WO1999038993A1

WO1999038993A1 - Combinatorial quinolone library

Info

Publication number: WO1999038993A1
Application number: PCT/US1999/001738
Authority: WO
Inventors: Roland Ellwood Dolle; Hongzhi Zhang; Christopher D'agostino; Christopher T. Louer
Original assignee: Pharmacopeia, Inc.
Priority date: 1998-01-29
Filing date: 1999-01-28
Publication date: 1999-08-05
Also published as: AU2345999A

Abstract

A combinatorial library comprising a plurality of members represented by formula (a), is disclosed in which (b) is a solid support and -Z is a compound residue. In this library containing quinolone compounds, Z is (c). R1 is a primary amine substituent; W is selected from a group consisting of stannanes and amines. The combinatorial library can be optionally encoded with identifiers T'-L, which are covalently attached to the solid support. A method of synthesizing such a library is also disclosed. The use of such library in assays to discover biologically active compounds is also disclosed.

Description

Combinatorial Quinolone Library

Cross-Reference Combinatorial Hydroxypropylamine Library, U.S. Ser. No. 08/592,654, filed

January 26, 1996, is incorporated herein by reference.

All patents and other references cited herein are hereby incorporated by reference.

Field of the Invention The present invention relates to the synthesis of chemical compounds for biological assay, and more particularly, to the synthesis of combinatorial libraries of substituted quinolones.

Background of the Invention Methodologies for the synthesis of vast numbers of diverse compounds that can be screened for various possible physiological or other activities are advantageous (Ellman, et al., Chem. Rev.. 555-600 (1996)). Procedures have been developed in which individual units are added sequentially as part of the chemical synthesis to produce all or a substantial number of the possible compounds that can result from all the different choices possible at each sequential stage of the planned synthesis. Many diverse compounds are produced by a series of reactions of a multiplicity of synthons in various combinations. Each compound in a combinatorial library results from the reaction of a subset of the synthons. For these techniques to be successful, the compounds should be amenable to methods by which one can determine the structure of the compounds so made. A premier example of such techniques is the production of oligonucleotide "tags" in parallel with oligopeptide compounds of interest (Brenner and Lemer PNAS USA 81. 5381-5383 (1992); WO 93/20242). Methods for particle- based synthesis of random oligomers wherein identification tags on the particles are used to facilitate identification of the synthesized oligomer sequence are known (WO 93/06121). A detachable tagging system that is useful in sequential synthesis of large numbers of compounds has been disclosed (Ohlmeyer et al.. Proc. Natl. Acad. ScL 90, 10922-10926(1993)).

Summary of the Invention

The present invention relates to a combinatorial library of substituted quinolone compounds optionally encoded with tags. The present invention also relates to the use of this library containing substituted quinolone compounds in assays to discover biologically active compounds.

I. Preferred Embodiments

In one aspect, the invention relates to a combinatorial chemical library for biological assay comprising a plurality of members of the Formula I:

(T-L)_Q- -Z

wherein: T^* is a tag; L- is a first linker;

T'-L when taken together, form an identifier residue; q is 0-15;

S is a solid support; wherein — is — ^ , wherein

L'- is a second linker;

— is a solid support to which the second linker attaches;

0 0

W N

Z is a compound of formula R , wherein R¹ is chosen from the group consisting of C ι.₂o alkyl, aryl, arylalkyl, either aryl or heteroaryl fused to a 3- or 4-membered moiety to form a non-aromatic second ring, cycloalkyl, cycloalkylalkyl, heterocycloalkylalkyl, heteroaryl, heteroarylalkyl, substituted alkyl, substituted aryl and substituted heteroaryl;

W is R² or R³NHCH₂Y-, wherein R² is chosen from the group consisting of alkenyl, alkynyl, aryl, heteroaryl, substituted alkenyl, substituted alkynyl, substituted aryl and substituted heteroaryl; R³ is chosen from the group consisting of alkyl, alkoxyalkyl, arylalkyl, cycloalkylalkyl, heteroalkyl, heteroarylalkyl and heterocycloalkylalkyl; and

Y is chosen from the group consisting of aryl, heteroaryl, substituted aryl and substituted heteroaryl.

A preferred embodiment of the invention is a library comprising a plurality of members of Formula I wherein: R¹ is chosen from the group consisting of the amine residues of Table 1;

W is R² or R³NHCH₂Y- wherein:

R" is chosen from the group consisting of the stannane residues of Table 2; R³ is chosen from the group consisting of the amine residues of Table 3; and

Another preferred embodiment of the invention is a combinatorial library comprising a plurality of members of Formula I wherein: R¹ is chosen from the group consisting of benzyl, butyl, 2-cyanoethyl, cyclopropylmethyl, cyclohexylmethyl, 2-dimethylaminoethyl, 3,3-dimethylbutyl, furfuryl, 3-(l-imidazoyl)ethyl, isobutyl, 2-methylbutyl, 4-methylphenethyl, 2-methylthioethyl, 2-(l-morpholinyl)ethyl, myrantyl, 1-naphthylenemethyl, piperonyl 2-phenylphenethyl, 2-( 1 -pyrrolidiny l)ethyl, 2-(2-aminoethyl)- 1 -pyrrolidinyl, 2-pyridylethyl, 3-pyridylmethyl,tetrahydrofurfuryl and 2-thienylmethyl.

Yet another preferred embodiment of die invention is a combinatorial library comprising a plurality of members of Formula I wherein:

R² is chosen from the group consisting of allyl, 1-ethoxyvinyl, 4-formylphenyl, 3-formyl-2-thienyl, 2-furanyl, phenyl, 2-phenylethynyl,thienyl and vinyl.

A further preferred embodiment of the invention is a combinatorial library comprising a plurality of members of Formula I wherein:

R³ is chosen from the group consisting of butyl, benzyl, cyclohexylmethyl, 3-isopropoxypropyl, 2-methoxyethyl, phenethyl, 2-phenylphenethyl, piperonyl, 3-pyridylmethyl and 3-(l-pyrrolidi-2-one^propyl.

An additional preferred embodiment of the invention is a combinatorial library comprising a plurality of members of Formula I wherein:

Y is chosen from the group consisting of m-substituted phenyl, p-substituted phenyl, and 3-substituted thienyl.

Another preferred embodiment of the invention is a combinatorial library comprising a plurality of members of Formula I wherein: (T'-L) is a compound of Formula II

— o-Ar

π

wherein: n = 3-12; and

Ar is halophenyl.

A preferred embodiment of the invention is a combinatorial library comprising a plurality of members of Formula I wherein:

(T'-L) is a compound of Formula II wherein: n = 3-12; and

Ar is pentachlorophenyl.

Yet another preferred embodiment of the invention is a combinatorial library comprising a plurality of members of Formula I wherein:

-L' is an acid-cleavable linker. Another preferred embodiment of the invention is a combinatorial library comprising a plurality of members of Formula I wherein:

-L' is selected from

and Ύ O-IΠC °H,

Depending on the choice of L\ the compounds -Z of Formula I may be detached by acidic, oxidative, or other cleavage techniques. For example, when L' of ^o-

Formula I is , acidic cleavage may be represented by:

(T'-L)q-( sJ-L (T'-L)q-( s -L'-H W

wherein:

O O

R m, represents a cleaved compound of Formula HI.

Another aspect of the invention is the use of the herein described combinatorial library in assays to discover biologically active compounds of Formula III. Thus, another aspect of the invention is a method for identifying a compound having a desired characteristic which comprises testing a combinatorial library comprising a plurality of members of Formula I, either attached to, or detached from, the solid supports, in a biological assay which identifies compounds of Formula HI having the desired characteristic.

A further aspect of the invention is determining the structure of any compound identified as having the desired biological activity.

Within the scope of the present invention, the chemical structures of compounds that are identified by biological assays as having a desired characteristic can be determined either by decoding the tags ( , T'-L- of Formula I) (Still et al., "Complex Combinatorial Chemical Libraries Encoded With Tags", WO 94/08051) or by deconvolution of the library (Smith et a Biomed. Chem. Lett. 4. 2821 (1994); Kurth et al., J. Ore. Chem. 59, 5862 (1994); Murphy et al., J. Am. Chem. Soc. 117, 7029 (1995); Campbell et al., J. Am. Chem. Soc. 118, 5381 (1995); and Erb et al., Proc. Natl. Acad. Sci. USA 91, 11422 (1994)).

Another preferred embodiment of the invention is the use of hydroxyl-

functionalized polyethylene glycol-grafted polystyrene resin as — in Formula I.

s v__0H s ^v — is — , as defined and ^^ is ^—^ , as shown in the schemes for simplicity. A further embodiment of the invention is a method of synthesizing a library comprising a plurality of members of Formula I which comprises: a. attaching an acyl Meldrum's acid to a solid support via its carbonyl group to form a resin-linked β-keto ester;

b. attaching a dimethyl acetal to the resin-linked β-keto ester to provide a first resin- linked enamine; c. displacing a dimemyl amino moiety from the first resin-linked enamine with a primary amine to form a second resin-linked enamine; d. cyclizing the second resin-linked enamine with a bis(trimethylsilyl)amide salt in an anhydrous solvent to form a resin-linked quinolone derivative; and e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether with a trialkyl stannane in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a resin-linked quinolone.

An additional preferred embodiment of the invention is a method of synthesizing a library comprising a plurality of members of Formula I which comprises: a. attaching an acyl Meldrum's acid to a solid support via its carbonyl group to form a resin-linked β-keto ester;

b. attaching a dimethyl acetal to the resin-linked β-keto ester to provide a first resin-

linked enamine; c. displacing a dimethyl amino moiety from the first enamine resin with a primary amine to form a second resin-linked enamine; 10

d. cyclizing the second resin-linked enamine with a bis(trimethylsilyl)amide salt in an anhydrous solvent to form a resin-linked quinolone derivative; e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether with a trialkyl stannane in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a resin-linked quinolone containing an aldehyde moiety; f. coupling the resin-linked quinolone containing an aldehyde moiety with a primary amine in the presence of a trialkyl orthoformate and a cyanoborohydride followed by treatment with an acid to produce an intermediate resin-linked amino quinolone salt; and g. neutralizing the resin-linked amino quinolone salt with a tertiary amine, followed by washing with anhydrous solvents to produce a resin-linked amino quinolone.

A preferred embodiment of the invention is a method of synthesizing a library comprising a plurality of members of Formula I which comprises: cleaving a resin-linked amino quinolone from the resin to provide an amino quinolone of Formula HI.

A further embodiment of the invention is a method of synthesizing a library comprising a plurality of members of Formula I which comprises: a. attaching an acyl Meldrum's acid of the formula:

⁰ O

10 1 1

in toluene to a solid support suspended in toluene to form a resin-linked β-keto ester

of the formula:

b. attaching a dimethyl acetal in diethyl ether to the resin-linked β-keto ester to

provide a first resin-linked enamine of the formula:

(T'-L)_Q

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R'NHI to form a second resin-linked enamine of the formula:

O O

Br' > d. cyclizing the second resin-linked enamine with lithium bis(trimethylsilyl)amide in THF to form a resin-linked quinolone derivative of the formula:

L--j S QJ-rr''-T ) 'q_Q

; and 12

e. treating the resin-linked quinolone derivative in ethylene glycol diedryl ether with a tributyl stannane of the formula Bu₃SnR" in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to afford a resin-linked quinolone of the formula: O

wherein:

R^l is chosen from the group consisting of C ι._2u alkyl, aryl, arylalkyl, either aryl or heteroaryl fused to a 3- or 4-membered moiety to form a non-aromatic second ring, cycloalkyl, cycloalkylalkyl, heterocycloalkylalkyl, heteroaryl, heteroarylalkyl, substituted alkyl, substituted aryl and substituted heteroaryl; and

R² is chosen from the group consisting of alkenyl, alkynyl, aryl, heteroaryl, substituted alkenyl, substituted alkynyl, substituted aryl and substituted heteroaryl.

Another preferred embodiment of the invention is a method of synthesizing a library comprising a plurality of members of Formula I which comprises: a. attaching an acyl Meldrum's acid of the formula:

0 o o 13

of the formula:

O O σ- )_Q

b. attaching a dimethyl acetal in diethyl ether to the resin-linked β-keto ester to provide a first resin-linked enamine of the formula:

O O

(T'-L),

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R^!NH₂ to form a second resin-linked enamine of the formula:

O O

L s -(T'-L)_q

Br

d. cyclizing the second resin-linked enamine with lithium bis(trimethylsilyl)amide in THF to form a resin-linked quinolone derivative of the formula:

O O

*7 \ S τ^*-L)_q

Br

; and 14

e. treating the resin-linked quinolone derivative in ethylene glycol diethyl ether with a tributyl stannane of the formula Bu₃SnR² in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a resin-linked quinolone of me formula:

O O

wherein:

R¹ is chosen from the group consisting of the amine residues of Table 1; and R² is chosen from the group consisting of the stannane residues of Table 2.

A further preferred embodiment of the invention is a method of synthesizing a library comprising a plurality of members of Formula I which comprises: a. attaching an acyl Meldrum's acid of the formula:

O

of the formula: 15

O

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R^f H₂ to form a second resin-linked enamine of the formula:

O O

; and

e. treadng the resin-linked quinolone derivative in ethylene glycol diethyl ether with a tributyl stannane of the formula Bu₃SnR" in the presence of triphenylphosphine 16

and palladium dibenzylidene acetone complex to afford a resin-linked quinolone of the formula:

0 O

wherein:

R¹ is chosen from the group consisting of benzyl, butyl, 2-cyanoethyl, cyclopropylmethyl, cyclohexylmethyl, 2-dimethylaminoethyl, 3,3-dimethylbutyl, furfuryl, 3-(l-imidazoyl)ethyl, isobutyl, 2-methylbutyl, 4-methylphenethyl, 2-methylthioethyl, 2-(l-morpholinyl)ethyl, myrantyl, 1-naphthylenemethyl, piperonyl 2-phenylphenethyl, 2-(l-pyrrolidinyl)ethyl, 2-(2-aminoethyl)-l-pyrrolidinyl, 2-pyridylethyl, 3-pyridylmethyl, tetrahydrofurfuryl, and 2-ti ienylmethyl; and

R is chosen from the group consisting of allyl, 1-ethoxy vinyl, 4-formylphenyl, 3-formyl-2-thienyl, 2-furanyl, phenyl, 2-phenylethynyl,thienyl and vinyl.

Yet another preferred embodiment of the invention is a method of synthesizing a library comprising a plurality of members of Formula I which comprises: a. attaching an acyl Meldrum's acid of the formula:

17

of the formula:

0 0

^Br N(Me)₂

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R'NHT to form a second resin-linked enamine of the formula:

O O

L S -( -L

0 0 s (T'-L)_q

Br

R¹ 18

e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether with a tributyl stannane of the formula Bu₃SnR~ in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to afford a resin-linked quinolone containing an aldehyde moiety of formula:

O 0

f. coupling the resin-linked quinolone containing an aldehyde moiety with a primary amine of the formula R³NH₂ in the presence of a trialkyl orthoformate and a cyanoborohydride, followed by treatment with an acid to produce an intermediate resin-linked amino quinolone salt of the formula:

O O

; and

g. neutralizing the resin-linked amino quinolone salt with a tertiary amine, followed by washing with anhydrous solvents to produce a resin-linked amino quinolone of the formula:

O O L^T s \(T'-L)_q

R D3^J.NHCH₃Y 19

wherein:

R¹ is chosen from the group consisting of C ι._2o alkyl, aryl, arylalkyl, either aryl or heteroaryl fused to a 3- or 4-membered moiety to form a non-aromatic second ring, cycloalkyl, cycloalkylalkyl, heterocycloalkylalkyl, heteroaryl, heteroarylalkyl, substituted alkyl, substituted aryl and substituted heteroaryl;

R² is chosen from the group consisting of alkenyl, alkynyl, aryl, heteroaryl, substituted alkenyl, substituted alkynyl, substituted aryl and substituted heteroaryl;

R³ is chosen from the group consisting of alkyl, alkoxyalkyl, arylalkyl, cycloalkylalkyl, heteroalkyl, heteroarylalkyl and heterocycloalkylalkyl; and

O

20

Docket No. PH 1031

of the formula: 0

provide a first resin-linked enamine of the formula:

O 0

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R'NH₂ to form a second resin-linked enamine of the formula:

O O

L s ( -L)_q

O O S -(T'-L)_q

21

e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether with a tributyl stannane of the formula Bu SnR² in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a resin-linked quinolone containing an aldehyde moiety of formula:

O 0

O O

-L S

R³NH₂CH₂Y

; and

0 O

*7 ^"N s V(T'-L)_q

R³NHCH,Y

22

wherein:

R^l is chosen from the group consisting of the amine residues of Table 1; R² is chosen from the group consisting of the stannane residues of Table 2; R³ is chosen from the group consisting of the amine residues of Table 3; and Y is chosen from the group consisting of aryl, heteroaryl, substituted aryl, and substituted heteroaryl.

J^ JL JL

Br' o\

of the formula:

O O

provide a first resin-linked enamine of the formula: 23

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R'NHa to form a second resin-linked enamine of the formula:

O O

e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether with a tributyl stannane of the formula Bu₃SnR² in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a resin-linked quinolone containing an aldehyde moiety of formula:

O 0

'LA s T-L)_Q

HCY

II ^'N^* o 24

f. coupling the resin-linked quinolone containing an aldehyde moiety with a primary amine of the formula R³NH₂ in the presence of a trialkyl orthoformate and a cyanoborohydride, followed by treatment with an acid to produce a resin-linked intermediate amino quinolone salt of the formula:

O O

R³ H₂CH₂Y'^^::^N

F> ; and

O 0

R³NHCH₂Y 7' "^ "N

R7¹ wherein:

R¹ is chosen from the group consisting of benzyl, butyl, 2-cyanoethyl, cyclopropylmethyl, cyclohexylmethyl, 2-dimethylaminoethyl, 3,3-dimethylbutyl, furfuryl, 3-(l-imidazoyl)ethyl, isobutyl, 2-methylbutyl, 4-methylphenethyl,

2-methylthioethyl, 2-(l-morpholinyl)ethyl, myrantyl, 1-naphthylenemethyl, piperonyl 2-phenylphenethyl, 2-( 1 -py rrolidinyl)ethy 1, 2-(2-aminoethy 1)- 1 -pyrrolidinyl, 2-pyridylethyl, 3-pyridylmethyl,tetrahydrofurfuryl and 2-thienylmethyl. 25

R² is chosen from the group consisting of allyl, 1 -ethoxyvinyl, 4-formylphenyl, 3-formyl-2-thienyl, 2-furanyl, phenyl, 2-phenylethynyl,thienyl and vinyl.

R³ is chosen from the group consisting of butyl, benzyl, cyclohexylmethyl, 3-isopropoxypropyl, 2-methoxyethyl, phenethyl, 2-phenylphenethyl, piperonyl, 3-pyridylmethyl and 3-(l-pyrrolidi-2-one)propyl.

Y is chosen from the group consisting of m-substituted phenyl, p-substituted phenyl and 3-substituted thienyl.

An additional embodiment of each of the above inventions further comprises cleaving a resin-linked quinolone from said resin to provide a quinolone.

DETAILED DESCRIPTION OF THE INVENTION

II. Abbreviations and Definitions

The following abbreviations and terms have the indicated meanings throughout:

ACN = acetonitrile aq. = aqueous

CAN = eerie ammonium nitrate dba = dibenzylidene acetone

DCM = dichloromethane = methylene chloride = CH;C1₂ 26

DEAD diethyl azodicarboxylate

DEE ethylene glycol diethyl ether

DMAP 4-dimethylaminopyridine

DMF N,N-dimethylformamide ECD electron capture detector

Et ethyl

EtOAc ethyl acetate

GC gas chromatography

HPLC high performance liquid chromatography Hunig's base = diisopropylethylamine

Kj inhibition coefficient

Meldrum's acid = 2,2-dimethy 1- 1 ,3-dioxane-4,6-dione

Me methyl

MSTFA N-methyl-N-trimethylsilyltrifluoroacetamide

NAD β-Nicotinamide adenine dinucleotide

NADH β-Nicotinamide adenine dinucleotide, reduced form

NADP β-Nicotinamide adenine dinucleotide phosphate

NADPH β-Nicotinamide adenine dinucleotide phosphate, reduced form

NBD 7-Nitrobenz-2-oxa- 1 ,3-diazo-4-yl PEG polyethylene glycol psi = pounds per square inch r.t. = room temperature

SHPTP1 protein tyrosine phosphatase 27

TFA = trifluoroacetic acid

THF = tetrahydrofuran

TMOF = trimethyl orthoformate

TMS = trimethylsilyl TPP = triphenylphosphine wt. = weight

"Alkoxy" means alkoxy groups of from 1 to 8 carbon atoms of a straight- branched or cyclic configuration and combinations thereof. Examples of alkoxy groups include methoxy, ethoxy, propoxy, isopropoxy, cyclopropyloxy.cyclohexyloxy and the like.

"Alkyl" is intended to include linear or branched hydrocarbon structures and combinations thereof. "Lower alkyl" means alkyl groups of from 1 to 12 carbon atoms. Examples of alkyl groups include methyl, ethyl, propyl, isopropyl, butyl, s-and t-butyl, pentyl, hexyl, octyl and the like. "Alkenyl" includes C₂ -C₈ hydrocarbons of a linear, branched or cyclic

(C₅ - C₈) configuration and combinations thereof containing at least one carbon- carbon double bond. Examples of alkenyl groups include vinyl, allyl, isopropenyl, pentenyl, hexenyl, c-hexenyl, 1-propenyl, 2-butenyl, 2-methyl-2-butenyl, 2,4-hexadienyl and the like. "Alkynyl" includes C₂ -C₈ hydrocarbons of a linear, branched, or cyclic

(C₅ - C₈) configuration and combinations thereof containing at least one carbon- carbon triple bond. Examples of alkynyl groups include ethynyl, propynyl, butynyl, pentynyl, 3-methyl-l -butynyl, 3,3-dimethyl-l-butynyl and the like. 28

"Aryl" means a 5- or 6-membered aromatic ring; a bicyclic 9- or 10-membered aromatic ring system; or a tricyclic 13- or 14-membered aromatic ring; each of which is optionally substituted with 1-3 groups selected from lower alkyl, substituted alkyl, alkenyl, alkynyl, halogen, hydroxy, alkoxy, methylenedioxy, alkoxyethoxy, cyano, acylamino, phenyl, benzyl, phenoxy, napthyloxy, aryloxy, benzyloxy .heteroaryl and heteroaryloxy. Each of the foregoing phenyl, benzyl, phenoxy, benzyloxy.heteroaryl and heteroaryloxy substituents may be optionally substituted with 1-3 substituents selected from lower alkyl, alkenyl, alkoxy, alkynyl, benzyl, benzyloxy, carboxamido, cyano, formyl, halogen, heteroaryl, heteroaryloxy, hydroxy.nitro and phenyl. "Arylalkyl" means an alkyl containing an aryl ring. For example: benzyl, phenethyl, 4-chlorobenzyl and the like.

"Aryloxy" means a phenoxy group where the aryl ring is optionally substituted with 1 to 2 groups selected from halo, alkoxy or alkyl.

"Cycloalkyl" includes cyclic hydrocarbon groups of from 3 to 12 carbon atoms. Examples of "cycloalkyl" groups include c-propyl, c-butyl, c-pentyl, c-hexyl, 2-methylcyclopropyl, norbornyl.adamantyl and the like.

"Cycloalkylalkyl" means an alkyl substituted with a cycloalkyl functionality. Examples include cyclopropylmethyl, cyclohexylmethyl and myrantyl.

"Halogen" includes F, Cl, Br, and I, with F and Cl as the preferred groups. "Halophenyl" means phenyl substituted by 1-5 halogen atoms. Halophenyl includes pentachlorophenyl, pentafluorophenyl, and 2,4,6-trichlorophenyl.

"Heterocycloalkyl" means a cycloalkyl ringcontaining 0-2 heteroatoms selected from O, N, and S; where the methylene H atom may be optionally substituted with alkyl, alkoxy, formyl or halogen. Both methylene hydrogens on a particular carbon atom may be replaced with carbonyl.

"Heteroaryl" means a 5- or 6-membered heteroaromatic ring containing 0-2 heteroatoms selected from O_rN, and S: or abicyclic 9- or 10-membered heteroaromatic ring system containing 0-2 heteroatoms selected from O, N, and S; where the methine H atom may be optionally substituted with alkyl, alkoxy, formyl or halogen. The 5- to 10-membered aromatic heterocyclic rings include imidazole, pyridine, indole, thiophene, benzopyranone, thiazole, furan, benzimidazole, quinoline, isoquinoline, quinoxaline, pyrimidine, pyrazine, tetrazole and pyrazole. "Heteroarylalkyl" means an alkyl containing a heteroaryl ring. For example: pyridinylmethyl, pyrimidinylethyl and the like.

"Substituted alkyl" means an alkyl or branched alkyl optionally substituted with groups such as alkoxy, alkylsulfide, amino, cyano, formyl, halogen, hydroxy and nitro. For the purpose of the present invention, the term "combinatorial library" means a collection of molecules based upon a logical design and involving the selective combination of building blocks by means of simultaneous chemical reactions. Each species of molecule in the library is referred to as a member of the library. The combinatorial library of the present invention represents a collection of molecules of sufficient number and diversity of design to afford a rich molecular population from which to identify biologically active members. 30

For the purpose of the present invention, "residue" shall mean the portion of the reagent that is incorporated into the product molecule after the reaction between the reagent and die molecule designated as the starting material. For examples, the residues of primary amines (R^lNH₂) are shown in Table I; residues of trialkyl stannanes (R²SnR₃) in Table TJ.; and residues of primary amines (R'NHi) in Table IE.

For the present invention, R¹ chosen from the "amine residues" shall mean R¹ chosen from those residues as shown in Table I. Examples of the substituents of "amine residues" are R¹ as C ι.₂₀ alkyl, aryl, arylalkyl, either aryl or heteroaryl fused to a 3- or 4-membered moiety to form a non-aromatic second ring, cycloalkyl, cycloalkylalkyl, heterocycloalkylalkyl, heteroaryl, heteroarylalkyl, substituted alkyl, substituted aryl and substituted heteroaryl as substituents on amino groups. Specific examples are benzyl, butyl, 2-cyanoethyl, cyclopropylmethyl, 2-dimethylaminoethyl, 3,3-dimethylbutyl, furfuryl, 3-(l-imidazoyl)ethyl, isobutyl, 2-methylbutyl, 4-methylphenethyl, 2-(l-morpholinyl)ethyl, myrantyl, 1-naphthylenemethyl, piperonyl 2-phenylphenethyl, 2-(l-pyrrolidinyl)ethyl, 2-(2-aminoethyl)-l-pyrrolidinyl, 2-pyridylethyl, 3-pyridylmethyl,tetrahydrofurfuryl and 2-thienylmethyl as substituents on amino groups.

For the present invention, R" chosen from the "stannane residues" shall mean

R² chosen from those residues as shown in Table II. Examples of the substituents of "stannane residues" are R² as alkenyl, alkynyl, aryl, heteroaryl, substituted alkenyl, substituted alkynyl, substituted aryl and substituted heteroaryl. Specific examples are 31

allyl, 1 -ethoxyvinyl, 4-formylphenyl, 3-formyl-2-thienyl, 2-furanyl, phenyl, 2- phenylethynyl, thienyl and vinyl.

For die present invention, R chosen from the "amine residues" shall mean R chosen from those residues as shown in Table HI. Examples of the substituents of "amine residues" are R³ as alkyl, alkoxyalkyl, arylalkyl, cycloalkylalkyl, heteroalkyl, heteroarylalkyl and heterocycloalkylalkyl as substituents on amino groups. Specific examples are butyl, benzyl, cyclohexylmethyl, 3-isopropoxypropyl, 2-methoxyethyl, phenethyl, 2-phenylphenethyl, piperonyl, 3-pyridylmethyl and 3-(l-pyrrolidi-2-one)propyl as substituents on amino groups.

In the present invention, the "reagent" shall mean the chemical entity shown in the reaction scheme or named and described in the specification. The reagent is reacted with the molecule designated as the starting material. For examples, reagents are the primary amines shown in Scheme 3 and Scheme 5, respectively, and the tributyl stannanes in Scheme 4.

The linker L' preferably is an acid-cleavable linker. For the structures as shown, the compound Z is attached to the benzylic oxygen where the oxymethylene is para to the methylene oxygen attached to the support. If another substituent is present, for example methoxy, then this particular substituent is placed ortho to the oxymethylene group that is attached to the compound Z. 32

The identifiers or tags of this invention, T'-L of Formula I are chemical entities which possess several properties. The identifiers are detachable from the solid supports, preferably by oxidative cleavage. They are individually differentiable, and preferably separable from one anotiier. The identifiers must be stable under the synthetic conditions and capable of being detected at very low concentrations (i.e., IO^"15 to IO^"9 mole). Preferred identifiers are discerned with readily available technical equipment operated by someone with the capabilities of one skilled in analytical techniques. The identifiers are relatively economical, and each is usually found attached to the solid supports at concentrations of at least 0.01 picomol, usually 0.1-10 pmol per bead after synthesis of the combinatorial library. The tags may be structurally related or unrelated , e.g. a homologous series, repetitive functional groups, related member of the Periodic Chart, different isotopes, combinations thereof or the like. Distinguishing features may be the number of repetitive units, such as methylene groups in an alkyl moiety; alkyleneoxy groups in a polyalkyleneoxy moiety; halo groups in a polyhalo compound; α- and/or β-substituted ethylene groups

where the substituents may be alkyl, alkoxy, carboxy, amino, halo or the like; isotopes; etc. Suitable tags and methods for their employment are described in US patent application 08/743,960 filed October 5, 1996, herein incorporated by reference.

TentaGel™ S-PHB (available from Rapp Polymere, Tubingen, Germany) is the hydroxyl-functionalized polyethylene glycol-grafted polystyrene resin. The material upon which the syntheses of the present invention are performed are referred to as solid supports, beads and resins. These terms are intended to include: beads, pellets, disks, fibers, gels or particles such as cellulose beads, pore-glass beads, silica gels, polystyrene beads optionally cross-linked with divinylbenzene and optionally grafted with polyethylene glycol and optionally functionalized with amino, hydroxy, carboxy, or halo groups, grafted co-poly beads, polyacrylamide beads, latex beads, dimethylacrylamide beads optionally cross-linked with N,N'-bis-acryloyl ethylene diamine, glass particles coated with hydrophobic polymer, etc.( i.e., material having a rigid or semi-rigid surface) and soluble supports such as low molecular weight, non- cross-linked polystyrene.

III. Optical Isomers - Diastereomers - Geometric Isomers Some of the compounds described herein contain one or more asymmetric centers and may thus give rise to enantiomers, diastereomers and other stereoisometric forms which may be defined in terms of absolute stereochemistry as (R> or (S)- or as (D)- or (L)- for amino acids. The present invention is meant to include all such possible diastereomers as well as their racemic and optically pure forms. Optically active (R)- and (S)- or (D)- and (L)- isomers may be prepared using chiral synthons or chiral reagents, or resolved using conventional techniques. When the compounds" described herein contain olefinic double bonds or other centers of geometric asymmetry and unless specified otherwise, it is intended to include both (E)- and (Z)- geometric isomers. Likewise, all tautomeric forms are intended to be included.

IV. Utility

The library of the present invention is useful as a screening tool for discovering new lead structures by evaluation across an array of biological assays, including the discovery of selective inhibition patterns across isoenzymes. The library 34

is thus a tool for drug discovery defined as a means to identify novel lead compounds by screening the library against a variety of biological targets and to develop structure-activity relationships (SAR) in large families of related compounds. The combinatorial libraries may be tested with the compounds attached to the solid supports as depicted in Formula I, or the compounds may be detached prior to evaluation as shown by Formula HI. With the compounds of Formula I, an array of library beads can be positioned on a Petri dish. Release of the compounds from the solid supports and addition of an enzyme, such as bacterial DNA topoisomerase, allows for detection of zones of inhibition. Quinolones are well known as antibacterial agents inhibiting DNA topoisomerase (Brown, E., Reeves, D., Antibiot. Chemother.. (7^th Ed.), 419-452, (1997); Shen, L.; Chu, D.; Current Pharmaceutical Design.2, 195-208 (1996); Sanchez et al., J. Med. Chem..38, 4478-4487, (1995); MacDonald, et al.. Tetrahedron Lett..37, 4815-4818 (1996); US 5,324,483, 1994). Fluorescence activated cell sorting (FACS) is a screening method that can be used to detect compound activity. A particularly useful method for identifying activities with respect to a large variety of enzymes and molecular targets is the lawn assay disclosed in US patent application 08/553,056, filed November 3, 1995.

In a lawn assay, a library of solid supports, preferably beads, is screened for the ability of compounds on the supports to affect the activity of an enzyme. Using the lawn assay, supports containing the active compounds are quickly and easily located merely by viewing zones of inhibition in a matrix. In one embodiment, the solid supports are contacted with a colloidal matrix such as agarose. The compounds are linked to the supports by a cleavable linker and released, e.g., by 35

exposure to light. As they slowly diffuse out of the solid supports, zones of high concentration of the compounds are created in the supports' immediate vicinity. The compounds contact enzyme contained in the matrix. Substrate is contacted with the matrix and reacts with the enzyme. Conversion of substrate to product is measured by monitoring a photometric change in the substrate or in a coenzyme or cofactor involved in reaction. For example, the substrate can be fluorogenic, i.e., becoming fluorescent when converted to product. In this case, compounds mat are active inhibitors of the enzyme reaction are detected as dark zones of inhibition. The less active, or inactive, compounds are contained in the lighter areas.

Using this assay, positive results from an assay of a combinatorial library can be detected very quickly. Furthermore, compound activity can be quantitated by, e.g., comparing the sizes of zones of activity. Once zones of activity have been determined, the relevant supports at the center of the zones can be located and the active compounds on those supports identified. The lawn assay thus allows large libraries of compounds to be quickly and easily screened. Very little effort is required to array the solid supports or assay the compounds released from the supports.

In another embodiment, the lawn assay is used to determine compounds that bind to a target molecule and thereby affect a detectable signal generated by a labeled compound bound to the target molecule. This assay allows screening of compounds that, e.g., act as agonists or antagonists of a receptor or that disrupt a protein: protein interaction. It also allows detection of binding to DNA. RNA or complex carbohydrates. For example, neurokinin receptor binds to a 36

7-nitrobenz-2-oxa-l,3-diazol-4-yl(NBD) - labeled peptide ligand. The labeled ligand has the following formula: PhCO-2,4-diaminobutyric acid(gamma-NBD)-Ala-D-trp- Phe-D-pro-Pro-NH2. NBD is a fluorophore and binding of the labeled ligand to the neurokinin receptor increases NBD's fluorescence. When a compound displaces the NBD-labeled ligand from the neurokinin receptor, fluorescence of the NBD fluorophore is reduced (G. Turcatti, H. Vogel, A. Chollet (1995) Biochemistry 34, 3972-3980). A library of solid supports can be screened for compounds that bind to neurokinin receptor in a colloidal matrix using this method. Active compounds are found in zones of decreased fluorescence. As another example, a radioligand (tritium or ¹²⁵lodine-labeled) can be used to screen for compounds binding to a receptor with the lawn assay by using Scintillation Proximity Assay beads (SPA™, Amersham Corp.) or scintillant coated plates (Flashplates™, DuPont NEN Research Products). Receptor is bound to SPA™ beads or to a Flashplate™ surface, and radiolabeled ligand in a colloidal matrix is allowed to interact with the receptor. This interaction brings the radiolabel in close proximity with the scintillant and results in a scintillation signal. The signal can be detected using x-ray film, or other commercially available film that is specifically designed to detect tritium dependent scintillations. Compounds released into the matrix from the solid supports that bind to receptor and displace the radioligand reduce the scintillation signal, i.e., result in a zone of reduced scintillation. The receptor used in the assay can be, e.g., membrane- bound, tethered to a solid phase or solubilized.

When using the assay to find compounds that affect enzyme activity, it is advantageous to employ a substrate or product of the enzymatic reaction that 37

generates a detectable signal. The difference in signal between the substrate and product should be significant. It is particularly preferred to use a substrate which generates little or no signal and which converts to a product which generates a strong signal. If the substrate produces detectable signal which cannot be distinguished from that of the product, it can create background noise, thereby reducing the overall sensitivity of the assay. For this reason, non-fluorescent substrates that convert to fluorescent products, i.e., fluorogenic substrates, are preferred. One well-known fluorogenic substrate is fluorescein diacetate, which converts to fluorescein in me presence of an esterase, such as carbonic anhydrase. Other fluorogenic substrates include 7-amino-trifluoromethyl coumarin (AFC), 4-trifluoromethylumbelliferyl

(HFC), 7-amino-4-methylcoumarin (AMC) and 4-methoxy-2-naphthylamine (MNA).

Alternately, a fluorescent substrate can be used that converts to a product having different excitation and emission characteristics. By using band-pass filters so that only the product is excited and detected, the substrate can be effectively screened out. An example of such a fluorescent substrate is peptidylaminomethylcoumarin, which is converted by an appropriate protease such as thrombin, to free aminomethyl- coumarin. The free aminomethylcoumarin excites and emits at different wavelengths than does the peptidylaminomethylcoumarin (S. Kawabata et al, (1988) Eur. J. Biochem. 172, 17).

It is also possible to use a substrate containing internally quenched fluorophores that become fluorescent when converted to product. Such quenching reactions are well known (Ε. Matayushi et al.. Science 247, 954). For example, a 38

peptide substrate can be produced having two fluorophores at opposite ends, one absorbing the fluorescence of the other. The substrate therefore emits a negligible amount of light. Upon cleavage of the peptide by a suitable protease, the absorbing fluorophore is released and no longer quenches the other fluorophore, resulting in an increase in fluorescence. One such substrate is 4-(dimethylaminophenylazo)-benzoic acid (DAB CYL)-Gabu-glu-arg-met-phe-leu-ser-phe-pro-5-[(2-aminoethyl)- aminonaphthalene]-l-sulfonic acid (EDANS), which when cleaved by an aspartyl protease (e.g., plasmepsin 11 of Plasmodiumfalcioarum) becomes fluorescent. In screening a library of aspartyl protease inhibitors using the lawn assay, those that are active inhibit cleavage of the substrate, allowing quenching to be maintained. Active compounds are found in dark zones of inhibition.

Fluorescence can be detected, e.g., using a field format fluorescence detection instrument such as Fluorimager™ from Molecular Dynamics. This type of fluorimeter is capable of determining fluorescence over a large area. It is also possible to detect fluorescence using a CCD camera and to transfer the image data to a computer. The image can be generated by illumination of the fluorophore with light of the wavelength that specifically excites it. Detection can be optimized by using a bandpass filter between the camera and the assay that is specific for the emission wavelength of the fluorophore.

Assays that measure a change in fluorescence are preferred as they are believed to result in the greatest sensitivity. Any method, however, can be used that measures a change in signal from one of the compounds involved in the reaction as a 39

result of conversion of the substrate to product or displacement of the labeled ligand from the target. An example of an assay for compounds that affect a chromogenic substrate, p-nitrophenylphosphate, is described in the examples. It is also possible, for example, to measure a change in absorbance. For example, NADP is a common cofactor in many enzymatic reactions. Absorbance changes as NADPH is converted to NADP by, for example, neutrophil NADPH oxidase (such as during an oxidative burst associated with an immune response). This change can be monitored to determine zones of inhibition for compounds that inhibit this and other enzymes that use NADP, NADPH, NAD and NADH as co-factors. The sensitivity of assays that measure a change in absorbance is believed to be generally lower than those that measure a change in fluorescence.

Other examples of detectable changes resulting from conversion of substrate to product include chemiluminescent changes and scintillation changes. Scintillation changes can be detected as described above for receptor binding with the exception that a substrate is attached to the scintillant (i.e., to the bead or plate containing scintillant). For example, a radioactive reagent such as tritiated famesyl pyrophosphate, can be added to the substrate by an enzyme such as farnesyl protein transferase. Transferase inhibitors prevent addition of the tritiated farnesyl pyrophosphate to the substrate, resulting in a reduction in detectable scintillations; i.e., transferase inhibitors are found in zones of reduced scintillation. In an alternative assay, removal of the radioactive portion of a substrate attached to the scintillant such as by cleaving with a protease, releases the radiolabeled portion (i.e., moves it away from the scintillant). In such an assay, protease inhibitors cause an increase in 40

scintillation, i.e., are found in zones of increased scintillation. As noted above, the scintillation signal can be detected using x-ray film or film that is specifically designed to detect tritium dependent scintillations.

For assaying binding to a target molecule, a labeled ligand provides a signal that indicates such binding. The label is preferably a fluorescent moiety that alters its signal as a result of target molecule binding. Examples of such fluorescent moieties are NBD and 5-(dimethylamino)-l-naphthalenesulfonyl (Dansyl) chloride.

Colloidal matrices that are useful for the lawn assay include silica gel, agar, agarose, pectin, polyacrylamide, gelatin, starch, gellan gum, cross-linked dextrans (such as Sephadex™ , available from Supelco, Bellefonte PA) and any other matrix that allows diffusion of compound from the solid supports in a limited region. Low melting-temperature agarose is preferred, generally in an amount of 0.5-2.0%, wt./vol. The colloidal matrix can be chosen to obtain a desired rate of diffusion. It is generally preferred to use a matrix that allows a high concentration of compounds to be easily obtained.

In carrying out the assay to determine compounds that affect enzyme activity, the solid supports are preferably embedded in a matrix containing the relevant enzyme. Following cleavage, compound diffuses from the support into the matrix and contacts the enzyme. Substrate is then added and, as it diffuses into the colloidal matrix, active compounds inhibit conversion to product. By following such a procedure, compounds to be screened are allowed to interact with enzyme before the 41

enzyme contacts substrate. This is believed to be advantageous because it allows compounds the best opportunity to inhibit the enzyme and thus results in the clearest zone of inhibition.

It is also possible, however, to embed the solid supports in a matrix that contains dispersed substrate. Following cleavage, the matrix can be contacted with enzyme. This procedure is not believed to be as sensitive since the compounds may not efficiently bind to the enzyme.

Solid supports can also be applied to the matrix's surface and the compounds allowed to diffuse into the matrix. This can be done, for example, by arraying the solid supports on the surface of a stretched sheet of plastic film (e.g., Parafilm™, available from Aldrich Co., Milwaukee, WI), and then applying the sheet to the surface of the matrix.

In assaying for compounds that affect enzyme activity, it may be desirable to use two colloidal matrices. For example, one matrix can contain enzyme and beads and the other can contain substrate. Contacting the surfaces of the matrices with each other allows the substrate to come into contact with the enzyme. It is also possible to add a solution of substrate over the surface of a matrix containing enzyme and embedded supports. Adding solution is prefeπed when, e.g., the substrate interferes with detection. Solution containing the substrate can be removed prior to determining the zones of activity. 42

When using the lawn assay to screen for binding to a target molecule, there is generally no need for more than one matrix. A matrix contains the target molecule bound to the labeled ligand which emits a detectable signal indicating binding to the target molecule. Compounds from the solid supports are diffused into d e matrix, preferably from embedded supports using photolysis. Alternatively, however, labeled ligand can be diffused into the matrix from a second matrix (or liquid layer) after release of the compounds in the matrix. This allows the compounds to contact the receptor before interaction with the labeled ligand, which can be advantageous.

Compounds can be cleaved from the solid supports either before or after the supports are contacted with the colloidal matrix. For example, solid supports may contain acid cleavable linkers, as further described below. These linkers can be cleaved in a gaseous acidic atmosphere before placing the supports on the matrix. The compounds, although cleaved, remain on the surface of the supports and diffuse into the matrix when the supports are placed on it. It is even possible to cleave the compounds prior to pouring low-melt liquid agarose over the solid supports. While some of the compounds will be washed away, sufficient compound can remain on the support's surface to result in a recognizable zone of activity.

Where the compounds are cleaved after the beads are embedded in the colloidal matrix, it is preferred to use photolysis, e.g., cleaving by exposure to UV light. By adjusting light exposure, it is possible to control the amount of compound that diffuses into the matrix. If more light is applied, by increasing intensity or duration, more cleavage results, in turn releasing more compound into the matrix. 43

This allows the amount of active compound released to be adjusted, so mat zones of activity are only produced for compounds that are most active. The amount of compound released can also be optimized to produce zones that are most distinct.

The solid supports can be in a random arrangement, or in an ordered one.

Preparing a random arrangement of solid supports requires little effort. For example, a library of beads can be suspended in a solvent, such as ethanol, and deposited on the bottom of a Petri plate. After the solvent has completely evaporated, a layer of agarose containing the relevant enzyme or target molecule can be poured over the beads. Alternatively, an ordered aπay can be used to space beads apart and allow easier identification of those that are active. In one example of an ordered aπay, beads are aπayed on a rigid template such as a thin glass disk having tapered holes. The tapered holes are sized to allow only single beads to settle into them. Beads are suspended in a solvent such as ethanol, and washed over the top of the template to fill each hole with one bead. The beads can then be cleaved in the dry state and the template set down on the colloidal matrix. Capillary action wets the beads, facilitating diffusion of the cleaved compounds into the matrix. Zones of activity can be observed immediately below beads containing active compounds. It is possible to remove the template prior to detecting zones of activity if an image of the template on the matrix is made. This image can later be used to coπelate the zones of inhibition in the matrix with the positions of beads on the template.

Ordered aπays also may be useful in identifying the compounds on supports that are associated with zones of activity. Specifically, the aπay can be ordered so that 44

the position of the solid support on the aπay coπesponds to the identity of the compound. Thus, once an assay has been carried out and the position on the aπay determined for a support carrying an active compound, the identity of that compound can be easily determined.

Preferably, however, the identity of active compounds is determined using d e encoding system described above, which employs tags T' encoding the identities of the compounds on the solid supports.

The assay is preferably carried out so that there is slow diffusion of d e compound from the solid support following cleavage. This results in a high concentration of compound in the vicinity of me bead. Thus, very little compound is required to cause a distinct zone of activity. Most of the compound remains on the support for any subsequent assays that are required. Such further assays may be needed if more than one solid support is found in the zone of activity. It may then be necessary to retest the supports from the zone to determine which one releases the active compound. Reassaying may be required as a matter of course if many thousands of beads are screened at high density. Reassaying may also be desirable to test for selectivity, i.e. to determine which active compounds are inactive in a second assay that tests for a different property.

With combinatorial libraries containing d ousands of related compounds, many compounds may be found that have some degree of activity. It therefore may be useful to use the lawn assay to distinguish the most potent compounds. In the assay, if the amount of compound released from each support is approximately the same, potent compounds have a detectable effect further from the bead than weak compounds do at any given time. Thus, the more active compounds create a larger zone of activity. Furthermore, the zone of activity of the most active compounds lasts longer. Thus, it is possible to quantify the activity of the compound eluted from the solid support by the size of the zone of activity as well as by the duration of the zone following cleavage.

Reducing photolysis time reduces the amount of compound released from the support. As the concentration of the compounds is lowered, diose that are less active become more difficult to detect. As a result, the number of active compounds drops. In experiments described in the Examples that follow, compounds that were detectable at the shortest elution times, i.e., that were most potent, were also identified as most potent using conventional solution-phase screening. The activity of die inhibitors was found to coπelate with the size and duration of the zone of activity: the most potent compounds produced the largest zones for the longest time for any given amount of photolysis.

When assaying a library containing many active compounds, it may be desirable to screen using a low density of solid supports, i.e., a low number of supports per cm³ of matrix. While requiring more assays to screen the entire library, it is less likely that supports will have to be retested to determine which contains the active compound. Screening a large library containing many active members at a low density is often more efficient than screening at high density since rescreening supports is time consuming. The optimum density for screening can be determined for a given library by comparing the throughput in the initial assay with the effort required to retest active supports. Other factors which affect optimum screening density include the cost of the target and the size of the library.

When several large libraries are available for testing, it may be advantageous to incompletely evaluate each library by "scouting" each at high density for active compounds. Screening at high density allows one to statistically evaluate the number and potency of active compounds in each library. Libraries which contain the most active compounds can be more thoroughly tested.

If the proportion of active compounds screened in the assay is high, a second assay of the active compounds may be performed to choose those mat should be further evaluated. The second assay can determine whether there is cross reactivity with other targets, i.e., a "selectivity screening". For example, a given library of compounds can be screened for activity against HTV protease, a member of the aspartyl protease family, using DABCYL-gAbu-Ser-Gln-Asn-Tyr-Pro-lle-Val-Gln- EDANS. Compounds found active in the initial assay can be counterscreened against a second, different aspartyl protease, such as cathepsin D. Alternately, all compounds screened in the assay for activity against H1N protease could be simultaneously screened in the counter assay.

It is also possible to test for compounds mat interfere with proteins that inhibit enzyme activity. In such an assay, the most active compounds prevent enzyme inhibition, resulting in more enzymatic catalysis. Thus, when a fluorogenic substrate is used, active compounds result in a brighter zone of activity. For example, PI 6 is a known protein inhibitor of cyclin-dependent kinase-4 (Cdk-4). Using the lawn assay, Cdk-4, Cyclin Dl, P16, a fluorogenic substrate and a library of beads to be screened can be included in a layer of low-melt agarose. Following photocleavage and after allowing sufficient time to convert substrate to product, die gel can be subjected to an electrophoretic separation. Product migrates to the anode, where it is preferably trapped on an anode filter. The location of product on die filter indicates die position in the gel of compound that disrupts P16 inhibition of Cdk4.

In another embodiment of the lawn assay, an electrophoretic procedure is used to separate substrate from product to increase the sensitivity of the assay. In this embodiment, a substrate is used which changes charge when converted to product. An example of such a substrate is the peptide Leu-Arg-Arg-Ala-Ser-Leu-Gly attached to a fluorophore, sold commercially as Pep-Tag™ (Promega Corp.). Protein kinase A (PKA) phosphorylates this substrate, which has net +1 charge, to form a phosphopeptide which has a net - 1 charge. A lawn assay is performed in which PKA is contacted in a colloidal matrix with substrate and a library of potential inhibitors. An electrophoretic separation is then carried out across the width of (i.e., perpendicular to) the matrix. The phosphopeptide (i.e., product) moves towards the anode and the dephosphopeptide (i.e., substrate) moves towards the cathode. If a membrane is applied to one or both sides of the matrix during electrotransfer, electroblotting can be achieved. For example, the phosphopeptide can be electroblotted to a suitable membrane such as an Immobilon™ CD membrane. Alternately, the dephosphopeptide can be electrotransfeπed to an appropriate paper such as Whatman™ 3MM paper. In another embodiment, the substrate and product can be chosen so that one is neutral and one is charged. Application of die electrophoretic field will remove the charged moiety. The resulting matrix will contain only the neutral moiety, thereby allowing detection of compounds that affect the conversion to product. The position of die bead containing die active compound can be determined by fluorescent imaging of d e substrate or product using, e.g., photography or video imaging. This technique increases sensitivity of the lawn assay by separating fluorescent substrate from fluorescent product, concentrating the fluorescent image and by eliminating compounds from the matrix that might cause background signal. Other protein kinases and phosphatases such as protein kinase C, cyclin dependent kinases, MAP kinases and inositol monophosphatase can also be used with appropriate substrates in this method. A protease can also be screened by this method by using a substrate consisting of an appropriate peptide linked to a labeling moiety such as a fluorophore. The peptide sequence is selected so mat the substrate and product will migrate differentially in an electric field.

Enzymes that can be used in the assay include, but are not limited to, the following: Acid Phosphatase

Activated Protein C Alkaline Phosphatase Aminopeptidases B & M Amyloid A4-Generating Enzyme 49

Angiotensinase Aryl Sulfatase β-Galactosidase

β-Glucosidase

5 β-Glucuronidase

Calpains I & H Cathepsins B, C, D & G Cholinesterase Chymotrypsin 10 Collagenase

Dipeptidyl Peptidases I-1N Elastase

Endodielin Converting Enzyme Factor Xa 15 Factor Xla

Factor XI la D/-Protease

Furin γ-Glutamyltranspeptidase

Granzymes A & B 20 H1N Protease

IL- IB Convertase

Kallikrein

Lysozyme 50

Mast Cell Protease Peroxidase Plasmin

Prohormone Convertase r ANP Precursor Processing Enzyme

Renin

Spleen Fibrinolytic Proteinase Staphylocoagulase Thrombin Tissue Plasminogen Activator

Trypsin Tryptase Urokinase The assay procedure is further illustrated by the Examples below.

Examples of the Use of the Assay

The lawn assay is performed in Petri plates using two layers of agarose, each about 1.5 mm thick. The first layer contains TentaGel S-NH™ beads (Rapp Polymere, Tubingen, Germany) and enzyme. The TentaGel S-NH>™ beads have compounds to be screened attached thereto by a photocleavable linker and chemical tags attached for identifying the compounds, prepared according to me ods described herein. The beads are either placed on the Petri plate and agarose poured over them or beads and agarose are first mixed and then poured together onto the piate. A second layer of agarose containing the fluorescein diacetate is contacted with the first layer to initiate the reaction.

More specifically: 50 mM sodium phosphate, pH 7.4, is used as a buffer and all solutions equilibrated in a 37°C water bath immediately prior to initiation of the

assay. 0.1 mL of 5.3 μM bovine carbonic anhydrase (Sigma Chemical Co.) is diluted in 2.15 mL of buffer and 1.25 mL of 2.5% low-gelling agarose added (SeaPlaque™, FMC BioProducts). Library beads suspended in methanol are added to a 6 cm polystyrene Petri plate and, if necessary, distributed with a flat pipette tip. After evaporation of the methanol, the agarose solution is poured over the beads and allowed to gel at room temperature for 2-3 minutes. (Alternatively, dry beads can be added to a mixing tube and then enzyme and agarose added; the mixture is then vortexed and poured.) The plate is then placed under a long wave (360 nm) UV lamp (Blackray™ UVP, Inc.) for from 5 seconds to 1 hour. After iπadiation, 0.01 mL of fluorescein diacetate (10 mM in DMF, Molecular Probes, Eugene. Oregon) is combined with 2.25 mL buffer and 1.25 mL of 2.5% agarose and poured over the first agarose layer. Detection is achieved by illumination using a short wavelength UV lamp (UVX, 254 nm) and image capture using a CCD camera coupled to a computer with NIH Image software obtained from the National Institutes of Health.

Fluorescein diacetate is hydrolyzed to produce fluorescein as me reaction proceeds. The plate then becomes significantly brighter except in the vicinity of beads that release inhibitors, thereby forming zones of inhibition. Beads at the center of these zones are removed with a hollow glass tube or a spatula and washed in meu anol/methylene chloride (1: 1) or with hot water (80°C) to remove most of the agarose. After a final rinse in methanol, beads are either retested in a separate assay using the methods described above to confirm activity or analyzed to determine the relevant compound structures by tag decoding.

Example 1: Assay of Two Known Inhibitors

In this example, two compounds were tested for inhibition of carbonic anhydrase by the lawn assay. Carbonic anhydrase inhibitors are useful in treating, e.g., glaucoma. Results were compared with ose obtained using a conventional solution phase assay.

There is a high coπelation shown between compounds that inhibit binding of dansylamide to carbonic anhydrase and those that inhibit conversion of fluorescein diacetate to fluorescein by carbonic anhydrase. This coπelation is believed to result from dansylamide and fluorescein diacetate occupying the same active site (a zinc atom) on carbonic anhydrase. The solution phase assay measured inhibition of dansylamide binding. The lawn assay measured inhibition of me conversion of fluorescein diacetate to fluorescein.

Two aryl sulfonamide-containing compounds (compounds "A" and "B") were synthesized on TentaGel™ beads (Rapp Polymere) and assayed in the standard solution-phase assay and in the lawn assay. Compounds containing aryl sulfonamide substituents are known to be potent inhibitors of carbonic anhydrase. In the solution phase assay, Ki's were determined to be 4 and 660 nM for compounds A and B respectively. In the lawn assay, beads containing each compound were embedded in agarose in a series of Petri plates. The right side of each plate contained beads with compound A and the left side contained beads with compound B. Separate plates were iπadiated for 2.5, 5, 10, 20 and 30 minutes. The more potent inhibitor of carbonic anhydrase (compound A) showed a clear zone of inhibition after only 2.5 minutes of photolysis. The weaker inhibitor (compound B) caused only a weak zone of inhibition after five minutes of photolysis. Ten minutes of photolysis was required to obtain a distinct zone. The clearest zones of inhibition were observed at the shortest time after photolysis. Zones at five minutes after photolysis were all sharper man at 15 minutes after photolysis. At 30 minutes after photolysis, all zones were much less distinct; some zones (for compound B) had disappeared.

In a second experiment, a plate containing beads with compounds A and B was iπadiated for a predetermined period of time. The size and duration of the resulting zones of inhibition were determined. The zones resulting from compound A were larger than those resulting from compound B. Furthermore, the zones for compound A could be observed for a longer time. The signal from compound A persisted for more than two hours (although the zones became very diffuse) while signal for compound B all but disappeared after 90 minutes. In addition, zones of inhibition for compound A were more distinct, i.e., there was greater contrast between the zones and the suπounding areas. 54

Example 2: Lawn Assay for Inhibitors of Inositol Monophosphate

An assay for inhibitors of inositol monophosphate is carried out in the same manner as described above for carbonic anhydrase inhibitors wim die following substitutions: the buffer used is 20mM Tris, lmM EGTA, pH 7.8. The agarose layer contains 1 mg/mL of recombinant human inositol monophosphate, purified from E. coli, and lOmM MgCl₂. The substrate is methylumbelliferyl phosphate (Sigma Chemical Company, St. Louis MO, M-8883), CSPD (Tropix, Bedford MA) or CDP- Star (Tropix). CSPD and CDP-Star^® are chemiluminescent substrates. The

prefeπed substrate is CSPD . Inositol monophosphate is believed to be the molecular

target of lithium therapy in bipolar disease.

Example 3: Lawn Assay for Compounds that Affect Tyrosine Phosphatase

This test chromogenically assays compounds for their effect on the catalytic domain of human SHPTP1, a protein tyrosine phosphatase [Pei et al., (1993) PNAS 90, 1092] using p-nitrophenylphosphate as a substrate. This enzyme is assayed as described above for carbonic anhydrase with the following substitutions. The buffer used is lOOmM N, N-bis(2hydroxyethyl)glycine, pH 8. The first (lower) agarose layer contains 0.5 mg/mL recombinant human SHPTP1 catalytic domain, purified fromE. Coli, and die substrate is 4-nitrophenylphosphate (Sigma Chemical Corp.). Enzyme activity coπesponds with the release of the 4-nitrophenolate anion (λm 400nm,

18,300 M^"1 cm^'1), which appears as a yellow color on a clear background. Areas where affectors of the SHPTP1 catalytic domain are found are distinguished by either clear zones of inhibition or more colored zones of stimulation. When a compound is detached prior to evaluation, the relationship of the compound to die solid support is maintained, for example, by location widiin the grid of a standard multi-well plate or by location of activity on lawn cells. Whether the compounds are tested in the biological assays attached or detached from the solid supports, the identifiers attached to the solid support associated with the bioactivity may be decoded to reveal the structural or synthetic history of the active compound (Still et al., Complex Combinatorial Libraries Encoded With Tags, WO 94/08501) or the structures may be determined by deconvolution. The effectiveness of such a library as a screening tool is demonstrated by the example of screening encoded combinatorial libraries for carbonic anhydrase inhibition (Burbaum et al.. Proc. Nad. Acad. Sci. 92, 6027-6031 (1995)).

Assays for evaluating the library of the present invention are well known in the art. Although the specific assays in which a particular library compound or group of library compounds will demonstrate activity are not known a priori, useful screening systems for assaying libraries of the format described herein have been developed.

Biological assays for a wide variety of enzymes and molecular targets can identify activity among the entities of a combinatorial library.

V. Methods of Synthesis

The compounds of the present invention can be prepared according to die following methods. During each step in the synthesis, the solid support upon which a compound is being synthesized is uniquels tagged to define the particular chemical event(s) occurring during that step. The tagging is accomplished using identifiers such as those of Formula H which record the sequential events to which the support is exposed during the synthesis. Tagging thus provides a reaction history for the compound produced on each support. The identifiers are used in combination with one another to form a binary or higher order encoding scheme permitting a relatively small number of identifiers to encode a relatively large number of reaction products. For example, when N identifiers are used in a binary code, up to 2^N - 1 different compounds and/or conditions can be encoded. By associating each variable or combination of variables at each step of the synthesis with a combination of identifiers which uniquely defines the chosen variables such as reactant, reagent, reaction conditions or combinations of these, one can use identifiers to define the reaction history of each solid support.

In carrying out the syntheses, one begins with at least 10 , desirably at least 10⁷, and generally not exceeding 10¹⁰ solid supports. Depending on d e predetermined number of R¹ choices for the first combinatorial step, one divides the supports accordingly into as many containers. The appropriate reagents and reaction conditions are applied to each container and the combination of identifiers which encode for each R¹ choice is applied. Depending on the chemistry involved, the tagging may be done prior to, concomitantly with or following the reactions which comprise each choice. As a control, a sample of the tagged support can be taken at any stage during the reaction sequence and analyzed to obtain information about the synthesized compound. As needed, one may wash the beads free of any excess reagents or by-products before proceeding. At the end of each step, the supports are combined, mixed and again divided, this time into as many containers as predetermined for the number of R² choices for die second step in the synthesis. This procedure of dividing, reacting, tagging and remixing is repeated until the combinatorial synthesis is completed.

A. Scheme 1: Attachment of acyl Meldrum's acid to resin.

A batch of hydroxyl-functionalized PEG-grafted polystyrene beads such as TentaGel S-PHB ™ 1 is coupled to acyl Meldrum's acid 2 through alcoholysis. (The acyl Meldrum's acid is prepared by the condensation of 4-bromo-2-fluorobenzoyl chloride with Meldrum's acid.) Coupling is achieved by reacting a suspension of 1 in toluene with 2. The suspension is shaken with heating, drained and washed in succession with toluene and DCM. The derivatized resin 3 so obtained is men dried overnight under vacuum.

B. Scheme 2: Attachment of N.N-dimethylformamide dimethyl acetal to β-keto ester on resin

N,N-dimethylformamide dimethyl acetal is attached to resin derivative 3 to form enamine 4. This is accomplished by adding N,N-dimethylformamide dimethyl acetal to a suspension of solid sample 3 in ethyl ether. The mixture is shaken overnight, washed with ether and DCM and then dried overnight under vacuum to give enamine 4 on resin. After the reaction, a small portion of enamine 4 may be cleaved off the resin with TFA and quantitated by HPLC. JO

C. Scheme 3; Amine fl^NH-) displacement on resin

The N,N-dimethyl amino moiety in resin derivative 4 is displaced with various primary amines (e.g., see Table 1) to generate compounds. The displacement occurs through a substitution with other amines (i.e., R'NH∑) that results in the formation of new enamines. The amine (R'NHa) is added to a suspension of resin 4 in THF and d e mixture is shaken overnight. The mixture is drained, washed with THF as well as DCM and dried overnight under vacuum to yield resin 5. The procedure is repeated for each of the twenty-four amines.

D. Scheme 4: Coupling of stannanes (Bu₃SnR² to quinolone derivatives on resin

The enamine 5 on resin is cyclized to produce a quinolone derivative 6. The cyclization of each enamine involves the displacement of the aromatic fluoride and formation of a new carbon-nitrogen bond intramolecularly. A THF solution of lithium bis(trimethylsilyl)amide is added to a suspension of resin 5 in THF. The mixture is shaken overnight, washed with THF and DCM and then dried under vacuum to give quinolone derivative 6. This is repeated for each of the twenty-four enamines. A small portion of the product on resin may be cleaved off using TFA and quantitated by HPLC. 59

Identifiers are then added to each quinolone derivative. Unique tagging of the quinolone bound supports is achieved by using combinations of identifiers encoded in a binary scheme (the identifiers are not shown in the schematics for die purpose of simplicity) for the all choices of R¹. The identifiers are attached by adding a solution of up to three identifiers in DCM (10% wt. ratio of each identifier to the solid support) to a batch of the supports suspended in DCM and shaking the mixture for an hour. A dilute solution of rhodium trifluoroacetate dimer in DCM is added and the mixture is then shaken overnight, washed in DCM and dried under vacuum. The coding for each quinolone derivative is confirmed by GC analysis. A second cycle of tagging may be applied when it is necessary.

The twenty-four batches of tagged quinolone resin 6 are pooled, mixed by shaking in DCM and then divided into a pre-determined number of reaction vessels. At mis stage, since 9 stannane reagents are used in the subsequent combinatorial step, equal portions of resin are placed in 9 vessels.

The mixtures of tagged quinolone derivatives 6 are coupled with a stannane reagent, coπesponding to one of the 9 R² choices shown in Table 2, by carbon-carbon bond formation. This is accomplished by adding TPP and Pd₂(dba)₃, followed by die addition of each stannane, to the respective suspensions of resin 6 in DEE. The mixture is shaken overnight with heating, drained and washed with DEE, DCM, EtOH and again with DCM. Compound 7 is obtained after drying overnight under vacuum. This is repeated for all 9 vessels. After coupling, a small portion of each batch of resin may be cleaved off and quantitated by HPLC. 60

E. Scheme 5: Attachment of amines (R³NH^ to quinolone derivatives on resin

Identifiers are introduced to resin 7. Unique tagging of the supports in each of the 9 reaction vessels is achieved with combinations of identifiers encoded in a binary scheme for all 9 choices of R². The identifiers are attached by adding a solution of up to three identifiers in DCM (10% wt. ratio of each identifier to the solid support) to a batch of the supports suspended in DCM and shaking the mixture for an hour. A dilute solution of rhodium trifluoroacetate dimer in DCM is added and the mixture is then shaken overnight, washed in DCM and dried under vacuum. The coding for each aldehyde derivative is confirmed by GC analysis.

The 3 batches of tagged aldehyde resin 7 (only some R² reagents generate aldehyde products, i.e., stannanes #7, 8 and 9 in Table 2) are pooled, mixed by shaking in DCM and divided into a pre-determined number of reaction vessels. In this case, since 10 amine reagents are used in this third combinatorial step, equal portions of resin are placed in the 10 vessels.

The mixtures of tagged quinolone aldehyde 7 are coupled with an amine coπesponding to one of the 10 R³ choices shown in Table 3 by a carbon-nitrogen bond formation. The aldehydes 7 are treated with the solution of an amine in TMOF which is in the presence of sodium cyanoborohydride. The mixture is shaken and drained. The solid product is then washed with 1% HCl/MeOH, MeOH. 10% Hunig's base/DCM, MeOH, DMF, and DCM. The washed product is finally dried under o l

vacuum. This procedure is repeated for each of the 10 amines to complete the synthesis of the library resin product 8.

VI. PREPARATION AND USE OF IDENTIFIERS

Ten compounds of die general formula II:

wherein n is 3-12 and Ar is pentachlorophenyl were prepared in accordance with Scheme 6 and die following illustrative example as disclosed in U.S. patent application 08/743,960, filed 11/5/96. The protocol for tagging resin in the present invention was taken from the same reference.

Step 1- l-hydroxy-9(2, 3, 4, 5, 6-pentachlorophenoxy)nonane (1.634 g,

4.0 mmol), methyl vanillage (0.729 g, 4.0 mmol) and triphenylphosphine (1.258 g, 4.8 mmol) were dissolved in 20 mL dry toluene under argon. DEAD (0.76 mL, 0.836 g, 4.8 mmol) was added dropwise and die mixture was stiπed at 25°C for one hour. The solution was concentrated to half volume and purified by flash chromatography, eluting with DCM to yield 1.0 g (1.7 mmol, 43%) of the methyl ester as a white crystalline solid. 62

Step 2- The methyl ester product of step 1 (1.0 g, 1.7 mmol) was dissolved in 50 mL THF, followed by addition of 2 mL water and LiOH (1.2 g, 50 mmol). The mixture was stiπed at 25°C for one hour then refluxed for five hours. After cooling to 25°C, the mixture was poured onto ethyl acetate (200mL) and washed widi 1 M HCl (3 x 50 mL), then saturated aqueous NaCl (1 x 50 mL) and dried over sodium sulfate. The solvent was removed and the crude acid azeotroped once with toluene.

Step 3- The crude material from step 2 was dissolved in 100 mL toluene. Thionyl chloride (10 mL, 1.63 g, 14 mmol) was added and die resulting mixture was refluxed for 90 minutes. The volume of the solution was reduced to approximately 30 mL by distillation, then the remaining toluene was removed by evaporation. The crude acid chloride was dissolved in 20 mL dry DCM and cooled to -70°C under argon. A solution of approximately 10 mmol diazomethane in 50 mL anhydrous ether was added. The mixture was warmed to room temperature and stiπed for 90 minutes. Argon was bubbled through the solution for ten minutes, then the solvents were removed by evaporation. The crude material was purified by flash chromatography, eluting with 10-20% ethyl acetate in hexane. The diazoketone (0.85 g, 1.4 mmol, 82% yield over three steps) was obtained as a pale yellow solid.

An improvement was made to the final diazomethylation step whereby the acid chloride was reacted with (trimethylsilyl)-diazomethane and triethylamine to give the identifier which was then used without further purification. This was a significant improvement over the original reaction with diazomethane as the identifier was now obtained in high yield with no chloromethylketone by-product. Also, purification by 63

flash chromatography was unnecessary which, in some cases, had resulted in significant acid-catalyzed decomposition of the identifier.

ALTERNATE Step 3- A 2.0 M solution of (trimethylsilyl)-diazomemane (5.7 mL, 11.4 mmol, 3.00 eq.) in hexanes was added to a solution of die acyl chloride (3.8 mmol, 1.00 eq.) and triethylamine (1.85 mL, 13.3 mmol, 3.5 eq.) in anhydrous THF/acetonitrile (1:1) at 0°C under argon. The resulting orange solution was stiπed at 0°C for two hours, then at 25°C for 17 hours. (If a precipitate immediately formed upon addition of the (trimethylsilyl)-diazomedιane, DCM was added until die precipitate dissolved). EtOAc (250 mL) was added and die organic layer was washed with 100 mL each of saturated aqueous NaHCO.. and water, then dried with anhydrous MgS0₄. Removal of the volatile materials in vacuo produced a yellow crystalline product at 60-100% yield.

In synthesis Example 1, nine identifiers were used to encode the combinatorial library. In Step 1, pentachlorophenyl identifiers where n =7-11 (abbreviated C₇Cl₅,C₈Cl₅, C₉CI₅, C₁₀CI₅ and C₁₁CI₅) were used in the following binary encoding scheme: 00001 = (n = 11), 00010 = (n = 10), 00100 = (n = 9), 01000 = (n = 8) and 10000 = (n = 7). In Step 2, pentachlorophenyl identifiers where n = 3, 4, 6 and 12 (abbreviated C₃C1₅, C4CI₅, C₆C1₅ and C₁₂CI₅) were used and encoded as follows: 0001 = (n = 12), 0010 = (n = 6), 0100 = (n = 4), and 1000 = (n = 3).

Thus, in Step 1, for example, 2-methyl butylamine (Table 1. entry 3) is encoded "00011", which represents tagging this choice in the synthesis with the two 64

pentachlorophenyl identifiers where n = 10 and 11. Likewise, in Step 2, thienyl (Table 2, entry 6) is encoded "0110" which represents tagging this choice in die synthesis with the pentachlorophenyl identifiers where n = 4, 6.

EXAMPLE 1

800 Plus COMPOUND LIBRARY

Step (l). Attachment of acyl Meldrum's acid to resin.

(a). Preparation of 4-bromo-2-fluorobenzoyl Meldrum's acid 2. DMAP (5.69 g, 46.57 mmol) was added to a solution of Meldrum's acid (3.22 g, 23.1 mmol) in

CH₂C1₂ (50 mL) in an ice-salt bath. A solution of 4-bromo-2-fluorobenzoyl chloride (6.82 g, 28.72 mmol) in CH₂C1₂(80 mL), prepared from a quantitative reaction of 4-bromo-2-fluorobenzoic acid witii oxalyl chloride in CH₂C1₂, was added dropwise over a period of 55 minutes to the above solution with stirring. The temperature was maintained for an additional hour before the reaction was allowed to warm to room temperature. The mixture was stiπed at room temperature for 2.5 hours. The resulting mixture was washed with 2N HCl (2 x 30 mL) and water (30 mL). The organic layer was then separated, dried over MgS0 , filtered and concentrated to give crude 2. NMR analysis indicated a yield of over 90%.

(b). Attachment of acyl Meldrum's acid to resin. To a suspension of TentaGel ™ resin 1 (S-PHB, 9 g, 0.28 mmol/g, 2.52 mmol, 180 mm) in toluene (130 mL) was added a toluene solution (20 mL) of the crude 2 (7.1 g. > 20 mmol) as prepared in (a). The mixture was heated to 100 °C and shaken until bubbling subsided (-1.5 hours). 65

After cooling to room temperature, the mixture was drained, washed with toluene and DCM and dried overnight under vacuum to yield resin 3.

Step (2). Attachment of N.N-dimethylformamide dimethyl acetal. To a suspension of resin 3 (9 g, 2.52 mmol) in anhydrous ethyl ether (100 mL) was added N,N-dimedιylformamide dimethyl acetal (9 g, 75 mmol). The mixture was shaken under Argon at room temperature for 16 hours. The resin was then drained, washed witii ether (2 x 50 mL) and DCM (2 x 50 mL) and dried overnight under vacuum to give resin 4. A small portion of the resin (20 mg) was suspended in TFA (5 mL) and shaken for one hour to cleave the compound off the solid support. The compound solution was collected, concentrated and analyzed by HPLC. A yield of 95% was obtained.

Step (3). Attachment of primary amines NH;R*. A portion of resin 4 was placed into each of twenty-four reaction vessels.

Each of the twenty-four batches of the resin was reacted with one of the amines shown in Table 2. For example, to a suspension of resin 4 (4 g, 1.12 mmol) in anhydrous THF (100 mL) was added n-butylamine (1.1 mL, 11.2 mmol). The mixture was shaken under Argon at room temperature for 16 hours. The resin was drained, washed with THF (2 x 40 mL) and DCM (2 x 40 mL) and dried overnight under vacuum to afford resin 5. Since the subsequent cyclized product 6 was also generated under this condition, the reaction was not quantitated. 66

Step (4). Intramolecular cvclization. encoding and coupling with stannanes

Bu₃SnR². (a). Intramolecular cvclization of resin 5. Each of the twenty-four batches of resin 5 was cyclized with lithium bis(trimethylsilyl)amide. For example, resin 5 (300 mg, 0.084 mmol) which was obtained from (-)-cis-myrtanylamine (amine 6 in Table 1), was suspended in anhydrous THF (8 mL). A 1.0 M solution of LifTMS^N in THF (0.25 mL, 0.25 mmol) was added and the mixture was shaken for 18 hours under Argon. The sample was then drained, washed with THF (2 x 5 mL) and DCM (2 x 5 mL) and dried overnight under vacuum to give the cyclized resin 6. A small portion (-20 mg) of this resin was suspended in TFA (5 mL) and shaken for one hour to cleave off the compound, 82% yield. The compound was analyzed by HPLC.

(b) Encoding of resin 6. Each batch of resin to be encoded was tagged by an amount of each tag weighing 10% by mass of resin. Each of the twenty-four batches of resin 6 was encoded with one or more of the C₁1Q5-, C10CI₅-, C₉CI₅-, CgCl₅ - and G₇CI₅ - linker-diazoketones to produce die appropriate binary code. For example, to 200 mg of resin batch 20 (coπesponding to 2-tiιienylmetiιylamine in Table 1, entry 20) suspended in DCM (4 mL) were added CuCl₅-, C₉CI₅- and C₇C1₅ - linker-diazoketones (20 mg each). Each tag reagent was dissolved in DCM (0.5 mL) and then added to the suspension. After agitating for one hour, rhodium trifluoroacetate dimer (0.2 mL of a 5 mg/mL solution in methylene chloride) was added, and the mixture was agitated at room temperature for an additional eight hours. The resin was then drained and washed with 5 mL portions of DCM (3x), MeOH (3x) and DCM (5x). The washed resin was dried overnight under vacuum. The tagging efficiency was determined by GC analysis (see Example 3).

(c) Coupling of stannanes B SnR². The twenty-four batches of tagged resin 6 were pooled, mixed to homogeneity and divided into a pre-determined number of reaction vessels according to the number of R". The twenty-four batches were transfeπed to a reaction vessel (v = 100 mL) and suspended in DCM (50 mL). The mixture was shaken for an hour, drained and dried under vacuum for 48 hours. Six 110 mg portions and tiiree 1.1 g portions of resin 6 were placed into six small and three medium reaction vessels, respectively. Each of the smaller portions was coupled separately, with each of the first six stannanes listed on Table 2 while each of die larger portions was coupled separately, with each of the bifunctional stannanes in Table 2 (reagent #'s: 7, 8 and 9). A larger amount of resin 6 was used for die bifunctional stannanes since an additional synthetic step was performed after the Stille coupling. This would ensure that the final compound amounts in terms of the weight of resin for all the sub-libraries are the same. For example, a suspension of resin 6 (288.5 mg, 0.081 mmol) in DEE (5 mL) was degassed by Argon (bubbled for 15 minutes). Pd₂(dba)₃ (7.2 mg, 0.008 mmol) and TPP (7.2 mg, 0.027 mmol) were then added to the degassed suspension. The mixture was shaken for five minutes at room temperature before 2-(tributylstannyl)thiophene (0.08 mL, 0.252 mmol, stannane 6 in Table 2) was added. The reaction mixture was then heated to 90 °C and shaken for 16 hours. The cooled mixture was drained, washed with DEE (6 mL), DCM (2 x 6 mL), EtOH 68

(6 mL) and DCM (2 x 6 mL) and dried overnight under vacuum to give thiophenyl quinolone resin 7. A small portion (-20 mg) of this resin was cleaved in TFA (5 mL, one hour shaking). The off-resin compound was analyzed by HPLC and a yield of 93% was determined.

Step (5) Encoding and attachment of amines NH>R³.

(a) Encoding of resin 7. Each of the 9 batches of resin 7 was encoded with one or more of the Cι₂Cls-, CβCls-, C Cls- and C₃Cl₅- linker-diazoketones to produce die appropriate binary code. Identifiers were incorporated at a 10% wt. ratio of each identifier to the solid support. For example, to a 100 mg portion of resin batch 6 (coπesponding to stannane 6 in Table 2) suspended in DCM (4 mL) were added CβCls- and C C1₅- linker-diazoketones (10 mg each). Each tag reagent was dissolved in DCM (0.5 mL) and then added to the suspension. After agitating for one hour, rhodium trifluoroacetate dimer (0.2 mL of a 5 mg/mL solution in methylene chloride) was added and the mixture was agitated at room temperature for an additional 8 hours. The resin was then drained and washed with 5 mL portions of DCM (3x), MeOH (3x) and DCM (5x). The washed resin was dried overnight under vacuum. The tagging efficiency was determined by GC analysis (see Example 3).

(b) Attachment of amines NH_?R³. The three batches of tagged resin 7 which contain aldehyde functional groups (i.e., those generated from stannane reagents #7, 8, and 9 in Table 2) were pooled and divided into a pre-determined number of reaction vessels according to the number of R³ substituents. The three batches were transfeπed to a reaction vessel (v = 100 mL) and suspended in DCM (50 69

mL). The mixture was shaken for two hours, drained and dried under vacuum for 48 hours. Since ten amine reagents were used in this reductive amination, a portion (300 mg) of the pooled resin 7 was placed into each often reaction vessels. Each of the ten batches of the resin was reacted with one of the amines shown in Table 3. For example, a suspension of 7 (300 mg, 0.084 mmol) in

TMOF (8 mL) was shaken for ten minutes and l-(3-aminopropyl)-2-pyπolidinone (235 mL, 1.68 mmol) was added. The suspension was shaken for 2 hours and NaBH3CN (106 mg, 1.68 mmol) was added. The mixture was then shaken for an additional 2 hours. The resin was washed with 1% HCl/MeOH (5 mL). MeOH (5 mL, 5x), 10% Hunig's base/DCM

(5 mL), MeOH (5 mL, 5x), DMF (5 mL, 5x) and DCM (5 mL, 5x) to yield the final quinolone resin 8. The washed resin 8 was dried under vacuum overnight. Each of these 10 final resin 8 batches was individually stored as a separate sub- library, obviating the need for encoding in this final synthetic step.

EXAMPLE 2 VERIFICATION OF SYNTHESIS

Several members from this library were synthesized in parallel on the solid phase to confirm the validity of the synthetic route and the identity of die final products. The target compounds were cleaved from the resin via one hour shaking in neat TFA. The structures were confirmed by ^lH NMR and mass spectroscopy. 70

mass spectrum (FAB): m/z = 328 (MH^*)

O 0

mass spectrum (ESI): m/z = 409 (MlT)

EXAMPLE3 DECODING PROCEDURE

An encoded bead is transfeπed into an insert placed in a GC sampling vial. 2 μL of oxidation solution (H20/CAN's solution (0.1 M aq. solution)/ACN = 1:3:16) and 10 mL of octane are added and die two-phase mixture centrifuged briefly (IK x one minute). The vial is screw capped and left at 35°C for four hours, then opened.

The organic layer is removed by syringe and mixed with 1 μL of MSTFA. The silated

tag solution (1 μL) is analyzed by GC with an ECD detector.

The GC analysis is performed with a Hewlett Packard 5890 plus . TM gas chromatograph using a column injection method (5 , 0.32 mm retention gap connected to a 25 m, 0.2 mm crosslinked 5% phenylmethyl silicone column). The 71

temperature and pressure programs for the analysis are 200-320°C ramp at

15°C/minute, then hold at 320°C for ten minutes and 20-40 psi at two psi/minute,

then 40 psi for ten minutes. The ECD detector is maintained at 400°C and the

auxiliary gas (He) is set at 35 psi.

The identity of the library compound attached to the bead is ascertained based on the reagents utilized in the synthesis of the compound which are readily determined from the binary codes associated, respectively, with each of the identifiers for such reagents as characterized through the above procedure. The binary codes for the identifiers are assigned to the various reagents (Tables 1 and 2) prior to initiating synthesis of the combinatorial library

72

Scheme 1 Attachment of Acyl Meldrum's Acid to Resin

0

If 0 π o o

^-OH ^*J0 toluene

^FO^

1 2 7 ^ ti ^^^<ϊ-&

Br'

1 : TentaGel S - PHB resin

2 : prepared by condensation of benzoyl chloride with Meldrum's acid

73

Scheme 2 Attachment of N,N-Dimethylformamide Dimethyl Acetal

^T ^^O-^ (MeO)₂CHN(Me)₂

B ^^^F ethyl ether

Scheme 3 Attachment of Primary Amines (NH₂R¹)

o

0 ^ vflpP R nΉ^H. _

F ^THF

(see Table 1 for a selection of R¹ amines)

Scheme 4 Coupiing of Stannanes (Bu₃SnR²) to Quinolone Derivatives on Resin

Li(TMS) N

^-

«-•

Br' ^"F THF Br'

Bu₃SnR² G-φ

Pd₂(dba)₃, TPP, DEE

N

R¹

(see Table 2 for a selection of R² stannanes)

76

Scheme 5 Attachment of Amines (NH₂R³)

0 o O 0 o-Q (1) NH₂R³, NaBH₃CN, TMOF

γ-

(2) 1% HCl / MeOH

R³HNH₂C

(R² = YCHO in some cases)

(see Table 3 for a selection of R amines)

77

Scheme 6 Preparation of Identifiers

7^7 OH ,0-(CH₂)_n-0-Ar

HO-(CH₂)_n-0-Ar

H₃CO. H₃CO.

"OCH₃ PPh₃, DEAD, Toluene ^{^}OCH₃

1. LiOH, THF/MeOH J5-(CH₂)_n-0-Ar CH₂N₂, DCM, Et₂0

2. SOCI₂,Toluene refluuxx Cl,

OCHa

TMS-CHN₂, Et₃N 0°C, THF/MeCN (1:1)

78

Table 1. Primary Amines (R¹NH₂)

R¹N R¹N

O^/^M MeS ^^

1. 8.

^ _/

2. 9.

H ^0=^\ _N

3. 10.

O^N J^^N

4. 11. ^<^-

CT^N r— N— ^N

5. 12.

N

6. 13.

NC— ^N

7. 14.

79

Table 1. Primary Amines (R¹NH₂) (continued)

R¹N RN

XX (!θ^N

15. 20.

16. N 21. cx

(θ

17. 22. cr^N

18. coo Ph

Ph —^N

23.

19. 24.

80

Table 2. Stannanes (Bu₃SnR )

R² R²

-^

1. 6.

o OCH₂CH₃

- ^~CH0

2. 7.

3. -o X CHO 8.

4. — Ό V X^sIv^CHO

9.

7)

5.

81

Table 3. Amine Reagents (R³NH₂)

R³NH R³NH

.NH

^'NH

2.

NH

,O ^^

NH

4.

NH NH

<

6.

NH

C NT^NH

NH X cOχO^_NH

10.

Claims

82ClaimsWe claim:

1. A combinatorial chemical library comprising a plurality of members of the

Formula I:

(T-L)_q- -Z

I

wherein: is a tag; L- is a first linker;

T'-L when taken together, form an identifier residue; q is 0-15;

s is a solid support; wherein - ΓÇö is ΓÇö , wherein

L'- is a second linker;

ΓÇö is a solid support to which the second linker attaches; 83

is a compound of formula

, wherein

R is chosen from the group consisting of C╬╣.₂o alkyl, aryl, arylalkyl, either aryl or heteroaryl fused to a 3- or 4-membered moiety to form a non-aromatic second ring, cycloalkyl, cycloalkylalkyl, heterocycloalkylalkyl, heteroaryl, heteroarylalkyl, substituted alkyl, substituted aryl and substituted heteroaryl;

W is R² or R³NHCH₂Y-, wherein

R" is chosen from the group consisting of alkenyl, alkynyl, aryl, heteroaryl, substituted alkenyl, substituted alkynyl, substituted aryl and substituted heteroaryl;

R³ is chosen from the group consisting of alkyl, alkoxyalkyl, arylalkyl, cycloalkylalkyl, heteroalkyl and heterocycloalkylalkyl; and

Y is chosen from the group consisting of aryl, heteroaryl, substituted aryl and substituted heteroaryl. 84

2. A combinatorial chemical library of claim 1 wherein:

R¹ is chosen from the group consisting of the amine residues of Table 1;

W is R² or R³NHCH₂Y- wherein:

R² is chosen from the group consisting of the stannane residues of Table 2;

R³ is chosen from the group consisting of the amine residues of Table 3; and

3. A combinatorial chemical library of claim 2 wherein: R¹ is chosen from the group consisting of benzyl, butyl, 2-cyanoethyl, cyclopropylmethyl, cyclohexylmethyl 2-dimethylaminoethyl, 3,3-dimethylbutyl furfuryl, 3-(l-imidazoyl)ethyl, isobutyl,2-methylbutyl, 4-methylphenethyl, 2-methylthioethyl, 2-(l-morpholinyl)ethyl, myrantyl, l-naphthylenemethyl,piperonyl 2-phenylphenethyl, 2-(l-py╧Çolidinyl)ethyl, 2-(2-aminoethyl)-l-py╧Çolidinyl, 2-pyridylethyl, 3-pyridylmethyl,tetrahydrofurfuryl and 2-thienylmethyi. 85

4. A combinatorial chemical library of claim 2 wherein:

R² is chosen from the group consisting of allyl, 1 -ethoxyvinyl, 3-formylphenyl, 4-formylphenyl, 3-formyl-2-thienyl, 2-furanyl, phenyl, 2-phenylethynyl, diienyl, and vinyl.

5. A combinatorial chemical library of claim 2 wherein:

R³ is chosen from the group consisting of butyl, benzyl, cyclohexylmethyl, 3-isopropoxypropyl, 2-methoxyethyl, phenethyl, 2-phenylphenethyl, piperonyl, 3-pyridylmethyl and 3-(l-py╧Çolidi-2-one)propyl.

6. A combinatorial chemical library of claim 2 wherein:

7. A combinatorial chemical library according to claim 1 wherein: T'-L is of the Formula II

^(CH₂)n ΓÇö o-Ar OCH₃

E wherein n = 3-12; Ar is halophenyl; 86

8. A combinatorial chemical library according to claim 7 wherein: Ar is pentachlorophenyl.

9. A combinatorial chemical library of claim 1 wherein: -L' is an acid cleavable linker.

10. A combinatorial chemical library of claim 9 wherein:

-L' is chosen from the group consisting of ^^ and

^<~"-^╧Ç3 wherein the benzyloxy group is attached to die compound Z

and the oxymethylene is attached to the solid support, ┬⌐ ΓÇö ^*

11. A combinatorial chemical library of claim 10 wherein:

-L' is ho ^-^r┬░- wherein -L' is chosen from

3-oxymethylene-l -benzyloxy and 4-oxymethylene-l -benzyloxy.

12. A combinatorial chemical library of claim 10 wherein:

-L' is ╧ÇO^{╬╢ u╧à}r^╧Ç3 wherein -L' is chosen from

3-methoxy-4-benzy loxy- 1 -oxymethylene and 3-benzyloxy-4-methoxy- 1 -oxymethylene. 87

13. A combinatorial chemical library of claim 1 wherein: q is zero; and

╬ÿ -L'-Z

ΓÇö is the compound -Z attached by the linker L' to d e solid support,

14. A method of synthesizing a library comprising a plurality of members of Formula I which comprises:

-Z

a. attaching an acyl Meldrum"s acid to a solid support via its carbonyl group to form a resin-linked ╬▓-keto ester;

b. attaching a dimethyl acetal to the resin-linked ╬▓-keto ester to provide a first

resin-linked enamine; c. displacing a dimethyl amino moiety from the first resin-linked enamine with a primary amine to form a second resin-linked enamine; d. cyclizing the second resin-linked enamine with a bis(trimethylsilyl)amide salt in an anhydrous solvent to form a resin-linked quinolone derivative; and e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether witii a trialkyl stannane in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a rpsin-linked quinolone. 88

15. The method of claim 14, further comprising: cleaving the resin-linked quinolone from the resin to provide a quinolone of Formula m;

HO'

W i , in.

16. A method of synthesizing a library comprising a plurality of members of Formula I which comprises:

-Z

a. attaching an acyl Meldrum's acid to a solid support via its carbonyl group to form a resin-linked ╬▓-keto ester;

resin-linked enamine; c. displacing a dimethyl amino moiety from the first enamine resin witii a primary amine to form a second resin-linked enamine; d. cyclizing the second resin-linked enamine with a bis(trimethylsilyl)amide salt in an anhydrous solvent to form a resin-linked quinolone derivative; e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether with a trialkyl stannane in the presence of triphenylphosphine and palladium 89

dibenzylidene acetone complex to produce a resin-linked quinolone containing an aldehyde moiety; f. coupling the resin-linked quinolone containing an aldehyde moiety with a primary amine in the presence of a trialkyl orthoformate and a cyanoborohydride, followed by treatment with an acid to produce an intermediate resin-linked amino quinolone salt; and g. neutralizing the resin-linked amino quinolone salt with a tertiary amine, followed by washing witii anhydrous solvents to produce a resin-linked amino quinolone.

17. The method of claim 16, further comprising: cleaving the resin-linked amino quinolone from the resin to provide an amino quinolone of Formula IH;

W m.

18. The method of claim 14 which comprises:

a. attaching an acyl Meldrum's acid of the formula:

⁰ O s C J 90

in toluene to a solid support suspended in toluene to form a resin-linked ╬▓-keto ester

of the formula:

O 0

b. attaching a dimethyl acetal in diethyl ether to die resin-linked ╬▓-keto ester to provide a first resin-linked enamine of the formula:

O

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R^lNH₂ to form a second resin-linked enamine of the formula:

L S (T'-D_Q

Br' ^*F NHR¹

d. cyclizing the second resin-linked enamine with lithium bis(trimethylsilyl)amide in THF to form a resin-linked quinolone derivative of the formula: O o j^L s (T'-L)_q

; and

e. treating the resin-linked quinolone derivative in ediylene glycol diediyl ether with a tributyl stannane of the formula Bu₃SnR² in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to afford a resin-linked quinolone of the formula:

O O (T'-L)_q

wherein

R is chosen from the group consisting of C ╬╣.₂o alkyl, aryl, arylalkyl, eitiier aryl or heteroaryl fused to a 3- or 4-membered moiety to form a non-aromatic second ring, cycloalkyl, cycloalkylalkyl, heterocycloalkylalkyl, heteroaryl, heteroarylalkyl, substituted alkyl, substituted aryl and substituted heteroaryl; and

19. The method of claim 18, further comprising: cleaving the resin-linked quinolone from the resin to provide a quinolone of Formula ni; 92

0

w m, wherein W = R^z

20. The method of claim 18 which comprises: a. attaching an acyl Meldrum's acid of me formula:

O

in toluene to a solid support suspended in toluene to form a resin-linked ╬▓-keto ester of the formula: O

b. attaching a dimethyl acetal in diethyl ether to the resin-linked ╬▓-keto ester to provide a first resin-linked enamine of the formula:

O

93

O O

; and

e. treating the resin-linked quinolone derivative in ediylene glycol diethyl ether with a tributyl stannane of the formula Bu₃SnR" in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a resin-linked quinolone of the formula:

O O

wherein: R , ╬╣ is chosen from the group consisting of the amine residues of Table 1; and 94

R is chosen from the group consisting of the stannane residues of Table 2.

21. The method of claim 20, further comprising: cleaving the resin-linked quinolone from the resin to provide a quinolone of Formula m;

HO

W m, wherein W = R².

22. The method of claim 20 Formula I which comprises: a. attaching an acyl Meldrum's acid of me formula:

O

of the formula:

O 0

b. attaching a dimethyl acetal in diethyl ether to the resin-linked ╬▓-keto ester

to provide a first resin-linked enamine of the formula:

O O S \(T'-L)_q

d. cyclizing die second resin-linked enamine with lithium bis(trimethylsilyl)amide in THF to form a resin-linked quinolone derivative of the formula: O l╬¢ S )-(T'-L)_q

; and

e. treating the resin-linked quinolone derivative in ethylene glycol diethyl ether with a tributyl stannane of the formula Bu₃SnR² in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to afford a resin- linked quinolone of the formula: 96

0

X

L > -(T-L)_q

R" ^ N

wherein: R¹ is chosen from the group consisting of benzyl, butyl, 2-cyanoethyl, cyclopropylmethyl, cyclohexylmethyl, 2-dimed╬╣ylaminoethyl, 3,3-dimed╬╣ylbutyl, furfuryl, 3-(l-imidazoyl)ethyl, isobutyl, 2-methylbutyI, 4-methylphenethyl, 2-med╬╣yld╬╣ioethyl, 2-(l-mo╧åholinyl)ethyl, myrantyl, 1-naphthylenemethyl, piperonyl 2-phenylphenethyl, 2-( 1 -pyrrolidinyl)ethyl, 2-(2-aminoethyl)- 1 -py╧Çolidinyl, 2-pyridylethyl, 3-pyridylmethyl,tetrahydrofurfuryl and 2-thienylmethyl; and

23. The method of claim 22, further comprising: cleaving die resin-linked quinolone from the resin to provide a quinolone of Formula m_; o o

HO

N W

R¹ m, wherein W = R". 97

24. The method of claim 16 which comprises a. attaching an acyl Meldrum's acid of die formula:

0 O o

of the formula:

O O

b. attaching a dimethyl acetal in diethyl ether to die resin-linked ╬▓-keto ester

to provide a first resin-linked enamine of the formula:

c. displacing a dimethyl amino moiety from the first resin-linked enamine in THF with a primary amine of the formula R╬ùH₂ to form a second resin-linked enamine of the formula:

O 0

^Γûá L-H S }-(T'-L)_╬▒

98

d. cyclizing the second resin-linked enamine with lithium bis(trimethylsilyl)amide in THF to form a resin-linked quinolone derivative of die formula:

O O

e. treating the resin-linked quinolone derivative in ed ylene glycol dialkyl ether witii a tributyl stannane of the formula Bu₃SnR² in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to afford a resin-linked quinolone containing an aldehyde moiety of formula:

f. coupling the resin-linked quinolone containing an aldehyde moiety with a primary amine of the formula R³NH₂ in the presence of a trialkyl orthoformate and a cyanoborohydride, followed by treatment witii an acid to produce an intermediate resin-linked amino quinolone salt of the formula:

O 0 x: S V(T'-L)_q

R³NH₂CH₂Y

; and 99

O O

R³NHCH₂Y

wherein:

R is chosen from the group consisting of C ╬╣.₂o alkyl, aryl, arylalkyl, either aryl or heteroaryl fused to a 3- or 4-membered moiety to form a non-aromatic second ring, cycloalkyl, cycloalkylalkyl, heterocycloalkylalkyl, heteroaryl, heteroarylalkyl, substituted alkyl, substituted aryl and substituted heteroaryl; and

R² is chosen from the group consisting of alkenyl, alkynyl, aryl, heteroaryl, substituted alkenyl, substituted alkynyl, substituted aryl and substimted heteroaryl.

Y is chosen from the group consisting of aryl, heteroaryl, substituted aryl and substituted heteroaryl. 100

25. The method of claim 24, further comprising: cleaving the resin-linked amino quinolone from the resin to provide a quinolone of the

Formula El;

O O HO

w m, wherein W = R^jNHCH₂Y-.

26. The method of claim 24 which comprises: a. attaching an acyl Meldrum's acid of die formula:

O

in toluene to a solid support suspended in toluene to form a resin-linked ╬▓-keto ester of the formula:

O O

to provide a first resin-linked enamine of the formula:

O O

101

O O

e. treating the resin-linked quinolone derivative in ethylene glycol dialkyl ether with a tributyl stannane of the formula Bu₃SnR" in the presence of triphenylphosphine and palladium dibenzylidene acetone complex to produce a resin-linked quinolone containing an aldehyde moiety of formula:

O 0

f. coupling the resin-linked quinolone containing an aldehyde moiety with a primary amine of the formula R³NH₂ in the presence of a trialkyl orthoformate and a 102

cyanoborohydride, followed by treatment with an acid to produce an intermediate resin-linked amino quinolone salt of the formula:

O 0 x-

R³NH₂CH₂Y

; and

O O

R³NHCH₂Y

wherein: R^l is chosen from the group consisting of the amine residues of Table 1 ; R² is chosen from the group consisting of the stannane residues of Table 2; R³ is chosen from the group consisting of the amine residues of Table 3; and Y is chosen from the group consisting of aryl, heteroaryl, substituted aryl and substituted heteroaryl.

27. The method of claim 26, further comprising: cleaving the resin-linked amino quinolone from the resin to provide a quinolone of the

Formula III; 103

m, wherein W = R NHCH₂Y-.

28. The method of claim 26 which comprises: a. attaching an acyl Meldrum's acid of the formula:

O

of the formula:

O O

b. attaching a dimethyl acetal in diethyl ether to d e resin-linked ╬▓-keto ester

to provide a first resin-linked enamine of the formula:

O O

104

O O

d. cyclizing die second resin-linked enamine with lithium bis(trimethylsilyl)amide in THF to form a resin-linked quinolone derivative of the formula:

O O

O O

(T'-L)_Q

f. coupling the resin-linked quinolone containing an aldehyde moiety with a primary amine of the formula R³NH₂ in the presence of a trialkyl orthoformate and a 105

cyanoborohydride, followed by treatment with an acid to produce a resin-linked intermediate amino quinolone salt of the formula:

O O x-

R³NH₂CH₂Y

and

O O

R³NHCH₅,Y

wherein:

R¹ is chosen from the group consisting of benzyl, butyl, 2-cyanoethyl, cyclopropylmethyl, cyclohexylmethyl, 2-dimethylaminoed╬╣yl, 3,3-dimethylbutyl, furfuryl, 3-(l-imidazoyl)ethyl, isobutyl, 2-methylbutyl, 4-methylphenethyl, 2-methylthioethyl, 2-(l-morpholinyl)ethyl, myrantyl, 1-naphthylenemethyl, piperonyl 2-phenylphenethyl, 2-( 1 -pyrrolidinyl)ethyl, 2-(2-aminoethyl)- 1 -py╧Çolidinyl, 2-pyridylethyl, 3-pyridylmethyl,tetrahydrofurfuryl and 2-thienylmethyl.

R is chosen from the group consisting of allyl, 1 -ethoxyvinyl, 4-formylphenyl,

3-formyl-2-thienyl, 2-furanyl, phenyl, 2-phenylethynyl, thienyl and vinyl. 106

R³ is chosen from the group consisting of butyl, benzyl, cyclohexylmediyl, 3-isopropoxypropyl, 2-memoxyethyl, phenethyl, 2-phenylphenethyl, piperonyl, 3-pyridylmethyl and 3-(l-py╧Çolidi-2-one^ropyl; and

Y is chosen from the group consisting of m-substituted phenyl, p-substituted phenyl and 3-substituted diienyl.

29. The method of claim 28, further comprising: cleaving the resin-linked quinolone from the resin to provide a quinolone of Formula m; O HO^'

W R¹ m, wherein W = R³NHCH₂Y-