Genetically encoded bicyclic peptide libraries

Abstract

The present invention relates to a bicyclic peptide complex comprising a peptide construct, the construct comprising: (i) a polypeptide having a free terminus (N or C); (ii) optionally a nucleic acid encoding the polypeptide; and (iii) a dyad-symmetric linker (TSL) compound attached to the polypeptide, wherein the linker is attached to the terminus of the polypeptide and to at least two distinct side chains of the peptide via a covalent bond. The present invention also relates to a library and methods for making the complex, and to screening methods using the same.

Description

[Technical Field]

[0001] This application relates to the fields of receptor-ligand interactions and molecular recognition. More particularly, this application relates to bicyclic peptide libraries, methods for producing libraries of bicyclic peptides, and the use of such libraries in various assays. [Background technology]

[0002] The generation of libraries of small molecules and the selection of those molecules that uniquely bind to a target of interest are important for drug discovery. The production of genetically encoded libraries, in which each library member is linked to an information template such as DNA or RNA, allows for the processing of large chemical libraries without the need to separate individual library members into separate solutions or reaction vessels. Target molecules can be selected from a mixture of genetically encoded molecules, and the information template can be used to identify or amplify the selected molecule of interest.

[0003] Peptides containing intramolecular covalent bonds and exhibiting bicyclic topology are known to be more stable to proteolysis and other forms of degradation than monocyclic or linear peptides. They are also known to exhibit stronger binding interactions with protein targets; this increased interaction is hypothesized to be due to the reduced conformational penalty of the constrained bicyclic system compared to linear or cyclic systems.

[0004] Phage display is a well-known technique used to analyze, display, and produce genetically encoded libraries of peptides and proteins (Scott et al., 1990). Phage display is a process during which phage are engineered to expose, or "display," different peptides or proteins, such as human antibodies, on their surface. Through genetic engineering, peptides or proteins of interest are individually bound to phage cell surface protein molecules (usually gene III protein or g3p). In such phage populations (phage libraries), each phage carries a gene for a different fusion peptide or protein of g3p and exposes it on its surface. Through various selection procedures, phage that "display" binders to specific target molecules of interest can be identified and isolated. These binders may include protein interaction partners for determining new functions or mechanisms of function of proteins, peptides that recognize and bind antigens (e.g., for use in diagnostic and therapeutic targets), and proteins involved in protein-DNA interactions (e.g., novel transcription factors).

[0005] Phage display can be extremely useful in the discovery and development of pharmaceutical and / or diagnostic polypeptides. In phage display, whole phages can bind and be eluted from immobilized target molecules. Phages remain infectious and can be amplified by injecting their DNA into bacterial cells. Phage display methods are typically limited to the production of libraries that can be encoded by direct DNA-RNA-protein information transfer. These methods are typically limited to linear sequences of peptides made only of the 20 natural amino acids or cyclic peptides cross-linked through disulfide bonds.

[0006] RNA and ribosome display are other techniques known in the art that allow the display of naturally occurring peptides on information templates. Amplification of libraries of RNA-bound peptides requires an in vitro translation system to generate or re-amplify the library. Suga et al. (US20100168380 A1) teach the production of cyclic peptides containing N-methyl amino acids and other unusual (non-standard) amino acids by encoding unnatural amino acids into RNA sequences. There are no examples of direct encoding of bicyclic peptides in RNA and the production of bicyclic libraries by translation alone.

[0007] It is known to produce libraries of bicyclic peptides or displays of bicyclic peptides on phage, DNA, or RNA by modifying the encoded display of peptide-derived molecules via chemical post-translational modification (cPTM). Typically, these methods use organic synthesis of peptides to create peptide derivatives. It is known that entire peptide libraries can be modified by uniform chemical modification. Selection from the modified library and DNA sequencing yields peptide sequences from which modified peptide derivatives can be created. Several methods exist that involve the conversion of peptide libraries, phage-displayed polypeptide libraries, and RNA-displayed polypeptide libraries into libraries of peptide derivatives.

[0008] Suga et al. described a methodology for synthesizing bicyclic peptides displayed on RNA by using cysteine (Cys) and three different non-proteinogenic amino acids, Cab, Aha, and Pgl, simultaneously incorporated into the peptide chain. The first cyclization occurred in situ during translation between the chloroacetyl group of Cab and the sulfhydryl group in Cys, and the second cyclization at the side chain of Aha-Pgl was carried out via Cu(I)-catalyzed azide-alkyne cycloaddition. (J. Am. Chem. Soc., 2008, 130(23), pp. 7232-7234) Hartman and coworkers used a different approach, combining two Cys residues with non-proteinogenic amino acids, azidohomoalanine (AzHA) and p-ethynylphenylalanine (F-yne) (ACS Chem. Biol. 2017, 12, 795-804). The first cyclization occurred by crosslinking the cysteine with dibromo-m-xylene, and the second cyclization at the side chain of AzHA-F-yne was performed via Cu(I)-catalyzed azide-alkyne cycloaddition. Expressing unnatural amino acids (UAAs), such as Aha, Pgl, and F-yne, requires specialized non-natural translation systems; the expression of multiple unnatural side chains is difficult in display systems such as phage display, resulting in low incorporation efficiency of UAAs. Therefore, it is of interest to develop methods using peptides composed of natural amino acid residues.

[0009] U.S. Patent Publication WO 2004 / 077062 describes a method for modifying a plurality of unprotected peptides consisting of natural amino acid residues in water with symmetric linkers to produce libraries of cyclic and bicyclic peptides. This method can be extended to other three-fold symmetric linkers with three thiol-reactive groups, but the number of linkers with such high symmetry is limited. Prior art examples (WO2009098450A2, WO2004077062, WO2011018227A2) and peer-reviewed literature ((Nat. Chem. Biol. 2009, 5, 502-507; Angew. Chem. Int. Ed. 2014, 53, 1602-1606) describe the synthesis of bicyclic peptides limited to three-fold symmetric connector compounds with sulfur-containing side chains and three identical reactive groups, such as tris-(bromomethyl)benzene (TBMB). The production of bicyclic peptides with linkers of lower symmetry (two-fold) is not evident from such prior art. Prior art examples using two-fold symmetric linkers to modify multiple unprotected peptides consisting of natural amino acids are limited to the production of monocyclic peptides (ACS Chem.Biol.2016,11,1422-1427;J.Am.Chem.Soc.2014,136,5880-5883;Bioconj.Chem.2 016,27,509-514;Chem.Sci.2016,7,3785-3790;Org.Biomol.Chem.2016,14,5539-5545).

[0010] To access a broad chemical diversity space and maximize the chances of finding bicyclic structures with desired chemical or biological properties, it is interesting to generate libraries using linkers with lower symmetry, such as two-fold symmetric linkers. While the application of lower symmetry modifications to unprotected peptides has been reported, all previous examples demonstrate that when applied to modifying unprotected peptides, lower symmetry linkers produce complex mixtures of bicyclic peptides. Heinis and coworkers (Nature Chemistry 2018, 10, 715) specifically demonstrated that the use of two-fold symmetric linkers reacting with amino acid side chains produces complex mixtures of multiple bicyclic structures. Such mixtures may not be separable by chromatographic or other techniques. Liu, Heinis et al. demonstrated that two-fold symmetric linkers applied to side chain modifications produce complex mixtures of products (Angew. Chem. Int. Ed. 2017, 56, 4458), and that to avoid such heterogeneous mixtures, it is necessary to use unconventional amino acids with unnatural side chains. These and other examples generally emphasize the well-understood fact that when two or fewer symmetric modifiers are used to modify three or more similar reactive groups in a peptide, such reactions produce uncontrolled and complex mixtures of products.

[0011] Unconventional methods for synthesizing bicyclic peptides are known, often requiring the use of one or more of the following factors: (i) amino acids containing unusual reactive groups not found in natural amino acids, (ii) protecting groups for the amino acids, (iii) solid supports for organic synthesis, (iv) organic solvents, and (v) reaction conditions that are incompatible with biomolecules such as DNA, RNA, and biomolecular complexes such as phages. There appear to be no examples of the use of less symmetric linkers for the synthesis of bicyclic peptides that lack protecting groups and involve modifications of peptides composed of natural amino acids. Such methods are of interest because they could provide an unobstructed route to the synthesis of a wide variety of genetically encoded bicyclic structures.

[0012] U.S. Patent Publication No. WO2009098450 A3 describes a method for modifying a genetically encoded peptide library displayed on phage using a three-fold symmetric linker. This method produces a library of bicyclic peptides with a free amino terminus and requires the use of a three-fold symmetric linker with a thiol-reactive group. The difficulty of generating a phage library with an odd number of cysteines (here, three) is known in the art.

[0013] The above methods described in the prior art are believed to produce libraries of bicyclic peptides, each with a free amino terminus. This terminus is known to be susceptible to proteolytic cleavage. Methodologies for producing polycyclic peptides without a free N-terminus are known, but such methods require the incorporation of non-proteinogenic amino acids containing an N-chloroacetyl (ClAc) group. No methods are known for producing genetically encoded bicyclic libraries with blocked N-terminus from natural amino acids.

[0014] This background information is provided for the purpose of making known information believed to be of possible relevance to the present invention. It is not necessarily intended, nor should be construed, that any of the preceding information constitutes prior art against the present invention. Summary of the Invention

[0015] In general terms, the present invention includes bicyclic peptides without an N-terminal residue, created by modifying a peptide or a plurality of peptides (libraries) with a linker. Preferably, the peptide or peptide library is unprotected. Natural protein and native amino acids It includes a polypeptide comprising or consisting of:

[0016] In some embodiments, the present invention includes a genetically encoded bicyclic peptide library that does not have an N-terminal residue, created by modifying a genetically encoded peptide library with a linker.

[0017] Thus, in some embodiments, the present invention provides a method for treating a cancer cell comprising: (a) a polypeptide; (b) a nucleic acid encoding a polypeptide, and optionally, an identification tag that uniquely identifies the polypeptide; and (c) a bicyclic structure resulting from attaching to a polypeptide a linker having a first terminal reactive group (A) and a second terminal containing two reactive groups (B2, B2), wherein the first terminal is covalently ligated to a terminal end of the polypeptide and both reactive groups at the second terminals of the linker are covalently attached to side chain residues of the polypeptide. The peptide may comprise a genetically encoded peptide construct comprising:

[0018] In some embodiments, the linker has two-fold symmetry in that the two second terminal reactive groups are identical. In some embodiments, the polypeptide terminus is the N-terminus, and the side chain residues are cysteine, lysine, or tyrosine, respectively. Alternatively, the polypeptide terminus is the C-terminus, and the side chain residues are cysteine, lysine, or tyrosine, respectively. Then, the first terminal reactive group of the linker is a group that has a unique reactivity to the C-terminus, such as C-terminus-selective photoredox decarboxylation conjugate addition at acidic pH (Org. Lett., 2015, 17, 4830-4833 and Nature Chemistry 2018, 10, 205-211). In some embodiments, the first terminal reactive group of the linker is an aldehyde-reactive group such as an oxime, hydrazine, 2-aminobenzamidoxime, phosphonium ylide, sulfur ylide, nitrogen ylide, or any other carbon nucleophile and carbenoid reagent known to be reactive with aldehydes and stable in aqueous environments.

[0019] In some embodiments, the second terminal reactive group of the linker is an electrophilic group reactive with thiols, amines, or phenols. Examples of thiol-reactive groups include haloketones, haloacetamides, halobenzyls, maleimides, acrylates, carbonyl acrylate reagents, 3-arylpropiolonitriles, alenamidofluoroarenes, chlorotetrazines, Julia-Koscienski-like reagents, 2-azidoacrylates, organometallic palladium reagents, Michael acceptors containing conjugated C-C double bonds, including organogold(I) reagents, conversion of thiols to dehydroalanine (Dha) followed by conjugate addition to Dha, or any other method that specifically reacts with thiol residues in water. Examples of groups that specifically react with lysine amine residues in water include N-hydroxysuccinimide esters, aryl esters, perfluoroaryl esters, perfluoroarenes, ketenes, ortho-phthalaldehyde strain-releasing amine modifiers, or other groups that specifically react with lysine amine residues in water. Examples of groups reactive with tyrosine or other aromatic side chains include allylpalladium (J. Am. Chem. Soc. 2006, 128, 1080-1081), diazodicarboxylate (J. Am. Chem. Soc. 2010, 132, 1523-1525), diazonium salts, aniline-formaldehyde hemiaminal, rhodium carbenoids, dirhodium metallopeptide catalysts, manganese-catalyzed C—H alkynylation, Wother's reagent, 1-[(triisopropylsilyl)ethynyl]-1,2-benziodoxol-3(1H)-one (TIPS-EBX), gold(I)-catalyzed, selective ruthenium-(II)-catalyzed C—H activation, palladium(II) acetate-catalyzed C—H activation with aryl iodides in water, and other reagents known in the art for specifically modifying the phenol of a tyrosine residue (Biochemistry 1999, 128, 1080-1081). 2017,56,3863-3873).

[0020] In some embodiments, the polypeptide comprises two cysteine residues and an N-terminal serine or threonine residue, the N-terminal serine or threonine being first converted to an aldehyde by selective oxidation for reaction with a first terminal reactive group of the linker, and the second terminal reactive group of the linker being covalently attached to the cysteine residue.

[0021] In some embodiments, the complex may be attached to a carrier such as a phage particle that carries the polypeptide externally and includes a nucleic acid encoding the polypeptide. Alternatively, the complex may be an RNA display compound that carries the polypeptide, encodes the polypeptide, and includes an RNA sequence linked to the polypeptide. Alternatively, the complex may be a DNA display compound, where the DNA display includes a DNA that carries the polypeptide and encodes the polypeptide linked to the polypeptide. Alternatively, the complex may be a polypeptide linked to a polymer or protein carrier along with an identification tag such as another peptide.

[0022] In another aspect, the present invention may include a method for producing a phage display complex, the method comprising: (i) providing a phage particle comprising a polypeptide having a terminus; (ii) providing a linker having a first terminus reactive group and a second terminus comprising two reactive groups, and ligating the first terminus reactive group of the linker to the polypeptide terminus to form a covalent bond to form an intermediate complex (the "ligation step"); and (iii) forming a bicyclic structure from the intermediate complex by reacting both reactive groups at the second terminus of the linker with side chain residues of the polypeptide (the "bicyclization step"). The ligation step and the bicyclization step are preferably separate and sequential steps and / or are preferably carried out at different pHs.

[0023] In some embodiments, the terminus of the polypeptide is N-terminal and is oxidized to an aldehyde prior to the ligation step ("oxidation step"), and the ligation step comprises mixing the polypeptide containing the N-terminal aldehyde with the linker in an aqueous buffer at an acidic pH. Preferably, the intermediate complex produced after the ligation step is purified by size exclusion purification, e.g., gel filtration or dialysis in an aqueous buffer at an acidic pH, preferably less than about pH 5.

[0024] In some embodiments, the intermediate complex, which can be purified, is reduced with a reducing agent such as TCEP ("reduction step") prior to the bicyclization step. The reduced intermediate complex can then be exposed to alkaline pH (>7) to induce bicyclization, with the second terminal reactive groups each containing a thiol-reactive group that reacts with two cysteine residues in the polypeptide.

[0025] In some embodiments, peptides are modified to introduce a diketone group that is orthogonally reactive to the first cycle. The diketone group can then be used to carry out a reaction (bicyclization reaction) to create a second cycle. The intermediate complex containing the first cycle and the diketone group can be purified and stored without degrading the reactivity of the diketone, which may be a 1,3-diketone. The intermediate complex can then be subjected to a number of reactions that utilize the unique reactivity of 1,3-diketones with hydrazine in biocompatible aqueous conditions.

[0026] In some embodiments, the 1,3-diketone functional group is introduced by reacting a peptide containing two cysteine residues with 1,5-dichloropentane-2,5-dione under neutral aqueous conditions (about pH 7 to about pH 8) to form a monocyclic peptide bearing the 1,3-diketone functional group.

[0027] In some embodiments, the second cyclic structure is formed by combining a monocyclic peptide containing a 1,3-diketone functional group with a molecule containing a hydrazine functional group and a functional group uniquely reactive toward the N- or C-terminus of the peptide. The reaction between these molecules creates a connection between the 1,3-ketone functional group and the terminus of the peptide, forming a second cycle in the peptide. Alternatively, hydrazine can be reacted with a 1,3-diketone to present a new functional or reactive group, which can then be linked to the N- or C-terminus of the peptide via a linear linker.

[0028] In another aspect, the present invention may include a method for measuring the yield of a reaction after at least one chemical reaction step described herein, comprising exposing the complex or polypeptide to a capture reagent reactive with the unreacted polypeptide but not the reacted polypeptide, and measuring the uptake of the capture agent by the affinity reagent. The capture agent comprises an affinity handle paired with a reactive group and an affinity reagent. The affinity handle-affinity reagent pair may include any affinity ligand pair known to have sufficiently high specific binding to allow quantification of the binding pair. Exemplary pairs include biotin-streptavidin, FLAG peptide-anti-FLAG antibody, sulfonamide-carbonic anhydrase, and methotrexate-dehydrofolate reductase (DHFR). Preferably, the affinity handle or affinity reagent is immobilized on a solid support, such as agarose beads.

[0029] In some embodiments, the scavenger reactive group is an aldehyde-reactive group, such as an aminooxy group, or a thiol-reactive group, such as an iodoacetamide group.

[0030] In some embodiments, the reaction measured is an oxidation step, in which the phage or complexes are exposed to aminooxybiotin (AOB) after the oxidation step, which reacts with the N-terminal groups oxidized to aldehydes, the reaction is diluted by at least one order of magnitude, streptavidin beads are added to retain the reacted complexes, and the difference in the number of phage retained on the streptavidin beads compared to the number of phage not retained is measured to determine the yield of the oxidation step (the percentage of phage particles that have acquired an aldehyde group).

[0031] In some embodiments, the reaction measured is a ligation step, where the phage or complexes are exposed to aminooxybiotin (AOB) before and / or after the ligation step, the reaction is diluted by at least one order of magnitude, streptavidin beads are added to retain the phage or complexes with unligated aldehyde groups, and the difference in the number of phage particles retained on the beads compared to the number of unretained phage is measured to determine the yield of the ligation step (percentage of phage particles that have lost their aldehyde groups).

[0032] In some embodiments, the reaction measured is a reduction step, in which phage or complexes are exposed to biotin-iodoacetamide (BIA) before and / or after the reduction step, the reaction is diluted by at least one order of magnitude, streptavidin beads are added to retain the reduced phage or complexes to expose thiol groups, and the difference in the number of phage particles retained on the beads compared to the number of unretained phage is measured to determine the yield of the reduction step (the proportion of phage particles that acquired thiol groups during the reduction step).

[0033] In some embodiments, the reaction measured is a bicyclization step, in which the phage or complexes are exposed to biotin-iodoacetamide (BIA) before and / or after the bicyclization step, the reaction is diluted by at least one order of magnitude, streptavidin beads are added to retain the phage or complexes with unreacted thiol groups, and the difference in the number of phage particles retained on the beads compared to the number of unretained phage is measured to determine the yield of the bicyclization step (the proportion of phage particles that lost a thiol group during the bicyclization step). Alternatively, the phage or complexes may be exposed to biotin-thiol (BSH), which reacts with benzyl chloride (or other thiol-reactive groups) before or after the bicyclization step, (i) the reaction is diluted by at least one order of magnitude, (ii) streptavidin beads are added, and (iii) the number of phage particles remaining after capture is measured, whereby the number of phage retained on the streptavidin beads constitutes the yield of the bicyclization step (the proportion of phage particles that lost a thiol-reactive group during the bicyclization step).

[0034] In another aspect, the present invention can include a library of genetically encoded ligands comprising a plurality of different polypeptide sequences, each of which forms a bicyclic structure with a linker as described herein. In some embodiments, at least two polypeptides P1 and P2 are encoded separately by DNA sequences. The plurality of different polypeptides can be modified with the same or different linkers. Each linker can be associated with a unique identification tag, such as a silent genetic barcode.

[0035] A mixed library of genetically encoded complexes can be formed by pooling libraries together to produce a library of complexes described herein. The mixed library can include at least two different peptide sequences (P1 and P2), each separately modified with at least two different linkers (L1, L2) to form at least four distinct complexes (P1L1, P1L2, P1L2, and P2L2). In some embodiments, linker L1 can react with two peptides related to a predefined nucleic acid code B1, and linker L2 can react with two peptides containing a predefined nucleic acid code B2. The nucleic acid code can include a silent genetic barcode, such as those described in WO2016061695-A1.

[0036] In another aspect, the invention includes a method for identifying a complex described herein in which a polypeptide bicyclic structure is capable of binding to a ligand, comprising: (i) contacting a library of complexes with a ligand; (ii) selecting those complexes that bind to the ligand; and (iii) identifying the structure of the binding complex by sequencing the nucleic acid encoding the polypeptide and / or linker structure. In some embodiments, the library of complexes can include at least four distinct bicyclic structures, P1L1, P1L2, P1L2, P2L2, as described above.

[0037] In another aspect, the invention includes a Nodal antagonist peptide comprising a bicyclic complex formed from a polypeptide comprising SPCQRGHMFC or SYCKRAHKNC and a linker comprising TSL6.

[0038] For a better understanding of the present invention, as well as other aspects and further features thereof, reference is made to the drawings, which together with the description form a part hereof, and in which: [Brief explanation of the drawings]

[0039] [Figure 1A]FIG. 1 shows a schematic diagram of a carrier with a peptide, a linker, and an encoding entity such as DNA or RNA.

[0040] [Figure 1B] FIG. 1 shows a conceptual overview of the bicyclization process starting from a phage library of different peptides.

[0041] [Figure 1C] FIG. 1 shows two schemes for reacting polypeptides with a carrier at the C-terminus (left) and N-terminus (right). [Figure 1D] FIG. 1C is a continuation of FIG.

[0042] [Figure 2] FIG. 1 shows a scheme of chemical bicyclization with linkers TSL1 and TSL2 and simultaneous encoding of the different linkers with two different silent gene barcodes, TCT and AGT.

[0043] [Figure 3] FIG. 1 shows a specific example of a linker using an N-terminal serine as reactive groups A and B and oxime formation on benzyl chloride.

[0044] [Figure 4A] Figure 1 shows linkers TSL1 (1), TSL3 (2), and TSL6 (3). Sequences 1-6 are variable peptides with which these linkers can react to form bicyclic structures.

[0045] [Figure 4B] FIG. 1 shows a synthetic scheme for linker TSL1.

[0046] [Figure 4C] FIG. 1 shows the synthesis schemes of linkers TSL3 and TSL6.

[0047] [Figure 5A]FIG. 1 shows modification of the oxidized peptide SHCVWWDC by TSL6. [Figure 5B] FIG. 5B is a diagram showing a continuation of FIG. 5A.

[0048] [Figure 6] FIG. 1 shows modification of reduced peptide SVCFDNGC with TSL6.

[0049] [Figure 7A] FIG. 1 shows bicyclic modifications with medium length (3 carbon atoms) linkers. [Figure 7B] FIG. 1 shows bicyclic modifications with short linkers (one carbon atom).

[0050] [Figure 8A] FIG. 1 shows SFCDWYGC modification with TSL-6 (six carbon atom linker). [Figure 8B] FIG. 1 shows SLCFDNGC modification by TSL-6. [Figure 8C] FIG. 1 shows SHDCYEC modification by TSL-6. [Figure 8D] FIG. 1 shows SWDYRECYLEC modification by TSL-6. [Figure 8E] FIG. 1 shows SWCDYRC modification by TSL-6. [Figure 8F] FIG. 1 shows SHCVWWDC modification by TSL-1. [Figure 8G] FIG. 1 shows SHCVWWDC modification by TSL-3.

[0051] [Figure 9A] FIG. 1 shows bicyclic SHCVWWDC-TSL6. [Figure 9B] FIG. 1 shows a graph demonstrating the stability of the complex in buffered media at pH 4, 7, and 8.5 for one month at room temperature.

[0052] [Figure 10A]Figure 9A shows the stability of the bicyclic ion shown in Figure 9A upon incubation in fetal bovine serum (FBS) at 37°C. The mixture was analyzed by ion-selective LCMS. The intensity of the bicyclic ion does not decrease even after 3 days (approximately 70 hours) of incubation in FBS at 37°C, whereas the 90% disulfide SHCVWWDC peptide degrades in less than 5 hours. [Figure 10B] Figure 1 shows bicyclic stability upon incubation in a Pronase™ protease mixture at 37°C. A linear disulfide and two bicyclic compounds made from TSL-1 and TSL-6 linkers were incubated with Pronase™ for the indicated times (up to 5 hours). Reactions were diluted and analyzed by ion-selective MS-UPLC. Numbers indicate the % of intact compound remaining after 5 hours of digestion.

[0053] [Figure 11] FIG. 1 shows the stability of several bicyclic peptide compositions under two different degradation conditions.

[0054] [Figure 12] FIG. 1 shows a schematic diagram of the modification of a phage library of peptides with TSL6.

[0055] [Figure 13] Figure 1 shows a schematic example of the modification of the large phage library SxCxxxxxxC with TSL1, TSL3, or TSL6 linkers, illustrating the generality of this approach to any library and any linker geometry. [Figure 14] Figure 1 shows a schematic example of the modification of the large phage library SxCxxxxxxC with TSL1, TSL3, or TSL6 linkers, illustrating the generality of this approach to any library and any linker geometry.

[0056] [Figure 15A]FIG. 1 shows an example of bicyclization using a molecule corresponding to two halves of a linker introduced at two distinct locations in a peptide: the first half by reaction with the peptide terminus and the second half introduced by reaction with two side chains of the peptide. [Figure 15B] (Continuation of 15A) An example of bicyclization using a molecule corresponding to two halves of a linker introduced at two separate locations in a peptide: the first half by reaction with the peptide terminus and the second half introduced by reaction with two side chains of the peptide. An intermediate complex, which can be purified, is induced to undergo cyclization upon reaction of the two halves of the linker.

[0057] [Figure 16A] FIG. 1 shows a bicyclic peptide library scaffold. [Figure 16B] FIG. 1 shows three panels A, B, and C illustrating the generation of a genetically encoded bicyclic library by modification of a library of phage-displayed peptides and selection of target-binding bicyclic peptides from this library.

[0058] [Figure 17] FIG. 1 shows four panels demonstrating the ability of bicyclic structures to inhibit known signaling events induced by Nodal. DETAILED DESCRIPTION OF THE INVENTION

[0059] In one aspect, the present invention includes a method for synthesizing a bicyclic peptide library, which includes two separate modification steps. The first step involves creating an intermediate complex by ligating a first end of a linker to a terminus of a peptide displayed on a support using ligation chemistry in an aqueous environment. The peptide terminus can be the N-terminus or the C-terminus. After an optional purification step, the second step involves exposing the intermediate complex to reaction conditions to induce an intramolecular bicyclization reaction, in which two reactive groups at the second end of the linker each independently react with the side chains of amino acids in the peptide sequence.

[0060] In some embodiments, the bicyclic carrier may comprise a phage, mRNA, DNA, ribosome, bacteria, yeast, a bead made of a synthetic polymer such as PEG or polystyrene, or any other genetically encoded biological display technology or synthetically encoded peptide library technology known in the art. In some embodiments, the carrier comprises a display set of gene sequences encoding a random peptide library of different chemical compositions. In another embodiment, the display set of gene sequences comprises a focused gene library encoding a focused subset of peptide sequences of different chemical compositions, such as those generated by random mutagenesis.

[0061] A display set of gene sequences can be paired with specific linkers such that a second set is linked with a linker of different size or composition, or with the chemical structure of the linker molecule, such as a linker that is a stereoisomer, diastereomer, or enantiomeric.

[0062] Either or both of the two steps of ligation and bicyclization can be chemical or enzymatic modification of peptide.In some embodiments, both modifications are chemical conjugation techniques specific to the N-terminus or specific N-terminus amino acid and a separate set of amino acids in peptide.For example, the chemical modification used for ligation can form oxime at oxidized N-terminus serine.Ligation can also use other N-terminus or C-terminus specific chemical reactions known in the art.

[0063] A schematic diagram of a carrier with a peptide and linker is shown in Figure 1A. The peptide end Z, which can be either the N-terminus or the C-terminus, reacts with the first end A of the linker L. The peptide has reactive side chains X1 and X2 that react with two functional groups B1 and B2 on the second end of the linker. The linker L can include an aliphatic chain, and can also include esters, phenyls, amines, etc.

[0064] In some embodiments, the first terminus A is first ligated with the peptide terminus Z, resulting in the linear intermediate complex identified in Figure 1A, followed by bicyclization. The ligation reaction may include modifying the oxidized serine (oxaloyl) through a carbon-carbon bond-forming process, such as the Wittig reaction. One preferred embodiment of the bicyclization step involves alkylation of cysteine, or any other suitable method for modifying peptides or proteins at specific locations.

[0065] In an alternative embodiment, shown schematically in Figures 1C and 1D, the first step involves reaction with a diketone, which forms the first ring and simultaneously introduces a unique reactive group. The diketone is preferably a 1,3-diketone. The intermediate monocyclic peptide formed in this reaction is uniquely stable and can be purified and stored without degrading the reactive group. The diketone-bearing intermediate monocyclic peptide can then be reacted with a linker molecule containing a 1,3-diketone-reactive group, such as an alkyl or aryl hydrazine, at one end and another group at the other end that can undergo chemical or enzyme-catalyzed ligation with the terminus of the peptide to induce the formation of a second ring ("bicyclization"). Ligation at the peptide terminus can occur first, resulting in a linear intermediate. Alternatively, reaction with the diketone can occur first, resulting in a branched intermediate.

[0066] For example, a display peptide with two internal cysteine residues in its sequence can be reacted with 1,5-dichloropentanedione-2,4 at alkaline pH, e.g., about pH 8 or 9, to introduce a 1,3-diketone group into the peptide. The 1,3-diketone group can then be reacted with a linker containing an alkyl or aryl hydrazine, introducing a functional group that completes a subsequent bicyclization reaction with the N-terminus of the peptide. For example, the peptide containing the 1,3-diketone group can be modified at the N-terminus to introduce a hydrazine linker, which can then be bicyclized via an intramolecular reaction between the N-terminal hydrazine and a 1,3-diketone ligated to the side chain. Alternatively, the reaction order can be reversed, with a linker containing an N-terminal reactive group reacting with a 1,3-diketone at acidic pH. Oxidation of the terminus at neutral pH and a change in environment to acidic pH 4.5 then triggers bicyclization via reaction with the N-terminus.

[0067] In all cases, it is preferable that the two steps, ligation and bicyclization, occur under conditions independent of each other. Such independence allows for purification of intermediate products and minimizes side reactions. Some embodiments involve the use of reactions requiring different pH values. For example, oxime formation at an oxidized N-terminal serine occurs at an acidic pH of approximately pH 3 and can be catalyzed by 0.1% trifluoroacetic acid (TFA). These conditions are tolerated by phage-displayed libraries and other genetically encoded peptide libraries, such as RNA-displayed libraries. The subsequent bicyclization reaction can be any intermolecular reaction with an amino acid side chain that occurs at a higher pH. Suitable reactions may include nucleophilic substitution between a thiol and a thiol-reactive group, such as halobenzyl, haloacetamide, nucleophilic aromatic substitution, or Michael addition of a thiol to conjugated alkenes and / or allenamides. Other known reactions that occur at alkaline pH with Lys, Tyr, or other amino acid side chain residues are also possible.

[0068] In an alternative embodiment, the ligation and bicyclization steps are separated by the use of a protection-deprotection reaction. For example, the formation of a C-C bond at an oxidized N-terminal serine via the Wittig reaction occurs at about pH 7 to about pH 8. The Wittig reaction is known not to disrupt the S-S disulfide that protects the thiol residue. The Wittig reaction is tolerated by other genetically encoded peptide libraries, such as phage-displayed libraries and RNA-displayed libraries. After ligation and purification, the bicyclization reaction is triggered by disulfide reduction and ligation between the thiol and a thiol-reactive group, which can occur at about pH 7 to about pH 8. Many linkers can be designed that combine the stabilized ylide used in the Wittig ligation reaction with a thiol-reactive group for nucleophilic substitution, nucleophilic aromatic substitution, or Michael addition, or other well-known reactions in the bicyclization step.

[0069] In some embodiments, the carrier comprises an identifier, preferably a variable nucleic acid-encoded identifier. The identifier can be silent so that it does not encode any peptides carried on the carrier exterior. Alternatively, the identifier can be such that all variants of the identifier encode the same or substantially similar peptides. This latter case, referred to as "silent barcode" technology, involves producing a bacteriophage display system on particles that contain DNA of different compositions within the bacteriophage particle and display peptides of the same composition. In some embodiments, the carrier is a virus or bacteriophage virion of the same external chemical composition, packaged within these particles and containing a variable nucleic acid code comprising a degenerate DNA tag within the genome. The genome of the virus or phage is manipulated in a way that does not result in changes to the chemical composition of the virion coat, such as the use of degenerate codons in the virion coat coding region, changes to the DNA sequence encoding excised sequences, changes to DNA sequences that do not encode expressed protein sequences, or changes to DNA sequences encoding components not incorporated into the virion coat. Thus, a carrier library can be provided that includes a plurality of carriers (such as phages or viruses), all of which are externally chemically identical (e.g., before modification for attachment of any ligands), but contain distinct nucleic acid identifiers therein.

[0070] In another aspect, the present invention can include a method for selecting genetically encoded modifications of a peptide library by using a unique identifier, preferably a silent genetic barcode, associated with a specific bicyclic structure. In some embodiments, a library of multiple carriers is generated, each with a unique silent genetic barcode and each displaying a polypeptide. Each library is then modified with a different linker, as described herein, to produce a unique bicyclic structure for the linker used. The libraries are then combined to produce a mixed library, and the unique bicyclic structure of each linker can be identified by a barcode. The mixed library can then be screened to select peptides with the desired sequence and bicyclic topology, which can then be identified by sequencing the genetic barcode (or otherwise identifying the identifier).

[0071] For example, as shown schematically in Figure 2, a first carrier carries a first silent nucleic acid code (barcode) and a DNA sequence encoding a peptide. A second carrier carries a second silent nucleic acid code different from the first code and a DNA sequence encoding a peptide that may be the same as or different from the first peptide. The first carrier peptide is modified with a first linker (TSL1), and the second carrier is modified with a different linker (TSL2). After ligation and bicyclization, the resulting two different bicyclic structures are distinguishable by the first and second nucleic acid codes.

[0072] In another aspect, the present invention can include a method for identifying drug candidates comprising preparing a genetically encoded bicyclic peptide mixed library as described herein and screening the library with a putative receptor molecule to identify those bicyclic peptides that bind to the receptor molecule. The receptor-binding bicyclic peptides are then identified by enriching and sequencing the silent nucleic acid code and the peptide-encoding sequence.

[0073] In another aspect, the invention may include a method of synthesizing a genetically encoded chemical bicyclic peptide library, comprising inserting a redundant set of gene sequences encoding peptide linkers into multiple independent vectors in a substrate, such that the gene sequences produce identical or closely related peptide sequences ("linkers") upon translation; inserting a second set of gene sequences encoding genetically diverse inserts into each vector, such that a diverse set of peptides ("library") is expressed upon translation; expressing and amplifying the first and second gene sequences, such that the translation products include a non-variable linker, and a variable peptide library is synthesized; and modifying each peptide library with a separate TSL and combining the multiple modified libraries to produce a library in which the chemical modifications are genetically encoded.

[0074] U.S. Patent Publication No. 2013 / 050083 to Derda et al., the entire contents of which are incorporated herein by reference where permitted, describes quantification of chemical modification of genetically encoded peptide libraries and selection of new strategies for effective modification. These methods can be used to quantify the yield of any one or all of the reactive steps described herein based on the presence or absence of reactive groups consumed in the reaction.

[0075] Techniques for genetically encoding chemical post-translational modifications of phage-displayed libraries are described in PCT Patent Application No. WO2016061695A1, the entire contents of which are incorporated herein by reference where permitted.

[0076] example In order to better understand the invention described herein, the following examples are set forth. It should be understood that these examples are for illustrative purposes only. As such, they should not be construed as limiting the scope of the invention in any way.

[0077] Example 1.Synthesis of linkers for bicyclization. As shown in Figure 4A, crosslinkers 1 (TSL1), 2 (TSL3), and 3 (TSL6) were designed based on oxime formation and cysteine S-alkylation chemistry. These reactions involve fast, clean, and high-yielding bioorthogonal reactions. The synthesis steps are outlined in Figures 4B and 4C. The difference between the three is primarily the number of carbon atoms in the linking chain.

[0078] Example 2. Modification of the reduced peptide SHCVWWDC with a TSL6 linker. Modification details are outlined in Figures 5A and 5B. The ligation step occurs at low pH. A size-exclusion column allows for purification after the aldehyde step or ligation. The low pH during the reduction step maximizes the yield of the bicyclization reaction. Increasing the pH after reduction afforded the bicyclic product cleanly, with no visible unreacted starting material or by-products by LCMS. Characterization by LCMS confirmed the identity of all intermediates at each step.

[0079] Example 3. Modification of the reduced peptide SVCFDNGC with a TSL6 linker. Modification details are outlined in Figure 5. Key details of the chemistry: low pH during the ligation step. A size exclusion column allows for purification after the aldehyde step or ligation. Low pH during the reduction step maximizes the yield of the bicyclization reaction. Increasing the pH after reduction afforded the bicyclic product cleanly, with no visible unreacted starting material or by-products by LCMS. Characterization by LCMS confirmed the identity of all intermediates at each step.

[0080] Example 4.Modification of a diverse set of peptides with several TSL linkers. A series of peptide sequences containing a serine residue at the N-terminus and two cysteines downstream was selected. Oxime ligation-based chemistry was used to recognize the oxoaldehyde generated from periodate-mediated oxidation of the N-terminal serine, followed by a crosslinking strategy to capture the two cysteines and finally form a conformationally rigid peptide macrobicycle using the conditions described in Examples 2 or 3. A summary of the reactions with several different peptides, as shown in Figures 8A-8G, confirms that the ligation and bicyclization chemistry can be effectively performed with a variety of peptide sequences. Modification with TSL linkers with a decreasing number of atoms proceeded effectively even when the size of the resulting second ring was formed with only one amino acid (Figure 7B). These results demonstrate that there are no constraints on the size of the peptide loop and little constraint on the geometry of the TSL.

[0081] Example 5. Stability of the bicycle in buffers and biological media. Bicyclic peptides formed from the sequence SHCVWWDC by modification with the TSL3 linker were incubated at room temperature for one month in buffered media at pH 4, 7, and 8.5 (Figure 9). We observed no change in LCMS, indicating no degradation of the bicyclic product under any of these conditions. The SHCVWWDC-TSL3 bicycle was incubated in fetal bovine serum (FBS) at 37°C, and the integrity of the bicycle was tested by LCMS. Using ion-selective LCMS, the bicycle remained unchanged in FBS, even after 3 days of incubation in FBS at 37°C. In contrast, ion-selective LCMS demonstrated that 90% of the cyclic peptide disulfide SHCVWWDC was degraded in FBS after 300 minutes (Figure 10A).

[0082] In another rigorous stability study, we used an aggressive cocktail of endo- and exoproteases (Pronase) to digest the linear sequence SWDYRECYLEC, its disulfide derivative, and two bicyclic derivatives of this sequence modified with TSL1 and TSL6 linkers (Figure 10B). After 5 h of incubation at 37 °C, only 0.4 ± 0.1% of the linear peptide and 0.9 ± 0.4% of the linear and disulfide peptides remained undigested. Under the same conditions, after 5 h of proteolytic digestion, 68 ± 14% of the TSL6-SWDYRECYLEC bicycle and 82 ± 13% of the TSL1-SWDYRECYLEC bicycle remained intact (Figure 10B). The latter observation estimates a half-life stability of approximately 24 h for the TSL1-SWDYRECYLEC bicycle.

[0083] Figure 11 demonstrates the generality of this observation across 12 other bicyclic peptide compositions under two different degradation conditions. Figure 11a describes the sequences and abbreviations of the reagents used for modification (TSL-1, TSL-3, and TSL-6 correspond to structures 1, 3, and 6 in Figure 4A). Alphanumeric designations are used to describe the products. Figure 11b shows examples of time-resolved stability measurements at 36°C for 5 hours and endpoint stability measurements after 5 hours. Figure 11c summarizes endpoint measurements (36°C for 5 hours) under two different proteolytic conditions: Pronase and fresh mouse serum. Figures 11c and 11d compare the stability of the bicyclic peptides with some cyclic and bicyclic structures known in the art. This comparison demonstrates the surprising and unobvious advantages of 14f, the compound produced by the TSL-6 modifier reported herein, and 16j, the known state-of-the-art compound produced from a similar peptide sequence with TBMB modifier (J. Med. Chem. 2018, 61(7), 2823-2836). Bicyclic compound 14f is 10-fold more stable to Pronase treatment than 16j. Another comparison between 13i and 13e again demonstrates the advantages of TSL-6 over art-known cyclizations via perfluoroarenes, as described in WO2014052650A2.

[0084] Example 6. Modification of a library of phage-displayed peptides and validation of the modification with capture agents to generate a genetically encoded bicyclic library.

[0085] A library of phage-displayed peptides of the structure SxCxxxC, where S is serine, C is cysteine, and x is a random amino acid, was used as the starting point for the production of a genetically encoded bicyclic library. Figure 12 illustrates an example of the modification of clones from such a phage library using the TSL6 linker. Figures 13 and 14 illustrate the modification of the large phage library SxCxxxxxxC using the TSL1, TSL3, or TSL6 linkers, demonstrating the generality of this approach to any library and any linker geometry. We used conditions optimized for the ligation of synthetic peptide sequences: specifically, the library was exposed to an ice-cold solution of 60 micromolar sodium periodate in PBS for 8 minutes, and the reaction was quenched by the addition of a 0.5 mM solution of methionine. The oxidized library was exposed to a solution of 1 mM TSL6 linker in 0.1% aqueous trifluoroacetic acid at room temperature for 1 hour. The ligated library can be purified by size-exclusion chromatography using Zeba™ Spin Desalting Columns, 7K MWCO, using pH 4 buffer as the eluent. The purified library was exposed to TCEP at pH 4 to reduce disulfide bonds. Finally, Tris-buffered medium was added to the solution to raise the pH to 8 and promote bicyclization (Figure 13A). Each step of the reaction can be effectively monitored using various capture reagents (Figure 13B). For example, to quantify the oxidation step, we mixed the oxidized library with aminooxybiotin (AOB) in aniline acetate buffer and measured the phage titer before and after exposure to streptavidin-coated beads (Figure 13C, F). AOB capture demonstrated that 73% of the library was oxidized. After exposure to TSL6, only 10% of the phage population contained AOB-reactive aldehydes, indicating that 87% of the oxidized library was ligated with TSL6. The same procedure can be used to test the integrity of the reactive groups.For example, modification with AOB followed by capture demonstrated that aldehydes remained reactive when the library was incubated in 0.1% TFA without the TSL linker (Figure 13F). Similarly, exposure to BIA and capture with streptavidin ("BIA capture") can be used to quantify the number of thiols, and exposure to BSH followed by capture with streptavidin ("BSH capture") can be used to quantify the percentage of the library containing thiol-reactive benzyl chloride groups. Exposure to BIA (Figure 13G) after ligation and bicyclization (Figure 13H) demonstrates that thiols are present in the library after ligation but disappear after bicyclization. Similarly, exposure to BSH can be used to check the presence or absence of benzyl chloride groups. Exposure to BSH can also be used to check the integrity of the benzyl chloride groups. For example, prolonged exposure of the ligated library to pH 7 or purification in a pH 7 buffer leads to hydrolysis of the benzyl chloride groups, as determined by BSH capture. This observation was used to select specific conditions for purifying the library ligated to TSL6. Without such quantification, it is unclear how the purification and integrity of the library modified with TSL7 can be checked. Optimization of other peptide and linker sizes follows the same steps (Figure 11, Figure 12) and shows the same observations.

[0086] Example 7Synthetic or phage-displayed peptides containing two Cys residues in the sequence can be reacted with 1,5-dichloropentanedione-2,4 at pH 8.5 to quantitatively introduce a 1,3-diketone group into the peptide. For synthetic peptides, LCMS confirms the completion of the reaction across multiple diverse peptide sequences. For phage-displayed libraries, the presence of the 1,3-diketone in the displayed peptide can be confirmed by reaction of the 1,3-diketone-modified product with hydrazine-biotin followed by capture with streptavidin beads. The 1,3-diketone group in the peptide or phage-displayed peptide can then be modified with alkyl or aryl hydrazines under controlled conditions, such as ammonium acetate buffer at pH 4.5 (Figure 15A). This reactivity can be used to introduce functional groups such as aldehyde-reactive ylides, which subsequently react with the N-terminus and complete bicyclization (Figure 15B). For example, the peptide containing a 1,3-diketone group can be modified at the N-terminus to introduce an aldehyde, and a change in the environment to acidic pH induces bicyclization via an intramolecular reaction between the N-terminal hydrazine and the 1,3-diketone ligated to the side chain (right pathway in Figure 15B). In another example, the order of reactions can be reversed, and a linker containing an N-terminal reactive group is reacted with a 1,3-diketone at acidic pH. Oxidation of the terminus at neutral pH triggers bicyclization via reaction with the N-terminus (left pathway in Figure 15B).

[0087] Example 8Genetically encoded bicyclic peptide libraries were generated by modifying a phage-displayed peptide library and selecting target-binding bicyclic peptides from this library. The screening step was performed using the complete sequence of the representative human protein Nodal (GenBank: BC104976.1; Proc. Natl. Acad. Sci. USA, 2002, 99(26), 16899-16903). Specifically, the 40-338 amino acid sequence PLAYMLSLY[..]VLLDHHKD encoded by BC104976, hexahistidine-tagged Nodal (His-Nodal) catalog number ag21882 from Proteintech, was used. Selection can easily be performed against any other target using known techniques for selecting phage-displayed libraries. The results of screening, panning, and validation of the TSL-6-modified SXCX6C are shown in Figure 13.

[0088] Targets immobilized on nitrilotriacetic acid (NTA)-functionalized agarose beads were panned using a mixed bicyclic library modified with a TSL-6 (6-carbon) linker (Figure 16B-A). After washing the beads with a KingFisher Duo, the beads were boiled in water to release Nodal, and the bound ligands / phage were subjected to PCR and amplification. After three rounds of selection, the desired convergence was observed: the bicyclic library in round 3 (R3) specifically enriched with Nodal immobilized on agarose beads when compared to the R1 and R2 libraries. No enrichment was observed with blank beads or beads bearing a control His-tagged target. Importantly, the unmodified R3 library panned on Nodal-modified beads showed no enrichment, confirming that selected peptide sequences bind to Nodal only when constrained to the bicyclic scaffold (Figure 16A). dsDNA amplicons from each round of selection were sequenced using Illumina NextSeq, and informatics analysis suggested numerous sequences (Figure 16B) and at least three families of sequence motifs (Figure 16C) as potential ligands for Nodal. A list of peptide sequences that bind to Nodal is shown in Table 1. [Table 1]

[0089] To verify the binding ability of the predicted ligand, we tested the ability of the bicyclic compound to inhibit known signaling events induced by Nodal. Specifically, we used Western blot analysis with an anti-phospho-Smad antibody to detect Nodal-induced phosphorylation of the effector protein Smad2 in embryonic carcinoma P19 cells. Figure 17a shows that treatment of P19 cells with 100 ng / mL Nodal for 1 hour in Dulbecco's modified Eagle's medium (DMEM) supplemented with 2.5% fetal bovine serum (FBS) and 7.5% bovine serum albumin (BSA) increased pSmad2 activity. Incubation of P19 cells with 100 ng / mL Nodal and the known kinase inhibitor SB-431542 inhibited the Nodal-induced increase in pSmad2 activity. Similarly, cotreatment of 100 ng / mL Nodal with 100 μM bicyclic peptides 2 (SPCKAGTGQC), 3 (SPCKGPSATC), 4 (SPCKGRHHNC), 5 (SPCKKAHGAC), 7 (SPCQRGHMFC), and 11 (SYCKRAHKNC) did not increase pSmad2 phosphorylation above background levels. The bicyclic peptides function as Nodal antagonists at 100 micromolar concentrations. Figure 17b illustrates that only two of the six bicyclic peptides retain potency at 10 μM: bicyclics 7 (SPCQRGHMFC) and 11 (SYCKRAHKNC) inhibit pSmad2 phosphorylation, whereas bicyclics 2, 3, 4, and 5 do not inhibit pSmad2 phosphorylation at 10 μM. Figure 17c-d illustrates the dose response of bicyclics 7 and 11, suggesting half-inhibitory concentrations of 1-3 μM for antagonism of Nodal signaling.

[0090] SPCQRGHMFC-TSL6 and SYCKRAHKNC-TSL6 and derivative compounds have the ability to antagonize the signaling function of the Nodal protein in cancer cells. Derivative compounds of SPCQRGHMFC-TSL6 and SYCKRAHKNC-TSL6 retain similar structural features of these compounds and exhibit similar or enhanced ability to antagonize Nodal protein function. The only known Nodal antagonist peptide is the anti-human Nodal monoclonal antibody 3D1 (WO2016057683A2). Small molecule compounds that can antagonize Nodal include SB431542 (listed as SB in Figure 17) and its derivatives. However, these compounds do not interact with Nodal and are inhibitors of ALK5, ALK4, and ALK7 kinases, which act downstream of Nodal. References

[0091] All publications, patents, and patent applications mentioned in this specification are indicative of the level of skill of those skilled in the art to which this invention pertains and, where permitted, are incorporated by reference herein to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. JPEG0007723599000002.jpg214154 JPEG0007723599000003.jpg140154

Claims

1. (a) a carrier; (b) a polypeptide linked to the carrier; and (c) a linker having a first end and a second end comprising two reactive groups, the linker having two-fold symmetry; (d) optionally, an identification tag that identifies the linker; 1. A library of a plurality of bicyclic peptide constructs comprising: the carrier comprises a nucleic acid encoding the polypeptide; the polypeptide consists of unprotected natural amino acids; A library of multiple bicyclic peptide constructs, wherein the first end of the linker is covalently ligated to the N-terminus of the polypeptide, and both reactive groups at the second end of the linker are covalently bonded to a side chain residue of lysine or tyrosine of the polypeptide, or to a side chain residue of cysteine of the polypeptide through a thioester bond, respectively, to form a bicyclic structure by binding the linker to the polypeptide, and the bicyclic structure does not have an N-terminus.

2. The library described in claim 1, wherein in each of the plurality of bicyclic peptide constructs, the first end of the linker comprises an aldehyde-reactive group such as an oxime, hydrazine, 2-aminobenzamidoxime, phosphonium ylide, sulfur ylide, nitrogen ylide, or any other carbon nucleophile and carbenoid reagent known to be reactive with aldehydes and stable in an aqueous environment.

3. A library as described in claim 1 or 2, wherein in each of the plurality of bicyclic peptide constructs, the polypeptide comprises two cysteine residues and an N-terminal serine or threonine residue, the N-terminal serine or threonine is first converted to an aldehyde by selective oxidation, and the reactive group at the second terminal is covalently bonded to the cysteine residue.

4. In each of the plurality of bicyclic peptide constructs, the carrier comprises: (a) a phage particle; (b) an RNA display compound comprising an RNA encoding the polypeptide and linked to the polypeptide; (c) a DNA display compound comprising DNA encoding the polypeptide and linked to the polypeptide; or (d) Polymer or Protein Carrier The library according to any one of claims 1 to 3, comprising:

5. A library described in any one of claims 1 to 4, wherein in each of the plurality of bicyclic peptide constructs, the polypeptide comprises an N-terminal serine residue and two cysteine residues separated by at least 2, 3, 4, or 5 amino acid residues.

6. A library described in any one of claims 1 to 5, wherein in each of the multiple bicyclic peptide constructs, the linker comprises an alkyl chain having 1 to 6 carbon atoms.

7. The method of claim 1, wherein the linker comprises: 、 , or The library of claim 6,

8. 1. A method for producing a phage display complex, comprising: (i) providing a phage particle comprising an identification tag that identifies a polypeptide consisting of natural amino acids having an N-terminus, a nucleic acid encoding said polypeptide, and a linker; (ii) providing a linker having a first end and a second end comprising two reactive groups, and ligating the first end of the linker to the N-terminus of the polypeptide to form a covalent bond to form an intermediate complex; (iii) forming a bicyclic structure from the intermediate conjugate by reacting both reactive groups at the second end of the linker with a thiol, amine, or phenol side chain residue of the polypeptide; A method wherein steps (ii) and (iii) are independent and sequential.

9. 9. The method of claim 8, wherein steps (ii) and (iii) are carried out at different pHs.

10. A library described in any one of claims 1 to 7, comprising different bicyclic peptide constructs produced from different peptide sequences and / or different linkers, the different linkers being associated with different identification tags comprising silent genetic barcodes, and each bicyclic structure being identified by identifying the peptide sequence by sequencing a nucleic acid encoding the peptide sequence and identifying the linker by sequencing the silent genetic barcode.

11. 11. A method for identifying a construct according to any one of claims 1 to 7 that is capable of binding to a ligand, the method comprising: (i) contacting the library according to claim 10 with the ligand; (ii) selecting the construct from the library that binds to the ligand; and (iii) identifying the structure of the selected construct by sequencing a nucleic acid that encodes the peptide sequence.

12. 12. The method of claim 11, further comprising identifying the linker by sequencing the silent genetic barcode.

Images