WO1998035923A1

WO1998035923A1 - Process for creating molecular diversity

Info

Publication number: WO1998035923A1
Application number: PCT/US1998/002812
Authority: WO
Inventors: Julius Rebek, Jr.; Thomas Carell; Edward A. Wintner
Original assignee: Massachusetts Institute of Technology
Current assignee: Massachusetts Institute of Technology
Priority date: 1997-02-14
Filing date: 1998-02-13
Publication date: 1998-08-20
Anticipated expiration: 1999-08-14
Also published as: AU6163898A

Abstract

Methods for forming combinatorial libraries and the libraries produced thereby are provided. According to a preferred aspect of the invention, a plurality of core molecules are reacted with a plurality of different tool molecules to form a library of molecules having non-naturally occurring molecular diversity. The libraries are useful for identifying lead compounds which modulate the functional activity of a biological molecule.

Description

PROCESS FOR CREATING MOLECULAR DIVERSITY

Govemnent Sunoort The invention described herein was supported in part by grant number GM 27932 froa the National Institute of Health.

Field of the Invention This invention relates to Methods for generating combinatorial libraries and to the libraries produced thereby. The libraries are useful for identifying pharmaceutical and agricultural lead compounds.

Background of the Invention A need exists to identify novel lead compounds that are capable of modulating pharmaceutical and agricultural molecular reactions. Historically, such novel compounds were identified by individually synthesizing a compound and testing it for biological activity. This time- consuming process has been replaced, in part, by a process referred to as "rational drug design". This traditional route to drug discovery requires a knowledge of the structure of the target (i.e., ligate) and/or its cognate (i.e., ligand) in order to synthesize only those molecules which are structurally-related to known ligands of the biological target. However, despite this more focussed approach to identifying lead compounds, the time requirements for conventional structural analysis and chemical synthesis of complex molecules continues to contribute to the substantial periods of time consumed by research and development in the discovery of novel drugs. The application of recombinant technology and automated methods for solid phase chemical synthesis to the generation of olecularly diverse libraries has circumvented many of the most time-consuming steps associated with the traditional drug discovery process, λs a result of these technological advances, "rational drug design" rapidly is being replaced by "irrational drug design", i.e., the process of screening libraries of molecularly diverse molecules to identify biologically active molecules contained therein (Brenner, S., and Lβrnβr, R. , Proc. Natl. Acad. set, re?* £2:5381-5383 (1992)).

Recombinant molecular libraries are generated by inserting random segments of nucleic acid into a vector, such as a phage, and allowing the vector to replicate, transcribe and express the inserted sequence. In a phage library, the nucleic acid is inserted into the phage genome in a manner which permits expression of the inserted material on the phage surface (see, e.g., Pavia, M. , et al., Biooro. i Mad. Cham. Ltr«. 3 tm . ιa7- 396 (1993) and references cited therein; Devlin, J. , et al., Science 2A2-404-406 (1990)). Typically, the virus particles are screened for the presence of a lead compound by contacting the particles with an immobilized ligate and isolating the virus particles which express a ligand that binds to the immobilized ligate.

The primary advantage of using a recombinantly-produced library for drug discovery is the ability to clone and amplify the nucleic acids encoding the lead compounds which are identified during the screening process, unfortunately, recombinant libraries inherently are limited in diversity to linear polymers (e.g., oligonucleotides, peptideβ) formed of naturally-occurring monomers (e.g., nucleotides, L-amino acids). Typically, these naturally-occurring polymers exhibit metabolic instability and poor absorption properties vivo. Accordingly, pharmaceutical lead compounds that are isolated from recombinant libraries frequently require substantial modification to obtain a clinically useful drug.

To overcome these limitations, a number of chemical methods have been used to generate molecular libraries. In general, such methods utilize a solid support (e.g. , cellulose paper, cotton, polystyrene-grafted polyethylene film, polymeric beads and resins) upon which a diverse collection of linear compounds are synthesized using conventional coupling chemistries (e.g., M^βrrifiβld, J. Am. Chen. Soc. (1963) 85:2149-2154). The synthesis of libraries containing molecularly diverse peptideβ has become commonplace since Geyβen et al. first reported the chemical synthesis of a peptide library on polyaerylie acid grafted polyethylene pins (Geysen, H. , et al., J. munoloα. Meth. 102:259-274 (1987)). More recently, Houghten et al. rBioTechniσues 4f6):522-528 (1986) and Nature 354:84-86 (1991)) described a method for generating peptide libraries in which small amounts of a resin support were encapsulated in a plurality of porous polypropylene bags. By sequentially immersing the bags in solutions of individual amino acids under conditions for forming a peptide linkage, Houghten et al. generated a collection of bags, each of which contained a unique resin- immobilized peptide. The generation of combinatorial bead-immobilized peptide libraries also has been reported using a split synthesis procedure (Lam, K., et al., Nature 354:82-84 (1991). Solid-phase peptide synthesis also has been combined with photolithography to prepare arrays of peptides or oligonucleotideβ attached to solid supports (Fodor, S., et al., Science 251:767-773 (1991)).

Although chemically-synthesized libraries offer nearly unlimited molecular diversity, it typically is necessary that small library nolecules be cleaved from the solid support prior to assessing biological activity. Moreover, the above-recited methods typically produce libraries of linear polymers in which monomers are connected via a biological backbone (e.g., a polypeptide backbone of amide linkages connecting amino acid monomers) . Because such structures are recognized by degradative enzymes in _y^, many chemically-synthesized library molecules also are metabolically unstable

^fi, _™""nrγ fff he Tn_Yffn lpn

The present invention overcomes the limitations of the prior art by providing a simple, solution phase process for generating tens of thousands of molecularly diverse molecules. In the preferred embodiments, the library molecules have structures that are not recognized by degradative enzymes i XiXS, a feature which contributes to the metabolic stability and improved absorption properties of these molecules i vivo. _By selecting the appropriate core molecules (discussed below) and tool molecules (discussed below) for forming the library, the conformation (e.g., linear, spherical, disc-like) of the library molecules can be preselected.

According to one aspect of the invention, a method for forming a combinatorial library is provided. The method includes (a) admixing a plurality of core molecules having at least one reactive center with a plurality of different tool molecules, each having at least one functional group to form a reaction mixture, and (b) reacting the reactive centers of the core molecules with the functional groups of the tool molecules to form a plurality of library molecules. To identify pharmaceutical and/or agricultural lead compounds, the combinatorial library is screened for the presence of molecules having biological activity, for example, by identifying library molecules which modulate the biological activity of a ligate (i.e., a molecule which is capable of specifically recognizing and associating with a ligand) .

The core molecule primarily serves as a scaffold to which different tool molecules can be attached at a fixed spatial orientation relative to one another. By combining naturally and/or non-naturally occurring core molecules and tool molecules in accordance with the methods disclosed herein, the instant invention provides a simple, one-step process for creating a combinatorial library of vast molecular diversity. Complementary pairs of reactive centers and functional groups are selected so that the reactive centers are capable of reacting with the functional groups to form the library molecules. In a particularly preferred embodiment, the reactive center is an acid halide, the functional group is an amine and the library molecule is a non-naturally occurring molecule in which the core molecule and the tool molecule(s) are linked via an amide bond. Alternative complementary pairs of reactive centers and functional groups are familiar to those skilled in the art.

According to another aspect of the invention, a method for manufacturing a combinatorial library having molecular diversity is provided. The method includes (a) admixing a plurality of core molecules having at least one reactive center with a plurality of different tool molecules to form a reaction mixture, each tool molecule including a first functional group for reacting with the reactive center and a second functional group attached to a removable, fat-soluble protecting group; (b) reacting the reactive centers of the core molecules with the functional groups of the tool molecules to form a pro-library of fat-soluble molecules; and (c) extracting the unreacted core molecules and the unreacted tool molecules from the pro-library molecules. The tool molecule may contain more than two functional groups (e.g., a third functional group, a fourth functional group) , each of which may be attached to a protecting group. Optionally, the protecting groups are removed from the pro-library of molecules to form a second combinatorial library.

According to yet another aspect of the invention, an alternative method for manufacturing a combinatorial library is provided. The method includes reacting a plurality of core molecules, each having at least two reactive centers at a fixed spatial orientation relative to one another, with a plurality of different tool molecules, so that the reactive centers of the core molecules react with the functional groups of the tool molecules to form a library of molecules, each containing tool molecules located at a fixed spatial orientation relative to one another.

According to still another aspect of the invention, a fat-soluble combinatorial pro-library is provided. As will be apparent to one of ordinary skill in the art, the terms "fat-soluble" and "water-soluble¹ express the relative degree of solubility of a compound in an organic ("fat") phase or in an aqueous ("water") phase. Thus, a compound or molecule is said to be "fat-soluble" if it is more soluble in an organic phase than it is in an aqueous phase. The fat-soluble pro-library contains a plurality of non-naturally occurring molecules, each molecule including at least one tool molecule linked to a core molecule. In the preferred embodiments, the tool molecules further include a removable (e.g., clβavable) protecting group, the inclusion of which group in the non-naturally occurring molecule renders the latter molecule fat-soluble. The fat-soluble pro-library is useful for identifying pharmaceutical and/or agricultural lead compounds which modulate the functional activity of, for example, a membrane-associated biological molecule. Removal of the fat- soluble protecting group from the library molecule (e.g., by cleaving the group from the library molecule) and removal of the cleaved protecting group from the reaction mixture (e.g., by extraction) yields a water- soluble library which also can be screened for the presence of pharmaceutical and/or agricultural lead compounds.

According to another aspect of the invention, a library containing a plurality of structurally-diverse non-naturally occurring molecules is provided. The non-naturally occurring molecules include a first tool molecule and a second tool molecule covalently coupled to a core molecule at a fixed spatial orientation. Within a combinatorial library, the first and second tool molecules may be coupled to the core molecule at the same fixed spatial orientation or at a different fixed spatial orientation. Such combinatorial libraries are useful, for example, for determining the optimal spatial orientation of tool molecules relative to on. another in a lead compound or group of lead compounds, e.g., by comparing the biological activities of lead compound, identified in combinatorial libraries which di far fro. on. another only in the spatial orientation of the tool molecule, or alternatively, by empirically determining the spatial oriantation in lead compound(s) identified in a library containing .olecule. in which the first and second tool molecules are coupled to the core molecule at different fixed spatial orientations.

According to yet another aspect of the invention, a kit for forming a combinatorial library i. provided. The kit includes a plurality of core molacules, each having at least one reactive center and instruction, for reacting the core molecule, with a plurality of tool molecule, to form a combinatorial library. Preferably, the kit further include, the plurality of different tool molecule., each having at l.a.t one functional group.

According to .till another a.pect of the invention, a combinatorial library of non-naturally occurring molecule, i. provided. The library include, a plurality of non-naturally occurring molecules, each including et least one tool molecule covalently coupled to a xanthene molecule. The tool molecule, are .elected from the group con.i.ting of an amino acid and a nucleoβide.

These and other aspects of the invention as well as various advantages and utilities will be -or. apparent with reference to the detailed description of the preferred embodiments and in the accompanying drawing.

Figure 1 .hows exemplary core molecules for generating the combinatorial libraries of the invention (in which the structures represent carbon cores except were noted) ;

Figure 2 show, exemplary amino acid and other primary amine-containing tool molecule, for generating the combinatorial librarie. of the invention;

Figure 3 .how. exemplary nucleoba.e and modified nucl.oba.e tool molecule, for generating the combinatorial librarie. of the invention;

Figure 4 .chematically illu.trate. the reaction of an acid chloride reactive center with a primary amine functional group to form a library molecule containing an amide linkage;

Figure 5 .chematically illu.trate. the reaction of an acid chloride reactive center with an alcohol functional group to for. a library molecule containing an e.ter linkage;

Figure 6 .chematically illu.trate. the reaction of an alkyl chloride reactive center with a primary amine functional group to form a library molecule containing a secondary amine linkage;

Figure 7 schematically illu.trate. the reaction of an alkyl chloride reactive center with an alcohol functional group to form a library molecule containing an ether linkage;

Figure 8 .chematically illu.trate. the reaction of an alcohol reactive center with an acid chloride functional group to for. a library molecule containing an ester linkage;

Figure 9 .chematically illu.trate. the reaction of an alcohol reactive center with an alkyl chloride functional group to form a library molecule containing an ether linkage;

Figure 10 .chematically illu.trate. the reaction of a primary amine reactive center with an acid chloride functional group to form a library molecule containing an amide linkage? Figure 11 schematically illustrates the reaction of a primary amine reactive center with an alkyl chloride functional group to form a library molecule containing a secondary amine linkage;

Figure 12 shows the synthesis and HPLC analysis of a library of 136 theoretical molecules;

Figure 13 shows the synthesis and HPLC analysis of a library of 1225 theoretical molecules;

Figure 14 shows the synthesis and HPLC analysis of a library of 10,968 theoretical molecules;

Figure 15 shows the synthesis and HPLC analysis of a library of 99,141 theoretical molecules;

Figure 16 shows the synthesis of protecting group-derivatized amino acid tool molecules; and

Figure 17 shows the synthesis of 9 , -dimethylxanthene 2,4,5,7-di- and tetra-acid chloride core molecules.

Detailed Description of the Invention 1^. Glossary of terms.

As used herein, the following terms are intended to have the following meanings :

Chemical and Biological Terms: have their common meanings known to one of ordinary skill in the art unless otherwise indicated.

Combinatorial Library: refers to a collection of structurally-diverse molecules. Hence, the libraries of the invention are said to have molecular diversity. As will be explained in more detail below, the extent of this diversity is dictated by the number, the nature and the ratio of the reactants (i.e., the core molecules and tool molecules) which react to form the library molecules. Although the core molecules and tool molecules may be naturally-occurring or non-naturally occurring, the reaction product of the reaction between the core and tool molecules typically is a non- naturally occurring molecule. Thus, in general, the libraries of the invention are said to have non-naturally occurring molecular diversity.

T,1b Mo_lecul s: refers to the plurality of molecules which comprise a combinatorial library. Library molecules are the products of the reaction between the core molecules and the tool molecules (defined below).

_ffnrf, _M»ι«cule: refers to a molecule having a rigid or relatively rigid molecular structure (discussed below) and including at least one reactive center for reacting with a functional group of a tool molecule (defined below) . The core molecule serves as a scaffold to which the tool molecules can be linked in a fixed spatial orientation (defined below) relative to one another.

Examples of core molecules that can be used in accordance with the methods of the invention, include, but are not limited to, chemical compounds (e.g., monocyclic molecules, polycyclic molecules, heterocyclic molecules, aromatic molecules) and biochemical compounds (e.g., adenine, thymine, guanine, cytidine, uracil, inoβine, as well as analogs, nucleosides, and nucleotides of the foregoing nucleobases) . Thus, core molecules also include pharmaceutical molecules which have a known biological activity and fragments of such pharmaceutical molecules. Examples of the foregoing categories of chemical compounds are provided in figure 1. In general, the planar aromatic and planar heterocyclic molecules shown in figure 1 exemplify a class of rigid core molecules and the non-aromatic monocyclic and polycyclic molecules shown in the figure exemplify a class of relatively rigid molecules. In a particularly preferred embodiment, the core molecule is selected from the group consisting of 9,9-dimethylxanthene-2, ,5,7-tetraacid chloride and 1,3,5,7-cubane tetraacid chloride (see, e.g., A. Bashir-Hashemi , Anoe . Chem. (Intl. Ed. in English) 32:612-613 (1993) for a synthesis procedure for the cubane tetraacid chloride). In the preferred embodiments, the core molecule includes two or more reactive centers (described below) for forming a covalent linkage with the functional groups of the tool molecules. Thus, following formation of the covalent linkage, the core molecules have a sufficiently rigid structure to maintain two or more tool molecules at a fixed spatial orientation relative to one another.

Reactive Center: refers to a reactive group of the core molecule which is capable of forming a linkage (e.g., a covalent bond) with a complementary functional group (described below) of the tool molecule. For a linkage- forming reaction which follows a nucleophilic/electrophilic mechanism (discussed below), the reactive center can be a nucleophile (e.g., an alcohol, an amine, a thiol) or an electrophile (e.g., an acid halide, an alkyl halide). In a particularly preferred embodiment, the reactive center is an acid chloride and the functional group of the tool molecule is an amine, an alcohol or a thiol group. Alternative types of reaction mechanisms can be used to form the linkage between the core molecule and the tool molecule(s). For example, palladium can be used to catalyze the reaction between the reactive centers of the core molecules and the functional groups of the tool molecules to form the combinatorial library molecules. (See, e.g., N. Miyaura, H. Sugino e, and A. Suzuki, Tetrahedron T.trs. 22:127-130 (1981)). Other mechanisms for forming a covalent linkage between the core molecules and tool molecules of the invention are known to those of ordinary skill in the art. (See, e.g., March, J., Advanced Organin chemistry. 4th Ed., New York, NY, Wiley and Sons, 1985), pp.326-1120) . Tftf?} Hfr^lecule; refers to the different molecules having at least one functional group which reacts with the reactive centers of the core molecules to form the combinatorial library molecules. Similar to libraries produced using recombinant methods (e.g., phage libraries containing recombinantly-produced peptides), the tool molecules of the invention can be the naturally-occurring L-amino acids or nucleoba^βes (e.g., adenine, guaninβ, thymine, cytidine, uracil). In contrast to such recombinant libraries, the tool molecules also can be the non-naturally occurring molecules, such as a non-naturally occurring nucleoba^βe, nucleoside or nucleotide analog, a D-amino acid, an L-amino acid analog, a non-naturally occurring carbohydrate (e.g., p^βnto^βe and hexo^βe sugar moieties) and carbohydrate analog. Examples of amino acids and other primary amines that are suitable for use as tool molecules are illustrated in figure 2. Examples of nuclβobaβeβ and modified nucl^βobases that are suitable for use as tool molecules are illustrated in figure 3.

mn tiπ ₁ croup; refers to a reactive group of the tool molecule which is capable of reacting with the reactive center of a core molecule to form a linkage. Preferably, the linkage is a covalent bond. For a linkage- forming reaction which follows a nucleophilic/electrophilic mechanism (discussed below), the functional group can be a nucleophile (e.g.-, an alcohol, an amine, a thiol) or an electrophile (e.g., an acid halide, an acyl halide). Thus, for a nucleophilic/electrophilic reaction mechanism, complementary pairs of reactive centers and functional groups (e.g., a nucleophilic reactive center and an electrophilic functional group, an electrophilic reactive center and a nucleophilic functional group) are selected to form the combinatorial libraries. Of course, other types of tool molecules (non-nucleophilic/non-electrophilic) can be reacted with the core molecules depending upon the type of mechanism employed for the coupling reaction (discussed below).

The tool molecules optionally include more than one functional group, e.g., a first functional group for reacting with the core molecule and a second functional group for reacting with a second molecule, such as a protecting group (described below).

F xed Spatial Orientation: refers to the placement in space of at least two molecules (attached to the same molecule, such as a core molecule) relative to one another. In a most preferred embodiment, the core molecules of the invention include two or more reactive centers. The centers are positioned at a fixed spatial orientation relative to one another to permit attachment of two or more tool molecules at a fixed spatial orientation relative to one another. The phrase "fixed spatial orientation" embraces orientations which are entirely restricted in movement, as well as orientations which are partially restricted in movement such as in a library molecule in which the tool molecules on a core molecule are free to move within a limited sphere (e.g., to rotate about the linkage which attaches the tool molecule to the core) .

Hon-naturallv occurring molecule: refers to a molecule which is not found in nature. Such molecules may be produced by synthetic or recombinant methods. For example, a non-naturally occurring peptide (i.e., a peptide which is not found in nature) can be prepared using recombinant methods (see, e.g., Pavia, M. , et al., Bioorg. & Ned. Chea. Ltrs. 3(3):387-396 (1993) or using chemical synthetic methods (see, e.g., Lam, K. , et al., Nature 354:82-84 (1991)).

protecting Group: refers to a material which is bound to a functional group or reactive center and which may be selectively removed therefrom to expose the functional group or reactive center in a reactive form. Preferably, the protecting groups are reversibly attached to the functional groups and can be removed (e.g., chemically or otherwise cleaved) from the functional groups and/or library molecules. The protecting groups include a hydrophobic moiety, which facilitates the separation of the unreacted core molecules and unreacted tool molecules from the combinatorial library molecules (discussed below).

Water-soluble/Fat-soluble: refers to the relative degree of solubility of a compound in an organic ("fat") phase or in an aqueous ("water") phase. A compound or molecule is said to be "fat-soluble" if it is more soluble in an organic phase than it is in an aqueous phase. A compound or molecule is said to be "water-soluble" if it is more soluble in an aqueous phase than it is in an organic phase.

Bit-logical: γ-active/-inactive: refers to the functional activity of a molecule (e.g., ligate or ligand) or a structure (e.g., membrane). Different categories of ligates (described below) exhibit different functional activities. The following example, are illustrative only: The functional activity of a receptor ligate i. the ability to .pacifically recognize and bind to a ligand. The functional activity of an antibody i. the ability to .pacifically recognize and bind to an antigen, an antigenic determinant, or an epitope. The functional activity of an enzyme i. the ability to catalyze a .p^βcific reaction. The functional activity of a nucleic acid i. the ability to hydrogen bond to an oligonucleotide having a substantially complementary nucleic acid sequence. Some nucleic acids (e.g., ribozymeβ) also exhibit an enzymatic activity. Thus, for example, a library molecule is said to be biologically active if it ha. the ability to modulate the functional activity of a apecific ligate or class of ligate.. Different type, of assay, are necessary to screen the combinatorial librarie. of the invention for the presence of biologically-active molecules.

screening; refers to the process by which library molecules are tested for biological activity. For example, the ability to modulate the functional activity of a ligand or ligate can be used as a screening assay to identify lead compounds. A preferred screening method involves contacting the library with an immobilized ligate and identifying the library molecules which bind to the ligate. Alternatively, solution-phase competition assays can be used to screen combinatorial libraries (e.g., by contacting the library with an immobilized ligand in the presence of a soluble ligate for the ligand) and to assess the relative affinity of the library molecule for the ligate (Zucker ann, R. , et al., Proc. Natl. Acad. Sci. USA aq;*gos-ιsn (1992)). Such assays are well known to those of ordinary skill in the art (see e.g., Lam, K. , et al. , su^pra . for a discussion of opiate peptide receptor assays) . Other assays useful for identifying pharmaceutical and/or agricultural lead compounds include assaying for antimicrobial activity, assaying for antiviral activity (e.g., by measuring plaque inhibition) and assaying for anti-fungal activity.

iaatβ. refers to a molecule that has ah affinity for a ligand. Ligateβ may be naturally-occurring or non-naturally occurring and can be used in a soluble or an immobilized state, e.g., attached to a solid support. Categories of ligate. for which the combinatorial librarie. of the invention are u.eful for identifying lead compound, include, but are not limited to, receptor., antibodie., enzyme, and nucleic acid..

ReCWtor Uqateg; The preferred receptor ligate. of the invention include receptor, which modulate a humoral immune response, receptors which modulate a cellular immune response (e.g., T-cβll receptors) and receptors which modulate a neurological response (e.g., glutamatβ receptor, glycine receptor, gamma-amino butyric acid (GABA) receptor). Other receptors for which the combinatorial libraries are useful for identifying pharmaceutical lead compounds include the cytokine receptors (implicated in arthritis, septic shock, transplant rejection, autoimmune disease and inflammatory diseases) , the major hiβtocompatibility (MHC) class I and II receptors associated with presenting antigen to cytotoxic T-cell receptors and/or T- helper cell receptors (implicated in autoimmune diseases) and the thrombin receptor (implicated in coagulation, cardiovascular disease) .

Ajitinodies.: The preferred antibody ligatβs of the instant invention include antibodies which recognize self-antigens βuch a. those antibodies implicated in autoimmune disorders and antibodies which recognize viral (e.g., HIV, herpes simplex virus) and/or microbial antigens. Ens *⁸? ^τhe Preferred enzyme ligates of the instant invention inclu_de polymera^βe^{β (}e.g., RNA polymeraβes, DNA polymeraβeβ) , reverse tran.cripta.es and kineses^, other enzymes for which the combinatorial librarie. are u.eful for identifying pharmaceutical lead compound, include enzyme, implicated in arthriti., o.t.oporo.i., inflaαβatory di.ea.es, diabetes, allergies, organ transplant rejection, oncogene activation ₍e.g., dihydrofolat. reducta.e⁾ , .ignal transduction, .elf-cycle regulation, transcription, DNA replication and repair.

Huclelς A ids; The preferred nucleic acid ligate. of the instant invention include any .egment of DNA or RNA containing natural or non-naturally occurring nucleosides. Nucleic acid, are capable of .pecifically binding to other nucleic acid, or oligonucleotide. via complementary hydrogen- bonding and also are capable of binding to non-nucleic acid ligates. ₍See, e.g.. Bock, L. , et al., Nature 355:564-566 (1992) which reports inhibition of the thrombin-catalyz^βd conversion of fibrinogen to fibrin using aptamer DNA).

iaand. refers to a molecule that is recognized by a ligate.

ead Molec le; refers to a molecule which is capable of modulating the functional activity of a biological molecule. Screening assays are used to identify lead molecules in the combinatorial libraries of the invention. Examples of lead molecules that can be synthesized and selected in accordance with the invention include, but are not limited to, agonists an_d antagonists for cell membrane receptors, toxins and venoms, viral epitopes, hormones, opiates, steroids, peptides, enzyme substrates, cofactors, drugs, lectins, sugars, oligonucleotide^β, nucleic acids, oligosaccharides, lipids, proteins, and analogs of any of the foregoing molecules. Analo^g: refers to a molecule which shares a common functional activity with the molecule to which it is deemed to be an analog and may share common structural features as well.

Pro-library; refers to a type of combinatorial library of molecules which may be further processed into another state to form a second combinatorial library. In a particularly preferred embodiment, the pro-library molecules contain protecting groups which enhance the fat-solubility of the library molecules. The pro-library i. purified by extracting the watβr-.oluble core and tool molecule, from the fat-.olublβ pro-library molecule.. The pro-library can be further procea.ed by, for example, removing the protecting groups from the pro-library molecules, to yield a second combinatorial library which contains water-soluble molecules.

___ Introduction to the Preferred Embodiments.

One method for forming a combinatorial library of the invention includes (a) admixing a plurality of core molecules, each having at least one reactive center with a plurality of different tool molecule., each having at least one functional group to form a reaction mixture; and (b) reacting the reactive center, of the core molecule, with the functional group, of the tool molecule, to orm a combinatorial library of molecules.

As used herein, the phrase "combinatorial library" refer, to a collection of atructurally-diver.e molecule.. As will be explained in sore detail below, the extent of this diversity i. dictated by the number, the nature and the ratio of the core molecules and tool molecules from which the library molecule, are aynthe.ized. Although the core molecules and tool molecule, may have natural or .ynthetic origins, typically the product of the reaction between the core and tool molecules i. a non-naturally occurring molecule. In the preferred embodiments, the library molecules have structures that are not recognized by degradative enzymes in vivo, a feature which contributes to the metabolic stability and improved absorption properties of these molecules in vivo.

The core molecules include at least one reactive center for reacting with the functional group of a tool molecule. Preferably, the core molecules serve a. a rigid or relatively rigid scaffold to which tool molecules can be linked in a fixed spatial orientation relative to one another. Example, of core molecules that can be used in accordance with the methods of the invention, include, but are not limited to, the chemical compounds illustrated in figure 1 (e.g., monocyclic, polycyclic, heterocyclic and aromatic molecules) . Figure 1 illustrates two classes of core molecules: (1) rigid core molecules, exemplified by the planar aromatic and planar heterocyclic molecules shown in the figure and (2) relatively rigid core molecules, exemplified by the non-aromatic monocyclic and polycyclic molecules shown in the figure. The core molecules have a sufficiently rigid structure to maintain two or more tool molecules in a fixed spatial orientation relative to one another following covalent attachment of the tool molecules to the core molecule.

Examples of core molecules also include biochemical compounds (e.g., adenine, thymine, guanine, cytidine, uracil, ino.ine, a. well a. analog., nucleoside., and nucleotides of the foregoing nucleoba.e.) . in a particularly preferred embodiment, the core molecules are pharmaceutical molecules having a known biological activity. Optionally, the pharmaceutical core molecule, are derivatized (e.g., by converting a reactive center of the pharmaceutical core molecule into a more potent nucleophile or electrophile (di.cua.ed below) to enhance reaction of the core molecule with a tool molecule containing a complementary functional group. For example, a pharmaceutical core molecule having an antibiotic activity can be reacted with a plurality of tool molecule, to form a library of derivatized antibiotic molecule.. The derivatized pharmaceutical molecule library can be screened using well-known colony inhibition assays to identify lead compounds which, for example, exhibit enhanced antibiotic potency and/or which confer aultidrug resistance. Conventional and/or novel separation techniques and analytical methods (e.g., HPLC and mass spectroscopy) can be used to elucidate the structure of the novel lead molecules. (See, e.g., the Examples).

The invention also embrace, a method for forming a combinatorial library in which two or more different core molecules are reacted with a plurality of different tool molecules. As will be apparent to one of ordinary .kill in the art, the inclusion of two or more different core molecule, substantially increases the molecular diversity of the libraries formed in accordance with the methods of the invention.

In the preferred embodiments, the reactive centers of the core molecule form a covalent linkage with the functional groups of the tool molecules, although other types of linkages (e.g., a metal coordinate linkage in which a metal is used as the core molecule and the tool molecules form a coordinate bond to the metal, an ionic linkage) also fall within the scope of the invention. Although different types of reaction mechanisms (discussed below) can be used to form the combinatorial libraries of the invention, the preferred reaction forms a covalent linkage between the reactive centers of the core molecule and the functional group of the tool molecule.. In general, almost any reaction for forming a covalent linkage can be used to link a tool molecule to a core molecule. For a summary of covalent coupling reactions, see, e.g., March, J. , Advanced Organic Chemistry. 4th Ed., New York, NY, Wiley and Sons, 1985), pp.326-1120) . The reactions for forming the following specific types of linkages are found in the March text at the pages indicated: (a) reaction of an alcohol and an acid chloride to form an ester bond (pp.346-347) ; (b) reaction of an alcohol and an alkyl chloride to form an ether bond (pp.342- 343) ; (c) reaction of an amine and an acid chloride to form an amide bond (pp.370-371) ; and (d) reaction of an amine and an alkyl chloride to form a secondary amine (pp.364-366) . Alternatively, covalent linkages can be formed using a palladium or other transition metal catalyzed coupling reaction (see, e.g., N. Miyaura, H. Sugino e, and A. Suzuki, Tetrahedron irs^-Z-Z: 127-130 (1981)):

Core Molecule Tools Molecular Library A part.-ularly preferred reaction follows a nucleophile/βlβctrophile mechanism for forming the covalent linkage between the core molecule and the tool molecule. A. used herein, the terms "nucleophile" and "electrophile" have their common meanings (see, e.g., March, J., ibid.). In a nucleophilic/electrophilic reaction, one member of the complementary pair of reactants is an electrophile (such as a carbon atom of an activated carbonyl group) and the other member is a nucleophile (such as an alcohol, an amine or a thiol) .

For example, the core molecule can include one or more electrophilic reactive center(β) which react with a nucleophilic functional group of a tool molecule (described in more detail below) to form a library molecule in which the core molecule and the tool molecule are linked via a covalent bond. The core molecule can include an acid halide (e.g., an acid chloride) reactive center which is capable of reacting with a nucleophilic functional group, such as a primary amine or an alcohol, to form a library molecule in which the core molecule and the tool molecule are linked via an amide (figure 4) or an ester (figure 5) linkage, respectively. Likewise, the core molecule can include an alkyl halide (e.g., an alkyl chloride) reactive center which reacts with, for example, a primary amine or an alcohol, to form a library molecule including a secondary amine (figure 6) or ester linkage (figure 7) , respectively.

Alternatively, the core molecule can include one or more nucleophilic reactive center(β) which are capable of reacting with an electrophilic functional group of a tool molecule to form a library molecule in which the core molecule and the tool molecule are coupled via a covalent bond. For example, the core molecule can include an alcohol reactive center which reacts with an electrophilic functional group, such as an acid halide or an alkyl halide on a tool molecule, to form a library molecule in which the core molecule and the tool molecule are linked via an e.ter (figure 8) or an ether linkage (figure 9), reβpectivβly. Lik^βwi.^β, the core molecule can include a primary amine reactive center which, upon reaction with an acid chloride functional group or an alkyl halide, form, an amide linkage (figure 10) or a secondary amine linkage (figure 11) , respectively. Complementary pair, of core molecule reactive center, and tool molecule functional group, for generating combinatorial librarie. are disclosed in Table 1. As illustrated in the table, exemplary reactive canters include, but are not limited to, an acid halide, an amine, an alcohol and a thiol. Other reactive center, and functional group, will be apparent to the arti.an of ordinary .kill in the art in view of the broad categories of reactive groups disclosed herein. In a particularly preferred embodiment, the reactive center is an acid chloride and the functional group of the tool molecule is an amine, an alcohol or a thiol group.

TABLE 1.

COMPLEMENTARY PAIRS OF CORE MOLECULE REACTIVE CENTERS AND TOOL MOLECULE FUNCTIONAL GROUPS FOR GENERATING COMBINATORIAL LIBRARIES .

Reactive Center on the Functional Group on the Core Molecule Tool Molecule

R s core molecule. R s tool

R-COα R-OH RR'-NH RNH₂ R-SH R-COOH

R-OH RR'-NH RNH₂ R-SH R-COOH

R-X R-CH_rx

X = Cl, Br. I or leaving groups such ts

tosylate. mesyLate. uiflate

R-COH R-COHal

R-CHO RR-.NH R Hj R-C≤CH

R-CH^ by Wittig reaction _. _An AJ = COR. COOR. ^ul2AB COOH. CN. NO2

R-OH RR'-NH RNHj R-SH R-coα R-COOH

R-OH RR'-NH RNH₂ R-SH R-X R-CH_rX ' ^

R-COOH X s P. Br, I or leaving groups like tosyiate. mesylate, trifla-e

RR'-NH RNH₂ R-C=_CH

R-CH₂Q by Wittig reaction

CH₂AB R-CHO

A.B = COR, COOR. COOH. CN.

N0₂ Exemplary tool molecules include, but are not limited to, the naturally-occurring nucleobases, nucleosides, nucleotides, oligonucleotide., amino acid., peptide. and carbohydrates, a. well as non- naturally occurring analogs of the foregoing molecules. Examples of amino acid, and other primary amine. that are suitable for use a. tool molecules in the generation of combinatorial libraries are shown in figure 2. Examples of nucleobases and modified nucleobases that are suitable as tool molecules are shown in figure 3.

The tool molecules can be entirely or partly of synthetic origin. For example, the tool molecule, (e.g., amino acid.) can be purchased in their naturally occurring state (e.g., not derivatized) or in a derivatized state. In a particularly preferred embodiment, a plurality of amino acid tool molecules are derivatized (e.g., to attach a fat-soluble protecting group to a second functional group of the tool molecule) prior to reacting the tool molecule with a core molecule reactive center. (See, e.g., the Examples. )

The diverse tool molecules of the invention have in common a functional group which is capable of reacting with a reactive center of the core molecule to form a linkage. In general, the same coupling reaction conditions can be used to react any electrophile (e.g., an acyl halide, an alkyl halide) with any nucleophile to covalently couple the βlectrophile- containing molecule (e.g., the core) to the nucleophile-containing molecule (e.g., the tool). However, the incubation times for completing the coupling reaction will vary depending on the reactivity of the particular complementary nucleophilβ/electrophile pair participating in the reaction. Thus, for example, a strong nucleophile (e.g., an amine) will react with a strong electrophile (e.g., an acyl chloride) to form an amide linkage in less than thirty minutes using the reaction conditions provided in the Examples. In general, a longer incubation time or an increase in the incubation temperature (e.g., from room temperature to 40 *C) can be used to enhance the reactivity of a weak nucleophile/electrophile pair. The optimization of reaction conditions for each complementary pair of βlβctrophileβ and nucleophilβs is within the ordinary skill of the art.

Preferably, the tool molecules that are used to form the combinatorial library include more than one functional group, e.g., a first functional group for reacting with the core molecule and a second functional group for reacting with a second molecule, such as a protecting group, to form a second linkage. Accordingly, in a particularly preferred embodiment, a method for forming a combinatorial library further includes the step of reacting the secondary functional group with a protecting group prior to admixing the plurality of core molecules with the plurality of tool molecules. A particularly preferred procedure for synthesizing a protecting group-derivatized amino acid tool molecule (in which the protecting group is covalently coupled to an amino acid side chain functional group) is provided in the Examples. Exemplary protecting groups which can be cleaved from the library molecules under strongly acidic conditions ("acid sensitive protecting groups") or basic conditions ("base sensitive protecting groups") are listed in Table 2. A fatty acid also may serve as a removable protecting group, e.g., a fatty acid coupled to a tool molecule via an ester linkage may be cleaved from the tool molecule upon exposure to aqueous basic condition..

TABLE 2.

PROTECTING GROUPS THAT CAN «* cτ.VMVD FROM FUNCTIONAL GROUPS

Acid Sensitive orotflctinα, σrouos:

Functional GrouD Protec inα Grouo

OH t-Bu (tertiary butyl ether)

COOH Ot-Bu (tertiary butyl ester)

NH2 BOC (tertiary butyloxycarbony1)

Guanidine MTR (4-Methoxy-2 ,3,6- trimethy1- benzene-sulfonyl) or

PMC (2,2,5,7,8- Pentamethychroman

6-βulfonyl)

Hi.tidine Trt (Triphenylmβthy1) SH Trt (TriphenyImethy1)

Base sensitive protectinα [ σrouoβ:

Functional Grouo Protectinα GrouD

NH2 FMOC (fluorenylmethoxycarbony1)

The protecting group primarily serve, to modulate the solubility properties of the protecting group-derivatized tool molecule and the library molecules formed therefrom. For example, the reaction of a core molecule (having a hydrophilic acid chloride reactive center) with a lysine tool molecule (in which a fat-soluble, protecting group is covalently linked to the epsilon amine group) yields a library molecule in which the core molecule and the tool molecule are coupled via an amide bond, i.e., the hydrophilic groups of the core and tool molecule are not available to enhance the water-solubility of the library molecule. This difference in solubility allows the fat-soluble, protecting group-derivatized library molecules to be separated from the unreacted core molecules and unreacted tool molecules by extracting the core and tool molecules into an aqueous phase while the library molecules remain soluble in the organic pha.e. If, however, the ly.inβ tool molecule portion of the library molecule contain, an unprotected amine functional group, the library molecule amine group vill be protonatβd under the extraction condition, used to remove the core and tool molecule, and the separation procedure will be made substantially more difficult.

In general, a combinatorial library is generated by loading a plurality of core molecules, a plurality of different tool molecules and 5 ml of dichloromethane (or other solvent in which the core molecules and tool molecules are soluble) into a round bottom flask equipped with a magnetic stir bar. The addition of base to the reaction mixture in the flask (typically about 1 ml triethylamine) starts the coupling reaction, which then is allowed to continue for 1 to 24 hours (depending on the reactivities of the particular electrophiles/nucleophiles present in the reaction) at room temperature under an argon atmosphere.

In general, the amounts of core molecules and tool molecules used in the coupling reaction are selected so that the molar ratio of tool molecule functional groups to core molecule reactive centers is slightly greater than, or equal to, 1. This selection criterion enhances the likelihood that each core reactive center will be linked to a tool molecule and that each tool molecule (regardless of whether it contains a strong or a weak electrophilic or nucleophilic functional group) will react with a core molecule reactive center. For example, the optimum total moles of a plurality of amino acid tool molecules (assuming each amino acid has a single functional group) that should be reacted with 10 moles of a core molecule having 4 reactive centers is slightly greater than, or equal to, 40 (because there are 40 moles reactive centers in the reaction mixture) . If the molar ratio of tool molecule functional groups to reactive centers is significantly greater than 1.0, the diver.ity of the combinatorial library may be limited bβcau.e only the most reactive functional group. _vHl react with the reactive centers. If the molar ratio of tool molecule functional groups to reactive center, i. lβ.β than 1.0, the combinatorial library molecules may include unreacted reactive center, which will likely exhibit .imilar βolubility properties as the unreacted core molecule and extraction of the unreacted core molecules likely will be more difficult.

A particularly preferred procedure for separating the unreacted tool molecules and core molecules from the library molecules, based upon the _βolubility differences between the unreacted molecules and the library molecules, is described herein. Following completion of the coupling reaction, the unreacted core molecules, unreacted tool molecules and library molecules containing core molecules coupled to one or more tool molecules are dissolved in approximately 50 ml of an organic solvent (preferably dichloromethane) . The addition of a weak acid (e.g. , 1 M citric acid) to the organic phase protonates the amine functional group of each unreacted tool molecule to form a quaternary amine, thereby rendering the tool molecules soluble in the weak acid (and in^βoluble in the dichloromethane) . Accordingly, in a particularly preferred embodiment, the unreacted tool molecule, are extracted from the organic pha.e by washing the organic phase with two 75 ml washes of a 1 M citric acid solution.

Exposure to citric acid also results in converting the acid chloride reactive center of each core molecule into its corresponding free acid (which remains soluble in the organic phase) . The addition of a weak base (e.g., 100 ml of saturated sodium hydrogen carbonate solution) to the organic phase deprotonates the unreacted core molecule acid moiety (i.e., the core molecule now carries a negative charge) , thereby rendering the unreacted core molecules soluble in the weak base but insoluble in the organic phase and allowing extraction of the unreacted core molecules. After removal of the weak base solution, the organic phase (containing the pro-library) is dried over MgSO, to remove trace amounts of aqueous phase and is concentrated to yield an oil. The oil comprising the fat-soluble pro-library is dried under high vacuum to yield a foam which can be screened directly for the presence of a lead compound or which can be processed into a second, water-soluble library prior to screening.

In the preferred embodiments, the fat-soluble pro-library is converted into a water-soluble library by removing the protecting groups from the pro-library molecules (e.g., by chemically cleaving the protecting groups from the molecules under acidic or basic conditions, see Table 2). Because of its solubility properties, the fat-soluble pro-library is useful for identifying pharmaceutical lead compounds which modulate, for example, membrane-associated biological processes. Removal of the fat-soluble protecting group from the pro-library molecules yields a library of water- soluble molecules which is useful for identifying pharmaceutical lead compounds that modulate aqueous biological processes. Accordingly, to identify lead molecules which modulate aqueous biological processes, the method for forming a combinatorial library further includes the step of converting the fat-soluble pro-library into a biologically-active state, e.g. , by removing the fat-soluble protecting groups from the pro-library molecules (as described above) or by derivatizing the fat-soluble library molecules to render them water-soluble.

Because the principle utility of the invention is the generation of combinatorial libraries for the identification of pharmaceutical and/or agricultural lead compounds, the methods of the invention further include the step of screening the combinatorial libraries for biologically-active molecules. Numerous screening methods have been reported for identifying lead compounds which modulate the activity of a particular ligate. Such methods include selecting the library molecules which bind to an immobilized ligate (e.g., an antibody), selecting molecules which compete with a soluble ligand for binding to an immobilized or membrane-associated ligate (e.g., a receptor) and selecting molecule, which modulate the activity of the ligate (e.g., an enzyme, a microbe, a virus) in a dose- dependent manner. For example, molecules having antimicrobial activity and/or antiviral activity can be identified using well-known colony inhibition or plaque inhibition assay., respectively. Following identification of a fraction containing one or more lead compound(β) , conventional and/or novel separation techniques and analytical method, (e.g. , HPLC and ma., .pectro.copy) are used to elucidate the structure of the novel lead molecule. (See, e.g., the Examples).

As will be apparent to the artisan of ordinary skill in the art, the extant of molecular diversity of the libraries described herein is delineated by several variables: (1) the number of reactive centers per core molecule; (2) the number of different types of core molecules; (3) the number of functional groups per tool molecule; (4) the number of different tool molecules; and (5) the ratio of tool molecules to core molecules. The dependence of molecular diversity upon these variables is illustrated in the following example.

The generic formula for the reaction of a core molecule having n reactive centers with x different tool molecules, each tool molecule including at least one functional group, is represented by the equation: functional group

Ofi Random <- mbιnaιior

"Core Molecule' X Number of Tools" Molecular library

The reaction yields a molecular library which has diversity, the extent of which is, at least in part, defined by the values of n and X. By selecting a particular set of different tool molecule, (e.g., a .at of amino acids or peptides, a set of nucleoside. or oligonucleotide., a .et of tool molecules including a mixture of both amino acids and nucleosides) , combinatorial libraries containing molecules having the potential to bind to a particular type of ligate (e.g., an amino acid-binding ligate, nucleoside-binding ligate) are generated.

According to related aspects of the invention, an automated process for generating the combinatorial libraries of the invention, as well as the equipment for performing the automated process, also are provided, in view of the simplicity of the reactions disclosed herein and the state of the art with respect to the automation of more complex methods for generating combinatorial libraries, automation of the processes disclosed herein is enabled by the instant disclosure (see, e.g.. Pavia, M. , et al., ______;

Jung, G. and Beck-Sickinger, A., Anαewandte Cheaie tlntl . Ed. in Enσ.i 31Ϊ4) :367-486 (1992) for a discussion of fully automated peptide synthesizers and Zuckermann, R. , et al., supr . for a description of the robotic synthesis of peptide mixtures) .

According to still another aspect of the invention, a fat-soluble combinatorial pro-library of non-naturally occurring molecules is provided. The library molecules include at least one tool molecule linked to a core molecule. In the preferred embodiments, the tool molecules further include a removable, fat-soluble protecting group. The fat-soluble pro-library is useful for identifying pharmaceutical and/or agricultural lead compounds which modulate organic phase-associated biological processes. Optionally, the pro-library can be further processed into a water-soluble library that is useful for identifying pharmaceutical and/or agricultural lead compounds which modulate aqueous-associated biological processes.

According to another aspect of the invention, a molecular library containing a plurality of non-naturally occurring molecule, i. provided. Each molecule includes a first tool molecule and a second tool molecule covalently coupled to a core molecule. The first and second tool molecules are positioned at a fixed spatial orientation relative to one another. Combinatorial libraries which differ solely in the spatial orientation of the first and second tool molecules are useful for identifying the optimal spatial orientation between tool molecules for binding to a particular ligate. For example, a first library can be generated by reacting a first core molecule (having reactive centers positioned at a first spatial orientation) with a plurality of tool molecules to form a first combinatorial library. A second combinatorial library can be generated by reacting a second core molecule (having reactive centers positioned at a second spatial orientation) with the same plurality of tool molecules used to generate the first combinatorial library. Each library is screened to identify a lead compound(β) and the relative biological activities of the lead compound(β) present in the first and second libraries are compared. A comparison of the biological activities for the lead compounds found in each library makes it possible to optimize the spatial orientation of the lead molecules for optimum binding to a particular ligate.

Also within the scope of the invention are kits for forming a combinatorial library. The kits contain a plurality of core molecules, each having at least one reactive center and instructions for reacting the core molecules with a plurality of tool molecules to form a combinatorial library. Preferably, the plurality of tool molecules are included in the kit. More preferably, the kits contain separate pluralities of core molecules which differ from one another in the nature and/or spatial orientation of the core reactive centers. The kits are useful for generating more than one combinatorial library which differ in the type (e.g., amide, aldehyde) and/or spatial orientation of the linkage connecting the core and tool molecules.

According to yet another aspect of the invention, a combinatorial library including a plurality of non-naturally occurring molecules is provided. The library molecules include at least one tool molecule that is covalently coupled to a xanthene molecule. In a particularly preferred embodiment, the tool molecules are selected from the group consisting of amino acids and nucleosides. Specific reactions for preparing the particularly preferred amino acid-derivatized xanthene molecules are disclosed in the Examples.

The utilities of the invention include the identification of novel pharmaceutical and/or agricultural lead compounds which antagonize, agonize or otherwise modulate the physiological activity of natural ligands. Thus, the libraries disclosed herein contain compounds which have the potential to modulate the functional activity of biological processes implicated in (but not limited to) disease progression, immune system modulation, and neurological signal transmission. The libraries of potential lead compounds include compounds which mimic the active determinants (i.e., the ligate-binding portion of the molecule) on hormones, cytokines, enzyme substrates, enzyme cofactors, virus, microbes and fungi. The combinatorial libraries of the invention also are potentially rich sources of novel antimicrobial, anti-viral and anti-fungal agents. In view of the state of the art with respect to the use of combinatorial libraries for epitope mapping (see e.g., Jung, G. and Beck- Sickinger, A., supra . which reports the epitope mapping of the immunogen region of cytomegalovirus (hCMV) using a library of 49 peptides simultaneously prepared on cellulose paper sheets), use of the combinatorial libraries disclosed herein for epitope mapping is supported by the instant disclosure and is within the skill in the art. Accordingly, the libraries are useful for identifying the epitopes of numerous types of Uganda and ligates (e.g., receptors, antibodies, enzymes, nucleic acids, carbohydrates, lipids), including both continuous epitopes (i.e., the epitope is formed of contiguous molecules) and discontinuous epitopes (i.e., the epitope is formed by the juxtaposition of non-adjacent ligand monomers as a result of, for example, secondary/tertiary structure folding) . The majority of naturally-occurring ligates are believed to recognize discontinuous epitopes (Pavia, M. , et al., supra ...

These and other representative utilities of the invention are illustrated in the following non-limiting Examples. The present invention provides methods for forming combinatorial libraries of molecules which are useful for identifying pharmaceutical and/or agricultural lead compounds. The following Examples illustrate a particularly preferred embodiment of the method and representative utilities of the present invention.

EXAMPLES

The disclosure of the present invention illustrates the reaction of a number of different core molecule/tool molecules combinations. According to one embodiment, a plurality of core molecules (xanthene-di-acid chloride) having two reactive centers (n-2) was reacted with ten different tool molecules to form a combinatorial library* Because each tool molecule included a single amine functional group (i.e., X-10), the library theoretically contained fifty-five different library molecules. Preparative HPLC separation, combined with fast atom bombardment masβ βpectromβtry, demonstrated that at least 41 different library molecules (or about 75% of the theoretical value) were formed.

More complex combinatorial libraries also have been prepared by reacting a plurality of core molecules (xanthene-tetra-acid chloride) having four reactive centers (n-4) with 4 (figure 12), 7 (figure 13), 12 (figure 14) or 21 (figure 15) different tool molecules (each including a single amine functional group). A proportional increase in molecular diversity for libraries formed by reacting increasing numbers of different tool molecules with a constant number of core molecules was demonstrated using high pressure liquid chromatography (HPLC) and mass spectroscopy. Theoretically, the reaction of one core molecule (including four reactive centers) with twenty-one different tool molecules yielded a library containing 99,141 different molecules. Although the exact numbers of library molecules formed could not be determined from the HPLC results, the HPLC elution profile (figure 15) demonstrated the presence of substantial diversity in this library. Materials and Methods:

Amino acids were purchased from Nova Biochem, La Jolla, CA.

FMOC-derivatized amino acids were purchased from Advanced

Che tech . Inc. , Louisville, KY.

Solvents, acids and bases were purchased from Fluka Chemie, Ag, Buchs, Switzerland, unless otherwise noted.

Reagent grade solvents were used throughout the procedure unless otherwise noted.

HPLC Chromatography was performed using a silica column (preparative or analytical, as described below) on a Waters 600E HPLC System with a Waters 490E UV detector (UV detection at 270 nm) and a Waters 717 autosampler (Millipore Corp., Waters Chromatography Division, Milford, MA).

The gradient used was threefold: 100% hexane to 100% ethyl acetate to 6% methanol/94% ethyl acetate. All HPLC solvents were purchased from EM Science (Gibbstown, NJ) and were EM Science's Omni Solve • grade. The analytical silica column was purchased from Beckman Instruments, Inc., Fullerton, CA (part #235341, ULTRASPHERE • SI column, 4.6 mm i.d. x 25 cm length, 80 angstrom, 5 micrometer). The preparative silica column was purchased from Millipore, Waters Chromatography Division, Milford, MA (part #25823, NOVA PAR • HR silica, 19 mm i.d. x 30 cm length, 68 angstrom, 6 micrometer) .

Samples were analyzed by Fast Atom Bombardment Mass Spectrometry in dimethyl sulfoxide (DMSO)/glycerol matrix (Aldrich, Milwaukee, WI) on an Ion Tech Mass Spectrometer (Teddington, UK).

Samples were analyzed by proton NMR Spectroscopy in deuterated DMSO (Aldrich, Milwaukee, WI) on a Varian XL 301 NMR instrument (Varian, San Fernando, CA) . EXAMPT.F 1. Synthesis of a Combinatorial Mbrarv Th«r.r»»<r._nn _γ Containing 136 Different Lihrnrv Mole^m**.,

This combinatorial library was generated by reacting the core molecule 9,9-dimethylxanthene-2,4,5,7-tetracarboxylic acid chloride with the following derivatized amino acid tool molecules:

H-L-Try-OMe, 60.9 mg;

H-L-Phe-OMe, 51.6 mg;

H-L-Val-OMe, 40.1 mg; and

H-L-Ala-OMe, 33.4 mg The reaction mixture contained 0.239 mmol of each amino acid.

A 10 ml one-neck, round bottomed flask equipped with a magnetic stir bar was charged with xanthene-2,4,5,7-tetracarboxylic acid chloride (100 ag, 0.217 mmol) (synthesized as described in Example 10), the mixture of derivatized amino acid tool molecules (shown above) and 5 ml dichloromethane. The flask was stoppered with a rubber septum containing an argon inlet. One ml of triethylamine was added to the reaction mixture with a syringe and the reaction mixture was stirred under an argon atmosphere for 3 hours. The mixture was diluted with 50 ml of dichloromethane, washed twice with 75 ml of citric acid solution (1 M) and washed once with 100 ml of saturated sodium hydrogen carbonate solution. The organic phase was dried over MgS0₄ and concentrated to yield a. tan oil. The oil was dried under a high vacuum to yield a tan foam. Analytical HPLC analysis of the tan foam is shown in figure 12.

EXAMPLE 2. S^ynthesis of a Combinatorial Library Thq_?rf,: πw_{l γ} Containing 1225 Different Library M l 0^1^ ,

This combinatorial library was generated by reacting the core molecule 9,9-dimethylxanthene-2,4,5,7-tetracarboxylic acid chloride with the following mixture of derivatized amino acid tool molecules:

H-L-Trp-OMe, 34.8 mg H-L-Val-OMe, 22.9 mg H-L-Ala-OMe, 19.1 mg H-L-Phe-OMe, 29.5 mg H-L-Met-OMe, 27.3 mg H-L-Pro-OMe, 22.6 mg H-L-Leu-OMe, 24.8 mg The reaction mixture contained 0.136 mmol of each amine.

This library was generated using the same procedure as described in Example 1. Analytical HPLC analysis of the tan foam is shown in figure 13.

Synthesis of a Combinatorial Library Th^ g i^ny

H-L-Ala-OMe, 11.1 mg H-L-Phe-OMe, 17.2 mg H-L-Met-OMe, 15.9 mg H-L-Pro-OMe, 13.2 mg H-L-Leu-OMe, 14.5 mg H-L-Lys(Boc)-OMe , 23.6 mg H-L-Ser(tBut)-OMe, 16.9 mg H-L-His(Trt)-NHR,, 34.9 mg H-L-Asp(OtBut)- Me, 19.1 mg H-L-Glu(OtBut)-OMe, 20.2 mg H-L-Thr(tBut)-OMe, 18.0 mg Fur urylamine, 8.4 mg The reaction mixture contained 0.0785 mmol of each amine.

This library was generated using the same procedure as described in Example 1. Analytical HPLC analysis of the tan foam is shown in figure 14.

H-L-Trp-OMe, 10.5 mg H-L-Val-OMe, 6.9 mg H-L-Ala-OMe, 5.7 mg H-L-Phe-OMe, 8.9 mg H-L-Met-OMe, 8.3 mg H-L-Pro-OMe, 6.9 mg H-L-Lβu-OMe, 7.5 mg H-L-Lyβ(Boc)-Me, 12.3 mg H-L-Ser(tBut)-OMe, 18.8 mg H-L-His(Trt)-NHR,, 18.1 mg H-L-Asp(OtBut)-OMe, 9.9 mg H-L-Glu(OtBut)-OMβ, 10.5 mg H-L-Thr(tBut)-OMe, 9.3 mg H-L-Ilβ-OtBut, 9.3 mg H-L-Cys(Trt)-NHR,, 18.7 mg H-L-Arg(Mtr)-NHR,, 21.0 mg H-L-Tyr(tBut)-OMe, 11.9 mg H-L-Val-NHR, , 8.2 mg 4-Methoxybenzylamine, 3.7 mg l-Methylpyrrol-2-ethylamine, 5.1 mg Furfuryla ine, 4.2 mg The reaction mixture contained 0.0413 mmol of each amine.

This library was generated using the same procedure as described in Example 1. Analytical HPLC analysis of the tan foam is shown in figure 15.

Rva»ple 5. Synthesis of a fat-soluble Pro-Librarv twhich own «- denrotected to yield a water-soluble Combinatorial Libra y^

IQA The Synthesis of Amino Acid Tool Molecules containing a Fat- soluble Protecting Group.

Selected amino acid derivatives which were not commercially available were synthesized (e.g., see figure 16, compounds 11, 18, 21 and 22). The synthesis of these compounds from the appropriate FMOC-protected amino acid compoud (see figure 16, compounds 31, 32, 33 and 34) is described below and is illustrated in figure 16. 111 Activation of the FMOC-protected amino ari der v ^j_Yf«- ^ conversion into the amides identi ied in figure 16 »_ff compounds 35. 36. 37 and 38.

The FMOC-protected amino acid derivatives were activated with benzotriazol-l-yloxytris-(dimethylamine ) phosphonium hexafluoro-phosphate

(BOP) and converted into the amides identified in figure 16 as compounds

35, 36, 37 and 38 by reacting with the amines cyclohexylamine, benzylamine,

4-methoxybenzylamine and n-propylamine, respectively (see figure 16, the

R.-NH,).

A description of the general reaction protocol is provided below, followed by the specific protocols for synthesizing compounds 35, 36, 37 and 38.

General Protocol: A one-neck, round bottomed flask equipped with a rubber septum with an argon inlet and a magnetic stir bar was charged with the FMOC- and side chain-protected amino acid derivative (500 mg), 3 ml of dimethyl format-tide and the appropriate amine (recited above). BOP was added and the reaction mixture was stirred at room temperature for 3 hours. The mixture was diluted with 70 ml of dichloromethane, washed twice with 100 ml of citric acid solution (1 M) and once with saturated sodium hydrogen carbonate solution. After drying over MgS0₄, the mixture was concentrated in vacuo to yield a clear oil. Crude compound 35 (see figure 16) was purified by suspending the compound in an appropriate organic solvent (e.g., n-hexane/ethyl acetate (1:1). Compounds 36, 37 and 38 were purified by chromatography on silica gel (described below). The chromatography fractions containing these compounds were concentrated in vacuo to yield a white foam. Synthesis of Compound *..

N"-Fluorenylmethoxycarbonyl-L-valine-cyclohexylamide (JJS w s synthesized by reaction of N"-fluorenylmethoxycarbonyl-L-valine (50₀ mg 1.5 mmol) with cyclohexylamine (150 mg, 1.5 mmol) and BOP (680 ag, 1.54 mmol) as the coupling reagent. Isolation of product was achieved by suspending the crude product in n-hβxane/ethyl acetate (1:1) and filtration. The residue was collected and concentrated to half of its volume. The slurry thus obtained was again filtered. The combined residues were dried under reduced pressure. The product was obtained as a white solid, 410 mg (66%).

Η NMR (300 MHz,DMSO-d 6 ^« 7.88 (d,2H,J - 7.6 Hz,ar-H), 7.76 (a,3H,ar- H,NH), 7.43 ("t",2H,J - 7.3 Hz,ar-H), 7.33 (m,3H,ar-H,NH) , 4.22 (m,4H,CH.,α-CH, flUorβnyl-H), 3.77 ("t",lH,J - 7.0 H , (CH,).-H), 3.50 (br.s,lH,cylohexyl-H), 1.90 (m,lH,cyclohexyl-H) , 1.70 (br.β,4H,cyclohexyl- H), 1.55 (br.s,lH,cyclohexyl-H), 1.10 - 1.35 (m,4H,cyclohβxyl-H) , ₀.85 (S,3H,CH.), 0.82 (S,3H,CH,).

Synthesis of Compound 36:

N-Fluorenylmethoxycarbonyl-S-Trityl-cysteine-benzylamide (2&) was synthesized by reaction of N-fluorenylmethoxycarbonyl-S-trityϊcysteine (500 ag, 0.85 mmol) with benzylamine (100 mg, 0.93 mmol) and BOP (500 ng, 1.1 mmol) as the coupling reagent. The product was purified by chromatography on silica gel with n-hexane/ethyl acetate (2:l)eluant, R, -0.59. Yield - 442 mg (80%).

Η NMR (300 MHz, DMSO- L.) * - 8.45 (br.s, 1H, NH) , 7.87 (d, 2H, J >= 7.6 Hz, ar-H), 7.74 (m, 3H), 7.45 - 7.08 ( , 24 H) , 4.35 - 4.05 (m, 4H, a-CH, CH₂, f luorenyl-H) , 2.40 (d, 2H, J » 7.3 Hz, CH,-ar). Synthesis of Compound 17.

-T-Fluorenylmethoxycarbonyl-Nx-( 4-methoxy-2 , 3 , 6-trimethyl-benzene- sulfonyl)-arginine-4-methoxybenzylamide (3_Z) was synthesized by reaction of N"-fluorenylmethoxycarbonyl-Nx-( 4-methoxy-2 , 3 , 6-trimethyl-benzene- sulfonyl)-arginine (550 mg, 0.9 mmol) with 4-methoxybenzylamine (130 mg, 0.95 mmol) and BPO (500 mg, 1.1 mmol) as the coupling reagent. The product was purified by chromatography on silica gel with ethyl acetate as eluant, R, - 0.70. Yield - 460 mg (70%).

Η NMR (300 MHZ, DMSO-d,) 6 - 8.31 (br.s,lH,NH) , 7.94 (S,1H,NH), 7.87 (d,2H,J - 7.3 Hz,ar-H), 7.71 (d,2H,J ■ 7.1 Hz,ar-H), 7.50 (d,lH,J - 7.7 HZ,NH), 7.37 ("t",2H,J - 7.3 Hz,ar-H), 7.31 ("t",2H,J - 7.3 Hz,ar-H), 7.13 (d,2H,J - 8.5 Hz,ar-H), 6.82 (d,2H,J « 8.3 Hz, ar-H), 6.66 (s,lH,ar-H), 6.40 (br.s,2H,N*--H), 4.20 (m,5H,CH,, fluorenyl-H) , 3.98 (m,lH,β-CH) , 3.76 (β,3H,OCH,), 3.69 (s,3H,OCH,), 3.03 (m,2H,CH,), 2.87 (s,3H,CH,), 2.72 (S,3H,CH,), 2.03 (s,3H,CH,), 1.70 - 1.30 (br.s,4H,CH,) .

Synthesis of Compound 38:

N*-Fluorenylmethoxycarbonyl-N*'"-trityl-histidine-prop-l-ylamide (38) was synthesized by reaction of N"-fluorenyaethoxycarbonyl-N*"ⁱ'-trityl- histidine (500 mg, 0.81 mmol) with n-propylamine (60 mg, 1.0 mmol) and BOP (500 mg, 0.9 mmol) as the coupling reagent. The product was purified by chromatography on silica gel with ethyl acetate as eluant, R, - 0.80. Yield - 315 mg (59%).

Η NMR (300 MHZ, DMSO-d,) δ 7.87 (m,3H,ar-H,NH) , 7.67 (t,lH,J - 7.11 Hz,NH), 7.50 - 7.20 (m,15H), 7.00 (m,7H), 6.66 (s,lH,his-H) , 4.25 - 4.10 (m,2H,α- CH,fluorenyl-H), 2.70 (m,2H,CH,), 2.85 (m,2H,CH,), 2.96 (m,2H,CH,), 1.34 (m,2H,CR.), 0.76 (t,3H,J = 7.3 Hz,CH,). 121 Removal of the FMOC-group from compound* ^ _< -)₆ - . -_fl to form the Protecting Group-derivatig»d a ino _ΠJ_{H tnn}i molecules identified in fi^gure 16 as eo»p _UndB ,,, _{1ff ?}ι and 22. The FMOC-protecting group was removed from compounds 35, 36 37 and 38 by treatment with diethylamine (DEA) to form the Protecting Group- derivatized amino acids identified in figure 16 as compounds 11, is, 21 and 22. The general reaction protocol is provided below, followed by the structural characteristics for each of compounds 11, 18, 21 and 22.

General Protocol: A 10 ml round bottomed flask equipped with a magnetic stir bar and a septum containing an argon inlet was charged with the FMOC-protected amino acid derivatives 35, 36, 37 or 38; 3 ml of dichloromethane and 3 ml of diethylamine. The reaction mixture was stirred under argon atmosphere at room temperature for 2 hours. The reaction mixture was diluted with 50 ml of dichloromethane and, after the addition of 5 g FLORISIL • (Fisher Scientific Co. Springfield, NJ, a course silica gel), concentrated in vacuo. The product, which precipitated on florisil, was added to the top of a small silica gel column (d - 2 cm, 1 - 15 cm). The column was rinsed with ethyl acetate/n-hexane (1:1) in order to remove the cleaved FMOC.

The free amine (i.e., FMOC-deprotected) product was eluted with methanol/triethylamine (99:1) and was detectable on a kieselgel-precoated TLC plate (Merck * Co., Inc., Rahway, NJ) by staining with ninhydrin solution. Fractions containing the free amine product were concentrated in vacuo to afford a white foam with a yield greater than 90 %.

Compound 11: (L-Valine-cyclohexylamide (H)): Η NMR (300 MHz,DMSO-d δ 7.60 (d,lH,J - 7.4 Hz,NH), 3.55 (m,2H,o- CH,cyclθhexyl-H), 2.85 (d,lH,J - 5.6 Hz, (CH,).CH), 1.77 - 1.60 (br.s,6H,NH₂, cyclohexyl-H) , 1.60 - 1.50 (br.s,2H,cyclohexyl-H) , 1.45 - 1.07 (m,4H,cyclohexyl-H), 0.81 (d,3H,J - 6.7 Hz,CH,), 0.76 (d,3H,J - 6.6 Hz,CH,).

_r,τ_np un 18: (S-Trityl-L-cystein-benzylamide (lfi) ) : Η NMR (300 HHz,DMSO-d«) * 8.34 (t,lH,J - 5.9 HZ,NH), 7.48 - 7.15 (m,20H,ar- H), 4.35 - 4.15 (m,2H,CH,Ph), 3.19 (t,lH,J - 6.1 Hz,o-CH), 2.39 (dd,lH,J - 11.2 and 6.2 Hz, CH,) , 2.36 (dd,lH,J - 11.7 and 7.3 Hz,CH,), 2.00 (br.s,2H,NH,).

romnoun_d 21: (Nx-( 4-Methoxy-2 , 3 , 6-trimβthyl-bβnzβnβ-βulfonyl )-L- arginine-4-methoxybenzylamidβ (21)): Η NMR (300 MHz, DMSO-d * 8.24 (br.β,lH,NH) , 7.15 (d,2H,J - 8.5 Hz,ar-H), 6.85 (d,2H,J - 7.3 Hz,ar-H), 6.67 (β,lH,ar-H), 6.40 (br.^β,3H, guanidin^β-H) , 4.18 (d,2H,J - 6.2 H_Z,CH,), 3.78 (β,3H,OCH,), 3.71 (^β,3H,OCH,), 3.12 (br._S,lH,α-CH), 3.00 (m,2H,CH₁), 2.59 (β,3H,CH.), 2.51 (^β,3H,CH,), 2.35 - 2.20 (br.ε,2H,NH,), 2.04 (β,3H,CH,), 1.60 - 1.25 (br.s^^CH, .

compound 22: (N^-Trityl-L-hiβtidine-prop-l-ylamide (22)) : Η NMR (300 MHz^MSO-dL,) δ 7.79 (t,lH,J - 5.8 Hz,NH), 7.40 (m,9H,trityl-H) , 7.24 (d,lH,J - 1.2 Hz, imid-H) , 7.05 (m, 6H, trityl-H), 6.61 (d,lH,J - 1.5 H_Z, imid-H), 3.33 (m,lH,α-CH), 2.95 (dd,2H,J - 13.4 and 7.0 Hz,CH-), 2.73 (dd,lH,J - 13.9 and 4.8 Hz,lH,hiβ-CH.) , 2.50 (dd,lH,J - 14 and 7.2 Hz, his- CH,), 1.93 (br._S,2H,NH.), 1.33 (m,2H,CH-), 0.77 (t,3H,J - 7.4 Hz,CH,).

tbl s hesis fTf t^hw Pro-Librarv.

This combinatorial library was generated by reacting the core molecule xanthene-2,4,5,7-tetracarboxylic acid chloride with the following mixture of derivatized amino acid tool molecules: H-L-Ala-Ot-But, 15.7 mg; H-L-Tyr(t-But)-OH, 20.5 mg; H-L-Arg(Mtr)-OH, 33.3 mg; H-L-Trp-OMe, 22.0 mg; H-L-Ser(t-But)-OH, 18.3 mg; H-L-Glu(Ot-But)-Ot-But, 25.5 mg; H-L-Asn-Ot-But, 19.4 mg; H-L-Val-Ot-But, 18.1 mg; H-L-Asp(Ot-But), 24.3 mg; H-L-Pro-Ot-But, 18.0 mg; H-L-Thr(t-But)-OH, 15.1 mg; H-Hiβ(Trt)-OH, 34.3 ag.

The reaction mixture contained 0.0865 mmol of each derivatized amino acid tool molecule.

This library was generated using the procedure disclosed in Example l.

ISX nenrotecti n of the Pro-Llbrarv Molecules.

The above-described foam was dissolved in 3 al reagent K (82.5 % trifluoro acetic acid, 5 % phenol, 5 % water, 5 % thioanisol, 2.5 % βthanedithiol (Aldrich, Milwaukee, WI) and stirred from 1 to 4 hours (preferably about 2 hours) at room temperature. Reagent K contains a strong acid which is capable of cleaving the protecting groups from the library molecules.

The reaction mixture was concentrated to yield a tan oil which was dried under high vacuum conditions. The dried oil was treated with cold diethylether to extract the cleaved protecting groups, leaving behind a white solid that was collected and dried in vacuo. The white solid (50 mg) is dissolved in water (e.g., 0.5 ml) or a buffered aqueous solution for analysis.

Fvamole 6. Screening procedures using an immobilized ligate.

An appropriate amount of the water-soluble combinatorial library is dissolved in a minimum amount of water or buffered solution for performing the screening assay. As used herein, "appropriate amount" refers to an amount which is within the detection limits of the screening assay. In general, the detection limits for an ELISA screening assay are between about 0.1 nM - 0.05 «. Thus, the "appropriate amount" of a library molecule for dissolution will be less for a screening method which employs a more sensitive detection method (e.g., a radiolabeled tag or amplification of the signal prior to detection) and will be more for a screening method which employs a less sensitive detection method.

An affinity matrix containing a plurality of ligate molecules immobilized to an insoluble support (e.g., polyacrylamide, agarose, sepharose) is purchased or synthesized according to methods known to one of ordinary skill in the art. The affinity matrix is washed with buffer prior to applying the combinatorial library molecules. An aliquot of serial dilutions of the combinatorial library (or a fraction thereof, such as a fraction eluted from an HPLC column) is added to the affinity matrix, followed by washing the matrix to remove unbound or non-specifically bound library molecules. The matrix is contained in a column or is contained in, for example, a microcentrifuge tube for performing a batch separation.

The elution of unbound or non-specifically bound library molecules from the affinity matrix is determined by monitoring, for example, the absorbance at the wavelength at which the library molecules are known to absorb light, or by other detection methods (e.g., thin layer chromatography). Elution fractions containing a detectable amount of library molecules are collected, concentrated and analyzed. Optionally, the affinity matrix is rinsed with a concentrated solution of a ligand that is known to bind to a specific region of the ligate. Accordingly, washing the matrix with a solution of the ligand specifically elutes compounds that bind to the same region of ligate. The eluted fraction(s) are subjected to the analysis procedure (described below). Alternatively, or additionally, the affinity matrix is rinsed with a solution of a denaturant which denatures the immobilized ligate, thereby releasing library molecules which have specifically bound to the immobilized ligate. The eluted fraction(s) then are subjected to the analysis procedure.

The following example is intended to illustrate the above-disclosed general process for using an immobilized ligate to identify lead compounds in a combinatorial library. The example is not intended to limit the invention to a particular embodiment.

The carbohydrate-binding protein concanavalin A (Con A) covalently bound to sepharose is purchased (e.g., Sigma Chemical Co., St. Louis, MO). An appropriate amount of the library molecules are added to 20 ul of Con A- Sepharose contained in a aicrocentifuge tube, in a final volume of 200 ul Con A Buffer (50 mM NaCl/20 mM Mops, pH 6.8/2 mM MgClj/2 mM CaCi o.2 mM EOTA). (See, e.g., Oldenburg, K., et al., Proc. Nat. Acad. Sci . OSA £2:5393-5397 (1992)). The library molecules are allowed to bind to the Con A-Sβpharose for 1 - 24 hours at room temperature with agitation. Unbound or non-specifically bound library molecules are removed by washing the Con A-Sepharose with Con A Buffer (e.g., three 5 minute washes) (described above). Specifically-bound library molecules are eluted with either 200 mM methyl alpha-D-mannopyranoside, 1 % mannan, or 100 mM citrate buffer (pH 3.0) for at least about 30 minutes at room temperature and are subjected to the analysis procedure (described below).

Fvamole 7. Screening procedures using a soluble ligate.

Combinatorial library molecules are assayed in a competition ELISA format over a concentration range of about o.l nM - 0.05 mM for each library molecule using procedures familiar to the artisan of ordinary skill in the art (see, e.g., Zuckermann, R. , et al., Proc. Nat. Acad. Sci. USA fl9_:4505-4509 (1992). Microtiter plates are coated (e.g., about 0.2 ug ligand per well in 50 mM borate, pH 9.0, overnight at 4 *C) with a known ligand of the antibody ligate for which lead compounds are being screened. To the ligand-coated microtiter wells is added a 50 ul aliquot of serial dilutions of the combinatorial library (or a fraction thereof, such as a fraction eluted from an HPLC column) with 50 ul of a diluted ligate antibody solution that is known to give a positive, detectable ELISλ signal when incubated with the ligand-coated microtiter well in the absence of a soluble competitor of the immobilized ligand. Wells to which no combinatorial library molecules are added are included as assay controls. Typically, the incubations are performed in a Trie-buffer for about one hour at 37 ^*C. The microtiter plates are washed to remove unbound or non- specifically bound antibody ligate, followed by incubating with a detection reagent (e.g., 100 ul of horseradish peroxidase-conjugated goat anti-mouse antibody (stock solution of 1 ag/ml diluted 1:1000, Boehringer Mannheim)) for about one hour at 37 *C and washed as above to remove unbound or non- specifically bound conjugated antibody. The amount of bound conjugated antibody is quantitated by color development with 100 ul of o- phenylenediamine at 5 mg/ml in 50 mM sodium citrate/0.02% H.O., pH 5.1 and measurement of the absorbance at 450 nm. The presence of a library molecule which specifically inhibits the binding of the antibody ligate to its known (immobilized) ligand is indicated by a reduction in the absorbance at 450 nm. The library molecules which specifically inhibit the binding are subjected to the analysis procedure described below.

Fvamole 8. Iterative Screening Procedure.

A known number of tool molecules (X) are used to generate a first combinatorial library. The library is screened to identify library molecules which modulate the biological activity of an immobilized ligate (see, e.g., Example 6) or a soluble ligate (see, e.g., Example 7). For example, the ability of a library molecule to modulate the biological activity of an enzyme is determined by observing a change in enzyme activity as increasing amounts of library molecules are included in the enzyme assay reaction mixture. Thus, a number of different screening assays can be used to quantitate the effect of the first combinatorial library on the biological activity of the ligate.

A second combinatorial library is prepared using the same mixture of tool molecules as used to generate the first combinatorial library with one exception: one of the tool molecules included in the generation of the first library is absent from the mixture of tool molecules used to generate the second library. By comparing the effect of the first and second libraries on the biological activity of the ligate, one can determine whether the omitted tool molecule is necessary or unnecessary for ligate binding. Moreover, by applying this iterative procedure to each of the tool molecules used to generate the first combinatorial library, the identity of each tool molecule that is involved in binding to the ligate is determined.

^pya pie 9. Analysis Procedure.

A fraction containing the library molecules of interest (e.g., a fraction which has exhibited binding to an immobilized ligate) is subjected to mass spectroscopy and the empirically obtained mass values are compared to the theoretical mass values calculated for each possible compound in the combinatorial library. By matching the empirically obtained mass values and calculated mass values, the molecular weights of library molecules that are able to bind to the ligate are deduced.

Ideally, the structures of the library molecules of interest are deduced from their molecular weights. However, if the structure of a library molecule is not uniquely determined by its molecular weight, an iterative process is used to identify the structure of the biologically active library molecule. The iterative process involves generating a second combinatorial library using only those tool molecules having molecular weights which the mass spectroscopy data indicates are present in the biologically active fraction. As a result, the second library contains a higher concentration of the biologically-relevant library molecules, a factor which facilitates mass spectroscopy analysis.

The second library is screened, subjected to mass spectroscopy and the empirical mass values are compared to the theoretic mass values (as described above) to deduce the structures of the library molecules of interest. This process is repeated until there is sufficient molecular weight data to determine the structure of the biologically active library molecule.

EXAMPLE 10- Synthesis of 9. -dimethvlxanthene diacvl- and tetraacyl- chloride.

Figure 17 schematically illustrates this synthesis procedure. Synthesis of 9.9 - Dlmethγlxanthene-2.4.5.7-tetrabromide:

A 50-mL, one-necked, recovery flask fitted with an addition funnel and magnetic stirring bar was charged with Br, (14.971 g, 93.674 mmol) and 25 mL of CH.-C1. and cooled in an ice bath. 9 ,9Dimethylzanthene (4.924 g, 23.42 mmol) was added over 10 min. After 45 min, Fe powder (0,030 g, 0.54 mmol) was added, and the reaction mixture was allowed to warm to room temperature over 2.5 h. The recovery flask was fitted with a condenser attached to a mineral oil bubbler, and the reaction mixture was heated at reflux until the condensate became colorless (2.5 h). The resulting solution was extracted with 30 mL of H,0, dried over MgS0₄, filtered, and concentrated to afford a white solid. The solid was boiled with 30 mL of ethanol, and the resulting suspension allowed to cool to room temperature. Filtration, followed by drying in vacuo afforded 11.477 (93%) of 4 as a white solid: mp 152-154'C; Η NMR (CDC1,, 5090 MHz) 6 7.64 (d,J - 2.0 Hz,2H), 7.44 (d,J - 2.5 Hz,2H), 1.60 (s,6H); HRMS a/e calculated for CMΕrfiS 521,7476, found 521.7465.

Synthesis of 9.9 - Dimethylxanthene- 2.4.S.7 - e a it-^f..

Copper cyanide (25.89 t, 4.4 eq) was weighed out in a flask fitted with a reflux condenser and drying tube. Mβthyl-2-pyrrolidione (lOO L) and tetrabromo dimethylxanthene (34.6 g, 1 βq) were added, and the mixture was refluxed for 2 h.

The olive green reaction mixture was filtered. The solid washed well with water and added to 250 mL of 20% nitric acid, evolving red gas (NO,) vigorously. The brilliant green mixture was stirred overnight, than filtered to yield a beige solid (20.36 g, 99.6% yield.)

Η NMR (CDCL,) 6 8.55 (d,J « 1.9 Hz,2H), 8.47 (d,J - 1.9 Hz,2H), 1.66 (8,6H). Synthesis of 9.9 - Dimethylxanthene 2.4.5.7 - tetraacid.

The xanthene tetranitrile (5.00 g, 1 eq) was suspended in H,0 (20 mL). Sodium hydroxide (3.21 g) in 20 mL water was added, and the brown mixture was stirred at reflux for 14 h.

The clear brown solution was acidified (pH ? 0) with 2.0 M HC1, precipitating the product. The aqueous layer was repeatedly extracted with THF/E70 λc(l:l) until clear. The combined organic portions were dried over a^O_, and stripped, giving a beige solid (6.127 g, 98.8% yield).

Η NMR (DMSO) 6 13.19 (β broad, 4H), 8.25 (d,J - 2.0 Hz,2H), 8.15 (d,J - 2.0 HZ,2H), 1.66 (s,6H).

synthesis of 9.9 - Dlmethvlxanthene-2.4.5.7-tetracarboxvllc acid chloride. A 250 mL one-necked round bottomed flask fitted with a magnetic stir bar and a reflux condenser was charged with phosphoropentachloride (7 g, 0.03 mol), 9,9-dimethylxanthene-2 , 4 , 5 ,7-tetracarboxylic acid (3 g, 7.8 mmol) and 100 mL of benzene. The reaction mixture was refluxed for 12 h and concentrated to a final volume of 25 mL. The mixture was kept at 4«c overnight. The acid chloride precipitate was filtered off and washed three times with 3 mL of benzene. The product was obtained in the form of small brownish platelets. Yield - 2.3 g (65%).

'H NMR (300 MHz, DMSO-d δ 8.25 (d,2H,J - 2.0 Hz,ar-H), 8.15 (d,2H,J - 2.0 Hz,ar-H), 1.68 (S,6H,CH,): MS (El) m/Z (%) - 460 (3) [M+], 445 (100) [M+- CH,], 423 (54), 380 (31), 317 (18), 254 (11), 194 (16), 163 (21).

Synthesis of 9.9-Dimethvlxanthane-2.4.5.7-tetrabenzvl ester;

5.43 g (14.056 mmol) 9 ,9-Dimβthylxanthene 2 ,4 , 5 , 7-tetra acid chloride was stirred with 10 al (excess) benzyl alcohol and 10 al pyridine in 100 ml CH_jCl,. The dark brown solution was stirred overnight and then extracted with 200 ml 1.0 N HC1 and 200 ml brine. The organic layer was dried over MgSO, and poured into 100 ml methanol (MeOH) . The precipitate was broken up and stirred in solution for one hour, after which it was filtered off and washed with MeOH. The product was an off-white colored solid, 6.98 g, 67% yield.

synt_hesis of 9. -Dimethylχanthene-2.7-dlbenzvl- .5-dicarboxylic aeid;

3.0 g 9 , 9-Dimethylxanthene-2 ,4 ,5 , 7-tetrabenzyl ester were dissolved in 100 ml dry CH,C1,. HBr was bubbled through the stirring solution for 15 minutes, and the solution was stoppered and allowed to stir 24 hours. The cloudy solution was quenched with 100 al H,0 and shaken well. 100 ml tetrahydrofuran (THF) were added, and the solution was extracted with 3 x 200 ml hγ> and 100 ml brine. The organic layer was dried over MgSO, and rotoevaporated to a light yellow solid. Recrystalization from THF, Hexaneε yielded 2.0 g product, 90 %. Synthesis of 9.9-Dimethvlxanthene-2.7-dibenzvl-4.«>-diacid e ln jrtf.

60 mg (.11 mmol) 9,9-Dimethylxanthene-2,7-dib^βnzyl-4,5-dicarboxylic acid were suspended in 1 al CH.C1. and .5 ml Oxalyl chloride. The reaction flask was fitted with a condenser, 3 drops Dimethylformamide were added, and the solution was refluxed for 2 hrs. The solution was diluted with 10 al CHjCl, and stripped to a yellow-white solid. The yield was 55 mg product, 91 %.

Each reference identified above is incorporated herein in its entirety by reference. It should be understood that the preceding is aerely a detailed description of certain preferred embodiments. It therefore should be apparent to those skilled in the art that various modifications and equivalents can be made without departing from the spirit or scope of the invention.

What is claimed is:

Claims

1. A method for forming a combinatorial library, the method comprising:

(a) admixing a plurality of core molecules having at least one reactive center with a plurality of different tool molecules having at least one functional group to form a reaction mixture; and

(b) reacting the reactive centers of the core molecules with the functional groups of the tool molecules to form a plurality of library molecules.

2. The method of claim 1, wherein the tool molecules include a primary functional group for reacting with the reactive centers of the core molecules and a secondary functional group, the method further comprising the step of protecting the secondary functional group with a protecting agent prior to admixing the plurality of core molecules with the plurality of different tool molecules.

3. The method of claim 1, further comprising the step of: (c) separating the unreacted core molecules and the unreacted tool molecules from the library molecules.

4. The method of claim 3, wherein the core molecules and the tool molecules are water-soluble and the library molecules are fat-soluble and wherein separating the core molecules and the tool molecules from the library molecules comprises extracting the unreacted core molecules and the unreacted tool molecules from the fat-soluble library molecules.

5. The method of claim 3, wherein the library molecules are biologically-inactive, further comprising the step of:

(d) converting the library molecules into a biologically-active state.

6. The method of claim 5, wherein the library molecules include a protecting group and wherein converting the library molecules into a biologically-active state comprises removing the protecting group from the library molecules.

7. The method of claim 5, further comprising the step of: (e) screening the library molecules for biological activity.

8. The method of claim 7, wherein screening comprises identifying the library molecules which modulate the functional activity of a ligate.

9. The method of claim 8, wherein identifying the molecules is by selecting the library molecules which bind to an immobilized ligate.

10. The method of claim 8, wherein the ligate is selected from the group consisting of a receptor, a hormone, an antibody, an antigen, an enzyme, a nucleic acid, a carbohydrate and a lipid.

11. The method of claim 10, wherein the ligate is a receptor which modulates an immune system response.

12. The method of claim 11, wherein the receptor is a T-cell receptor.

13. The method of claim 10, wherein the ligate is a receptor which modulates a neurological response.

14. The method of claim 13, wherein the receptor is selected from the group consisting of a glutamate receptor, a glycine receptor and a gamma- amino glutaaic acid receptor.

15. The method of claia 10, wherein the ligate is an enzyme.

16. The method of claia 15, wherein the enzyme selected from the group consisting of a polymerase, a reverse transcriptase and a kinase.

17. The method of claim 1, wherein the plurality of core molecules includes at least two different core molecules.

18. The method of claim 1, wherein the core molecules include at least two reactive centers.

19. The method of claim 18, wherein the two reactive centers are positioned at a fixed spatial orientation relative to one another.

20. The method of claim 1, wherein the core molecules are selected from the group consisting of a monocyclic molecule, a polycyclic molecule a heterocyclic molecule, and an aromatic molecule.

21. The method of claim 1, wherein the core molecule is selected from the group consisting of 9 , 9-dimethylxanthene 2 , 4 , 5 , 7-tetraacid chloride and 1 , 3 ,5 , -cubane tetraacid chloride.

22. The method of claim 1, wherein the core molecule is selected from the group consisting of a nucleoside, a nucleoside analog, a nucleotide, and a nucleotide analog.

23. The method of claim 1 , wherein the reactive center is selected from the group consisting of an acid halide, an amine, an alcohol, a thiol ester and a sulfonate.

24. The method of claim 1 , wherein the reactive center is an acid chloride and the functional group is selected from an amine, an alcohol and a thiol.

25. The method of claim 1, wherein the tool molecules include at least two functional groups.

26. The method of claim 1, wherein the tool molecules are selected from the group consisting of an amino acid, an amino acid analog, a nucleoside, a nucleoside analog, a nucleotide, a nucleotide analog,- a carbohydrate, a carbohydrate analog, a lipid and a lipid analog.

27. The method of claim 1, wherein the functional groups are selected from the group consisting of an acid halide, an amine, an alcohol, a thiol, a sulfonate and an oxalate.

28. The method of claim 27, wherein the functional group is an amine.

29. A method for manufacturing a combinatorial library having molecular diversity, the method comprising:

(a) admixing a plurality of core molecules having at least one reactive center with a plurality of different tool molecules to form a reaction mixture, the tool molecule including a first functional group for reacting with the reactive center and a second functional group attached to a removable protecting group;

(b) reacting the reactive centers of the core molecules with the functional groups of the tool molecules to form a pro-library of molecules having the protecting group attached thereto, wherein attachment of the protecting group renders the pro-library of molecules fat-soluble;

(c) extracting the unreacted core molecules and the unreacted tool molecules from the pro-library of molecules.

30. The method of claim 29, further comprising the step of:

(d) removing the protecting groups from the pro-library of molecules to form a second combinatorial library.

31. The method of claim 30, further comprising the step of:

(e) screening the second combinatorial library for biological activity.

32. A method for manufacturing a combinatorial library having molecular diversity, the method comprising:

(a) reacting a plurality of core molecules having at least two reactive centers at a fixed spatial orientation relative to one another with a plurality of different tool molecules, each tool molecule having at least one functional group for reacting with the reactive centers, so that the reactive centers of the core molecules react with the functional groups of the tool molecules to form a non-naturally occurring molecule containing tool molecules located at a fixed spatial orientation relative to one another.

33. A fat-soluble combinatorial pro-library of non-naturally occurring molecular diversity comprising: a plurality of non-naturally occurring molecules, each of the non-naturally occurring molecules including at least one tool molecule linked to a core molecule, the tool molecule further including a removable fat-soluble protecting group which renders the non-naturally occurring molecule fat-soluble.

34. A molecular library comprising: a plurality of non-naturally occurring molecules, each of the non-naturally occurring molecules including a first and a second tool molecule covalently coupled to a core molecule, the first and the second tool molecule being positioned at a fixed spatial orientation relative to one another.

35. A kit for forming a combinatorial library, comprising: a plurality of core molecules having at least one reactive center; and instructions for reacting the core molecules with a plurality of tool molecules to form a combinatorial library.

36. The kit of claim 35, further including: a plurality of different tool molecules having at least one functional group.

37. The kit of claim 36, wherein the tool molecules include a first functional group for reacting with the reactive center of the core molecule and a second functional group attached to a protecting agent, the kit further including: a cleaving agent for removing the protecting agent from the second functional group.

38. A combinatorial library comprising: a plurality of non-naturally occurring molecules, each of the non-naturally occurring molecules including at least one tool molecule covalently coupled to a xanthene molecule, the tool molecule being selected from the group consisting of an amino acid and a nucleoside.