms from each molecule is then outputs the RMS deviation value between the selected atoms of the superimposed molecules. The closer the relationship between the three- dimensional structures, the smaller the RMS deviation value. For example, the three- dimensional relationship between the C-α backbone trace of two identical protein structures solved from two different crystals is between 0.0 - 0.5 A RMS deviation. Therefore, one embodiment of this invention is the three-dimensional structures of the BCATc peptide complexes of the invention. An additional embodiment is a "structurally related" peptide, crystals of the structurally related peptide and the three-dimensional structures thereof.

As used herein, "structurally related" protein or peptide refers to a protein or peptide that is defined by the atomic coordinates set forth in Table 1 , Table 2, and/or Table 3, or by a related set of atomic coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A from the core C alpha atoms of the atomic coordinates set forth in Table 1 , Table 2, and/or Table 3. Preferably the root mean square deviation is not more than about 0.50A, more preferably not more than about 0.75 A, even more preferably not more than about 1.00 A, and most preferably not more than about 1.25 A.

Similarly, as used herein, "related set of structural coordinates" or "related set of atomic coordinates" refers to a set of structural (e.g. atomic) coordinates having a root mean square deviation of from not more than about 1.5 A to not more than about 0.50 A from the core C alpha atoms of the structural coordinates a set forth in Table 1 , Table 2, and/or Table 3. Preferably the root mean square deviation is not more than about 0.50A, more preferably not more than about 0.75 A, even more preferably not more than about 1.00 A, and most preferably not more than about 1.25 A.

As used herein, "chemical entity" refers to a chemical compound, a complex of at least two chemical compounds, or a fragment of such a compound or complex. Such entities can be, for example, potential drug candidates and can be evaluated for their ability to inhibit or enhance the activity of BCATc peptide, or a structurally related peptide.

As used herein, the term "inhibitor" or "inhibit" (or variations thereof) refers to a ligand such as a compound or substance that lowers, reduces, decreases, prevents, diminishes, stops or negatively interferes with BCATc activity. Often the terms "inhibitor" and "antagonists" can be used interchangeably. Inhibition is typically expressed as a percentage of the enzymes activity in the presence of the inhibitor over the enzymes activity without the inhibitor. Or it may be expressed in terms of IC50, the inhibitor concentration at which 50% of the original enzyme activity is observed.

As used herein, the term "enhancer" or "enhance" (or variations thereof) refers to a ligand such as a compound or substance that improves, increases, stimulates, raises or positively interferes with BCATc's activity. Often the terms "enhancer" or "agonists" can be used interchangeably. An enhancer would increase the enzyme's activity.

As used herein, "binding pocket," also referred to as, for example, "binding site," "binding domain," "substrate-binding site," or "catalytic domain," refers to a region or regions of a molecule or molecular complex, that, as a result of its surface features, including, but not limited to, volume (both internally in cavities or in total), solvent accessibility, and surface charge and hydrophobicity, can associate with another chemical entity or compound (e.g. a ligand or co-factor). Such regions are of utility in fields such as drug discovery.

As used herein, a "BCATc-like" peptide binding pocket refers to a peptide binding pocket defined by the atoms found in the structural coordinates as set forth in Table 1 , Table 2, and/or Table 3, or defined by structural coordinates having a root mean square deviation ranging from not more than about 1.5 A to not more than about 0.50 A from the binding pocket C alpha atoms of any one of the BCATc binding pockets (e.g. the BCATc co-factor or ligand binding pockets defined above in the Summary of the Invention section), or a conservatively substituted variant thereof. Preferably the root mean square deviation is not more than about 0.50A, more preferably not more than about 0.75 A, even more preferably not more than about 1.00 A, and most preferably not more than about 1.25 A.

As used herein, the term "activity" refers to all BCATc activities, e.g., glutamate synthesis, catalysis of the formation of BCAAs from corresponding ketoacids, etc., as well as to the enzyme's potency. The terms "activity" and "function" are used interchangeably herein. As used herein, the term "associate" refers to the process wherein at least two molecules reversibly interact with each other by, for example, binding with each other. This term may also refer to the process in which the conformation of a protein changes in response to the presence of a ligand (also referred to as "compound," "drug" or "substance") to better accommodate the steric and electrostatic properties of the ligand. Associations between BCATc and a ligand may occur with all or a part of a BCATc binding pocket. The association(s) may be non-covalent, e.g., wherein the juxtaposition is energetically favored by hydrogen bonding, van der Waals interactions or electrostatic interactions, or the association(s) may be covalent.

As used herein, the terms "model" and "modeling" mean the procedure of evaluating (also referred to as "assessing") the affinity of the interaction between a BCATc or BCATc-like binding pocket and a chemical entity (also referred to as a "candidate compound") based on, for example, steric constraints and surface/solvent electrostatic effects. BCATc Crystal Structure It has been discovered that the human BCATc:PLP crystal has an asymmetric unit of four monomers organized into two dimers. Each dimer has a molecular weight of approximately 90,000 Da and the approximate dimensions 80 A x 65 A x 40 A (Figure 1). The dimer interface is extensive and includes, but is not limited to, some interdigitation of polypeptide loops of one monomer in clefts in the second monomer. Each monomer contains two domains and a single, obligate bound PLP co-factor, and has the structure of an oblate ellipsoid of revolution with the approximate dimensions of 40 A x 65 A x 40 A. The amino terminal domain of human BCATc (amino acid residues 19-188 of SEQ ID NO:2) comprises a central 6-strand antiparallel beta sheet with four flanking short helical segments. The carboxyl terminal domain of human BCATc (amino acid residues 198-384 of SEQ ID NO:2) is comprised of a central 10-strand antiparallel beta barrel with four flanking external helical segments. There is a linker region of human BCATc (amino acid residues 189-197 of SEQ ID NO:2) which forms the link between the two domains in each monomer. Thus, each monomer contains 384 amino acid residues.

It has also been discovered that human BCATc has a PLP binding site, a catalytic binding site and an alpha-ketoglutarate (AKG) binding site, as defined below. PLP Binding Pocket The PLP co-factor is bound at the bottom of a deep rectangular pocket in the BCATc surface at the interface of the N-terminal and C-terminal domains with dimensions of approximately 14 A x about 10 A and about 6 A. The co-factor is covalently bound to the enzyme through the formation of a Schiff base to the side chain of Iys220. The 3-hydroxyl group of PLP forms a hydrogen bond to the phenolic hydroxyl of tyr225. The pyridinium nitrogen forms a salt bridge to the side chain of glu255. The phosphate oxygen is involved in a series of hydrogen bonds via argl 17, the main chain amide groups of val287 and thr288, as well as the side chain and amide of thr331 as illustrated in Figure 2. In addition, residues thr258, met259, asn260, and gly330 contact PLP and define the shape of its binding site. The pyridoxal-5'-phosphate (PLP) binding pocket is defined by the atomic coordinates of the following amino acid residues within about 5 A of a cofactor located in the PLP binding pocket: thr258, met259, asn260, gly330, Iys220, arg117, val287, thr288, thr331 , tyr225, glu255, and ser329.

Catalytic Binding Pocket

Adjacent to the PLP binding pocket is a large rectangular catalytic pocket/binding site with dimensions of about 14 A x about 10 A x about 19 A deep. One wall of the catalytic pocket is formed by the bound PLP co-factor. The rest of the catalytic pocket is bounded by the following residues: Ieu45, val46, phe47, gly48, thr49, phe93, glu94, gly95, Ieu96, Iys97, tyr159, arg161 , val188, gly189, pro190, tyr191 , phe192, ser193, val200, ser201 , Ieu202, arg210, asn224, ser228, Ieu229, gln232, val236, cys240, gln241 , gln242, val243, trp245, val256, gly257, ser329, ala332, cys333 and cys336 from one BCATc peptide monomer and by residues ser84, his87, tyr88, glu168, ser170, Ieu171 , gly172, val173, Iys174 and Iys175 from the second BCATc peptide monomer as illustrated in Figure 3. Of these residues, arg210, asn224, tyr225, ser228, Ieu229, gln232, trp245, gly257 and thr258 from the first BCATc monomer and val173 from the BCATc second monomer define the terminus of the channel to the AKG binding site. Thus, the catalytic binding pocket can be defined by the atomic coordinates of the following amino acid residues within about 5 A of a co-factor located in the catalytic binding pocket: (a) amino acid residues Ieu45, va!46, phe47, gly48, thr49, phe93, glu94, gly95, Ieu96, Iys97, tyr159, arg161 , val188, gly189, pro190, tyr191 , phe192, ser193, val200, ser201 , Ieu202, arg210, asn224, ser228, Ieu229, gln232, val236, cys240, gln241 , gln242, val243, trp245, val256, gly257, ser329, ala332, cys333 and cys336 from a first BACTc peptide monomer; and

(b) amino acid residues ser84, his87, tyr88, glu168, ser170, leu171 , gly172, val173, Iys174 and Iys175 from a second BCATc peptide monomer.

Alpha-Ketoglutarate (AKG) Binding Pocket

Adjacent to the PLP and catalytic binding pockets, there is a deep pocket in the surface of the enzyme that binds the substrate AKG! This pocket is rectangular in shape and about 18 A x about 7 A. AKG binds deep in this pocket with the keto-acid end of the molecule buried most deeply. The keto-acid is located about 5 A from the PLP phosphate group and is hydrogen bonded to the side chain of asp114 and thr331 and to the main chain amide of Iys97. The carboxylic acid group makes a salt bridge with the side chain of Iys97 and is hydrogen bonded to the side chain of gln371 as illustrated in Figure 4. In addition, residues Ieu96, phe99, phe109, val287, thr288 and val334 contact AKG and are within about 5 A of its binding pocket. The pocket is located at the bottom of a channel that is defined by the atomic coordinates of by the following residues within about 5 A of a substrate located in the channel: phe47, gly95, gln110, pro111 , asn112, Ieu113, ser155, ser157, val188, pro190, tyrl 91 , phel 92, cys291 , gly330, ala332, cys333, Ieu364, Ieu367, thr368 and tyr372. Crystallization of the BCATc Peptide Complexes

The present invention provides methods for growing mammalian, e.g., human, BCATc peptide crystals including, but not limited to, BCATc peptide:PLP complexes and BCATc peptide:PLP:ligand complexes, preferably, wherein the BCATc peptide is in a catalytically active configuration. Generally, any ligand that forms a complex with a BCATc peptide can be used to form a BCATc peptide:PLP:ligand complex of the present invention. Preferably, the ligand comprises a BCATc activity inhibitor or enhancer which binds to the BCATc catalytic binding site, thus, possibly displacing substrates such as BCAAs; binds to the PLP binding site, thus, possibly displacing PLP; or binds to the AKG binding site, thus, possibly displacing AKG. Preferred ligands are BCATc activity inhibitors such as 2- benzofurancarboxylic acid, 5-chloro-2-[[2-(trifluoromethyl)phenyl]sulfonyl]hydrazide, [2-(2,6- dichlorophenylamino)-4-oxo-4H-thiazol-5-ylidine]acetic acid methyl, dibenzofuran-2-carboxylic acid 2-(phenylsulfonyl)hydrazine, benzoic acid, [4-[phenylsulfonyl]amino]-2,2- (phenylsulfonyl)hydrazide and 2-benzofurancarboxylic acid, 5-methoxy-3-(1-methylethoxy)-2- (phenylsulfonyl)hydrazide.

Methods of crystallization are well-known in the art and numerous different methods and conditions may be employed to grow BCATc peptide crystals. See, e.g., McPherson, A., Preparation and Analysis of Protein Crystals, Krieger Press (1989). The presence of PLP in the BCATc peptide crystals results in the production of binary crystals and the presence of PLP and a ligand in the BCATc peptide crystals results in the production of ternary or tertiary crystals. Generally, the crystallization process comprises contacting a BCATc peptide with a co-factor or with a co-factor and ligand, wherein a stable binary or ternary, respectively, complex of a BCATc peptide:co-factor or BCATc peptide:co-factor:ligand is formed, and then growing a crystal of the BCATc peptide:co-factor or BCATc peptide:co-factor:ligand complex by adding a solution of the complex (BCATc solution) to a precipitating solution. The concentration of the BCATc peptide ranges from about 1.0 mg/ml to about 20 mg/ml, and, more preferably, is about 4.0 mg/ml. The concentration of co-factor ranges from about 1.0 mM to about 100 mM, preferably, ranges from about 2.5 mM to about 25 mM and, more preferably, is about 5.0 mM. PLP is preferably present in a molar excess of the BCATc peptide. The concentration of ligand, if present, ranges from about 2- to about 20-fold in excess of the BCATc peptide concentration, preferably, ranges from about 5- to about 15-fold in excess of the BCATc peptide concentration, and more preferably, is about 10-fold in excess of the BCATc peptide concentration. The pH of the BCATc solution preferably ranges from about 4.0 to about 6.0, and, more preferably, ranges from about 4.5 to about 5.5.

Ingredients in the BCATc solution include, but are not limited to, a BCATc peptide and co-factor, buffering agent, reducing agent, chelating agent and ionic source. The buffering agent may include, but is not limited to, phosphate, MES, HEPES, Tris, bis-Tris and bis-Tris propane. Preferably, HEPES is the buffering agent and, if present, has a pH ranging from about 6.8 to about 8.8, more preferably, ranging from about 7.0 to about 8.0, and, even more preferably, of about 7.5, and has a concentration ranging from about 10 mM to about 100 mM, more preferably, ranging from about 10 mM to about 50 mM and, even more preferably, of about 20 mM.

The reducing agent may be, but is not limited to, 2-mercaptoethanol, TCEP and DTT. Preferably, the reducing agent is TCEP and, if present, has a concentration ranging from about 0.1 mM to about 10 mM, more preferably, ranging from about 2.0 mM to about 10 mM and, even more preferably, of about 5.0 mM.

The chelating agent may be, for example, sodium citrate or EDTA. The preferred chelating agent is EDTA having a concentration ranging from about 1.0 mM to about 10 mM, and, more preferably, of about 2.0 mM.

The ionic source may be a salt such as NaCI, KCI, ammonium acetate or sodium acetate. Preferably, the salt is ammonium acetate at a concentration ranging from about 5.0 mM to about 500 mM, and, more preferably, of about 150 mM. Alternatively, the preferred salt is NaCI having a concentration ranging from about 20 mM to about 100 mM and, more preferably, of about 50 mM.

The precipitant solution, when mixed with the BCATc solution described above, preferably causes the BCATc peptide complex to form well-diffracting crystals. The precipitant solution may comprise a variety of components designed to stabilize the formation of the BCATc peptide complex as a crystalline solid. For example, the precipitant solution may include, but is not limited to, an ionic source such as ammonium acetate, a precipitating agent such as polyethylene glycol (PEG), a buffering agent such as trisodium citrate, a chelating agent such as EDTA, an additive to improve crystal clarity such as 2-methyl,2,4-pentanediol (MPD), and a reducing agent such as DTT.

More specifically, ingredients in the precipitant solution may include, but are not limited to, polyethylene glycol (PEG), tri-sodium citrate (can function as both a chelating agent and a buffer), ammonium acetate and MPD. Preferably, PEG has a molecular weight ranging from about 1000 to about 10,000, more preferably, ranging from about 3000 to about 10,000, and, even more preferably, of about 4000. The preferred concentration of PEG ranges from about 10% (w/v) to about 20% (w/v), and, more preferably, is about 15% (w/v). Preferably, trisodium citrate has a pH of about 5.6 and a concentration ranging from about 50 mM to about 200 mM, and, more preferably, of about 100 mM. Preferably, ammonium acetate has a concentration ranging from about 100 mM to about 300 mM, and, more preferably, of about 200 mM. Preferably, MPD has a concentration ranging from about 0.2% (v/v) to about 1.0% (v/v), and, more preferably, of about 0.6% (v/v).

Many possible methods may be used to grow the crystals of the BCATc peptide complexes, including, but not limited to, hanging-drop vapor diffusion, sitting-drop vapor diffusion, microbatch, batch, or counter diffusion in gels or oils. Preferably, crystallization is performed by hanging-drop vapor diffusion wherein a droplet of the BCATc solution is mixed with a droplet of the precipitant solution to obtain a mixed droplet solution. The mixed droplet solution is then suspended over a well of precipitant solution in a sealed container. The mixed droplet solution is preferably placed on a glass slide prior to inclusion in the sealed container. In a preferred embodiment, about 1.0 μl of the BCATc solution is mixed with the precipitant solution in a ratio ranging from about 1 :4 to about 4:1 , preferably, ranging from about 1 :2 to about 2:1 and, even more preferably, of about 1 :1. In one embodiment, the mixed droplet may be suspended over a well containing between about 0.5 ml and about 1.2 ml of precipitant solution, and, more preferably, containing about 0.75 ml of precipitant solution. The vapor pressure of the precipitant solution in the well must be lower than the vapor pressure of the mixed droplet solution in order for crystals to form. The crystallization temperature may be between about 4°C and about 20°C, and, preferably, is about 4°C. The mixed droplet solution is allowed to stand suspended over the well containing the precipitant solution at the crystallization temperature for a period of about 5 days to about 5 weeks, preferably, about 4 weeks, until the crystals reach a size appropriate for crystallographic data collection such as about 0.4 mm x 0.2 mm x 0.05 mm. If the crystals are of good quality, but too small, they can be used to grow larger crystals. The process is the same as described above except that each hanging droplet is seeded with a few small crystals.

At this point in the crystallization process, the BCATc peptide:PLP crystals may be soaked in a ligand- or ligands-containing solution, wherein the ligand is present at a concentration ranging from about 0.1 mg/ml to about 10 mg/ml and, more preferably, ranging from about 1.0 mg/ml to about 4.0 mg/ml, depending upon the solubility of the particular ligand. The ligand solution may include, but is not limited to, PEG, tri-sodium citrate and ammonium acetate. Preferably, PEG has a molecular weight ranging from about 1000 to about 10,000, more preferably, ranging from about 3000 to about 10,000, and, even more preferably, of about 4000. The preferred concentration of PEG ranges from about 12% (w/v) to about 20% (w/v), and, more preferably, is about 15% (w/v) in water. Preferably, tri-sodium citrate has a pH of about 5.6 and a concentration ranging from about 50 mM to about 200 mM, and, more preferably, of about 100 mM. Preferably, ammonium acetate has a concentration ranging from about 50 mM to about 300 mM, and, more preferably, of about 200 mM. After soaking for about 1 to about 48 hours at about 4oC, the BCATc peptide:PLP:ligand crystals are harvested and dipped in a cryoprotective solution including, but not limited to, ammonium acetate, tri-sodium citrate, PEG and glycerol. The cryoprotective solution comprises components designed to stabilize the formation of a vitreous solid containing the BCATc complex as a crystalline solid at a temperature of about 100oK. Preferably, the cryoprotective solution comprises ammonium acetate having a concentration ranging from about 100 mM to about 200 mM, and, more preferably, of about 170 mM; tri-sodium citrate having a pH of about 5.6 and a concentration ranging from about 50 mM to about 100 mM, and, more preferably, of about 85 mM; and PEG having a molecular weight ranging from about 1000 to about 10,000, more preferably, ranging from about 3000 to about 10,000, and, even more preferably, of about 4000. PEG, such as PEG 4000, has a concentration ranging from about 20% (w/v) to about 30% (w/v) and, more preferably, of about 25.5% (w/v) in water. Glycerol, preferably, anhydrous glycerol, is also present having a concentration ranging from about 12% (v/v) to about 20% (v/v), and, more preferably, of about 15% (v/v). The solution is then flash-frozen by immersion in a stream of cold nitrogen at, for example, 100oK. Alternatively, the crystals may be dipped directly into liquid nitrogen or liquid propane. Alternatively, a ligand or ligands may be added to the BCATc solution in the beginning of the crystallization process in a process known as co-crystallization.

Crystals of the present invention may take a variety of forms, all of which are included in the present invention, such as triclinic, monoclinic, orthorhombic, tetragonal, cubic, trigonal or hexagonal. In a preferred embodiment, the crystals of human BCATc:PLP are orthorhombic and possess a space group P212121 (#19) with unit cell dimensions of: a =

109.9 N b = 113.5 A and c = 149.2 A. See, e.g., Example 2.

X-Ray Data Collection and Structural Analysis

The present invention further includes methods for resolving the three-dimensional structures of BCATc peptide complexes using analyzed X-ray diffraction data. Preferably, the three-dimensional structures of the crystals are resolved to about 3.0 A or better, more preferably, about 2.5 A or better, even more preferably, about 2.0 A or better, and, most preferably, about 1.9 A or better. The data collection methods and conditions cited herein are provided to elucidate the approach used for the structural determination of BCATc peptide complexes. One of ordinary skill in the art would be aware of other methods and conditions that may be suitable for X-ray data collection and structural determination of BCATc peptide complexes. See, e.g., Glusker, J., Crystal Structure Analysis for Chemists and Biologists, Wiley-VCH Press (1994).

Generally, collecting the X-ray diffraction data for the BCATc peptide complex crystals comprises mounting the crystals in a cryoloop, bathing the crystals in a cryoprotectant solution as described above and rapidly cooling the crystals to about 100oK, followed by collecting diffraction data in the oscillation mode. The source(s) of X-rays includes, but is not limited to, a standard rotating anode home source such as a Rigaku® Ru- H3R or Ru-200B generator (Rigaku Corp., Tokyo, Japan), a sealed tube or a synchrotron source.

The preferred method of data collection is to collect an initial data set using a home source to evaluate crystal quality and then collecting a complete data set at IMCA-CAT (insertion device at beamline 17 (17 ID) at the Argonne, IL National Laboratory Advanced Photon Source). The method of detecting and quantitating the diffraction data, i.e., diffraction pattern produced by the diffracted X-rays, may be performed using, for example, a standard image plate such as the R-Axis IV++ (Rigaku/MSC, Inc., The Woodlands, TX) or a charge- coupled device such as the MAR-CCD X-ray detector. The data is generally corrected for Lorentz and polarization effects and converted to indexed structure factor amplitudes using data processing software such as DENZO®, HKL-2000 or SCALEPACK (HKL Research, Inc., Chariottesville, VA) (Otwinowski, Z. et al., Methods Enzymol. 276: 307-326 (1997)), d*Trek (Rigaku/MSC, Inc. (Pflugrath, J.W., Acta Cryst. D55: 1718-1725 (1999)), or MOSFILM (Leslie, A.G.W., Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography, No. 26 (1992)), or other related computer programs, alone or in combination. The preferred processing software is DENZO® and/or HKL-2000 (HKL Research, Inc.).

The three-dimensional image generated from the X-ray diffraction data, more specifically, generated from the intensities of the diffracted X-rays, is referred to as an electron density map of the repeating unit of the crystal. However, the electron density map cannot be generated until the amplitudes and phases of the diffracted X-rays are known.

Amplitudes may be obtained directly from the intensities as described above.

Phases may be obtained indirectly by, for example, any one or a combination of the following methods: computational methods, molecular replacement analysis (if a homologous structure is known), heavy atom substitution techniques, e.g., isomorphous replacement, synchrotron radiation at multiple wavelengths, Patterson difference, single-wavelength anomalous scattering, etc. Software that can aid in generating the electron density map includes, but is not limited to, SHARP (Statistical Heavy Atom Refinement and Phasing) (de la Fortelle, E. et al., Meth. Enzymol. 276: 472-494 (1997)) and SOLOMON (Abrahams, J.P. et al., Acta Cryst. D52: 30-42 (1996)), and other related computer programs, alone or in combination. The map may then be used, via model building, to build a model of the enzyme. A molecular model of the amino acid or nucleotide sequence is then fit into the electron density map and the map is refined. Refinement establishes a set of atomic coordinates representing every non-hydrogen molecule of the enzyme or enzyme complex and results in a three-dimensional structure. Thus, another aspect of the invention involves using the atomic coordinates to generate a three-dimensional shape. Atomic coordinates (also referred to as "structure coordinates," "structural coordinates" and "crystal coordinates") are Cartesian coordinates derived from mathematical equations related to the patterns obtained on diffraction of a monochromatic beam of X-rays by the atoms of a protein or protein complex in crystal form.

Molecular replacement analysis, mentioned above, involves using a known three- dimensional structure as a model to assist in determining the structure of an identical or closely related protein or protein-ligand complex in a new crystal form. See, e.g., Example 3. The measured X-ray diffraction intensities of the new crystal are compared with those calculated from the known related structure to compute the positions and orientations of the molecules in the new crystal. Computer programs that can be used to perform this task include, but are not limited to, X-PLOR (Accelrys Inc.), EPMR (Kissinger et al., Acta Cryst. D55: 484-491 (1999)), ProLSQ, AMORE (Navaza, J., Acta Cryst. AS0: 157-163 (1994)), X- SITE, QUANTA (Accelrys Inc.), INSIGHT-II (Accelrys Inc.), ARP/wARP (Automated Refinement Procedure) (Lamzin, U.S. et al., Acta Cryst. D49: 129-147 (1993)), and ICM, and other related computer programs, alone or in combination. Using this approach, it is possible to use a known structure to solve the three dimensional structures of related proteins or protein complexes.

The model building and refinement of a BCATc peptide complex may be performed using, for example, GRIN/GRID (Molecular Discovery Ltd., London, England), MolCad (Tripos, Inc.), QUANTA (Accelrys Inc.), CHARMm® (Accelrys Inc.), INSIGHT®-II (Accelrys Inc.), SYBYL® (Tripos, Inc.), MacroModel® (Trustees of Columbia Univ., New York, N.Y.), ICM (MolSoft LLC, San Diego, CA), CNX (Accelrys Inc.), CAVEAT (P.A. Bartlett, et al., Royal Chem. Soc. 78: 182-196 (1989), available from the University of California, Berkley, CA)), GRASP (A. Nicholls, Columbia University), or SitelD (Tripos, Inc.), or other related computer programs. The preferred programs are QUANTA (Accelrys Inc.), SYBYL® (Tripos, Inc.) and/or CNX (Accelrys Inc). Any of the programs may be used individually or in combination.

The computer software may be used alone or combined with a docking computer program such as GOLD (commercially available via Cambridge Crystallographic Data Centre, Cambridge, UK; Jones, G., J. Mol. Biol. 245: 43-53 (1995)), FlexX (Tripos, Inc.), GRAMM (llya A. Vakser, Rockefeller Univ.), Flexidock (Tripos, Inc.), Dock (Ewing, T.J.A. et al., J. Comput.-Aided Mol. Des. 15: 411-428 (2001)), or AutoDock (Molecular Graphics Laboratory (Scripps Research Inst.); Goodsell, D.S., J. Mol. Recognit. 9: 1-5 (1996)), or other related computer programs. These docking computer programs scan known databases of small molecules to find core compounds that roughly fit the binding sites. Any of the programs may be used alone or in combination.

If necessary, the three-dimensional structure may be "cleaned up" by modifying the atom types of the ligand, if present, and the co-factor and any water molecules that are present so that the water molecules find their lowest energy rotamer. The software may also be used to add hydrogens in standardized geometry with optimization of orientations of OH, SH, NH3+, Met methyls, Asn and Gin sidechain amides, and His rings. Suitable software for performing this "clean up" includes, but is not limited to, SYBYL® (Tripos, Inc.), WHATCHECK (part of CCP4 suite, COLLABORATIVE COMPUTATIONAL PROJECT, No. 4, Acta Cryst. D50: 760-763 (1994)), and REDUCE (Word et al., J. Mol. Bio. 285: 1733-45 (1999)), and other related computer programs. Any suitable docking computer program may be used to further validate the refined peptide crystal structure by placing hydrogens in the most favorable protonation state and/or by rotating water molecules into orientations that give optimal interactions with the peptide. Any of the programs may be used alone or in combination.

For human BCATc, DENZO® and HKL-2000 (HKL Research, Inc.) were initially used to process the X-ray diffraction data. QUANTA (Accelrys Inc.), CNX (Accelrys Inc.), and SYBYL® (Tripos, Inc.) were then employed to determine the atomic coordinates provided in Tables 1-3, to determine the characteristics of the BCATc crystals as described above and to determine that the BCATc active site is segmented into three subsites as described in detail above: (1) a PLP binding pocket that covalently binds PLP; (2) a catalytic pocket, i.e., a large active site channel that allows BCAAs access to the PLP co-factor; and (3) a narrow AKG binding pocket which specifically binds AKG. The PLP- and AKG-binding pockets, described in detail above, are directly connected by a short (approximately 6 angstroms) channel.

One of ordinary skill in the art would recognize that a set of atomic coordinates for a BCATc peptide complex, or a subset thereof, is a relative set of points in space that defines a complex three-dimensional surface. See, e.g., Tables 1-3. As such, it is possible to represent the same surface using an entirely different set of coordinates. Also, due to small errors in the measurement and resolution of crystallographic data, slight variations in individual coordinates may exist which have little or no effect on the overall surface. Thus, a binding pocket could be generated from the coordinates provided herein or from some variation of the coordinates that still retains similar surface features, including, but not limited to, volume (both internally in cavities and in total), solvent accessibility, surface charge and hydrophobicity. Variations of the coordinates found in Tables 1-3 (described in Examples 3-5, respectively) may be generated by one skilled in the art because of mathematical manipulations and crystallographic permutations, including, but not limited to, fractionalization, integer addition or subtraction, inversion or any combination thereof. Alternatively, modifications in the crystal structure due to mutations, additions, substitutions, and/or deletions of amino acids (as described below), or other changes in any of the components that make up the crystal could also account for variations in atomic coordinates. If such variations are within an acceptable standard error as compared to the original atomic coordinates, then the resulting three-dimensional shape is considered to be the same. For example, a ligand that binds to a BCATc binding site would also be expected to bind to another binding site whose atomic coordinates, when compared to those described, have a root mean square difference of from not more than about 1.5 A to not more than about 0.5 A. Preferably the root mean square deviation is not more than about 0.50A, more preferably not more than about 0.75 A, even more preferably not more than about 1.00 A, and most preferably not more than about 1.25 A. In addition, it would be apparent to one skilled in the art that the binding pockets described in detail above could be modified in order to obtain somewhat different three-dimensional coordinates.

It should also be recognized that minor modification of any or all of the components of the peptide: ligand: cofactor complexes that results in the generation of structural coordinates that still retain the basic features of the three-dimensional structure should be considered part of the invention.

Once the atomic coordinates are known and thus, the three-dimensional structure, the BCATc binding site(s) may be manipulated to create new images or representations and structures. Such manipulation may be desirable to, for example, enhance bioactivity in the binding site(s), modify stability such that functionality of the enzyme is not affected, reduce immunogenicity by modifying surface residues, and alter enzyme solubility.

The atomic coordinates and thus, three-dimensional structure, may also be used in homology modeling or NMR spectroscopy for drug design, for example.

Computers, Computer Software, Computer Modeling

As discussed above, once the atomic coordinates are known, a computer may be used for producing a three-dimensional representation of the BCATc peptide or structurally related peptide and BCATc peptide or BCATc-like binding pockets. Thus, another aspect of the invention involves using the structure coordinates generated from the BCATcPLP and BCATc:PLP:ligand complexes to generate three-dimensional representations of BCATc peptide, a structurally related peptide, or a BCATc or BCATc-like peptide binding pocket. This is achieved through the use of commercially available software that is capable of generating three-dimensional graphical representations of molecules or portions thereof from a set of atomic coordinates.

Suitable computers are known in the art and typically include a central processing unit (CPU), and a working memory, which can be random-access memory, core memory, mass-storage memory, or a combination thereof. The CPU may encode one or more programs. Computers also typically include display, input and output devices, such as one or more cathode-ray tube display terminals, keyboards, modems, input lines and output lines. Further, computers may be networked to computer servers (the machine on which large calculations can be run in batch) and file servers (the main machine for all the centralized databases).

Machine-readable media containing data, such as the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, and Table 3, or a related set of atomic coordinates, may be inputted using various hardware, including modems, CD-ROM drives, disk drives, or keyboards.

Machine-readable data medium can be, for example, a floppy diskette, hard disk, or an optically-readable data storage medium, which can be either read only memory, or rewritable, such as a magneto-optical disk.

Output hardware, such as a CRT display terminal, may be used for displaying a graphical representation of the three-dimensional structural coordinates of the BCATc peptide as set forth in Table 1 , Table 2, and/or Table 3, of a structurally related peptide, or of a BCATc or BCATc-like binding pocket, as described herein. Output hardware may also include a printer and disk drives. The CPU coordinates the use of the various input and output devices, coordinates data access from storage and access to and from working memory, and determines the sequence of data processing steps. A number of programs may be used to process the machine-readable data. Such programs are discussed herein in reference to the computational methods of drug discovery.

In a preferred embodiment of the invention, atomic coordinates capable of being processed into a three-dimensional representation of a molecule or molecular complex that comprises a BCATc peptide or BCATc-like binding pocket are stored in a machine-readable storage medium. As described below, the three-dimensional structure of a molecule or molecular complex comprising a BCATc or BCATc-like binding pocket is useful for a variety of purposes, such as in drug discovery and drug design. For example, the three-dimensional structure derived from the atomic coordinate data may be computationally evaluated for its ability to associate with chemical entities.

BCATc Peptide Activity Inhibitors and/or Enhancers

The association of natural ligands with their corresponding binding pockets on receptors or enzymes is the basis of many biological mechanisms of action. Similarly, many drugs exert their biological effects via an interaction with the binding pockets of a receptor or enzyme. An understanding of such associations can lead to the design of drugs having more favorable and specific interactions with their target receptors or enzymes, and thus, improved biological effects. Therefore, information related to ligand association with the BCATc or BCATc-like peptide binding sites is valuable in designing and/or identifying potential inhibitors or enhancers of BCATc or structurally related peptides. Further, the more specific the design of a potential drug, the more likely that the drug will not interact with similar proteins, thus, minimizing potential side effects due to unwanted cross interactions.

Computer programs can be employed to estimate the attraction, repulsion, and steric hindrance of a ligand to a BCATc or BCATc-like binding pocket. For example, one can screen computationally small molecule databases for chemical entities or compounds that can bind in whole, or in part, to a BCATc binding pocket. In this screening, the quality of fit of such entities or compounds to the binding site may be judged either by shape complementarity or by estimated interaction energy (Meng, et al., J. Comp. Chem., 73:505- 524 (1992)). Generally, the tighter the fit, e.g., the lower the steric hindrance and/or the greater the attractive force, the more potent the drug is projected to be since these properties are consistent with a tighter-binding constant.

The present invention provides methods for screening ligands as potential therapeutic agents for the treatment of various disease conditions associated with BCATc such as AD and other neurodegenerative diseases and certain behavioral disorders associated with glutamatergic mechanisms. See, e.g., Beal, M.F., Current Opinions in Neurobiology 2: 657 (1992). More specifically, the present invention provides methods for screening and identifying potential BCATc activitfy inhibitors and/or enhancers, or inhibitors or enhancers of a peptides structurally related to BCATc, by de novo design of novel compounds or the modification of known compounds that associate with a BCATc peptide to inhibit or enhance BCATc activity. The availability of specific activity inhibitors and/or enhancers, particularly, inhibitors, can facilitate the analysis of the role of BCATc in certain diseases and disorders and may ultimately provide a new therapeutic approach for treating such diseases and disorders since, as noted above, the inhibition of BCATc is believed to have therapeutic utility in the treatment of AD and other neurodegenerative diseases, and in the treatment of certain behavioral disorders associated with glutamatergic mechanisms. See, e.g., Beal, M.F., Current Opinions in Neurobiology 2: 657 (1992).

The compound design or modification process begins after the structure of the target, e.g., a human BCATc peptide, is resolved to at least a resolution of about 3.0 A or better, preferably, about 2.5 A or better, more preferably, about 2.0 A and, even more preferably, about 1.9 A or better as described above. Refinement of the structure to a resolution of about 2.0 A or better with 266 water molecules provides optimal conditions for compound design. As described above, the data generated from the resolved crystal structure is applied to a computer algorithm to generate a three-dimensional representation and, ultimately, model, of the BCATc molecule and BCATc or BCATc-like binding pockets.

After a three-dimensional representation of the BCATc molecule, a structurally related peptide, or a BCATc or BCATc-like binding pocket is generated, a ligand having the potential to associate with the peptide or binding pocket is generated by, for example, (i) assembling molecular fragments into the chemical entity; (ii) de novo design of the chemical entity; (iii) selecting a chemical entity from a small molecule database; or (iv) modifying a known inhibitor, or portion thereof, of BCATc activity.

If a chemical entity is designed, the following factors may be considered. First, the entity must be capable of physically and structurally associating with some or the entire BCATc or BCATc-like binding pocket. Second, the entity must be able to assume a conformation that allows it to associate with the BCATc or BCATc-like binding pocket directly. Although certain portions of the entity will not directly participate in these associations, those portions of the entity may still influence the overall conformation of the molecule. This, in turn, may have a significant impact on potency. Such conformational requirements include the overall three-dimensional structure and orientation of the chemical entity in relation to all or a portion of the binding pocket, and the spacing between functional groups of an entity comprising several chemical entities that directly interact with the BCATc or BCATc-like binding pocket.

The design of new compounds or the modification of known compounds may involve synthesizing or modifying compounds, or fragments thereof, via computer programs which build and link fragments or atoms into a target binding site(s) based upon steric and electrostatic complementarity, without reference to substrate analog structures. Selected compounds, or fragments thereof, may be positioned in a variety of orientations, or docked, within the BCATc or BCATc-like binding pocket(s) as defined by the atomic coordinates. If compounds have been selected, then they may be assembled into a single complex. If fragments have been selected, then they may be assembled into a single compound. Assembly may be preceded by visual inspection of the relationship of the compounds or fragments to each other on the three-dimensional BCATc or BCATc-like representation displayed on a computer screen in relation to the atomic coordinates. This visual image step may be followed by manual model building using appropriate software programs. Alternatively, compounds may be designed as a whole using either empty binding pocket(s) or binding pocket(s) containing the natural iigand(s).

Computer programs that may be used in the design or modification of the potential ligand include, but are not limited to, alone or in combination, QUANTA (Accelrys Inc.) and/or SYBYL® (Tripos, Inc.) and/or a docking computer program such as GOLD (commercially available via Cambridge Crystallographic Data Centre, Cambridge, UK; Jones, G., J. Mol. Biol. 245: 43-53 (1995)), FlexX (Tripos, Inc.), GRAMM (llya A. Vakser, Rockefeller Univ.), Flexidock (Tripos, Inc.), Dock (Ewing, T.J.A. et al., J. Comput.-Aided Mol. Des. 15: 411-428 (2001)), or AutoDock (Molecular Graphics Laboratory (Scripps Research Inst.); Goodsell, D.S., J. Mol. Recognit. 9: 1-5 (1996)). In addition, other related computer programs may be used. The potential inhibitory or binding effect of the chemical entity on a BCATc peptide, a structurally related peptide, or BCATc or BCATc-like binding pocket may be analyzed prior to its actual synthesis and testing through the use of computer modeling techniques. The "modeling" includes applying an iterative or rational process to individual or multiple potential ligands, or fragments thereof, to evaluate their association with the BCATc or related peptide structure and to evaluate their inhibition and/or enhancement of BCATc or related peptide activity. This procedure may include computer fitting a potential ligand into a BCATc or BCATc-like binding site(s) to ascertain how well the shape and chemical structure of the potential ligand complements or interferes with the peptide. Computer programs may also be used to estimate the attraction, repulsion and steric hindrance of the ligand to the BCATc or BCATc-like binding sites. Generally, the tighter the fit, e.g., the lower the steric hindrance and/or the greater the attractive force, the more antagonistic or agonistic the potential ligand will be since these properties are consistent with a tighter-binding constant. If the theoretical structure, i.e., computational structure, indicates insufficient interaction and association, further testing may not be necessary. However, if computer modeling indicates a strong interaction, then the ligand may be synthesized and tested for its ability to bind to a BCATc or BCATc-like binding site(s). Thus, a potential inhibitor or enhancer may be identified and selected, based on its computational ability to positively associate with the amino acid residues found within any one or all of the binding sites, e.g., the potential inhibitor or enhancer binds to the catalytic binding pocket in the presence or absence of a substrate such as a BCAA and, possibly, forms a covalent adduct with the PLP co-factor; binds to the PLP binding pocket in the presence or absence of PLP such that access of substrates to the bound PLP is possibly denied as illustrated in Figure 2; binds to the AKG binding pocket thus, displacing the substrate AKG as illustrated in Figure 4; or binds to the AKG binding pocket in addition to AKG.

The screening method and subsequent identification of potential ligands, may be accomplished in vivo, in vitro or ex vivo. Initial ligand computation analysis is optional. Instead, or additionally, high-throughput screening may be employed which may be capable of full automation at robotic workstations such that large collections of compound libraries may be screened.

In one embodiment of the screening and identification method, the initial computer modeling is performed with one or more of the following docking computer modeling programs: Dock (Ewing, T.J.A. et al., J. Comput.-Aided Mol. Des. 15: 411-428 (2001)), AutoDock (Molecular Graphics Laboratory; Goodsell, D.S., J. Mol. Recognit. 9: 1-5 (1996)), GOLD (commercially available via Cambridge Crystallographic Data Centre, Cambridge, UK; Jones, G., J. Mol. Biol. 245: 43-53 (1995)) or FlexX (Tripos, Inc.). Potential ligands initially identified by the docking program(s) are elaborated using standard modeling methods as found in, for example, SYBYL® (Tripos, Inc.), QUANTA (Accelrys Inc.), INSIGHT®-II (Accelrys Inc.), GRIN/GRID (Molecular Discovery Ltd.), UNITY® (Tripos, Inc.), LigBuilder (Want, R., J. Mol. Model 6: 498-516 (2000)), or SPROUT (developed and distributed by ICAMS (Institute for Computer Applications in Molecular Sciences) at the University of Leeds, United Kingdom (Gillet, V. et al., J. Comput. Aided Mol. Design 7: 127-153 (1993))).

After a potential activity inhibitor and/or enhancer is identified, it can either be selected from commercial libraries of compounds or alternatively the potential inhibitor and/or enhancer may be synthesized and assayed to determine its effect(s) on the activity of BCATc peptide or a related peptide. Optionally, the assay may be radioactive. However, in a preferred embodiment, the assay is a non-radioactive ELISA. Secondary assays, for example, rate-limiting assays, may be employed which measure the conversion of AKG to glutamate in vitro using leucine as the nitrogen source.

In one embodiment of screening and identifying potential ligands via computer modeling, the method comprises: (a) generating a three-dimensional model of BCATc or a structurally related peptide; (b) designing and building (e.g. computationally) de novo potential ligands; and (b) identifying the ligands that associate with the BCATc or BCATc-like binding site(s). In an alternative embodiment of screening and identifying potential ligands via computer modeling, the method comprises: (a) generating a three-dimensional model of

BCATc or a structurally related peptide; (b) building (e.g. computationally) and, optionally, modifying, known potential ligands; and (b) identifying the ligands that associate with the BCATc or BCATc-like binding site(s).

In an alternative embodiment, the compound screening and identification method comprises evaluating the ability of de novo compounds to function as BCATc peptide or a structurally related peptide activity inhibitors and/or enhancers by, for example: (a) generating a BCATc or BCATc-like virtual binding cavity, the binding cavity defined by the binding sites; (b) designing (e.g. computationally) a compound structure that spatially conforms to the binding cavity; (c) synthesizing the compound and, optionally, analogs thereof, and (d) testing to determine whether the compound binds to at least one of the binding sites by, for example, determining whether the compound inhibits the production of glutamate by BCATc or the structurally related peptide in a biochemical assay known in the art, if desired.

In an alternative embodiment, the compound screening and identification method comprises evaluating the ability of known compounds to function as BCATc peptide or a structurally related peptide activity inhibitors and/or enhancers by, for example: (a) generating a BCATc or BCATc-like virtual binding cavity defined by the binding sites; (b) generating (e.g. computationally) and, optionally, modifying, a known compound structure; (c) determining whether that compound spatially conforms to the binding cavity; (d) synthesizing the compound and, optionally, analogs thereof; and (e) testing to determine whether the compound binds to at least one of the binding sites by, for example, determining whether the compound inhibits the production of glutamate by BCATc or the structurally related peptide in a biochemical assay known in the art, if desired.

In another embodiment, wherein a potential ligand has been selected, the identification method comprises: (a) generating a three-dimensional BCATc structure or three-dimensional structure of a structurally related peptide with the potential ligand bound thereto; (b) modifying the potential ligand based on the three-dimensional peptide structure; and (c) generating a second three-dimensional peptide structure with the modified potential ligand bound thereto. Then, one can test the potential ligand in a biochemical assay known in the art, if desired.

In addition, when a potential ligand is identified, a supplemental crystal may be grown comprising the BCATc (or structurally related peptide):PLP:ligand complex. Molecular replacement analysis, for example, may be used to determine the three-dimensional structure of the supplemental crystal. Molecular replacement analysis may also be used in the initial crystal structure determination as described above and in Example 3. It should be understood that in all of the structure-based drug design strategies provided herein, a number of iterative cycles of any or all of the steps may be performed to optimize the selection.

Thus, according to another embodiment, the invention provides compounds that associate with a BCATc or BCATc-like binding pocket(s) produced or identified by any one or a combination of the methods set forth above. BCATc Nucleic Acid Sequences

A nucleic acid sequence of the polynucleotide encoding the full-length protein of human BCATc is published in Naylor, S.L. et al., Somatic Cell Genet. 6: 641-652 (1980); Schuldiner, O. et al., PNAS 93: 7143-7148 (1996); and Bledsoe, R.K. et al., Biochim. Biophys. Acta 1339(1): 9-13 (1997). GenBank lists the Accession No. as NM_005504 derived from Accession No. U21551.1. The nucleic acid sequences described therein are identified herein as SEQ ID NO:1 as the sequences in the two accession numbers are identical. Nucleotides 1-1155 of SEQ ID NO:1 encode amino acid residues 1-384 of SEQ ID NO:2 (and a stop codon) which is the corresponding peptide sequence of the full length protein. The peptide sequence has GenBank Accession No. AAB08528.1 (Swiss-Prot Accession No. P54687).

Another BCATc nucleic acid sequence is set forth in SEQ ID NO:4. SEQ ID NO:4 is identical to the published sequence (SEQ ID NO:1 ) except for the codon at nucleotides 1096- 1098. In SEQ ID NO:1 , the codon at nucleotides 1096-1098 is AAA. In SEQ ID NO:4, the codon at nucleotides 1096-1098 is GAA.

There are 22 known amino acids, but 64 possible permutations of nucleic acid triplets ("codons"), thus, many amino acids are specified by more than one codon. For example, both UUU and UUC code for phenylalanine, and serine is encoded by UCU, UCC, UCA, UCG, AGU and AGC (Molecular Biology of the Gene, 4th edition, Watson, J.D. et al. (eds.) (1987) at pages 437-438). Due to this degeneracy of the genetic code, there are many functionally equivalent nucleic acid sequences that can encode the same peptide. Thus, the human BCATc amino acid sequences set forth in SEQ ID NOS: 2 and 5 can be encoded by multiple nucleotide sequences and are not limited to the nucleotide sequences set forth in SEQ ID NOS: 1 and 4, respectively. Further, functionally equivalent nucleotide sequences can be readily prepared using known methods such as modified primer PCR, site-directed mutagenesis and chemical synthesis.

As used herein, the nucleic acid molecules are "purified" or "isolated," i.e., separated from other nucleic acid molecules present in the natural source of the nucleic acid molecules. Preferably, an isolated nucleic acid molecule is free of sequences which naturally flank the nucleic acid molecule, i.e., sequences located at the 5' and 3' ends of the nucleic acid molecule in the genomic DNA or cDNA of the organism from which the nucleic acid molecule is derived. However, there can be some flanking nucleotide sequences, for example up to about 5 KB, particularly contiguous peptide encoding sequences and peptide encoding sequences within the same gene, but separated by introns in the genomic sequence. Thus, the nucleic acid molecule can be fused to other coding or regulatory sequences and still be considered isolated. For example, recombinant nucleic acid molecules contained in a vector are considered isolated. The important point is that the nucleic acid molecule is isolated from remote and unimportant flanking sequences such that it can be subjected to the specific manipulations described herein such as recombinant expression, preparation of probes and primers and other uses specific to nucleic acid sequences. In addition, the isolated nucleic acid molecules can encode a BCATc peptide of the present invention plus additional amino or carboxy-terminal amino acids or amino acids interior to the mature peptide (when the mature peptide has more than one peptide chain, for example). Such sequences may, for example, facilitate protein trafficking, prolong or shorten protein half-life, facilitate manipulation of a protein for assay or production, facilitate or enhance crystallization and/or solubility, among other things. As generally is the case in situ, the additional amino acids may be processed away from the mature peptide by cellular enzymes.

Thus, the isolated nucleic acid molecules include, but are not limited to, the sequence encoding a BCATc peptide of the present invention alone or in combination with coding sequences, such as a leader or secretory sequence, e.g., a pre-pro or pro-protein sequence, and the sequence encoding a BCATc peptide of the present invention, with or without the additional coding sequences, plus additional non-coding sequences, for example introns and non-coding 5' and 3' sequences such as transcribed, but non-translated sequences that play a role in transcription, mRNA processing (including, but not limited to, splicing and polyadenylation signaling), ribosome binding and mRNA stability. In addition, the nucleic acid molecules may be fused to a marker sequence encoding, for example, a peptide that facilitates purification.

The isolated nucleic acid molecules are substantially free of other cellular material, or culture medium when produced by recombinant techniques, or chemical precursors or other chemicals when chemically synthesized. For example, isolated nucleic acid molecules include, but are not limited to, recombinant DNA molecules maintained in heterologous host cells or purified (partially or substantially) DNA molecules in solution.

The isolated nucleic acid molecules can be in the form of RNA, such as mRNA, or in the form DNA, including, but not limited to, cDNA and genomic DNA, obtained by cloning or produced by chemical synthesis techniques or a combination thereof. The nucleic acid, especially DNA, can be double-stranded or single-stranded. Single-stranded nucleic acid molecules can be the coding strand (sense strand) or the non-coding strand (anti-sense strand). The preferred forms of nucleic acid molecules are the full-length cDNA molecules, genes and genomic clones. Full-length genes may be cloned from known sequences using any of a number of methods known in the art. For example, a method which employs XL- PCR (Perkin-Elmer, Foster City, CA) to amplify long pieces of DNA may be used. Other methods for obtaining full-length genes are known in the art.

The invention further provides nucleic acid molecules that encode active fragments of a BCATc peptide of the present invention. A fragment comprises a contiguous nucleotide sequence greater than about 12 or more nucleotides. Preferably, the fragment comprises at least about 30, about 40, about 50, about 100, about 250 or about 500 nucleotides in length. The length of the fragment will be based on its intended use. For example, the fragment may encode epitope-bearing regions of the peptide or may be useful as a DNA probe or primer. Such fragments may be isolated using the known nucleotide sequence to synthesize an oligonucleotide probe. The probe may then be labeled and used to screen a cDNA library, genomic DNA library or mRNA library to isolate a nucleotide sequence corresponding to the coding region. Further, primers may be used in PCR reactions to clone specific regions of a gene. A probe or primer typically comprises a purified oligonucleotide or oligonucleotide pair. The oligonucleotide typically comprises a nucleotide sequence that hybridizes under stringent conditions (described below) to at least about 12, about 20, about 25, about 40 or about 50 or more consecutive nucleotides. The isolated nucleic acid molecules, including variants described below, are useful as, for examples, probes, primers and chemical intermediates, and in, for example, biological assays. For example, the nucleic acid molecules are useful as hybridization probes for cDNA and genomic DNA to isolate full-length cDNA and genomic clones encoding a BCATc peptide described herein and to isolate cDNA and genomic clones that correspond to variants (alleles, orthologs, etc.) producing the same or related BCATc peptides described herein. A probe can correspond to any sequence along the entire length of the nucleic acid molecule provided in, for example, SEQ ID NO:1 or 4. Accordingly, it could be derived from the 5' non-coding region(s), the coding region(s) and/or the 3' non-coding region(s). Alternatively, the nucleic acid molecules may be used as primers for PCR to amplify any given region of a nucleic acid molecule or may be used to synthesize antisense molecules of any desired length and sequence. Alternatively, the nucleic acid molecules may be used as probes for determining the chromosomal positions of a nucleic acid molecule by means of in situ hybridization methods.

In addition, the isolated nucleic acid molecules, including variants described below, may be used as hybridization probes for determining the presence, level, form and distribution of nucleic acid expression. Accordingly, the probes can be used to detect the presence of, or to determine levels of, a specific nucleic acid molecule in cells, tissues and organisms. The nucleic acid molecule whose level is determined can be DNA or RNA. Accordingly, probes corresponding to the BCATc peptides described herein can be used to assess expression and/or gene copy number in a given cell, tissue or organism. These uses are relevant for the diagnosis of disorders involving an increase or decrease in BCATc expression relative to normal results.

In addition, the isolated nucleic acid molecules, including variants described below, may be used as probes as part of a diagnostic test kit for identifying cells or tissues that express BCATc, such as by measuring the level of a receptor-encoding nucleic acid, e.g., mRNA or genomic DNA, in a sample of cells from a subject, or by determining if a receptor gene has been mutated. In vitro techniques for detecting mRNA include, but are not limited to, Northern hybridization and in situ hybridization. In vitro techniques for detecting DNA include, but are not limited to, Southern hybridization and in situ hybridization.

The isolated nucleic acid molecular sequences, including variant sequences described below, can further be used as "query sequences" to perform a search against sequence databases to, for example, identify other family members or related sequences. Such searches can be performed using commercially available search engines such as the BLASTN and BLASTX programs (version 2.0) of Altschul et al. (J. Mol. Biol. 215:403-10 (1990)). BLAST nucleotide searches can be performed with the BLASTN program, score = 100, word length = 12, to obtain nucleotide sequences homologous to the isolated nucleic acid molecules. To obtain gapped alignments for comparison purposes, Gapped BLAST can be utilized as described by Altschul et al. (Nucleic Acids Res. 25: 3389-3402 (1997)). When using BLAST programs, the default parameters of the respective programs, e.g., BLASTN and BLASTX, can be used.

BCATc Amino Acid Sequences The published amino acid sequence (amino acids 1-384) of the BCATc peptide is set forth in SEQ ID NO:2. A nearly identical sequence is set forth in SEQ ID NO:5. Both sequences contain an optional amino terminal His6 tag (amino acid residues -23 to -4 of SEQ ID NOS: 2 and 5) and an optional thrombin cleavage sequence (MAC) (amino acid residues - 3 to -1 of SEQ ID NOS: 2 and 5). The sequence of the His6 tag is MGSSHHHHHHSSGLVPRGSH (SEQ ID NO: 3). SEQ ID NO: 5 is identical to the published sequence (SEQ ID NO: 2) except for a single amino acid change (K366E) (corresponds to nucleotides 1096-1098 in SEQ ID NOS: 1 and 4). It will be readily apparent to those of skill in the art that the numbering of amino acids in other forms, e.g., isoforms, of BCATc may be different than that set forth herein. Corresponding amino acids in other forms of BCATc are easily identified by inspection of the amino acid sequences, for example, through the use of commercially available homology software programs. BCATc variants are also described below. As used herein, the term "peptide" refers to any peptide, polypeptide or protein comprising two or more amino acids joined to each other by peptide bonds or modified peptide bonds, i.e., peptide isosteres. The term "peptide" refers to both short chains, generally referred to as peptides, oligopeptides or oligomers, and to longer chains, generally referred to as proteins. The terms "peptide," "polypeptide" and "protein" are used interchangeably herein. Peptides may contain amino acids other than the 20 naturally occurring amino acids. Further, many amino acids in the peptide, including the terminal amino acids, may be modified by natural processes, such as processing and other post-translational modifications, or by chemical modification techniques well-known in the art. Common modifications that occur naturally in peptides are described in basic texts, detailed monographs and the literature, and are well-known to those of skill in the art. See, e.g., Wold, F., Posttranslational Covalent Modification of Proteins, B.C. Johnson (ed.), Academic Press, N.Y., pages 1-12 (1983); Seifter et al., Meth. Enzymol. 182: 626-646 (1990); Rattan et al., Ann. N.Y. Acad. Sci. 663:48-62 (1992); and Proteins - Structure and Molecular Properties, 2nd Ed., T.E. Creighton, W.H. Freeman and Company, N.Y. (1993).

Known modifications include, but are not limited to, acetylation, acylation, ADP- ribosylation, amidation, covalent attachment of flavin, covalent attachment of a heme moiety, covalent attachment of a nucleotide or nucleotide derivative, covalent attachment of a lipid or lipid derivative, covalent attachment of phosphotidylinositol, cross-linking, cyclization, disulfide bond formation, demethylation, formation of covalent crosslinks, formation of cystine, formation of pyroglutamate, formylation, gamma carboxylation, glycosylation, GPI anchor formation, hydroxylation, iodination, methylation, myristoylation, oxidation, proteolytic processing, phosphorylation, phenylation, racemization, selenoylation, sulfation and transfer- RNA-mediated addition of amino acids to proteins such as arginylation and ubiquitination. General modifications such as additional heavy atom derivatives common in X-ray crystallographic studies may also be performed.

The present invention further provides for peptides that consist of, consist essentially of or comprise active fragments of a BCATc peptide. As used herein, a fragment comprises at least about 8 or more contiguous amino acid residues from BCATc or a protein structurally related to BCATc. Such fragments are chosen based on the ability to retain one or more of the biological activities of BCATc or for the ability to perform an activity, e.g., act as an immunogen. Particularly important fragments are catalytically active fragments, e.g., peptides which are, for example, about 8 or more amino acids in length. Such fragments typically comprise at least a part of a domain or motif of a BCATc peptide, e.g., active site or binding site. Additional fragments contemplated by the present invention include, but are not limited to, domain or motif containing fragments, soluble peptide fragments and fragments containing immunogenic characteristics. The BCATc peptides, including variants described below, can be attached to heterologous sequences to form chimeric or fusion proteins. In general, the two peptides linked in a fusion peptide are typically derived from two independent sources and therefore, a fusion peptide comprises two linked peptides not normally found linked in nature. The two peptides may be from the same or different genomes. More specifically, a chimeric or fusion protein comprises a desired peptide operatively linked to a heterologous peptide having an amino acid sequence not substantially homologous to the desired peptide. The phrase "operatively linked" indicates that the peptides are fused in-frame. For example, the heterologous peptide can be fused to the N-terminus or C-terminus of the desired peptide. In some cases, the fusion protein does not affect the activity of the desired peptide per se. For example, the fusion protein may include enzymatic fusion proteins, for example, beta- galactosidase fusions, yeast two-hybrid GAL fusions, poly-His fusions, MYC-tagged fusions, Hl-tagged fusions and Ig fusions. In certain host cells, e.g., mammalian host cells, expression and/or secretion of a protein can be increased by using a heterologous signal sequence fused thereto.

A chimeric or fusion protein can be produced by standard recombinant DNA techniques. For example, DNA fragments coding for the different amino acid sequences are ligated together in-frame in accordance with conventional techniques. In another embodiment, the fusion protein can be synthesized by conventional techniques including, but not limited to, automated DNA synthesizers. Alternatively, PCR amplification of gene fragments can be carried out using anchor primers which give rise to complementary overhangs between two consecutive gene fragments which can subsequently be annealed and re-amplified to generate a chimeric gene sequence. See, e.g., Ausubel et al., Current Protocols in Molecular Biology (1992). Moreover, many expression vectors are commercially available that already encode a fusion moiety, e.g., a GST protein.

Accordingly, BCATc peptides, including variants described below, encompass, for example, derivatives or analogs in which a substituent group is included such as a substituted amino acid residue coded or not coded by the genetic code; in which the mature polypeptide is fused with another compound such as a compound to increase the half-life of the polypeptide, e.g., polyethylene glycol; and in which the additional amino acids are fused to the mature polypeptide such as a leader or secretory sequence or a sequence for purification of the mature polypeptide or a pro-protein sequence.

The amino acid sequences encoding the BCATc peptides, including variants described below, can be optionally used as "query sequences" to perform a search against sequence databases to, for example, identify other family members or related sequences. Such searches can be performed using commercially available search engines such as the BLASTP and BLASTX programs (version 2.0) of Altschul et al. (J. Mol. Biol. 215:403-10 (1990)). BLAST protein searches can be performed with the BLASTP program, score = 50, word length = 3, to obtain amino acid sequences homologous to the peptides of the invention. To obtain gapped alignments for comparison purposes, Gapped BLAST can be utilized as described by Altschul et al. (Nucleic Acids Res. 25: 3389-3402 (1997)). When utilizing BLAST programs, the default parameters of the respective programs, e.g., BLASTP and BLASTX, can be used.

BCATc Polynucleotide and Peptide Variants

The BCATc peptides encompass nucleic acid molecules that encode active variants of the BCATc peptides and the active variants so encoded. Such nucleic acid molecules and resultant peptides may be naturally occurring, such as allelic variants (same locus), homologs, paralogs (different locus) or orthologs (different organism), or may be constructed by recombinant DNA methods or by chemical synthesis. The non-naturally occurring variants may be made by mutagenesis techniques, including, but not limited to, those applied to nucleic acid molecules, cells or organisms. Accordingly, the variants can contain nucleotide substitutions, deletions, inversions and/or insertions. Variation can occur in either or both the coding and non-coding regions. The variations can produce both conservative and non- conservative amino acid substitutions.

Variant BCATc peptides can be fully functional or can lack function in one or more activities. Fully functional variants typically contain only conservative variation or variation in non-critical residues or in non-critical regions. Functional variants can also contain substitution of similar amino acids which results in no change or in an insignificant change in function. Non-functional variants typically contain one or more non-conservative amino acid substitutions, deletions, insertions, inversions or truncations or a substitution, deletion, insertion, inversion or truncation in a critical residue or critical region. Amino acids that are essential for function can be identified by methods known in the art such as site-directed mutagenesis or alanine-scanning mutagenesis (Cunningham et al., Science 244:1081-1085 (1989)). The latter procedure introduces single alanine mutations at every residue in the molecule. The resulting modified molecules are then tested for biological activity using, for example, receptor binding or in vitro proliferative activity assays. Sites that are critical for binding can also be determined by structural analysis such as crystallography, NMR or photoaffinity labeling (Smith et al., J. Mol. Biol. 224:899-904 (1992); de Vos et al. Science 255:306-312 (1992)).

Paralogs, orthologs, homologs and allelic variants can be identified using methods known in the art. These variants comprise a nucleotide sequence encoding a peptide that is typically about 70%, preferably, about 75%, even more preferably, about 80% and, most preferably, about 85% or more homologous to the nucleotide sequence provided in SEQ ID NO:1 or 4 or a fragment thereof. In a preferred embodiment, the variants comprise a nucleotide sequence encoding a peptide that is typically about 90% and, most preferably, about 95% or more homologous to the nucleotide sequence provided in SEQ ID NO:1 or 4 or a fragment thereof. Such nucleic acid molecules can be readily identified as being able to hybridize under moderate to stringent conditions to the nucleotide sequence shown in SEQ ID NO:1 or 4 or a fragment thereof.

As used herein, the phrase "hybridize under moderate or stringent conditions" is intended to describe conditions for hybridization and washing under which nucleotide sequences encoding peptides at least about 50%, preferably, at least about 55%, homologous to each other typically hybridize to each other. The conditions can be such that sequences at least about 65%, preferably, at least about 70%, and, more preferably, at least about 75% or more homologous to each other typically hybridize. Such stringent conditions are known to those skilled in the art and can be found in, for example, Current Protocols in Molecular Biology, John Wiley & Sons, N.Y. (1989), 6.3.1-6.3.6. Moderate hybridization conditions are defined as equivalent to hybridization in 2X sodium chloride/sodium citrate (SSC) at 30oC, followed by one or more washes in 1 X SSC, 0.1% SDS at about 50oC to about 60oC. Highly stringent hybridization conditions are defined as equivalent to hybridization in 6X sodium chloride/sodium citrate (SSC) at about 45oC, followed by one or more washes in 0.2 X SSC, 0.1% SDS at about 50oC to about 65oC.

Paralogs have some degree of significant sequence homology, i.e., identity, to at least a portion of a given peptide, are encoded by a gene from the same species and have similar activity or function. Two peptides are typically considered paralogs when the amino acid sequences are at least about 70%, preferably, about 75%, more preferably, about 80%, and, even more preferably, about 85% or more homologous through a given region or domain. In one preferred embodiment, two peptides are typically considered paralogs when the amino acid sequences are about 90% or more, and, preferably, about 95% or more homologous through a given region or domain. A paralog of a BCATc peptide is encoded by a nucleic acid molecule that hybridizes to a BCATc peptide encoding nucleic acid molecule under stringent conditions.

Orthologs have some degree of significant sequence homology to at least a portion of a given peptide and are encoded by a gene from another organism. Preferred orthologs are isolated from mammals, preferably humans, for the development of human therapeutic targets and agents, or other invertebrates, particularly insects of economical and/or agriculture importance, e.g., members of the Lepidopteran and Coleopteran orders, for the development of insecticides and insecticidal targets. An ortholog of a BCAT peptide is encoded by a nucleic acid molecule that hybridizes to a BCATc peptide encoding nucleic acid molecule under moderate to stringent conditions depending on the degree of relatedness of the two organisms yielding the proteins. Allelic variants have a high degree of sequence homology to at least a portion of a given peptide. As used herein, two allelic variants (or regions thereof) have significant homology when the amino acid sequences are typically at least about 70%, preferably, at least about 75%, more preferably, at least about 80%, even more preferably, at least about 85%), even more preferably, at least about 90% and, most preferably, at least about 95% or more homologous. A significantly homologous amino acid sequence is encoded by a nucleic acid molecule that hybridizes to a BCATc peptide encoding nucleic acid molecule under stringent conditions.

Non-naturally occurring variants of BCATc peptides can be readily generated using recombinant techniques. Such variants can be readily identified and/or created using molecular techniques and the sequence information disclosed herein. The degree of homology present is based primarily on whether the peptide is a functional variant or nonfunctional variant, the amount of divergence present in the paralog family and/or the evolutionary distance between the orthologs. Such variants include, but are not limited to, peptides having amino acid sequences containing, for example, deletions, additions and substitutions in the amino acid sequence of the BCATc peptide, e.g., SEQ ID NO: 2 or 5. For example, one type of substitution is a conserved amino acid substitution. Such substitutions are those that substitute a given amino acid in the known BCATc peptide amino acid sequence with another amino acid of like characteristics. Typically seen as conservative substitutions are substitution (or replacement), one for another, among the aliphatic amino acids alanine, valine, leucine and isoleucine; interchange of the hydroxyl residues serine and threonine; exchange of the acidic residues aspartic acid and glutamic acid; substitution between the amide residues asparagine and glutamine; exchange of the basic residues lysine and arginine; and substitution of aromatic residues phenylalanine and tyrosine one for the other. Guidance concerning which amino acid changes are likely to be phenotypically silent is found in Bowie et al., Science 247:1306-1310 (1990).

To determine the homology of two amino acid sequences or two nucleic acid sequences, the sequences are aligned for optimal comparison purposes, e.g., gaps can be introduced in one or both of a first and a second amino acid or nucleic acid sequence for optimal alignment, and non-homologous sequences can be disregarded for comparison purposes. In a preferred embodiment, the length of a sequence aligned for comparison purposes is at least about 30%, about 40%, about 50%, about 60%, about 70%, about 80% or about 90% or more of the length of the reference sequence. The amino acid residues or nucleotides at corresponding amino acid or nucleotide positions are then compared. When a position in the first sequence is occupied by the same amino acid residue or nucleotide as the corresponding position in the second sequence, then the molecules are identical at that position (as used herein amino acid or nucleic acid "identity" is equivalent to amino acid or nucleic acid "homology"). The homology, or percent identity, between the two sequences is a function of the number of identical positions shared by the sequences, taking into account the number of gaps and the length of each gap introduced for optimal alignment of the two sequences. The comparison of sequences and determination of percent identity between two sequences can be accomplished using a mathematical algorithm (Computational Molecular

Biology, Lesk, A.M. (ed.), Oxford University Press, N.Y. (1988); Biocomputing: Informatics and Genome Projects, Smith, D.W. (ed.), Academic Press, N.Y. (1993); Computer Analysis of

Sequence Data, Part 1 , Griffin, A.M., and Griffin, H.G. (eds.), Humana Press, N.J. (1994); Sequence Analysis in Molecular Biology, von Heinje, G., Academic Press, N.Y. (1987); and

Sequence Analysis Primer, Gribskov, M. and Devereux, J. (eds.), M. Stockton Press, N.Y.

(1991)). In a preferred embodiment, the percent identity between two amino acid sequences is determined using the Needleman and Wunsch algorithm (J. Mol. Biol. 48:444-453 (1970)) which has been incorporated into commercially available computer programs such as GAP in the Accelrys Inc. software package (Devereux, J., et al., Nucleic Acids Res. 12:387 (1984)) using either a Blossom 62 matrix or a PAM250 matrix, and a gap weight of 16, 14, 12, 10, 8, 6 or 4 and a length weight of 1 , 2, 3, 4, 5 or 6.

In another preferred embodiment, the percent identity between two nucleotide or amino acid sequences is determined using commercially available computer programs including, but not limited to, the NWS gap DNA CMP matrix with a gap weight of 40, 50, 60,

70 or 80 and a length weight of 1 , 2, 3, 4, 5 or 6.

In another preferred embodiment, the percent identity between two amino acid or nucleotide sequences is determined using the algorithm of Meyers and Miller (CABIOS 4:11-

17 (1989)) which has been incorporated into commercially available computer programs such as ALIGN (version 2.0) using a PAM120 weight residue table, a gap length penalty of 12 and a gap penalty of 4.

An alternative method to using the primary sequence for describing the structural relationship between two peptides is to use the three-dimensional structures of the peptides.

In this method, two structures are resolved by crystallography or by NMR, and then the similarity is determined by comparing the RMS deviation of the C-α backbone trace of the two corresponding peptides.

BCATc Peptide Isolation

The present invention further provides methods of obtaining BCATc peptides. The peptides are comprised of, consist of or consist essentially of the amino acid sequences of the peptides encoded by the nucleic acid sequences disclosed in SEQ ID NO:1 or 4, for example, as well as all obvious variants of these peptides that are within the art to make and use. The present invention further provides methods for expressing BCATc peptides comprising culturing a cell that expresses a BCATc peptide in an appropriate cell culture medium under conditions that provide for peptide expression by the cell. Any of the cells mentioned below or known in the art may be employed in this method. The BCATc peptides of the present invention can be isolated and purified to homogeneity or other degrees of purity. The level of purification will be based on the intended use of the peptide. The important point regarding purity is that the peptide is able to function in the desired manner, even if considerable amounts of other components are present. However, BCATc peptide preparations are preferably substantially free of other cellular material, i.e., have less than about 30% (by dry weight) of other proteins, e.g., contaminating proteins, more preferably, less than about 20% of other proteins, even more preferably, less than about 10% of other proteins and, most preferably, less than about 5% of other proteins. When a BCATc peptide is recombinantly produced, it is preferably substantially free of culture medium, i.e., culture medium represents less than about 5% of the volume of the protein preparation.

In addition, BCATc peptide preparations are preferably substantially free of chemical precursors or other chemicals, e.g., chemical precursors or other chemicals involved in protein synthesis, having less than about 30% (by dry weight) of chemical precursors or other chemicals, preferably, less than about 20% of chemical precursors or other chemicals, more preferably, less than about 10% of chemical precursors or other chemicals, and, most preferably, less than about 5% of chemical precursors or other chemicals.

Once BCATc has been isolated from a mammalian source, preferably, a human, it may be purified from cells that naturally express it or purified from cells that have been altered to express it, e.g., recombinant cells. For example, a nucleic acid molecule encoding a BCATc peptide may be cloned into an expression vector, the expression vector introduced into a host cell and the peptide expressed using the host cell's expression mechanism(s). The peptide may then be isolated from the host cell by an appropriate purification scheme using standard protein purification techniques. For example, in Example 1 below, BCATc was expressed in E. coli cells and purified by a combination of IMAC, SEC and anion-exchange chromatography using routine methods in the art. Methods of expressing peptides via expression vectors and host cells are well-within the purview of the skilled artisan. Similarly, protein purification using chromatography or any known protein purification technique is well- within the purview of the skilled artisan.

As used herein, the term "vector" refers to a vehicle, preferably, a nucleic acid molecule, that can transport nucleic acid molecules. The invention provides vectors for the maintenance (cloning vectors) or expression (expression vectors) of BCATc encoding nucleic acid molecules. The vectors can function in prokaryotic or eukaryotic cells or in both (shuttle vectors). Moreover, a vector can be maintained in a host cell as an extrachromosomal element where it replicates and produces additional copies of the desired nucleic acid molecule. Alternatively, the vector can integrate into the host cell genome and produce additional copies of the desired nucleic acid molecule when the host cell replicates. In a preferred embodiment, the desired nucleic acid molecule is operatively associated with an expression control sequence.

A variety of expression vectors can be used to express a desired nucleic acid molecule. Such vectors include, but are not limited to, chromosomal, episomal and viral vectors, for example, vectors derived from bacterial plasmids, bacteriophages, yeast episomes, yeast chromosomal elements, including, but not limited to, yeast artificial chromosomes, and viruses such as baculoviruses, papovaviruses such as SV40, vaccinia viruses, adenoviruses, poxviruses, pseudorabies viruses and retroviruses. Vectors may also be derived from combinations of these sources such as those derived from plasmid and bacteriophage genetic elements, e.g., cosmids and phagemids. Appropriate cloning and expression vectors for prokaryotic and eukaryotic hosts are described in, for example, Sambrook et al., Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. (1989).

When the vector is a nucleic acid molecule, the desired BCATc nucleic acid molecule may be covalently linked to the vector nucleic acid molecule. Such vectors include, but are not limited to, an episomal vector, plasmid, single- or double-stranded phage, single- or double-stranded RNA or DNA viral vector or artificial chromosome, such as BAG, PAC, YAC or MAC (Qiagen, Valencia, CA). Viral vectors can be replication-competent or replication- defective wherein replication occurs in host cells providing functions that complement the defects. Expression vectors contain cis-acting regulatory regions that are operably linked to the desired nucleic acid molecule such that transcription of the desired nucleic acid molecule occurs in the host cell. Optionally, the desired nucleic acid molecule may be introduced into the host cell with a separate nucleic acid molecule capable of affecting transcription. Thus, the second nucleic acid molecule may provide a trans-acting factor interacting with the cis- regulatory control to allow transcription of the desired nucleic acid molecule. Alternatively, a trans-acting factor may be supplied by the host cell or produced by the vector.

When two nucleic acid molecules are present, they can be introduced on different vectors in the same cell. The desired nucleic acid molecule may be introduced either alone or with nucleic acid molecules that are not related to the desired nucleic acid molecules such as those providing trans-acting factors for expression vectors. When more than one vector is introduced into a cell, the vectors can be introduced independently, co-introduced or joined. The regulatory sequence to which the desired nucleic acid molecule may be operably linked includes, but is not limited to, a promoter for directing mRNA transcription, including, but not limited to, the left promoter from bacteriophage λ, the lac, TRP and TAC promoters from E. coli, the early and late promoters from SV40, the CMV immediate early promoter, the adenovirus early and late promoters and retrovirus long-terminal repeats (LTRs).

In addition to regions that promote transcription, expression vectors may contain regions that modulate transcription, such as repressor binding sites and enhancers. Examples include, but are not limited to, the SV40 enhancer, the cytomegalovirus immediate early enhancer, the polyoma enhancer, adenovirus enhancers and retrovirus LTR enhancers. In addition to containing sites for transcription initiation and control, expression vectors may contain sequences necessary for transcription termination and, in the transcribed region, a ribosome binding site for translation. Other regulatory control elements for expression include, but are not limited to, initiation and termination codons and polyadenylation signals. A person of ordinary skill in the art would be aware of the numerous regulatory sequences that are useful in expression vectors. Such regulatory sequences are described, for example, in Sambrook et al., Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. (1989).

The regulatory sequence may provide constitutive expression in one or more host cells, i.e., tissue specific, or may provide for inducible expression in one or more cell types such as by temperature, nutrient additive or exogenous factor such as the presence of a hormone. A variety of vectors providing for constitutive and inducible expression in prokaryotic and eukaryotic hosts are known to those of ordinary skill in the art.

The desired nucleic acid molecule may be inserted into the vector nucleic acid molecule by well-known methodology. Generally, the DNA sequence that will ultimately be expressed is joined to an expression vector by cleaving the DNA sequence and the expression vector with one or more restriction enzymes and then ligating the fragments together. Procedures for restriction enzyme digestion and ligation are well-known to those of ordinary skill in the art.

The invention also encompasses vectors wherein the desired nucleic acid molecule is cloned into a vector in reverse orientation, but is operably linked to a regulatory sequence that permits transcription of antisense RNA. Thus, an antisense transcript can be produced to all, or to a portion, of the desired nucleic acid molecule, including both coding and non-coding regions. Expression of the antisense RNA is subject to each of the parameters described above, e.g., regulatory sequences, constitutive or inducible expression or tissue-specific expression.

Vectors generally include selectable markers that enable the selection of a subpopulation of cells that contain the recombinant vector constructs. The marker may be found on the vector that contains the desired nucleic acid molecule or on a separate vector. Markers include, but are not limited to, tetracycline or ampicillin-resistance genes for prokaryotic host cells and dihydrofolate reductase or neomycin resistance for eukaryotic host cells. However, any marker that selects for a phenotypic trait will be effective. Alternatively, where secretion of the BCATc peptide is desired, appropriate secretion signals are incorporated into the vector. The signal sequence can be endogenous or heterologous.

Nucleic acid molecules may be expressed in yeast such as S. cerevisiae. Examples of vectors for expression in yeast include, but are not limited to, pYepSed (Baldari et al., EMBO J. 6:229-234 (1987)), pMFa (Kurjan et al., Cell 30:933-943 (1982)), pJRY88 (Schultz et al., Gene 54:113-123 (1987)) and pYES2 (Invitrogen Corp., San Diego, CA).

Nucleic acid molecules may also be expressed in bacterial cells, such as E. coli, Streptomyces or Salmonella such as Salmonella typhimurium. Examples of suitable inducible E. coli expression vectors include, but are not limited to, pTrc (Amann et al., Gene 69:301-315 (1988)), pET11d (Studier et al., Gene Expression Technology: Methods in Enzymology 785:60- 89 (1990)) and pET28a (Novagen, Madison, WI).

Nucleic acid molecules may also be expressed in insect cells such as Drosophila cells using, for example, baculovirus expression vectors. Baculovirus vectors available for expression of proteins in cultured insect cells, e.g., Sf9 cells, include, but are not limited to, vectors of the pAc series (Smith et al., Mol. Cell Biol. 3:2156-2165 (1983)) and the pVL series (Lucklow et al., Virology 170:31-39 (1989)).

It may be desirable to express the nucleic acid molecules in mammalian cells using mammalian expression vectors. Examples of mammalian expression vectors include, but are not limited to, pCDM8 (Seed, B., Nature 329:840 (1987)) and pMT2PC (Kaufman et al., EMBO J. 6:187-195 (1987)). The expression vectors cited herein are provided by way of example only. Preferred vectors include, but are not limited to, pET28a (Novagen), pET24b (Novagen), pAcSG2 Baculovirus Transfer Vector (Pharmingen, San Diego, CA), and pFastBac (Invitrogen Corp.). A person of ordinary skill in the art would be aware of other vectors suitable for maintenance propagation or expression of the nucleic acid molecules described herein. Examples are found in, for example, in Sambrook, J. et al., Molecular Cloning: A Laboratory Manual, 2nd, ed., Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. (1989).

Alternatively, it may be desirable to express the desired nucleic acid molecule as a fusion protein. Accordingly, the invention provides fusion vectors that allow for the production of such proteins. Fusion vectors can increase the expression of the desired protein, increase the solubility of the desired protein and aid in the purification of the desired protein by acting, for example, as a ligand for affinity purification. A proteolytic cleavage site may be introduced at the junction of the fusion moiety so that the desired protein can ultimately be separated from the fusion moiety. Proteolytic enzymes include, but are not limited to, factor Xa, thrombin and enterokinase. Typical fusion expression vectors include, but are not limited to, pRS (Sikorski et al., Genetics 122: 19-27 (1989)), pGEX (Smith et al., Gene 67:31-40 (1988)), pMAL (New England Biolabs, Beverly, MA) and pRIT5 (Pharmacia, Piscataway, N.J.) which fuse glutathione S-transferase (GST), maltose E binding protein and/or protein A to the target recombinant protein.

The vector containing the desired nucleic acid molecule may be introduced into an appropriate host cell for propagation or expression using well-known techniques. Bacterial host cells include, but are not limited to, E. coli, Streptomyces and S. typhimurium. Eukaryotic host cells include, but are not limited to, yeast such as S. cerevisiae, insect cells such as Drosophila cells and Sf9 cells, mammalian cells such as animal cells, for example, COS and CHO cells, and plant cells. Thus, the invention also relates to recombinant host cells containing the vectors described herein. Preferred host cells of the present invention include, but are not limited to, prokaryotic cells such as E. coli cells and eukaryotic cells such as Sf9 cells. Preferred E. coli cell sources include, but are not limited to, E. coli strains BL21 (DE3), JM109 and DH5 alpha.

The recombinant host cells may be prepared by introducing the vector constructs described above into the cells by techniques known to a person of ordinary skill in the art as described above. Other known techniques include, but are not limited to, calcium phosphate transfection, DEAE-dextran-mediated transfection, cationic lipid-mediated transfection, electroporation, transduction infection, lipofection and other techniques such as those found in Sambrook, J. et al., Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. (1989). In the case of bacteriophage and viral vectors, these can be introduced into cells as packaged or encapsulated virus by standard procedures for infection and transduction. Recombinant protein expression can be maximized in a host bacterial cell by providing a genetic background wherein the host cell has an impaired capacity to proteolytically cleave the recombinant protein (Gottesman, S., Gene Expression Technology: Methods in Enzymology 185: 119-128, Academic Press, San Diego, CA (1990)). Alternatively, the sequence of the desired nucleic acid molecule can be altered to provide preferential codon usage for a specific host cell, for example E. coli (Wada et al., Nucleic Acids Res. 20:2111-21 18 (1992)).

Depending upon the host cell employed in the recombinant production of a BCATc peptide, the peptide can have various glycosylation patterns or even be non-glycosylated, as when produced in bacteria. In addition, the peptide may include an initial modified methionine possibly as a result of a host-mediated process. Host cells of the present invention expressing the BCATc peptide have a variety of uses. For example, the cells are useful for producing BCATc or a protein structurally related to BCATc which can be further purified to produce desired amounts of BCATc or a protein structurally related to BCATc or fragments of any of the above. Thus, host cells containing expression vectors are useful for peptide production.

Host cells of the prevention invention are also useful for conducting cell-based assays involving a BCATc peptide or fragments thereof. For example, a recombinant host cell expressing BCATc is useful for assaying ligands that inhibit or enhance BCATc peptide activity. Host cells of the present invention are also useful for identifying BCATc modifications in which BCATc activities are affected. If the modifications naturally occur and give rise to a pathology, host cells containing the modified BCATc are useful to assay ligands that have an effect on the modified BCATc, for example, inhibiting or enhancing BCATc activity, which may not be indicated by their effect on native BCATc. Genetically engineered host cells of the present invention can also be used to produce non-human transgenic animals. A transgenic animal is preferably a mammal, for example, a rodent such as a rat or mouse, in which one or more of the cells of the animal include a transgene. A transgene is exogenous DNA which is integrated into the genome of a cell from which a transgenic animal develops and which remains in the genome of the mature animal in one or more cell types or tissues. These animals are useful for studying the function of BCATc and identifying and evaluating modulators of its activity. Other examples of transgenic animals include, but are not limited to, non-human primates, sheep, dogs, cows, goats, chickens and amphibians.

While the BCATc peptide can be expressed in bacterial, yeast, mammalian and other cells under the control of appropriate regulatory sequences, cell-free transcription and translation systems may also be used to produce the desired peptide using RNA derived from BCATc DNA constructs. Such techniques are well-known in the art. BCATc Peptide Purification The purification conditions and methods listed herein are provided to elucidate the approach used in the purification of BCATc peptides. Those of ordinary skill in the art would be aware of other purification conditions and techniques that may be suitable for the purification of BCATc peptides. See, e.g., Methods in Enzymology, Volume 182; Guide to Protein Purification, M.P. Duetscher (ed.), Academic Press (1990).

The invention provides a multi-step method for purifying isolated BCATc peptides to near homogeneity. The BCATc peptide, preferably with a poly-histidine tag at the NH2- terminus, may be purified by employing immobilized metal affinity chromatography (IMAC). The immobilized metal may be nickel, zinc, cobalt or copper. The IMAC step may be accomplished with the following resins, including, but not limited to: nickel affinity columns such as Ni-NTA Superflow (Qiagen), HisTrap® (Pharmacia, Peapack, N.J.), Poros® MC

(PerSeptive Biosystems, Inc., Foster City, CA) or TALON® (Clontech, Palo Alto, CA).

Preferably, Ni-NTA Superflow (Qiagen) is used. The IMAC step may be performed on the soluble fraction of a fermentation broth containing the expressed BCATc peptide obtained after host cell lysis. Host cell lysis processes are known to those skilled in the art. For example, when E. coli is used as host cell, E. coli cell lysis may be enzymatically performed, for example, by using lysozyme.

Alternatively, mechanical processes such as using a french press, sonicator or glass bead mill, may be employed. Preferably, mechanical lysis using a glass bead mill (Dyno-Mill KDL)

(Glenn Mills Inc., Clifton, NJ.) is used to perform E. coli cell lysis.

The BCATc-binding and contaminant removal steps of the purification process may be performed in the presence of any one or more suitable buffering agent(s), including, but not limited to, Tris (tris(hydroxymethyl)nitromethane), HEPES (N-(2-hydroxyethyl)piperazine- N'-2-ethanesulfonic acid), potassium phosphate, citrate phosphate, sodium phosphate and

MOPS (3-(N-morpholino)propanesulfonic acid). Preferably, the buffering agent is Tris at a concentration of about 50 mM with a pH range of about 7.0 to about 9.0, and, more preferably, a pH of about 8.0.

The binding and contaminant removal steps of the purification process may be performed in the presence of any one or more suitable contaminating protein-displacing agent(s), including, but not limited to, imidazole and histidine. Preferably, the contaminating protein-displacing agent is imidazole at a concentration of between about 5.0 mM and about

25 mM, and, more preferably, of about 5.0 mM.

The binding and contaminant removal steps of the purification process may be performed in the presence of any one or more suitable reducing agent(s), including, not limited to, 2-mercaptoethanol and TCEP (Tris-(2-carboxyethyl)phosphine, hydrochloride).

Preferably, the reducing agent is 2-mercaptoethanol at a concentration of about 10 mM.

The binding and contaminant removal steps of the purification process may be performed in the presence of one or more detergent(s), including, but not limited to, CHAPS (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate), Triton® X-100 (octylphenoxy polyethoxyethanol), Tween® 20 (ICI American Holdings, Inc., Bridgewater, NJ.) and Tween®

80 (ICI American Holdings, Inc.). Preferably, the detergent is Triton® X-100 at a concentration of between about 0.1 % and about 1.0%.

The binding and contaminant removal steps of the purification process may be performed in the presence of one or more ion source(s), including, but not limited to, KCI,

NaCI or sodium sulfate. Preferably, the ion source is NaCI at a concentration of between about 0.2 M and about 0.5 M. The elution of the BCATc peptide from the column may be accomplished by -several modes known in the art. For example, elution of the BCATc peptide may be accomplished by using EDTA ((ethylenedinitrilo)tetraacetic acid), histidine or imidazole, or by reducing the pH.

Preferably, elution of the BCATc peptide is accomplished with imidazole at a concentration of between about 150 mM and about 500 mM.

The elution step of the purification process may be performed in the presence of one or more suitable buffering agent(s), including, but not limited to, Tris, phosphate, HEPES and MOPS. Preferably, the buffering agent is Tris at a concentration of about 50 mM with a pH of about 8.0. The invention also provides a method for further purifying the BCATc peptide using size exclusion chromatography (SEC) and/or anion-exchange chromatography. The SEC step may be performed using various types of chromatography resins, including, but not limited to, Sephadex® G-100 (Pharmacia), Sephadex® G-200 (Pharmacia), Sephacryl® S- 100 (Pharmacia), Sephacryl® S-200 (Pharmacia), Superdex® 75 (Pharmacia) and Superdex® 200 (Pharmacia). Preferably, the SEC resin is Superdex® 200 (Pharmacia).

The SEC step may performed in the presence of one or more suitable buffering agent(s), including, but not limited to, phosphate, HEPES, MES (4-morpholineethanesulfonic acid), Tris, bis-Tris and bis-Tris propane. Preferably, the buffering agent is HEPES at a concentration of about 25 mM with a pH of about 7.5. The SEC step may also be performed in the presence of one or more suitable reducing agent(s), including, but not limited to, 2-mercaptoethanol, TCEP and DTT (dithiothreitol). Preferably, the reducing agent is DTT at a concentration of about 10 mM.

The SEC step may also be performed in the presence of one or more suitable salt(s), including, but not limited to, NaCI, KCI, ammonium acetate and sodium sulfate. Preferably, the salt is NaCI at a concentration of about 150 mM.

The SEC step may also be performed in the presence of one or more suitable chelating agent(s), including, but not limited to, sodium citrate and EDTA. Preferably, the chelating agent is EDTA at a concentration of about 5.0 mM.

The next step is to fully charge the BCATc with pyridoxal-5-phosphate (PLP). This is preferably accomplished by adding about 0.1 mM to about 10 mM PLP to the BCATc protein solution and incubating for about one minute to about seven days. Preferably, a 150 mM excess of PLP relative to the BCATc (typically about 1.0 mM PLP) is added and incubated for about one hour.

The next step, the anion-exchange step, if performed, may be performed using several types of chromatography resins, including, but not limited to, Q-Sepharose®

(Pharmacia), DEAE-Sepharose® (Pharmacia), Mono Q® (Pharmacia), Poros® HQ (PerSeptive Biosystems, Inc.) and Poros® PI (PerSeptive Biosystems, Inc.). Preferably, the exchange resin is Poros® HQ (PerSeptive Biosystems, Inc.).

The anion-exchange step may be performed in the presence of one or more suitable buffering agent(s), including, but not limited to, HEPES, Tris, bis-Tris, bis-Tris propane, N- methyldiethanolamine, 1 ,3-diaminopropane, ethanolamine, piperazine and phosphate.

Preferably, the buffering agent is Tris at a concentration of about 50 mM with a pH of about

8.6.

The anion-exchange step may be performed in the presence of one or more suitable reducing agent(s), including, but not limited to, 2-mercaptoethanol, TCEP and DTT. Preferably, the reducing agent is DTT at a concentration of about 10 mM.

Elution of the BCATc peptide from the anion-exchange resin may be accomplished by several different methods known in the art. For example, the BCATc peptide may be eluted by decreasing the pH. Alternatively, the BCATc peptide may be eluted by increasing the salt concentration by using, for example, NaCI, KCI, ammonium acetate or sodium sulfate. Preferably, a gradient NaCI from about 0.1 M to about 0.5 M is used to cause elution.

Overall, the preferred method of purification is IMAC chromatography using a Ni-NTA Superflow resin (Qiagen), followed by SEC using a Superdex® 200 resin (Pharmacia), followed by charging the BCATc peptide with PLP and conducting anion-exchange chromatography using a Poros® HQ resin (PerSeptive Biosystems, Inc.). The BCATc suitable for crystallization should be fully charged with PLP at the end of the above-defined process as indicated by a ratio of absorbance at 280 nm to 416 nm of 10:16.

The following examples illustrate preferred embodiments of the invention.

EXAMPLES Example 1 - Expression and Purification of Human BCATc

Human BCATc (SEQ ID NO:2) was expressed from the E. coli BL21 (DE3) clone described in Davoodi et al. (Journal of Biological Chemistry 273: 4982-4989 (1998)) using SEQ ID NO:4. The expression was initiated from a 20% glycerol stock stored at -80oC. One ml of the glycerol stock was used to inoculate 200 ml of LB media containing 10 mg/ml ampicillin in a 500 ml shake flask. The flask was incubated at 30oC in a shaking incubator at 250 RPM for 6-8 hours, until the culture became cloudy. One ml of the shake flask culture was used to inoculate a 15 liter working volume Biolafitte® fermentor (Cellex BioSciences, Inc., Coon Rapids, MN) which had been prepared with the following sterile media: 260 g yeast extract (Difco 0127), 260 g acidicase peptone (BBL 211843), 260 g casitone (Difco 0259), 260 g gelysate peptone (BBL 211870), 26 g K2HP04 (anhydrous), 26 g Na2HP04.7H20, and 10 g/l ampicillin (pH 6.5-6.8). The fermentation temperature was set at 30oC, the fermentor impeller speed was set at 600 RPM and air was sparged into the fermentor at about 8 liters per minute. pH was maintained at 6.8 + 0.2 by the addition of 85% lactic acid (USB 18140) and foam was controlled by the addition of Antifoam 289 (Sigma). Seventeen hours after induction, the turbidity reached an OD600 of 10. The fermentor was induced by the addition of IPTG (isopropyl-thiogalactopyranoside, AIC). Three hours after induction, the fermented material was harvested and the cell paste collected by centrifugation.

The cell paste (400 g wet weight) was resuspended to 1.3 I in 50 mM Tris (pH 8.0), 1.0% Triton® X-100 and 10 mM MgCI2. Forty μl of benzonase (EM Industries, Inc., Hawthorne, NJ.) was added and the cells were lysed by two passages through a Dyno-Mill Type KDL (Glenn Mills Inc.) using the 600 ml chamber and 500 ml of 0.25-0.5 mm glass beads. The impeller speed was 4200 RPM and the flow rate was 100 ml/min.

The lysate was clarified by centrifugation and the lysate supernatant was batch loaded on 50 ml of fresh Ni-NTA Superflow resin (Qiagen) for one hour at 4oC. The resin was then packed in a 5 cm diameter column and washed with 50 mM Tris (pH 8.0), 1.0% Triton® X-100, and 10 mM 2-mercaptoethanol. The column was next washed with 50 mM Tris/5.0 mM imidazole (pH 8.0), 0.5 M NaCI, and 10 mM 2-mercaptoethanol until the UV absorbance at 280 nm of the eluted solution reached a baseline. The column was subsequently washed with 50 mM Tris/20 mM imidazole (pH 8.0), 0.5 M NaCI, and 10 mM 2- mercaptoethanol until the UV absorbance at 280 nm of the eluted solution reached a baseline. Human BCATc was eluted from the column with 50 mM Tris/200 mM imidazole (pH 8.0), 0.5 M NaCI, and 10 mM 2-mercaptoethanol. The elution was monitored by UV absorbance at 280 nm. Denaturing gel electrophoresis (SDS-PAGE) was used to determine which fractions contained human BCATc and needed to be pooled.

The pooled human BCATc fractions were dialyzed into 25 mM HEPES, 5.0 mM EDTA (pH 7.5), 150 mM NaCI, and 10 mM DTT. Based on UV absorbance, 300 mg of protein was recovered at this stage.

Sixty-two mg of the above human BCATc was concentrated to 10 mg/ml using an Amicon® YM-10 stirred cell (Millipore Investment Holdings, Ltd., Wilmington, DE). The protein was loaded on a 3.2 x 90 cm column of Superdex®-200 (Pharmacia) which had been equilibrated with 25 mM HEPES, 5.0 mM EDTA (pH 7.5), 150 mM NaCI, and 10 mM DTT. The column was run at 1.5 ml/min with the above buffer and 6 ml fractions were collected. Fractions were checked by UV absorbance and SDS-PAGE, and those containing human BCATc were pooled and concentrated to 2.5 mg/ml using an Amicon® YM-10 stirred cell (Millipore Investment Holdings, Ltd.). 37.5 mg of human BCATc was recovered at this stage. The human BCATc above was diluted to 0.31 mg/ml using 50 mM Tris, (pH 8.6) and 10 mM DTT. The solution was adjusted to 1.0 mM with PLP and incubated for one hour at 4oC. The solution was then passed through a 0.22 micron filter and loaded onto an 18 ml POROS® 50 HQ column (PerSeptive Biosystems, inc.) equilibrated with 50 mM Tris (pH 5.6) and 10 mM DTT. The column was washed with 50 mM Tris (pH 8.6) and 10 mM DTT for 16 hours at a flow rate of 0.5 ml/min. The human BCATc was eluted from the column using a gradient of zero to 0.5 mM NaCI in 50 mM Tris (pH 8.6) and 10 mM DTT. The duration of the gradient was 120 minutes and the flow rate was 3 ml/min. Fractions were checked by UV and visible absorbance and SDS-PAGE and those containing human BCATc (and having a ratio of absorbance of 280 nm to absorbance of 416 nm of 11 ) were pooled and concentrated to 4 mg/ml using an Amicon® YM-30 stirred cell (Millipore Investment Holdings, Ltd). Six mg of human BCATc was recovered at this stage.

Example 2 - Crystallization of BCATc:PLP Complex

Crystallization of BCATc was achieved by the hanging drop vapor-diffusion method at 4°C. The drops consisted of 2.0 μl of BCATc solution and 2.0 μl of precipitant solution in equilibrium with 750 μl of precipitant solution. The BCATc solution contained 4.0 mg/ml human BCATc isolated as described in Example 1 , 20 mM HEPES (pH 7.5), 2.0 mM EDTA, 5.0 mM TCEP and 50 mM NaCI. The precipitant solution contained 200 mM ammonium acetate, 100 mM tri-sodium citrate (pH 5.6), 400 mM PLP, 0.6% (v/v) MPD and 15% (w/v) PEG-4000. Crystals of the enzyme appeared after about 4 days and continued to grow as clusters of yellow plates for up to about one month, reaching maximum dimensions of 0.5 x 0.5 x 0.05 mm.

Prior to X-ray diffraction data collection, the crystals were harvested, dipped in a cryoprotective solution containing 170 mM ammonium acetate, 85 mM tri-sodium citrate (pH 5.6), 25.5%) (w/v) PEG 4000 and 15% (v/v) anhydrous glycerol, and flash-frozen by immersion in a stream of cold nitrogen at 100oK.

X-ray diffraction data to 1.9 A resolution was collected by IMCA-CAT insertion device beamline (17-ID) at the Argonne, IL National Laboratory Advanced Photon Source using a MAR-CCD X-ray detector. The data were corrected for Lorentz and polarization effects converted to indexed structure factor amplitudes using DENZO® and HKL-2000 software (HKL Research, Inc.) (Otwinowski, Z. et al., Methods Enzymol. 276: 307-326 (1966)). The crystals of BCATc were determined to be orthorhombic, space group P212121 (#19), with unit cell dimensions of: a = 109.9, b = 113.5 and c = 149.2.

Example 3 - Three Dimensional Structural Determination of BCATc.-PLP Crystal Complex

The crystals obtained in Example 2 were used to collect X-ray diffraction intensities. Phases, and thus, the crystal and molecular structure of the enzyme, were determined by molecular replacement analysis (AMORE software, Navaza, J., J. Acta Cryst. A50: 157-163 (1994)) using a molecular homology model of human BCATc derived from the crystal structure of BCATm (mitochondrial BCAT) (Yennawar et al., Acta Cryst. D57: 506-515 (2001)). The molecular homology model includes both main chain and side chain atoms. Molecular rotation and translation searches using the BCATm model produced four equally probable solutions indicating that there are four independent molecules of BCATc in the asymmetric unit of structure. This is consistent with the calculated volume density of the crystal form. The best resolution for the structure with four independent molecules in the asymmetric unit structures yields a crystallographic R-factor of 0.448 and a correlation coefficient of 0.61.

The position and conformation of the enzyme-bound PLP co-factor molecules were determined from subsequent difference electron density maps as were the positions of approximately 500 tightly bound water molecules in the structure. The entire structure, including four BCATc monomers, four enzyme-bound PLP molecules and 514 water molecules, was refined to a crystallographic R = 0.257 (R-free = 0.286) using CNX (Accelrys Inc), QUANTA (Accelrys Inc.) and SYBYL® (Tripos, Inc.). The estimated standard deviations in bond lengths and angles are 0.015 A and 2.1 degrees, respectively. Atomic coordinates are provided in Table 1. The coordinates therein describe a crystalline BCATc peptide structure containing: 2 BCATc dimers (obtained as described in Examples 1 and 2), 4 PLP, 2 AKG and 350 waters. The PLP binding site, catalytic binding site and an alpha-ketoglutarate (AKG) binding site are described above.

Example 4 - BCATc:PLP:2-benzofurancarboxylic acid, 5-chloro-, 2-[ [2- (trifluoromethyl)phenyl]sulfonyl]hydrazide Crystalline Complex

A single yellow rod-shaped crystal of the BCATc:PLP complex of Example 2 was soaked for 24 hours at 4oC in a solution containing 5.0 mM saturated 2-benzofurancarboxylic acid, 5-chloro-, 2-[ [2-(trifluoromethyl)phenyl]sulfonyl] hydrazide, 200 mM ammonium acetate, 100 mM tri-sodium citrate (pH 5.6) and 15% (w/v) PEG-4000. The crystal was dipped in cryoprotectant solution containing 170 mM ammonium acetate, 85 mM tri-sodium citrate (pH 5.6), 25.5% (w/v) PEG-4000 and 15% (v/v) anhydrous glycerol, and flash-frozen by dipping the crystal directly into liquid nitrogen. X-ray diffraction data to 2.2 A resolution were collected on the frozen crystal (maintained at 100oK in a cold nitrogen stream) by IMCA-CAT insertion device beamline (17-ID) at the Argonne, IL National Laboratory Advanced Photon Source using a MAR-CCD X-ray detector (tuned to a wavelength of 1.0 A). The data were corrected for Lorentz and polarization effects converted to indexed structure factor amplitudes using DENZO® and HKL-2000 software (HKL Research, Inc.) (Otwinowski, Z. et al., Methods Enzymol. 276: 307-326 (1966)). The structure was resolved by difference Fourier methods using structure factors calculated from the BCATc:PLP complex. The complete structure, including 262 bound water molecules, was refined to R = 0.247 (R-free = 0.279) using CNX (Accelrys Inc), QUANTA (Accelrys Inc.) and SYBYL® (Tripos, Inc.).

As illustrated in Figure 5, 2-benzofurancarboxylic acid, 5-chloro-, 2-[ [2- (trifluoromethyl)phenyl]sulfonyl]hydrazide binds in the catalytic pocket of the BCATc:PLP complex. The trifluoromethyl phenyl moiety stacks edgewise against the pyridoxal ring of the PLP co-factor; the sulfonamide portion forms a series of hydrogen bonds to the peptide backbone of the protein; and the chlorobenzofuran portion of the inhibitor stacks between the side chains of tyr191 and met259 and makes hydrogen-bonded contact with the side chains atoms of cys333, tyr159, thr258 and gln242. The inhibitor also makes hydrogen-bonded contact with the main chain atoms of cys333, ala332 and thr258 as well as with a phosphate oxygen of PLP.

A secondary assay was run to measure the conversion of AKG to glutamate in vitro. It was determined that 2-benzofurancarboxylic acid, 5-chloro-, 2-[ [2- (trifluoromethyl)phenyl]sulfonyl]hydrazide is a potent (IC50 = 0.6 μM) BCATc inhibitor. In addition to the bound 2-benzofurancarboxylic acid, 5-chloro-, 2-[ [2-

(trifluoromethyl)phenyl]sulfonyl]hydrazide, the binding of AKG was observed in each of the four monomers in the BCATc structure.

Atomic coordinates are provided in Table 2 for the BCATc:PLP: 2- benzofurancarboxylic acid, 5-chloro-, 2-[ [2-(trifluoromethyl)phenyl]sulfonyl]hydrazide crystalline complex. The coordinates describe a crystalline BCATc peptide structure bound to 2-benzofurancarboxylic acid, 5-chloro-, 2-[ [2-(trifluoromethyl)phenyl]sulfonyl]hydrazide containing: 1 BCATc dimer, 2 PLP, 2 AKG, 2 2-benzofurancarboxylic acid, 5-chloro-, 2-[ [2- (trifluoromethyl)phenyl]sulfonyl]hydrazide and 76 waters.

Example 5 - BCATc PLP: [2-(2,6-dichlorophenylamino)-4-oxo-4H-thiazol-5 ylidinejacetic acid methyl ester Crystalline Complex

Crystals of the BCATcPLP complex of Example 2 were soaked for 24 hours at 4oC in 5.0 mM saturated [2-(2,6-dichlorophenylamino)-4-oxo-4H-thiazol-5-ylidine]acetic acid methyl ester, 200 mM ammonium acetate, 100 mM tri-sodium citrate (pH 5.6) and 15% (w/v) PEG-4000. The crystals were dipped in cryoprotective solution containing 170 mM ammonium acetate, 85 mM tri-sodium citrate (pH 5.6), 25.5% (w/v) PEG-4000 and 15% (v/v) anhydrous glycerol, and flash-frozen by immersion in a stream of liquid nitrogen. X-ray diffraction data to 2.6 A resolution were collected on the frozen crystals (maintained at 100oK in a cold nitrogen stream) by IMCA-CAT insertion device beamline (17-ID) at the Argonne, IL National Laboratory Advanced Photon Source using a MAR-CCD X-ray detector (tuned to a wavelength of 1.0 A). The data were corrected for Lorentz and polarization effects converted to indexed structure factor amplitudes using DENZO® and HKL-2000 software (HKL Research, Inc.) (Otwinowski, Z. et al., Methods Enzymol. 276: 307-326 (1966)). The structure was resolved by difference Fourier methods using structure factors calculated from the BCATc:PLP complex. The complete structure, including 2 BCATc dimers, four PLP cofactors and 163 waters was refined to R=0.233 (R-free = 0.297) using CNX (Accelrys Inc), QUANTA (Accelrys Inc.) and SYBYL® (Tripos, Inc.). It was determined that [2-(2,6-dichlorophenylamino)-4-oxo-4H-thiazol-5-ylidine]acetic acid methyl ester binds in the AKG binding pocket. The compound made contact with the following residues of BCATc: thr368, phe109, val287, asp114, Ieu96, Iys97, thr331 and tyr373 (Figure 6). The compound binding site is the same as AKG binding site and the mechanism of inhibition is ascribed to the blocking of the AKG binding to BCATc. A secondary assay was run to measure the conversion of AKG to glutamate in vitro. It was determined that [2-(2,6-dichlorophenylamino)-4-oxo-4H-thiazol-5-ylidine]acetic acid methyl ester is a potent (IC50 = 0.6 μM) BCATc inhibitor.

Atomic coordinates for the BCATcPLP: [2-(2,6-dichlorophenylamino)-4-oxo-4H- thiazol-5-ylidine]acetic acid methyl ester crystalline complex are provided in Table 3. The coordinates describe a crystalline BCATc peptide structure bound to [2-(2,6- dichlorophenylamino)-4-oxo-4H-thiazol-5-ylidine]acetic acid methyl ester containing: 1 BCATc dimer, 2 PLP, 1 AKG, 1 [2-(2,6-dichlorophenylamino)-4-oxo-4H-thiazol-5-ylidine]acetic acid methyl ester and 120 waters.

The practice of the present invention generally employs conventional techniques of molecular biology, microbiology, recombinant DNA, immunology, protein chemistry and crystallography, which are well-within the purview of the skilled artisan. Such techniques are explained fully in the literature. See, e.g., Sambrook, J. et al., Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. (1989) and the numerous citations listed above. As used throughout the application and in the claims, the terms "comprising" and

"including" are used in the conventional, non-limiting sense.

All articles, books, patents, patent applications and patent publications cited herein are incorporated by reference in their entirety. While the invention has been described in conjunction with examples and preferred embodiments, it is to be understood that the foregoing description is exemplary and explanatory in nature and is intended to illustrate the invention and its preferred embodiments. Through routine experimentation, one of ordinary skill in the art will recognize apparent modifications and variations that may be made without departing from the spirit of the invention. Thus, the invention is intended to be defined not by the above description, but by the following claims and their equivalents.

Claims

WHAT IS CLAIMED IS:

1. A crystal structure comprising a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide.

2. The crystal structure according to claim 1 having the structural coordinates set forth in Table 1 , Table 2, or Table 3.

3. A crystal structure comprising a peptide that is defined by the atomic coordinates as set forth in Table 1 , Table 2, or Table 3 or a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the core C alpha atoms of the atomic coordinates as set forth in Table 1 , Table 2, or Table 3.

4.^! A crystal structure of a peptide structurally related to a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide comprising a pyridoxal-5'- phosphate (PLP) binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the C alpha atoms of a PLP binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co-factor located in the PLP binding pocket: thr258, met259, asn260, gly330, Iys220, arg117, val287, thr288, thr331 , tyr225, glu255, and ser329, or a conservatively substituted variant thereof.

5. A crystal structure of a peptide structurally related to a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide comprising a catalytic binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the C alpha atoms of a catalytic binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co-factor located in the catalytic binding pocket:

(a) amino acid residues Ieu45, val46, phe47, gly48, thr49, phe93, glu94, gly95, Ieu96, Iys97, tyrl 59, arg161 , val188, gly189, pro190, tyr191 , phe192, ser193, val200, ser201 , Ieu202, arg210, asn224, ser228, Ieu229, gln232, val236, cys240, gln241 , gln242, val243, trp245, val256, gly257, ser329, ala332, cys333 and cys336 from a first peptide monomer; and (b) amino acid residues ser84, his87, tyr88, glu168, ser170, Ieu171 , gly172, val173, Iys174 and Iys175 from a second peptide monomer.

6. A crystal structure of a peptide structurally related to a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide comprising a channel between a catalytic binding pocket and an alpha-ketoglutarate binding pocket, wherein the channel is defined by the atoms found in the atomic coordinates of the

BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the C alpha atoms of a channel that is defined by the atomic coordinates of the following amino acid residues within about 5 A of the channel:

(a) arg210, asn224, tyr225, ser228, Ieu229, gln232, trp245, gly257 and thr258 from a first peptide monomer; and (b) val173 from a second peptide monomer.

7. A crystal structure of a peptide structurally related to a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide, comprising an alpha- ketoglutarate (AKG) binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the C alpha atoms of a AKG binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a substrate located in the AKG binding pocket: Ieu96, phe99, phe109, val287, thr288 and val334 of the peptide.

8. A crystal structure of a peptide structurally related to a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide, comprising an alpha- ketoglutarate (AKG) binding pocket comprising a channel that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the C alpha atoms of a channel that is defined by the atomic coordinates of the following amino acid residues within about 5 A of the channel: phe47, gly95, glnl 10, pro111 , asn 112, Ieu113, ser155, ser157, val188, pro190, tyr191 , phe192, cys291 , gly330, ala332, cys333, Ieu364, Ieu367, thr368 and tyr372 of the peptide.

9. Three-dimensional atomic coordinates of a peptide:co-factor complex comprising: a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide, and a co-factor, wherein the complex has the atomic coordinates set forth in Table 1 , Table 2 or Table 3 or a related set of structural coordinates of any of the foregoing Tables having a root mean square deviation of not more than about 1.25 A from the core C alpha atoms of the atomic coordinates set forth in Table 1 , Table 2 or Table 3.

10. A method of growing crystals of a peptide:co-factor complex comprising a human cytosolic branched-chain amino acid aminotransferase (BCATc), wherein the method comprises: (a) providing a BCATc peptide solution comprising a human BCATc peptide, co- factor, buffering agent, reducing agent, chelating agent, and ionic source; (b) providing a precipitant solution comprising an ionic source, precipitating agent, buffering agent, chelating agent, and reducing agent;

(c) mixing the BCATc peptide solution with the precipitant solution to form a third solution; (d) suspending the third solution over a container housing additional precipitant solution, wherein the vapor pressure of the additional precipitant solution is lower than the vapor pressure of the third solution;

(e) allowing the suspended third solution to stand at a temperature of from about 4°C to about 20°C for a period of time until BCATc peptide:co-factor complex crystals grow to a predetermined size;

(f) dipping the complex crystals into a cryoprotective solution; and

(g) freezing the dipped complex crystals.

11. A method of making crystallizable BCATc peptide, comprising: (a) isolating a BCATc peptide; (b) charging the isolated BCATc peptide with a molar excess of pyridoxal-5'- phosphate (PLP);

(c) subjecting the charged BCATc peptide to anion-exchange chromatography; and

(d) recovering the charged BCATc peptide.

12. A method for generating a three-dimensional computer representation of a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide or structurally related peptide comprising applying the atomic coordinates set forth in Table 1 , Table 2, or Table 3, or a related set of structural coordinates having a root mean square deviation of not more than about 1.25 A from the core C alpha atoms of the atomic coordinates set forth in Table 1 , Table 2, Table 3, to a computer algorithm to generate a three-dimensional representation of the peptide.

13. A method for screening and identifying a potential inhibitor or enhancer of the activity of a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide or a structurally related peptide comprising: (a) generating a three-dimensional representation of a binding pocket selected from the group consisting of:

(1 ) a pyridoxal-5'-phosphate (PLP) binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co- factor located in the PLP binding pocket: thr258, met259, asn260, gly330, Iys220, argl 17, val287, thr288, thr331 , tyr225, glu255, and ser329; (2) a catalytic binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co-factor located in the catalytic binding pocket:

(i) amino acid residues Ieu45, val46, phe47, gly48, thr49, phe93, glu94, gly95, Ieu96, Iys97, tyr159, arg161 , val188, gly189, pro190, tyr191 , phe192, ser193, val200, ser201 , Ieu202, arg210, asn224, ser228, Ieu229, gln232, val236, cys240, gln241 , gln242, val243, trp245, val256, gly257, ser329, ala332, cys333 and cys336 from a first peptide monomer; and (ii) amino acid residues ser84, his87, tyr88, glu168, ser170,

Ieu171 , gly172, val173, Iys174 and Iys175 from a second peptide monomer;

(3) an alpha-ketoglutarate (AKG ) binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A a substrate located in of the AKG binding pocket: Ieu96, phe99, phel 09, val287, thr288 and val334; and

(4) a binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the C alpha atoms of the binding pocket of (a)(1), (a)(2), or (a)(3), by applying the atomic coordinates set forth in Table 1 , Table 2, or Table 3, or a related set of structural coordinates having a root mean square deviation of not more than about 1.25 A from the core C alpha atoms of the atomic coordinates set forth in Table 1 , Table 2, or Table 3, to a computer algorithm to generate a three- dimensional representation of the peptide binding pocket; (b) generating a potential inhibitor or enhancer by (i) assembling molecular fragments into a chemical entity; (ii) de novo design of a chemical entity; (iii) selecting a chemical entity from a small molecule database; or (iv) modifying a known chemical entity; and

(c) evaluating by computer modeling whether the potential inhibitor or enhancer associates with the binding pocket.

14. A method for evaluating the potential of a chemical entity to associate with a human cytosolic branched-chain amino acid aminotransferase (BCATc) peptide or structurally related peptide, comprising

(a) generating a three-dimensional representation of a binding pocket selected from the group consisting of:

(1 ) a pyridoxal-5'-phosphate (PLP) binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co- factor located in the PLP binding pocket: thr258, met259, asn260, gly330, Iys220, arg117, val287, thr288, thr331, tyr225, glu255, and ser329;

(2) a catalytic binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a co-factor located in the catalytic binding pocket:

(i) amino acid residues Ieu45, val46, phe47, gly48, thr49, phe93, glu94, gly95, Ieu96, Iys97, tyr159, arg161 , val188, gly189, pro190, tyr191 , phe192, ser193, val200, ser201 , Ieu202, arg210, asn224, ser228, Ieu229, gln232, val236, cys240, gln241 , gln242, val243, trp245, val256, gly257, ser329, ala332, cys333 and cys336 from a first peptide monomer; and

(ii) amino acid residues ser84, his87, tyr88, glu168, ser170, Ieu171 , gly172, val173, Iys174 and Iys175 from a second peptide monomer;

(3) an alpha-ketoglutarate (AKG ) binding pocket that is defined by the atomic coordinates of the following amino acid residues within about 5 A of a substrate located in the AKG binding pocket: Ieu96, phe99, phe109, val287, thr288 and val334; and

(4) a binding pocket that is defined by the atoms found in the atomic coordinates of the BCATc peptide as set forth in Table 1 , Table 2, or Table 3, or in a related set of atomic coordinates having a root mean square deviation of not more than about 1.25 A from the C alpha atoms of the binding pocket of (a)(1), (a)(2), or (a)(3), by applying the atomic coordinates set forth in Table 1 , Table 2, or Table 3, or a related set of structural coordinates having a root mean square deviation of not more than about 1.25 A from the core C alpha atoms of the atomic coordinates set forth in Table 1 , Table 2, or Table 3, to a computer algorithm to generate a three- dimensional representation of the peptide binding pocket;

(b) applying a chemical entity to the three-dimensional representation; and

(c) quantifying the association between the chemical entity and the binding pocket.

15. A machine-readable medium having stored thereon data comprising the atomic coordinates as set forth in Table 1 , Table 2, or Table 3, or a related set of structural coordinates having a root mean square deviation of not more than about 1.25 A from the core C alpha atoms of the atomic coordinates set forth in Table 1 , Table 2, or Table 3.