WO1998046551A1

WO1998046551A1 - Synthesis and use of biased arrays

Info

Publication number: WO1998046551A1
Application number: PCT/US1998/007870
Authority: WO
Inventors: David L. Coffen; Joseph C. Hogan, Jr.
Original assignee: Arqule, Inc.
Priority date: 1997-04-16
Filing date: 1998-04-16
Publication date: 1998-10-22
Also published as: AU7135598A

Abstract

The invention provides methods for making a new type of library of compounds, referred to herein as a 'biased array'. The biased array is based on the premise that by preparing compounds that include (i) a common molecular core that is targeted or biased to interact with or bind to a defined class of target molecules, and (ii) one or more of a wide variety of so-called 'structural diversity elements', one obtains an array that simultaneously achieves broad diversity and a very high probability of including compounds that will bind to or otherwise interact with specific members of the defined target class, but that will not interact with molecules outside of the defined target class. The clear benefit is that the new arrays will provide compounds that can serve as superior lead compounds for the development of new pharmaceutical agents that are highly specific and unlikely to have undesired general biologic activity.

Description

SYNTHESIS AND USE OF BIASED ARRAYS

Field of the Invention The invention relates to the manufacture and use of arrays of molecular compounds .

Background of the Invention Specific molecular recognition between a biological macromolecule and an effector, such as between an enzyme and a substrate, or a receptor and a ligand, is a hallmark of biological systems. These molecular recognition events can lead to binding interactions between the macromolecules and their effectors, thereby resulting in biological activity. For example, the interaction between a receptor and its ligand in a cell can cause an intracellular signaling cascade, ultimately resulting in observable biological effects on the cell. In spite of the enormous complexity of biological systems, evolution has ensured that the many biological activities needed to maintain life can occur simultaneously in conjunction with one another because each activity is modulated by a unique, specific molecular recognition event .

Chemical compounds can influence the molecular recognition events that modulate a given biological activity. For example, a chemical compound can inhibit the function of an enzyme or act as an agonist or antagonist of a receptor. Often, this result occurs because a biological macromolecule recognizes and binds a chemical compound as a substitute for a natural biomolecular effector. If the interaction between the biological macromolecule and the chemical compound is sufficiently strong and specific, the compound can produce a very specific biological effect, potentially with minimal side effects.- Such compounds can therefore make effective and safe pharmaceutical products.

The search for new pharmaceutical products typically begins with the discovery of a linkage between a biological activity and a human disease state. For example, rejection of a transplanted organ by the immune system has been shown to be caused by T cells that are activated by major histocompatibility complex antigens on the surface of cells within the organ. The complex phenomenon of organ transplant rejection is therefore linked to numerous biological activities, each of which reflects numerous molecular recognition events between potential biomolecular targets. Modulation of a molecular recognition event within an appropriate biological activity can therefore modulate the disease state itself. This concept forms the basis for most pharmaceutical research.

The search for chemical compounds that can interact specifically with a biomolecular target is traditionally undertaken by random screening of compounds to find an effect on the biological activity of the target. Historically, chemical compounds with specific biological activity were identified by testing extracts of natural products (such as plants, bacteria, and fungi) against whole organisms to find the desired activity, and then isolating the active component from the extract. More recently, the wide availability of molecular targets, the development of high-throughput screening methods, and the introduction of combinatorial chemistry technology has enabled the efficient synthesis and testing of many thousands of new chemical compounds against hundreds of biomolecular targets every year.

Because the search for biologically active compounds is essentially random, collections of chemical compounds, or libraries, for screening must exhibit a high degree of chemical diversity to maximize the probability that even a single active compound will emerge. In other words, the goal is to accumulate libraries containing a large number of dissimilar compounds so that a screen against a biomolecular target can expose the target to the broadest possible sample of physiochemical features as represented by the compounds in the library. This strategy has traditionally been expected to maximize the probability that at least one compound in the library will have the necessary physiochemical features to interact with the target . After one or more lead compounds are identified from the diverse chemical library, their structures and other physiochemical features can be examined to develop a hypothesis regarding the structure-activity relationship, or SAR. The SAR can then be used to determine changes to the lead compound that can be expected to improve the molecular recognition between the compound and the target. The hypothesis is tested and refined until an optimized compound is developed.

Because the molecular recognition events in biological systems are very specific, a low resolution (i.e., relatively few compounds) sampling of chemical diversity would have a low probability of revealing a useful lead compound. The probability theoretically increases if one screens a library that represents complete chemical diversity. Clearly, the exploration of total chemical diversity space would be ideal; however, this approach would require the synthesis and testing of billions, or even trillions, of compounds. Even with the assistance of high-throughput screening and combinatorial chemistry technologies, such an effort would be prohibitively laborious, expensive, and time-consuming. The logical goal is to determine the minimum resolution that yields an acceptable result, and assemble a library that is intermediate between this minimum resolution and total chemical diversity space. However, the minimal acceptable resolution has not been defined and, because of differences among biomolecular targets, can vary within a given screen.

The random screening method has been successful for the discovery of new pharmaceutical products, and has consequently been widely accepted. However, the method suffers several drawbacks. Despite the assistance provided by high-throughput screening and combinatorial chemistry technologies, the random screening method is laborious, time-consuming, and expensive, as described above, because of the enormous numbers of compounds that need to be synthesized and tested to find a lead compound. Even if a relatively large and diverse compound library is used, the random screening method often fails to identify a lead compound because current compound libraries have many "holes" in their diversity space and lack sufficient resolution in the diversity space that is represented. Even if "holes" in the diversity space are filled and the resolution is increased by adding more compounds to the library, it will remain cost-prohibitive to use current methods to screen the numbers of compounds required to ensure that at least one lead compound will be identified for each biomolecular target tested.

An alternative to the random screening method is structure-guided molecular design. This approach uses information about the physiochemical properties of the molecular recognition site on a biomolecular target to identify a chemical compound that will specifically interact with that site. In effect, this approach enables one to conduct a "virtual screen" of a compound library. The approach requires access to detailed information about the molecular recognition site on the target. The information is entered into a computer and modelled by special -purpose software. The necessary information about the molecular recognition site can be obtained by, for example, x-ray diffraction studies of a crystallized sample of the target, nuclear magnetic resonance studies of the target, or quantitative structure-activity relationship (QSAR) studies of compounds that bind to the target .

Like the random screening method, the structure- guided molecular design approach has had some success in the identification of lead compounds. However, this method also suffers from several drawbacks. Often, the detailed target information required for structure-guided design is not available, which means that the necessary information must first be developed. The development of this detailed target information is very expensive, laborious, and time-consuming, if it is possible to develop at all . Even assuming that the required information is available or can be developed, the computer models of the molecular recognition site are often slightly inaccurate, due, for example, to differences between solid-state crystal structures and solution-phase structures. Even slight inaccuracies in the computer models can have a large effect on the molecular recognition of the target for a compound. To overcome these inaccuracies, one must synthesize, test and model different compounds with the target until an acceptable lead compound is discovered.

A new approach to lead compound identification was described in U.S. Patent No. 5,712,171. This patent discloses, inter alia, a method of generating a library of chemical compounds in a logically ordered, spatially arranged array format. In this method, each compound in the library has a randomly selected common molecular core structure and one or more variable structural diversity elements. In one embodiment of the invention, the compounds in each library are logically ordered in a spatially addressable array format before they are screened against a biomolecular target . The compounds exhibit structural diversity elements that can cover the maximum chemical diversity space available within the constraints imposed by the common molecular core. The compounds in such an array are related to each other by their common molecular core, yet differ in a known manner because of the differences in their respective diversity elements. When such an array is tested under the random screening method, a field of relational binding data results if a molecular recognition event occurs between the target and any of the compounds in the array. The relational binding data facilitates development of an SAR hypothesis that allows rapid and efficient optimization of a lead compound.

Although this method allows the systematic exploration of chemical diversity space within the constraint of a molecular core, the method does not overcome the drawbacks of the random screening method. To accumulate a chemical library that represents total diversity space, one must find molecular cores that collectively represent the total of chemical diversity space and then construct an array based on each of these molecular cores. As with the traditional random screening method, it would be prohibitively expensive, laborious, and time-consuming to synthesize and test the astronomical numbers of compounds required to represent total chemical diversity space, and smaller numbers of compounds would decrease the probability of identifying a lead compound (i.e., because of diversity "holes" and lack of sufficient resolution in the compound library) . Summary of the Invention The invention provides methods for making a new type of library of compounds, referred to herein as a "biased array. " The biased array is based on the premise that by preparing compounds that include (i) a common molecular core that is targeted or biased to interact with or bind to a defined class of target molecules, and (ii) one or more of a wide variety of so-called "structural diversity elements," one obtains an array that simultaneously achieves broad diversity and a very high probability of including compounds that will bind to or otherwise interact with specific members of the defined target class, but that will not interact with molecules outside of the defined target class. The clear benefit is that the new arrays will provide compounds that can serve as superior lead compounds for the development of new pharmaceutical agents that are highly specific and unlikely to have undesired general biologic activity. In one embodiment, the invention features a method for generating a biased array of compounds that preferentially bind to or interact with a particular class of biomolecular targets. The method includes the steps of: (1) selecting a molecular core structure based on its affinity for or reactivity towards the molecules of the class of biomolecular targets, where the molecular core structure does not significantly bind to or interact with molecules outside of the desired class of biomolecular targets, and where the molecular core structure has at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets; and (2) reacting a set of molecular components to generate an array of compounds, where each compound in the array includes a combination of the molecular core structure and at least one structural diversity element.

By saying that the molecular core structure "does not significantly bind to molecules outside of the desired class" is meant that the core structure binds at a level less than or equal to the compounds of a random library screened against the same molecules outside the class of biomolecular targets.

Examples of classes of biomolecular targets include proteins, DNA, RNA, carbohydrates, and lipids, as well as more specific subsets of these classes.

In some cases , each compound in the array can include two or more (e.g., two, three, four, or more) structural diversity elements. The method can also include the step of screening the array to identify those compounds in the array that exhibit a property of interest (e.g., biological activity) .

In another embodiment, the invention features a biased array that includes a plurality (e.g., ten, 100, 1000, or more) of compounds. Each compound in the array includes a molecular core structure and at least one structural diversity element bonded to the molecular core structure. It is essential that the core structure bind to or otherwise interact with a plurality of molecules of a specific class of biomolecular targets, but not significantly bind to or interact with molecules outside that class of biomolecular targets. Also, the molecular core structure must have at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets.

Still another embodiment of the invention features a method of making a logically ordered, spatially addressable, biased array of compounds having a common molecular core structure and n (e.g., one, two, three, or more) variable structural diversity elements. The method includes the steps of: (1) selecting the molecular core structure based on its affinity for or reactivity towards the molecules of a class of biomolecular targets, where the molecular core structure does not significantly bind to or interact with molecules outside of that class of biomolecular targets, and where the molecular core structure has at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets; (2) providing a plurality of reaction vessels arranged into n (e.g., one, two, three, or more) sub-arrays; (3) adding a plurality of reaction components to each of the reaction vessels in a manner such that when reacted the components form the compounds of the array, including the common molecular core structure, and such that the compounds of each sub-array differ from one another by either zero or one change in a single structural diversity element; and (4) reacting the contents of each reaction vessel under appropriate conditions to form the biased array.

Yet another embodiment of the invention features a method of making a biased array of compounds . The method includes the steps of: (1) selecting a molecular core structure based on its affinity for or reactivity towards the molecules of the class of biomolecular targets, where the molecular core structure does not significantly bind to or interact with molecules outside of said class of biomolecular targets, and where the molecular core structure has at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets; (2) apportioning into reaction vessels that are identifiable by their spatial addresses, (i) a plurality of different first reaction components, with one different first compound per reaction vessel; and (ii) a second component, with one second component per reaction vessel; and (3) reacting the first and second components under solution phase conditions to form compounds that each include a common molecular core structure, thus forming the biased array. Optionally, the method can also include the step of formatting the contents of at least a portion of the reaction vessels into a spatially addressable array.

Each of the plurality of first reaction components can include a same first reactive group and a different first structural diversity element such that the first components composing the plurality differ from one another, with one first compound per reaction vessel; and each second component comprises a second reactive group and a second structural diversity element, with one second component per reaction vessel. The first and second components can be reacted under solution phase conditions in an addition reaction to form compounds that each include the molecular core structure, thus forming the biased array.

Each component used in the preceding methods can be unique or can be reused repeatedly. The arrays made by any of the above methods are also contemplated as an aspect of the invention. Methods of identifying compounds having a property of interest (e.g., biological activity, or a binding interaction) is also an aspect of the invention; this method includes the steps of providing any of the biased arrays described above, and then identifying which compounds in that biased array exhibit the property of interest .

An "array" is a collection of compounds (e.g., as a mixture or as many individual compounds) synthesized from one or more components, or starting materials. An array can include 5 or more, 10, 50, 100 or more, 1,000 or more, or even 10,000, 50,000, or 100,000 or more compounds that differ from each other (i.e., 100 different compounds and not 100 copies of the same molecule) . Each compound in the array includes an invariant portion, termed a "molecular core structure, " and one or more variable portions, called "structural diversity elements . "

A "molecular core structure" is an invariant structural moiety common to each compound in an array. The molecular core structure can be selected based on known activity or other interaction with several members of a given class or subclass of target molecules. In general, the molecular core structures are linear, branched, or cyclic organic moieties, typically including at least three carbon atoms and at least two sites for reaction with other reagents that can form structural diversity elements.

"Structural diversity elements" are chemical functional groups and can include any combination of substituted or unsubstituted alkyl, carbocyclic, or aryl groups. Structural diversity elements can also be chemical bonds, connecting groups, or heteroatomic functional groups, and can be different in adjacent units.

Examples of "alkyl groups" include lower alkyls, for example, methyl, ethyl, n-propyl, iso-propyl, n-butyl, iso-butyl, tert-butyl; higher alkyls, for example, octyl, nonyl, and decyl; lower alkenyls, for example, ethenyl , propenyl , propadienyl, butenyl, butadienyl; higher alkenyls such as 1-decenyl, 1-nonenyl, 2 , 6 -dimethyl-5-octenyl, and 6-ethyl-5-octenyl ; and alkynyls such as 1-ethynyl, 2-butynyl, and 1-pentynyl. Other linear and branched alkyl groups are also within the scope of the present invention. In addition, one or more hydrogen or carbon atoms of the alkyl groups can been replaced by a functional group. Functional groups include, but are not limited to, tertiary amines, amides, esters, ethers, and halogens, i.e., fluorine, chlorine, bromine and iodine. Specific substituted alkyl groups can be, for example, alkoxy such as methoxy, ethoxy, butoxy, and pentoxy; dimethylamino, diethylamino, cyclopentylmethylamino, benzylmethylamino, and dibenzylamino; formamido, acetamido, or butyramido; methoxycarbonyl or ethoxycarbonyl ; or dimethyl or diethyl ether groups .

Substituted and unsubstituted "carbocyclic groups" include cyclic carbon-containing moieties such as cyclopentyl, cyclohexyl, cycloheptyl, and adamantyl . One or more hydrogen or carbon atoms of these moieties can be replaced by a heteroatoms or other functional groups, resulting in epoxides, aziridines, tetrahydrofurans, oxazolones, etc. Suitable functional groups include those described above, as well as alkyl groups as described above. Heterocyclyl compounds are also contemplated .

Substituted and unsubstituted "aryl groups" include a hydrocarbon ring bearing a system of conjugated double bonds, usually comprising (4n + 2) pi bond electrons, where n is an integer equal to or greater than 0. One or more hydrogen or carbon atoms of these moieties can be replaced by a functional group. Suitable functional groups include those described above, as well as alkyl or carbocyclic groups as described above. Examples of aryl groups include, but are not limited to, phenyl, naphthyl , anisyl, tolyl, xylyl, aryloxy, aralkyl (e.g., benzyl), aralkyloxy (e.g., benzyloxy) , and heteroaryl groups (e.g., pyrimidine, morpholine, piperazine, piperidine, benzoic acid, toluene, and thiophene) . In addition to the functional groups described above in connection with substituted alkyl groups and carbocyclic groups, suitable functional groups on the aryl groups include heteroatomic functional groups such as nitro and sulfonyl groups, and other nitrogen, oxygen, sulfur, or halogen bearing groups.

The terms "bonded," "binding," "binds," or "bound, " as used herein, can refer to, for example, covalent, ionic, van der Waals, or hydrophobic interactions. Coordination complexes and hydrogen bonding are also contemplated. Typically, the bonding interactions are reversible, but can be irreversible in some cases.

As used in this application, the terms "molecular core" (or "molecular core structure") and "structural diversity element" describe the structural moieties or domains of a chemical compound in a library and are not necessarily related to the chemical components that reacted to form the compound. In some cases, the chemical components that reacted to form the compound can exhibit a one-to-one correspondence with the molecular core and structural diversity elements. For example, in an addition reaction between two chemical components, one component can correspond to the molecular core structure and the other to a structural diversity element . In other cases, the chemical components that reacted to form the compound may have little or no correspondence with the molecular core and structural diversity elements. An example of the latter is a multicomponent condensation reaction in which the various chemical components react to form the molecular core, as well as some or all of the structural diversity elements.

Furthermore, "molecular core" (or "molecular core structure) and "structural diversity element" are relative terms. For example, a peptide mimetic that resembles the natural substrate of a class of protease enzymes can be classified "as a "molecular core" with respect to that broad class, but this same "molecular core" and a "structural diversity element" can together constitute a new "molecular core" in relation to a subclass of proteases within the broad class. This situation arises because a domain that is classified as a "structural diversity element" in relation to the entire class can be used to select between the subclasses, and can thus be reclassified as part of the molecular core structure specific to a particular subclass.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and are not intended to be limiting. An advantage of the new methods and arrays is that they provide a systematic method for exploring biologically relevant chemical diversity space, rather than total chemical diversity space. The new methods use available information about a target class to concentrate efforts and resources on arrays that are likely to produce many "hits," while simultaneously maintaining diversity within each array.

Another advantage of the new methods and arrays is that they address molecular recognition and binding events at two levels. The first level involves rational selection of a molecular core structure, and the second level is the random screening of compounds having a diverse range of structural diversity elements. The result of this two-level approach is an array that is biased towards an entire target class or subclass, including compounds that selectively bind to specific members of the target class or subclass. Accordingly, screening for compounds that bind or interact with a specific target within that class can be greatly accelerated, especially if the solution phase structure of the target is unknown or partially known.

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

Brief Description of the Drawings

Figs. 1A-1D are, respectively, drawings of a generic α-ketoamide molecular core structure, and aminomethylene oxazolone (AMO) , triazine, and thiourea derivatives thereof. Fig. 2 is a synthetic scheme for making a- ketoamide molecules and AMO derivatives thereof.

Fig. 3 is a table that lists primary screening results for arrays screened against five proteases.

Fig. 4 is an illustration of six compounds (1-6) and their respective IC₅₀ values for inhibition of cruzain and cathepsin B.

Fig. 5 is a table that lists percentage inhibition of cruzain and five other proteases by compounds 1-6. Fig. 6 is a table illustrating the activity pattern that results when a logically ordered, spatially arranged array is screened for biological activity. Detailed Description The invention is based on the observation that related biomolecular targets (i.e., members of the same class or subclass of targets) share common physiochemical attributes in the molecular recognition sites that interact with their normal biological effector molecules. These common physicochemical attributes define the relevant chemical diversity space for identifying molecular core structures that have affinity for (i.e., will bind to or otherwise interact with) the molecular recognition sites of a class or subclass of biomolecular targets. Once an appropriate molecular core structure has been identified, a logically ordered, spatially arranged array of compounds can be constructed, based on the molecular core structure. Each compound in the array includes the molecular core structure and at least one structural diversity element attached to the molecular core in a position that allows the structural diversity elements to interact with a specific member or group of members of the biomolecular target class.

The molecular recognition sites of the targets define a very specific subset of chemical diversity space. The invention enables the systematic exploration of this biologically relevant chemical diversity space, rather than total chemical diversity space, by addressing molecular recognition at two distinct levels.

Molecular Core Structures

One level of molecular recognition occurs between a common molecular recognition site shared by many or all members of a target class and either a single molecular core structure or a small number of molecular core structures that specifically interact with that common molecular recognition site as distinct from the common molecular recognition site of other target classes. On this level of molecular recognition, the use of available information about the common molecular recognition site of a target class collapses the infinite possibilities of total chemical diversity space into the finite diversity space that is relevant to the specific target class. The molecular core in the biased array can be based, for example, on a chemical structure that is a mimetic of a natural biological effector molecule.

A molecular core selected from within the biologically relevant chemical diversity space is used as the common molecular core structure to construct a logically ordered, spatially addressable compound array. At this first level of molecular recognition, therefore, the selection of a molecular core structure is more rational than random.

Structural Diversity Elements

The second level of molecular recognition occurs between structural diversity elements displayed on the molecular core structure and the individual members of the target class. At this level of molecular recognition, the various structural diversity elements displayed on the compounds in an array can be selected to maximize chemical diversity in the array subject to the constraint provided by the common molecular core . The individual members of the target class exhibit specific molecular recognition of different structural diversity elements, resulting in a selective interaction between one or more compounds in the array and an individual biomolecular target in the target class . Lead compounds can be optimized by using the SAR data generated by the logically ordered, diverse compounds in the array. At this level of molecular recognition, therefore, the selection of structural diversity elements is more random than rational but the logical ordering of the compounds in the array facilitates optimization of a lead compound after initial screens are completed.

This blend of design and serendipity enables the generation of a biased array in which there is an enhanced probability, compared with a random screening library, that at least one compound in the library will interact selectively with a particular biomolecular target in a target class.

Selection of a Molecular Core Structure An appropriate molecular core structure for a given target class can be identified through review of medicinal chemistry literature, computational chemistry predictions, correlation of actual data obtained from screening known molecular constructs with multiple targets within a defined target class, structure-activity relationships (SARs) identified using directed libraries, or combinations of these methods.

The core structure is selected on the basis of its biological function; the core can serve as a mimic or substitute for naturally occurring or synthetic molecules that are known to bind to enzymes, receptors, or other macromolecules . Ideally, the core structure itself can bind to every member of the target class; in other words, the core structure would be a consensus binding motif for the target class. In this ideal scenario, the role of the structural diversity elements is to confer selectivity and specificity for particular members of the target class.

A further consideration in the selection of a core structure is that the sites on the core to which the structural diversity elements are attached must be positioned to affect the binding of the compounds to the target. Otherwise, either all compounds in the array will likely bind to the target or else none of the compounds will bind. The -correct positioning of the structural diversity elements can be determined experimentally or can be derived from existing structural information. It is desirable, although not essential, that suitable molecular core structures and compounds prepared from them can be constructed in a rapid and concerted fashion. The molecular core structure can, for example, be able to present the structural diversity elements in controlled, varying spatial arrangements.

Binding patterns provide SAR data that is useful for further optimizing the molecular cores that bind to the target class. Insights gained through analysis of data from biased arrays can then be used to create second-generation based arrays that include compounds that are even more targeted or biased towards the target class .

There are a finite number of families of compounds (e.g., α-ketoamides, piperazines, and diazepines) of which most known drugs for treating disease are members. A number of biased arrays, each based on a core selected from one of these families, can be synthesized; later, when the target associated with a disease has been identified (e.g., through functional genomics studies), the appropriate biased array or arrays can be screened for binding or other activity towards the target or target class.

For example, serine and cysteine proteases are molecular targets for the treatment of both autogenous and infectious diseases. Several members of the class of compounds known as α-ketoamides were previously known to be protease inhibitors. Thus, pursuant to the strategy described above, biased arrays of compounds having an o>- ketoamide molecular core were made. In one biased array, for example, each member has the following structure:

where groups A, B, C, D, and Y are structural diversity elements attached to the molecular core structure. Each of the structural diversity elements was varied independently. When this biased array was screened for activity against a sample of protease targets, the hit rate was found to be much higher that the hit rate observed from screening non-ketoamide mapping arrays.

Selection of Structural Diversity Elements The structural diversity elements are ideally selected to reflect maximum diversity within the limits imposed by the molecular core structure. In practice, the range of structural diversity elements that can be used is also limited by the current state of synthetic chemical methodology. It is important, but not essential, that the synthetic methods used in preparing the arrays be amenable to automated synthesis, especially using commercially available robotic synthesis systems, and use commercially available reagents. In addition, the structural diversity elements can be selected based on practical characteristics such as non-toxicity (i.e., when looking for a drug compound) , insolubility, and charge . Making the Arrays

The synthesis of each individual compound of the array is accomplished by reacting components to form the molecular core structure and structural diversity elements. Thus, during synthesis, "components" are used to make the "members" or "individual compounds" of an array, and the terms "molecular core" (or "molecular core structure") and "diversity element" (or "structural diversity element") are used herein to describe the parts of the completed compounds of an array.

The members of the new arrays can be constructed from a wide variety of reaction components. Each component can form a part or all of a molecular core structure or structural diversity element. Thus, components can be added to reactive sites on a preexisting molecular core structure to form or attach structural diversity elements.

On the other hand, the molecular core structure and the structural diversity elements can, in some cases, be formed from the components in one or more steps, e.g., simultaneously in a single reaction step. For example, one component can include a portion of a molecular core structure and also a partial or complete structural diversity element, while a second component can include the remainder of the molecular core structure together with any remaining structural diversity elements.

Examples of molecular core structures include, but are not limited to, the oxazolones disclosed in PCT Application No. WO 95/17903, the aminimides disclosed in PCT Application No. WO 95/18186, and combinations and other scaffolds disclosed in PCT Application Nos. WO 95/32184, WO 95/22529, WO 95/12482, WO 96/35953, and U.S. Serial No. 09/009,846, all of which are incorporated herein by reference in their entirety. In general, the most appropriate synthetic conditions for use in making the individual members of the arrays can be determined by reviewing the chemical literature; there are no limits on the synthetic pathways that can be used to make the compounds of the arrays, classes of components used in the synthetic pathways, the size of the arrays, or the molecular weight of the compounds in the arrays .

In multi-component syntheses, a considerable gain in efficiency may be achieved through the introduction of convergence and parallel synthesis. Convergence refers to a synthetic strategy wherein several reaction components are brought together to form a compound ( cf . linear synthesis, in which a single component is transformed into the final compound by a sequential set of reactions with various reagents) . Convergence can be highly desirable in a multi-step synthesis because of its significant impact on yields and the overall efficiency of the synthesis. In particular, convergence can provide a considerable improvement in efficiency when utilized in combinatorial synthesis. Parallel synthesis refers to the batch synthesis of multiple compounds. Parallel synthesis is often amenable to automation. In parallel synthesis, the geometrical expansion of both the number and diversity of compounds in a product or intermediate set can be achieved by the synthesis of a matrix or array of compounds in which rows of structurally diverse, but functionally identical (i.e., from the same chemical class) , molecular components are combined with columns of diverse molecular components having complementary functionalities .

For example, rows of primary amines (R-NH₂) , each having a different R group, can be combined with columns of isocyanates (0=C=N-R' ) , each having a different R' group, resulting in an array of compounds having the structure (R-N- (C=0) -N-R' )-, where urea (-N- (C=0) -N- ) is the molecular core structure and R and R' are the structural diversity elements.

These methods therefore incorporate multiple parallel processing into the synthesis of single compounds. Because the synthesis involves a repeating set of unit operations, the synthetic organic chemistry and analysis of the products can be adapted for automation, possibly using robotics and automated data acquisition and storage.

Ordering of the Arrays

The logical layout of the arrays can be important for construction and testing. In general, the biased arrays are logically ordered, spatially addressable arrays. The differences among the compounds in the biased array can be, although are not necessarily, logically reflected in the spatial arrangement of those compounds within the array. This method enables the generation of a library of a plurality of compounds, in known locations within a 2- or 3 -dimensional array, that exhibit an enhanced probability of binding to, or otherwise interacting with, a target within a defined target class. Moreover, the desired properties of each compound in the array can be measured and correlated to specific ordered changes in the structural diversity elements .

In the construction of a biased array of the invention, the array is ordered in a manner that expedites assembly; this order can be maintained after construction of the array, or the array can be rearranged

(e.g., to maximize the informational content obtained by testing, or to facilitate the rapid extraction of data from the testing process) . A computer can be used to keep track of the compounds in the array to create a virtual logical pattern based on what amounts to a random physical pattern. This method, coupled with high- throughput screening, can accelerate the discovery of compounds within the biased array that have optimal activity towards or affinity for individual members of the class of target compounds.

The new methods and arrays provide for the rapid determination and optimization of desired biological or physical activity. Although each compound in the array will not necessarily become a lead compound, each provides information that is useful in designing subsequent arrays and in determining or developing the optimum compounds. The new methods thus provide a powerful tool for the rapid screening and testing of compounds, e.g., for the inhibition of proteases or other classes of biomolecular targets. Once an SAR pattern has been established, this information can be used to formulate an optimization plan to further develop lead compounds. Optimization of the biased arrays can be very rapid, since the structural diversity elements at each location of the array are known with certainty and can thus provide SAR data. Both positive and negative results are of value in the ultimate construction of the array and in establishing SARs . The arrays can also be constructed to permit analysis of synthesized compounds for the assurance of purity and quality. By testing a series of loci within any given sub-array, it becomes possible to determine the efficacy of construction of the compounds of that array, and eliminate those reaction components that do not provide satisfactory synthetic results. The new methods therefore indicate effectiveness of given synthetic reagents from previous results, and either advise against their further use or suggest general conditions for their efficient incorporation into the array. Again, both positive and negative results can be of value in the construction of the array and in establishing SARs. A 96-well microtiter plate can serve as the foundation for the management of both high throughput screening data and chemical data. Organic compounds arranged in numbered 96-well plates are specified by descriptors derived from row, column, and plate numbers. The descriptors are ideally suited for electronic storage and retrieval of structural databases . Each compound in the array can share at least some homology with neighboring array members (unless the compounds have been randomized) ,^' providing intuitive insight into SARs for arrays containing biologically active compounds, based on activity patterns within the array. The process can be still more useful if the design of the array incorporates structural diversity elements that are commonly found in known pharmaceutical compounds; incorporation of such elements can make the compounds within the array more "drug-like . "

General Considerations

It is to be understood that this invention is not limited to the particular target classes, compounds, molecular core structures, and structural diversity elements described herein; such target classes, compounds, core structures, and diversity elements, as well as the methods used in their manufacture and use, can, of course, vary. The invention can be applied to any specific binding interaction that involves a macromolecule . Throughout this description and the appended claims, it must be noted that the singular forms "a, " "an, " and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a molecular core structure" includes groups or subgroups of core structures. Similarly, reference to "solvent" includes reference to mixtures of solvents, and reference to "the method" includes a plurality of methods . The methods described herein can be used to create biased arrays of compounds that differ from the specific biased arrays described below, but which are also within the scope of the invention.

In this description, a compound is a molecule made up, for example, of carbon and any one or more of hydrogen, nitrogen, oxygen, iodine, bromine, chlorine, fluorine, sulphur, or phosphorous atoms. A compound can be a cyclic or an acyclic compound formed entirely of carbon and hydrogen, or it can include one or more heteroatoms, including oxygen, nitrogen, sulphur, halogen, or phosphorus, or other structural diversity elements .

Structural diversity elements can include one or more carbon, oxygen, hydrogen, iodine, bromine, chlorine, fluorine, nitrogen, sulfur, or phosphorus atoms, or any combination of these atoms. Typically, structural diversity elements are attached to the molecular core structures, and can include alkyls, alkenyls, alkynyls, and aryls, each of which can be either unsubstituted or substituted (e.g., as ketones, aldehydes, esters, carboxylic acids, nitriles, ethers, or amides) , and may be cyclic, polycyclic, heterocyclic, or acyclic. The general structure of each of these groups is well known. Structural diversity elements can also be drawn from any other groups that can be bonded to an organic compound, for example, via a carbon, oxygen, or nitrogen atom. A non-limiting list of examples of structural diversity elements includes all known protein or nucleic acid recognition elements (e.g., hydrophobic moieties, hydrophilic moieties, nucleic acid analogs, peptide mimics, and polar compounds) , as well as hydrogen, hydroxy, R_a, -0R_a, -NR_aR_b, -S0_1>2,3ι4R_a, -C(0)R_a, -C(0)OR_a, - OC(0)R_a, -0C(0)0R_a, -NR_bC(0)R_a, -C(0)NR_aR_b, -0C(0)NR_aR_b, -NR_cC(0)NR_aR_b, -NR_bC(0)OR_a, -R_a-0-R_b, -R_a-NR_bR_c, -R_a-S-R_b, -R_a-S(0)-R_b, -R_a-S(0)₂-R_b, -OR_a-0-R_b, -NR_aR_b-0-R_c, -S0_lf2>3ι4R_a- 0-R_b, -C(0)R_a-0-R_b, -C(0)OR_a-0-R_b, -OC (0) R_a-0-R_b, -0C(0)0R_a- 0-R_b, -NR_bC(0)R_a-0-R_c, -C (0) NR_aR_b-0-R_c, -OC (0) NR_aR_b-0-R_c, -NR_cC(0)NR_aR_b-0-R_d, -NR_bC (0) 0R_a-0-R_c, -0R_a-S-R_b, -NR_aR_b-S-R_c, -S0_1>2ι3ι4R_a-S-R_b; -C(0)R_a-S-R_b, -C (0) 0R_a-S-R_b, -OC (0) R_a-S-R_b, -0C(0)0R_a-S-R_b, -NR_bC(0)R_a-S-R_c, -C (0) NR_aR_b-S-R_c,

-0C(0)NR_aR_b-S-R_c, -NR_cC(0)NR_aR_b-S-R_d, -NR_bC (0) 0R_a-S-R_c, -0R_a- NR_bR_d, -NR_aR_b-NR_cR_d, -S0_1;2,3ι4R_a-NR_bR_d, -C (0) R_a-NR_bR_d, -C(0)0R_a- NR_bR_d, -0C(0)R_a-N-R_bR_d, -OC (0) 0R_a-NR_bR_d, -NR_bC (0) R_a-NR_cR_d, -C(0)NR_aR_b-NR_cR_d, -OC (0) NR_aR_b-NR_cR_d, -NR_CC (0) NR_aR_b-NHR_d, and -NR_bC (0)0R_a-NR_cR_d; where R_a, R_b, R_c, and R_d are each independently alkyl, alkenyl, alkynyl, aryl, aralkyl, aralkenyl, or aralkynyl groups having, e.g., 1 to 6, 10, 20, or even 30 carbon atoms. R_a, R_b, R_c and R_d can each be substituted, for example, with halo (e.g., 1 to 6 halogen atoms), nitro, hydroxyl , alkyl (e.g., having 1 to 6 carbon atoms), mercapto, sulfonyl, amino, acyl, acyloxy, alkylamino, dialkylamino, trihalomethyl, nitrilo, nitroso, alkylthio, alkylsufinyl , or alkylsulfonyl . The substituents can include electron withdrawing groups, electron donating groups, Lewis acids, Lewis bases, as well as polar, nonpolar, hydrophilic, and hydrophobic functional groups.

An electron withdrawing group is a moiety that is capable of decreasing electron density in other parts of a compound to which it is covalently attached. Non- limiting examples of electron withdrawing groups useful in the invention include nitro, carbonyl, cyano, iodo, bromo, chloro, fluoro, and sulfone groups.

An electron donating group is a moiety that is capable of increasing electron density in other parts of a compound to which it is -covalently attached. Non- limiting examples of electron donating groups useful in the invention include alkyl, amine, hydroxyl , and alkoxy. Examples of structural diversity elements also include aliphatic (e.g., alkyl or alkenyl), alicyclic (e.g., cycloalkyl or cycloalkenyl) , or aromatic (e.g., phenyl or naphthyl) substituents, aliphatic and alicyclic-substituted aromatic nuclei (e.g., p- (n-butyl) - phenyl or o-xylyl) , as well as cyclic functional groups wherein the ring is completed through another portion of the molecule (i.e., for example, any two indicated structural diversity elements together form an alicyclic radical) .

Hetero-substituted structural diversity elements are also contemplated. These are diversity elements that contain atoms other than carbon in a ring or chain otherwise composed of carbon atoms. Suitable heteroatoms include, for example, sulfur, oxygen, and nitrogen. Hetero-substituted diversity elements therefore include groups such as epoxides, ethers, pyridines, piperazines, furans, pyrrolidines, and imidazoles.

Biological Screening Assays

The screening of the new arrays against specific targets can result in the generation of complete relational structural information. A positive result, for example, provides information on a compound at a given spatial address. This information can be juxtaposed upon a set of systematically structural congeners (i.e., the results obtained from several arrays of related compounds or analog sets are compared with each other) , and relational structural information can be extracted from negative results in the presence of positive results. Therefore, by analyzing the activity towards or affinity for members of the class of target compounds determined during screening, and correlating this information to the known structures of the compounds at each address, an SAR can be established rapidly.

Compounds that exhibit activity and selectivity in initial screening studies can later be screened for bioavailability and non-toxicity .

Any assay for biological activity can be used to screen the compounds of the array. What is an appropriate assay for a given target depends on the target itself, the compounds being screened, and the nature of the activity being screened for. An illustrative example of a screening procedure is included in Example 1 below.

Applications of the Arrays Biased arrays can be used to find a lead compound for any specific interaction. The lead compounds can be used, for example, in the discovery of preventative agents or treatments for any disease. The arrays also have applications in agriculture, diagnostics, pharmaceuticals, bioseparations, and every other area in which there is binding between a biological molecule and any other molecule.

The invention is further described in the following examples, which do not limit the scope of the invention described in the claims.

EXAMPLES Example 1

The α-ketoamide molecular core structure shown in Fig. 1A was selected as the basis of a set of biased arrays for use in screening for inhibition of cysteine proteases. An array of 1,600 α-ketoamide-AMO compounds (Fig. IB) was synthesized using the convergent parallel synthesis scheme shown in Fig. 2, which allowed the entire array to be synthesized from just 38 components (i.e., 10 α-ketoesters, 8 diamines, and 20 ethoxymethylene oxazolones) . Combinations of stock solutions at standard concentration were prepared at 50 μM for the automated steps of the synthesis.

The α-ketoesters used in the preparation of the array were :

1. methyl 3 -trifluoromethylbenzoylformate,

2. methyl benzoylformate, 3. methyl 4 -bromobenzoylformate,

4. methyl 2, 4-difluorobenzoylformate,

5. methyl 4 -nitrobenzoylformate,

6. methyl 4 -tert-butylbenzoylformate,

7. methyl 3-methylbenzoylformate, 8. methyl 3 -methoxybenzoylformate,

9. methyl 3 -fluorobenzoylformate, and

10. methyl 4 -methoxybenzoylformate .

The diamines used were :

1. N,N' -dimethyl-1, 2-ethylenediamine, 2. N,N' -dimethyl-1, 3 -propanediamine,

3. N,N' -dimethyl-1, 6-hexanediamine,

4. piperazine,

5. homopiperazine,

6. 1, 4-diaminobutane, 7. 1, 3-cyclohexanebis (methylamine) , and

8. 1, 3-diaminopropane.

The alkoxymethylene oxazolones (AMO) used were:

1. 2-thienyl AMO,

2. 2-naphthyl AMO, 3. p-biphenyl AMO,

4. m-tolyl AMO,

5. 4-trifluoromethylphenyl AMO,

6. 2-furyl AMO, 7. 2-chlo ophenyl -AMO,

8. o-tolyl AMO,

9. 4-t-butylphenyl AMO,

10. 3-methoxyphenyl AMO, 11. 2,4-dichlorophenyl AMO,

12. 3-nitrophenyl AMO,

13. 4-bromophenyl AMO,

14. 1-naphthyl AMO,

15. 3-furyl AMO, 16. 3, 4-methylenedioxyphenyl AMO,

17. 3-pyridyl AMO,

18. p-tolyl AMO,

19. 4-chlorophenyl AMO, and

20. 4-nitrophenyl AMO.

To carry out the syntheses, methyl α-ketoesters were reacted with diamines. Stock solutions of the α- ketoesters and the diamines were prepared at 0.25 M concentration in methanol. The α-ketoester solutions were dispensed (200 μl , 0.05 mmol) , then the diamine solutions (200 μl, 0.05 mmol) were added. The reactions were incubated at 25°C for 5 days and were then concentrated. The resulting intermediate products were dissolved iri DMSO (200 μl, 0.25 M) , and then treated with a solution of the appropriate alkoxymethylene oxazolone compounds (0.25 M in DMSO, 200 μl, 0.05 mmol). The reactions were then heated to 80 °C and agitated for 24 hours .

Example 2

Arrays of α-ketoamide-triazine conjugates (9600 compounds; Fig. IC) and α-ketoamide-thiourea conjugates (3200 compounds; Fig. ID) were also synthesized using the convergent parallel synthesis described in Example 1, substituting the alkoxymethylene oxazolone precursors with triazolinyl and isothiocyanate precursors, respectively.

To carry out the syntheses, methyl α-ketoesters were reacted with diamines. Stock solutions of the < - ketoesters and the diamines were prepared as above . The α-ketoester solutions were dispensed, then the diamine solutions were added. The reactions were incubated for 5 days and were then concentrated at temperatures suitable for maintaining the stability of the triazolinyl or isothiocyanate compounds. The resulting intermediate products were dissolved in DMSO (200 μl, 0.25 M) , and then treated with a solution of the appropriate triazolinyl .or isothiocyanate compounds (0.25 M in DMSO, 200 μl, 0.05 mmol) . The reactions were then heated and agitated for 24 hours.

Example 3

High-throughput cathepsin B assay: A fluorometric high throughput cathepsin B assay for detecting inhibitory activity was developed in 96-well microtiter plate format. Benzyloxy-Phe-Arg-AMC (Z-F-R-AMC;

Molecular Probes, Inc., Eugene, OR) was the substrate used in the assay (K,,, = 150 μM) . Cathepsin B (EC 3.4.22.1, from human liver) was purchased from Sigma Chemical Company (St. Louis, MO) in a stock solution containing 25 μg of the protein in 25 μl buffer (i.e., 20 mM sodium acetate, pH 5.0, and 1 mM EDTA) .

The bioassay of the compounds was performed on a Sagian/Beckman Integrated robotic system (Beckman, Fullerton, CA) . Inhibition of enzyme activity was measured via fluorescent detection. Fluorescence readings were taken at an excitation wavelength of 390 nm (λ-_ma = 350 nm) and an emission wavelength of 430 nm (λ_max = 430 nm) on a BMG FluoStar (BMG, Durham, NC) . Chemical compounds solubilized in 100% dimethyl sulfoxide (DMSO) were pre-transferred to 96 or 384 well assay plates to yield the indicated final concentration of compound. The assay was carried out at a substrate concentration of 30 μM in an assay buffer (88 mM KH₂P0₄, 12 mM Na₂HP0₄, 1.33 mM Na₂EDTA, pH 6.0) at room temperature (22°C) . All of the reagents and the enzyme were maintained at 22 °C during the assay.

The reaction was initiated by the addition of the cathepsin B (0.46 ng/ml) in the assay buffer containing dithiothreitol (giving a final concentration of 1.5 mM) . The final concentration of DMSO in the each well was 14% to promote dissolution of the inhibitors. Within the linear range of the plot of substrate hydrolysis versus time, the reaction was quenched by trifluoroacetic acid (TFA) to a final concentration of 0.3% in each well. The activity was obtained by observing the end point absorbance reading after quenching. The percentage of inhibition was calculated by dividing the absorbance measured in the presence of inhibitor plus enzyme (inhibited signal) by the absorbance measured in the presence of enzyme alone (full signal) minus the absorbance obtained in the absence of both inhibitors and the enzyme (background) .

High-throughput cruzain assay: Cruzain protein has been expressed in bacteria as an inactive, insoluble fusion protein that is easily isolated from the bacterial lysate. The inactive enzyme has been successfully refolded and processed to yield an active form (Eakin et al., J. Biol . Chem . , 267:7411-7420, 1992). This recombinant enzyme (Eakin et al . , J. Biol . Chem . ,

2 8: 6115-6118, 1993) was used in the following assay. The recombinant enzyme used in these experiments was provided by Dr. Charles Kraik (University of California, San Francisco) . A stock solution o cruzain was prepared, having a concentration of 435 μM in 20 mM Bis-Tris, pH 5.8. A fluorometric high throughput cruzain assay similar in concept to the cathepsin B assay described above was developed. The same substrate, Z-F-R-AMC, was also used in the cruzain assay (K-, = 1.0 μM) . The assay conditions were the same as in the cathepsin B assay except as noted below. The substrate concentration in the assay was 3 μM, and the assay buffer was composed of 50 mM sodium acetate containing 5 mM EDTA, pH 5.5.

In each well, the reaction was activated by the addition of cruzain (final concentration, 0.4 nM) that had been previously activated by incubating with 5 mM dithiothreitol at 4°C. The final assay concentration of the compound was noted. The final concentration of DMSO was 10%. The reaction was quenched by the addition of 100 nM of the irreversible cruzain inhibitor E-64 (Bohringer Mannheim, Indianapolis, IN) dissolved in 1:1 water : ethanol . The assay was carried out at room temperature (22°C) . All reagent stocks were left at 22 °C, except the cruzain stock was kept at 4 °C.

Other high-throughput assays were performed using thrombin (such testing can be performed, e.g., according to the method described in Balasubramian et al . , J. Med . Chem . , 3_6: 300-303 (1993)), factor Xa (such testing can be performed, e.g., according to the method described in Rezaie et al . J. Biological Chem. , 270 (27) : 16176-16181 (1995) ) , and 92 kDa gelatinase (such testing can be performed, e.g., according to the method described in Knight et al . , FEBS Lett . , 296:263-266 (1992)) as targets .

Example 4

A chemical set of approximately 168,000 low molecular weight organic compounds was assayed for inhibitory activity against cathepsin B, cruzain, thrombin, factor Xa, and 92 kDa gelatinase using the assays described in Example 3. The 168,000 compounds were representative of 33 arrays each based on a different molecular core structure, and included 38,000 of-ketoamide derivatives and 130,000 non-α-ketoamides .

Fig. 3 summarizes the number and percentage (i.e., "hit rate") of compounds in each array that produced greater than 75% inhibition at a 10 μM compound concentration. As shown in Fig. 3, the hit rates for the 38,000 α-ketoamides against the cysteine proteases (i.e., cruzain and ^■cathepsin B) were significantly higher than those observed for the same compounds against the serine proteases (i.e., thrombin and factor Xa) or the metalloprotease (i.e., 92 kDa gelatinase). A similar difference in hit rate did not exist for the 130,000 other compounds. Specifically, Fig. 3 shows that the hit rate for the α-ketoamides against cathepsin B was 27 times higher than the hit rate for the other compounds against cathepsin B (0.754% v. 0.0273%). For cruzain, the α-ketoamide hit rate was 3.2 times higher (8.33% v. 2.60%). Three specific arrays of α-ketoamides (i.e., α- ketoamide-AMO, α-ketoamide-triazine, and α-ketoamide- thiourea) , in particular, demonstrated higher hit rates as compared to the 130,000 other compounds; the hit rates were 1.83, 16.5, and 2.2 times higher for cathepsin B, and 3.4, 11.5, and 3.68 times higher for cruzain, respectively.

Example 5 Of all of the α-ketoamide derivatives studied (see Example 4), the thiourea conjugated compounds (Fig. IB) demonstrated the greatest potency towards cruzain, and the greatest selectivity over other proteases. 307 of the 3200 compounds in the α-ketoamide-thiourea array (more than 9%) exhibited greater than 75% inhibition at 10 μM, as compared to 2 compounds with greater than 75% inhibition of cathepsin B and 0 compounds for 92 kDa gelatinase (Fig. 3) . The 307 α-ketoamide-thioureas that produced greater than 75% inhibition of cruzain were re-tested in quadruplicate at 1 μM, and six compounds were identified that produced more than 50% inhibition. Based on the structures of these compounds, and on synthetic feasibility considerations, four of these six, and two additional, structurally related compounds, were resynthesized. The six resulting compounds are shown in Fig. 4 (compounds 1-6).

To test the potency and specificity of compounds 1-6, they were tested in triplicate in the in vi tro enzyme inhibition assays described in Example 3 at dosages ranging from 10 nM to 50 μM. The percentage of inhibition was plotted against the logarithm of inhibitor concentration, and the inhibitor concentration at 50% inhibition was determined (IC₅₀) . Compounds 1-6 in Fig. 4 all demonstrated single digit micromolar potency or better against cruzain; compounds 4 and 5 were the most potent, with IC₅₀ values of 70 and 80 mM, respectively (see Fig. 4) . Compound 4 demonstrated the greatest selectivity over cathepsin B (i.e., greater than 20- fold) .

Compounds 1-6 also displayed selectivity over five serine and metalloproteases as shown in Fig. 5. Fig. 5 lists the percentage inhibition measured in quadruplicate from in vi tro assays for activity against thrombin, trypsin, plasmin, 92 kDa gelatinase (92K) , and tumor necrosis factor alpha converting enzyme (TACE) , as well as cruzain. No significant inhibitory activity was noted against the serine and metalloproteases as compared with cruzain. These data show that the α-ketoamide-based compounds are generally more potent inhibitors against cysteine proteases than against other protease classes. The data also show that certain α-ketoamide thiourea- based compounds are potent and specific inhibitors of cruzain. The discovery of these inhibitors was accelerated by the integration of high throughput testing and synthesis methods, and the logically ordered spatially addressable nature of the array (see, e.g., Fig. 6) .

Fig. 6 shows primary screening results for the α- ketoamide thiourea array at 10 μM concentration screened against cruzain. Array locations corresponding to compounds with inhibition greater than 75% are darkened. Fig. 6 illustrates the pattern of activity that results when a logically ordered, spatially arranged array is screened for biological activity. The activity pattern facilitates understanding of the structure-activity relationship between the compounds and the target.

Other Embodiments It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

What is claimed is:

Claims

1. A method for generating a biased array of compounds that preferentially bind to or interact with a particular class of biomolecular targets, the method comprising: selecting a molecular core structure based on its affinity for or reactivity towards the molecules of the class of biomolecular targets, wherein said molecular core structure does not significantly bind to or interact with molecules outside of said class of biomolecular targets, and wherein the molecular core structure has at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets; and reacting a set of molecular components to generate an array of compounds, wherein each compound in the array includes a combination of the molecular core structure and at least one structural diversity element.

2. The method of claim 1, wherein each compound in the array includes two or more structural diversity elements.

3. A biased array generated by the method of claim 1.

4. The method of claim 1, wherein the class of biomolecular targets is selected from the group consisting of proteins, DNA, RNA, carbohydrates, and lipids.

5. The method of claim 1, further comprising screening the array to identify compounds that exhibit a property of interest .

6. The method of -claim 5, wherein the property of interest is biological activity.

7. A biased array, comprising a plurality of compounds, wherein each compound includes: a molecular core structure that (i) binds to or interacts with a plurality of molecules of a class of biomolecular targets but does not significantly bind to or interact with molecules outside said class of biomolecular targets, and (ii) has at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets; and at least one structural diversity element bonded to the molecular core structure.

8. The biased array of claim 7, comprising at least ten compounds.

9. The biased array of claim 7, comprising at least 100 compounds.

10. The biased array of claim 7, comprising at least 1000 compounds.

11. A method of identifying a compound having a property of interest, said method comprising: providing a biased array of claim 7; and identifying which compounds in the array exhibit the property of interest .

12. The method of claim 11, wherein the property of interest is biological activity.

13. The method of- claim 11, wherein the property of interest is a binding interaction.

14. A method of making a logically ordered, spatially addressable, biased array of compounds having a common molecular core structure and n variable structural diversity elements, said method comprising: selecting the molecular core structure based on its affinity for or reactivity towards the molecules of a class of biomolecular targets, wherein said molecular core structure does not significantly bind to or interact with molecules outside of said class of biomolecular targets, and wherein the molecular core structure has at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets; and providing a plurality of reaction vessels arranged into n sub-arrays; adding a plurality of reaction components to each of the reaction vessels in a manner such that when reacted the components form the compounds of the array, including the common molecular core structure, and such that the compounds of each sub-array differ from one another by either zero or one change in a single structural diversity element; and reacting the contents of each reaction vessel under appropriate conditions to form the biased array.

15. The method of claim 14, wherein each component is unique.

16. The method of claim 14, wherein n is 1.

17. The method of claim 14, wherein n is 2.

18. A biased array generated by the method of claim 14.

19. A method of identifying a compound having a property of interest, said method comprising: providing a biased array of claim 18; and identifying which compounds in the array exhibit the property of interest .

20. The method of claim 19, wherein the property of interest is biological activity.

21. A method of making a biased array of compounds, said method comprising: selecting a molecular core structure based on its affinity for or reactivity towards the molecules of a class of biomolecular targets, wherein said molecular core structure does not significantly bind to or interact with molecules outside of said class of biomolecular targets, and wherein the molecular core structure has at least one attachment site for a structural diversity element, positioned such that the structural diversity element can bind to or interact with the biomolecular targets; apportioning into reaction vessels that are identifiable by their spatial addresses (i) a plurality of different first reaction components, with one different first compound per reaction vessel; and (ii) a second component, with one second component per reaction vessel; and reacting said first and second components under solution phase conditions to form compounds that each include a common molecular core structure, thus forming the biased array.

22. The method of- claim 21, further comprising: formatting the contents of at least a portion of the reaction vessels into a spatially addressable array.

23. The method of claim 21, wherein each of the plurality of first reaction components comprises a same first reactive group and a different first structural diversity element such that the first components composing the plurality differ from one another, with one first compound per reaction vessel; and each second component comprises a second reactive group and a second structural diversity element, with one second component per reaction vessel; and wherein the first and second components are reacted under solution phase conditions in an addition reaction to form compounds that each include the molecular core structure, thus forming the biased array.

24. The method of claim 21, wherein each component is unique.

25. A biased array generated by the method of claim 21.

26. A method of identifying a compound having a property of interest, said method comprising: providing a biased array of claim 25; and identifying which compounds in the array exhibit the property of interest .

27. The method of claim 26, wherein the property of interest is biological activity.