This application claims priority under 35 U.S.C. § 119 (e) from United States Provisional Application Serial No. 62/237,382, filed April 25, 2016 which is hereby incorporated by reference in its entirety.

FIELD

Disclosed herein are methods of identifying target binding ligands from tagged combinatorial libraries. The methods generally involve contacting a combinatorial library with a target, washing the target to remove non-binding ligands, eluting ligands which bind to the target and identifying the ligands which were bound to the target. In some embodiments, the target may be a cell, tissues or membrane preparations. In other embodiments, the target may be attached to a matrix. In still other embodiments, the matrix may be packed in a column. The ligands may be operatively linked to monovalent oligonucleotides or multivalent oligonucleotides. Also, disclosed herein are various multivalent oligonucleotide architectures and methods of preparing and using combinatorial libraries which include multivalent oligonucleotides.

BACKGROUND

Combinatorial libraries, which were first developed over thirty years ago, are now routinely used to identify novel, high affinity ligands for a wide variety of biological targets (e.g. , receptors, enzymes, nucleic acids, etc.) and hence are of increasing importance in drug discovery. Tagged combinatorial libraries, which use DNA as a tag to record the synthetic steps undergone by ligands operatively attached to DNA or use DNA to direct and record the synthetic steps undergone by ligands operatively attached to DNA, are of particular current interest. Advances in DNA sequencing, PCR technology and ligand assay development, provide methods to identify and select monovalent ligands operatively linked to DNA that bind to a biological target, from complex mixtures of monovalent ligands operatively linked to DNA (Harbury, et al , U.S. Patent No. 7,479,472; Liu et al, U.S Patent No. 7,070,928; Liu et al, U.S Patent No. 7,223,545; Liu et al, U.S. Patent No. 7,442,160; Liu et al, U.S. Patent No. 7,491,160; Liu et al, U.S. Patent No. 7,557,068; Liu et al, U.S. Patent No.

7,771,935; Liu et al, U.S. Patent No. 7,807,408; Liu et al, U.S. Patent No.

7,998,904; Liu et al, U.S. Patent No. 8,017,323; Liu et al, U.S. Patent No.

8,183,178; Pedersen et al , U.S. Patent No. 1, 211,113; Pedersen et al, U.S. Patent No. 7,413,854; Gouliev et al, U.S. Patent No. 7,704,925; Franch et al , U.S. Patent No. 7,915,201; Gouliev et al , U.S. Patent No. 8,722,583; Freskgard et al , U.S. Patent Application No. 2006/0269920; Freskgard et al , U.S. Patent Application No. 2012/0028812; Hansen et al , U.S. Patent No. 7,928,211; Hansen et al , U.S. Patent No. 8,202,823; Hansen et al , U.S. Patent Application No. 2013/0005581; Hansen et al , U.S. Patent Application No. 2013/0288929; Neri et al , U.S. Patent No. 8,642,514; Neri et al , U.S. Patent No. 8,673,824; Neri et al , U.S. Patent Application No. 2014/01288290; Morgan et al , U.S. Patent No. 7,972,992; Morgan et al , U.S. Patent No. 7,935,658; Morgan et al , U.S. Patent Application No. 2011/0136697; Morgan et al , U.S. Patent No. 7,972,994;

Morgan et al , U.S. Patent No. 7,989,395; Morgan et al , U.S. Patent No.

8,410,028; Morgan et al , U.S. Patent No. 8,598,089; Morgan et al , U.S. Patent Application Serial No. 14/085,271; Wagner et al , U.S. Patent Application No. 2012/0053901; Keefe et al , U.S. Patent Application No. 2014/0315762; Dower et al , U.S. Patent No. 6,140,493; Lerner et al , U.S. Patent No. 6,060,596; Dower et al , U.S. Patent No. 5,789,162; Lerner et al, U.S. Patent No. 5,723,598; Dower et al; U.S. Patent No. 5,708,153; Dower et al , U.S. Patent No. 5,639,603; and Lerner et al , U.S. Patent No. 5,573,905). Although, combinatorial libraries, such as those referenced above, often provide ligands with superior binding affinity (i.e. , less than or equal to 1 μΜ) for many biological molecules, many targets of current interest such as, for example, protein-protein interfaces, when challenged with combinatorial libraries have not yielded any binding ligands, let alone high affinity ligands. Furthermore, combinatorial libraries which provide ligands with binding affinity for cells, tissues or membrane preparations have not yet been successfully identified. In phage display, multivalent display of ligands on the phage coat identifies low affinity ligands, whose recovery from binding selections is enhanced by the chelate or avidity effect (Phage Display: A Practical Approach, Clackson and Lowman (editors), Oxford University Press, 2004, pp 63-64). The low affinity ligands identified from multivalent display often serve as a starting point for discovery of high affinity ligands.

Accordingly, what is needed are methods which identify ligands from tagged combinatorial libraries, including libraries which incorporate multivalent oligonucleotides, which have affinity for targets, such as, for example, protein- protein interfaces, cell, tissues or membrane preparations. Such ligands, even those of low affinity, provide, inter alia, a starting point for optimizing affinity for biological targets.

SUMMARY

The present invention satisfies this and other needs by providing methods for identifying ligands which bind to target. In a first aspect, a method for identifying ligands which bind to cells, tissues or membrane preparations is provided. The method includes the steps of contacting cells, tissues or membrane preparations with a combinatorial library, washing to remove members of the combinatorial library which do not bind to cells, tissues or membrane preparations, eluting members of the combinatorial library which bind to cells, tissues or membrane preparations and identifying members of the combinatorial library which bind to cells, tissues or membrane preparations. In some embodiments, the combinatorial library includes multivalent oligonucleotides.

In a second aspect, a method for identifying ligands internalized by cells is provided. The method includes the steps of contacting the cells with a combinatorial library, which includes ligands operatively linked to recognition elements, washing the cells to remove members of the combinatorial library not internalized, disrupting the cell line to release internalized members of the combinatorial library and identifying the members of the combinatorial library which are internalized. In some embodiments, the combinatorial library includes multivalent oligonucleotides.

In a third aspect, a method for identifying ligands which bind to a target is provided. The method includes the steps of attaching the target to a matrix, contacting the members of a combinatorial library with the matrix, washing the matrix to remove the members of the combinatorial library which do not bind to the target, eluting the members of the combinatorial library which bind to the target and identifying the members of the combinatorial library which bind to the target. In some embodiments, the combinatorial library includes multivalent oligonucleotides.

In a fourth aspect, a method for identifying ligands which bind to cells, tissues or membrane preparations is provided. The method includes the steps of attaching a biological target present in cells, tissues or membrane preparations to a matrix, contacting a combinatorial library with the matrix washing the matrix to remove members of the combinatorial library which do not bind to the biological target, eluting members of the combinatorial library which bind to the biological target, contacting combinatorial library with cells, tissues or membrane preparations which contain the biological target, washing the cells to remove members of the combinatorial library which do not bind to the cells, tissues or membrane preparations, eluting multi-valent oligonucleotides which bind to cells, tissues or membrane preparations and identifying ligands which bind to cells, tissues or membrane preparations. In some embodiments, the combinatorial library includes multivalent oligonucleotides.

BRIEF DESCRIPTION OF THE FIGURES Figure la illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5 ' end of an oligonucleotide. Figure lb illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 3 ' end of an oligonucleotide.

Figure lc illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of an oligonucleotide.

Figure 2 illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5 ' end of an oligonucleotide and a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 3' end of the oligonucleotide.

Figure 3 illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5 ' end of an oligonucleotide and a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide.

Figure 4 illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5 ' end of an oligonucleotide and a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide.

Figure 5 illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide and a linking entity or multivalent linking entity operatively linked to a different internal nucleic acid subunit of the oligonucleotide.

Figure 6 illustrates an oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5 ' end of an oligonucleotide, a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide and a linking entity or multivalent linking entity operatively linked attached to a nucleic acid subunit at the 3' end of the oligonucleotide. Figure 7 illustrates a hairpin oligonucleotide with a linker and a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5' end of an oligonucleotide and a linking entity or multivalent linking entity operatively attached to a nucleic acid subunit at the 3 ' end of the oligonucleotide.

Figure 8 illustrates a hairpin oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5 ' end of an oligonucleotide and a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide.

Figure 9 illustrates a hairpin oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 3 ' end of an oligonucleotide and a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide.

Figure 10 illustrates a hairpin oligonucleotide with a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide and a linking entity or multivalent linking entity operatively linked to a different internal nucleic acid subunit of the oligonucleotide.

Figure 11 illustrates a hairpin oligonucleotide with a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 5 ' end of an oligonucleotide, a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the oligonucleotide and a linking entity or multivalent linking entity operatively linked to a nucleic acid subunit at the 3 ' end of the oligonucleotide.

Figure 12 illustrates two oligonucleotides each with a linking entity or multivalent linking entity operatively linked to internal nucleosides hybridized to a contiguous linear oligonucleotide.

Figure 13 illustrates two oligonucleotides each with a linking entity or multivalent linking entity operatively linked to internal nucleosides hybridized to a contiguous circular oligonucleotide. Figure 14 illustrates a circular oligonucleotide with a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the circular oligonucleotide and a linking entity or multivalent linking entity operatively linked to a different internal nucleic acid subunit of the circular oligonucleotide.

Figure 15a illustrates an exemplary oligonucleotide with a monovalent linker.

Figure 15b illustrates an exemplary oligonucleotide with a divalent linker.

Figure 15c illustrates an exemplary oligonucleotide with a trivalent linker. Figure 16a illustrates an exemplary oligonucleotide with a monovalent linker.

Figure 16b illustrates an exemplary oligonucleotide with a divalent linker.

Figure 16c illustrates an exemplary oligonucleotide with a trivalent linker.

Figure 17a illustrates the structure of N3-TEG-K17F. Figure 17b illustrates the structure of FAM-TEG-K17F.

Figure 18a illustrates FACS analysis using the FAM-TEG-K17F ligand with negative control and APJ-expressing CHO cells.

Figure 18b illustrates ELISA standard curve for FAM.

Figure 18c illustrates cell surface ELISA for quantitation of APJ with the FAM-TEG-K17F ligand.

Figure 19a illustrates reduction of background spike DNA.

Figure 19b illustrates the number of cells lost by washing stringency.

Figure 20a illustrates Specific fold enrichment (ER) of K17F conjugates for multiple washes. Figure 20b illustrates loss of enrichment by using excess K17F-FAM competitor.

Figure 20c illustrates specific enrichment for K17F species using PBS- based panning protocol.

Figure 21 illustrates a comparison between heat and lysis elution.

Figure 22 illustrates the specific enrichment of K17F conjugates with a 10 ⁶ -fold excess of background DNA.

Figure 23 illustrates the specific enrichment of ligand-conjugates which were incubated with cells, allowed to internalize, and when the cell surface was treated with and without benzonase.

Figure 24 illustrates the performance of monovalent vs multivalent conjugates (15 washes, +3E7).

DETAILED DESCRIPTION

Definitions Unless defined otherwise, all technical and scientific terms used herein have the same meaning as is commonly understood by one of ordinary skill in the art to which this invention belongs. In the event that a plurality of definitions for a term exists, those in this section prevail unless stated otherwise.

It must be noted that as used herein and in the appended claims, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a tag" includes a plurality of such tags and reference to "the compound" includes reference to one or more compounds and equivalents thereof known to those skilled in the art, and so forth.

It is further noted that the claims may be drafted to exclude any optional element. As such, this statement is intended to serve as antecedent basis for use of such exclusive terminology as "solely", "only" and the like in connection with the recitation of claim elements, or the use of a "negative" limitation.

As used herein, and unless otherwise specified, the terms "about" and "approximately," when used in connection with a property with a numeric value or range of values indicate that the value or range of values may deviate to an extent deemed reasonable to one of ordinary skill in the art while still describing the particular property. Specifically, the terms "about" and "approximately," when used in this context, indicate that the numeric value or range of values may vary by 5%, 4%, 3%, 2%, 1%, 0.9%, 0.8%, 0.7%, 0.6%, 0.5%, 0.4%, 0.3%, 0.2% or 0.1 % of the recited value or range of values while still describing the particular solid form.

"Antibody" as used herein refers to a protein comprising one or more polypeptides substantially or partially encoded by immunoglobulin genes or fragments of immunoglobulin genes, e.g. , a fragment containing one or more complementarity determining region (CDR). The recognized immunoglobulin genes include the kappa, lambda, alpha, gamma, delta, epsilon and mu constant region genes, as well as myriad immunoglobulin variable region genes. Light chains are typically classified as either, e.g., kappa or lambda. Heavy chains are typically classified e.g. , as gamma, mu, alpha, delta, or epsilon, which in turn define the immunoglobulin classes, IgG, IgM, IgA, IgD and TgE, respectively. A typical immunoglobulin (antibody) structural unit comprises a tetramer. In nature, each tetramer is composed of two identical pairs of polypeptide chains, each pair having one "light" (about 25 kD) and one "heavy" chain (about 50-70 kD). The N-terminus of each chain defines a variable region of about 100 to 110 or more amino acids primarily responsible for antigen recognition. The terms variable light chain (VL) and variable heavy chain (VH) refer to these light and heavy chains respectively. Antibodies exist as intact immunoglobulins or as a number of well characterized fragments produced by digestion with various peptidases. Thus, for example, pepsin digests an antibody below the disulfide linkages in the hinge region to produce F(ab)'2 (fragment antigen binding) and Fc (fragment crystallizable, or fragment complement binding). F(ab)'2 is a dimer of Fab, which itself is a light chain joined to VH-CH1 by a disulfide bond. The F(ab)'2 may be reduced under mild conditions to break the disulfide linkage in the hinge region thereby converting the (Fab')?, dimer into a Fab' monomer. The Fab' monomer is essentially a Fab with part of the hinge region. The Fc portion of the antibody molecule corresponds largely to the constan region of the immunoglobulin heavy chain, and is responsible for the antibody's effector function (see, Fundamental Immunology, 4 ^th edition. W.E. Paul, ed., Raven Press, N.Y. (1998), for a more detailed description of antibody fragments). While various antibody fragments are defined in terms of the digestion of an intact antibody, one of skill will appreciate that such Fab' or Fc fragments may be synthesized de novo either chemically or by utilizing recombinant DNA methodology, peptide display, or the like. Thus, the term antibody, as used herein, also includes antibody fragments either produced by the modification of whole antibodies or synthesized de novo using recombinant DNA methodologies. Antibodies also include single-armed composite monoclonal antibodies, single chain antibodies, including single chain Fv (sFv) antibodies in which a variable heavy and a variable light chain are joined together (directly or through a peptide linker) to form a continuous polypeptide, as well as diabodies, tribodies, and tetrabodies (Pack et al. (1995) J Mol Biol 246:28; Biotechnol 11 :1271;

and Biochemistry 31 : 1579). The antibodies are, e.g. , polyclonal, monoclonal, chimeric, humanized, single chain, Fab fragments, fragments produced by a Fab expression library, or the like.

"Chromatographic matrix" as used herein, refers to conventional chromatography resins, such as, for example, reverse phase, silica, ion exchange or "mixed mode" supports. The chromatographic matrix may also include resins which include immobilized proteins (e.g. , serum albumin, alpha-acid

glycoprotein or immobilized artificial membranes (many are available from Regis Technologies, Inc., Morton Grove, IL) such as, for example, phospholipids (Pidgeon et al., Anal. Chem. 176 (1989)). "Combinatorial library" as used herein refers to a library of molecules containing a large number, typically between 10 ³ and 10 ⁱ⁵ or more different compounds typically characterized by different sequences of subunits, or a combination of different side chains functional groups and linkages. In some embodiments, the combinatorial library includes multivalent oligonucleotides. In other embodiment, the combinatorial library consists entirely of multivalent oligonucleotides.

"Depsipeptide" as used herein refers to a peptide as defined herein where one or more of amide bonds are replaced by ester bonds.

"Functional group" as used herein, refers to a chemical group such as, for example, an electrophilic group, a nucleophilic group, a diene, a dienophile, etc. Examples of functional groups include, but are not limited to, -NH ₂ , -SH, -OH, - CO2H, halo, -N3, -CONH2, etc. and may also include dendrimers with the above functional groups.

"Gel matrix" as used herein, includes, refers to various gels such as, cyrogels, agarose, superagarose or polyacrylamide gels. Typically, a gel matrix will include a lipid phase, such as, for example, vesicles, liposomes, micelles, lipophilic compounds, lipophilic polymers, artificial membranes or combinations thereof.

"Ligand" as used herein refers to an oligonucleotide, single stranded RNA, single stranded DNA, a DNA binding protein, a RNA binding protein, a peptide nucleic acid, a peptide, a depsipeptide, a polypeptide, an antibody, a peptoid, a polymer, a polysiloxane, an inorganic compound of molecular weight greater that 50 daltons, an organic compound of molecular weight of less than about 1500 daltons.

"Linking entity" as used herein, refers to any linking molecule which is operatively linked to a tagging oligonucleotide and which includes at least one functional group. The functional group of the linking entity, in some instances, serves as the initiation site for commencing ligand synthesis. In other instances, the functional group of the linking entity may be the site for connecting to another linking entity or a dendrimer. In some embodiments, the functional group of the linking entity may be protected, by methods well known to those of skill in the art. The linking entity may vary in structure and length. The linking entity may be hydrophobic or hydrophilic, long or short, rigid, semirigid or flexible, etc. The linking entity can comprise, for example, a polymethylene chain, such as a— (CH ₂ ) _R — chain or a poly(ethylene glycol) chain, such as a— (CH ₂ CH ₂ 0)n chain, where in both cases n is an integer from 1 to about 40, 5'-0- Dimethoxytrityl- ,2'-Dideoxyribose-3'-[(2-cyanoethyl)-(N,N-diisopropyl)]- phosphoramidite; 9-O-Dimethoxytrityl-trieihylene glycol, l-[(2-c anoethyl)- (N,N-diisopropyl)]-phosphoramidite; 3 -(4,4'-Dimethoxy trityloxy)propyl- 1 - [(2- cyanoethyl)-(N,N-diisopropyl)]-phosphoramidite; and 18-0-

Dimethoxytritylhexaethyleneglycol, l ,-[(2-cyaiioethyl)-(N,N-diisopropyl)]- phosphoramidite, amino-carboxylic linkers (e.g. , peptides (e.g., Z-Gly-Gly-Gly- Osu or Z-Gly-Gly-Gly-Gly-Gly-Gly-Osu), PEG (e.g. , Fmoc-aminoPEG2000- NHS or amino-PEG (12-24)-NHS), or alkane acid chains (e.g., Boc-ε- arninocaproic acid-Osu)), click chemistry linkers (e.g. , peptides (e.g. , azidohomalanine-Gly-Gly-Gly-OSu or propargylglycine-Gly-Gly-Gly-OSu), PEG (e.g. , azido-PEG-NHS), or alkane acid chains (e.g. , 5-azidopentanoic acid, (S)-2-(azidomethyl)- l-Boc -pyrrolidine, or 4-azido-butan- l-oic acid N- hydroxysuccinimide ester)), thiol-reactive linkers (e.g. , PEG (e.g. , SM(PEG)n NHS-PEG-maleimide), alkane chains (e.g. , 3 -(pyridin- 2 -yldisulfanyl)- ropionic acid-Osu or sulfosuccinimidyl 6-(3'-[2-pyridyldithio]propionaniido)hexanoate))), amidites for oligonucleotide synthesis (e.g. , amino modifiers (e.g. , 6- (trifluoroacetylarnino)-hexyl-(2-cyanoethyl)-(N,N-diisopropy l)- phosphoramidite), thiol modifiers (e.g. , S-trityl-6-mercaptohexyl-l-[(2- cyanoethyl)-(N,N-diisopropyl)]-phosphoramidite, or chick chemistry modifiers (e.g. , 6-hexyn- l-yl-(2-cyanoethyl)-(N,N-diisopropyl)-phosphoramidite, 3- dimethoxytrityloxy-2-(3-(3-propargyloxypropanamido)propanami do)propyl-l-0- succinoyl, long chain alkylamino CPG, or 4-azido-butan-l-oic acid N- hydroxysuccinimide ester)). In some embodiments, the linking entity may include a functionalized dendrimer which are available from a number of commercial suppliers such as, for example, Sigma Aldrich, St. Louis, MO., Polymer Factory Sweden AB, Stockholm, Sweden, Dendritech, Inc. Midland, MI, 48642 or NanoSynthons LLC, Mt. Pleasant, MI 48858. The dendrimer may be, for example, a PANAM dendrimer or polypropylenimine dendrimer.

"Linker" as used herein, is any molecule or substance which links the ligand to the oligonucleotide. The linker will, for example, generally have the structure of a linking entity where the ligand or a chemical entity has reacted with the functional group.

"Lipid matrix" as used herein refers to lipophilic vesicles, artificial membranes, beads coated with lipophilic or amphophilic material, lipophilic compounds, lipophilic polymers, liposomes or micelles.

"Matrix" as used herein refers to, in general, at least three different type of matrixes: "chromatographic," "gel," or "lipid." In some embodiments, the matrix may be a hydrophilic liquid contacting a hydrophobic liquid (e.g. , octanol- water).

"Monovalent linking entity" as used herein, refers to a tagging oligonucleotide which is operative] y linked to one linking entity

"Multivalent linking entity" as used herein, refers to a tagging

oligonucleotide which is operatively linked to more than one linking entity.

"Multivalent oligonucleotide" as used herein refers to a tagging oligonucleotide which is operatively linked to more than one ligand. In some embodiments, the multivalent oligonucleotide is a member of a DNA encoded library.

"Nucleic acid" as used herein refers to an oligonucleotide analog as defined below as well as a double stranded DNA and RNA molecule. A DNA and RNA molecule may include the various analogs defined below.

"Oligonucleotides" or "oligos" as used herein refer to nucleic acid oligomers containing between about 3 and up to about 500, and typically from about 5 to about 250 nucleic acid subunits. In the context of oligos (e.g. , hybridization sequence) which may direct the synthesis of library compounds, the oligos may include or be composed of naturally-occurring nucleotide residues, nucleotide analog residues, or other subunits capable of forming

sequence-specific base pairing, when assembled in a linear polymer, with the proviso that the polymer is capable of providing a suitable substrate for

strand-directed polymerization in the presence of a polymerase and one or more nucleotide triphosphates, e.g. , conventional deoxyribonucleotides. A

"known-sequence oligo" is an oligo whose nucleic acid sequence is known.

"Oligonucleotide analog" as used herein refers to nucleic acids that have been modified and which are capable of some or all of the chemical or, biological activities of the oligonucleotide from which it was derived. An oligonucleotide analog will generally contain phosphodiester bonds, although in some cases, oligonucleotide analogs are included that may have alternate backbones.

Modifications of the ribose-phosphate backbone may facilitate the addition of additional moieties such as labels, or may be done to increase the stability and half-life of such molecules, in addition, mixtures of naturally occurring nucleic acids and analogs can be made. Alternatively, mixtures of different nucleic acid analogs, and mixtures of naturally occurring nucleic acids and analogs may be made. The oligonucleotides may be single stranded or double stranded, as specified, or contain portions of both double stranded or single stranded sequence. The oligonucleotide may be DNA, R A or a hybrid, where the nucleic acid contains any combination of deoxyribo-and ribo-nucleotides, and any combination of bases, including uracil, adenine, thymine, cytosine, guanine, inosine, xathanine hypoxathanine, isocytosine, isoguanine, etc. "Operatively linked, ^' " as used herein, means at least two chemical structures joined together in such a way as to remain linked through the various manipulations described herein. Typically, a ligand and the encoding oligonucleotide are linked covalently via an appropriate linker. The linker is at least a bi valent moiety with a site of attachment for the oligonucleotide and a site of attachment for the ligand or chemical moiety. For example, when the functional moiety is a polyamide compound, the polyamide compound can be attached to the linking group at the N-terminus, the C-terminus or via a functional group on one of the side chains. The linker is sufficient to separate the ligand and the oligonucleotide by at least one atom and in some embodiments by more than one atom. In most embodiments, the linker is sufficiently flexible to allow the ligand to bind target molecules in a manner which is independent of the oligonucleotide.

"Peptide" as used herein refers to a polymer of amino acid residues between about 2 and 50 amino acid residues, between about 2 and 20 amino acid residues, or between about 2 and 10 residues. Peptides include modified peptides such as, for example, glycopeptides, PEGylated peptides, !ipopeptides, peptides conjugated with organic or inorganic ligands, peptides which contain peptide bond isosteres (e.g., /i O i -S j. ψ! θ ί.·Ν Π · |. \|/[NHCO], ψ[(ΓΟ€Η ₂ ], ψ[(Ε) or (Ζ) CH=CH], etc. and also include cyclic peptides. In some embodiments, the amino acid residues may be any L- a- amino acid, D--a~amino residue, N-alkyl variants thereof or combinations thereof. In other embodiments, the amino acid residues may any L-a- amino acid, D- -amino residue, β-amino acids, x-amino acids, N-alkyl variants thereof or combinations thereof.

"Peptide nucleic acid" as used herein refers to oligonucleotide analogues where the sugar phosphate backbone of nucleic acids has been replaced by psuedopeptide skeleton (e.g. , N-(2-aminoethyl)-glycine)(Nielsen et al, U.S.

Patent No. 5,539,082; Nielsen et al, U.S. Patent No. 5,773,571; Burchardt et al, U.S. Patent No. 6,395,474).

"Peptoid" as used herein refers to polymers of poly N-substituted glycine (Simon et al , Proc. Natl. Acad. S t (1992) 89(20) 9367-9371) and include cyclic variants thereof.

"Polypeptide" as used herein refers to a polymer of amino acid residues typically comprising greater than 50 amino acid residues and includes cyclic variants thereof. Polypeptide includes proteins (including modified proteins such as glycoproteins, PEGylated proteins, lipoproteins, polypeptide conjugates with organic or inorganic ligands, etc.) receptor, receptor fragments, enzymes, structural proteins (e.g. , collagen) etc. In some embodiments, the amino acid residues may be any L-a-amino acid, D-a-araino residue, or combinations thereof. In other embodiments, the amino acid residues may be any L-a-amino acid, D-a-amino residue, N-alkyl variants thereof or combinations thereof. "Polymer" includes copolymers, and the term "monomer ^' " includes co-monomers. Polymers include, for example, polyarnides, phospholipids, polycarbonates, polysaccharides, polyurethanes, polyesters, polyureas, polyacetates, polyarylene sulfides, polyethylenimines, polyimides, etc.

"Recognition Element" as used herein refers to an oligonucleotide, single or double- stranded RNA, single or double-stranded DNA, a DNA binding protein, a locked nucleic acid, a RNA binding protein, a peptide nucleic acid, a peptide, a depsipeptide, a polypeptide, an antibody, a peptoid, a polymer, a polysiloxanes, an inorganic compounds of molecular weight greater that 50 daltons, organic compounds of molecular weight between about 2000 daltons and about 50 daltons or a combination thereof.

Reference will now be made in detail to embodiments of the invention. While the invention will be described in conjunction with these embodiments, it will be understood that it is not intended to limit the invention to the

embodiments, infra. To the contrary, it is intended to cover alternatives, modifications, and equivalents as may be included within the spirit and scope of the invention as defined by the appended claims.

Methods of Identifying Target Binding Ligands from Tagged

Combinatorial Libraries

Disclosed herein are methods of identifying target binding ligands from tagged combinatorial libraries. The methods generally involve contacting a combinatorial library with a target, washing the target to remove non-binding ligands, eluting ligands which bind to the target and identifying the ligands which were bound to the target. In some embodiments, the target may be a cell, tissues or membrane preparations. In other embodiments, the target may be a receptor, enzyme, protein-protein interface, etc. In some embodiments, the target may be attached to a matrix. In other embodiments, the matrix may be packed in a column. The ligands may be operatively linked to monovalent oligonucleotides or multivalent oligonucleotides. Also, disclosed herein are various multivalent oligonucleotide architectures and methods of preparing and using combinatorial libraries which include multivalent oligonucleotides.

In some embodiments, a method for identifying ligands which bind to cells, tissues or membrane preparations is provided. The method includes the steps of contacting cells, tissues or membrane preparations with a combinatorial library, washing to remove members of the combinatorial library which do not bind to cells, tissues or membrane preparations, eluting members of the combinatorial library which bind to cells, tissues or membrane preparations and identifying members of the combinatorial library which bind to cells, tissues or membrane preparations. In some embodiments, the combinatorial library includes multivalent oligonucleotides.

In other embodiments, a method for identifying ligands internalized by cells is provided. The method includes the steps of contacting the cells with a combinatorial library, which includes ligands operatively linked to recognition elements, washing the cells to remove members of the combinatorial library not internalized, disrupting the cell line to release internalized members of the combinatorial library and identifying the members of the combinatorial library which are internalized. In some embodiments, the combinatorial library includes multivalent oligonucleotides.

In still other embodiments, a method for identifying ligands which bind to a target is provided. The method includes the steps of attaching the target to a matrix, contacting the members of a combinatorial library with the matrix, washing the matrix to remove the members of the combinatorial library which do not bind to the target, eluting the members of the combinatorial library which bind to the target and identifying the members of the combinatorial library which bind to the target. In some embodiments, the combinatorial library includes multivalent oligonucleotides. In still other embodiments, a method for identifying ligands which bind to cells, tissues or membrane preparations is provided. The method includes the steps of attaching a biological target present in cells, tissues or membrane preparations to a matrix, contacting a combinatorial library with the matrix washing the matrix to remove members of the combinatorial library which do not bind to the biological target, eluting members of the combinatorial library which bind to the biological target, contacting combinatorial library with cells, tissues or membrane preparations which contain the biological target, washing the cells to remove members of the combinatorial library which do not bind to the cells, tissues or membrane preparations, eluting multi-valent oligonucleotides which bind to cells, tissues or membrane preparations and identifying ligands which bind to cells, tissues or membrane preparations. In some embodiments, the combinatorial library includes multivalent oligonucleotides.

In some of the above embodiments, the cells, tissue or membrane preparations are localized in a hollow fiber reactor. In other of the above embodiments, the cells, tissue or membrane preparations are localized in a microfluidic device (see e.g., Wang et al, Proc. Nat. Acad. Sci. 108, 17, 6909; Yu et al, Biomed Microdevices 11, 547, 2009). In some of the above embodiments, a biological target is assembled into nanodiscs as is well known in the art (Bayburt et al, FEBS Lett 2010, 584 (9), 1721-1727). The cells, tissues and membrane preparations used in the embodiments above, may either be purchased from commercial suppliers or prepared by conventional means known to those of skill in the art.

Ligands are operative! y linked to a recognition element with a linker, which generally is any molecule or substance which performs the function of connecting the ligands to the recognition element. The distance between the ligand and the recognition element may be greater than about 10A, about 25 A, about 50 A or about 100 A.

Linkers, in addition to the ones describe in the definitions may also include UV activated cross linkers. UV activated cross linkers are well known in the art and many examples thereof may be purchased from commercial suppliers, such as, for example, ThermoFisher Scientific, San Jose, CA and Santa Cruz Biotechnology, Inc. Dallas, TX. Exemplary UV activated cross linkers include, for example, N-5-azido-2-nitrobenzoyloxysuccinimide and succinimidyl 6-(4,4'- azipentamido)hexanoate.

In many of the above embodiments, nuclease inhibitors are included in the contacting step. The nuclease inhibitors may prevent oligonucleotide degradation. The inhibitor may be a divalent metal chelator, such as, for example, EDTA.

In some of the above embodiments, the ligand is a peptide, a peptide macrocycle, a depsipeptide, a peptoid or an organic compound of molecular weight of less than 1500 daltons. In other embodiments, the ligand is a peptide, a peptide macrocycle or an organic compound of molecular weight of less than 1500 daltons.

In some embodiments, the recognition element is an oligonucleotide, a double stranded oligonucleotide, single stranded RNA, a double stranded RNA- DNA hybrid, single stranded DNA, double stranded DNA or a peptide nucleic acid. In other embodiments, the recognition element is an oligonucleotide, single stranded DNA, single stranded RNA or double stranded DNA.

Generally, the target can be any substance, including any molecule, for which identification of compounds with affinity is desirable. In some embodiments, the target is a biological molecule. In other embodiments, the target is an antibody, enzyme, a receptor, an ion channel, a nucleic acid, a carbohydrate, protein-protein interface, a virus, bacteria, a eukaryotic cell or a prion.

Biological targets include, but are not limited to, CD3, GPII/IIIA, CD20, IL2R, RSV F, TNF, Her2, CD33, CD52, CDlla, EGFR, IgE, VEGF, VEGF Fab, VLA4, C5, ILlbeta, EPCAM, P40 (IL12R), IL6R, RANKL, B7.1/2, CD30, B anthrasis PA, Alpha IV/Beta VII, BAFFAPRIL, CD2, CTLA4, ile Enterotoxin, ILla and b, IL5, IL6, IL6R, IL13, IL17a, IL17R, IL23, BLyS, CD20. Amyloid Beta, PCSK9, ST4, HGF, Folate Ralpha, CD22, EGFR2, PD1, cMET, NaV1.7 and GM3. Other biological targets include, for example, whole fixed or living cells, tissue sections and membrane preparations. The biological targets may be immobilized on a matrix by ionic, covalent, magnetic or high affinity ligand- receptor (e.g. , bio tin and streptavidin) means.

Columns used herein may be any sort of column contemplated by the skilled artisan, such as, for example, a high-pressure liquid chromatography column, a gravity column, a capillary column, etc. As is well known in the chromatography arts, various neutral solvents or buffers may be used to elute combinatorial library non-binders. Combinatorial library members, including binders, may be eluted from the column by solvents which include chaotropic agents of increasing percentage or solvents of increasing ionic strength.

Washes may be conducted using low stringency conditions, medium stringency or high stringency conditions. Low stringency conditions, for example, are moderate number of repetitive washes (e.g. , 1-3 washes) using neutral solvents or buffers. Medium stringency conditions, for example, are an intermediate number of repetitive washes (e.g. , 4-10 washes) using neutral solvents or buffers. High stringency conditions, for example, are a large number of repetitive washes (e.g., >10 washes) using neutral solvents or buffers.

Stringency can also be increase by increasing the ionic strength of the solvent or the percentage of chaotropic agent in successive washes.

As is known to the skilled artisan, identification of member of combinatorial libraries where the ligands are operatively linked to

oligonucleotides can be accomplished by identification of the oligonucleotide. Generally, the oligonucleotide may be identified by any method known to those of skill in the art, including, for example, but not limited to, biological methods (e.g., affinity binding, sequencing, etc.) and chemical methods (e.g. , NMR, mass spectroscopy, etc.). Ligand identification generally involves amplification of the operatively linked oligonucleotides and sequencing the amplified

oligonucleotides. Oligonucleotides may be amplified by polymerase chain reaction, linear chain reaction or rolling circle amplification. Such methods are well known in the art and within the ambit of the skilled artisan. The amplified nucleic acids may be sequenced by well-known methods, including, but not limited to, Next Gen Sequencing.

Figures 1-14 exemplify a number of multivalent oligonucleotides architectures. In Figures 1-14, the straight line represents an oligonucleotide. In some implementations, the helical line represents a linking entity and X is a functional group. In other implementations, the helical line represents a linker and X is Y, which is multivalent linking entity. In principle, the multivalency of the linking entity is unlimited. As will be apparent to the skilled artisan, these implementations may be used as starting materials for combinatorial library synthesis, which will result in multivalent oligonucleotides which have the architecture depicted in Figures 1-14.

It should be further understood that multivalent oligonucleotides, include, but are not limited to the architectures disclosed in Figures 1-14 and that other architectures which are not depicted are within the scope of the present disclosure. Thus, the architectures depicted in Figures 1-14 are exemplary rather than limiting. It should also be also understood that, although Figures 1-14 refer to oligonucleotides, that the structures exemplified in Figures 1-14 may include, in other embodiments, different recognition elements, such as, for example, peptide nucleic acids.

Referring now to Figures la, lb and lc, depicted are oligonucleotides, where the linking entity (Y) includes multiple functional groups and is a multivalent linking entity. Exemplary examples of multivalent linking entities can be found in Figures 15b, 15c, 16b and 16c. Figures la and lb exemplify attachment of a multivalent linking entity to the 5' or 3' nucleoside of an oligonucleotide, respectively, while Figure lc exemplifies attachment of a multivalent linking entity to an internal nucleoside of an oligonucleotide. Other multivalent linkers, such as various complex dendrimers will be immediately obvious to the skilled artisan. In the representations above, one linking entity is attached to one nucleic acid subunit. Because the linking entity is multivalent, after conclusion of combinatorial library synthesis, more than one ligand will be attached to one nucleic acid subunit of the oligonucleotide to provide a multivalent oligonucleotide.

Referring now to Figures 2-14 depicted are oligonucleotides where a linking entity or a multivalent linking entity is attached to at least two different nucleic acid subunits. It should be understood that the linking entity optionally may be multivalent (i.e. , X can optionally be Y) (Figures 15b, 15c, 16b and 16c) or monovalent (Figures 15a and 16a).

Figure 3 illustrates a linking entity or a multivalent linking entity attached to the nucleic acid subunit at the 5' end and an internal nucleic acid subunit of an oligonucleotide. Figure 4 illustrates a linking entity or a multivalent linking entity attached to the nucleic acid subunit at the 3' end and an internal nucleic acid subunit of an oligonucleotide. Figure 5 illustrates a linking entity or a multivalent linking entity attached to two internal nucleic acid subunits of an oligonucleotide. Figure 6 illustrates a linking entity or a multivalent linking entity attached to a nucleic acid subunit at the 5 'end, an internal nucleic acid subunit and the nucleic acid subunit at the 3 ' end of an oligonucleotide

Figure 7 illustrates a linking entity or a multivalent linking entity attached to both the nucleic acid subunits at the 5' and 3' end of an oligonucleotide, which includes a hairpin. Figure 8 illustrates a linking entity or a multivalent linking entity attached to the nucleic acid subunit at the 3' end and an internal nucleic acid subunit of an oligonucleotide, which includes a hairpin. Figure 9 illustrates a linking entity or a multivalent linking entity attached to the nucleic acid subunit at the 5 ' end and an internal nucleic acid subunit of an oligonucleotide, which includes a hairpin. Figure 10 illustrates a linking entity or a multivalent linking entity attached to two internal nucleic acid subunits of an oligonucleotide, which includes a hairpin. Figure 11 illustrates a linking entity or a multivalent linking entity attached to the nucleic acid subunit at the 5' end, at the nucleic acid subunit at the 3 ' end and an internal nucleic acid subunit of an oligonucleotide which includes a hairpin. Figure 12 illustrates two oligonucleotides each with a linking entity or multivalent linking entity operatively linked to internal nucleosides hybridized to a contiguous linear oligonucleotide. Figure 13 illustrates two oligonucleotides each with a linking entity or multivalent linking entity operatively linked to internal nucleosides hybridized to a contiguous circular oligonucleotide. Figure 14 illustrates a circular oligonucleotide with a linking entity or multivalent linking entity operatively linked to an internal nucleic acid subunit of the circular oligonucleotide and a linking entity or multivalent linking entity operatively linked to a different internal nucleic acid subunit of the circular oligonucleotide. It should be understood that any of the single stranded oligonucleotides depicted in Figure 1-14 can be hybridized with a complementary oligonucleotide to provide double stranded oligonucleotides. It should also be apparent that any of the single stranded oligonucleotides depicted in Figures 1-14 can be converted into multivalent oligonucleotides or can be incorporated as multivalent oligonucleotides into combinatorial libraries by use of the various methods of combinatorial library synthesis referenced below.

The oligonucleotide portion of many of the structures depicted in Figures

1-14 may, in some embodiments, be based on oligonucleotides previously used in

DNA encoded libraries (Harbury, et al , U.S. Patent No. 7,479,472; Liu et al, U.S Patent No. 7,070,928; Liu et al, U.S Patent No. 7,223,545; Liu et al, U.S.

Patent No. 7,442,160; Liu et al, U.S. Patent No. 7,491,160; Liu et al, U.S.

Patent No. 7,557,068; Liu et al, U.S. Patent No. 7,771,935; Liu et al, U.S.

Patent No. 7,807,408; Liu et al, U.S. Patent No. 7,998,904; Liu et al, U.S.

Patent No. 8,017,323; Liu et al, U.S. Patent No. 8,183,178; Pedersen et al , U.S. Patent No. 1, 211,113; Pedersen et al , U.S. Patent No. 7,413,854; Gouliev et al ,

U.S. Patent No. 7,704,925; Franch et al , U.S. Patent No. 7,915,201; Gouliev et al , U.S. Patent No. 8,722,583; Freskgard et al , U.S. Patent Application No.

2006/0269920; Freskgard et al , U.S. Patent Application No. 2012/0028812;

Hansen et al , U.S. Patent No. 7,928,211; Hansen et al , U.S. Patent No.

8,202,823; Hansen et al , U.S. Patent Application No. 2013/0005581; Hansen et al , U.S. Patent Application No. 2013/0288929; Morgan et al , U.S. Patent No. 7,972,992; Morgan et al , U.S. Patent No. 7,935,658; Morgan et al , U.S. Patent Application No. 2011/0136697; Morgan et al , U.S. Patent No. 7,972,994;

Morgan et al , U.S. Patent No. 7,989,395; Morgan et al, U.S. Patent No.

8,410,028; Morgan et al , U.S. Patent No. 8,598,089; Morgan et al , U.S. Patent Application Serial No. 14/085,271; Wagner et al, U.S. Patent Application No. 2012/0053901; and Keefe et al , U.S. Patent Application No. 2014/0315762).

Synthesis of multivalent linking entities and monovalent linking entities such as those described in Figures 1-14 is conventional and within the ambit of the skilled artisan. Examples of both multivalent and monovalent linking entities attached to nucleic acids are provided in Figures 15 and 16. For example, attachment of linkers to the 5' nucleoside and/or or the 3' nucleoside is well known in the art. In some instances, the 5' hydroxyl and/or 3' hydroxyl of an oligonucleotide may be acylated or carbamylated to provide a linking or multivalent entity. In other implementations, the 5' hydroxyl and/or 3' hydroxyl of an oligonucleotide can be converted to another functional group (e.g., -SH or NH2) which may then be acylated or carbamylated to yield a linking or multivalent linking entity. In other embodiments, a nucleotide functionalized with 5' or 3' linker may be used as a starting material for oligonucleotide synthesis or may be attached enzymatically to the 5 ' end or 3 ' end of an oligonucleotide to provide a linking or multivalent linking entity attached to oligonucloetide. Other methods for attaching linkers or multivalent linkers to the 5' end or 3' end of an oligonucleotide are known in the art.

Attachment of linkers to internal nucleosides to provide linking or multivalent linking entities may be accomplished by use of functionalized nucleic acid subunits in oligonucleotide synthesis. Modified nucleotides with a free primary amine such as, for example, 5-aminoallyl-UTP, 5-acryl-6-amino hexamide-UTP, 7-deaza-7-propargylamin-dATP, N6-(6-Amino)hexyl-dATP and 5-propargylamin-dCTP can be incorporated into oligonucleotides during synthesis and then subsequently reacted with a linker or multivalent linker to yield a linking or multivalent entity to provide a linking entity. Many other modified nucleic acid subunits are known in the art and may be incorporated at internal nucleic acid subunits as well as at the 3' or 5' end of an oligonucleotide (e.g., Abendroth et al, Angew Chem. Int. Ed., 2011, 50, 8592).

Multivalent linking entities may be used in lieu of monovalent linking entities in combinatorial library synthesis (Harbury, et al, U.S. Patent No.

7,479,472; Pedersen et al, U.S. Patent No. 1, 211,113; Pedersen et al, U.S. Patent No. 7,413,854; Gouliev et al, U.S. Patent No. 7,704,925; Franch et al, U.S. Patent No. 7,915,201; Gouliev et al, U.S. Patent No. 8,722,583; Freskgard et al, U.S. Patent Application No. 2006/0269920; Freskgard et al, U.S. Patent Application No. 2012/0028812; Hansen et al, U.S. Patent No. 7,928,211; Hansen et al , U.S. Patent No. 8,202,823; Hansen et al , U.S. Patent Application No. 2013/0005581; Hansen et al, U.S. Patent Application No. 2013/0288929;

Morgan et al, U.S. Patent No. 7,972,992; Morgan et al, U.S. Patent No.

7,935,658; Morgan et al, U.S. Patent Application No. 2011/0136697; Morgan et al, U.S. Patent No. 7,972,994; Morgan et al, U.S. Patent No. 7,989,395; Morgan et al, U.S. Patent No. 8,410,028; Morgan et al, U.S. Patent No. 8,598,089;

Morgan et al, U.S. Patent Application Serial No. 14/085,271; Wagner et al, U.S. Patent Application No. 2012/0053901; and Keefe et al, U.S. Patent Application No. 2014/0315762). Use of multivalent linking entities such as those described above in place of monovalent linking entities, will provide combinatorial libraries which include multivalent oligonucleotides.

In general, the above methods are variations of the "split and pool" strategy. In some of the methods, the oligonucleotide only records the sequence of the synthesized ligand. In these methods, a solution a first building block linked to an coding oligonucleotide, is divided ("split") into multiple fractions. In each fraction, the first building block linked to the coding oligonucleotide is reacted with a second, unique, building block and a second, unique

oligonucleotide is added which identifies the second building block. These reactions can be simultaneous or sequential and, if sequential, either reaction can precede the other. The dimeric molecules produced in each of the fractions are combined ("pooled") and then divided again into multiple fractions. Each of these fractions is then reacted with a third unique (fraction -specific) building block and a third unique oligonucleotide which codes the building block. The number of unique molecules present in the library is a function of (1 ) the number of different building blocks used at each step of the synthesis, and (2) the number of times the pooling and dividing process is repeated. A variation of the split and pool method is described by Harbury, et al. ,

U.S. Patent No. 7,479,472. Here, an oligonucleotide is used to both direct ligand synthesis and record the sequence of the ligand. Splitting is accomplished by hybridization of an oligonucleotide with n discrete unique hybridization sequences with a complementary immobilized oligonucleotide with one of the n hybridization sequences in each round. As above, the number of unique molecules present in the library is a function of (1) the number of different building blocks used at each step of the synthesis, and (2) the number of times the pooling and splitting process is repeated. Importantly, in this method, iteration and evolution of the ligand are possible. Multivalent linking entities can be use as starting materials in any of the above methods to yield combinatorial libraries which include multivalent oligonucleotides. For example, a multivalent oligonucleotide such as those depicted above could replace the headpiece used in the procedures described in Morgan et al. , Wagner et al. or Keefe et al, above to provide multivalent ligands in combinatorial libraries. Similarly, use of a multivalent linking entity (see Example 7) can provide combinatorial libraries which include multivalent oligonucleotides.

The distance between ligands may be an important variable in multivalent oligonucleotides. Ligands which are spatially adjacent or distant may not be able to simultaneously bind the target of interest, which diminishes the chelate effect of multivalency. Accordingly, ligands in multivalent oligonucleotides may need to arranged at an optimal distance to maximize the chelate effect. In some embodiments, the distance between ligands is between about 3 nm and about 30 nm. In other embodiments, the distance between ligands is between about 5 nm and about 25 nm. In other embodiments, the distance between ligands is between about 10 nm and about 15 nm. At least two variables may be manipulated to achieve an optimal distance between ligands in a multivalent oligonucleotide. First, the spatial distance may be adjusted by varying the number of bases between the ligands in a multivalent oligonucleotide. Second, linkers of variable length may be used to adjust the distance between ligands in a multivalent oligonucleotide. Third a combination of the above methods may be used to adjust the distance between ligands in a multivalent oligonucleotide. Finally, stem-loop multivalent oligonucleotides such as those depicted in Figures 7-11 with or without linkers of variable length may be used to vary the distance between ligands in a multivalent

oligonucleotide.

Identification of low affinity ligands is an important facet of preparing and using combinatorial libraries which include multivalent oligonucleotides. In some embodiments, ligands of binding affinity of 1 mM are identified by screening combinatorial libraries which include multivalent oligonucleotides against biological targets including cells, tissues or membrane preparations. In other embodiments, ligands of less than 100 μΜ are identified by screening combinatorial libraries which include multivalent oligonucleotides against biological targets including cells, tissues or membrane preparations. In still other embodiments, ligands of less than 10 μΜ are identified by screening

combinatorial libraries which include multivalent oligonucleotides against biological targets including cells, tissues or membrane preparations. In still other embodiments, ligands of less than 1 μΜ are identified by screening combinatorial libraries which include multivalent oligonucleotides against biological targets including cells, tissues or membrane preparations. Low affinity ligands, such as those above, may be optimized by a number of methods including conventional medicinal chemistry, preparation of hit expansion combinatorial libraries or evolution.

The following examples are provided for illustrative purposes only and are not intended to limit the scope of the invention.

EXAMPLES Example 1: Generation of AP.T tool compounds

Tool compounds were used for direct detection of APJ on the surface of expression cells (to confirm ligand binding and optimize expression) and as surrogates for DEL components. Apelin-17 (K17F) was chosen as the apelin surrogate because of a strong affinity for APJ. In addition, the length of Apelin- 17 is sufficient to minimize steric hindrance between conjugated molecules (such as FAM and DNA) and the receptor. Ligand and conjugated molecules were separated by a TEG linker to reduce the impact of steric hindrance (Figure 17 A: N ₃ -TEG-K17F, Figure 17B: FAM-TEG-K17F). Two affinity measurements were used to analyze the strength of the K17F-APJ interaction: a radioligand binding assay (conducted at GVK-Bio (India)) and the DiscoverX reverse agonist bioassay (enzyme complementation assays for reverse cAMP agonism and β- arrestin recruitment were conducted at DiscoverX (Fremont, CA)) as disclosed in Table 1, below. An azide was included for conjugation to DNA using the K17F peptide. Incorporation of TEG, FAM or azide did not appear to significantly reduce the affinity of K17F for APJ and the above K17F peptides conferred similar β-arrestin affinities to APJ.

Table 1

FAM-TEG- K17F 1.41 2.6 3.8 K17F derivative

K17F-TEG- Primer ND 4.1* 6.7* F primer conjugate

K17F-TEG- Primer ND 5* 3.8* R primer conjugate

K17-Spike 220 bp gene ND >10 >10 conjugate

K17-Spike- 220 bp gene ND 5.3** >10 K17 conjugate

Corrected for yield

* Not corrected for yield

To generate a ligand-DNA conjugate compatible with panning, K17F was conjugated to DBCO-TEG-XA (5 ' GCTCGTCGCATTCGGC ACGC ) and DBCO-TEG-XF' (5 ' ACGGCGCGTGAGGAGGCAC A) oligonucleotides (TriLink Biotechnologies) using the Cu ²⁺ -free click reaction. DBCO-TEG-XA primer and DBCO-TEG-XF' primer were used in the Cu ²⁺ -free click reaction to link to K17F-TEG-N3, with desired products produced in 46% and 38% yield, respectively, as described in Examples 5 and 6, respectively. Primer conjugates were then purified by filtration and incorporated into PCR, containing the indicated Spike gene (Table 2). The resulting 220 bp genes were then purified by PCR purification kit (Qiagen). These oligonucleotides were included in PCR, using different 220 bp spike genes as a template, to generate monovalent and bivalent ligand-DNA conjugates (Table 2). It should be noted that K17F-DNA conjugations had yields of around 50%, so for the monovalent K17F-conjugates, about half of the DNA molecules had a K17F conjugate attached. For the bivalent K17F-DNA conjugate, only around 25% of the DNA was conjugated to two K17F peptides (although most of this DNA had at least one peptide). The K17F-conjugated full double- stranded 220 bp constructs were submitted for affinity assays using the DiscoverX platform (Table 1). The ligand- oligonucleotide conjugates had similar affinities to APJ as their precursor molecules. However, the monovalent K17F-spike conjugate had a lower affinity than K17F alone, although this affinity was increased somewhat by bivalency (Table 1).

Table 2

Example 2: Confirmation and Quantitation of APJ cell surface

APJ expression cell lines were provided (Sanofi), in CHO and HEK293 expression cells. Expression of APJ in these cell lines is inducible by tetracycline using the Flp-In TREX expression system (ThermoFisher, San Jose (CA)). These cell lines were accompanied by matched negative control cell lines, containing

Flp-In plasmids for inducible expression but lacking the plasmid containing APJ.

CHO cells were cultured in Ham's F12 media (Sigma- Aldrich, St. Louis MO) with lx pen/strep (Fisher) and 10% FBS (Hyclone, VWR). HEK293 cells were cultured in DMEM high glucose media, again with lx pen/strep and 10% FBS.

Selective media for Flp-In cells also contained 100 μg/mL Zeocin (Fisher).

Selective media for Flp-In TRex (FITR) cells contained both 100 μg/mL Zeocin and 15 μg/mL blasticidin. Selective media for Flp-In TRex APJ cells contained 15 μg/mL blasticidin and 100 μg/mL hygromycin. For APJ induction, cells were seeded, typically in a T150 flask, at a lOVmL concentration, grown overnight, then 24 hrs later induced with 1 μg/mL (final concentration) tetracycline or 1 μg/mL doxycycline, incubated for another 24 hrs and then removed from the flask with Cell Stripper (VWR). Cells were washed once with PBS before proceeding to cell surface ELISA or panning.

To confirm expression of APJ after the cells were induced, fluorescence conferred by the K17F-TEG-FAM tool compound was used. Analysis by fluorescent activated cell sorting (FACS) was performed using cells induced using tetracycline for 24 hrs. and 1 μΜ K17F-TEG-FAM (Figure 18A). FACS was conducted by Aragen, Inc. (Morgan Hill, CA) with 10 ⁶ cells and 1 μΜ K17F-TEG-FAM in ice-cold PBS, with a 30 min incubation. For cell surface ELISAs, K17F-TEG-FAM (25 nM) and 0.1% NaN ₃ in incubation buffer (50 mM HEPES pH 7.4, 10 mM MgCh, 1 mM CaCh, 0.5% BSA, 1 mg/mL sssDNA) was incubated with 2xl0 ⁶ cells in a total volume of 100 μΐ, for 1 hr. at 4°C. Cells were then washed 3x cold with 50 mM Tris-HCl, pH 7.5, 0.5% BSA. Cells were incubated with 1/5000 anti-FAM-HRP in 250 μΐ, incubation buffer (without azide) for 1 hr at 4°C. Cells were washed 3x again with Tris/BSA, 100 uL, TMB buffer was added, color was developed for 4 min and then 0.2 N H2SO4 was added. The cells were removed by centrifugation and the reaction read at 450 nm. For magnetic bead FAM standards, 2xl0 ⁶ negative control cells were added to these tubes, and control magnetic beads were also added to the APJ-induced cells to be analyzed.

Fluorescence by the APJ-expressing CHO cell line was roughly 10-fold more intense than that of the HEK293 line, although both cell lines had a significant shift in fluorescence. Concurrently, a cell-based ELISA using buffers derived from the radioligand assay was developed using the same cells and reagents, and an anti-FAM-HRP antibody for detection. A magnetic bead-based FAM standard curve was applied to the ELISA to serve as a standard to quantify APJ on the cell surface (Figure 18B). The APJ CHO cell line expresses roughly 50,000 receptors per cell, and 5,000 for the HEK293 cell line (Figure 18C). Correspondingly, the APJ CHO cell line was used for all cell-based panning experiments. Induction using 1 μg/mL tetracycline or 1 μg/mL doxycycline was equivalent (data not shown). Example 3: Cell-based Panning

A three-component sample library (Table 2), with a 1000-fold excess of a control DNA gene (Spike B l, at 1 pmole/tube) to each ligand-DNA conjugate (at 1 fmole/tube), was used to develop a protocol for APJ-specific cell-based panning. A non-enzymatic cell dissociation agent was used to remove the cells from their culture flasks, then centrifuged to concentrate the cells, and centrifugation was used to wash the cells as well (Wu et al. , ACS Comb Sci. 2015;17(12):722-731). Passivating agents were included in the ligand-DNA conjugate binding mixture to reduce non-specific binding: sheared salmon sperm DNA (sssDNA) at 1 mg/mL and bovine serum albumin (BSA) at 0.5%. 0.5% BSA was also included in the wash buffer. Two methods of DNA "elution" (disruption of ligand-target binding to release the DNA gene) were applied: heat (95 °C for 10 min.) and base-detergent treatment (lysis using ES2 solution, followed by 95°C for 10 min.). ES2 solution is 20 mM NaOH, 15 mM NaCl, 0.02% SDS, 0.005% TritonXlOO.

More specifically, 10 ⁶ washed cells were used for each panning reaction. Cells were resuspended in HEPES-based (50 mM HEPES, pH 7.4, 10 mM MgCh, 1 mM CaCh, 0.5% BSA, 1 mg/mL sssDNA, 0.1% NaN ₃ ) or PBS-based (20 mM phosphate buffer, pH 7.4, 150 mM NaCl, 10 mM MgCh, 1 mM CaCh, 0.5% BSA, 1 mg/mL sssDNA, 0.1% NaN ₃ ) panning buffer containing 1 pmole of Spike B l and 1 fmole each of the ligand conjugates in Table 2 (this is the 1000- fold excess of background DNA). For the 10 ⁶ -fold excess of background DNA, 1 pmole of Spike Bl and 1 amole of each ligand conjugate was used. Axygen low retention tubes were used for panning. The cells in the panning mix were incubated at 4°C for 1 hr while shaking. The cells were then washed by centrifugation (1 min, 5 kG) four times (2 and 6 for optimization as well), transferring all cells to a new tube before the final wash. Wash buffer was 50 mM Tris-HCl, pH 7.5, 0.5% BSA for the HEPES-based protocol and lxPBS, pH 7.4, 0.5% BSA for the PBS-based protocol. After removing the final wash, cells were lysed by either heat (95°C for 10 min.) or ES2+heat (with 10 1 M Tris H 7.5 added, then incubated at 95°C for 10 min.) as indicated. Debris was then centrifuged and removed, and the remaining milieu was prepared for NGS.

Two-step and one-step NGS prep strategies were employed. For both the two-step and one-step NGS prep, between 30 and 35 total cycles were performed, using Phusion master mix (New England Biolabs). Prepped DNA was then purified by PCR purification kit and sent to Seqmatic (Fremont, CA) for sequencing. Numbers of Spike genes were counted and to get a specific enrichment, counts from APJ-expressing cells were divided by counts from control cells. Prior to panning, an experiment was conducted to determine the optimal number of washes for this suspension cell-based panning procedure. Maximum removal of background DNA gene (Spike B l), as analyzed by qPCR, was achieved after approximately four washes (Figure 19A). This resulted in roughly a 10,000-fold reduction in amount of the background DNA gene. After four washes, roughly half of the cells were retained, which was deemed an acceptable recovery (Figure 19B).

With the three-component library, washing stringency using the APJ cell ELISA conditions (HEPES-based low salt binding buffer, salt-free Tris-based washing buffer) was tested in a full-fledged cell-based panning experiment. Four washes were again optimal (Figure 20 A, ES2 treatment for cell lysis and elution). Despite the abundance of blocking agents, there was a considerable amount of library DNA binding to the polypropylene tubes. In subsequent experiments, low-retention tubes were used, and the panning sample was moved to a separate tube before the final wash to eliminate most vessel-bound library. To ensure that the observed enrichment of K17F conjugates was specific to APJ, an

overwhelming excess of K17F-FAM (a 100,000-fold higher concentration K17F- FAM than K17F conjugates) was used to compete away the conjugate. When competitor was added, enrichment of K17F conjugate species was negligible (Figure 20B). A different binding/washing regimen (PBS-based high salt binding buffer, PBS-based washing buffer) was then used. This new regimen enhanced the binding of the bivalent conjugate to APJ cells, resulting in a reproducible specific fold enrichment of above 20 (Figure 20C). Two methods of DNA "elution" from the cells were investigated using the previously-described HEPES-based protocol: elution by heating (95°C) and by lysis (ES2, then 95°C). These two methods are similar in terms of specific enrichment (Figure 21), but certain experiments can only employ one or the other (psuedorounds can only be performed by heat elution, while adherent cell panning can only be eluted by lysis).

Cell-based panning was also attempted while using a 10 ⁶ -fold excess of background. In this regimen, only 1 amole (6xl0 ⁵ molecules) of each ligand- conjugate was included per sample. Only an 8-fold specific enrichment of the bivalent K17F conjugate was observed, while the monovalent K17F conjugate did not appear to be enriched (Figure 22).

Example 4: Panning by internalization

Panning was done using adherent cells on 6-well plates, including serum- free media, 1 mg/mL sssDNA, and the three-component library. Cells were washed once with serum-free media, then panning mix was added at 1 mL/well. Cells were incubated at 37°C for 15 min., then washed with TBS twice, and 1 mL of benzonase solution was added to half the wells (1/10,000 din in TBS + 2 mM MgC12) (the other half of the wells received buffer but no benzonase). Plates were incubated for 15 min. @ 37°C, then washed twice with TBS + 10 mM

EDTA. The cells were then lysed with 200 uL/well ES2, 20 1 M Tris, pH 7.5 was added, and the mixture was incubated at 95°C for 10 min. Debris was spun out, and lysate prepared for NGS.

Internalization of receptor and ligand can be utilized to protect ligand conjugates, allowing the receptor-ligand interaction to be somewhat

unidirectional, and insulating bound ligand-DNA conjugates from the environment outside the cell. As part of the proof of concept for APJ cell-based panning, the three-component library plus passivating agents was incubated with adherent cells (both APJ-expressing and control) under conditions where K17F would be internalized. A 15 min incubation of library with cells, 0.5% BSA, and 1 mg/mL sssDNA in serum-free media was conducted at 37°C as described above. This was followed a 15 min incubation of cells with benzonase to remove cell surface bound DNA. After a series of EDTA washes (to inactivate and remove the benzonase), cells were lysed and processed by NGS. As with cell surface panning, internalization panning yielded around 20-fold specific enrichment, with or without benzonase treatment (Figure 23). Accordingly, this strategy can still be a valuable alternative approach to cell-based panning. As this ligand-target is theoretically less-dependent on a slow off-rate, it could end up being an indispensable strategy for uncovering lower affinity ligands. Example 5: Preparation of DBCO-amine modified oligonucleotide coupled with K17F

Into a filter plate with DEAE Sepharose resin was added an amine modified oligonucleotide solution. In a separate Eppendorf was mixed dibenzocyclooctyne-acid (DBCO) and Ν,Ν-Diisopropylethylamine (DIPEA). To the DBCO/DIPEA solution was added 4-(4,6-Dimethoxy-l,3,5-triazin-2-yl)-4- methylmorpholinium tetrafluoroborate (DMTMM). This solution was added to the amine modified oligonucleotide bound on resin. After the reaction was complete the product was eluted from the resin. The DBCO-amide modified oligonucleotide and K17F-TEG-azide were mixed in an Eppendorf tube. After the reaction was complete the product was 46% pure by HPLC.

Example 6: Preparation of DBCO-amine modified reverse oligonucleotide coupled with K17F

Into a filter plate with DEAE Sepharose resin was added an amine modified reverse oligonucleotide solution. In a separate Eppendorf was mixed dibenzocyclooctyne-acid (DBCO) and Ν,Ν-Diisopropylethylamine (DIPEA). To the DBCO/DIPEA solution was added 4-(4,6-Dimethoxy-l,3,5-triazin-2-yl)-4- methylmorpholinium tetrafluoroborate (DMTMM). This solution was added to the amine modified reverse oligonucleotide bound on resin. After the reaction was complete the product was eluted from the resin. The DBCO-amide reverse modified oligonucleotide and K17F-TEG-azide were mixed in an Eppendorf tube. After the reaction was complete the product was 38% pure by HPLC.

Example 7: Preparation of multivalent conjugates using DNA- programmed combinatorial chemistry

Monovalent and multivalent DPCC conjugates were prepared by sequentially building tetrapeptides on 220-mer D A 'genes' with monovalent amine or multivalent amine groups at their 5' ends, respectively. Monovalent (D05.3. (Figure 16A, Glen Research)) and multivalent (equimolar mixture of di and tri amine forms, i.e. , D018.1 (Figure 16, Glen Research) and D019.1 (Figure 16C, Glen Research)), XA oligonucleotides

(5 ' GCTCGTCGCATTCGGCACGC' ) (TriLink Biotechnologies) were extended to 220-mer genes through reverse-transcription as described in (Harbury et ai, U.S. Patent No. 7,479,472). Different DPCC gene sequences were used to encode the six test conjugates (three tetra-pe tides in mono-valent and multivalent formats). The test tetrapeptide conjugates were built as part of a library in three steps. Each step consisted of separating the different DPCC genes into different wells using preparative hybridization, eluting the DNA from each well and immobilizing it on DEAE sepharose in a cognate well, performing a chemical modification with the appropriate mono- or di-peptide and eluting the modified DNA from the DEAE sepharose before pooling it with the others for the next step. The chemical modification step consisted of exposing the DEAE- immobilized DNA with one or more reactive amines to FMOC-protect mono- or di-peptide activated with DMTMM in 50% DMF/methanol for 30 rains at room temperature two times, washing with DMF, removing the protective FMOC group with 20%; piperide in DMF and eluting the modified DN A using 0.5M NaCl. Three test tetrapeptides YGGF (K _D ~0.19uM), YGGL (K _D ~3.8uM), and YGFL (KD~28UM) were synthesized in monovalent and multivalent forms in three chemical steps: first coupling with FMOC-F or FMOC-L, second coupling with dipeptides FMOC-GG or FMOC-GF, and third coupling with FMOC-Y.

Example 8: Selection of multivalent vs monovalent display compounds After synthesis, the DPCC conjugates were converted to full double- stranded format through primer-extension with Dream Taq (ThermoFisber, San Jose, CA). A positive control test-gene was made by directly coupling FMOC- YGGFL peptide (KD~0.007uM) to a monovalent amine oligo and two negative control genes consisted of an unreacted monovalent or multivalent oligos. A test panning mixture was made consisting of a background DPCC gene (5 H) at 10,000-fold excess compared to the test conjugates. The panning experiment was performed using Phynexus tips (5ul ProA resin) with a Rainin PureSpeed pipet system. Tips were incubated with panning buffer (50mM Tris-HCl pH7.5, 150mM NaCl, 0.1% Tween-20, 0.25mg/ml yeast tRNA (Ambion), Img/ml BSA (Ambion), lniM 2-mercaptoethanol) with ('+3E7') or without ('-3E7') 22ug (300pmol) of 3E7 antibody for lh (-17 cycles at low speed) and washed to remove any unbound target. Panning mixture (2pmol) was applied in 60ul of binding buffer and washed either 3 or 15 times with lOOul washes. Conjugates retained on the tips were eluted using glycine pH2.2 (3xl5ul), which was neutralized with 5ul of 1M Tris pH 9.5. Aliquots of the pre- and eluate mixtures were amplified for NGS by PGR. Enrichment for a particular gene was defined as ratio of the frequency of that gene in the eluate to its frequency in the input. As shown in Figure 24, in the '+3E7, 15 washes' selection condition, the monovalent conjugates showed about 10-fold lower enrichment compared to the multivalent conjugate with the same peptide. The difference was very stark for the lower affinity peptide (YGFL), which showed no obvious enrichment (Ix) in the monovalent form but showed 17-fold enrichment in the multivalent form. The enrichment was 3E7-dependent because all the genes showed no significant enrichment (0.25x-1.5) in the '-3E7' selection. Similar results were observed in the '+3E7 and -3E7, 3 washes' selection condition (data not shown), where the multivalent format performed 5- to 10-fold better than the monovalent format.

All publications and patents cited herein are incorporated by reference in their entirety to disclose and describe the methods and/or materials in connection with which the publications are cited. It is understood that the present disclosure supersedes any disclosure of an incorporated publication to the extent there is a contradiction. The publications discussed herein are provided solely for their disclosure prior to the filing date of the present application. Nothing herein is to be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior invention. Further, the dates of publication provided may be different from the actual publication dates which may need to be independently confirmed.