Note: Within nine months from the publication of the mention of the grant of the European patent, any person may givenotice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed ina written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art.99(1) European Patent Convention).

Printed by Jouve, 75001 PARIS (FR)

(19)E

P0

736

105

B1

(Cont. next page)

��&������������(11) EP 0 736 105 B1

(12) EUROPEAN PATENT SPECIFICATION

(45) Date of publication and mention of the grant of the patent: 21.11.2007 Bulletin 2007/47

(21) Application number: 94928621.5

(22) Date of filing: 16.09.1994

(51) Int Cl.: �C12N 15/10 (2006.01) C12Q 1/68 (2006.01)

C07H 21/04 (2006.01) C07H 21/02 (2006.01)

(86) International application number: PCT/US1994/010562

(87) International publication number: WO 1995/008003 (23.03.1995 Gazette 1995/13) �

(54) SYSTEMATIC EVOLUTION OF LIGANDS BY EXPONENTIAL ENRICHMENT: PHOTOSELECTION OF NUCLEIC ACID LIGANDS

SYSTEMATISCHE EVOLUTION VON LIGANDEN DURCH EXPONENTIELLE ANREICHERUNG: FOTOSELEKTION VON NUKLEINSÄURELIGANDEN

EVOLUTION SYSTEMATIQUE DE LIGANDS PAR ENRICHISSEMENT EXPONENTIEL : PHOTOSELECTION DE LIGANDS D’ACIDE NUCLEIQUE

(84) Designated Contracting States: AT BE CH DE DK ES FR GB GR IE IT LI LU MC NL PT SE

(30) Priority: 17.09.1993 US 12393525.10.1993 US 143564

(43) Date of publication of application: 09.10.1996 Bulletin 1996/41

(60) Divisional application: 07016085.8

(73) Proprietor: Somalogic, Inc. �Boulder, CO 80301 (US) �

(72) Inventors: • GOLD, Larry

Boulder, CO 80303 (US) �• WILLIS, Michael

Louisville, CO 80027 (US) �• KOCH, Tad

Boulder, CO 80303 (US) �• RINGQUIST, Steven

Lyons, CO 80540 (US) �• JENSEN, Kirk

Boulder, CO 80302 (US) �• ATKINSON, Brent

Boulder, CO 80302 (US) �

(74) Representative: Clegg, Richard IanMewburn Ellis LLP York House 23 KingswayLondon WC2B 6HP (GB) �

(56) References cited: WO- �A- �91/19813 WO-�A- �93/05182WO- �A- �95/07364 US-�A- 5 035 996

• KATOUZIANSAFADI M (REPRINT) ET AL: "DETERMINATION OF THE DNA- �INTERACTING REGION OF THE ARCHAEBACTERIAL CHROMOSOMAL PROTEIN MC1. PHOTOCROSSLINKS WITH 5- �BROMOURACIL- �SUBSTITUTED DNA" NUCLEIC ACIDS RESEARCH, (1991) VOL. 19, NO. 18, PP. 4937-4941., XP002065572

• Y.J.K. FARRAR ET AL.: "Interactions of photoactive DNAs with terminal deoxynucleotidyl transferase: Identification of peptides in the DNA binding domain" BIOCHEMISTRY, vol. 30, 1991, AM. CHEM SOC., US, pages 3075-3082, XP002065573

• J.M. GOTT ET AL.: "A specific, UV- �induced RNA- �protein cross- �link using 5- �Bromouridine- �substituted RNA" BIOCHEMISTRY, vol. 30, 1991, AM. CHEM. SOC., US, pages 6290-6295, XP002065574

2

EP 0 736 105 B1

• DIETZ T M ET AL: "PHOTOCHEMICAL COUPLING OF 5 BROMOURACIL BU TO A PEPTIDE LINKAGE A MODEL FOR BU DNA PROTEIN PHOTOCROSSLINKING." J AM CHEM SOC 109 (6). 1987. 1793-1797. CODEN: JACSAT ISSN: 0002-7863, XP002065575

• AKHLYNINA T V ET AL: "Using in vitro selection to direct the covalent attachment of human immunodeficiency virus type 1 Rev protein to high- �affinity RNA ligands." PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, (1995 DEC 19) 92 (26) 12220-4. JOURNAL CODE: PV3. ISSN: 0027-8424., XP002065576

• SCIENCE, Volume 249, issued 03 August 1990, C. TUERK et al., "Systematic Evolution of Ligands by Exponential Enrichment: RNA Ligands to Bacteriophage T4 DNA Polymerase", pages 505-510.

EP 0 736 105 B1

3

5

10

15

20

25

30

35

40

45

50

55

Description

FIELD OF THE INVENTION

�[0001] This invention relates to a method for selectingnucleic acid ligands which bind and/or photocrosslink toand/or photoinactivate a target molecule. The target mol-ecule may be a protein, pathogen or toxic substance, orany biological effector. The nucleic acid ligands of thepresent invention contain photoreactive or chemically re-active groups and are useful, inter alia, for the diagnosisand/or treatment of diseases or pathological or toxicstates.�[0002] The underlying method utilized in this inventionis termed SELEX, an acronym for Systematic Evolutionof Ligands by Exponential enrichment. An improvementof the SELEX method herein described, termed SolutionSELEX, allows more efficient partitioning between oligo-nucleotides having high and low affinity for a target mol-ecule. An improvement of the high affinity nucleic acidproducts of SELEX are useful for any purpose to whicha binding reaction may be put, for example in assay meth-ods, diagnostic procedures, cell sorting, as inhibitors oftarget molecule function, as therapeutic agents, asprobes, as sequestering agents and the like.

BACKGROUND OF THE INVENTION

�[0003] The SELEX method (hereinafter termed SE-LEX), described in U.S. Patent Application Serial No.07/536,428 (WO 91/19813), filed June 11, 1990, entitledSystematic Evolution of Ligands By Exponential Enrich-ment, now abandoned. U.S. Patent Application Serial No.07/714,131 (US 5,475,096), filed June 10, 1991, entitledNucleic Acid Ligands, and U.S. Patent Application SerialNo. 07/931,473, filed August 17, 1992, entitled NucleicAcid Ligands, issued as U.S. Patent No. 5,270,163 (re-ferred to herein as the SELEX Patent Applications), pro-vides a class of products which are nucleic acid mole-cules, each having a unique sequence, each of whichhas the property of binding specifically to a desired targetcompound or molecule. Each nucleic acid molecule is aspecific ligand of a given target compound or molecule.SELEX is based on the unique insight that nucleic acidshave sufficient capacity for forming a variety of two- andthree-�dimensional structures and sufficient chemical ver-satility available within their monomers to act as ligands(form specific binding pairs) with virtually any chemicalcompound, whether monomeric or polymeric. Moleculesof any size can serve as targets.�[0004] The SELEX method involves selection from amixture of candidates and step-�wise iterations of struc-tural improvement, using the same general selectiontheme, to achieve virtually any desired criterion of bindingaffinity and selectivity. Starting from a mixture of nucleicacids, preferably comprising a segment of randomizedsequence, the method includes steps of contacting themixture with the target under conditions favorable for

binding, partitioning unbound nucleic acids from thosenucleic acids which have bound to target molecules, dis-sociating the nucleic acid- �target pairs, amplifying the nu-cleic acids dissociated from the nucleic acid-�target pairsto yield a ligand-�enriched mixture of nucleic acids, thenreiterating the steps of binding, partitioning, dissociatingand amplifying through as many cycles as desired.�[0005] While not bound by theory, SELEX is based onthe inventors’ insight that within a nucleic acid mixturecontaining a large number of possible sequences andstructures there is a wide range of binding affinities for agiven target. A nucleic acid mixture comprising, for ex-ample a 20 nucleotide randomized segment can have420 candidate possibilities. Those which have the higheraffinity constants for the target are most likely to bind tothe target. After partitioning, dissociation and amplifica-tion, a second nucleic acid mixture is generated, enrichedfor the higher binding affinity candidates. Additionalrounds of selection progressively favor the best ligandsuntil the resulting nucleic acid mixture is predominantlycomposed of only one or a few sequences. These canthen be cloned, sequenced and individually tested forbinding affinity as pure ligands.�[0006] Cycles of selection, partition and amplificationare repeated until a desired goal is achieved. In the mostgeneral case, selection/�partition/�amplification is contin-ued until no significant improvement in binding strengthis achieved on repetition of the cycle. The method maybe used to sample as many as about 1018 different nu-cleic acid species. The nucleic acids of the test mixturepreferably include a randomized sequence portion aswell as conserved sequences necessary for efficient am-plification. Nucleic acid sequence variants can be pro-duced in a number of ways including synthesis of rand-omized nucleic acid sequences and size selection fromrandomly cleaved cellular nucleic acids. The variable se-quence portion may contain fully or partially random se-quence; it may also contain subportions of conservedsequence incorporated with randomized sequence. Se-quence variation in test nucleic acids can be introducedor increased by mutagenesis before or during the selec-tion/�partition/�amplification iterations.�[0007] Photocrosslinking of nucleic acids to proteinshas been achieved through incorporation of photoreac-tive functional groups in the nucleic acid. Photoreactivegroups which have been incorporated into nucleic acidsfor the purpose of photocrosslinking the nucleic acid toan associated protein include 5- �bromouracil, 4-�thiouracil,5-�azidouracil, and 8-�azidoadenine (see Fig. 1).�[0008] Bromouracil has been incorporated into bothDNA and RNA by substitution of bromodeoxyuracil (Br-dU) and bromouracil (BrU) for thymine and uracil, respec-tively. BrU-�RNA has been prepared with 5-�bromouridinetriphosphate in place of uracil using T7 RNA polymeraseand a DNA template, and both BrU-�RNA and BrdU-�DNAhave been prepared with 5-�bromouracil and 5-�bromode-oxyuracil phosphoramidites, respectively, in standardnucleic acid synthesis (Talbot et al. (1990) Nucleic Acids

1 2

EP 0 736 105 B1

4

5

10

15

20

25

30

35

40

45

50

55

Res. 18: �3521). Some examples of the photocrosslinkingof BrdU- �substituted DNA to associated proteins are asfollows: BrdU- �substituted DNA to proteins in intact cells(Weintraub (1973) Cold Spring Harbor Symp. Quant. Bi-ol. 38: �247); BrdU-�substituted lac operator DNA to lac re-pressor (Lin and Riggs (1974) Proc. Natl. Acad. Sci.U.S.A. 71:�947; Ogata and Wilbert (1977) Proc. Natl.Acad. Sci. U.S.A. 74:�4973; Barbier et al. (1984) Biochem-istry 23:�2933; Wick and Matthews (1991) J. Biol. Chem.266: �6106); BrdU- �substituted DNA to EcoRI and EcoRVrestriction endonucleases (Wolfes et al. (1986) Eur. J.Biochem. 159:�267); Escherichia coli BrdU- �substitutedDNA to cyclic adenosine 3’,�5’-�monophosphate receptorprotein (Katouzian-�Safadi et al. (1991) Photochem. Pho-tobiol. 53: �611); BrdU-�substituted DNA oligonucleotide ofhuman polyomavirus to proteins from human fetal brainextract (Khalili et al. (1988) EMBO J. 7:�1205); a yeastBrdU- �substituted DNA oligonucleotide to GCN4, a yeasttranscriptional activator (Blatter et al. (1992) Nature 359:650); and a BrdU-�substituted DNA oligonucleotide ofMethanosarcina sp CHT155 to the chromosomal proteinMc1 (Katouzian-�Safadi et al. (1991) Nucleic Acids Res.19:�4937). Photocrosslinking of BrU-�substituted RNA toassociated proteins has also been reported: BrU- �substi-tuted yeast precursor tRNAPhe to yeast tRNA ligase (Tan-ner et al. (1988) Biochemistry 27:�8852) and a BrU-�sub-stituted hairpin RNA of the R17 bacteriophage genometo R17 coat protein (Gott et al. (1991) Biochemistry 30:6290).�[0009] 4-�Thiouracil-�substituted RNA has been used tophotocrosslink, especially, t- �RNA’s to various associatedproteins (Favre (1990) in: Bioorganic Photochemistry,Volume 1: Photochemistry and the Nucleic Acids, H.� Mor-rison (ed.), John Wiley & Sons: New York, pp. 379-425;Tanner et al. (1988) supra). 4-�Thiouracil has been incor-porated into RNA using 4-�thiouridine triphosphate andT7 RNA polymerase or using nucleic acid synthesis withthe appropriate phosphoramidite; it has also been incor-porated directly into RNA by exchange of the amino groupof cytosine for a thiol group with hydrogen sulfide. Yetanother method of site specific incorporation of photore-active groups into nucleic acids involves use of 4-�thiou-ridylyl-(3’-�5’)-guanosine (Wyatt et al. (1992) Genes & De-velopment 6:�2542).�[0010] Examples of 5-�azidouracil-�substituted and 8-azidoadenine-�substituted nucleic acid photocrosslinkingto associated proteins are also known. Associated pro-teins that have been crosslinked include terminal deox-ynucleotidyl transferase (Evans et al. (1989) Biochemis-try 28:�713; Farrar et al. (1991) Biochemistry 30: �3075);Xenopus TFIIIA, a zinc finger protein (Lee et al. (1991)J. Biol. Chem. 266: �16478); and E. coli ribosomal proteins(Wower et al. (1988) Biochemistry 27:�8114). 5-�Azidour-acil and 8- �azidoadenine have been incorporated intoDNA using DNA polymerase or terminal transferase. Pro-teins have also been photochemically labelled by exciting8-�azidoadenosine 3’, �5’- �biphosphate bound to bovinepancreatic ribonuclease A (Wower et al. (1989) Biochem-

istry 28: �1563) and 8- �azidoadenosine 5’-�triphosphatebound to ribulose-�bisphosphate carboxylase/�oxygenase(Salvucci and Haley (1990) Planta 181:�287).�[0011] 8-�Bromo- �2’- �deoxyadenosine as a potentialphotoreactive group has been incorporated into DNA viathe phosphoramidite (Liu and Verdine (1992) Tetrahe-dron Lett. 33:�4265). The photochemical reactivity has yetto be investigated.�[0012] Photocrosslinking of 5-�iodouracil- �substitutednucleic acids to associated proteins has not been previ-ously investigated, probably because the size of the iodogroup has been thought to preclude specific binding ofthe nucleic acid to the protein of interest. However, 5-iodo-�2’-�deoxyuracil and 5-�iodo-�2’-�deoxyuridine triphos-phate have been shown to undergo photocoupling to thy-midine kinase from E. coli (Chen and Prusoff (1977) Bi-ochemistry 16:�3310).�[0013] Mechanistic studies of the photochemical reac-tivity of the 5-�bromouracil chromophore have been re-ported including studies with regard to photocrosslinking.Most importantly, BrU shows wavelength dependentphotochemistry. Irradiation in the region of 310 nm pop-ulates an n,�π* singlet state which decays to ground stateand intersystem crosses to the lowest energy triplet state(Dietz et al. (1987) J. Am. Chem. Soc. 109:�1793), mostlikely the π,�π* triplet (Rothman and Kearns (1967) Pho-tochem. Photobiol. 6: �775). The triplet state reacts withelectron-�rich amino acid residues via initial electrontransfer followed by covalent bond formation. Photo-crosslinking of triplet 5-�bromouracil to the electron richaromatic amino acid residues tyrosine, tryptophan andhistidine (Ito et al. (1980) J. Am. Chem. Soc. 102: �7535;Dietz and Koch (1987) Photochem. Photobiol. 46:�971),and the disulfide bearing amino acid, cystine (Dietz andKoch (1989) Photochem. Photobiol. 49: �121), has beendemonstrated in model studies. Even the peptide linkageis a potential functional group for photocrosslinking totriplet BrU (Dietz et al. (1987) supra). Wavelengths some-what shorter than 308 nm populate both the n, �π* and π,π* singlet states. The π, �π* singlet undergoes carbon- �bro-mine bond homolysis as well as intersystem crossing tothe triplet manifold (Dietz et al. (1987) supra); intersystemcrossing may occur in part via internal conversion to then, �π* singlet state. � Carbon-�bromine bond homolysis likelyleads to nucleic acid strand breaks (Hutchinson andKöhnlein (1980) Prog. Subcell. Biol. 7:�1; Shetlar (1980)Photochem. Photobiol. Rev. 5: �105; Saito and Sugiyama(1990) in: Bioorganic Photochemistry, Volume 1: Photo-chemistry and the Nucleic Acids, H. Morrison, ed., JohnWiley and Sons, New York, pp. 317-378). The wave-length dependent photochemistry is outlined in the Jab-lonski Diagram in Figure 2 and the model photocrosslink-ing reactions are shown in Figure 3.�[0014] The location of photocrosslinks from irradiationof some BrU-�substituted nucleoprotein complexes havebeen investigated. In the lac repressor- �BrdU-lac operatorcomplex a crosslink to tyrosine-�17 has been established(Allen et al. (1991) J. Biol. Chem. 266:�6113). In the ar-

3 4

EP 0 736 105 B1

5

5

10

15

20

25

30

35

40

45

50

55

chaebacterial chromosomal protein MC1-�BrdU- �DNAcomplex a crosslink to tryptophan-�74 has been implicat-ed. In yeast BrdU-�substituted DNA-�GCN4 yeast tran-scriptional activator a crosslink to alanine-�238 was re-ported (Blatter et al. (1992) supra). In this latter examplethe nucleoprotein complex was irradiated at 254 nmwhich populated initially the π, �π* singlet state.�[0015] The results of some reactivity and mechanisticstudies of 5- �iodouracil, 5- �iodo- �2’-�deoxyuracil, 5-�iodo-�2’-deoxyuracil-�substituted DNA, and 5- �iodo-�2’-�deoxycyto-sine-�substituted DNA have been reported. 5- �Iodouraciland 5-�iodo- �2’-�deoxyuracil couple at the 5-�position to al-lylsilanes upon irradiation in acetonitrile-�water bearingexcess silane with emission from a medium pressuremercury lamp filtered through Pyrex glass; the mecha-nism was proposed to proceed through initial carbon-iodine bond homolysis followed by radical addition to theπ-�bond of the allylsilane (Saito et al. (1986) J. Org. Chem.51:�5148).�[0016] Aerobic and anaerobic photo- �deiodination of 5-iodo- �2’-�deoxyuracil- �substituted DNA has been studiedas a function of excitation wavelength; the intrinsic quan-tum yield drops by a factor of 4 with irradiation in theregion of 313 nm relative to the quantum yield with irra-diation in the region of 240 nm. At all wavelengths themechanism is proposed to involve initial carbon- �iodinebond homolysis (Rahn and Sellin (1982) Photochem.Photobiol. 35:�459). Similarly, carbon-�iodine bond homo-lysis is proposed to occur upon irradiation of 5-�iodo-�2’-deoxycytidine- �substituted DNA at 313 nm (Rahn andStafford (1979) Photochem. Photobiol. 30:�449). Strictlymonochromatic light was not used in any of these studies.Recently, a 5-�iodouracil-�substituted duplex DNA wasshown to undergo a photochemical single strand break(Sugiyama et al. (1993) J. Am. Chem. Soc. 115:�4443).�[0017] Also of importance with respect to the presentinvention is the observed direct population of the tripletstates of 5-�bromouracil and 5-�iodouracil from irradiationof the respective So→T absorption bands in the regionof 350-400 nm (Rothman and Kearns (1967) supra).�[0018] Photophysical studies of the 4-�thiouracilchromophore implicate the π, �π* triplet state as the reac-tive state. The intersystem crossing quantum yield is uni-ty or close to unity. Although photocrosslinking within 4-thiouracil-�substituted nucleoprotein complexes has beenobserved, amino acid residues reactive with excited 4-thiouracil have not been established (Favre (1990) su-pra). The addition of the α-�amino group of lysine to ex-cited 4-�thiouracil at the 6-�position has been reported;however, this reaction is not expected to be important inphotocrosslinking within nucleoprotein complexes be-cause the α-�amino group is involved in a peptide bond(Ito et al. (1980) Photochem. Photobiol. 32:�683).�[0019] Photocrosslinking of azide-�bearing nucleotidesor nucleic acids to associated proteins is thought to pro-ceed via formation of the singlet and/or triplet nitrene(Bayley and Knowles (1977) Methods Enzymol. 46:�69;Czarnecki et al. (1979) Methods Enzymol. 56:�642; Hanna

et al. (1993) Nucleic Acids Res. 21:�2073). Covalent bondformation results from insertion of the nitrene in an O- �H,N-�H, S-�H or C- �H bond. Singlet nitrenes preferentially in-sert in heteroatom-�H bonds and triplet nitrenes in C-�Hbonds. Singlet nitrenes can also rearrange to azirineswhich are prone to nucleophilic addition reactions. If anucleophilic site of a protein is adjacent, crosslinking canalso occur via this pathway. A potential problem with theuse of an azide functional group results if it resides orthoto a ring nitrogen; the azide will exist in equilibrium witha tetrazole which is much less photoreactive.�[0020] The coat protein-�RNA hairpin complex of theR17 bacteriophage is an ideal system for the study ofnucleic acid-�protein photocrosslinking because of thesimplicity of the system in vitro. The system is well char-acterized, consisting of a viral coat protein that binds withhigh affinity to an RNA hairpin within the phage genome.In vivo the interaction of the coat protein with the RNAhairpin plays two roles during phage infection: the coatprotein acts as a translational repressor of replicase syn-thesis (Eggens and Nathans (1969) J. Mol. Biol. 39:�293),and the complex serves as a nucleation site for encap-sidation (Ling et al. (1970) Virology 40:�920; Beckett etal. (1988) J. Mol Biol. 204:�939). Many variations of thewild- �type hairpin sequence also bind to the coat proteinwith high affinity (Tuerk & Gold (1990) Science 249:�505;Gott et al. (1991) Biochemistry 30:�6290; Schneider et al.(1992) J. Mol. Biol. 228:�862).�[0021] The selection of nucleic acid ligands accordingto the SELEX method may be accomplished in a varietyof ways, such as on the basis of physical characteristics.Selection on the basis of physical characteristics mayinclude physical structure, electrophoretic mobility, sol-ubility, and partitioning behavior. U. S. patent applicationSerial No. 07/960,093 (WO 94/09168), filed October 14,1992, entitled Method for Selecting Nucleic Acids on theBasis of Structure, herein specifically incorporated by ref-erence, describes the selection of nucleic acid sequenc-es on the basis of specific electrophoretic behavior. TheSELEX technology may also be used in conjunction withother selection techniques, such as HPLC, column chro-matography, chromatographic methods in general, sol-ubility in a particular solvent, or partitioning between twophases.

BRIEF SUMMARY OF THE INVENTION

�[0022] In one embodiment, the present invention in-cludes a method for selecting and identifying nucleic acidligands from a candidate mixture of randomized nucleicacid sequences on the basis of the ability of the rand-omized nucleic acid sequences to bind and/or photo-crosslink to a target molecule. This embodiment istermed Covalent SELEX generally, and PhotoSELEXspecifically when irradiation is required to form covalentlinkage between the nucleic acid ligand and the target.�[0023] In one variation of this embodiment, the methodcomprises preparing a candidate mixture of nucleic acid

5 6

EP 0 736 105 B1

6

5

10

15

20

25

30

35

40

45

50

55

sequences which contain photoreactive groups; contact-ing the candidate mixture with a target molecule whereinnucleic acid sequences having increased affinity to thetarget molecule bind the target molecule, forming nucleicacid-�target molecule complexes; irradiating the nucleicacid-�target molecule mixture, wherein some nucleic ac-ids incorporated in nucleic acid-�target molecule complex-es crosslink to the target molecule via the photoreactivefunctional groups; taking advantage of the covalent bondto partition the crosslinked nucleic acid-�target moleculecomplexes from free nucleic acids in the candidate mix-ture; and identifying the nucleic acid sequences that werephotocrosslinked to the target molecule. The process canfurther include the iterative step of amplifying the nucleicacids that photocrosslinked to the target molecule to yielda mixture of nucleic acids enriched in sequences that areable to photocrosslink to the target molecule.�[0024] In another variation of this embodiment of thepresent invention, nucleic acid ligands to a target mole-cule selected through SELEX are further selected fortheir ability to crosslink to the target. Nucleic acid ligandsto a target molecule not containing photoreactive groupsare initially identified through the SELEX method. Pho-toreactive groups are then incorporated into these se-lected nucleic acid ligands, and the ligands contactedwith the target molecule. The nucleic acid-�target mole-cule complexes are irradiated and those able to photo-crosslink to the target molecule identified.�[0025] In another variation of this embodiment of thepresent invention, photoreactive groups are incorporatedinto all possible positions in the nucleic acid sequencesof the candidate mixture. For example, 5-�iodouracil and5-�iodocytosine may be substituted at all uracil and cyto-sine positions. The first selection round is performed withirradiation of the nucleic acid-�target molecule complexessuch that selection occurs for those nucleic acid se-quences able to photocrosslink to the target molecule.Then SELEX is performed with the nucleic acid sequenc-es able to photocrosslink to the target molecule to selectcrosslinking sequences best able to bind the target mol-ecule.�[0026] In another variation of this embodiment of thepresent invention, nucleic acid sequences containingphotoreactive groups are selected through SELEX for anumber of rounds in the absence of irradiation, resultingin a candidate mixture with a partially enhanced affinityfor the target molecule. PhotoSELEX is then conductedwith irradiation to select ligands able to photocrosslink tothe target molecule.�[0027] In another variation of this embodiment of thepresent invention, SELEX is carried out to completionwith nucleic acid sequences not containing photoreactivegroups, and nucleic acid ligands to the target moleculeselected. Based on the sequences of the selected lig-ands, a family of related nucleic acid sequences is gen-erated which contain a single photoreactive group ateach nucleotide position. PhotoSELEX is performed toselect a nucleic acid ligand capable of photocrosslinking

to the target molecule.�[0028] In a further variation of this embodiment of thepresent invention, a nucleic acid ligand capable of mod-ifying the bioactivity of a target molecule through bindingand/or crosslinking to a target molecule is selectedthrough SELEX, photoSELEX, or a combination of thesemethods.�[0029] In a further variation of this embodiment of thepresent invention, a nucleic acid ligand to a unique targetmolecule associated with a specific disease process isidentified. In yet another variation of this embodiment ofthe present invention, a nucleic acid ligand to a targetmolecule associated with a disease state is used to treatthe disease in vivo.�[0030] The present invention further encompasses nu-cleic acid sequences containing photoreactive groups.The nucleic acid sequences may contain single or mul-tiple photoreactive groups. Further, the photoreactivegroups may be the same or different in a single nucleicacid sequence. The photoreactive groups incorporatedinto the nucleic acids of the invention include any chem-ical group capable of forming a crosslink with a targetmolecule upon irradiation. Although in some cases irra-diation may not be necessary for crosslinking to occur.�[0031] The nucleic acids of the present invention in-clude single- and double- �stranded RNA and single- anddouble-�stranded DNA. The nucleic acids of the presentinvention may contain modified groups such as 2’-�amino(2’-�NH2) or 2’- �fluoro (2’- �F)-modified nucleotides- The nu-cleic acids of the present invention may further includebackbone modifications.�[0032] The present invention further includes themethod whereby candidate mixtures containing modifiednucleic acids are prepared and utilized in the SELEXprocess, and nucleic acid ligands are identified that bindor crosslink to the target species. In one example of thisembodiment, the candidate mixture is comprised of nu-cleic acids wherein all uracil residues are replaced by 5-halogenated uracil residues, and nucleic acid ligands areidentified that form covalent attachments to the selectedtarget.�[0033] Also disclosed, and termed solution SELEX,are several improved methods for partitioning betweenligands having high and low affinity nucleic acid- �targetcomplexes is achieved in solution and without, or priorto, use of a partitioning matrix. Generally, a central themeof the method of solution SELEX is that the nucleic acidcandidate mixture is treated in solution and results in pref-erential amplification during PCR of the highest affinitynucleic acid ligands or catalytic RNAs. The solution SE-LEX method achieves partitioning between high and lowaffinity nucleic acid-�target complexes through a numberof methods, including (1) Primer extension inhibitionwhich results in differentiable cDNA products such thatthe highest affinity ligands may be selectively amplifiedduring PCR. Primer extension inhibition is achieved withthe use of nucleic acid polymerases, including DNA orRNA polymerases, reverse transcriptase, and Qβ-�repli-

7 8

EP 0 736 105 B1

7

5

10

15

20

25

30

35

40

45

50

55

case. (2) Exonuclease hydrolysis inhibition which alsoresults in only the highest affinity ligands amplifying dur-ing PCR. This is achieved with the use of any 3’→ 5’double-�stranded exonuclease. (3) Linear to circle forma-tion to generate differentiable cDNA molecules resultingin amplification of only the highest affinity ligands duringPCR.�[0034] In one embodiment of the solution SELEXmethod, synthesis of cDNAs corresponding to low affinityoligonucleotides are preferentially blocked and thus ren-dered non- �amplifiable by PCR. In another embodiment,low affinity oligonucleotides are preferentially removedby affinity column chromatography prior to PCR amplifi-cation. Alternatively, high affinity oligonucleotides maybe preferentially removed by affinity column chromatog-raphy. In yet another embodiment of the SELEXES meth-od, cDNAs corresponding to high affinity oligonucle-otides are preferentially rendered resistant to nucleaseenzyme digestion. In a further embodiment, cDNAs cor-responding to low affinity oligonucleotides are renderedpreferentially enzymatically or chemically degradable.�[0035] Solution SELEX is an improvement over priorart partitioning schemes. Partitioning is achieved withoutinadvertently also selecting ligands that only have affinityfor the partitioning matrix, the speed and accuracy of par-titioning is increased, and the procedure may be readilyautomated.�[0036] The present disclosure provides non-�limitingexamples which are illustrative and exemplary of the in-vention. Other partitioning schemes and methods of se-lecting nucleic acid ligands through binding and photo-crosslinking to a target molecule will be become apparentto one skilled in the art from the present disclosure.

BRIEF DESCRIPTION OF THE FIGURES

�[0037]

FIGURE 1 shows structures of photoreactivechromophores which have been incorporated intonucleic acids.

FIGURE 2 shows a Jablonski energy level diagramfor the 5- �bromouracil chromophore and the reactivityof the various excited states.

FIGURE 3 shows the model reactions for photo-crosslinking of the 5- �bromouracil chromophore toamino acid residues such as tyrosine, tryptophan,histidine, and cystine.

FIGURE 4 compares UV absorption by thymidine,5-�bromouracil, 5- �iodouracil, and L-�tryptophan inTMK pH 8.5 buffer (100 mM tris �(hydroxymethyl)�ami-nomethane hydrochloride, 10 mM magnesium ace-tate, and 80 mM potassium chloride). The emissionwavelengths of the XeCl and HeCd lasers are alsoindicated. Of particular importance is absorption by

5-�iodouracil at 325 nm without absorption by tryp-tophan or thymidine. The molar extinction coefficientfor 5-�iodouracil at 325 nm is 163 L/mol·cm.

FIGURE 5 shows structures of photoreactivechromophores which can be incorporated into ran-domized nucleic acid sequences.

FIGURE 6 shows the structures of hairpin RNA se-quences RNA-�1 (SEQ ID NO:�1), RNA-�2 (SEQ ID NO:2), and RNA-�3 (SEQ ID NO:�3) containing 5-�bromour-acil, 5-�iodouracil, and uracil, respectively. These arevariants of the wild-�type hairpin of the R17 bacteri-ophage genome which bind tightly to the R17 coatprotein.

FIGURE 7 shows binding curves for RNA-�1 (SEQ IDNO: �1), RNA-�2 (SEQ ID NO: �2), and RNA-�3 (SEQ IDNO: �3) to R17 coat protein. The binding constantscalculated from the binding curves are also given.

FIGURE 8 shows the percent of RNA-�1 (SEQ ID NO:1) and RNA-�2 (SEQ ID NO:�2) photocrosslinked toR17 coat protein with monochromatic irradiation at308 nm from a XeCl excimer laser as a function oftime. Photocrosslinking of RNA-�1 (SEQ ID NO:�1)maximized at 40% because of competitive photo-damage to the coat protein during the irradiation pe-riod. Less photodamage to coat protein occurredwith RNA-�2 (SEQ ID NO:�2) because of the shorterirradiation time.

FIGURE 9 shows the percent of RNA-�2 (SEQ ID NO:2) photocrosslinked to R17 coat protein with mono-chromatic irradiation at 325 nm from a HeCd laseras a function of time. The data are also presented inthe original electrophoretic gel format. The symbolIU XL marks RNA crosslinked to protein. A near-quantitative yield of photocrosslinking was obtained.

FIGURE 10 shows the percent of RNA-�1 (SEQ IDNO: �1) and RNA-�2 (SEQ ID NO:�2) photocrosslinkedto R17 coat protein with broad-�band irradiation in theregion of 312 nm from a Transilluminator as a func-tion of time. Less than quantitative yields of photo-crosslinking were obtained because of photodam-age to the protein and possibly to the RNA sequenc-es.

FIGURE 11 shows formation of the same product,Structure 6, from irradiation at 308 nm of 5-�iodouraciland 5-�bromouracil in the presence of excess N-acetyltyrosine N-�ethylamide (Structure 5).

FIGURE 12 pictures photocrosslinking of RNA-�7(SEQ ID NO:�4) to R17 coat protein with 308 nm lightfollowed by enzymatic digestion of most of the coatprotein.

9 10

EP 0 736 105 B1

8

5

10

15

20

25

30

35

40

45

50

55

FIGURE 13 shows formation of a complementaryDNA from the RNA template after enzymatic diges-tion of the coat protein of Figure 12 (SEQ ID NO: �4).

FIGURE 14 shows the polyacrylamide gel of Exam-ple 8 showing production of a cDNA from an RNAtemplate bearing modified nucleotides as shown inFigures 12 and 13. The modified nucleotides were5-�iodouracil and uracil substituted at the 5- �positionwith a small peptide. Based upon model studiesshown in Figure 11, the peptide was most likely at-tached to the uracil via the phenolic ring of a tyrosineresidue.

FIGURE 15 shows the photocrosslinking of [α-32P]GTP labelled pool RNA to HIV- �1 Rev protein usingtRNA competition.

FIGURE 16 (SEQ ID NOS: �5-55) shows the se-quence of the previously identified RNA ligand toHIV-�1 Rev protein that is referred to herein as 6a(SEQ ID NO:�5). Also shown are 52 sequences fromround 13 selected for photocrosslinking to HIV-�1 Revprotein.

FIGURE 17 (SEQ ID NOS:�56-57) shows the con-sensus for class 1 ligands and class 2 ligands. Class1: Consensus secondary structure for class 1 andclass 2 molecules. N1-N1’ indicate 1-2 complemen-tary base pairs; N2-N2’ indicates 1-4 complementarybase pairs, D-�H’ is an A- �U, U- �A, or G- �C base pair;K-�M’ is a G- �C or U- �A base pair (16). Class 2: Boldsequences represent the highly conserved 10 nucle-otides that characterize class 2 molecules; base-pairing is with the 5’ fixed end of the molecule.

FIGURE 18 (SEQ ID NO:�58) shows the sequenceand predicted secondary structure of trunc2 (A). Alsoshown (B) is a gel demonstrating the specificity oftrunc2 photocrosslinking to ARM proteins.

FIGURE 19 shows the sequence and predicted sec-ondary structure of trunc24 (SEQ ID NO: �59) (A). Alsoshown (B) is a gel demonstrating the specificity oflaser independent crosslinking to ARM proteins.

FIGURE 20 shows the trunc24 (SEQ ID NO:�59) pho-to-�independent crosslinking with HIV- �1 Rev in thepresence of human nuclear extracts.

FIGURE 21 illustrates the cyclical relationship be-tween SELEX steps. A single-�stranded nucleic acidrepertoire of candidate oligonucleotides is generatedby established procedures on a nucleic acid synthe-sizer, and is amplified by PCR to generate a popu-lation of double-�stranded DNA molecules. Candi-date DNA or RNA molecules are generated throughasymmetric PCR or transcription, respectively, puri-

fied, and allowed to complex with a target molecule.This is followed by partitioning of bound and unboundnucleic acids, synthesis of cDNA, and PCR amplifi-cation to generate double-�stranded DNA.

FIGURE 22 illustrates one embodiment of the solu-tion SELEX method in which primer extension inhi-bition is used to create differentiable cDNA pools -an amplifiable high affinity oligonucleotide cDNApool and a non-�amplifiable low affinity oligonucle-otide cDNA pool. In this embodiment, the first cDNAextension is performed in the presence of chain ter-minating nucleotides such as ddG. After removal ofthe target molecule and dideoxynucleotides, the sec-ond cDNA extension is conducted in the presenceof four dNTPs. Full-�length cDNA is preferentially syn-thesized from the high affinity oligonucleotides andtherefore, the high affinity cDNA pool is amplified inthe subsequent PCR step.

FIGURE 23 illustrates the cyclic solution SELEXprocess for the embodiment shown in Figure 22.

FIGURE 24 illustrates one embodiment of the cyclicsolution SELEX process wherein partitioning be-tween oligonucleotides having high and low affinityto a target molecule is achieved by restriction en-zyme digestion. In this embodiment, the first cDNAextension is conducted with four dNTPs and resultsin cDNAs corresponding to the low affinity oligonu-cleotides. The target is then removed and a secondcDNA extension is conducted in the presence ofmodified nucleotides resistant to enzymatic cleav-age. The cDNA pools are then incubated with restric-tion enzyme and only the cDNA pool correspondingto high affinity oligonucleotides is amplifiable in thesubsequent PCR step.

FIGURE 25 illustrates one embodiment of the cyclicsolution SELEX process wherein partitioning be-tween oligonucleotides having high and low affinityto a target molecule is achieved by affinity chroma-tography. The first cDNA extension is performed inthe presence of a modified nucleotide such as a bi-otinylated nucleotide, which allows the cDNA poolcorresponding to the low-�affinity oligonucleotide tobe retained on an affinity column.

FIGURE 26 illustrates one embodiment of the solu-tion SELEX process wherein partitioning betweenoligonucleotides having high and low affinity to a tar-get molecule is achieved by exonuclease inhibitionand results in formation of a double-�stranded nucleicacid population with high affinity for the target mol-ecule.

FIGURE 27 illustrates one embodiment of the solu-tion SELEX process wherein catalytic nucleic acids

11 12

EP 0 736 105 B1

9

5

10

15

20

25

30

35

40

45

50

55

are selected and isolated.

DETAILED DESCRIPTION OF THE INVENTION

�[0038] The present invention includes a variation ofthe SELEX method for selecting nucleic acid ligands. Themethod of one embodiment of the present invention,termed covalent SELEX or photoSELEX, identifies andselects nucleic acid ligands capable of binding and/orphotocrosslinking to target molecules.�[0039] This application also presents a method for im-proved partitioning of nucleic acid ligands identifiedthrough the SELEX method.�[0040] In its most basic form, the SELEX process maybe defined by the following series of steps:�

1) A candidate mixture of nucleic acids of differingsequence is prepared. The candidate mixture gen-erally includes regions of fixed sequences (i.e., eachof the members of the candidate mixture containsthe same sequences in the same location) and re-gions of randomized sequences. The fixed se-quence regions are selected either: a) to assist inthe amplification steps described below; b) to mimica sequence known to bind to the target; or c) to en-hance the potential of a given structural arrangementof the nucleic acids in the candidate mixture. Therandomized sequences can be totally randomized(i.e., the probability of finding a base at any positionbeing one in four) or only partially randomized (e.g.,the probability of finding a base at any location canbe selected at any level between 0 and 100 percent).2) The candidate mixture is contacted with the se-lected target under conditions favorable for bindingbetween the target and members of the candidatemixture. Under these circumstances, the interactionbetween the target and the nucleic acids of the can-didate mixture can be considered as forming nucleicacid- �target pairs between the target and the nucleicacids having the strongest affinity for the target.3) The nucleic acids with the highest affinity for thetarget are partitioned from those nucleic acids withlesser affinity to the target. Because only an extreme-ly small number of sequences (and possibly only onemolecule of nucleic acid) corresponding to the high-est affinity nucleic acids exist in the candidate mix-ture, it is generally desirable to set the partitioningcriteria so that a significant amount of the nucleicacids in the candidate mixture (approximately5-10%) is retained during partitioning.4) Those nucleic acids selected during partitioningas having the relatively higher affinity to the targetare then amplified to create a new candidate mixturethat is enriched in nucleic acids having a relativelyhigher affinity for the target.5) By repeating the partitioning and amplifying stepsabove, the newly formed candidate mixture containsfewer and fewer unique sequences, and the average

degree of affinity of the nucleic acid mixture to thetarget will generally increase. Taken to its extreme,the SELEX process will yield a candidate mixturecontaining one or a small number of unique nucleicacids representing those nucleic acids from the orig-inal candidate mixture having the highest affinity tothe target molecule.

�[0041] The SELEX Patent Applications describe andelaborate on this process in great detail. Included aretargets that can be used in the process; methods for thepreparation of the initial candidate mixture; methods forpartitioning nucleic acids within a candidate mixture; andmethods for amplifying partitioned nucleic acids to gen-erate enriched candidate mixtures. The SELEX PatentApplications also describe ligand solutions obtained to anumber of target species, including both protein targetswherein the protein is and is not a nucleic acid bindingprotein.�[0042] Partitioning means any process whereby lig-ands bound to target molecules can be separated fromnucleic acids not bound to target molecules. More broad-ly stated, partitioning allows for the separation of all thenucleic acids in a candidate mixture into at least two poolsbased on their relative affinity to the target molecule. � Par-titioning can be accomplished by various methods knownin the art. Nucleic acid-�protein pairs can be bound to ni-trocellulose filters while unbound nucleic acids are not.Columns which specifically retain nucleic acid- �targetcomplexes can be used for partitioning. For example,oligonucleotides able to associate with a target moleculebound on a column allow use of column chromatographyfor separating and isolating the highest affinity nucleicacid ligands. Liquid- �liquid partitioning can be used as wellas filtration gel retardation, and density gradient centrif-ugation.

I. PhotoSELEX.

�[0043] The present invention encompasses nucleicacid ligands which bind, photocrosslink and/or photoin-activate target molecules. Binding as referred to hereingenerally refers to the formation of a covalent bond be-tween the ligand and the target, although such bindingis not necessarily irreversible. Certain nucleic acid lig-ands of the present invention contain photoreactivegroups which are capable of photocrosslinking to the tar-get molecule upon irradiation with light. Additional nucleicacid ligands of the present invention are capable of bondformation with the target in the absence of irradiation.�[0044] In one embodiment, the present invention en-compasses nucleic acid ligands which are single- or dou-ble- �stranded RNA or DNA oligonucleotides. The nucleicacid ligands of the present invention may contain photo-reactive groups capable of crosslinking to the selectedtarget molecule when irradiated with light. Further, thepresent invention encompasses nucleic acid ligands con-taining any modification thereof. Reference to a photo-

13 14

EP 0 736 105 B1

10

5

10

15

20

25

30

35

40

45

50

55

reactive group herein may simply refer to a relatively sim-ple modification to a natural nucleic acid residue that con-fers increased reactivity or photoreactivity to the nucleicacid residue. Such modifications include, but are not lim-ited to, modifications at cytosine exocyclic amines, sub-stitution with halogenated groups, e.g., 5’-�bromo- or 5’-iodo-�uracyl, modification at the 2’-�position, e.g., 2’- �amino(2’-�NH2) and 2’- �fluoro (2’- �F), backbone modifications,methylations, unusual base- �pairing combinations andthe like. For example, the nucleic acid ligands of thepresent invention may include modified nucleotides suchas 2’- �NH2-iodouracil, 2’-�NH2-iodocytosine, 2’-�NH2-io-doadenine, 2’-�NH2-bromouracil, 2’-�NH2-bromocytosine,and 2’-�NH2-bromoadenine.�[0045] In one embodiment of the photoSELEX methodof the present invention, a randomized set of nucleic acidsequences containing photoreactive groups, termed thecandidate mixture, is mixed with a quantity of the targetmolecule and allowed to establish an equilibrium bindingwith the target molecule. The nucleic acid-�target mole-cule mixture is then irradiated with light until photo-crosslinking is complete as indicated by polyacrylamidegel electrophoresis. Only some of those nucleic acidsbinding tightly to the target molecules will efficientlycrosslink with the target.�[0046] The candidate mixture of the present inventionis comprised of nucleic acid sequences with randomizedregions including chemically reactive or a photoreactivegroup or groups. Preferably the reactive groups are mod-ified nucleic acids. The nucleic acids of the candidatemixture preferably include a randomized sequence por-tion as well as conserved sequences necessary for effi-cient amplification. The variable sequence portion maycontain fully or partially random sequence; it may alsocontain subportions of conserved sequence incorporatedwithin the randomized sequence regions.�[0047] Preferably, each oligonucleotide member of thecandidate mixture contains at least one chemically reac-tive or photoreactive group. Further, each oligonucle-otide member of the candidate mixture may be partiallyor fully substituted at each position by modified nucle-otides containing reactive groups. The candidate mixturemay also be comprised of oligonucleotides containingmore than one type of reactive group.�[0048] The target molecules bound and/or photo-crosslinked by the nucleic acid ligands of the presentinvention are commonly proteins and are selected basedupon their role in disease and/or toxicity. Examples areenzymes for which an inhibitor is desired or proteins forwhich detection is desired. However, the target moleculemay be any compound of interest for which a ligand isdesired. A target molecule can be a protein, peptide, car-bohydrate, polysaccharide, glycoprotein, hormone, re-ceptor, antigen, antibody, virus, substrate, metabolite,transition state analog, cofactor, inhibitor, drug, dye, nu-trient, growth factor, etc., without limitation.�[0049] A photoreactive group for the purpose of thisapplication is preferably a modified nucleotide that con-

tains a photochromophore, and that is capable of photo-crosslinking with a target species. Although referred toherein as photoreactive groups, in some cases as de-scribed below, irradiation is not necessary for covalentbinding to occur between the nucleic acid ligand and thetarget. Preferentially, the photoreactive group will absorblight in a spectrum of the wavelength that is not absorbedby the target or the non-�modified portions of the oligonu-cleotide. This invention encompasses, but is not limitedto, oligonucleotides containing a photoreactive group se-lected from the following: 5- �bromouracil (BrU), 5-�iodou-racil (IU), 5-�bromovinyluracil, 5- �iodovinyluracil, 5-�azidou-racil, 4-�thiouracil, 5- �bromocytosine, � 5- �iodocytosine, 5-bromovinylcytosine, 5-�iodovinylcytosine, 5- �azidocyto-sine, 8-�azidoadenine, 8-�bromoadenine, 8- �iodoadenine,8-�azidoguanine, 8-�bromoguanine, 8-�iodoguanine, 8-�azi-dohypoxanthine, 8-�bromohypoxanthine, 8-�iodohypox-anthine, 8-�azidoxanthine, 8-�bromoxanthine, 8-�iodoxan-thine, 5- �bromodeoxyuridine, 8-�bromo-�2’-�deoxyadenine,5-�iodo-�2’- �deoxyuracil, 5-�iodo-�2’-�deoxycytosine, 5-[(4-azidophenacyl)�thio] �cytosine, 5-[(4-�azidophenacyl)�thio]uracil, 7- �deaza-�7-�iodoadenine, 7-�deaza-�7-�iodoguanine,7-�deaza-�7-�bromoadenine, and 7- �deaza- �7-�bromogua-nine (Fig. 5). In one embodiment, the photoreactivegroups are 5-�bromouracil (BrU) and 5- �iodouracil (IU).�[0050] The photoreactive groups of the present inven-tion are capable of forming bonds with the target speciesupon irradiation of an associated nucleic acid target pair.The associated pair is referred to herein as a nucleopro-tein complex, and in some caes irradiation is not requiredfor bond formation to occur. The photocrosslink that typ-ically occurs will be the formation of a covalent bond be-tween the associated nucleic acid and the target. How-ever, a tight ionic interaction between the nucleic acidand target may also occur upon irradiation.�[0051] In one embodiment, photocrosslinking occursdue to electromagnetic irradiation. Electromagnetic irra-diation includes ultraviolet light, visible light, X-�ray andgamma ray. 5-�Halo substituted deoxyuracils and deoxy-cytosines are known to sensitize cells to ionizing radiation(Szybalski (1974) Cancer Chemother. Rep. 58:�539).�[0052] Crosslinking experiments have shown that aprecise juxtaposition of either IU or BrU and a tyrosine,tryptophan, or histidine is required for a high yieldcrosslinking to occur. The present invention takes advan-tage of this finding with selection for crosslinking mole-cules with randomly incorporated photoreactive groups.In one embodiment of the present invention, the photo-reactive groups 5- �bromouracil (BrU) or 5-�iodouracil (IU)are incorporated into RNA by T7 polymerase transcrip-tion with the 5-�halouridine triphosphate present in placeof uridine triphosphate. Incorporation is achieved by us-ing a mixture of 5-�halouridine triphosphate and uridinetriphosphate or all 5-�halouridine triphosphate. A rand-omized set of 32P or 33P-�labeled or unlabeled RNA se-quences is obtained from a randomized set of DNA tem-plates, synthesized using standard methodology.�[0053] The randomized set of RNA oligonucleotides

15 16

EP 0 736 105 B1

11

5

10

15

20

25

30

35

40

45

50

55

containing BrU or IU are mixed with a quantity of a targetprotein. The photoreactive chromophore is incorporatedrandomly into RNA as BrU or IU in place of uracil usingstandard methodology. The RNA-�target protein mixtureis irradiated with near ultraviolet light in the range of 300to 325 nm until photocrosslinking is complete. Only thosephotoreactive groups adjacent to a reactive amino acidresidue in a nucleoprotein complex form a covalent bondto the protein. Excited BrU or IU, returns to the groundstate unless it is near a reactive functional group suchas an oxidizable amino acid residue. Amino acid residueswhich have been established as being reactive with thelowest triplet state of 5-�bromouracil include tyrosine, tryp-tophan, histidine, and cystine (see Fig. 3). Others of po-tential reactivity based upon mechanistic studies are phe-nylalanine, methionine, cysteine, lysine, and arginine.�[0054] Nucleoprotein complexes which do not formcrosslinks may be easily disrupted by adjusting the re-action medium such as by denaturing with heat and/orsalt. Nucleic acids covalently bound to the protein maybe separated from free nucleic acids on a nitrocellulosefilter or by other partitioning methods known to thoseskilled in the art. Alternate methods for separating nucleicacids covalently bound to targets from free nucleic acidsinclude gel electrophoresis followed by electroelution,precipitation, differential solubility, and chromatography.To one skilled in the art, the method of choice will dependat least in part on the target molecule of interest. Thecrosslinked nucleic acids are released from the nitrocel-lulose filter by digestion of the protein material with en-zymes such as Proteinase K. At this point 5- �halouracilgroups which have photocrosslinked to the target proteinare bound to a single amino acid or to a short peptide.The read-�through ability of reverse transcriptase is noteffected by incorporation of a substituent at the 5- �positionof uracil because reverse transcriptase (RT) does notdifferentiate the 5-�position of uracil from that of thymine.Derivatization of the 5-�position has been used to incor-porate groups as large as biotin into RNA molecules. Inone embodiment of the present invention, the target isremoved from the selected photocrosslinked nucleic acidby photo or chemical dissociation.�[0055] Complementary nucleic acid copies of the se-lected RNA sequences are prepared with an appropriateprimer. The cDNA is amplified with a DNA polymeraseand a second primer. 5-�Halo- �2’- �deoxyuracil is not em-ployed in the DNA synthesis and amplification steps. Theamplified DNAs are then transcribed into RNA sequenc-es using 5- �halouridine triphosphate in place of uridinetriphosphate in the same or different ratio of 5-�halouridineto uridine in the candidate mixture.�[0056] For the subsequent round of photoSELEX, thepartially selected RNA sequences are again allowed tocomplex with a quantity of the target protein.� Subse-quently, the nucleoprotein complexes are irradiated inthe region of 300-325 nm. RNA sequences which havecrosslinked to protein are again separated from RNA se-quences which have not crosslinked. cDNAs are pre-

pared and amplified and a third set of RNA sequencescontaining 5-�halouracil are prepared. The cycle is con-tinued until it converges to one or several RNA ligandswhich bind with high affinity and photocrosslink to thetarget protein. Shortening of the irradiation time in latercycles can further enhance the selection. The cDNAs ofthe selected RNA ligands are amplified, gel purified, andsequenced. Alternatively, the RNA sequences can besequenced directly. The selected RNA sequences arethen transcribed from the appropriate synthesized DNAtemplate, again using 5-�halouridine triphosphate in placeof uridine triphosphate (Example 11).�[0057] In another embodiment of the present inven-tion, photoSELEX is performed on oligonucleotide se-quences preselected for their ability to bind the targetmolecule (Example 12). SELEX is initially performed witholigonucleotides which do not contain photoreactivegroups. The RNA ligand is transformed into a photore-active ligand by substitution of photoreactive nucleic acidnucleotides at specific sites in the ligand. The photo-chemically active permutations of the initial ligand maybe developed through a number of approaches, such asspecific substitution or partial random incorporation ofthe photoreactive nucleotides. Specific substitution in-volves the synthesis of a variety of oligonucleotides withthe position of the photoreactive nucleotide changedmanually. The location of the substitution is directedbased upon the available data on binding of the ligandto the target molecule. For example, substitutions aremade to the initial ligand in areas of the molecule thatare known to interact with the target molecule. For sub-sequent selection rounds, the photoSELEX method isused to select for the ability to crosslink to the target mol-ecule upon irradiation.�[0058] In another embodiment of the present inven-tion, nucleic acid ligands are selected by photoSELEXfollowed by SELEX. PhotoSELEX is performed initiallywith oligonucleotide sequences containing photoreactivegroups. Sequences selected for their ability to crosslinkto the target molecule are then selected for ability to bindthe target molecule through the SELEX method (Exam-ple 13).�[0059] In another embodiment of the present inven-tion, a limited selection of oligonucleotides through SE-LEX is followed by selection through photoSELEX (Ex-ample 14). The initial SELEX selection rounds are con-ducted with oligonucleotides containing photoreactivegroups. After a number of SELEX rounds, photoSELEXis conducted to select oligonucleotides capable of bind-ing the target molecule.�[0060] In yet another embodiment of the present in-vention, nucleic acid ligands identified through SELEXare subjected to limited randomization, followed by se-lection through photoSELEX (Example 15). SELEX isfirst carried out to completion with nucleic acid sequencesnot containing photoreactive groups. The sequence ofthe nucleic acid ligand is used to generate a family ofoligonucleotides through limited randomization. Pho-

17 18

EP 0 736 105 B1

12

5

10

15

20

25

30

35

40

45

50

55

toSELEX is subsequently performed to select a nucleicacid ligand capable of photocrosslinking to the target mol-ecule.�[0061] In another embodiment of the present inven-tion, photoSELEX is used to identify a nucleic acid ligandcapable of modifying the biological activity of a targetmolecule (Example 16).�[0062] In a further embodiment of the present inven-tion,� the photoSELEX methodology is applied diagnosti-cally to identify unique proteins associated with specificdisease states (Example 17). In yet another embodimentof the present invention, nucleic acid ligands capable ofcrosslinking a target molecule associated with a specificdisease condition are used in vivo to crosslink to the tar-get molecule as a method of treating the disease condi-tion (Examples 18 and 19).�[0063] In one embodiment of the present invention,RNA ligands identified by photoSELEX are used to detectthe presence of the target protein by binding to the proteinand then photocrosslinking to the protein. Detection maybe achieved by incorporating 32P or 33P-�labels and de-tecting material which is retained by a nitrocellulose filterby scintillation counting or detecting material which mi-grates correctly on an electrophoretic gel with photo-graphic film. Alternatively, photoSELEX creates a fluo-rescent chromophore which is detected by fluorescenceemission spectroscopy. Fluorescence emission for theproducts of reaction of 5- �bromouracil to model peptides(as shown in Fig. 3) has been reported by Dietz and Koch(1987) supra. In another embodiment of the invention, afluorescent label is covalently bound to the RNA and de-tected by fluorescence emission spectroscopy. In anoth-er embodiment of the invention, RNA ligands selectedthrough photoSELEX are used to inhibit the target proteinthrough the same process. In yet another embodiment,the photoselected ligand is bound to a support and usedto covalently trap a target.�[0064] In a one embodiment of the invention, 5-�iodou-racil is incorporated into the RNA sequences of the can-didate mixture, and light in the range of 320-325 nm isused for irradiation. This combination assures regionse-lective photocrosslinking of the 5- �halouracil chromo-phore to the target protein without other non-�specific pho-toreactions.� In particular, tryptophan residues of proteinand thymine and uracil bases of nucleic acids are knownto be photoreactive. As shown in Figure 4, 5-�iodouracilabsorbs at 325 nm but tryptophan and the standard nu-cleic acid bases do not. The molar extinction coefficientfor 5- �iodouracil at 325 nm is 163 L/mol·cm. Monochro-matic light in the region of 320-325 nm is preferably sup-plied by a frequency doubled tunable dye laser emittingat 320 nm or by a helium cadmium laser emitting at 325nm.�[0065] In one embodiment of the invention a xenonchloride (XeCl) excimer laser emitting at 308 nm is em-ployed for the photocrosslinking of 5-�iodouracil-�bearingRNA sequences to a target protein. With this laser, a highyield of photocrosslinking of nucleoprotein complexes is

achieved within a few minutes of irradiation time.�[0066] In another embodiment of the invention, photo-crosslinking of 5-�iodouracil-�bearing RNA sequences toa target protein is achieved with wavelength filtered 313nm high pressure mercury lamp emission or with lowpressure mercury lamp emission at 254 nm absorbed bya phosphor and re- �emitted in the region of 300-325 nm.The latter emission is also carefully wavelength filteredto remove 254 nm light not absorbed by the phosphorand light in the region of 290-305 nm which could damagethe protein.�[0067] In a further embodiment of the invention, photo-crosslinking of BrU- or IU-�bearing RNA sequences to atarget protein is achieved with light in the region of350-400 nm which populates directly the triplet state fromthe ground state. Monochromatic light from the third har-monic of a Neodymium YAG laser at 355 nm or the firstharmonic from a xenon fluoride (XeF) excimer laser at351 nm may be particularly useful in this regard.�[0068] In yet another embodiment of the invention thephotoreactive nucleotides are incorporated into singlestranded DNAs and amplified directly with or without thephotoreactive nucleotide triphosphate.

A. Covalent SELEX and Nucleic Acid Ligands That Bind to HIV- �1 Rev Protein With and Without Irradiation.

�[0069] The target protein chosen to illustrate photo-SELEX is Rev from HIV- �1. Rev’s activity in vivo is derivedfrom its association with the Rev-responsive element(RRE), a highly structured region in the HIV-�1 viral RNA.Previous RNA SELEX experiments of Rev have allowedthe isolation of very high affinity RNA ligands. The highestaffinity ligand, known as "6a," (SEQ ID NO:�5) has a Kdof approximately 1nM and is shown in Figure 16. Thesecondary structure of 6a, and its interaction with Rev,have been well characterized.�[0070] The construction of the nucleic acid library forphoto-�SELEX was based upon the Rev 6a sequence(SEQ ID NO:�5). During the synthesis of the deoxyoligo-nucleotide templates for SELEX, the random region ofthe template was substituted by a "biased randomization"region, in which the ratio of the four input bases wasbiased in favor of the corresponding base in the 6a se-quence. (Actual ratios were 62.5:�12.5:�12.5: �12.5.) For ex-ample, if the 6a base for a particular position is G, thenthe base input mixture for this synthesis step is 62.5%G, and 12.5% of the other three bases.�[0071] The photoreactive uracil analogue 5-�iodouracil(iU), which has been used to generate high-�yield, region-specific crosslinks between singly-�substituted iU nucleicacids and protein targets (Willis et al. (1993) Science262: �1255) was used for this example. The iU chromo-phore is reactive under long- �wavelength ultraviolet radi-ation, and may photocouple to the aromatic amino acidsof protein targets by the same mechanism as 5-�bromour-acil (Dietz et al. (1987) J. Am. Chem. Soc. 109:�1793). Asdiscussed above, the target for this study is the HIV-�1

19 20

EP 0 736 105 B1

13

5

10

15

20

25

30

35

40

45

50

55

Rev protein, which is necessary for productive infectionof the virus (Feinberg et al. (1986) Cell 46:�807) and theexpression of the viral structural genes gag, pol and env(Emerman et al. (1989) Cell 57: �1155). The interaction ofRev with high affinity RNA ligands is well characterized.A single, high-�affinity site within the RRE (the IIB stem)has been identified (Heaphy et al. (1991) Proc. Natl.Acad. Sci. USA 88:�7366). In vitro genetic selection ex-periments have generated RNA ligands that bind withhigh affinity to Rev and have helped determine the RNAstructural elements necessary for Rev: �RNA interactions(Bartel et al. (1991) Cell 67:�529; Tuerk et al., In thePolymerase Chain Reaction (1993); Jensen et al. (1994)J. Mol. Biol. 235: �237).�[0072] A "biased randomization" DNA oligonucleotidelibrary, based upon the high affinity Rev ligand sequence6a (SEQ ID NO:�5), contains approximately 1014 uniquesequences. This template was used for in vitro T7 tran-scription with 5-�iUTP to generate fully-�substituted iU RNAfor selection. The photo-�SELEX procedure alternated be-tween affinity selection for Rev using nitrocellulose par-titioning and monochromatic UV irradiation of the nucle-oprotein complexes with denaturing polyacrylamide gelpartitioning of the crosslinked complexes away from non-crosslinked RNA sequences. The final procedure utilizeda simultaneous selection for affinity and crosslinking us-ing competitor tRNA. Each round constitutes a selectionfollowed by the conversion of recovered RNA to cDNA,polymerase chain reaction (PCR) amplification of theDNA, and in vitro transcription to generate a new pool ofiU-�RNA. To amplify RNA’s recovered as covalent nucle-oprotein complexes, the appropriate gel slice was isolat-ed and proteinase K treated.�[0073] The RNA pool was first subjected to threerounds of affinity selection with Rev protein, with parti-tioning of the higher affinity sequences by nitrocellulosefilters. Next, the evolving RNA pool was subjected to UVlaser irradiation in the presence of excess Rev protein toallow those RNA sequences with the ability to crosslinkwith the protein to do so. Crosslinked RNA sequenceswere then partitioned using polyacrylamide gel electro-phoresis (PAGE). These crosslinked RNAs were recov-ered from the gel material, the linked Rev protein digestedaway, and the RNAs used for cDNA synthesis and furtheramplification for the next round of photo- �SELEX. A308nm XeCl excimer laser was used for the first roundof photocrosslinking; thereafter, a 325nm HeCd laserwas employed.�[0074] Following four rounds of selection for laser-�in-duced crosslinking, the RNA pool was again put throughthree rounds of affinity selection. Finally, the RNA poolwas selected simultaneously for its ability to bind Revwith high affinity and to crosslink to the protein. This wasaccomplished by using high concentrations of a non-�spe-cific nucleic acid competitor in the photocrosslinking re-action.�[0075] Crosslinked product increased approximately30-�fold from the starting pool to round 13 (Fig. 15). Under

these conditions, the greatest increase in crosslinking iscorrelated with the greatest increase in affinity - fromround 7 to round 10.�[0076] After 13 rounds of selection, the PCR productswere cloned and 52 isolates sequenced (Fig. 16, SEQID NOS: �5-55). Class 1 molecules, which comprise 77%of the total sequences, contain a very highly conservedmotif, 5’KDAACAN ...N’UGUUH’M’3’ (SEQ ID NO:�56)(Fig. 17). Computer folding algorithms predict that thisconserved motif is base-�paired and lies in a stem-�loopstructure.� Subclasses a-�d (Fig. 16, SEQ ID NOS:�5-43)illustrate different strategies utilized in the "biased rand-omization" pool to obtain the consensus motif. Class 2molecules show a highly conserved 10- �base sequence(Fig. 17, SEQ ID NO. 57), which is predicted to fold withthe 5’ fixed region of the RNA and forms a structure dis-tinct from either the class 1 or the 6a (SEQ ID NO:�5)motif. All class 1 sequences exhibit biphasic binding toRev, with high affinity dissociation constants (Kds) rang-ing from 1-10nM. Class 2 sequences show monophasicbinding to Rev with Kds approximately of 30-50 nM. Anal-ysis of round 13 sequences reveal that the frequency ofthe consensus motifs for class 1 and class 2 populationswas very small in the starting pool, and some individualsequences arose only through the mutational pressuresof the photo-�SELEX procedure.�[0077] Cross- �linking behavior differs between the twoclasses. Under high Rev concentrations (500 nM), anda 4 min. of 325 nm irradiation, class 2 molecules producegreater crosslink yield and efficiency than class 1 mole-cules (data not shown); presumably, this behavior allowsthe class 2 molecules, with relatively low affinity for Rev,to compete under the photo- �SELEX procedures. Forclass 1 molecules, longer irradiation times will producehigher molecular weight crosslinked species. Althoughnot bound by theory, it is proposed that the RNAs, whichcontain both an evolved binding domain for Rev, and thefixed regions needed for amplification in SELEX, are ableto bind more than one Rev molecule per RNA. Sinceeach RNA contains on average 21 iU bases (RNA length- 86 bases), it is thought that there is a certain promiscuityof the photoreaction that allows crosslinking of a singleRNA to more than one Rev molecule at high protein con-centrations. Class 2 molecules produce fewer high mo-lecular weight species upon photocrosslinking; they are,on average, iU poor and may contain structures whichdo not allow binding/�crosslinking to additional Rev mol-ecules.�[0078] Analysis of individual round 13 RNAs revealedthat a subpopulation could crosslink to Rev without laserirradiation. Thus, the single set of experiments demon-strated that both covalent SELEX without irradiation andphotoSELEX with irradiation can be found in the samesystem. 4 of 15 round 13 sequences analyzed crosslinkwithout laser irradiation (Fig. 16). From these few se-quences, it was not readily possible to identify a se-quence motif that confers laser independent crosslinking,although all molecules considered to date belong to the

21 22

EP 0 736 105 B1

14

5

10

15

20

25

30

35

40

45

50

55

1a subclass.�[0079] To further investigate laser-�dependent and la-ser-�independent crosslinking (LD- �XL and LI-�XL, respec-tively) and avoid the secondary photoproducts formedwith full-�length class 1 molecules, several small RNAscontaining only the evolved sequences were construct-ed. Trunc2 and trunc24 (Figs. 18 and 19) (SEQ ID NOS:58 AND 59) are based upon clones #3 and #24, respec-tively, and show monophasic binding to Rev with Kds of0.5 nM (trunc2) and 20 nM (trunc24). Trunc2 (Fig. 18)exhibits Ld-�XL behavior, and trunc24 (Fig. 19) is capableof both LI-�XL and LD-�XL.�[0080] To explore the conformation and chemical re-quirements for LD-�XL and LI-�XL, crosslinking reactionswere performed with trunc2 (SEQ ID NO: �58) and trunc24(SEQ ID NO:�59) and several Arginine Rich Motif (ARM)proteins. The class of RNA-�binding proteins includes thetarget protein, HIV- �1 Rev, and also HIV-�1 Tat and thehighly similar HIV-�2 Rev. LD-�XL reactions with trunc2(Fig. 18, SEQ ID NO:�58) show that trunc2 is capable ofcrosslinking specifically to both HIV-�1 and HIV- �2 Rev pro-teins, but not HIV-�1 Tat. The two slightly different migrat-ing nucleoprotein complexes probably represent the abil-ity of trunc2 to use one of two iU nucleotides to crosslinkthe Rev proteins. Although not bound by theory, it is pro-posed that a tryptophan residue present in the highly sim-ilar ARMs of both Rev proteins is the amino acid neces-sary for the specific photo-�crosslinking of our high-�affinityRNA ligands.�[0081] Trunc24 LI-�XL (Fig. 19, SEQ ID NO:�59), per-formed with the same proteins, shows crosslinking onlyto HIV- �1 Rev. Like trunc3, trunc24 can photo- �crosslinkto HIV-�2 Rev (data not shown). It was also observed thatthis LI- �crosslink is reversible under highly denaturingconditions, or with high concentration of nucleic acidcompetitors. Although not bound by theory, these obser-vations lead to the postulation that LI-�XL proceeds by aMichael adduct between the 6 position of an IU and acystein residue, or possible a 5 position substitution re-action. This postulation is consistent both with the obser-vation that iU undergoes Michael adduct formation morereadily than U, and the fact that HIV-�1 Rev contains threecysteines, while HIV-�2 Rev contains none.�[0082] To test for the ability of trunc24 (SEQ ID NO:59) to discriminate HIV- �1 Rev in a complex mixture,trunc24 and 10 Pg of human fibroblast nuclear extracttogether with decreasing amounts of HIV-�1 Rev (Fig. 20).At 50 nM Rev and a 1:�100 weight ratio of Rev to nuclearextract, it was possible to see a very significantcrosslinked product between trunc24 and Rev. Nuclearextracts and trunc24 alone resulted in no crosslinkedproducts.�[0083] Example 1 describes the synthesis of hairpinRNA oligonucleotides RNA- �1 (SEQ ID NO:�1), RNA-�2(SEQ ID NO: �2), and RNA-�3 (SEQ ID NO:�3) using 5-�br-omouridine triphosphate, 5-�iodouridine triphosphate anduridine triphosphate, respectively. Experiments deter-mining the RNA-�protein binding curves for RNA-�1 (SEQ

ID NO: �1), RNA- �2 (SEQ ID NO:�2), and RNA-�3 (SEQ IDNO:�3) to the bacteriophage R17 coat protein are de-scribed in Example 2. Example 3 describes the photo-crosslinking of the RNA oligonucleotides to the R17 coatprotein. The amino acid residue of the R17 coat proteinphotocrosslinked by RNA-�1 (SEQ ID NO:�1) after illumi-nation via xenon chloride (XeCl) excimer laser at 308 nmis described in Example 4. Example 5 describes thephotocrosslinking of iodouracil-�substituted RNA-�2 (SEQID NO:�2) to the R17 coat protein by monochromatic emis-sion at 325 nm. Example 6 describes the photocrosslink-ing of RNA-�1 (SEQ ID NO: �1) and RNA-�2 (SEQ ID NO:�2)to the R17 coat protein achieved after broad-�band emis-sion illumination with a transilluminator. Example 7 de-scribes the photoreaction of 5-�iodouracil with N- �acetyl-tyrosine N-�ethyl amide, which appeared to yield a photo-crosslink similar to that achieved with 5-�bromouracil- �sub-stituted nucleic acids to associated proteins. The prepa-ration of a cDNA from a RNA photocrosslinked to theR17 coat protein is described in Example 8. Example 9describes the photocrosslinking of an IC-�substitutedRNA ligand to the R17 coat protein.�[0084] Example 10 describes the incorporation of hal-ogenated nucleotides into DNA. Examples 11-15 de-scribes photoSELEX protocols which may be used to pro-duce specific photoreactive nucleic acid ligands. Exam-ple 11 describes a continuous photoSELEX method. Ex-ample 12 describes a method in which nucleic acid lig-ands initially selected through SELEX are subsequentlyselected through photoSELEX for the capacity tocrosslink to the target molecule. Example 13 describesone embodiment in which nucleic acid ligands identifiedthrough photoSELEX are then subjected to selectionthrough SELEX and selected for ability to bind the targetmolecule. Example 14 describes another embodimentwherein a limited SELEX selection is followed by selec-tion through photoSELEX. Example 15 describes an em-bodiment of the present invention in which nucleic acidligands identified through SELEX are subjected to limitedrandomization, followed by selection through photoSE-LEX. Example 16 describes a method for selecting a nu-cleic acid ligand capable of modifying the biological ac-tivity of a target molecule. Example 17 describes a diag-nostic procedure which uses the SELEX and photoSE-LEX methods to identify proteins associated with specificdisease processes.�[0085] Example 18 describes a method for the in vivotreatment of disease through photoSELEX. A photoSE-LEX selected nucleic acid ligand able to bind andcrosslink to a target molecule associated with a diseasestate is introduced into a patient in a number of waysknown to the art. For example, the photoSELEX ligandmay be transiently or constitutively expressed in the ap-propriate cells of a patient with the disease. Alternatively,the photoSELEX ligand may be taken into a patient’scells as a double-�stranded DNA which is transcribed inthe cell in the presence of iodinated cytosine. Iodinatedcytosine may be administered to the patient, followed by

23 24

EP 0 736 105 B1

15

5

10

15

20

25

30

35

40

45

50

55

irradiation with X- �rays. IC incorporated into the nucleicacid ligand synthesized in the appropriate cells allowsthe ligand to crosslink and inactivate the target molecule.Further methods of introducing the photoSELEX ligandinto a patient include liposome delivery of the halogen-ated ligand into the patient’s cells.�[0086] Example 20 describes the production of modi-fied nucleic acid ligands the crosslink, with or withoutirradiation, to HIV- �1 Rev protein. Figure 15 shows theresults of crosslinking to the bulk candidate mixture atvarious rounds of SELEX. rd1-round 1 pool RNA;rd7-round 7 pool RNA; rd10-round 10 pool RNA;rd13-round 13 pool RNA; rd13/PK, photocrosslinkedround 13 pool RNA proteinase K treated (35ul of a 100ul reaction was incubated in 0.5% SDS, 50 ug/ml Protei-nase K and 1mM EDTA at 65 C. for one hour); rd13/noiU-round 13 pool RNA transcribed with UTP (no iU).R-free RNA; XL-crosslinked nucleoprotein complex.�[0087] Figure 16 shows the sequences sequenced af-ter 13 rounds of SELEX (SEQ ID NOS: �5-55). The se-quences are aligned for maximum homology to the 6asequence (SEQ ID NO:�5). Underlines represent potentialbase pairing as indicted by computer RNA folding algo-rithms. Dashed underlines represent the 6a ligand "bub-ble" motif. Sequences flanked by underline represent ei-ther loop or bulge regions. Dashes are placed to maxi-mize alignment with 6a. * denotes that two isolates wereobtained. + indicates laser independent crosslinking and- denotes the lack of laser independent crosslinking toHIV-�1 Rev. Figure 17 (SEQ ID NOS:�56-57) shows theconsensus for class 1 and class 2 ligands. Figures 18and 19 show the sequence of Trunc2 (SEQ ID NO:�58)and Trunc24 (SEQ ID NO:�59) and the specificity results.500 nM protein, 20Pg tRNA, and approximately 1 nM ofkinased trunc2 RNA were incubated for 10 min. at 37°Cand irradiated for 4 min at 325 nm. t2-trunc2 RNA irradi-ated without added protein; t2Rev1/O’ -trunc2 RNA, HIV-1 Rev protein and 0 min. of irradiation; t2/Rev1/4’ -trunc2RNA, HIV-�1 Rev protein, adn 4 min. of irradiation;t2/Rev1/4’/PK -trunc2 RNA, HIV- �1 Rev protein, 4 min. orirradiation, and proteinase K treated as in Fig. 1;t2/Rev2/4’ -trunc2 RNA, HIV-�2 Rev protein, and 4 min.of irradiation; t2/Tat/ �4’-trunc2 RNA, HIV-�1 Tat protein,and 4 min. of irradiation. R-free RNA; XL-crosslinked nu-cleoprotein complex. Figure 20 shows the trunc24 pho-toindependent crosslinking with HIV-�1 Rev in the pres-ence of human nuclear extract.

II. Solution SELEX.

�[0088] The following description of "Solution SELEX"does not form part of the invention claimed by this patent,but is useful for understanding the invention that isclaimed.�[0089] This disclosure presents several improvedmethods for partitioning between oligonucleotides hav-ing high and low affinity for a target molecule. The methodhas several advantages over prior art methods of parti-

tioning: (1) it allows the isolation of nucleic acid ligandsto the target without also isolating nucleic acid ligands tothe partitioning matrix; (2) it increases the speed and ac-curacy by which the oligonucleotide candidate mixture isscreened; and (3) the solution SELEX procedure can beaccomplished in a single test tube, thereby allowing thepartitioning step to be automated.�[0090] The materials and techniques required by themethod are commonly used in molecular biology labora-tories. They include the polymerase chain reaction(PCR), RNA or DNA transcription, second strand DNAsynthesis, and nuclease digestion. In practice, the tech-niques are related to one another in a cyclic manner asillustrated in Figure 21.�[0091] In the SELEX method, described by Tuerk andGold (1990) Science 249:�1155 and illustrated in Figure21, a single-�stranded nucleic acid candidate mixture isgenerated by established procedures on a nucleic acidsynthesizer, and is incubated with dNTP and Klenowfragment to generate a population of double-�strandedDNA templates. The double-�stranded DNA or the RNAtranscribed from them are purified, and contacted with atarget molecule. RNA sequences with enhanced affinityto the target molecule form nucleic acid-�target complex-es. This is followed by partitioning of bound and unboundnucleic acids, and separation of the target molecule fromthe bound nucleic acids. cDNA is synthesized from theenhanced affinity nucleic acids and double- �strandedDNA generated by PCR amplification. The cycle is re-peated until the complexity of the candidate mixture hasdecreased and its affinity as well as specificity to the tar-get has been maximized.�[0092] A novel feature of the solution SELEX methodis the means by which the bound and free members ofthe nucleic acid candidate mixture are partitioned. In oneembodiment of the method, generation of two physicallydistinct cDNA pools is accomplished by use of primerextension inhibition. One cDNA extension step is addedto the basic SELEX protocol between steps 2 and 3above, which allows the generation of two physically dis-tinct cDNA pools -- one having high affinity for the targetand one having low affinity for the target -- which areeasily distinguished and separated from each other.Primer extension inhibition analysis is a common tech-nique for examining site-�bound proteins complexed tonucleic acids (Hartz et al. (1988) Methods Enzymol- 164:419), and relies on the ability of high affinity complexesto inhibit cDNA synthesis. Examples of protein-�nucleicacid interactions studied by primer extension inhibitioninclude ribosome binding to the mRNA ribosome-�bindingsite (Hartz et al. (1988) Meth. Enzym. 164:�419) as wellas binding of the unique E. coli translation factor, SELBprotein, to the mRNA selenocysteine insertion sequence(Baron et al. (1993) Proc. Natl. Acad. Sci. USA 90:�4181).�[0093] In one embodiment of the solution SELEXscheme, the first cDNA extension is performed in thepresence of chain terminating nucleotide triphosphates.Under these conditions, oligonucleotides with low affinity

25 26

EP 0 736 105 B1

16

5

10

15

20

25

30

35

40

45

50

55

for the target which form fast dissociating complexes withthe target are converted into truncated cDNAs with a 3’-end terminated with a nonextendible nucleotide. Thetruncated cDNA chain is unable to anneal to the PCRprimers, and therefore, is non-�amplifiable. In contrast,tight complexes formed between high affinity oligonucle-otides and the target molecule, characterized by slowdissociation rates, inhibit cDNA extension- The chain ter-minators are not incorporated into the nascent cDNAchain synthesized from the high affinity oligonucleotidebecause cDNA synthesis is blocked by the tightly boundtarget molecule. Full length cDNA from the high affinitycomplexes are obtained during a second round of cDNAextension in which the target and chain terminators havebeen removed from the system. Thus, weak affinity com-plexes are converted into truncated cDNA lacking theprimer annealing site while tight complexes are convert-ed into full length cDNA and are amplified by PCR (Fig.22). The stringency of this method is easily modified byvarying the molar ratio of chain terminators and dNTPsor the concentration of the polymerase, as primer exten-sion inhibition is sensitive to polymerase concentration(Ringquist et al. (1993) Biochemistry in press). As usedin the present disclosure, the term "stringency" refers tothe amount of free RNA that will be converted into PCRproduct.�[0094] Therefore, one crucial feature is its ability to par-tition strong and weak affinity complexes into amplifiableand non-�amplifiable nucleic acid pools without requiringa partitioning matrix- It is the unique properties of thesecDNA pools that allow selective amplification of the highaffinity ligand.�[0095] The target molecule can be a protein (either nu-cleic acid or non-�nucleic acid binding protein), nucleicacid, a small molecule or a metal ion. The solution SELEXmethod allows resolution of enantiomers as well as theisolation of new catalytic nucleic acids.�[0096] Primer extension inhibition may be achievedwith the use of any of a number of nucleic acid polymer-ases, including DNA or RNA polymerases, reverse tran-scriptase, � and Qβ-�replicase.�[0097] The candidate mixture of nucleic acids includesany nucleic acid or nucleic acid derivative, from which acomplementary strand can be synthesized.�[0098] Prior art partitioning included use of nitrocellu-lose or an affinity column. One disadvantage of the priorart partitioning was the phenomenon of matrix binders inwhich nucleic acids that specifically bind the partitioningmatrix are selected along with those that specifically bindthe target. Thus, one advantage of the method is that itovercomes unwanted selective pressure originating withuse of a partitioning matrix by only using such matrixesafter nucleic acids with high affinity for the target havebeen partitioned in solution and amplified. Moreover, theability to partition strong and weak affinity complexes dur-ing cDNA synthesis, based on the ability of only thestrongest complexes to inhibit extension by a polymer-ase, results in the selection of only the highest affinity

nucleic acid ligands. It is estimated that complexes withdissociation constants in the nanomolar or less range willefficiently block cDNA synthesis. The method is expectedto preferentially screen nucleic acid candidate mixturesfor members that bind the target at this limit.�[0099] The use of primer extension inhibition allowspartitioning of the oligonucleotide candidate mixture intotwo pools - those oligonucleotides with high target affinity(amplifiable) and those with low target affinity (non- �am-plifiable). As described above, chain terminators may beused to poison the first extension product, rendering thelow affinity cDNAs non-�amplifiable.�[0100] In another embodiment of the method, restric-tion enzymes are used to selectively digest the cDNAgenerated from the low affinity nucleic acids. A numberof restriction enzymes have been identified that cleavesingle- �stranded DNA. These enzymes cleave at specificsequences but with varying efficiencies. Partitioning ofweak and strong affinity nucleic acids is accomplishedby primer extension in the presence of the four dNTPs,followed by removal of the target and a second extensionwith modified nucleotides that are resistant to enzymaticcleavage. The cDNA pools can then be incubated withthe appropriate restriction enzyme and the cDNA syn-thesized during the first extension cleaved to remove theprimer annealing site and yield a non-�amplifiable pool.Increased efficiency of cleavage is obtained using a hair-pin at the restriction site (RS) to create a localized double-stranded region (Fig. 24).�[0101] In another embodiment of the method, cDNAsequences corresponding to low affinity nucleic acids arerendered selectively degradable by incorporation of mod-ified nucleotide into the first cDNA extension productsuch that the resulting cDNA is preferentially degradedenzymatically or chemically.�[0102] In,�another embodiment of the method the firstextension product can be removed from the system byan affinity matrix. Alternatively, the matrix could be usedto bind the second extension product, e.g., the cDNAscorresponding to high affinity nucleic acids. This strategyrelies on the incorporation of modified nucleotides duringcDNA synthesis. For instance, the first cDNA extensioncould be performed in the presence of modified nucle-otides (e.g., biotinylated, iodinated, thiolabelled, or anyother modified nucleotide) that allow retention on an af-finity matrix (Fig- 25). In an alternate embodiment of themethod, a special sequence can also be incorporated forannealing to an affinity matrix. Thus,� first synthesis cD-NAs can be retarded on commercially obtainable matri-ces and separated from second synthesis cDNA, syn-thesized in the absence of the modified nucleotides andtarget.�[0103] In another embodiment, exonuclease hydroly-sis inhibition is used to generate a pool of high affinitydouble-�stranded nucleic acid ligands.�[0104] In yet another embodiment, the solution SELEXmethod is used to isolate catalytic nucleic acids.�[0105] In another embodiment, solution SELEX is

27 28

EP 0 736 105 B1

17

5

10

15

20

25

30

35

40

45

50

55

used to preferentially amplify single-�stranded nucleic ac-ids.�[0106] In a further embodiment, the solution SELEXmethod is automated.�[0107] Removal of the target to allow cDNA synthesisfrom the high affinity nucleic acids can also be accom-plished in a variety of ways. For instance, the target canbe removed by organic extraction or denatured by tem-perature, as well as by changes in the electrolyte contentof the solvent. In addition, the molecular repertoire of thecandidate mixture that can be used with the inventioninclude any from which a second complementary strandcan be synthesized. Single-�stranded DNA as well asRNA can be used, as can a variety of other modifiednucleotides and their derivatives.�[0108] The following non-�limiting examples 1-20 illus-trate the method of the present invention. Example 21describes the solution SELEX process wherein partition-ing between high and low affinity nucleic acids isachieved by primer extension inhibition. Example 22 il-lustrates the solution SELEX process wherein partition-ing is achieved by restriction enzyme digestion of lowaffinity RNA. Example 23 describes the solution SELEXprocess wherein low affinity nucleic acids are separatedfrom high affinity nucleic acids by affinity chromatogra-phy. Example 24 describes the isolation of high affinitydouble-�stranded nucleic acid ligands with the use of ex-onuclease inhibition. Example 25 describes the isolationof catalytic nucleic acids. Example 26 describes an au-tomated solution SELEX method.�[0109] Examples 1-20 are non-�limiting illustrations ofmethods of utilizing the present invention. Other methodsof using the invention will become apparent to thoseskilled in the art from the teachings of the present disclo-sure.

Example 1. Synthesis of RNA Sequences RNA-�1, RNA- �2, RNA-�3, and RNA- �7 and R17 Coat Protein.

�[0110] RNA-�1 (SEQ ID NO:�1), RNA-�2 (SEQ ID NO:�2),and RNA- �3 (SEQ ID NO:�3) shown in Figure 6 and RNA-7 (SEQ ID NO:�4) shown in Figure 12 were prepared byin vitro transcription from synthetic DNA templates orplasmids using methodology described by Milligan andco-�workers (Milligan et al. (1987) Nucleic Acids Res. 15:8783). Transcription reactions contained 40 mM tris�(hy-droxymethyl)�aminomethane hydrochloride (Tris- �HCl, pH8.1 at 37°C), 1 mM spermidine, 5 mM dithiothreitol (DTT),50 Pg/ml of bovine serum albumin (BSA), 0.1% (v/v) Tri-ton X- �100, 80 mg/ml of polyethylene glycol (mr 8000),and 0.1 mg/ml of T7 RNA polymerase. Larger quantitiesof RNA were prepared with 3-5 mM of each of the nucle-otide triphosphates (NTPs), 25 mM magnesium chloride,and 1 PM DNA template or 0.1 Pg/ml of plasmid. Body-labeled RNAs were prepared in 100 PM reactions with 1mM each of the three NTPs, 0.25 mM of the equivalentradiolabelled NTP ([α-32P] NTP, 5 PCi), 15 mM MgCl2,and 0.1 mg/ml of T7 RNA polymerase. Nucleotides, in-

cluding 5- �iodouridine triphosphate and 5-�bromouridinetriphosphate, were obtained from Sigma Chemical Co.,St. Louis, Missouri.� RNA fragments were purified by 20%denaturing polyacrylamide gel electrophoresis (PAGE).The desired fragment was eluted from the polyacryla-mide and ethanol-�precipitated in the presence of 0.3 Msodium acetate. R17 bacteriophage was propagated inEscherichia coli strain S26, and the coat protein was pu-rified using the procedure described by Carey and cow-orkers (Carey et al. (1983) Biochemistry 22: �4723).

Example 2. Binding Constants for RNA- �1 and RNA- �2 to R17 Coat Protein.

�[0111] RNA-�protein binding curves for hairpin variantsRNA-�1 (SEQ ID NO: �1), RNA-�2 (SEQ ID NO:�2) and RNA-3 (SEQ ID NO:�3) to the bacteriophage R17 coat proteinare shown in Figure 7. The association constants be-tween coat protein and the RNA hairpin variants weredetermined with a nitrocellulose filter retention assay de-scribed by Carey and co-�workers (Carey et al. (1983)supra). A constant, low-�concentration of 32P-�labeledRNA was mixed with a series of coat protein concentra-tions between 0.06 nM and 1 PM in 10 mM magnesiumacetate, 80 mM KCL, 80 Pg/ml BSA, and 100 mM Tris-HCl (pH 8.5 at 4°C) (TMK buffer). These were the samesolution conditions used in the crosslinking experiments.After incubation at 4°C for 45-60 min, the mixture wasfiltered through a nitrocellulose filter and the amount ofcomplex retained on the filter determined by liquid scin-tillation counting. For each experiment the data pointswere fit to a non-�cooperative binding curve and the Kdvalue shown in Figure 7 was calculated.

Example 3. Photocrosslinking of RNA- �1 and RNA- �2 to R17 Coat Protein at 308 nm.

�[0112] 32P-�Labeled RNA sequences RNA-�1 (SEQ IDNO:�1) and RNA-�2 (SEQ ID NO: �2) (5 nM) and R17 coatprotein (120 nM) � were each incubated on ice in 100 mMTris-�HCl (pH 8.5 at 4°C), 80 mM KCl, 10 mM magnesiumacetate, 80 Pg/ml of BSA for 15-25 min before irradia-tions. These are conditions under which the RNA is fullybound to the coat protein. The RNAs were heated in waterto 85°C for 3 min and quick cooled on ice before use toensure that the RNAs were in a hairpin conformation(Groebe and Uhlenbeck (1988) Nucleic Acids Res. 16:11725). A Lambda Physik EMG- �101 excimer lasercharged with 60 mbar of xenon, 80 mbar of 5% HCl inhelium and 2360 mbar of helium was used for 308 nmirradiations. The output of the XeCl laser was directedunfocused toward a 4 mm wide by 1 cm path length quartzcuvette containing the RNA-�protein complex. The laserwas operated in the range of 60 mJ/�pulse at 10 Hz; how-ever, only about 25% of the laser beam was incident uponthe reaction cell. Photocrosslinking yields of RNA-�1 (SEQID NO: �1) and RNA-�2 (SEQ ID NO:�2) to the R17 bacteri-ophage coat protein as a function of irradiation time are

29 30

EP 0 736 105 B1

18

5

10

15

20

25

30

35

40

45

50

55

shown in Figure 8. Crosslinked RNA was separated fromuncrosslinked RNA by PAGE, and the yields were deter-mined by autoradiography. Crosslinking of 5-�bromour-acil-�containing variant RNA- �1 (SEQ ID NO: �1) maximizedat about 40% because of competitive photodamage tothe coat protein which inhibits binding to the RNA (Gottet al. (1991) supra). Less photodamage to coat proteinoccurred with RNA 2 because of the shorter irradiationtime.�[0113] Crosslinking as a function of photons absorbedindicated that the quantum yield for crosslinking of BrU-RNA 1 is 0.014 and for crosslinking of IU-�RNA-�2 (SEQID NO: �2), 0.006 with irradiation at 308 nm. In spite of thelower quantum yield, a higher crosslinking yield was ob-tained with IU-�RNA 2 because of the seven times higherabsorption probability of the IU chromophore at 308 nm.BrU and IU absorb at 308 nm with molar extinction coef-ficients of 385 and 2640 L/mol·cm, respectively. Hence,a high level of photocrosslinking of the IU- �RNA wasachieved prior to protein damage.

Example 4. Identification of the Amino Acid Residue In-volved in the Crosslink of RNA-�1 to R17 Coat Protein

�[0114] Large scale crosslinking of RNA- �1 (SEO ID NO:1) to R17 coat protein. A 10 ml solution containing 300nM 5’-�end- �labeled RNA and 500 nM coat protein wasincubated on ice in the presence of 100 mM Tris-�HCl (pH8.5 at 4°C), 10 mM Mg �(OAc)2, 80 mM KCL, 80 mg bovineserum albumin (BSA), and 5 mM dithiothreitol (DTT) for10-90 min. A Lambda Physik EMG- �101 excimer laserwas used for monochromatic irradiation at 308 nm. Thebeam output was measured at 69 � 5 mJ/ �pulse at 10Hz. Approximately 50% of the beam was focused througha 7 mm-�diameter circular beam mask into a 1 cm pathlength quartz cuvette in a thermostated cell holder. Thelaser power was measured with a Scientech 360-001 diskcalorimeter power meter. The temperature was regulatedat 4 � 2°C with a Laude RC3 circulating bath.�[0115] The 10 ml reaction mixtures were prepared justprior to the irradiations which were performed in 2 mlfractions. After 5 min of irradiation the protein concentra-tion was brought to 1 PM. The reaction mixture was thenincubated for 3 min to allow exchange of photodamagedprotein for fresh protein in the nucleoprotein complex andirradiated for an additional 5 min. This step was repeatednine times to give 90 ml of irradiated sample. Thecrosslinking, analyzed by 20% denaturing PAGE, andquantitated on a Molecular Dynamics Phosphoimager,revealed 22% crosslinking.�[0116] The 90 ml sample contained 5.9 nmol ofcrosslinked RNA, 21 nmol of free RNA, 97 nmol of freecoat protein,� and 7.2 mg of BSA. The total volume wasreduced to 20 ml and split equally between two 50 mlpolypropylene screw cap centrifuge tubes (Nalgene) andethanol precipitated overnight at -20°C. The RNA andproteins were spun down to a pellet at 13,000 rpm in afixed angle J-�20 rotor with an Beckman J2-21 centrifuge.

Each pellet was resuspended in 1 ml of 0.5 M urea, 50mM Tris- �HCl pH 8.3, and 0.2% SDS for 48 h at 4°C withshaking. The fractions were combined, and the SDS wasthen removed by precipitation so as not to decrease theactivity of trypsin. This was achieved using 40 mM KC1,and the precipitate was removed by spinning through a0.22 Pm cellulose acetate spin filter. The trypsin condi-tions were optimized using 500 Pl of the solution.�[0117] Proteolytic Digestion. The remaining 1.5 ml ofcrosslinked RNA solution containing free RNA and pro-tein was brought to 6 ml to contain 1 M urea, 20 mMCaCl2, and 6 mM DTT, and then 1.61 mg (1: �5 w/w.)trypsin-�TPCK (251 units/mg) was added. The reactionproceeded at 36°C for 2 h at which time 1.61 mg moretrypsin was added. At 4 h a 100 PL aliquot was removedand the reaction stopped by quick freezing. The reactionwas analyzed by 20% polyacrylamide 19:�1 crosslinked,7 M urea, 90 mM Tris-�borate/�2 mM EDTA (TBE) gel elec-trophoresis (20% urea denaturing PAGE).�[0118] Purification of the digested crosslinked RNA.The trypsin reaction mixture was brought to 10 ml to re-duce the molar concentration of salt, and run through a240 Pl DEAE ion exchange centrifuge column. The col-umn was washed with 100 mM NaCl and spun dry in abench top centrifuge to remove free peptide. The columnbound material containing the RNA and crosslinked tryp-tic fragment was eluted from the column with 1 ml of 600mM NaCl and the column spun dry. An additional 200 Plof 600 mM NaCl was spun through the column. The twofractions were pooled, ethanol precipitated and pelletedat 10,000 rpm for 35 min at 4°C. The pellet was resus-pended in 25 Pl of 7 M urea- �TBE buffer, 10 mM DTT,0.1% bromophenol blue, 0.1% xylene cylanol, and heat-ed to 85°C for 4 min and purified by 20% denaturingPAGE. The gel ran for 3.5 h at 600 V. A 5 min phosphoim-age exposure was taken of the gel. The digested protein-RNA crosslink was then electrolytically blotted from thegel onto a PVDF protein sequencing membrane (0.2 mi-cron) from Bio-�RAD. The membrane was air dried,coomassie stained for 1 min, destained for 2 min in 50%MeOH: 50%H2O, and rinsed twice with deionized H2O.An autoradiogram was made of the membrane to visu-alize the digested protein RNA crosslink which was ex-cised from the membrane and submitted for Edman deg-radation. The immobilized peptide was sequenced by au-tomated Edman degradation, performed on an AppliedBiosystems 470A sequencer using manufacturer’s meth-ods and protocols (Clive Slaughter, Howard HughesMedical Institute, University of Texas, Southwestern).The Edman analysis indicated that the position of thecrosslink was tyrosine-�85 based upon the known aminoacid sequence (Weber (1983) Biochemistry 6:�3144).

Example 5. Photocrosslinking of RNA-�2 to R17 Coat Pro-tein at 325 nm.

�[0119] In an experiment analogous to that describedin Example 3, IU-�substituted RNA-�2 (SEQ ID NO:�2) was

31 32

EP 0 736 105 B1

19

5

10

15

20

25

30

35

40

45

50

55

photocrosslinked to R17 coat protein with monochromat-ic emission at 325 nm from an Omnichrome HeCd laser(model 3074-40M325). The power output of the HeCdlaser was 37 mW and the total beam of diameter 3 mmwas incident upon the sample. To increase excitation perunit time the beam was reflected back through the samplewith a dielectric-�coated concave mirror. Crosslinked RNAwas separated from uncrosslinked RNA by PAGE, andthe yields were determined with a PhosphoImager. Thepercent of the RNA crosslinked to the protein as a func-tion of irradiation time is shown in Figure 9. High-�yieldcrosslinking occurred without photodamage to the R17coat protein. In a separate experiment analogous irradi-ation of coat protein alone at 325 nm with yet a higherdose resulted in protein which showed the same bindingconstant to R17 coat protein. Irradiation at 325 nm ofBrU- �containing RNA- �1-�R17 coat protein complex did notresult in crosslinking because the BrU chromophore istransparent at 325 nm.

Example 6. Photocrosslinking of RNA-�1 and RNA- �2 to R17 Coat Protein with a Transilluminator.

�[0120] In an experiment analogous to that describedin Example 3, RNA-�1 (SEQ ID NO:�1) and RNA-�2 (SEQID NO: �2) were photocrosslinked to the R17 coat proteinwith broad- �band emission in the range of 312 nm from aFisher Biotech Transilluminator (model FBTIV-�816) fil-tered with polystyrene. Crosslinked RNA was separatedfrom uncrosslinked RNA by PAGE, and the yields weredetermined by autoradiography. Percent RNAscrosslinked to protein as a function of irradiation time isshown in Figure 10.

Example 7. Photoreaction of 5-Iodouracil with N- �Acetyltvrosine N- �Ethyl Amide.

�[0121] N-�acetyltyrosine N-�ethylamide was preparedas described by Dietz and Koch (1987) supra. Irradiationof a pH 7, aqueous solution of iodouracil and 10 molequivalent excess of N- �acetyltyrosine N- �ethyl amide at308 nm with a XeCl excimer laser gave a photoadductidentical to the photoadduct (structure 6) from irradiationof bromouracil and N-�acetyltyrosine N-�ethylamide (Dietzand Koch (1987) supra) as shown in Figure 11. Productcomparison was performed by C- �18 reverse phase HPLCand by 1H NMR spectroscopy. Although little is knownabout the mechanism of photocrosslinking of IU- �substi-tuted nucleic acids to associated proteins, this result sug-gests that it is similar to that of photocrosslinking of BrU-substituted nucleic acids to associated proteins.

Example 8. Preparation of a cDNA from an RNA Photo-crosslinked to a Protein.

�[0122] RNA-�7 (SEQ ID NO: �4) (Fig. 11) was preparedusing methodology as reported in Example 1 using aplasmid instead of a DNA template. The photocrosslink-

ing was performed as described in Example 3. A 4 mlreaction mixture consisting of 6.75 nM RNA and 120 nMR17 coat protein was irradiated, 2 ml at a time, at 308nm with unfocused emission from a XeCl excimer laser.The laser produced 50 mJ/ �pulse and was operated at 10Hz. The reaction proceeded to near quantitativecrosslinking, 85-90%, in 5 min of irradiation. Aftercrosslinking, 1 ml of the total reaction mixture was re-moved; EDTA (80 mM), SDS (0.1%), and CaCl2 (0.1 mM)were added; the free (unbound) RNA present was puri-fied away; and the protein digested with Proteinase K at60°C for 30 min. The RNA bound to residual protein wasethanol precipitated to remove salts and spun to a pellet.The pellet was washed three times with 70% ethanol toremove any residual salts. A reverse transcription reac-tion was employed to make a complementary DNA copyof the RNA template. A 13-�base promoter was annealedto the RNA and the reverse transcription reaction wasperformed under the standard conditions of the manu-facturer, Gibco (Gaithersburg, MD), and was stopped af-ter 1 hr. The cDNA was body labelled with 32P-�labelleddeoxycytidine triphosphate. The RNA template was thenremoved by hydrolyzing with 0.2 M sodium hydroxide at100°C for 5 min. The formation of the cDNA was followedby PAGE. A hydrolysis ladder and markers were addedto the gel to determine the length of the cDNA. The cDNAco-�migrated with the 44 nucleotide RNA template. If therehad been a stop in the cDNA as a result of crosslinkingmodification, a shortened product of 31 nucleotideswould have been observed. A small amount of a stopproduct was observed in the 22-25 nucleotide region ofthe gel, but this may have resulted from the hairpin sec-ondary structure which begins at position 25 of the cDNAon the RNA template. No stop in the 31 nucleotide regionof the gel appeared; this established that the reverse tran-scriptase had read through the position of the crosslink.A diagram of the gel appears in Figure 14.

Example 9. Iodocytosine Photocrosslinking.

�[0123] 5-�iodocytosine (IC) was incorporated in a hair-pin RNA (RNA 8) that contained cytosine at the -5 positionand bound the R17 coat protein with high affinity. TheIC-�bearing RNA is designated RNA 9. RNA 9 (5 nM) andR17 coat protein (120 nM) were incubated on ice in 100mM Tris-�HCl (pH 8.5 at 4°C) �/ �80 mM KCl/�10 mM magne-sium acetate/ 80 Pg/ml BSA for 15-25 min prior to irra-diation. The RNA in water was heated to 85°C for 3 minand quick cooled on ice before use to ensure that it wouldbe in a conformation that bound the coat protein (Groebeand Uhlenbeck (1988) supra). The complex was irradi-ated for 5 min at 4°C, and the experiment was comparedto control irradiations of RNA 2 and RNA 8 coat proteincomplexes. Irradiation of RNA 8-�coat protein complexresulted in no crosslinked product. Irradiation of RNA 9-coat protein complex resulted in the formation of acrosslink that formed in high yield (70-80%) similar to theyield of the control irradiation of RNA 2-�coat protein com-

33 34

EP 0 736 105 B1

20

5

10

15

20

25

30

35

40

45

50

55

plex (80-90%)� Crosslinking of RNA 9 is presumed to oc-cur through a similar mechanism as RNAs containing IUat position -5 of the loop hairpin (Figs. 6 and 12). Thisassumption is based on the specificity of the crosslinksince RNA 8 did not photocrosslink.

Example 10. Incorporation of Halogenated Nucleotides into DNA Ligands.

�[0124] Photoreactive nucleotides may be incorporatedinto a DNA ligand capable of crosslinking to a target mol-ecule upon irradiation by the methods discussed above.5-�Bromodeoxyuracil (BrdU), 8-�b romo-�2’-�deoxyadenine,and 5-�iodo-�2’-�deoxyuracil are examples of such photo-reactive nucleotides.

Example 11. PhotoSELEX.

�[0125] In one embodiment of the present invention, thephotoSELEX method is applied to completion in the se-lection of a nucleic acid ligand which binds and photo-crosslinks to a target molecule.�[0126] A randomized set of nucleic acid oligonucle-otides is synthesized which contain photoreactivegroups. The oligonucleotides of the candidate mixturemay be partially or fully saturated at each available po-sition with a photoreactive group. The candidate mixtureis contacted with the target molecule and irradiated atthe appropriate wavelength of light. Oligonucleotidescrosslinked to the target molecule are isolated from theremaining oligonucleotides and the target molecule re-moved. cDNA copies of the isolated RNA sequences aremade and amplified. These amplified cDNA sequencesare transcribed into RNA sequences in the presence ofphotoreactive groups, and the photoSELEX process re-peated as necessary.

Example 12. Selection of Enhanced Photocrosslinking Ligands: SELEX Followed by PhotoSELEX.

�[0127] In one embodiment of the method of the presentinvention, selection of nucleic acid ligands through SE-LEX is followed by selection through photoSELEX forligands able to crosslink the target molecule. This proto-col leads to ligands with high binding affinity for the targetmolecule that are also able to photocrosslink to the target.�[0128] Photoreactive nucleotides are incorporated intoRNA by T7 polymerase transcription with the reactivenucleotide triphosphate in place of a specified triphos-phate. For example, 5-�bromouridine triphosphate is sub-stituted for uridine triphosphate or 8-�bromo-�adenosinetriphosphate is substituted for adenosine triphosphate.A randomized set of RNA sequences containing photo-reactive nucleotides are generated and the SELEX meth-odology applied. The initial SELEX rounds are used toeliminate intrinsically poor binders and enhance the poolof molecules that converge to form a pool of RNAs thatcontain the photoreactive group�(s) and which bind to the

target molecule. Aliquots from the initial SELEX roundsare irradiated and the enhancement of photocrosslinkingfollowed via PAGE as the rounds proceed. As a slowermigrating band representing crosslinked products startsto become evident, the pool of RNAs are introduced intorounds of photoSELEX. RNAs that have a photoreactivegroup adjacent to a reactive amino acid residue in thenucleoprotein complexes form a crosslink and are se-lected and RNAs that do not have reactive nucleotidesin proximity to reactive target residues are eliminated.�[0129] This protocol selectively applies photoSELEXselection to previously identified ligands to a target mol-ecule.

Example 13. PhotoSELEX Followed by SELEX.

�[0130] In another embodiment of the method of thepresent invention, an RNA ligand able to photocrosslinka target molecule is preselected through the photoSE-LEX methodology. Subsequently, SELEX is performedto select a crosslinking oligonucleotide for ability to bindthe target molecule.

Example 14. Limited SELEX Followed by PhotoSELEX.

�[0131] In this embodiment of the present invention, nu-cleic acid ligands are selected through the SELEX proc-ess for a limited number of selection rounds. SELEX isnot applied to completion as in Example 12. Rather, thecandidate mixture is partially selected for oligonucle-otides having relatively enhanced affinity for the targetmolecule. The random oligonucleotides of the candidatemixture contain photoreactive groups and the initial SE-LEX selection is conducted in the absence of irradiation.PhotoSELEX is then performed to select oligonucle-otides able to crosslink to the target molecule.�[0132] This protocol allows the selection of crosslink-ing ligands from a pool of oligonucleotides with a some-what enhanced capacity to bind the target molecule andmay be useful in circumstances where selection to com-pletion through SELEX does not yield crosslinking lig-ands.

Example 15. Limited Directed PhotoSELEX.

�[0133] In one embodiment of the method of the presentinvention, in which nucleic acid ligands identified throughSELEX are subjected to limited randomization, followedby selection through photoSELEX.�[0134] The construction of the DNA template used totranscribe the partially randomized RNA is based on thesequence of the initially selected ligand and contains ateach position primarily the nucleotide that is complemen-tary to that position of the initial selected RNA sequence.However, each position is also partially randomized byusing small amounts of each of the other three nucle-otides in the sequencer, which varies the original se-quence at that position. A limited RNA pool is then tran-

35 36

EP 0 736 105 B1

21

5

10

15

20

25

30

35

40

45

50

55

scribed from this set of DNA molecules with a photore-active triphosphate replacing a specific triphosphate inthe reaction mix (i.e., BrU for U). The partially randomizedset of RNA molecules which contains the photoreactivenucleotides is mixed with a quantity of the target protein.Bound RNAs that have a photoreactive group adjacentto a reactive amino acid residue in the nucleoprotein com-plex form covalent crosslinks upon irradiation. RNAs thatbind and crosslink are selected through several roundsof photoSELEX and separated away from RNAs that bindbut do not crosslink.

Example 16. Methods for Modifying a Target Molecule.

�[0135] In another embodiment of the method of thepresent invention, photoSELEX is applied to develop aligand capable of modifying a target molecule. Underthese circumstances, incorporation of a photoreactivegroup onto or into a ligand selected by photoSELEX orSELEX may modify the target in several ways such thatthe biological activity of the target molecule is modified.For example, the target molecule may be inactivated byphotocrosslinked ligand. Mechanisms of inactivation in-clude electron or hydrogen abstraction from the targetmolecule or radical addition to the target molecule thatelicit a chemical modification. These different mecha-nisms may be achieved by changing the mode of irradi-ation.�[0136] A ligand selected through photoSELEX usedas a diagnostic for a target molecule with ultraviolet (UV)light may also inactivate the same target in vivo if thesource of irradiation is changed to X- �rays or gamma rays.The resultant vinyl radical may work similarly to a hy-droxyl radical, that is, by abstraction of hydrogen atomsfrom the binding domain of the target molecule.�[0137] X-�ray irradiation of the R17 coat protein boundto radio- �labelled IU- or BrU-�substituted RNA hairpin se-quences may result in the formation of a crosslink. TheBrU or IU chromophore may also be excited to a higherenergy state by X- �ray irradiation resulting in the formationof a vinyl radical (Mee (1987) in: Radiation Chemistry:Principles and Applications (Farhataziz and Rodgers,eds.), VCH Publishers, New York, pp. 477-499). The rad-ical abstracts a hydrogen from the binding domain of theR17 coat protein, thereby reducing or inhibiting its abilityto bind the RNA ligand. Inactivation is tested by X-�rayirradiation of the R17 coat protein in the presence andabsence of substituted RNAs. The formation ofcrosslinked complexes is analyzed by PAGE. The effectof X-�ray irradiation of RNA resulting in modification ofbinding by modification of the protein domain is followedby nitrocellulose binding assay.

Example 17. Diagnostic Use of PhotoSELEX To Identify Unique Proteins Associated with Specific Disease Proc-esses.

�[0138] A goal of diagnostic procedures is to correlate

the appearance of unique proteins with specific diseaseprocesses. Some of these correlations are obvious, e.g.,after bacterial or viral infections, one can detect antigenswhich are antigen specific or antibodies to such antigensnot found in the blood of uninfected subjects. Less obvi-ous correlations include the appearance in serum of α-foeto protein which is directly correlated with the pres-ence of the most common form of testicular cancer.�[0139] The photoSELEX method may be applied to thediscovery of heretofore unknown correlations betweenbiological proteins and important human diseases. In oneembodiment of the present invention, serum is taken froma patient with a disease, RNA ligands to all the proteinsin the serum are produced and adsorbed to normal sera.RNA ligands to serum proteins may be identified throughthe SELEX method, with subsequent incorporation ofphotoreactive groups, or may be identified through photo-SELEX, initially selected from a candidate mixture of ol-igonucleotides containing one or more photoreactivegroups. RNA ligands left unbound are those which spe-cifically bind only unique proteins in the serum from pa-tients with that disease. For example, RNA ligands areinitially identified to a limited number of serum proteins(e.g., 11). The RNA ligands identified contain a modifiedNTP having a reversible or photoreactive functionalgroup capable of crosslinking reversibly or non-�reversiblywith the target protein. Optionally, the presence of across- �linked ligand to every protein may be verified. TheRNA ligands are then removed and amplified. RNA isthen transcribed for a second SELEX round. RNA is nowbound to a large excess of 10 of the original 11 proteins,leaving an RNA ligand specific for the unique (11th) pro-tein. This RNA is then amplified. This is a subtractivetechnique.�[0140] In one embodiment of the diagnostic method ofthe present invention, the method described above isused to identify a ligand to an abnormal protein, for ex-ample, an α-�foeto protein. Sera from patients with impor-tant diseases is obtained and RNA ligands to all proteinspresent identified. The RNA ligands are adsorbed to nor-mal sera, leaving an unbound ligand. The unbound ligandis both a potential diagnostic agent and a tool for identi-fying serum proteins specifically associated with a dis-ease.

Example 18. Method of Treating Disease by In Vivo Use of Photocrosslinking Nucleic Acid Ligand.

�[0141] A nucleic acid ligand to a target molecule asso-ciated with a disease state is selected through the pho-toSELEX process (Example 11). The photoSELEX se-lected nucleic acid ligand may be introduced into a patientin a number of ways known to the art. For example, thenon-�halogenated photoSELEX ligand is cloned into stemcells which are transferred into a patient. The ligand maybe transiently or constitutively expressed in the patient’scells. IC administered to the patient is incorporated intothe oligonucleotide product of the cloned sequence. Up-

37 38

EP 0 736 105 B1

22

5

10

15

20

25

30

35

40

45

50

55

on irradiation, the ligand is able to crosslink to the targetmolecule. Irradiation may include visible, 325 nm, 308nm, X-�ray, ultraviolet, and infrared light.�[0142] Alternatively, the photoSELEX ligand may betaken into a patient’s cells as a double-�stranded DNAwhich is transcribed in the cell in the presence of iodinat-ed cytosine. Further methods of introducing the photoSE-LEX ligand into a patient include liposome delivery of thehalogenated ligand into the patient’s cells.

Example 19. PhotoSELEX Ligands for Use in In Vitro Diagnostic, In Vivo Imaging and Therapeutic Delivery.

�[0143] PhotoSELEX may be used to identify moleculesspecifically associated with a disease condition and/orabnormal cells such as tumor cells. PhotoSELEX-�iden-tified oligonucleotides may be produced that react cov-alently with such marker molecules.�[0144] In one embodiment of the present invention, thetarget for photoSELEX is the abnormal serum or tumorcell (e.g., the target mixture). A library candidate mixtureof oligonucleotides is generated containing photoreac-tive groups. Using one of the above-�described photoSE-LEX protocols, oligonucleotides able to photocrosslinkto the unique proteins in the abnormal serum or on thetumor cells are identified. Oligonucleotides able tocrosslink to a marker protein on a tumor cell are usefulas in vitro diagnostics or when coupled to enhancingagents for in vitro imaging. Further, oligonucleotides ableto crosslink to a marker protein on a tumor cell may beused therapeutically, for example, as a method for im-mune activation, as a method of inactivation, or as amethod of delivering specific target-�active pharmaceuti-cal compounds.

Example 20. PhotoSELEX and HIV-�1 Rev.

�[0145] At each position of the template deoxyoligonu-cleotide synthesis, the nucleotide reagent ratio was 62.5:12.5: �12.5: �12.5. The nucleotide added in greater amountat each position corresponds to the nucleotide found inthe 6a sequence (SEQ ID NO:�5) at the same position.�[0146] Cloning and Sequencing procedure: RNA’s iso-lated from each round were reverse transcribed to pro-duce cDNA and PCR amplified producing a 111 bp frag-ment with unique BamHI and HindIII restriction sites atthe ends. The phenol/�CHCl3 treated fragment and apUC18 vector were digested together overnight withBamHI and HindIII at 37°C., phenol/�ChCl3 treated andprecipitated. The digested vector and PCR product wasligated at room temperature for 4 hours and with T4 DNAligase and transformed to competent DH5α-�F’cells whichwere then grown on ampicillin-�containing LB plates. In-dividual colonies were grown overnight in LB- ampicillinmedia and plasmid was prepared using Wizard (Prome-ga) plasmid preparation kit. Sequencing was performedutilizing a Sequenase (USB) kit.�[0147] Conditions for nitrocellulose filter binding selec-

tions: All rounds utilized approximately 20 nM RNA.Round 1 and 2: 6 nM Rev. Round 3: 3 nM Rev. Round8: 1nM Rev. Round 9-10: 3 nM Rev. Binding reactionvolumes ranged from 5 mls to 1 ml.�[0148] Conditions for crosslinking selections: �

Approximately 50-100 nM of folded pool RNA wasadded to 0.2 (Rounds 4-6) or 0.5 (Round 7) PM Rev,1 PM BSA in 1x BB (50mM TrisAc pH 7.7, 200mMKOAc, 10 mM DTT) on ice and incubated 5 minutesat 37°C. The samples were then irradiated at 37°C,for 3 minutes at 308 nm by a XeCl excimer laser(round 4), 30 minutes at 325 nm by a HeCd laser(round 5), 10 minutes at 325 nm, (round 6), or 1minute at 325 nm, (round 7). Approximately one-�halfof the sample was heated in 50% formamide, 40 ugtRNA at 90°C. for 4 minutes and separated by elec-trophoresis in an 8 percent polyacrylamide-/�8 M ureagel.

�[0149] The following procedure was utilized to elutecrosslinked RNAs from acrylamide gels with approxi-mately 80% recovery: The nucleoprotein containing gelslice was crushed to a homogenous slurry in 1X PK buffer(100mM Tris-�Cl pH 7.7, 50mM NaCl and 10mM EDTA).Proteinase K was added to 1 mg/ml concentration andincubated at 42°C for 30 minutes. 15 minute incubationsat 42°C with increasing urea concentrations of approxi-mately 0.7 M, 1.9 M, and 3.3M were performed. The re-sulting solution was passed through DMCS treated glasswool and 0.2 um cellulose acetate filter. The filtered so-lution was extracted twice with phenol/�CHCl3 and thenprecipitated with a 1:�1 volume mixture of EtOH:�isopro-panol.�[0150] The crosslinked band from each round wasplaced in scintillation fluid and counted in a Beckman LS-133 Liquid Scintillation System. The percent crosslinked= cpms of crosslinked product from RNA + Rev after 4minutes irradiation at 325 nm minus cpms in crosslinkedregion for RNA only irradiated divided by total cpms.� Thefold increase in crosslinking is % R13 crosslinked dividedby % D37 crosslinked.�[0151] Simultaneous selection for affinity andcrosslinking using competitor tRNA was performed asfollows. 10uM yeast tRNA was added to 0.5PM Rev, 1PM BSA in 1x BB (50mM TrisAc (pH=7.5), 200mM KOAc,10 mM DTT) and incubated 10 minutes on ice. 200,000cpms (approximately 50-100 nM final concentrationRNA) was added and incubated an additional 15 to 60minutes on ice followed by 5 minutes at 37°C. The sam-ples were then irradiated 4 minutes at 325 nm by a HeCdlaser at 37°C. Approximately one- �third of the sample washeated in 50% formamide, 40 ug tRNA at 90 C. for 4minutes and separated by electrophoresis in an 8 percentpolyacrylamide- �8M urea gel.�[0152] The LI crosslinking RNA ligands form additionalcrosslinked product with a 4 minute 325 nm laser irradi-ation.

39 40

EP 0 736 105 B1

23

5

10

15

20

25

30

35

40

45

50

55

�[0153] The template oligos used to produce the trun-cated RNA’s are: PTS-�1; 5’-�TAATACGACTCACTATA-3’, (SEQ ID NO:�60) DNA-�2; 5’-�GAGTGGAAACACACGT-GGTGTTT- � CATACACCCTATAGTGAGTCGTATTA- � 3’(SEQ ID NO: 61), and DNA-�24; 5’- �AGGGTTAACAGGT-GTGCCTGTTAATCCCCTATAGT-� GAGTCGTATTA- � 3’(SEQ ID NO: �62). PTS- �1 was annealed with DNA-�2 orDNA-�24 to produce a template for T7 transcription.�[0154] To calculate the number of changes for individ-ual molecules compared to 6a (SEQ ID NO:�5), each wasaligned to 6a for maximum similarity. Gaps are calculatedas one change and truncated molecules were countedas unchanged. To calculate the average probability offinding molecules within each class; the average numberof specific (s) and non-�specific (ns) changes and un-changed (u) were calculated and used in the equation:�

(P)�=�(. �125)s �(.�375)ns�(. �625)u. Class Ia (P) �=9x10-15; Ib(P)�=3x10-15; Ic (P)�=7x10-13; Id (P)�=3x10-15; Class II(P)�=2x10-14. Since the starting population consistsof 1014 molecules, sequences with (P)�<10-14 will notbe represented. (s) are those changes required toproduce the uppercase, consensus nucleotides and(ns) are additional changes.

�[0155] Trunc24 (SEQ ID NO:�59) photo-�independentcrosslinking with HIV-�1 Rev in the presence of humannuclear extracts was determined as follows: Trunc24RNA, nuclear extracts, and Rev protein were combinedand incubated on ice for 10 min. Samples were mixed 1:1 with 8 M urea loading buffer and placed on a 7 M urea,8% polyacrylamide gel for analysis, XL indicates the nu-cleoprotein complex, RNA indicates free trunc24 RNA.�[0156] Examples 21-26 described below are not em-bodiments forming part of the invention claimed by thispatent, but are useful for understanding the invention thatis claimed.

Example 21. Primer Extension Inhibition Solution SE-LEX.

�[0157] Primer extension inhibition relies on the abilityof a tightly bound target molecule to inhibit cDNA syn-thesis of high affinity oligonucleotides and results in for-mation of an amplifiable cDNA pool corresponding to highaffinity oligonucleotides and a non-�amplifiable cDNA poolcorresponding to low affinity oligonucleotides. Thus, thePCR step of solution SELEX acts as a partitioning screenbetween two cDNA pools. General protocols for nucleicacid synthesis, primer extension inhibition and PCR areherein provided. Further, N-�acryloylamino phenyl mer-curic gel electrophoretic conditions for separation of se-lected nucleic acid ligands is described. The methods ofcloning and sequencing nucleic acid ligands is as de-scribed by Tuerk and Gold (1990) supra.�[0158] RNA Synthesis. The RNA candidate mixturewas generated by incubating RNA polymerase and DNAtemplates. The reaction conditions are 8% polyethylene

glycol 8000, 5 mM dithiothreitol, 40 mM Tris- �HCl (pH 8.0),12 mM MgCl2, 1 mM spermidine, 0.002% Triton X-�100,2 mM nucleotide triphosphates, and 1 unit/�Pl RNApolymerase. Reactions are incubated at 37°C for 2 hours.�[0159] The transcription protocol may be used to gen-erate RNAs with modified nucleotides. The transcriptionreaction may either be primed with a nucleotide triphos-phate derivative (to generate a modified 5’ end), modifiednucleotides may be randomly incorporated into the nas-cent RNA chain, or oligonucleotides or their derivativesligated onto the 5’ or 3’ ends of the RNA product.�[0160] Primer Extension Inhibition. Primer extensioninhibition is performed as described by Hartz et al. (1988)supra. Briefly, an oligonucleotide primer is annealed tothe 3’ end of the oligonucleotides of the candidate mixtureby incubating them with a 2-�fold molar excess of primerat 65°C for 3 min in distilled water. The annealing reactionis cooled on ice, followed by the addition of 1/10 volumeof 10X concentrated extension buffer (e.g., 10 mM Tris-HCl (pH 7.4), 60 mM NH4Cl, 10 mM Mg-�acetate, 6 mMβ-�mercaptoethanol, and 0.4 mM nucleotide triphos-phates). Primer extension is initiated by addition ofpolymerase and incubation at any of a variety of temper-atures ranging between 0-80°C, and for times rangingfrom a few seconds to several hours. In one embodimentof the method of the present invention, primer extensionis first conducted in the presence of chain terminatingnucleotide triphosphates such that low-�affinity nucleic ac-ids preferentially incorporate these chain terminators. Asecond primer extension is then conducted after remov-ing the target from high affinity nucleic acids and remov-ing the chain terminating nucleotides triphosphates.�[0161] Polymerase Chain Reaction. The polymerasechain reaction (PCR) is accomplished by incubating anoligonucleotide template, either single- or double-�strand-ed, with 1 unit/�Pl thermal stable polymerase in buffer (50mM KCl, 10 mM Tris- �HCl (pH 8.6), 2.5 mM MgCl2, 1.7mg/ml BSA, 1 mM deoxynucleotide triphosphates, and1 PM primers). Standard thermal cycles are 95° C for 30sec, 55°C for 30 sec, and 72°C for 1 min, repeated asnecessary. One modification of the PCR protocol gener-ates single- �strand DNA by incubating either single- ordouble-�stranded template with a single, elongated primeroligonucleotides and results in an elongated product.PCR preferentially amplifies the oligonucleotides ren-dered amplifiable in the primer extension steps describedabove.�[0162] �(N-�Acryloylamino)�phenyl mercuric gel electro-phoresis. Polyacrylamide gel electrophoresis using N-acryloylamine phenyl mercury (APM) was performed asdescribed by Igloi (1988) Biochemistry 27:�3842. APMwas synthesized by mixing 8 ml of acetonitrile to 0.35 gof (p- �aminophenyl) �mercuric acetate at 0°C, followed by2 ml of 1.2 M NaHCO3. A total of 0.2 ml of acryloyl chloridewas then added with vigorous stirring and the reactionincubated overnight at 4°C. The solid phase was collect-ed by centrifugation and washed with water, dissolvedby warming to 50°C in 8.5 ml of dioxane, followed by

41 42

EP 0 736 105 B1

24

5

10

15

20

25

30

35

40

45

50

55

filtration to remove undissolved contaminants. APM crys-tals were formed upon standing at room temperature andthe solid was washed again with water and dried undervacuum. APM was stored at 4°C. APM- �polyacrylamidegels were prepared by addition of a appropriate aliquotof a 1 mg/ml solution of APM in formamide to a solutioncontaining a given amount of acrylamide, bis�(acryla-mide), an urea in 0.1 M Tris-�borate/�EDTA (pH 8.3). Po-lymerization was initiated by addition of 0.5 ml of 1% am-monium persulfate and 7 ul of TEMED per 10 ml of gelsolution.

Example 22. Enzymatic or Chemical Degradation Solu-tion SELEX.

�[0163] Enzymes or chemicals may be used to selec-tively degrade the pool of cDNA corresponding to low-affinity oligonucleotides. In one embodiment of thepresent invention, restriction enzymes are used to selec-tively degrade the cDNA pool corresponding to low-�af-finity oligonucleotides. A number of restriction enzymeshave been identified that cleave single-�stranded DNA.These enzymes cleave at specific sequences but withvarying efficiencies.�[0164] Restriction enzyme digestion may be per-formed with a variety of sequence specific restriction en-donucleases. Endonucleases that cleave single-�strand-ed DNA include DdeI, HaeIII, HgaI, HinfI, HinPI, MnII,PstI, and RsaI. These enzymes are used under standardconditions known to those skilled in the field of molecularbiology. Double-�stranded nucleic acids may also becleaved using the proper combination of nucleic acid re-striction sequences and site specific restriction nucleas-es.�[0165] The basic solution SELEX procedure is fol-lowed as described in the SELEX Patent Applications.The first cDNA extension is performed in the presenceof four dNTPS, followed by removal of the target. Thesecond cDNA extension is performed with modified nu-cleotides that are resistant to enzymatic cleavage by re-striction endonucleases. The mixture of cDNA extensionproducts is incubated with the appropriate restriction en-zyme. The product of the first cDNA extension from freenucleic acid is cleaved to remove the primer annealingsite, rendering this cDNA pool non-�amplifiable by PCR.The efficiency of cleavage by restriction endonucleasesmay be improved using a hairpin at the restriction site(RS) to create a localized double- �stranded region, asshown in Figure 24.�[0166] Alternatively, the first cDNA extension productis rendered selectively degradable by other classes ofenzymes by incorporation of modified nucleotides. Forexample, cDNA corresponding to low affinity ligands maybe synthesized with nucleotides sensitive to uracil DNAglycosylase, while cDNA corresponding to high affinityligands may incorporate resistant nucleotides.�[0167] Chemical degradation of cDNA correspondingto low affinity ligands can be accomplished by incorpo-

ration of 7- �methylguanosine, 5- �bromouracil, or 5-�iodou-racil as described using piperidine or photodegradation(Sasse- �Dwight and Gralla (1991) Methods Enzymol.208: �146; Aigen and Gumport (1991) Methods Enzymol.208: �433; Hockensmith et al. (1991) Methods Enzymol.208: �211).

Example 23. Solution SELEX Followed by Affinity Chro-matography.

�[0168] Selective removal of either the first or secondcDNA extension products may be achieved through af-finity chromatography. Removal of the first cDNA exten-sion product preferentially removes the cDNA pool cor-responding to free or low- �affinity nucleic acids. Removalof the second cDNA extension product preferentially re-tains cDNA corresponding to the high-�affinity ligand. Thisstrategy relies on the incorporation of modified nucle-otides during cDNA synthesis.�[0169] Selective Removal of First Extension Product.Following the basic solution SELEX protocol, the first cD-NA extension is performed in the presence of modifiednucleotides (e.g., biotinylated, iodinated, thiolabelled, orany other modified nucleotide) that allow retention of thefirst cDNA pool on an affinity matrix (Fig. 25). The targetis then removed and the second cDNA extension per-formed in the presence of non-�modified nucleotides.� ThecDNAs that have incorporated the modified nucleotidesmay be removed by affinity chromatography using a col-umn containing the corresponding affinity ligand. The cD-NA pool corresponding to nucleic acids with high affinityfor the target remain and are then amplified by PCR.�[0170] Selective Removal of the Second ExtensionProduct. Following the basic protocol, the first cDNA ex-tension is performed in the presence of four dNTPs, andthe second cDNA extension is performed in the presenceof modified nucleotides (e.g., biotinylated, iodinated, thi-olabelled, or any other modified nucleotide) that allowretention of the second cDNA pool on an affinity matrixas described above.�[0171] Incorporation of Specific Sequences for An-nealing to An Affinity Matrix. In an alternate embodimentof the method of the present invention, a special se-quence can also be selectively incorporated for anneal-ing to an affinity matrix. Thus, either first or second syn-thesis cDNAs can be retarded and purified on commer-cially obtainable matrices as desired.

Example 24. Exonuclease Inhibition Solution SELEX.

�[0172] Exonuclease inhibition may be used to isolatedouble-�stranded ligands. Double-�stranded nucleic acidligands tightly bound to the target molecule will inhibitexonuclease hydrolysis at the 3’ edge of the binding site.This results in a population of nucleic acid molecules re-sistant to hydrolysis that also contain a long single-stranded 5’ overhang and a central base paired region(see Fig. 26). This nucleic acid molecule is a substrate

43 44

EP 0 736 105 B1

25

5

10

15

20

25

30

35

40

45

50

55

for any polymerase, and incubation with polymerase willgenerate the double-�stranded starting material. This mol-ecule is amplified by PCR. Members of the nucleic acidcandidate mixture that are not tightly bound to the targetmolecule are digested during the initial exonuclease step.�[0173] 3’→ 5’ hydrolysis of double-�stranded nucleic ac-id is accomplished by incubation with any double- �strand-ed specific 3’→ 5’ exonuclease. Exonuclease III specifi-cally hydrolyzes double-�stranded DNA 3’→ 5’ and is ac-tive in a variety of buffers, including 50 mM Tris-�HCl (pH8.0), 5 mM MgCl2, 10 mM β-�mercaptoethanol at 37°C.

Example 25. Solution SELEX Method for Isolating Cat-alytic Nucleic Acids.

�[0174] Solution SELEX may be used to isolate catalyticnucleic acid sequences. This embodiment of the inven-tion takes advantage of a linear to circular transformationto sort non-�catalytic nucleic acids from catalytic nucleicacids.�[0175] As shown in Figure 27, the PCR step may beexploited to screen the nucleic acid candidate mixturefor catalytic members. Catalytic nucleic acids that eitherself- �circularize, or alter their 5’ or 3’ ends to allow circu-larization with ligase, will amplify during PCR. The figureillustrates circle formation by catalytic members of thecandidate mixture; the non-�catalytic oligonucleotidemembers of the candidate mixture will remain linear. Aftercircularization, the candidate mixture is incubated with aprimer that anneals to the extreme 5’ end. In this embod-iment of the invention, only the circular oligonucleotidemembers will generate cDNA and be amplified duringthe PCR step.�[0176] This strategy isolates nucleic acids that eitherdirectly catalyze self- �circularization or that modify theirown ends so that the amplifiable form may be generatedby incubation with ligase. As shown in Figure 27, theunusual interaction of the cDNA primer with the 5’ endof the oligonucleotides of the candidate mixture permitsamplification of only the circular molecules. In a furtherembodiment of the method of the present invention, thisstrategy is modified to allow isolation of catalytic nucleicacids that catalyze novel reactions.

Example 26. Automation of Solution SELEX.

�[0177] The automated solution SELEX protocol repre-sents a modification of the technology used in the auto-mated DNA synthesizer. The nucleic acid candidate mix-ture is attached to a solid support by either the biotin/avidin interaction or a variety of covalent chromatograph-ic techniques (e.g., the condensation of modified nucle-otides onto maleimide or citraconic anhydride supports).The bound nucleic acid candidate mixture provides agood substrate for targeting binding, and the column al-lows use of a single reaction vessel for the SELEX pro-cedure. Primer extension inhibition is used to physicallysort low and high affinity ligands. Low affinity nucleic ac-

ids may be degraded by incorporation of modified nucle-otides during the first cDNA extension step that rendersthe cDNA degradable as described in Example 22, whilehigh affinity ligands are copied into non-�degradable cD-NA and amplified by PCR. For additional rounds of so-lution SELEX, the PCR generated candidate mixture ispurified or is transcribed into RNA and reattached to asecond solid support, in the same or a new reaction ves-sel as desired. The process is repeated as necessary.

SEQUENCE LISTING

�[0178]

(1) GENERAL INFORMATION: �

(i) APPLICANT: GOLD, LARRYWILLIS, MICHAELKOCH, TADRINGQUIST, STEVENJENSEN, KIRKATKINSON, BRENT

(ii) TITLE OF INVENTION: SYSTEMATIC EV-OLUTION OF LIGANDS BY EXPONENTIALENRICHMENT: PHOTOSELECTION OF NU-CLEIC ACID LIGANDS AND SOLUTION SE-LEX

(iii) NUMBER OF SEQUENCES: 64

(iv) CORRESPONDENCE ADDRESS:�

(A) ADDRESSEE: Swanson & Bratschun,L.L.C.(B) STREET: 8400 E. Prentice Avenue,Suite 200(C) CITY: Englewood(D) STATE: Colorado(E) COUNTRY: USA(F) ZIP: 80111

(v) COMPUTER READABLE FORM: �

(A) MEDIUM TYPE: Diskette, 5.25 inch, 360Kb storage(B) COMPUTER: IBM(C) OPERATING SYSTEM: MS- �DOS(D) SOFTWARE: WordPerfect 5.1

(vi) CURRENT APPLICATION DATA: �

(A) APPLICATION NUMBER:(B) FILING DATE:(C) CLASSIFICATION:

(vii)�PRIOR APPLICATION DATA:�

45 46

EP 0 736 105 B1

26

5

10

15

20

25

30

35

40

45

50

55

(A) APPLICATION NUMBER: 08/123,935(B) FILING DATE: 17 September 1993

(vii) �PRIOR APPLICATION DATA:�

(A) APPLICATION NUMBER: 08/143,564(B) FILING DATE: 25 October 1993

(viii) �ATTORNEY/�AGENT INFORMATION: �

(A) NAME: Barry J. Swanson(B) REGISTRATION NUMBER: 33,215(C) REFERENCE/ �DOCKET NUMBER:NEX10/PCT

(ix) TELECOMMUNICATION INFORMATION:�

(A) TELEPHONE: (303) 793-3333(B) TELEFAX: (303) 793-3433

(2) INFORMATION FOR SEQ ID NO:�1: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 19 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(ix) FEATURE:�

(D) OTHER INFORMATION: U at position13 is 5- �bromouracil

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�1: �

GGGAGCGAGC AAUAGCCGC 19

(2) INFORMATION FOR SEQ ID NO:�2: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 19 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(ix) FEATURE:�

(D) OTHER INFORMATION: U at position13 is 5- �iodouracil

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�2: �

GGGAGCGAGC AAUAGCCGC 19

(2) INFORMATION FOR SEQ ID NO:�3: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 19 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(ix) FEATURE: �

(D) OTHER INFORMATION: U at position13 has hydrogen molecule attached

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�3: �

GGGAGCGAGC AAUAGCCGC 19

(2) INFORMATION FOR SEQ ID NO:�4: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 44 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(ix) FEATURE: �

(D) OTHER INFORMATION: all U are 5- �io-douracil

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�4: �

GAACAUGAGG AUUACCCAUG AAUUC-GAGCU CGCCCGGGCU CUAG 44

(2) INFORMATION FOR SEQ ID NO:�5: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�5: �

GGGUGCAUUG AGAAACACGU UU-GUGGACUC UGUAUCU 37

(2) INFORMATION FOR SEQ ID NO:�6: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

47 48

EP 0 736 105 B1

27

5

10

15

20

25

30

35

40

45

50

55

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�6: �

AGGUACGAUU AACAGACGAC UGU-UAACGGC CUACCU 36

(2) INFORMATION FOR SEQ ID NO:�7

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�7: �

UAACGGCUUA ACAAGCACCA UUGU-UAACCU AGUGCCU 37

(2) INFORMATION FOR SEQ ID NO:�8: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�8: �

GAGUGGCUUA ACAAGCACCA UUGU-UAACCU AGUACCU 37

(2)�INFORMATION FOR SEQ ID NO:�9: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:�9: �

GUGCAGAUUA ACAACAACGU UGU-UAACUCC UCCUCU 36

(2) INFORMATION FOR SEQ ID NO:�10:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10: �

CUGUGGAUUA ACAGGCACAC CUGU-UAACCG UGUACCU 37

(2) INFORMATION FOR SEQ ID NO:�11:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:11: �

CUGUGGAUUA ACAGGCACAC CUGU-UAACCG UGUACCC 37

(2) INFORMATION FOR SEQ ID NO:�12:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:12: �

AGACGAUUAA CAUCCACGGA UGU-UAACGCG CUAGAA 36

(2) INFORMATION FOR SEQ ID NO:�13:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:13: �

AAGACGAUUA ACAAACACGU UUGU-UAACGC AACACCU 37

(2) INFORMATION FOR SEQ ID NO:�14:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:

49 50

EP 0 736 105 B1

28

5

10

15

20

25

30

35

40

45

50

55

14: �

GAUUGGAUUA ACAGGCACCC CUGU-UAACCU ACCACU 36

(2) INFORMATION FOR SEQ ID NO:�15:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:15: �

AGGAGGAUUA ACAACAAAGG UUGU-UAACCC CGUACCA 37

(2) INFORMATION FOR SEQ ID NO:�16:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 34 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:16: �

UGAAGGAUUA ACAACUAAUG UUGU-UAACCA UGUA 34

(2)�INFORMATION FOR SEQ ID NO:�17: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:17: �

UUGAGGAUUA ACAGGCACACCUGCUAACCG UGUACCC 37

(2) INFORMATION FOR SEQ ID NO:�18:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:18: �

AUGUGGCUUA ACAAGUACGC UUGU-UAACCC AAAAACG 37

(2) INFORMATION FOR SEQ ID NO:�19:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 35 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:19: �

AGGACGAUGA ACAAACACGU UUGUU-CACGC CAUGC 35

(2) INFORMATION FOR SEQ ID NO:�20:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 38 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:20: �

GACUGGCUUA ACAAACAUGU UUUGU-UAACC GUGUACCA 38

(2)�INFORMATION FOR SEQ ID NO:�21: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:21: �

CGGCGGAUUA ACACGACACA CUCGU-GUUAA CCAUAUC 37

(2) INFORMATION FOR SEQ ID NO:�22:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single

51 52

EP 0 736 105 B1

29

5

10

15

20

25

30

35

40

45

50

55

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:22: �

GCAUCAGAUG AACAGCACGU CUGUU-CACUA UGCACCC 37

(2) INFORMATION FOR SEQ ID NO:�23:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:23: �

GCAUCAGAUG AACAGCACGU CUGUU-CACUA UGCACCU 37

(2) INFORMATION FOR SEQ ID NO:�24:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:24: �

GCAUCAGAUG GACAGCACGU CUGUU-CACUA UGCACCU 37

(2) INFORMATION FOR SEQ ID NO:�25:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:25: �

CAGUGUAUGA AACACCACGU GUGUU-UCCAC UGUACCU 37

(2) INFORMATION FOR SEQ ID NO:�26:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 35 base pairs

(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:26: �

CAGUGUAUGA AACAACACGU UUGUU-UCCAC UGCCU 35

(2) INFORMATION FOR SEQ ID NO:�27:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 35 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:27: �

GAGUGUAUGA AACAACACGU UUGUU-UCCAC UCCCU 35

(2)�INFORMATION FOR SEQ ID NO:�28: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 35 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:28: �

GAGUGUAUGA AACAACACGU UUGUU-UCCAC UGUCU 35

(2) INFORMATION FOR SEQ ID NO:�29:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 35 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:29: �

GAUUGUAUGA AACAACGUGU UUGUU-UCCAC UCCCU 35

(2) INFORMATION FOR SEQ ID NO:�30:�

(i) SEQUENCE CHARACTERISTICS:�

53 54

EP 0 736 105 B1

30

5

10

15

20

25

30

35

40

45

50

55

(A) LENGTH: 35 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:30: �

GAAUGUAUGA AACAACACGU UUGUU-UCCAC UGCCU 35

(2) INFORMATION FOR SEQ ID NO:�31:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:31: �

GAUUGGACUU AACAGACACC CCUGU-UAACC UACCACU 37

(2) INFORMATION FOR SEQ ID NO:�32:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 34 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:32: �

UGCGACAGUU AGAAACACGA UUGUU-UACUG UAUG 34

(2) INFORMATION FOR SEQ ID NO:�33:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:33: �

UACAGGCUUA AGAAACACGU UUGU-UAACCA ACCCCU 36

(2) INFORMATION FOR SEQ ID NO:�34:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:34: �

UCGAGCAGUG UGAAACACGA UUGUG-UUUCC UGCUCA 36

(2) INFORMATION FOR SEQ ID NO:�35:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:35: �

UGAUGCCUAG AGAAACACAU UAGUG-UUUCC CUCUGU 36

(2) INFORMATION FOR SEQ ID NO:�36:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:36: �

ACGUGCCUCU AGAAACACAU CUGAU-GUUUC CCUCUCA 37

(2) INFORMATION FOR SEQ ID NO:�37:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:37: �

ACCCGCCUCG UGAAACACGC UUGAU-GUUUC CCUCUCA 37

55 56

EP 0 736 105 B1

31

5

10

15

20

25

30

35

40

45

50

55

(2)�INFORMATION FOR SEQ ID NO:�38: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 34 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:38: �

CGGUGACGUA UGAAACACGU UCGUU-GAUUU CCGU 34

(2) INFORMATION FOR SEQ ID NO:�39:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 30 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:39: �

GCUUGCGAAA CACGUUUGAC GUGU-UUCCCU 30

(2) INFORMATION FOR SEQ ID NO:�40:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 33 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:40: �

GCACCCUAGA AACGCGUUAG UA-GACGUUUC CCU 33

(2) INFORMATION FOR SEQ ID NO:�41:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:41: �

AGGAACCUAG AAACACACAG UGUUU-

CCCUC UGCCCAC 37

(2)�INFORMATION FOR SEQ ID NO:�42: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:42: �

GCCUGCAUGG AUUAACACGU AUGUG-UUAAC CGACUCC 37

(2) INFORMATION FOR SEQ ID NO:�43:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:43: �

UGAAACACUG AGAAACACGU GUUUC-CCCUU GUGUGAU 37

(2) INFORMATION FOR SEQ ID NO:�44:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:44: �

AGGAACCUCA AGCCGCCCCU AGAA-CACUCG GCACCU 36

(2) INFORMATION FOR SEQ ID NO:�45:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:45: �

57 58

EP 0 736 105 B1

32

5

10

15

20

25

30

35

40

45

50

55

AGGAACCUCA AGAAAGCCCC UGAAA-CACUC GAAGCCU 37

(2) INFORMATION FOR SEQ ID NO:�46:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:46: �

AGGAACCUCA AGAAACCCCC UGAAA-CACUC AUUACCG 37

(2) INFORMATION FOR SEQ ID NO:�47:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:47: �

AGGAACCUCA AGAAAUCCGA ACGA-CAACCC UACACCU 37

(2) INFORMATION FOR SEQ ID NO:�48:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 36 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:48: �

AGGAACCUCA AGAAACCCCG CCACG-GACCC CAACCA 36

(2)�INFORMATION FOR SEQ ID NO:�49: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:

49: �

GGGAACCUCA AUAAUCACGC ACGCA-UACUC GGCAUCU 37

(2) INFORMATION FOR SEQ ID NO:�50:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 34 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:50: �

GGGAACCUCA AGAGACCCGA CAGGA-UACUC GGAC 34

(2) INFORMATION FOR SEQ ID NO:�51:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:51: �

AAGUGGAACC UCAAUCCCGU AAGAA-GAUCC UGUACCU 37

(2) INFORMATION FOR SEQ ID NO:�52:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:52: �

AUGUGCAUAG AGAUGUACAU AUGAA-CCUC AGUAGAG 37

(2) INFORMATION FOR SEQ ID NO:�53:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

59 60

EP 0 736 105 B1

33

5

10

15

20

25

30

35

40

45

50

55

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:53: �

UCAUGCAUAG GCAUAGGCAG AUGGA-ACCUC AGUAGCC 37

(2) INFORMATION FOR SEQ ID NO:�54:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:54: �

AUGUGCAACA AGGCGCACGG AUAAG-GAACC UCGAAGU 37

(2)�INFORMATION FOR SEQ ID NO:�55: �

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 37 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:55: �

GAGUACAGCA CGCAACACGU ACG-GGGAACC UCAAAGU 37

(2) INFORMATION FOR SEQ ID NO:�56:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 20 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(ix) FEATURE:�

(D) OTHER INFORMATION: N at positions1 and 20 indicates 1-2 complementary basepairs

(ix) FEATURE:�

(D) OTHER INFORMATION: N at position3 indicates 1 or 3 nucleotides

(ix) FEATURE:�

(D) OTHER INFORMATION: N at positions10 and 12 indicates 1-4 complementarybase pairs

(ix) FEATURE: �

(D) OTHER INFORMATION: N at position11 indicates 4 or 5 nucleotides

(ix) FEATURE: �

(D) OTHER INFORMATION: U is iodouracil

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:56: �

NGNKDAACAN NNUGUUHMCN 20

(2) INFORMATION FOR SEQ ID NO:�57:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 23 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(ix) FEATURE: �

(D) OTHER INFORMATION: U is iodouracil

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:57: �

GGAACCUCAA UUGAUGGCCU UCC23

(2) INFORMATION FOR SEQ ID NO:�58:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 32 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(ix) FEATURE: �

(D) OTHER INFORMATION: U is iodouracil

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:58: �

GGGUGUAUGA AACACCACGU GUGU-UUCCAC UC 32

(2)�INFORMATION FOR SEQ ID NO:�59: �

61 62

EP 0 736 105 B1

34

5

10

15

20

25

30

35

40

45

50

55

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 29 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:59: �

GGGGAUUAAC AGGCACACCU GUUA-ACCCU 29

(2) INFORMATION FOR SEQ ID NO:�60:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 17 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:60: �

TAATACGACT CACTATA 17

(2) INFORMATION FOR SEQ ID NO:�61:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 49 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:61: �

GAGTGGAAAC ACACGTGGTG TTTCAT-ACAC CCTATAGTGA GTCGTATTA

49

(2) INFORMATION FOR SEQ ID NO:�62:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 46 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:62: �

AGGGTTAACA GGTGTGCCTG TTAATC-CCCT ATAGTGAGTC GTATTA 46

(2) INFORMATION FOR SEQ ID NO:�63:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 13 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:63: �

CTAGAGCCCG GGC 13

(2) INFORMATION FOR SEQ ID NO:�64:�

(i) SEQUENCE CHARACTERISTICS:�

(A) LENGTH: 44 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:64: �

CTAGAGCCCG GGCGAGCTCG AATTC-ATGGG TAATCCTCAT GTTC 44

Claims

1. A method for identifying nucleic acid ligands capableof covalently binding a target molecule upon irradi-ation, from a candidate mixture of nucleic acids, saidmethod comprising:�

(a) preparing a candidate mixture of nucleic ac-ids, each nucleic acid containing at least onephotoreactive group;(b) contacting said candidate mixture with saidtarget molecule, wherein nucleic acid sequenc-es having increased affinity to the target mole-cule relative to the candidate mixture form nu-cleic acid-�target molecule complexes;(c) irradiating said candidate mixture, whereinsome of said nucleic acid-�target molecule com-plexes covalently photocrosslink;(d) partitioning the crosslinked nucleic acid-�tar-get molecule complexes from free nucleic acidsin the candidate mixture; and(e) identifying the nucleic acid sequences thatphotocrosslinked to the target molecule.

2. A method according to claim 1, wherein a mixture ofnucleic acids is produced and which further compris-es after step (d):�

63 64

EP 0 736 105 B1

35

5

10

15

20

25

30

35

40

45

50

55

I. amplifying the nucleic acids that photo-crosslinked to the target molecule to yield a mix-ture of nucleic acids enriched in sequences thatare capable of photocrosslinking to the targetmolecule; andII. repeating steps (b)-(d) using the enriched mix-ture of each successive repeat.

3. A method according to claim 1, wherein a mixture ofnucleic acids is produced and which further compris-es after step (d):�

I. amplifying the nucleic acids that photo-crosslinked to the target molecule to yield a mix-ture of nucleic acids enriched in sequences thatare capable of photocrosslinking to the targetmolecule.

4. A method according to claim 1 wherein the targetmolecule is a protein and which further comprisesremoving the protein target molecule from the nucle-ic acid-�target molecule complexes by proteolytic di-gestion of the protein after step (d).

5. A method according to claim 2 or 3 wherein the targetmolecule is a protein, which further comprises re-moving the protein target molecule from the nucleicacid- �target molecule complexes by proteolytic diges-tion of the protein prior to amplifying the nucleic acidsthat photocrosslinked to the target molecule.

6. A method according to any preceding claim, whereinthe target molecule has a biological activity, whichis capable of modification by the nucleic acid ligand.

7. A method according to any preceding claim, whereinthe candidate mixture of nucleic acids in step (a) hasbeen produced by a process which comprises: �

(a’) preparing a candidate mixture of nucleic ac-ids, which contain photoreactive groups;(b’) contacting said candidate mixture with saidtarget molecule, wherein nucleic acid sequenc-es having an increased affinity to the target mol-ecule, relative to the candidate mixture, form nu-cleic acid-�target molecule complexes;(c’) irradiating said candidate mixture, whereinsaid nucleic acid-�target molecule complexesphotocrosslink;(d’) partitioning the increased affinity nucleic ac-ids from the remainder of the candidate mixture,or, when crosslinking has taken place in accord-ance with step (c’), � the crosslinked nucleic acid-target molecule complexes from free nucleic ac-ids in the candidate mixture; and(e’) amplifying the increased affinity nucleic ac-ids to yield a ligand- �enriched mixture of nucleicacids.

8. A method according to claim 2 wherein the irradiationtime is shortened in later cycles.

9. A method according to claim 1 wherein the candidatemixture of nucleic acids in step

(a) is produced by a process which comprises: �

(i) preparing a candidate mixture of nucleicacids of differing sequence each nucleic ac-id containing at least one photoreactivegroup;(ii) contacting the candidate mixture with theselected target under conditions favourablefor binding between the target and mem-bers of the candidate mixture;(iii) partitioning unbound nucleic acids fromthose nucleic acids which have bound to thetarget;(iv) amplifying the nucleic acids selected in(iii) as having the relatively higher affinity tothe target to create a new candidate mixturethat is enriched in nucleic acids having rel-atively higher affinity for the target;(v) reiterating steps (ii)-(iv) through as manycycles as desired.

10. A method for identifying nucleic acid ligands capableof covalently binding a target molecule upon irradi-ation, from a candidate mixture of nucleic acids, saidmethod comprising:�

(a) preparing a candidate mixture of nucleic ac-ids;(b) contacting said candidate mixture with saidtarget molecule, wherein nucleic acid sequenc-es having increased affinity to the target mole-cule relative to the candidate mixture form nu-cleic acid-�target molecule complexes;(c) partitioning the increased affinity nucleic ac-ids from the remainder of the candidate mixture;and(d) amplifying the increased affinity nucleic acidsto yield a ligand- �enriched mixture of nucleic ac-ids, whereby nucleic acid ligands of the targetmolecule may be identified;(e) incorporating at least one photoreactivegroup into said increased affinity nucleic acids;(f) repeating step (b);(g) irradiating the increased affinity nucleic ac-ids, wherein some of said nucleic acid-�targetmolecule complexes covalently photocrosslink;(h) repeating step (d) and (e).

11. The method of claim 10 wherein step (e) further com-prises limited randomisation of said increased affin-ity nucleic acids.

65 66

EP 0 736 105 B1

36

5

10

15

20

25

30

35

40

45

50

55

12. The method of claim 10 wherein step (e) further com-prises substitution of photoreactive nucleic acid nu-cleotides at specific sites in the ligand.

13. The method of claim 10 wherein said substitutionsare made to the ligand in areas of the molecule thatare known to interact with the target molecule.

14. The method of claim 10 wherein step (e) further com-prises partial random incorporation of the photore-active nucleotides.

15. A method according to any one of the precedingclaims wherein said photoreactive groups are select-ed from: �

5-�bromouracil, 5-�iodouracil, 5- �bromovinyluracil,5-�iodo- �vinyluracil, 5- �azidouracil, 4- �thiouracil, 5-bromocytosine, 5-�iodocytosine, 5-�bromovinyl-cytosine, 5- �iodovinylcytosine, 5- �azidocytosine,8-�azidoadenine, 8- �bromoadenine, 8-�iodoade-nine, 8-�azidoguanine, 8-�bromoguanine, 8-�io-doguanine, 8-�azidohypoxanthine, 8-�bromohy-poxanthine, 8-�iodohypoxanthine, 8-�azidoxan-thine, 8-�bromoxanthine, 8-�iodoxanthine, 5-�bro-modeoxyuracil, 8-�bromo-�2’- �deoxyadenine, 5-iodo-�2’-�deoxyuracil, 5-�iodo-�2’- �deoxycytosine,5-[(4-�azidophenacyl)�thio] �cytosine, 5-[(4- �azido-phenacyl)-thio]�uracil, 7- �deaza-�7- �iodoadenine,7-�deaza-�7-�iodoguanine, 7-�deaza- �7-�bromoade-nine and 7-�deaza-�7-�bromoguanine.

16. A method according to any one of claims 1 to 14wherein said photoreactive group absorbs light of awavelength that is not absorbed by the target.

17. A method according to any one of claims 1 to 14wherein said photoreactive group absorbs light of awavelength that is not absorbed by the non-�modifiedportions of the nucleic acid.

18. A method according to any one of the precedingclaims wherein said irradiation comprises ultravioletlight, visible light, X-�ray or gamma ray.

19. A method according to any one of the precedingclaims wherein said photoreactive group is 5-�iodou-racil and said nucleic acid is RNA and wherein lightin the range of 320-325nm is used for said irradiation.

20. A method according to any one of claims 1 to 18wherein said photoreactive group is 5- �iodouracil,said nucleic acid is RNA and said target is a protein,wherein light of 308nm is used for said irradiation.

21. A method according to any one of claims 1 to 18wherein said photoreactive group is 5-�bromouracilor 5-�iodouracil, said nucleic acid is RNA and said

target is a protein, wherein light in the range of350-400nm is used for said irradiation.

22. A method according to any one of the precedingclaims wherein said target molecule is not a nucleicacid binding protein.

23. A method according to any one of claims 1 to 21wherein said target molecule is selected from a pro-tein, peptide, carbohydrate, polysaccharide, glyco-protein, hormone, receptor, antigen, antibody, virus,substrate, metabolite, transition state analog, cofac-tor, inhibitor, drug, dye, nutrient, growth factor, path-ogen, toxic substance or biological effector.

24. A method according to any preceding claim whereinthe target molecule is specifically associated with adisease.

25. A method according to any one of the precedingclaims wherein said candidate mixture comprisessingle or double stranded RNA.

26. A method according to any one of claims 1 to 24wherein said candidate mixture comprises single ordouble stranded DNA.

27. A method according to any one of the precedingclaims wherein each nucleic acid of the candidatemixture is partially substituted, or fully substituted ateach position, by nucleotides containing a said pho-toreactive group.

28. A method according to any one of the precedingclaims wherein said candidate mixture is comprisedof oligonucleotides containing more than one typeof photoreactive group.

29. A method according to any one of the precedingclaims wherein said photoreactive group is a 5-halouridine, said method further comprising tran-scribing selected RNA sequences identified by saidmethod from a synthesised DNA template using 5-halouridine trisphosphate in place of uridine tri-sphosphate.

30. A method according to any one of claims 1 to 28wherein said method further comprises making andamplifying cDNA copies of the identified RNA se-quences.

31. A method according to claim 30 wherein said methodfurther comprises transcribing the amplified cDNAsequences into RNA sequences in the presence ofphotoreactive groups.

32. A method according to any one of the precedingclaims wherein nucleic acids of the candidate mix-

67 68

EP 0 736 105 B1

37

5

10

15

20

25

30

35

40

45

50

55

ture comprise modified nucleotides.

33. A method according to claim 32 wherein said mod-ification is selected from backbone modifications,methylation, modification of cytosine exocyclicamines, modification at the 2’-�position or substitutionwith halogenated groups.

34. A method according to claim 32 wherein said mod-ification is selected from 2’-�NH2, 2’- �F, 5-�bromo-uracil, 5-�iodo-�uracil, 2’-�NH2-iodouracil, 2’-�NH2-iodo-cytosine, 2’- NH2-iodoadenine, 2’-�NH2-bromouracil,2’- �NH2-bromocytosine and 2’-�NH2-bromoadenine.

35. An in vitro diagnostic method comprising the cova-lent photocrosslinking in vitro of a nucleic acid ligandcontaining at least one photoreactive group to a tar-get molecule, wherein said target molecule is not anucleic acid binding protein, and wherein the nucleicacid ligand has the property of binding specificallyto the target molecule, and of covalently binding thetarget molecule upon irradiation, and has been iden-tified by a process which comprises the steps of:�

(a) preparing a candidate mixture of nucleic ac-ids, each nucleic acid containing at least onephotoreactive group;(b) contacting said candidate mixture with saidtarget molecule, wherein nucleic acid sequenc-es having increased affinity to the target mole-cule relative to the candidate mixture form nu-cleic acid-�target molecule complexes;(c) irradiating said candidate mixture, whereinsome of said nucleic acid-�target molecule com-plexes covalently photocrosslink;(d) partitioning the crosslinked nucleic acid-�tar-get molecule complexes from free nucleic acidsin the candidate mixture; and(e) identifying the nucleic acid sequences thatphotocrosslinked to the target molecule.

36. A method according to claim 35 wherein said diag-nostic method comprises the diagnosis of a disease,pathological condition or toxic state.

37. A method according to claim 35 or 36 wherein saiddiagnostic method comprises the step of irradiatingthe nucleic acid ligand.

38. A method according to any one of claims 35 to 37wherein said diagnostic method comprises the stepof detecting the presence of a target protein by bind-ing a said RNA nucleic acid ligand to the protein andthen photocrosslinking said RNA nucleic acid ligandto the protein.

39. A method according to any one of claims 35 to 38wherein said ligand is bound to a support and is used

to covalently trap a target molecule.

40. A method according to any one of claims 35 to 37involving use of the nucleic acid ligand in in vitro im-aging wherein the nucleic acid ligand is capable ofcrosslinking to a marker protein on a tumor cell andis coupled to an enhancing agent.

41. A method according to any one of claims 35 to 40wherein said target molecule is selected from a pro-tein, peptide, carbohydrate, polysaccharide, glyco-protein, hormone, receptor, antigen, antibody, virus,substrate, metabolite, transition state analog, cofac-tor, inhibitor, drug, dye, nutrient, growth factor, path-ogen, toxic substance or biological effector.

42. A method according to any one of claims 35 to 41wherein said photoreactive groups are selectedfrom: �

5-�bromouracil, 5-�iodouracil, 5-�bromovinyluracil,5-�iodo-�vinyluracil, 5- �azidouracil, 4- �thiouracil, 5-bromocytosine, 5-�iodocytosine, 5-�bromovinyl-cytosine, 5- �iodovinylcytosine, 5- �azidocytosine,8-�azidoadenine, 8-�bromoadenine, 8-�iodoade-nine, 8-�azidoguanine, 8-�bromoguanine, 8-�io-doguanine, 8-�azidohypoxanthine, 8-�bromohy-poxanthine, 8-�iodohypoxanthine, 8-�azidoxan-thine, 8-�bromoxanthine, 8-�iodoxanthine, 5-�bro-modeoxyuracil, 8-�bromo-�2’- �deoxyadenine, 5-iodo-�2’-�deoxyuracil, 5-�iodo- �2’- �deoxycytosine,5-[(4-�azidophenacyl)�thio] �cytosine, 5-[(4- �azido-phenacyl)-thio]�uracil, 7- �deaza-�7- �iodoadenine,7-�deaza-�7-�iodoguanine, 7- �deaza- �7-�bromoade-nine and 7-�deaza-�7-�bromoguanine.

43. A method according to any one of claims 35 to 42wherein said ligand comprises single or doublestranded RNA.

44. A method according to any one of claims 35 to 42wherein said ligand comprises single or doublestranded DNA.

45. A method according to any one of claims 35 to 44wherein said ligand comprises modified nucleotides.

46. A method according to claim 45 wherein said mod-ification is selected from backbone modifications,methylation, modification of cytosine exocyclicamines, modification at the 2’-�position or substitutionwith halogenated groups.

47. A method according to claim 45 wherein said mod-ification is selected from 2’-�NH2, 2’- �F, 5-�bromo-uracil, 5-�iodo-�uracil, 2’-�NH2-iodauracil, 2’-�NH2-iodo-cytosine, 2’-�NH2- iodoadenine, 2’-�NH2-bromouracil,2’- �NH2-bromocytosione and 2’-�NH2-bromoadenine.

69 70

EP 0 736 105 B1

38

5

10

15

20

25

30

35

40

45

50

55

Patentansprüche

1. Verfahren zur Identifikation von Nucleinsäureligan-den, die bei Bestrahlung zur kovalenten Bindung ei-nes Zielmoleküls in der Lage sind, aus einem geeig-neten Nucleinsäuregemisch, wobei das VerfahrenFolgendes umfasst:�

(a) das Herstellen eines geeigneten Nucleinsäu-regemischs, wobei jede Nucleinsäure zumin-dest eine photoreaktive Gruppe enthält;(b) das Kontaktieren des geeigneten Gemischsmit dem Zielmolekül, worin die Nucleinsäurese-quenzen mit erhöhter Affinität zum Zielmolekülim Vergleich zum geeigneten Gemisch Nuclein-säure- �Zielmolekül- �Komplexe bilden;(c) das Bestrahlen des geeigneten Gemischs,worin sich einige der Nucleinsäure-�Zielmolekül-Komplexe kovalent photovernetzen;(d) das Abtrennen der vernetzten Nucleinsäure-Zielmolekül- �Komplexe von freien Nucleinsäu-ren im geeigneten Gemisch; und(e) das Identifizieren der Nucleinsäuresequen-zen, die sich mit dem Zielmolekül photovernetz-ten.

2. Verfahren nach Anspruch 1, worin ein Nucleinsäu-regemisch hergestellt wird und das nach Schritt (d)weiters Folgendes umfasst:�

I. das Amplifizieren der Nucleinsäuren, die sichmit dem Zielmolekül photovernetzten, um einNucleinsäuregemisch zu erhalten, das mit Se-quenzen angereichert ist, die zur Photovernet-zung mit dem Zielmolekül in der Lage sind; undII. das Wiederholen der Schritte (b)-(d) unterVerwendung des angereicherten Gemischs dereinzelnen aufeinander folgenden Wiederholun-gen.

3. Verfahren nach Anspruch 1, worin ein Nucleinsäu-regemisch hergestellt wird und das nach Schritt (d)weiters Folgendes umfasst:�

I. das Amplifizieren der Nucleinsäuren, die sichmit dem Zielmolekül photovernetzten, um einNucleinsäuregemisch zu erhalten, das mit Se-quenzen angereichert ist, die zur Photovernet-zung mit dem Zielmolekül in der Lage sind.

4. Verfahren nach Anspruch 1, worin das Zielmolekülein Protein ist und das weiters das Entfernen desProtein-�Zielmoleküls aus den Nucleinsäure- �Zielmo-lekül-�Komplexen durch proteolytischen Verdau desProteins nach Schritt (d) umfasst.

5. Verfahren nach Anspruch 2 oder 3, worin das Ziel-molekül ein Protein ist und das weiters das Entfernen

des Protein-�Zielmoleküls aus den Nucleinsäure-Zielmolekül-�Komplexen durch proteolytischen Ver-dau des Proteins vor der Amplifikation der Nuclein-säuren, die sich mit dem Zielmolekül photovernet-zen, umfasst.

6. Verfahren nach einem der vorangegangenen An-sprüche, worin das Zielmolekül eine biologische Ak-tivität aufweist, die zu einer Modifikation durch denNucleinsäureliganden in der Lage ist.

7. Verfahren nach einem der vorangegangenen An-sprüche, worin das geeignete Nucleinsäuregemischaus Schritt (a) durch ein Verfahren hergestellt wurde,das Folgendes umfasst: �

(a’) das Herstellen eines geeigneten Gemischsvon Nucleinsäuren, die photoreaktive Gruppenenthalten;(b’) das Kontaktieren des geeigneten Gemischsmit dem Zielmolekül, worin Nucleinsäurese-quenzen mit erhöhter Affinität zum Zielmolekülim Vergleich zum geeigneten Gemisch Nuclein-säure- �Zielmolekül-�Komplexe bilden;(c’) das Bestrahlen des geeigneten Gemischs,worin sich die Nucleinsäure-�Zielmolekül-�Kom-plexe photovernetzen;(d’) das Abtrennen der Nucleinsäuren mit erhöh-ter Affinität vom Rest des geeigneten Gemischsoder, wenn eine Vernetzung gemäß Schritt (c’)stattgefunden hat, der vernetzten Nucleinsäure-Zielmolekül-�Komplexe von freien Nucleinsäu-ren im geeigneten Gemisch; und(e’) das Amplifizieren der Nucleinsäuren mit er-höhter Affinität, um ein mit Liganden angerei-chertes Nucleinsäuregemisch zu erhalten.

8. Verfahren nach Anspruch 2, worin die Bestrahlungs-dauer in späteren Durchgängen kürzer ist.

9. Verfahren nach Anspruch 1, worin das geeigneteNucleinsäuregemisch aus Schritt (a) durch ein Ver-fahren hergestellt ist, das Folgendes umfasst:�

(i) das Herstellen eines geeigneten Gemischsvon Nucleinsäuren mit unterschiedlicher Se-quenz, wobei jede Nucleinsäure zumindest einephotoreaktive Gruppe umfasst;(ii) das Kontaktieren des geeigneten Gemischsmit dem gewählten Ziel unter Bedingungen, diefür die Bindung zwischen dem Ziel und Elemen-ten des geeigneten Gemischs vorteilhaft sind;(iii) das Abtrennen ungebundener Nucleinsäu-ren von jenen Nucleinsäuren, die an das Zielgebunden haben;(iv) das Amplifizieren der in (iii) als solche mitvergleichsweise höherer Affinität zum Ziel aus-gewählten Nucleinsäuren, um ein neues geeig-

71 72

EP 0 736 105 B1

39

5

10

15

20

25

30

35

40

45

50

55

netes Gemisch herzustellen, das mit Nuclein-säuren mit vergleichsweise höherer Affinität fürdas Ziel angereichert ist;(v) das Wiederholen der Schritte (ii)-(iv) für einegewünschte Anzahl an Durchgängen.

10. Verfahren zur Identifikation von Nucleinsäureligan-den, die bei Bestrahlung zur kovalenten Bindung ei-nes Zielmoleküls in der Lage sind, aus einem geeig-neten Nucleinsäuregemisch, wobei das VerfahrenFolgendes umfasst:�

(a) das Herstellen eines geeigneten Nucleinsäu-regemischs;(b) das Kontaktieren des geeigneten Gemischsmit dem Zielmolekül, worin die Nucleinsäurese-quenzen mit erhöhter Affinität zum Zielmolekülim Vergleich zum geeigneten Gemisch Nuclein-säure- �Zielmolekül- �Komplexe bilden;(c) das Abtrennen der Nucleinsäuren mit erhöh-ter Affinität vom Rest des geeigneten Gemischs;und(d) das Amplifizieren der Nucleinsäuren mit er-höhter Affinität, um ein mit Liganden angerei-chertes Nucleinsäuregemisch zu erhalten, wo-durch Nucleinsäureliganden des Zielmolekülsidentifiziert werden können;(e) das Inkorporieren zumindest einer photore-aktiven Gruppe in die Nucleinsäuren mit erhöh-ter Affinität;(f) das Wiederholen von Schritt (b);(g) das Bestrahlen der Nucleinsäuren mit erhöh-ter Affinität, worin sich einige der Nucleinsäure-Zielmolekül- �Komplexe kovalent photovernet-zen;(h) das Wiederholen der Schritte (d) und (e).

11. Verfahren nach Anspruch 10, worin Schritt (e) wei-ters eine begrenzte Randomisierung der Nuclein-säuren mit erhöhter Affinität umfasst.

12. Verfahren nach Anspruch 10, worin Schritt (e) wei-ters die Substitution von photoreaktiven Nucleinsäu-re-�Nucleotiden an bestimmten Stellen im Ligandenumfasst.

13. Verfahren nach Anspruch 10, worin die Substitutio-nen am Liganden in Bereichen des Moleküls durch-geführt werden, die bekannterweise mit dem Ziel-molekül wechselwirken.

14. Verfahren nach Anspruch 10, worin Schritt (e) wei-ters eine partielle Zufallsinkorporation der photore-aktiven Nucleotide umfasst.

15. Verfahren nach einem der vorangegangenen An-sprüche, worin die photoreaktiven Gruppen ausge-wählt sind aus:�

5-�Bromuracil, 5-�Ioduracil, 5-�Bromvinyluracil, 5-Iodvinyluracil, 5-�Azidouracil, 4- �Thiouracil, 5-Bromcytosin, 5- �Iodcytosin, 5- �Bromvinylcytosin,5-�Iodvinylcytosin, 5- �Azidocytosin, 8-�Azidoade-nin, 8-�Bromadenin, 8- �Iodadenin, 8-�Azidogua-nin, 8-�Bromguanin, 8-�Iodguanin, 8-�Azidohypo-xanthin, 8-�Bromhypoxanthin, 8-�Iodhypoxan-thin, 8- �Azidoxanthin, 8- �Bromxanthin, 8-�Iodxan-thin, 5-�Bromdesoxyuracil, 8-�Brom- �2’-�desoxya-denin, 5-�Iod-�2’- �desoxyuracil, 5-�Iod-�2’- �desoxy-cytosin, 5-[(4-�Azidophenacyl)�thio]�cytosin, 5-[(4-Azidophenacyl)�thio]�uracil, 7- �Deaza-�7- �iodade-nin, 7-�Deaza-�7-�iodguanin, 7-�Deaza- �7- �broma-denin und 7-�Deaza- �7-�bromguanin.

16. Verfahren nach einem der Ansprüche 1 bis 14, worindie photoreaktive Gruppe Licht mit einer Wellenlän-ge absorbiert, die vom Ziel nicht absorbiert wird.

17. Verfahren nach einem der Ansprüche 1 bis 14, worindie photoreaktive Gruppe Licht mit einer Wellenlän-ge absorbiert, die von den nichtmodifizierten Ab-schnitten der Nucleinsäure nicht absorbiert werden.

18. Verfahren nach einem der vorangegangenen An-sprüche, worin die Bestrahlung Ultraviolettlicht,sichtbares Licht, Röntgenstrahlen oder Gamma-Strahlen umfasst.

19. Verfahren nach einem der vorangegangenen An-sprüche, worin die photoreaktive Gruppe 5- �Ioduracilist und die Nucleinsäure RNA ist und worin Licht imBereich von 320-325 nm für die Bestrahlung verwen-det wird.

20. Verfahren nach einem der Ansprüche 1 bis 18, worindie photoreaktive Gruppe 5-�Ioduracil ist, die Nuclein-säure RNA ist und das Ziel ein Protein ist, worin Lichtmit 308 nm für die Bestrahlung verwendet wird.

21. Verfahren nach einem der Ansprüche 1 bis 18, worindie photoreaktive Gruppe 5-�Bromuracil oder 5-Ioduracil ist, die Nucleinsäure RNA ist und das Zielein Protein ist, worin Licht im Bereich von 350-400nm für die Bestrahlung verwendet wird.

22. Verfahren nach einem der vorangegangenen An-sprüche, worin das Zielmolekül kein Nucleinsäurebindendes Protein ist.

23. Verfahren nach einem der Ansprüche 1 bis 21, worindas Zielmolekül aus einem Protein, Peptid, Kohlen-hydrat, Polysaccharid, Glykoprotein, Hormon, Re-zeptor, Antigen, Antikörper, Virus, Substrat, Meta-bolit, Übergangszustandsanalogon, Cofaktor, Inhi-bitor, Arzneimittel, Farbstoff, Nährstoff, Wachstums-faktor, Pathogen, toxischen Stoff oder biologischenEffektor ausgewählt ist.

73 74

EP 0 736 105 B1

40

5

10

15

20

25

30

35

40

45

50

55

24. Verfahren nach einem der vorangegangenen An-sprüche, worin das Zielmolekül spezifisch mit einerKrankheit assoziiert ist.

25. Verfahren nach einem der vorangegangenen An-sprüche, worin das geeignete Gemisch einzel- oderdoppelsträngige RNA umfasst.

26. Verfahren nach einem der Ansprüche 1 bis 24, worindas geeignete Gemisch einzel- oder doppelsträngi-ge DNA umfasst.

27. Verfahren nach einem der vorangegangenen An-sprüche, worin jede Nucleinsäure des geeignetenGemischs mit Nucleotiden, die eine der genanntenphotoreaktiven Gruppen enthalten, teilweise substi-tuiert oder an jeder Position vollständig substituiertist.

28. Verfahren nach einem der vorangegangenen An-sprüche, worin das geeignete Gemisch aus Oligonu-cleotiden besteht, die mehr als einen Typ photore-aktiver Gruppen enthalten.

29. Verfahren nach einem der vorangegangenen An-sprüche, worin die photoreaktive Gruppe 5-�Halogen-uridin ist, wobei das Verfahren weiters das Transkri-bieren ausgewählter RNA-�Sequenzen, die durchdas Verfahren identifiziert wurden, aus einer synthe-tisierten DNA- �Matrize unter Verwendung von 5-�Ha-logenuridintrisphosphat anstelle von Uridintrisphos-phat umfasst.

30. Verfahren nach einem der Ansprüche 1 bis 28, worindas Verfahren weiters das Herstellen und Amplifi-zieren von cDNA-�Kopien der identifizierten RNA-�Se-quenzen umfasst.

31. Verfahren nach Anspruch 30, worin das Verfahrenweiters das Transkribieren der amplifizierten cDNA-Sequenzen in RNA-�Sequenzen in Gegenwart vonphotoreaktiven Gruppen umfasst.

32. Verfahren nach einem der vorangegangenen An-sprüche, worin Nucleinsäuren des geeigneten Ge-misches modifizierte Nucleotide umfassen.

33. Verfahren nach Anspruch 32, worin die Modifikationaus Rückgrat-�Modifikationen, Methylierung, Modifi-kation von exozyklischen Cytosinaminen, Modifika-tion an der 2’-�Position oder Substitution mit haloge-nierten Gruppen ausgewählt ist.

34. Verfahren nach Anspruch 32, worin die Modifikationaus 2’-�NH2, 2’- �F, 5-�Bromuracil, 5-�Ioduracil, 2’-NH2-Ioduracil, 2’-�NH2-Iodcytosin, 2’- �NH2-Iodade-nin, 2’-�NH2-Bromuracil, 2’-�NH2-Bromcytosin und 2’-NH2-Bromadenin ausgewählt ist.

35. In-�vitro-�Diagnoseverfahren, das die kovalente Pho-tovernetzung eines Nucleinsäureliganden, der zu-mindest eine photoreaktive Gruppe enthält, in vitroan ein Zielmolekül umfasst, worin das Zielmolekülkein Nucleinsäure bindendes Protein ist und worinder Nucleinsäureligand die Fähigkeit aufweist, spe-zifisch an das Zielmolekül zu binden und das Ziel-molekül bei Bestrahlung kovalent zu binden, unddurch ein Verfahren identifiziert wurde, das folgendeSchritte umfasst:�

(a) das Herstellen eines geeigneten Nucleinsäu-regemischs, wobei jede Nucleinsäure zumin-dest eine photoreaktive Gruppe enthält;(b) das Kontaktieren des geeigneten Gemischsmit dem Zielmolekül, worin die Nucleinsäurese-quenzen mit erhöhter Affinität zum Zielmolekülim Vergleich zum geeigneten Gemisch Nuclein-säure- �Zielmolekül-�Komplexe bilden;(c) das Bestrahlen des geeigneten Gemischs,worin sich einige der Nucleinsäure-�Zielmolekül-Komplexe kovalent photovernetzen;(d) das Abtrennen der vernetzten Nucleinsäure-Zielmolekül-�Komplexe von freien Nucleinsäu-ren im geeigneten Gemisch; und(e) das Identifizieren der Nucleinsäuresequen-zen, die sich mit dem Zielmolekül photovernetz-ten.

36. Verfahren nach Anspruch 35, worin das Diagnose-verfahren die Diagnose einer Krankheit, eines pa-thologischen Zustands oder eines toxischen Zu-stands umfasst.

37. Verfahren nach Anspruch 35 oder 36, worin das Dia-gnoseverfahren den Schritt des Bestrahlens des Nu-cleinsäureliganden umfasst.

38. Verfahren nach einem der Ansprüche 35 bis 37, wor-in das Diagnoseverfahren den Schritt des Nachwei-sens der Gegenwart eines Zielproteins durch Bin-dung eines der genannten RNA-�Nucleinsäureligan-den an das Protein und dann Photovernetzung desRNA-�Nucleinsäureliganden mit dem Protein um-fasst.

39. Verfahren nach einem der Ansprüche 35 bis 38, wor-in der Ligand an einen Träger gebunden ist und ver-wendet wird, um ein Zielmolekül kovalent einzufan-gen.

40. Verfahren nach einem der Ansprüche 35 bis 37, dasdie Verwendung des Nucleinsäureliganden für In-vitro-�Bildgebung umfasst, worin der Nucleinsäureli-gand in der Lage ist, sich mit einem Markerproteinauf einer Tumorzelle zu vernetzen, und an einen Ver-stärker gebunden ist.

75 76

EP 0 736 105 B1

41

5

10

15

20

25

30

35

40

45

50

55

41. Verfahren nach einem der Ansprüche 35 bis 40, wor-in das Zielmolekül aus einem Protein, Peptid, Koh-lenhydrat, Polysaccharid, Glykoprotein, Hormon,Rezeptor, Antigen, Antikörper, Virus, Substrat, Me-tabolit, Übergangszustandsanalogon, Cofaktor, In-hibitor, Arzneimittel, Farbstoff, Nährstoff, Wachs-tumsfaktor, Pathogen, toxischen Stoff oder biologi-schen Effektor ausgewählt ist.

42. Verfahren nach einem der Ansprüche 35 bis 41, wor-in die photoreaktiven Gruppen ausgewählt sind aus:�

5-�Bromuracil, 5-�Ioduracil, 5-�Bromvinyluracil, 5-Iodvinyluracil, 5-�Azidouracil, 4- �Thiouracil, 5-Bromcytosin, 5- �Iodcytosin, 5- �Bromvinylcytosin,5-�Iodvinylcytosin, 5- �Azidocytosin, 8-�Azidoade-nin, 8-�Bromadenin, 8-�Iodadenin, 8-�Azidogua-nin, 8-�Bromguanin, 8-�Iodguanin, 8-�Azidohypo-xanthin, 8-�Bromhypoxanthin, 8-�Iodhypoxan-thin, 8- �Azidoxanthin, 8- �Bromxanthin, 8-�Iodxan-thin, 5-�Bromdesoxyuracil, 8-�Brom- �2’-�desoxya-denin, 5-�Iod-�2’- �desoxyuracil, 5-�Iod-�2’- �desoxy-cytosin, 5-[(4-�Azidophenacyl)�thio]�cytosin, 5-[(4-Azidophenacyl)�thio]�uracil, 7- �Deaza-�7- �iodade-nin, 7-�Deaza-�7-�iodguanin, 7-�Deaza- �7- �broma-denin und 7-�Deaza- �7-�bromguanin.

43. Verfahren nach einem der Ansprüche 35 bis 42, wor-in der Ligand einzel- oder doppelsträngige RNA um-fasst.

44. Verfahren nach einem der Ansprüche 35 bis 42, wor-in der Ligand einzel- oder doppelsträngige DNA um-fasst.

45. Verfahren nach einem der Ansprüche 35 bis 44, wor-in der Ligand modifizierte Nucleotide umfasst.

46. Verfahren nach Anspruch 45, worin die Modifikationaus Rückgrat-�Modifikationen, Methylierung, Modifi-kation von exozyklischen Cytosinaminen, Modifika-tion an der 2’-�Position oder Substitution mit haloge-nierten Gruppen ausgewählt ist.

47. Verfahren nach Anspruch 45, worin die Modifikationaus 2’-�NH2, 2’- �F, 5-�Bromuracil, 5-�Ioduracil, 2’-NH2-Ioduracil, 2’-�NH2-Iodcytosin, 2’- �NH2-Iodade-nin, 2’-�NH2-Bromuracil, 2’-�NH2-Bromcytosin und 2’-NH2-Bromadenin ausgewählt ist.

Revendications

1. Procédé pour identifier des ligands acide nucléiquecapables de lier par covalence une molécule ciblelors d’une irradiation, dans un mélange candidatd’acides nucléiques, ledit procédé comprenant lesétapes consistant à: �

(a) préparer un mélange candidat d’acides nu-cléiques, chaque acide nucléique contenant aumoins un groupe photoréactif;(b) mettre en contact ledit mélange candidatavec ladite molécule cible, lesdites séquencesd’acide nucléique ayant une affinité accrue pourla molécule cible relativement au mélange can-didat formant des complexes acide nucléique -molécule cible;(c) irradier ledit mélange candidat, certains descomplexes acide nucléique - molécule cible su-bissant une photo-�réticulation covalente;(d) séparer les complexes réticulés acide nucléi-que - molécule cible des acides nucléiques li-bres dans le mélange candidat; et(e) identifier les séquences d’acide nucléiquequi ont subi une photo-�réticulation à la moléculecible.

2. procédé selon la revendication 1, dans lequel un mé-lange d’acides nucléiques est produit et qui com-prend en outre après l’étape (d):�

I. l’amplification des acides nucléiques qui ontsubi une photo-�réticulation à la molécule ciblepour donner un mélange d’acides nucléiquesenrichi en séquences qui sont capables de subirune photo-�réticulation à la molécule cible; etII. la répétition des étapes (b)-(d) en utilisant lemélange enrichi de chaque répétition successi-ve.

3. Procédé selon la revendication 1, dans lequel un mé-lange d’acides nucléiques est produit et qui com-prend en outre après l’étape (d):�

I. l’amplification des acides nucléiques qui ontsubi une photo-�réticulation à la molécule ciblepour donner un mélange d’acides nucléiquesenrichi en séquences qui sont capables de subirune photo-�réticulation à la molécule cible.

4. Procédé selon la revendication 1 dans lequel la mo-lécule cible est une protéine et qui comprend en outrel’élimination de la protéine molécule cible du mélan-ge de complexes acide nucléique - molécule ciblepar digestion protéolytique de la protéine après l’éta-pe (d).

5. Procédé selon la revendication 2 ou 3 dans lequella molécule cible est une protéine, qui comprend enoutre l’élimination de la protéine molécule cible dumélange de complexes acide nucléique - moléculecible par digestion protéolytique de la protéine avantl’amplification des acides nucléiques qui ont subi unephoto-�réticulation à la molécule cible.

6. Procédé selon l’une quelconque des revendications

77 78

EP 0 736 105 B1

42

5

10

15

20

25

30

35

40

45

50

55

précédentes, dans lequel la molécule cible a uneactivité biologique, qui est capable de modificationpar le ligand acide nucléique.

7. Procédé selon l’une quelconque des revendicationsprécédentes, dans lequel le mélange candidat d’aci-des nucléiques à l’étape (a) a été produit par un pro-cessus qui comprend les étapes consistant à:�

(a’) préparer un mélange candidat d’acides nu-cléiques, qui contient des groupes photoréac-tifs;(b’) mettre en contact ledit mélange candidatavec ladite molécule cible, lesdites séquencesd’acide nucléique ayant une affinité accrue pourla molécule cible relativement au mélange can-didat formant des complexes acide nucléique -molécule cible;(c’) irradier ledit mélange candidat, lesdits com-plexes acide nucléique - molécule cible subis-sant une photo-�réticulation;(d’) séparer les acides nucléiques d’affinité ac-crue du reste du mélange candidat ou, lorsquela réticulation a eu lieu conformément à l’étape(c’), les complexes acide nucléique - moléculecible réticulés, des acides nucléiques libresdans le mélange candidat; et(e’) amplifier les acides nucléiques d’affinité ac-crue pour donner un mélange d’acides nucléi-ques enrichi en ligand.

8. Procédé selon la revendication 2 dans lequel la du-rée d’irradiation est raccourcie dans les cycles ulté-rieurs.

9. Procédé selon la revendication 1 dans lequel le mé-lange candidat d’acides nucléiques à l’étape (a) estproduit par un processus qui comprend les étapesconsistant à:�

(i) préparer un mélange candidat d’acides nu-cléiques de séquence différente, chaque acidenucléique contenant au moins un groupe pho-toréactif;(ii) mettre en contact le mélange candidat avecla cible sélectionnée dans des conditions favo-rables à la liaison entre la cible et des élémentsdu mélange candidat;(iii) séparer les acides nucléiques non liés desacides nucléiques qui se sont liés à la cible;(iv) amplifier les acides nucléiques sélectionnésen (iii) comme ayant l’affinité relativement supé-rieure pour la cible pour créer un nouveau mé-lange candidat qui est enrichi en acides nucléi-ques ayant une affinité relativement supérieurepour la cible;(v) répéter les étapes (ii)-(iv) autant de fois quedésiré.

10. Procédé pour identifier des ligands acides nucléi-ques capables de se lier par covalence à une molé-cule cible lors de l’irradiation, dans un mélange d’aci-des nucléiques, ledit procédé comprenant les étapesconsistant à:�

(a) préparer un mélange candidat d’acides nu-cléiques;(b) mettre en contact ledit mélange candidatavec ladite molécule cible, lesdites séquencesd’acide nucléique ayant une affinité accrue pourla molécule cible, relativement au mélange can-didat, formant des complexes acides nucléique- molécule cible;(c) séparer les acides nucléiques d’affinité ac-crue du reste du mélange candidat; et(d) amplifier les acides nucléiques d’affinité ac-crue pour donner un mélange d’acides nucléi-ques enrichi en ligand, les ligands acide nucléi-que de la molécule cible pouvant ainsi être iden-tifiés;(e) incorporer au moins un groupe photoréactifdans lesdits acides nucléiques d’affinité accrue;(f) répéter l’étape (b);(g) irradier les acides nucléiques d’affinité ac-crue, certains desdits complexes acide nucléi-que - molécule cible subissant une photo-�réti-culation covalente;(h) répéter les étapes (d) et (e).

11. Procédé selon la revendication 10 dans lequel l’éta-pe (e) comprend en outre une randomisation limitéedesdits acides nucléiques d’affinité accrue.

12. Procédé selon la revendication 10 dans lequel l’éta-pe (e) comprend en outre la substitution de nucléo-tides d’acide nucléique photoréactifs dans des sitesspécifiques dans le ligand.

13. Procédé selon la revendication 10 dans lequel les-dites substitutions sont faites dans le ligand dansdes zones de la molécule qui sont connues pour in-teragir avec la molécule cible.

14. Procédé selon la revendication 10 dans lequel l’éta-pe (e) comprend en outre l’incorporation aléatoirepartielle des nucléotides photoréactifs.

15. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel lesdits groupes photoréac-tifs sont choisis parmi: �

le 5- �bromouracile, le 5- �iodouracyle, le 5-�bromo-vinyluracile, le 5-�iodo- �vinyluracile, le 5- �azidou-racile, le 4-�thiouracile, le 5-�bromocytosine, le 5-iodocytosine, le 5-�bromovinylcytosine, le 5- �io-dovinylcytosine, le 5-�azidocytosine, le 8-�azidoa-dénine, le 8-�bromoadénine, le 8-�iodoadénine,

79 80

EP 0 736 105 B1

43

5

10

15

20

25

30

35

40

45

50

55

le 8- �azidoguanine, le 8-�bromoguanine, le 8-�io-doguanine, le 8-�azidohypoxanthine, le 8- �bromo-hypoxanthine, le 8-�iodohypoxanthine, le 8-�azi-doxanthine, le 8-�bromoxanthine, le 8-�iodoxan-thine, le 5- �bromodésoxyuracile, le 8-�bromo-�2’-désoxyadénine, le 5- �iodo- �2’- �désoxyuracile, le 5-iodo-�2’- �désoxycytosine, le 5-[(4-�azidophénacyl)thio]�cytosine, le 5-[(4-�azidophénacyl)-thio]�ura-cile, le 7-�déaza-�7-�iodoadénine, le 7-�déaza-�7-�io-doguanine, le 7-�déaza- �7- �bromoadénine et le 7-déaza-�7-�bromoguanine.

16. Procédé selon l’une quelconque des revendications1 à 14 dans lequel ledit groupe photoréactif absorbela lumière d’une longueur d’onde qui n’est pas ab-sorbée par la cible.

17. Procédé selon l’une quelconque des revendications1 à 14 dans lequel ledit groupe photoréactif absorbela lumière d’une longueur d’onde qui n’est pas ab-sorbée par les parties non modifiées de l’acide nu-cléique.

18. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel l’irradiation comprend la lu-mière ultraviolette, la lumière visible, les rayons Xou les rayons gamma.

19. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel ledit groupe photoréactifest le 5-�iodouracile et ledit acide nucléique est l’ARNet dans lequel la lumière dans la plage de 320-325nm est utilisée pour ladite irradiation.

20. Procédé selon l’une quelconque des revendications1 à 18 dans lequel ledit groupe photoréactif est le 5-iodouracile, ledit acide nucléique est de l’ARN et la-dite cible est une protéine, la lumière de 308 nm étantutilisée pour ladite irradiation.

21. Procédé selon l’une quelconque des revendications1 à 18 dans lequel ledit groupe photoréactif est le 5-bromouracile ou le 5- �iodouracile, ledit acide nucléi-que est de l’ARN et ladite cible est une protéine, lalumière dans la plage de 350-400 nm étant utiliséepour ladite irradiation.

22. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel ladite molécule cible n’estpas une protéine de liaison à l’acide nucléique.

23. Procédé selon l’une quelconque des revendications1 à 21 dans lequel ladite molécule cible est choisieparmi une protéine, un peptide, un hydrate de car-bone, un polysaccharide, une glycoprotéine, unehormone, un récepteur, un antigène, un anticorps,un virus, un substrat, un métabolite, un analogued’état de transition, un cofacteur, un inhibiteur, un

médicament, un colorant, un nutriment, un facteurde croissance, un agent pathogène, une substancetoxique ou un effecteur biologique.

24. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel la molécule cible est spé-cifiquement associée à une maladie.

25. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel ledit mélange candidatcomprend de l’ARN monocaténaire ou bicaténaire.

26. Procédé selon l’une quelconque des revendications1 à 24 dans lequel ledit mélange candidat comprendde l’ADN monocaténaire ou bicaténaire.

27. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel chaque acide nucléique dumélange candidat est partiellement substitué, oucomplètement substitué en chaque position, par desnucléotides contenant un dit groupe photoréactif.

28. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel ledit mélange candidat estconstitué d’oligonucléotides contenant plus d’un ty-pe de groupe photoréactif.

29. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel ledit groupe photoréactifest la 5-�halouridine, ledit procédé comprenant enoutre la transcription de séquences d’ARN sélection-nées identifiées par ledit procédé à partir d’une ma-trice d’ADN synthétisée en utilisant le triphosphatede 5-�halouridine au lieu du triphosphate d’uridine.

30. Procédé selon l’une quelconque des revendications1 à 28 dans lequel ledit procédé comprend en outrela fabrication et l’amplification d’exemplaires d’ADNcdes séquences d’ARN identifiées.

31. Procédé selon la revendication 30 dans lequel leditprocédé comprend en outre la transcription des sé-quences d’ADNc amplifié en séquences d’ARN enprésence de groupes photoréactifs.

32. Procédé selon l’une quelconque des revendicationsprécédentes dans lequel les acides nucléiques dumélange candidat comprennent des nucléotides mo-difiés.

33. Procédé selon la revendication 32 dans lequel laditemodification est sélectionnée parmi des modifica-tions du squelette, la méthylation, la modificationd’amines exocycliques de cytosine, la modificationen position 2’ ou la substitution avec des groupeshalogénés.

34. Procédé selon la revendication 32 dans lequel ladite

81 82

EP 0 736 105 B1

44

5

10

15

20

25

30

35

40

45

50

55

modification est choisie parmi 2’-�NH2, 2’-�F, 5- �bro-mouracile, 5- �iodo-�uracile, 2’-�NH2-iodouracile, 2’-NH2-iodocytosine, 2’-�NH2-iodoadénine, 2’-�NH2-bro-mouracile, 2’- �NH2-bromocytosine et 2’-�NH2-bro-moadénine.

35. Procédé de diagnostic in vitro comprenant la photo-réticulation covalente in vitro d’un ligand acide nu-cléique contenant au moins un groupe photoréactifà une molécule cible, ladite molécule cible n’étantpas une protéine de liaison à l’acide nucléique et leligand acide nucléique ayant la propriété de se lierspécifiquement à la molécule cible et de lier par co-valence la molécule cible lors de l’irradiation, et ayantété identifié par un processus qui comprend les éta-pes consistant à: �

(a) préparer un mélange candidat d’acides nu-cléiques, chaque acide nucléique contenant aumoins un groupe photoréactif;(b) mettre en contact ledit mélange candidatavec ladite molécule cible, lesdites séquencesd’acide nucléique ayant une affinité accrue pourla molécule cible, relativement au mélange can-didat, formant des complexes acide nucléique -molécule cible;(c) irradier ledit mélange candidat, certains descomplexes acide nucléique - molécule cible su-bissant une photo-�réticulation covalente;(d) séparer les complexes réticulés acide nucléi-que - molécule cible des acides nucléiques li-bres dans le mélange candidat; et(e) identifier les séquences d’acide nucléiquequi ont subi une photo-�réticulation à la moléculecible.

36. Procédé selon la revendication 35 dans lequel leditprocédé de diagnostic comprend le diagnostic d’unemaladie, d’un état pathologique ou d’un état toxique.

37. Procédé selon la revendication 35 ou 36 dans lequelledit procédé diagnostique comprend l’étape consis-tant à irradier le ligand acide nucléique.

38. Procédé selon l’une quelconque des revendications35 à 37 dans lequel ledit procédé diagnostique com-prend l’étape consistant à détecter la présence d’uneprotéine cible en liant un dit ligand acide nucléiqueARN à la protéine puis en photoréticulant ledit ligandacide nucléique ARN à la protéine.

39. Procédé selon l’une quelconque des revendications35 à 38 dans lequel ledit ligand est lié à un supportet est utilisé pour piéger de manière covalente unemolécule cible.

40. Procédé selon l’une quelconque des revendications35 à 37 impliquant l’utilisation du ligand acide nucléi-

que dans l’imagerie in vitro, le ligand acide nucléiqueétant capable de réticulation à une protéine mar-queur sur une cellule tumorale et étant couplé à unagent de contraste.

41. Procédé selon l’une quelconque des revendications35 à 40 dans lequel ladite molécule cible est choisieparmi une protéine, un peptide, un hydrate de car-bone, un polysaccharide, une glycoprotéine, unehormone, un récepteur, un antigène, un anticorps,un virus, un substrat, un métabolite, un analogued’état de transition, un cofacteur, un inhibiteur, unmédicament, un colorant, un nutriment, un facteurde croissance, un agent pathogène, une substancetoxique ou un effecteur biologique.

42. Procédé selon l’une quelconque des revendications35 à 41 dans lequel lesdits groupes photoréactifssont choisis parmi: �

le 5- �bromouracile, le 5- �iodouracyle, le 5-�bromo-vinyluracile, le 5-�iodo- �vinyluracile, le 5- �azidou-racile, le 4-�thiouracile, le 5-�bromocytosine, le 5-iodocytosine, le 5-�bromovinylcytosine, le 5- �io-dovinylcytosine, le 5-�azidocytosine, le 8-�azidoa-dénine, le 8-�bromoadénine, le 8-�iodoadénine,le 8- �azidoguanine, le 8-�bromoguanine, le 8-�io-doguanine, le 8-�azidohypoxanthine, le 8- �bromo-hypoxanthine, le 8-�iodohypoxanthine, le 8-�azi-doxanthine, le 8-�bromoxanthine, le 8-�iodoxan-thine, le 5-�bromodésoxyuracile, le 8-�bromo-�2’-désoxyadénine, le 5- �iodo- �2’- �désoxyuracile, le 5-iodo-�2’- �désoxycytosine, le 5-[(4-�azidophénacyl)thio]�cytosine, le 5-[(4-�azidophénacyl)-thio]�ura-cile, le 7-�déaza-�7-�iodoadénine, le 7-�déaza-�7-�io-doguanine, le 7-�déaza- �7- �bromoadénine et le 7-déaza-�7-�bromoguanine.

43. Procédé selon l’une quelconque des revendications35 à 42 dans lequel ledit ligand comprend de l’ARNmonocaténaire ou bicaténaire.

44. Procédé selon l’une quelconque des revendications35 à 42 dans lequel ledit ligand comprend de l’ADNmonocaténaire ou bicaténaire.

45. Procédé selon l’une quelconque des revendications35 à 44 dans lequel ledit ligand comprend des nu-cléotides modifiés.

46. Procédé selon la revendication 45 dans lequel laditemodification est choisie parmi des modifications dusquelette, la méthylation, la modification d’aminesexocycliques de cytosine, la modification en position2’ ou la substitution avec des groupes halogénés.

47. Procédé selon la revendication 45 dans lequel laditemodification est choisie parmi 2’-�NH2, 2’-�F, 5- �bro-

83 84

EP 0 736 105 B1

45

5

10

15

20

25

30

35

40

45

50

55

mouracile, 5- �iodo-�uracile, 2’-�NH2-iodouracile, 2’-NH2-iodocytosine, 2’-�NH2-iodoadénine, 2’-�NH2-bro-mouracile, 2’- �NH2-bromocytosine et 2’-�NH2-bro-moadénine.

85 86

EP 0 736 105 B1

46

EP 0 736 105 B1

47

EP 0 736 105 B1

48

EP 0 736 105 B1

49

EP 0 736 105 B1

50

EP 0 736 105 B1

51

EP 0 736 105 B1

52

EP 0 736 105 B1

53

EP 0 736 105 B1

54

EP 0 736 105 B1

55

EP 0 736 105 B1

56

EP 0 736 105 B1

57

EP 0 736 105 B1

58

EP 0 736 105 B1

59

EP 0 736 105 B1

60

EP 0 736 105 B1

61

EP 0 736 105 B1

62

EP 0 736 105 B1

63

EP 0 736 105 B1

64

EP 0 736 105 B1

65

EP 0 736 105 B1

66

EP 0 736 105 B1

67

EP 0 736 105 B1

68

EP 0 736 105 B1

69

EP 0 736 105 B1

70

EP 0 736 105 B1

71

EP 0 736 105 B1

72

EP 0 736 105 B1

73

EP 0 736 105 B1

74

EP 0 736 105 B1

75

EP 0 736 105 B1

76

EP 0 736 105 B1

77

EP 0 736 105 B1

78

EP 0 736 105 B1

79

EP 0 736 105 B1

80

EP 0 736 105 B1

81

REFERENCES CITED IN THE DESCRIPTION

This list of references cited by the applicant is for the reader’s convenience only. It does not form part of the Europeanpatent document. Even though great care has been taken in compiling the references, errors or omissions cannot beexcluded and the EPO disclaims all liability in this regard.

Patent documents cited in the description

• US 536428 A [0003]• WO 9119813 A [0003]• US 714131 A [0003]• US 5475096 A [0003]

• US 93147392 A [0003]• US 5270163 A [0003]• US 960093 A [0021]• WO 9409168 A [0021]

Non-patent literature cited in the description

• TALBOT et al. Nucleic Acids Res., 1990, vol. 18,3521 [0008]

• WEINTRAUB. Cold Spring Harbor Symp. Quant. Bi-ol., 1973, vol. 38, 247 [0008]

• LIN ; RIGGS. Proc. Natl. Acad. Sci. U.S.A., 1974,vol. 71, 947 [0008]

• OGATA ; WILBERT. Proc. Natl. Acad. Sci. U.S.A.,1977, vol. 74, 4973 [0008]

• BARBIER et al. Biochemistry, 1984, vol. 23, 2933[0008]

• WICK ; MATTHEWS. J. Biol. Chem., 1991, vol. 266,6106 [0008]

• WOLFES et al. Eur. J. Biochem., 1986, vol. 159, 267[0008]

• KATOUZIAN-SAFADI et al. Photochem. Photobiol.,1991, vol. 53, 611 [0008]

• KHALILI et al. EMBO J., 1988, vol. 7, 1205 [0008]• BLATTER et al. Nature, 1992, vol. 359, 650 [0008]• KATOUZIAN-SAFADI et al. Nucleic Acids Res.,

1991, vol. 19, 4937 [0008]• TANNER et al. Biochemistry, 1988, vol. 27, 8852

[0008]• GOTT et al. Biochemistry, 1991, vol. 30, 6290 [0008]

[0020]• FAVRE. Bioorganic Photochemistry. Photochemis-

try and the Nucleic Acids, 1990, vol. 1, 379-425[0009]

• WYATT et al. Genes & Development, 1992, vol. 6,2542 [0009]

• EVANS et al. Biochemistry, 1989, vol. 28 [0010]• FARRAR et al. Biochemistry, 1991, vol. 30, 3075

[0010]• LEE et al. J. Biol. Chem., 1991, vol. 266, 16478

[0010]• WOWER et al. Biochemistry, 1988, vol. 27, 8114

[0010]• WOWER et al. Biochemistry, 1989, vol. 28, 1563

[0010]• SALVUCCI ; HALEY. Planta, 1990, vol. 181, 287

[0010]

• LIU ; VERDINE. Tetrahedron Lett., 1992, vol. 33,4265 [0011]

• CHEN ; PRUSOFF. Biochemistry, 1977, vol. 16,3310 [0012]

• DIETZ et al. J. Am. Chem. Soc., 1987, vol. 109, 1793[0013] [0071]

• ITO et al. J. Am. Chem. Soc., 1980, vol. 102, 7535[0013]

• DIETZ ; KOCH. Photochem. Photobiol., 1987, vol.46, 971 [0013]

• DIETZ ; KOCH. Photochem. Photobiol., 1989, vol.49, 121 [0013]

• HUTCHINSON ; KÖHNLEIN. Prog. Subcell. Biol.,1980, vol. 7, 1 [0013]

• SHETLAR. Photochem. Photobiol. Rev., 1980, vol.5, 105 [0013]

• Bioorganic Photochemistry. SAITO ; SUGIYAMA.Photochemistry and the Nucleic Acids. John Wileyand Sons, 1990, vol. 1, 317-378 [0013]

• ALLEN et al. J. Biol. Chem., 1991, vol. 266, 6113[0014]

• SAITO et al. J. Org. Chem., 1986, vol. 51, 5148[0015]

• RAHN ; SELLIN. Photochem. Photobiol., 1982, vol.35, 459 [0016]

• RAHN ; STAFFORD. Photochem. Photobiol., 1979,vol. 30, 449 [0016]

• SUGIYAMA et al. J. Am. Chem. Soc., 1993, vol. 115,4443 [0016]

• ITO et al. Photochem. Photobiol., 1980, vol. 32, 683[0018]

• BAYLEY ; KNOWLES. Methods Enzymol., 1977,vol. 46, 69 [0019]

• CZARNECKI et al. Methods Enzymol., 1979, vol. 56,642 [0019]

• HANNA et al. Nucleic Acids Res., 1993, vol. 21, 2073[0019]

• EGGENS ; NATHANS. J. Mol. Biol., 1969, vol. 39,293 [0020]

• LING et al. Virology, 1970, vol. 40, 920 [0020]

EP 0 736 105 B1

82

• BECKETT et al. J. Mol Biol., 1988, vol. 204, 939[0020]

• TUERK ; GOLD. Science, 1990, vol. 249, 505 [0020]• SCHNEIDER et al. J. Mol. Biol., 1992, vol. 228, 862

[0020]• SZYBALSKI. Cancer Chemother. Rep., 1974, vol.

58, 539 [0051]• WILLIS et al. Science, 1993, vol. 262, 1255 [0071]• FEINBERG et al. Cell, 1986, vol. 46, 807 [0071]• EMERMAN et al. Cell, 1989, vol. 57, 1155 [0071]• HEAPHY et al. Proc. Natl. Acad. Sci. USA, 1991,

vol. 88, 7366 [0071]• BARTEL et al. Cell, 1991, vol. 67, 529 [0071]• TUERK et al. In the Polymerase Chain Reaction,

1993 [0071]• JENSEN et al. J. Mol. Biol., 1994, vol. 235, 237

[0071]• TUERK ; GOLD. Science, 1990, vol. 249, 1155

[0091]• HARTZ et al. Methods Enzymol, 1988, vol. 164, 419

[0092]

• HARTZ et al. Meth. Enzym., 1988, vol. 164, 419[0092]

• BARON et al. Proc. Natl. Acad. Sci. USA, 1993, vol.90, 4181 [0092]

• MILLIGAN et al. Nucleic Acids Res., 1987, vol. 15,8783 [0110]

• CAREY et al. Biochemistry, 1983, vol. 22, 4723[0110]

• GROEBE ; UHLENBECK. Nucleic Acids Res.,1988, vol. 16, 11725 [0112]

• WEBER. Biochemistry, 1983, vol. 6, 3144 [0118]• MEE. Radiation Chemistry: Principles and Applica-

tions. VCH Publishers, 1987, 477-499 [0137]• IGLOI. Biochemistry, 1988, vol. 27, 3842 [0162]• SASSE-DWIGHT ; GRALLA. Methods Enzymol.,

1991, vol. 208, 146 [0167]• AIGEN ; GUMPORT. Methods Enzymol., 1991, vol.

208, 433 [0167]• HOCKENSMITH et al. Methods Enzymol., 1991, vol.

208, 211 [0167]