Libraries of macrocyclic peptides displayed on the surface of yeast cells

Abstract

The present invention concerns the generation of libraries of cysteine-rich macrocyclic peptide sequences expressed on the surface of yeast cells, the method of preparation thereof, and a population of yeast cells genetically modified to express a plurality of disulfide-tethered macrocyclic peptide ligands of different structure and amino acid sequence on the yeast cell surface.

Description

[0001] LIBRARIES OF MACROCYCLIC PEPTIDES DISPLAYED ON THE SURFACE OF YEAST CELLS FIELD OF THE INVENTION The present invention relates to the generation of libraries of cysteine-rich cyclic peptide sequences that are expressed on the surface of yeast cells. More particularly, the invention makes available a yeast cell-based macrocyclic peptide display system, a method of preparation thereof, and a population of yeast cells genetically modified to express on the cell surface a plurality of disulfide-tethered macrocyclic peptide ligands of different structure and amino acid sequence. INVENTION BACKGROUND Macrocyclic peptides are increasingly proving to be a valuable molecular format for drug development (Ji, Nielsen, and Heinis 2024). At present, there are around eighty peptide therapeutics on the global market, of which more than forty are macrocyclic peptides, with a growing number of macrocyclic peptide drugs approved per year (Muttenthaler et al. 2021; Fetse et al.2023). Macrocyclic peptides have a number of favourable properties that make them an attractive modality for the development of therapeutic agents (Zhang and Chen 2022). They can bind to macromolecular targets with high affinity and selectivity. Moreover, they often display good proteolytic stability, and in some cases even membrane permeability. Furthermore, they usually show a low inherent toxicity or antigenicity. Additionally, macrocyclic peptides can be efficiently produced by chemical synthesis and possess ease of modification. Besides, their modular structure and facile access to hundreds of different commercially available amino acid building blocks simplify the development of macrocyclic peptide variants with tailored properties. All these qualities well position macrocyclic peptides to bridge the gap between small molecule drugs and larger biologics such as antibodies (Zhang and Chen 2022). While numerous macrocyclic peptides continue to originate from the investigation and exploitation of naturally occurring peptides, recent technological advances and major breakthroughs in molecular biology paved the way for the development of de novo generated macrocyclic peptide ligands with desired qualities (Li, Craven, and Levine 2022). Indeed, the advent of powerful combinatorial technologies such as phage display (G. P. Smith and Petrenko 1997; Deyle, Kong, and Heinis 2017) and mRNA display have exponentially accelerated the generation of macrocyclic peptide ligands to diverse protein targets for which no natural peptide ligands have been discovered. All these in vitro directed evolution approaches rely on a physical linkage between the ‘phenotype’ (expressed macrocyclic peptide) and ‘genotype’ (encoding DNA or RNA sequence). By applying such technologies, macrocyclic peptide ligands with desired properties are usually evolved following a similar scheme, comprising the generation of large combinatorial libraries of random genetically encoded macrocyclic peptides, multiple iterative cycles of selection, amplification and diversification. Recently, the development of innovative strategies for post-translational chemical and enzymatic modifications have greatly expanded the number of macrocyclic peptide formats that, in addition to the canonical disulfide- tethered peptides, can be examined by the aforementioned combinatorial technologies (Passioura et al.2014; Sharma et al.2023; Jaroszewicz et al.2022; Dotter et al.2021). These, coupled with the latest technological advances in screening procedures and automatisation, the emergence of new reagents and tools, and the affirmation of next-generation sequencing techniques, have enabled the construction of larger libraries and the isolation of macrocyclic peptide ligands with exquisite binding properties towards increasingly challenging targets (T. P. Smith et al. 2023; Stellwagen et al. 2019; Sloth et al. 2023). Examples of successful macrocyclic peptides discovered using in vitro display technologies are Merck & Co.'s oral PCSK9 inhibitor MK-0616 for atherosclerotic cardiovascular disease (Kingwell 2023), and UCB's recently FDA-approved zilucoplan for generalized myasthenia gravis in adults who are anti-acetylcholine receptor antibody-positive (Mullard 2023). Though these technologies have been proven capable of generating and screening extremely large sizes of sequences and structurally diverse repertoires, thus enabling the rapid identification of macrocyclic peptide ligands against virtually any target, they still rely on rather difficult to control procedures as the performance of individual selected clones or populations cannot be easily monitored during the high-throughput screening. The success of a selection campaign is typically only seen after several weeks of work, once the isolated macrocyclic peptide molecules are identified. Moreover, the biophysical characterization of selected ligands often requires chemical synthesis or sub-cloning and recombinant production followed by purification, all additional phases which ultimately slow down and make the downstream process more complex and often costly (Jaroszewicz et al.2022; Vodnik et al.2011; Wang and Liu 2011). Finally, the number of selected macrocyclic peptide ligands that can be synthesised, purified and characterised is typically limited to 10 – 100 molecules. Herein, we present a novel approach based on yeast surface display technology that addresses the abovementioned concerns. Since its invention, yeast surface display has proven to be a transformative tool for the direct evolution of multiple immunoglobulin- and non- immunoglobulin-based proteins (Boder and Wittrup 1997; Angelini et al.2015). This technique was initially validated to enhance the binding affinity of an existing antibody fragment, but later proved highly effective also for isolating de novo proteins with fine-tuned binding affinities and specificities toward a wide range of targets from naïve combinatorial libraries (Cherf and Cochran 2015; Lopez-Morales, Vanella, Appelt, et al. 2023). An advantage of yeast surface display technology coupled to fluorescence-activated cell sorting (FACS) is that it offers real- time monitoring of the iterative screening of large combinatorial libraries of diverse macrocyclic peptide ligands. By applying flow cytometry, information on the affinity, specificity, stability, and enrichment of individual yeast clones encoding diverse macrocyclic peptide ligands can be monitored in a continuous and quantitative manner through successive rounds of sorting (Angelini et al. 2015; Teymennet-Ramírez, Martínez-Morales, and Trejo- Hernández 2022). Moreover, by applying different labelling approaches, including equilibrium, kinetic and competition binding approaches, yeast display combined with flow cytometry can allow accurate adjustment of the selection stringency, enables rapid and fine epitope mapping, and favours quantitative discrimination between single macrocyclic peptide variants directly as cell surface fusions without the need for their chemical synthesis or recombinant expression and purification (Cherf and Cochran 2015; Lopez-Morales, Vanella, Appelt, et al.2023; Lopez- Morales, Vanella, Kovacevic, et al.2023). Yeast display was previously applied for the screening of peptide libraires but mostly for the engineering of naturally occurring linear and cystine knot peptides (Linciano et al.2019) or the affinity maturation or binding characterisation of existing molecules (van Rosmalen et al.2017; Ishii et al.2012). Recently, first studies reported the exploitation of yeast display for the de novo generation of cyclic peptide ligands, such as the generation of binders to lysozyme and human interleukin-17 with dissociation constant (KD) values ranging from 300 nM to 3 μM (Bacon et al. 2020). Similarly, screening of library of post-translationally enzymatically modified octapeptides yielded cyclic peptide ligands capable of binding two distinct domains of the yes-associated protein (YAP) with apparent KD values ranging from 700 nM to 1.5 μM (Bowen et al.2021). However, the sizes of the libraries were still limited as well as the structural diversity of the peptides that were restricted to monocyclic peptides of 7 and 8 amino acids. Hence, in this work we attempted to thoroughly explore the full potential of yeast surface display for the isolation and characterisation of macrocyclic peptide ligands with fine binding properties. Toward this goal we generated large and structurally diverse naïve combinatorial libraries encoding macrocycle peptide ligands with different topologies. To provide a more robust procedure we turned to disulfide-cyclised peptide ligands. The spontaneous intra-molecular oxidation of cysteine residues, present in the genetically encoded peptide sequences, is significantly easier and more reliable than post-translational cyclisation. In this regard, we foresee the eukaryotic expression machinery of yeast cells to support precise disulfide isomerization of cysteine-rich sequences ultimately enabling the generation of unique disulfide-tethered patterns that may be otherwise refractory to other systems (Angelini et al. 2015; Teymennet-Ramírez, Martínez-Morales, and Trejo-Hernández 2022; Linciano et al. 2019). STATE OF THE ART Linciano S. et al. (‘Molecular evolution of peptides by yeast surface display technology’, Medicinal Chemistry Communications, 2019, 10, 1569-1580) describes recent developments and use of yeast display technology for the engineering of peptide molecules, highlighting its differences from other existing in vitro display systems. Angelini A. et al. (‘Protein Engineering and Selection Using Yeast Surface Display’, Methods in Molecular Biology, 2015, 1319, 3-36) describes a protocol for the modification and isolation of protein variants from combinatorial libraries and their subsequent characterisation by yeast display technology. Van Deventer J. A. et al. (‘A platform for constructing, evaluating, and screening bioconjugates on the yeast surface’, Protein Engineering Design & Selection, 2016, 29, 485- 493) describes the site-specific incorporation of a non-canonical amino acid on proteins expressed on the surface of yeast cells and their post-translation chemical modification. Stieglitz J T et al. (‘A robust and quantitative reporter system to evaluate noncanonical amino acid incorporation in yeast’, ACS Synthetic Biology, 2018, 7, 2256-2269) describes a display-based system in yeast involving the incorporation of non-canonical amino acids into proteins via an orthogonal translation system (OTS). Van Rosmalen, M. et al. (‘Affinity Maturation of a Cyclic Peptide Handle for Therapeutic Antibodies Using Deep Mutational Scanning’, Journal of Biological Chemistry, 2017, 292, 1477–1489) describes the combination of yeast display and deep mutational scanning techniques to improve the affinity of a known cyclic peptide, named meditope, towards the therapeutic antibody cetuximab. Ishii, J. et al. (‘Cell Wall Trapping of Autocrine Peptides for Human G-Protein-Coupled Receptors on the Yeast Cell Surface’, PLoS One, 2012, 7, e37136) describes a yeast-based cell wall trapping of autocrine peptides (CWTrAP) system to select agonistic peptides for human GPCRs. Bacon, K. et al. (‘Isolation of Chemically Cyclized Peptide Binders Using Yeast Surface Display’, ACS Combinatorial Science, 2020, 22, 519–532) describes a method to generate cyclic peptides displayed on the surface of yeast cells by the cross-linking two primary amine groups with disuccinimidyl glutarate. Bowen, J. et al. (‘Screening of Yeast Display Libraries of Enzymatically Treated Peptides to Discover Macrocyclic Peptide Ligands’, International Journal of Molecular Sciences, 2021, 22, 1634) describes a method to generate cyclic peptides displayed on the surface of yeast cells by using transglutaminase enzyme. US 9134309B2 describes a method for the generation and expression of peptide libraries displayed on the surface of bacterial cells. EP 2535346A1 describes a method for the generation and screening of peptide libraries displayed on the surface of bacterial cells for the identification of an enzyme ligand. US 5922545A describes a general method for generating libraries of polysomes displaying nascent peptides and antibodies derivates suitable for affinity interaction screening, based on E. coli S30 translation system for high efficiency translation of mRNA. WO2007109823A2 describes a method for the generation of a vector library comprising a multiplicity of different eukaryotic secretion vectors, wherein each vector encodes for an extracellular soluble fusion polypeptide. WO2020236146A1 describes a method for generation and screening of cyclic peptide libraries displayed on the surface of phage particles. Cyclization of the peptide occurs via a cysteine and a non-canonical amino acid. DESCRIPTION OF THE FIGURES Figure 1. Schematic representation of the yeast surface display disulfide-tethered macrocyclic peptide system developed in this work. The cysteine-rich macrocyclic peptide (MP) sequence of interest (black) is expressed as N-terminal (left) or C-terminal (right) fusion of a cell-surface protein (CSP, white). A long and flexible linker (FL) is placed between the MP and the immunofluorescence tag (IFT, light grey) to secure distance and minimise potential steric hindrance with detection reagents. Bridging of one pair of cysteines by disulfide bridge yields macrocycle peptide ligands with ‘one-ring’ and a unique topology. Bridging of two or more pairs of cysteines by disulfide bridges yields macrocycle peptide ligands with ‘two-rings’ or ‘multiple rings’, and different topologies that can be sampled for binding to protein targets of interest. Figure 2. Generation of yeast-encoded cysteine-rich macrocyclic peptide libraries with large backbone diversity. a) Generation of cysteine-rich macrocyclic peptide libraries with large sequence diversity encoding one- or two-ring disulfide-tethered macrocyclic peptide topologies fused to the N-terminus of a cell-surface anchor protein expressed on the surface of yeast cells. The insert is created by appending a randomised DNA sequence, encoding a macrocyclic peptide (MP), to the 5’ end of DNA encoding the flexible linker using a forward degenerate primer and a constant reverse primer that anneals on the 3’ portion of the gene encoding the cell-surface anchor protein (white). The MPs may have random amino acid sequences and lengths. The final construct from 5’ to 3’ ends includes the following genetic elements: an inducible promoter (P), a segment encoding a secretion signal (SS, grey), a segment encoding a randomised cysteines-rich macrocyclic peptide sequence (MP, black) a segment encoding a long and flexible spacer (FL, light grey), a segment encoding an immunofluorescence detection tag (IFT, dark grey) and a gene encoding a cell-surface anchor protein (CSP, white). b) Generation of cysteine-rich macrocyclic peptide libraries with large sequence diversity encoding ‘one-ring’ or ‘two-rings’ topologies fused to the C-terminus of a cell-surface anchor protein expressed on the surface of yeast cells. The insert is created by appending a randomised DNA sequence encoding MP sequences (black) with varying amino acid sequences and lengths to the 3’ end of a DNA sequence encoding the flexible linker using a reverse degenerate primer and a constant forward primer that anneals on a 5’ portion of the gene encoding the cell-surface anchor protein (white). The final DNA construct from 5’ to 3’ ends includes the following elements: an inducible promoter (P), a segment encoding a secretion signal (SS, grey), a gene encoding a cell-surface anchor protein (CSP, white), a segment encoding an immunofluorescence detection tag (IFT, dark grey), a segment encoding a long and flexible spacer (FL, light grey) and a segment encoding a randomised cysteines-rich macrocyclic peptide sequence (MP, black). Figure 3. Design and analysis of an exemplificative ‘one-ring’ cysteine-rich macrocyclic peptide library (X)nC(X)mC(X)t displayed on the yeast cell surface. a) The exemplificative ‘one-ring’ macrocyclic peptide library named CX7C has been generated to include two cysteine residues (C) at fixed positions between 7 random amino acids (X). The experimental size diversity of the library is reported. Random amino acid X are encoded by ‘NNK’ codons. The cysteine-rich macrocyclic peptide library has been constructed using homologous recombination-based methods (Van Deventer and Wittrup 2014; Angelini et al. 2015). The quality of the naïve CX7C library has been assessed by next generation sequencing (NGS). Analysis of the libraries confirmed that the encoded disulfide-tethered macrocyclic peptide ligands present the expected number of cysteines, amino acid frequency and topological diversity. Bar plot indicating the normalised frequency (%) of the number of cysteines (b) and the ‘TAG’ amber stop codon (c) determined by NGS analysis; d) Bar plot representing the experimental frequency of each individual amino acid determined by NGS analysis compared to the theoretical expected values for a ‘NNK’ codon. The individual amino acids (20) are indicated by a one-letter code. The amber stop codon is indicated by the letter ‘Z’. The height of each bar correlates with the frequency ratio. Enriched amino acids in naïve libraries have an experimental / theoretical ratio >1, whereas depleted amino acids show an experimental / theoretical ratio <1. Figure 4. Design and analysis of an exemplificative ‘two-rings’ cysteine-rich macrocyclic peptide library (X)nC(X)mC(X)tC(X)ydisplayed on the yeast cell surface. a) The exemplificative ‘two-rings’ macrocyclic peptide library named CX3CX9C has been generated to include three cysteine residues (C) at fixed positions and 12 random amino acids (X). The experimental size diversity of the library is reported. Random amino acid X are encoded by ‘NNK’ codons. Cysteine-rich macrocyclic peptide library has been constructed using homologous recombination-based methods (Van Deventer and Wittrup 2014; Angelini et al. 2015). The quality of the naïve CX3CX9C library has been assessed by next generation sequencing (NGS). Analysis of the libraries confirmed that the encoded disulfide-tethered macrocyclic peptide ligands present the expected number of cysteines, amino acid frequency and topological diversity. Bar plot indicating the frequency of the number of cysteines (b) and the ‘TAG’ amber stop codon (c) determined by NGS analysis; d) Bar plot representing the experimental frequency of each individual amino acid determined by NGS analysis compared to the theoretical expected values for a ‘NNK’ codon. The individual amino acids (20) are indicated by a one-letter code. The amber stop codon is indicated by the letter ‘Z’. The height of each bar correlates with the frequency ratio. Enriched amino acids in naïve libraries have an experimental / theoretical ratio >1, whereas depleted amino acids show an experimental / theoretical ratio <1. Figure 5. Design and analysis of an exemplificative ‘two-rings’ cysteine-rich macrocyclic peptide library (X)nC(X)mC(X)tC(X)jC(X)y displayed on the yeast cell surface. a) The exemplificative ‘two-rings’ macrocyclic peptide library named CX4CX4CX4C (SEQ ID NO:70) has been generated to include four cysteine residues (C) at fixed positions and 12 random amino acids (X). The experimental size diversity of the library is reported. Random amino acid X are encoded by ‘NNK’ codons. Cysteine-rich macrocyclic peptide library has been constructed using homologous recombination-based methods (Van Deventer and Wittrup 2014; Angelini et al. 2015). The quality of the naïve CX4CX4CX4C library has been assessed by next generation sequencing (NGS). Analysis of the libraries confirmed that the encoded disulfide-tethered macrocyclic peptide ligands present the expected number of cysteines, amino acid frequency and topological diversity. Bar plot indicating the frequency of the number of cysteines (b) and the ‘TAG’ amber stop codon (c) determined by NGS analysis; d) Bar plot representing the experimental frequency of each individual amino acid determined by NGS analysis compared to the theoretical expected values for a ‘NNK’ codon. The individual amino acids (20) are indicated by a one-letter code. The amber stop codon is indicated by the letter ‘Z’. The height of each bar correlates with the frequency ratio. Enriched amino acids in naïve libraries have an experimental / theoretical ratio >1, whereas depleted amino acids show an experimental / theoretical ratio <1. DESCRIPTION OF INVENTION The present invention relates to the design and generation of large cysteine-rich macrocyclic peptide libraries that have been optimised to be recombinantly expressed on the surface of yeast cells (Figure 1). The randomised DNA sequence encoding cysteine-rich macrocyclic peptide variants have been cloned into a vector for the recombinant expression on yeast cells. The vector has been designed and optimised to allow generation of cysteine-rich macrocyclic peptide (MPs) sequences that are expressed on the surface of yeast cells as N-terminal or C-terminal fusion of a cell-surface protein (CSP; Figure 2). The MP sequences can include a number of fixed cysteine residues varying between 2 and 4. The number of random amino acids ‘X’ can vary between 3 and 21. The random amino acids ‘X’ can be a natural (otherwise known as ‘canonical amino acid’ or cAA) and a non- natural amino acid (otherwise known as ‘non-canonical amino acid’ or NcAA). Overall, the length of the generated macrocyclic molecule herein described can range from 5 to 25 amino acid residues. The yeast cell-surface anchor protein (CSP) can possess i) one terminus (either N- or C- terminus) intracellular and one extracellular (either C- or N-terminus), the latter covalently linked to the cysteine-rich macrocyclic peptide sequence of interest, ii) both N- and C-terminus extracellular, one of which (either N- or C-terminus) is covalently linked to the cysteine-rich macrocyclic peptide sequence of interest, iii) the N- or C-terminus terminus not linked to the cysteine-rich macrocyclic peptide sequence of interest covalently or non-covalently bound to the membrane. The vector also includes a gene segment encoding one or more immunofluorescence tag (IFT) located nearby the cell-surface anchor protein (CSP). The IFT is used for identification and / or isolation of yeast cells expressing the MP-CSP (or CSP-MP) fusion on cell surface using different biochemical techniques such as flow cytometry, fluorescence activated cell sorting, magnetic beads, fluorescence microscopy, confocal microscopy, ELISA, microfluidic devices and systems. Moreover, the vector includes a long and flexible linker (FL) placed between the cysteine-rich peptide sequence and the cell-surface anchor protein to minimise potential steric hindrance with both cell-surface anchor protein and immunofluorescence tag detection reagents. Once inserted into the cell, the vector allows recombinant expression of multiple copies (104-106) of a unique disulfide-tethered macrocyclic peptide fused to anchor protein displayed on the surface of yeast cells. The cysteine-rich peptide libraries according to the invention encode either ‘one-ring’ or ‘multiple-rings’ macrocycle peptide topologies with sequence diversity comprised between 106and 109. ‘One-ring’ macrocyclic peptide ligands include 1 disulfide bridge, ‘two-rings’ macrocyclic peptide ligands include 2 disulfide bridges while ‘multiple-rings’ macrocyclic peptide ligands include >2 disulfide bridges. We designed yeast-encoded peptide libraries of the format (X)nC(X)mC(X)t (LIB ID NO:1, Figure 3) in which all macrocyclic peptide sequences contain predominantly two fixed cysteines (C), where X = any amino acid, including both canonical (cAA) and non-canonical (NcAA) amino acid with final length of peptide sequences varying from 5 to 23 amino acids. Similarly, we created yeast-encoded peptide libraries of the format (X)nC(X)mC(X)tC(X)y (LIB ID NO:2, Figure 4) and (X)nC(X)mC(X)tC(X)jC(X)y(LIB ID NO:3, Figure 5) in which all macrocyclic peptide sequences contain predominantly three or four constant cysteines (C) spaced by different random amino acids (X) wherein X = any amino acid, including both canonical (cAA) and non- canonical (NcAA) amino acid, with final length of peptide sequences varying from 7 to 25 amino acids. In the format (X)nC(X)mC(X)tC(X)y(LIB ID NO:2), the theoretical probability of encountering an additional cysteine residue was estimated to be 37%. A first disulfide bond can be formed between any of the two of the three fixed cysteines and a second disulfide bridge could be formed between the third and a fourth cysteine appearing in any of the randomised positions. In total, as many as 3(l – 4)! / 2(l – 6)! different macrocyclic peptide topologies, with a length of l amino acids, could be generated (Kong et al.2020). Such a large structural diversity should increase the chance of identifying multiple rings macrocycle peptides that have a shape complementary to the surface of target protein. Next generation sequencing (NGS) analysis of the naïve cysteine-rich macrocyclic peptide libraries confirmed that the peptide sequences present the expected number of cysteines, amino acid frequency and topological diversity. We showed that by applying quantitative flow cytometry-based selections yeast displaying disulfide-tethered macrocyclic peptide ligands with high structural diversity and fine binding properties against a panel of distinct protein targets can be rapidly identified. The technology appears capable to effectively pick and enrich rare disulfide-tethered macrocyclic peptide ligands with different motifs and topologies for each target tested. In a first aspect, the invention provides a yeast cell containing an expression vector encoding a fusion protein (I): (NH2- or COOH) – [MP] – [FL] – [IFT] – [CSP] – (COOH or NH2) (I) wherein: (i) (NH2or COOH) indicates the amino or carboxyl terminus of the cysteine-rich macrocyclic peptide [MP] of interest and (COOH or NH2) indicates the respective carboxyl or amino terminus of a protein [CSP] expressed on the surface of yeast cells, wherein the first terminus is located outside the cell; (ii) [MP] indicates a macrocyclic peptide (X)nC(X)mC(X)twherein the two cysteine residues ‘C’ form a disulfide bond; n, m, t are 0 or integers with n = 0-5; m = 3-21; t = 0-5 with the condition that n + m + t = 3-21 or [MP] indicates a macrocyclic peptide (X)nC(X)mC(X)tC(X)y wherein two cysteine residues ‘C’ form a disulfide bond; n, m, t and y are 0 or integers, with n = 0-5; m = 0- 22; t = 0-22; y = 0-5 with the conditions that n + m + t + y = 4-22; or [MP] indicates a macrocyclic peptide (X)nC(X)mC(X)tC(X)jC(X)y wherein each pair of Cys residues ‘C’ forms a disulfide bond; n, m, t, j and y are 0 or integers, with n = 0-5, m = 0-21, t = 0-21, j = 0-21, y = 0-5 with the conditions that n + m + t + j + y = 3- 21; ‘X’ denotes any natural (cAA) or non-natural (NcAA) amino acid, wherein: - the natural amino acid (cAA) is encoded by any of the 61 naturally occurring codons or it is encoded by a degenerate codon selected from: ‘NNN’ encoding for all 20 natural amino acids and the TAG, TAA, TGA stop codons; ‘NNK’ encoding for all 20 natural amino acids and the TAG stop codon; ‘NNS’ encoding for all 20 natural amino acids and the TAG stop codon; ‘NDT’ encoding for 12 natural amino acids: R, N, D, C, G, H, I, L, F, S, Y and V; ‘DBK’ encoding for 12 natural amino acids: A, R, C, G, I, L, M, F, S, T, W and V; ‘NRT’ encoding for 8 natural amino acids: R, N, D, C, G, H, S and Y; ‘NNC’ encoding for 15 natural amino acids: F, S, Y, C, L, P, H, R, I, T, N, V, A, D and G; ‘NWW’ encoding for 11 natural amino acids: D, E, F, H, I, K, L, N, Q, V, Y and the TAA stop codon; ‘RVK’ encoding for 10 natural amino acids: A, D,E ,G, H, K, N, R, S and T; ‘DVT’ encoding for 8 natural amino acids: A, C, D, G, N, S, T and Y; ‘NVT’ encoding for 10 natural amino acids: C, D, G, H, N, P, R, S, T and Y; ‘VVC’ encoding for 9 natural amino acids: A, D, G, H, N, P, R, S and T; ‘RST’ encoding for 4 natural amino acids: A, G, S and T; ‘NTT’ encoding for 4 natural amino acids: F, I, L and V, ‘TDK’ encoding for 5 natural amino acids: C, F, L, W, Y and the TAG stop codon; with the condition that at least one random amino acid is present in the sequence of the macrocyclic peptide of interest, preferably 20 to 50% of the amino acids in the sequence are random, more preferably > 50% of the amino acids are random; - the non-natural amino acid (NcAA) is encoded by a TAG, TAA or TGA stop codon present within the coding sequence for the peptide of interest in the expression vector; (iii) [FL] indicates a spacer sequence containing from 1 to 25 amino acids, preferably selected from the following sequences: (GGGGS)n(SEQ ID NO:71) where n ranges from 1 to 5; (G)n(SEQ ID NO:72) where n is an integer ranging from 1 to 25; (S)n(SEQ ID NO:78) where n is an integer ranging from 1 to 25; (P)n (SEQ ID NO:73) where n is an integer ranging from 1 to 25; (EAAAK)n (SEQ ID NO:74) where n is an integer ranging from 1 to 5; (A(EAAAK)4ALEA(EAAAK)4A)n(SEQ ID NO:75) where n is an integer ranging from 1 to 2; (XP)n(SEQ ID NO:77) where n is an integer ranging from 1 to 12 and ‘X’ can be any natural amino acids, preferably an alanine (A), a lysine (K) and a glutamic acid (E); (iv) [IFT] indicates a sequence suitable for the detection and / or isolation of yeast cells expressing the MP-CSP or CSP-MP fusion on yeast cell surface; (v) [CSP] indicates a protein expressed on the surface of yeast cells that can be fused to the N- or C-terminus of the macrocyclic peptide [MP] sequence, containing: (a) at least one extracellular portion; (b) a transmembrane portion or a portion binding the membrane and / or wall that is covalently or non-covalently anchored to molecules of the membrane and / or wall. In one embodiment, said [CSP] is Aga1 protein; Aga2 protein; a member of the glycosylphosphotidylinositol (GPI) binding protein family, preferably Cwp1p, Cwp2p, Tip1p, Flop1p, Sed1p, YCR89, Tir1p; or a synthetic, non-naturally occurring protein that is capable of anchoring to the yeast membrane and / or wall in a stable manner via covalent or non-covalent interactions. In another embodiment, the yeast cell contains one or more vectors encoding an orthogonal translation system (OTS) consisting of an aminoacyl-tRNA synthetase (aaRS) and the corresponding transfer RNA (tRNA), capable of inserting a non-natural amino acid (NcAA) at the TAG, TAA, TGA stop codons present in the expression vector encoding the fusion protein. In a further embodiment, the non-natural amino acid (NcAA) present in the macrocycle peptide [MP] is encoded by a stop codon (TAG, TAA, TGA) present within the macrocyclic peptide DNA sequence. The NcAA can be inserted at the level of a stop codon (TAG, TAA, TGA) into the macrocyclic peptide sequence by a process of natural amino acid depletion or through the action of a bio-orthogonal translation system (OTS) comprising an aminoacyl- tRNA synthetase (aaRS) and the corresponding transfer RNA (tRNA). The bio-orthogonal aaRS / tRNA system is encoded by a plasmid that is inserted into the yeast cell together with the vector encoding the cysteine-rich macrocyclic peptide of interest. A NcAA is defined as a chemical molecule consisting of a carbon atom bonded to a hydrogen, a carboxyl group, an amine group, and a side chain that does not occur naturally. The NcAA can undergo selective post-translational modification. In a further embodiment, [IFT] indicates a sequence suitable for the detection and / or isolation of yeast cells expressing the MP-CSP (or CSP-MP) fusion on cell surface using different biochemical techniques such as flow cytometry, fluorescence activated cell sorting, magnetic beads, fluorescence microscopy, confocal microscopy, ELISA, microfluidic devices and systems. Examples of IFT sequences that can be used include c-myc-tag (EQKLISEEDL_SEQ ID NO:1), HA-tag (YPYDVPDYA_SEQ ID NO:2), Flag-tag (DYKDDDK_SEQ ID NO:3), His-tag (HHHHHH_SEQ ID NO: 4), Strep-tag (WSAPGFEK_SEQ ID NO:5), TC-tag (CCPGCC_SEQ ID NO:6), S-tag (KETAAAKFERQHMDS_SEQ ID NO:7), AviTag (GLNDIFEAQKIEWH_SEQ ID NO:8), Strep-tag2 (AWRHPQFGG_SEQ ID NO:9). In a further embodiment, [CSP] indicates a protein expressed on the surface of yeast cells, containing: (c) at least one extracellular portion; (d) a transmembrane portion or a portion binding the membrane and / or wall that is covalently or non-covalently anchored to molecules of the membrane and / or wall, such as fatty acids, sugars, chemically synthesised molecules or recombinant synthesised biomolecules. Preferably the cell surface proteins [CSP] that can be fused to the N- or C-terminus of the macrocyclic peptide [MP] sequence of interest include, but are not limited to, Aga1 protein, Aga2 protein, or a member of the glycosylphosphotidylinositol (GPI) binding protein family, preferably Cwp1p, Cwp2p, Tip1p, Flop1p, Sed1p, YCR89, Tir1p, or a synthetic, non-naturally occurring protein that is capable of anchoring to the yeast membrane and / or wall in a stable manner via covalent or non-covalent interactions. In another embodiment, the invention provides an expression vector encoding the fusion protein (I) as herein disclosed. Preferably, a yeast cell may contain one or two expression vectors encoding for a cysteines-rich macrocycle peptide (MP) sequence fused to the amino (N-) or carboxyl (C-) terminus of a protein expressed on the surface of yeast cells [CSP], an immunofluorescence [IFT] tag and a spacer [FL], as defined above (I). The said vector is preferably a plasmid suitable for replication and expression in yeast cells; more preferably, it further comprises an inducible promoter [P] functionally linked to the fusion protein (I). In a preferred aspect, said vector further comprises a secretion signal [SS] at position 5' relative to the fusion protein (I). Preferably, the secretion signal [SS] is a sequence selected from Aga2 protein secretion signal (MQLLRCFSIFSVIASVLA; SEQ ID NO:10), alpha-mating factor secretion signal (MRFPSIFTAVLFAASSALAAPANTTTEDETAQIPAEAVIDYSDLEGDFDAAALPLSNS TNGLSSTNTTIASIAAKEEGVQLDKREA; SEQ ID NO: 11), phage secretion signal (MKYLLPTAAAGLLLLAAQPAMA; SEQ ID NO:12), chicken lysozyme (CLY) secretion signal (MRSLLILVLCFLPLAALG; SEQ ID NO:13), CLY-L8 (MRLLLLLLPLAALG; SEQ ID NO:14), S. cerevisiae invertase (SUC2) (MLLQAFLLAGFAAKISA; SEQ ID NO:15), MF- alpha (Prepro) (MRFPSIFTAVLFAASSALAAPVNTTTEEGVSLEKR; SEQ ID NO:16), MF- alpha (Pre)-apv (MRFPSIFTAVLFAASSALAAPV; SEQ ID NO:17), MF-alpha (Pre)-apv- SLEKR (MRFPSIFTAVLFAASSALAAPVSLEKR; SEQ ID NO:18), MF-alpha (Prepro)- (EA)3 (MRFPSIFTAVLFAASSALAAPVNTTTE-EGVSLEKREAEAEA; SEQ ID NO: 19), αF peptide signal (MRFPSIFTAVLFAASSALAAPVNTTTE- DETAQIPAEAVIGYSDLEGDFDVAVLPFSNSTNGLNGLLFINTTIASIAAKEEGVSLEK R; SEQ ID NO: 20), KILM1 peptide signal (MTKPTQVLVRSVSILFFITLLHLVVALNDVAGPAETAPVSLLPR; SEQ ID NO:21), repressible acid phosphatase (PHO1) peptide signal (MFSPILSLEIILALATLQSVFA; SEQ ID NO:22), GOX A. niger peptide signal (MQTLLVSSLVVSLAAALPHYIR; SEQ ID NO:23), S. occidentalis glucoamylase (GAM1) gene signal peptide (MIFLKLIKSIVIGLGLVSAIQA; SEQ ID NO:24), human serum albumin (HSA) signal peptide with pro sequence (MKWVTFISLLFLFSSAYSRGVFRR; SEQ ID NO:25), human serum albumin (HSA) signal peptide without pro sequence (MKWVTFISLLFLFSSAYS; SEQ ID NO:26), human serum albumin (HSA) signal peptide ISN (MALWMRLLPLLALLALWGPDPAA; SEQ ID NO:27), human serum albumin (HSA) signal peptide IFN (MKYTSYILAFQLCIVLGSLGCDLP; SEQ ID NO:28), HGH signal peptide (MAADSQTPWLLTFSLLCLLWPQEPGA; SEQ ID NO:29), phytohaemagglutinin (PHA) (MKKNRMMMIWSVGVWMLLLVGGSYG; SEQ ID NO:30), silkworm lysozyme (MQKLIIFALVVLCVGSEA; SEQ ID NO:31), human lysozyme (LYZ1) (MKALIVLGLVLLSVTVQG; SEQ ID NO:32), type I activin receptor (MVDGVMILPVLIMIALPSPS; SEQ ID NO:33), type II activin receptor (MGAAAKLAVFLISCSSG; SEQ ID NO:34), immunoglobulin-binding protein P. pastoris (PpBiP) immunoglobulin binding protein (MLSLKPSWLTLAALMYAMLLVVPFAKPVRA; SEQ ID NO:35), and human antibody 3D6 light chain leader (MDMRVPAQLLGLLWLPGAKC; SEQ ID NO:36) (Felgenhauer, Kohl, and Rüker 1990; Damasceno et al.2007; Daly and Hearn 2006; Koganesawa et al.2001; Raemaekers et al.1999; Gellissen et al.2005; OKA et al.1999; Hashimoto, Koyabu, and Imoto 1998; Kingwell 2023; Mullard 2023). In a further aspect, the invention provides a method of preparing the expression vector herein disclosed, in which the fusion protein (I) has the following orientation: (HOOC)-[CSP]-[IFT]-[FL]-[MP]-(NH2) (I), which comprises the following steps: (1) providing an expression vector containing, from 5' to 3', the following construct: 5'-[ip] – [ss] – [re] – [sr] – [fl] – [ift] – [csp]-3' wherein: [ip] indicates an inducible promoter (e.g., GAL1-10) [ss] indicates a DNA sequence encoding a secretion signal as defined above in relation to the expression vector; [re] indicates a DNA sequence that can be specifically and uniquely recognised by a restriction enzyme, preferably selected from, but not limited to, BamHI, NheI, EcoRI, XhoI, SalI, NdeI, NcoI, SfiI, SfoI, BglII, KpnI, HindIII, XbaI, NotI, wherein said site is located within and / or at the 5' end and / or the 3' end of the DNA spacer region [sr]; (iv) [sr] indicates a spacer region comprising 6 to 3000 nucleotides; (v) [fl] indicates a DNA sequence encoding for a flexible linker [FL] as defined above in relation to the fusion protein (I); (vi) [ift] indicates a DNA sequence encoding an immunofluorescence detection tag [IFT] as defined above in relation to the fusion protein (I); (vii) [csp] indicates a DNA sequence encoding a cell-surface anchor protein [CSP] as defined above in relation to the fusion protein (I) and contains at least one restriction site as defined in (iii) within and / or at the 5' end and / or the 3' end. (2) inserting the sequence encoding the macrocycle peptide of interest into the vector by PCR reaction using the following primer pair: (i) reverse primer that pairs to the antisense strand with respect to the sequence encoding the macrocyclic peptide [MP] of interest in a region about 10 to 60 base pairs downstream of the restriction enzyme [re] cut site contained within or at the ends of the cell-surface anchor protein encoding gene (csp); preferably the reverse primer has length ranging from 5 to 60 base pairs with a pairing temperature comprised within 45 °C to 80 °C, preferably within 55 °C to 70 °C, more preferably within 60 °C to 65 °C; (ii) forward primer that pairs to the sense strand with respect to the sequence encoding the macrocyclic peptide of interest [MP] and that encodes from the 5' to 3' terminus: 5'-[region of pairing to ss]-[mp of interest]-[region of pairing to sr]-3', where: (a) the region of pairing to the secretion signal (ss) is about 10 to 60 base pairs upstream the restriction enzyme (re) cut site located between the secretion signal and the DNA linker, with a temperature of annealing above 20 °C, preferably above 30 °C, more preferably above 40 °C; (b) the region of pairing to the spacer region (sr) has a length ranging from 5 to 60 base pairs and a pairing temperature comprised within 45 °C and 80 °C, preferably within 55 °C and 70 °C and, more preferably within 60 °C and 65 °C. In a further aspect, the invention provides a method of preparing an expression vector as herein defined, in which the fusion protein (I) has the following orientation: (NH2)-[CSP]-[IFT]-[FL]-[MP]-(COOH) (I), which comprises the following steps: (1) providing an expression vector containing, from 5' to 3', the following genetic elements: 5'-[ip] – [ss] – [csp] – [ift] – [fl] – [sr] – [re]-3' where: [ip] indicates an inducible promoter as defined above in relation to the fusion protein (I), [ss] indicates a DNA sequence encoding a secretion signal as defined above in relation to the fusion protein (I), [csp] indicates a DNA sequence encoding a cell-surface anchor protein [CSP] as defined above in relation to the fusion protein (I), [ift] indicates a DNA sequence encoding an immunofluorescence detection tag IFT as defined above in relation to the fusion protein (I), [fl] indicates a DNA sequence encoding for a flexible linker FL as defined above in relation to the fusion protein (I), [sr] indicates a spacer region as defined above in relation to the fusion protein (I) and [re] indicates a DNA sequence that can be specifically and uniquely recognised by a restriction enzyme as defined above in relation to the fusion protein (I). (2) inserting the coding sequence for the macrocyclic peptide of interest into the vector by PCR reaction using the following primer pair: (i) forward primer that pairs to the sense strand with respect to the sequence encoding the peptide of interest [MP], in a region with 10-60 base pairs upstream the restriction enzyme (re) cut site located at the 5' end or within the cell-surface anchor protein encoding gene (csp); preferably, the forward primer has a length ranging from 5 to 60 base pairs and a pairing temperature comprised within 45 °C to 80 °C, more preferably within 55 °C to 70 °C, more preferably within 60 °C to 65 °C; (ii) reverse primer that pairs to the antisense strand with respect to the sequence encoding the peptide of interest and that encodes 5' to 3' terminus: 5'-[backbone pairing region]-[mp of interest]-[region of pairing to sr]-3', where: (a) the backbone pairing region pairs to a sequence of the plasmid located 10 to 50 base pairs downstream the restriction site (re) at the end or within the linker sequence (fl), with a pairing temperature greater than 50 °C; (b) the region of pairing to the spacer (sr) has a length of 5 to 50 base pairs and a pairing temperature comprised within 45 °C and 80 °C, preferably within 55 °C and 70 °C, more preferably within 60 °C and 65 °C; with the condition that the difference in pairing temperature between the two primers must be less than 20 °C, preferably less than 10 °C, more preferably less than 5 °C. Preferably, the PCR reaction is conducted according to the following procedure: 1) initial denaturation at 95 – 98 °C (3 min) 2) 30 cycles consisting of three steps: 95 °C (30 sec), 60 – 55 °C (30 sec) and 72 °C (1 – 5 min) 3) final elongation at 72 °C (10 min). The method of the invention allows: - the generation of C- or N-terminus fused cysteine-rich macrocyclic peptide libraries in different yeast cell display systems including, but not limited to, Aga1-Aga2, Cwp1p, Cwp2p, Tip1p, Flop1p, Sed1p, YCR89, Tir1p, GPL-1, GPL-1 polypeptide derivatives, synthetic artificial proteins or fusion proteins. Broad compatibility is provided by the inclusion of homology regions that are universal and translatable on different genetic vectors; - the generation of high-diversity cysteine-rich macrocyclic peptide libraries, fused at the C- or N-terminus in different yeast cell display systems including, but not limited to, Aga1-Aga2, Cwp1p, Cwp2p, Tip1p, Flop1p, Sed1p, YCR89, Tir1p, GPL-1, GPL-1 polypeptide derivatives, artificial synthetic proteins or fusion proteins,; - direct next generation sequencing analysis of both naïve and selected libraries. The PCR conditions are based solely on the pairing regions, which depend only on the sequence of the chosen starting vector rather than the sequence, structure or length of the designed cysteine-rich macrocyclic peptide region. The procedure can be applied both for creating fully and / or partially randomised sequences as well as for cloning peptides with defined amino acid sequences. In a further aspect, the invention provides a cell population consisting of a plurality of yeast cells, wherein each single cell expresses on its surface a unique cysteine-rich macrocyclic peptide [MP] of interest whose sequence varies within the cell population. Different cells within the cell population may express different cysteine-rich macrocyclic peptide [MP] of interest, preferably a different peptide for each cell, wherein said peptides of interest are selected from the group comprising: (X)nC(X)mC(X)tt; (X)nC(X)mC(X)tC(X)y; (X)nC(X)mC(X)tC(X)jC(X)y. Each single yeast cell can express on the cell surface multiple copies (104– 106) of a unique cysteine-rich macrocyclic peptide [MP] sequence fused to cell-surface protein chosen from the previously reported libraries (LIB ID NO: 1-3). In a preferred embodiment, the yeast cell population of the invention may have a diversity, in terms of sequence of the macrocyclic peptide of interest, of at least 1 ^ 106, preferably superior to 1 ^ 108, more preferably superior to 1 ^ 109. The term ‘diversity’ refers to the number of different genetically modified yeast cells obtained upon creation of the library. The ‘diversity’ can be calculated by i) plating serial dilutions and counting colonies of successful yeast cells transformants containing vectors encoding the cysteine-rich macrocyclic peptide fused to a cell-surface protein and / or ii) applying massive parallel next generation sequencing (NGS) analyses on yeast-extracted DNA vectors encoding the cysteine-rich macrocyclic peptide sequences. The invention allows to efficiently generate and isolate de novo high-affinity and specificity disulphide-bound macrocyclic peptide ligands from large and topologically diverse combinatorial libraries expressed on the surface of yeast cells. Yeast surface display can also be applied to enhance the properties of macrocyclic peptide ligands occurring in nature and / or that have been previously developed using other techniques (hit-to-lead process). Moreover, yeast surface display can be combined with other in vitro directed evolution tools (e.g., phage and mRNA display) to further harness the unique advantages of each, thus enabling previously unexplored applications. These include the evaluation and improvement of binding and stability properties and the rapid and fine epitope mapping. Moreover, based on the enormous success that yeast surface display technology achieved in rapidly screening and assessing the biophysical properties of hundreds of computationally designed proteins (Pan and Kortemme 2021; Linciano et al. 2019; Listov et al. 2024), the tool is suitable also for effectively characterising in silico developed genetically encoded macrocyclic peptide ligands. Another use of a macrocyclic peptide library according to the invention is for screening of biologically active disulfide-tethered macrocyclic peptide ligands of relatively small size (5 – 25 amino acids length). Screening and selection of disulfide-tethered macrocyclic peptide ligands capable of binding a given target of interest (e.g., protein, DNA, RNA, cell) can be conveniently performed by using magnetic bead (MB) separation and / or fluorescence-activated cell sorting (FACS). The identity of isolated disulfide-tethered macrocyclic peptide ligands can be revealed by applying both Sanger and next generation DNA sequencing. Applications of selected disulfide-tethered macrocyclic peptide ligands include, but are not limited to, development of therapeutic agents to be used alone or in combination with other pharmacological active entities (e.g., radionuclides, chemotherapeutics, protein, antibodies, cells, DNA, RNA, lipid nanoparticles). Disulfide-tethered macrocyclic peptide ligands discovered using yeast display technology could be potentially utilised also to develop diagnostic tools or being applied in the veterinary, pant and food sectors. EXPERIMENTAL SECTION Generation of a cysteine-rich macrocycle peptide library 1. PCR reaction for the creation of the cysteine-rich macrocycle peptide insert fused to a cell-surface protein. The procedure disclosed above for the construction of an expression vector was followed. The vector contains the following elements: 5'-[ip] – [ss] – [re] – [sr] – [fl] – [ift] – [csp]-3' or 5'-[ip] – [ss] – [csp] – [ift] – [fl] – [sr] – [re]-3' The PCR reaction of the insert were performed in a 50 µL reaction volume containing the degenerate forward primer F1 (500 nM), a reverse primer R (500 nM), a mixture of dNTPs (200 µM each), 1 ng of pCT-CSP as templated DNA, 10X DreamTaq buffer (1X), 1.25 U of DreamTaq DNA polymerase (Thermo Fisher) and H2O milli-Q. The PCR reaction is conducted according to the following steps: 1) initial denaturation at 95 – 98 °C (3 min) 2) 30 cycles consisting of three steps: 95 °C (30 sec), 60-55 °C (30 sec) and 72 °C (1 – 5 min) 3) final elongation at 72 °C (10 min). The PCR conditions are based solely on the pairing regions, which depend only on the sequence of the chosen starting vector rather than the sequence, structure or length of the designed cysteine-rich macrocyclic peptide region. The procedure can be applied both for creating random sequences as well as for cloning known amino acid sequences. The backbone vector DNA is linearised using 1 to 5 different enzymes, preferably 2 – 3 restriction enzymes, more preferably 3 restriction enzymes. The digestion is performed in reaction volumes comprised between 20 and 200 µL and incubation time varying from 1 to 48 hours, preferably varying from 12 to 32 hours, more preferably varying from 16 to 30 hours. The enzymatic digestion product is purified by ethanol precipitation and resuspended in sterile H2O milli-Q at a concentration comprised within 0.5 – 3 µg / µL, preferably within 1 – 2 µg / µL. Absorbance ratio 260 / 280 is comprised within 1.4 – 2.0, preferably higher than 1.6, and ratio 260 / 230 is comprised within 1.6 – 2.5, preferably higher than 1.8. The reported library formats were obtained using the homologous recombination method described in the literature (Angelini et al.2015). To increase the diversity of the library, the following modifications were made to the method previously reported in the literature: - lack of penicillin and streptomycin in the media during all passages; - initial incubation of cells at an OD600 comprised within 0.25 – 0.3; - pre-incubation of the digested vector and PCR product for 5 – 15 minutes on ice, referred to as DNA mix. The DNA mix has a volume of 5 – 20 µL, preferably 8 – 15 µL. The mix has an insert to vector ratio of 50:1 with a total DNA amount between 5 – 20 µg, preferably between 10 – 15 µg, more preferably between 12 – 14 µg; - pre-incubation of the DNA mix with the yeast cells for 5 – 30 minutes on ice, before proceeding with electroporation; - rescuing of the electroporated yeast cells with YPD or YPD:sorbitol solution (1:1); - washing of the electroporated yeast cells and incubation at 30 °C with fresh SD-CAA media without penicillin and streptomycin. Libraries are obtained by performing 8 to 32 electroporations, preferably between 10 – 20 electroporations and have diversities comprised within 108– 109, preferably higher than 109. 2. Deep sequence analysis of the cysteine-rich macrocyclic peptide libraries displayed on the surface of yeast cells To gain a better insight into the genetic diversity and abundance of the cysteine-rich macrocyclic peptide naïve library, next generation sequencing (NGS) analysis is performed. The NGS data from naïve libraries were analysed as follows: 1) partitioning according to barcode sequence; 2) translation of the cysteine-rich macrocyclic peptide sequences; 3) quantification of the TAG codon; 4) statistical analysis of amino acid frequency. Amplicons were sequenced with a NovaSeq6000 (Illumina, IgaTech), obtaining paired- end reads of 150 base pairs. 3. Extraction of plasmid DNA from yeast cells of a polyclonal library 1) Growing a number of yeast cells encoding the naïve library at least 10 times higher than the estimated diversity of the library at a concentration comprised within 0.1 ^ 107– 108cells / mL (preferably between 1 – 3 ^ 107cells / mL), for 16 – 24 hours at 30 °C under shacking (250 rpm); 2) Dilute the yeast cells encoding the naïve library to a concentration of 107cells / mL by using 2 – 10 times more cells than the estimated naïve library diversity; 3) Growing the yeast cells encoding the naïve library to a concentration of 2 – 3 ^ 107cells / mL; 4) Extracting plasmid DNA from yeast cells encoding the naïve library using a commercial kit (e.g. Zymoprep yeast plasmid miniprep II, Zymo Research); 4. Generation of NGS amplicons by performing two sequential PCR reactions The first PCR is carried out in a 30 µL reaction volume containing: - 400 nM universal forward primer per vector; - 400 nM universal reverse primer per vector; - 250 µM of dATP; - 250 µM of dCTP; - 250 µM of dTTP; - 250 µM of dGTP; - 30% DMSO; - 10 µL of yeast-extracted plasmid DNA, corresponding to a total amount of plasmid DNA ranging from 0.1 – 100 ng (number of DNA molecules comprised within 107– 1010); - Taq DNA polymerase buffer, 1X working concentration; - 0.2 – 2 U of Taq DNA polymerase; - sterile H2O milli-Q. The forward primer of the first PCR has the following 5' – 3' structure: (a) an extra region, that is recognised and anneal to the 3' end of the second forward primer used during the second NGS-PCR. The length of the region can vary from 1 to 50 base pairs; b) pairing region with length varying from 5 to 50 base pairs and a pairing temperature ranging from 45 to 80 ° C, preferably from 70 to 55 °C, more preferably from 68 to 63 °C. The primer pairs in a region ranging from 1 to 200 base pairs, preferably from 1 to 100 base pairs, more preferably from 1 to 50 base pairs upstream the 5' end of the DNA encoding the cysteine- rich macrocyclic peptide sequence of interest. The reverse primer of the first PCR has the following 5' – 3' structure: (a) random region with length varying from 0 to 50 base pairs; b) pairing region with length varying from 5 to 50 base pairs with a pairing temperature ranging from 45 to 80 °C, preferably from 70 to 55 °C, more preferably from 68 to 63 °C. The primer pairs in a region preferably ranging from 10 to 200 base pairs downstream the 3' end of the DNA encoding for the macrocyclic peptide of interest. The product of the first PCR can be purified by ethanol precipitation, gel extraction, silica column or any commercial kit suitable for purification of PCR amplified DNA fragments. The DNA product is eluted in sterile H2O milli-Q or commercial recommended buffer at a concentration ranging from 0.01 to 1 µg / µL and absorbance ratios 260 / 280 from 1.4 to 2.0, preferably higher than 1.6, and ratio 260 / 230 ranging from 1.6 to 2.5, preferably higher than 1.8. The second PCR is performed in a 50 µL reaction volume containing: - 400 nM forward primer containing the barcoding region; - 400 nM reverse primer; - 250 µM of dATP; - 250 µM of dCTP; - 250 µM of dTTP; - 250 µM of dGTP; - 30% DMSO; - 10 ng of first PCR product; - High-fidelity DNA polymerase buffer, 1X working concentration; - 0.2-2 U of Phusion High-fidelity DNA polymerase; - sterile H2O milli-Q The forward primer of the second PCR has the following 5' – 3' structure: a) a region containing the nucleotide sequence of the Illumina forward adaptor; b) a barcoding nucleotide sequence with lengths varying from 4 to 20 base pairs, preferably from 8 to 15 base pairs and with C and G content from 20 to 70%, preferably from 40 to 60% (examples of barcoding used are shown in Table 1); c) pairing region with length varying from 5 to 50 base pairs with a pairing temperature ranging from 45 to 80 °C, preferably from 70 to 55 °C, more preferably from 68 to 63 °C. The primer pairs in a region ranging from 1 to 200 base pairs, preferably from 1 to 100 base pairs, more preferably from 1 to 50 base pairs upstream the 5' end of the DNA encoding the macrocyclic peptide of interest. ID SEQ NO Barcode Sequence (5’-3’) 37 A CTGCACATCGAT 38 B ATGTCCAACTAC 39 C GCGTACACTCAC 40 D TCTGAGGATCCA 41 E CCTTACGATGTG 42 F GACACTGATGGT 43 G CACAGAGATACT 44 H TGCTTGGATTCG 45 I AGTGTACACTGC Table 1. Example of DNA barcode sequences used for NGS analysis The reverse primer of the second PCR has the following 5' – 3' structure: a) a region containing the nucleotide sequence of the Illumina reverse adaptor; b) a pairing region with length varying from 5 to 50 base pairs with a pairing temperature comprised from 45 to 80 °C, preferably from 70 to 55 °C, more preferably from 68 to 63 °C. The primer pairs in a region ranging from 1 to 200 base pairs downstream the 3' end of the DNA sequence encoding for the macrocyclic peptide of interest. The PCR programme set up for the two reactions is: 1) initial denaturation at 95 – 98 °C (1- 3 min); 2) 30 cycles consisting of three steps: 95 – 98 °C (30 sec), 60 – 55 °C (30 sec) and 72 °C (1 – 5 min); 3) final elongation at 72 °C (2 – 10 min); The product of the second PCR is purified by gel extraction from agarose gel. The gel must have an agarose content ranging from 0.8 to 2.5% w / v, preferably from 1 to 2% w / v, more preferably from 1.2 to 1.5% w / v. The product is eluted in sterile H2O milli-Q or commercial recommended buffer at a concentration ranging from 0.01 to 1 µg / µL and absorbance ratios 260 / 280 from 1.4 to 2.0, preferably higher than 1.6, and ratio 260 / 230 from 1.6 to 2.5, preferably higher than 1.8. 5. Generation of yeast-encoded cysteine-rich macrocyclic peptide libraries with structure (X)nC(X)mC(X)t and (X)nC(X)mC(X)tC(X)y fused to the N-terminus of cell-surface anchored protein and including only canonical amino acids (cAA) Large and structurally diverse cysteine-rich macrocyclic peptide libraries (5 – 25 amino acids) displayed on the surface of yeast cells were generated by forming intra-molecular disulfide bridges between cysteine residues. To secure distance and minimise potential steric hindrance with detection reagents, we placed a flexible glycine-serine (G4S)3linker (SEQ ID NO:76) between the peptide sequence and the immunofluorescence tag (IFT). Detection of the IFT tag using fluorescently labelled antibody combined with flow cytometry allows not only rapid quantification of the expression level of disulfide-tethered macrocyclic peptide ligands on yeast cells, but also normalization of the binding signal to surface expression. We generated cysteine-rich peptide libraries preferentially encoding either ‘one ring’ or ‘two rings’ macrocycle topologies with sequence diversity comprised between 106and 109. We designed yeast-encoded ‘one ring’ macrocyclic peptide libraries of the format (X)nC(X)mC(X)tin which all sequences contain predominantly two fixed cysteines where X = any amino acid, including both canonical (cAA) and non-canonical (NcAA) amino acid, and n = 0-5, m = 3-21 and t = 0-5 (final length of ‘one-ring’ macrocyclic peptide sequences varies from 5 to 23 amino acids). Similarly, we created yeast-encoded ‘two-rings’ macrocyclic peptide libraries of the format (X)nC(X)mC(X)tC(X)y in which at least the 30% of the peptides contains four cysteines spaced by different numbers of random amino acids (X) wherein X = any amino acid, including both canonical (cAA) and non-canonical (NcAA) amino acid, and n = 0-5, m = 0-22, t = 0-22, y = 0-5 (final length of ‘two-rings’ macrocyclic peptide sequences varies from 7 to 25 amino acids). The random amino acid positions in all libraries were encoded by ‘NNK’ codons. Next generation sequencing (NGS) analysis of the naïve libraries confirmed that the macrocyclic peptides present the expected number of amber stop codon (‘TAG’; Table 2), amino acid frequency and topological diversity (Table 4). We showed that by applying quantitative flow cytometry-based selections, yeast displayed disulfide-tethered macrocyclic peptide ligands with high structural diversity and fine binding properties against a panel of distinct protein targets can be rapidly identified. The technology appears capable to effectively pick and enrich rare disulfide-tethered macrocyclic peptide ligands with different motifs and topologies for each target tested. frequency (%) amber stop codon (‘TAG’) naïve libraries 0 1 2 3 CX7C (SEQ ID: 79) 50.03% 39.49% 9.47% 1.00% CX3CX9C (SEQ ID: 80) 50.18% 32.90% 14.03% 2.85% Table 2. Experimental probability of identifying additional amber stop codon (‘TAG’) in two exemplificative yeast-encoded macrocyclic peptide naïve libraries: the ‘one-ring’ macrocyclic peptide library of the format (X)nC(X)mC(X)twhere n = 0, m = 7 and t =0 and the ‘two-rings’ macrocyclic peptide library of the format (X)nC(X)mC(X)tC(X)ywhere n = 0, m = 3, t = 9 and y = 0. The occurrence of ‘TAG’ codons is 1.5 times higher than the theoretical value expected for ‘NNK’. One (1), two (2) and three (3) additional ‘TAG’ codons occur at a rate of approximately 32 – 39%, 9 – 14% and 1 – 3%, respectively. Library diversities were calculated as previously reported in literature (Angelini et al. 2015) and are in the range of 1 ^ 108and 5 ^ 109. To verify the quality of the naïve library, the frequency of occurrence of each amino acid was calculated from the NGS data and compared to the theoretical frequency of an amino acid when using the degenerate NNK codon (Table 3 and 4) (Chen et al.2013). Amino acid NNK probability C 3.13 A 6.25 L 9.38 V 6.25 I 3.13 N 3.13 D 3.13 R 9.38 E 3.13 Q 3.13 Z 3.13 G 6.25 H 3.13 K 3.13 M 3.13 F 3.13 P 6.25 S 9.38 T 6.25 W 3.13 Y 3.13 Table 3. Theoretical probability of occurrence of amino acids in the presence of the degenerate codon NNK. The theoretical probability of having a given amino acid, within a macrocyclic peptide sequence of a given ‘n’ length, is defined by the formula: probability (n) = probability NNK ^ n Some amino acids were found to be over-represented (cysteine, valine, tryptophan and glycine) or under-represented (glutamine, histidine, proline and threonine) with an experimental / theoretical frequency ratio from 1.2 to 1.5 and from 0.8 to 0.5, respectively. However, frequency variations did not cause biases during selection operations, as amino acid abundance did not correlate with naïve abundance and were seen to be exclusively target- dependent (Table 4). Amino acidratio experimental and theoretical frequency of single amino acidCX7C CX3CX9CC 1.30 1.28 A 0.95 0.88 L 0.88 1.07 V 1.29 1.42 I 0.97 0.94Amino acidratio experimental and theoretical frequency of single amino acidCX7C CX3CX9CN 0.95 0.74 D 1.12 0.91 R 1.02 0.98 E 1.06 0.98 Q 0.82 0.80 Z 1.06 1.16 G 1.34 1.27 H 0.87 0.72 K 0.10 0.91 M 1.08 1.19 F 0.99 1.17 P 0.60 0.57 S 0.92 0.92 T 0.83 0.74 W 1.20 1.40 Y 1.05 0.98 Table 4. Ratio between experimental (NGS data) and theoretical (codon 'NNK') frequency of single amino acid. The amino acids are indicated as one letter code. The amber stop codon (‘TAG’) is shown as ‘Z’. A ratio less than 1 indicates that the amino acid is underrepresented, while a ratio greater than 1 indicates that the amino acid is overrepresented. 6. Generation of yeast-encoded cysteine-rich macrocyclic peptide libraries including a non-canonical amino acid (NcAA) named Z at TAG position in the sequence with structure (X)nC(X)mC(X)t fused to the N-terminus of cell-surface anchored protein To test the procedure for the generation of cysteine-rich macrocyclic peptide libraries fused to the N-terminus of a cell-surface anchored protein, compatible with site-specific chemical modifications conducted on non-canonical amino acids, ‘one-ring’ macrocyclic peptide library of the format (X)nC(X)mC(X)t where n = 0, m = 7 or 9 and t =0 were generated by homologous recombination methods as described above and further transformed into yeast cells presenting a plasmid coding for an aminoacyl tRNA transferase and an E. coli bio- orthologue tRNA. Library diversities were calculated as previously reported in literature (Angelini et al. 2015) and are in the range of 1 ^ 108and 1 ^ 109. The macrocyclic peptide libraries were designed to incorporate a non-canonical amino acid at a specific position, named with the letter Z, with a loop of 7 amino acids (CXZX5C, CX3ZX3C, CX5ZXC, AZGCX7C) or with a loop of 9 amino acids (CXZX7C, CX4ZX4C, CX7ZXC, AZGCX9C). Yeast strain and media preparation The yeast strain used is this specific study has the following genotype: MATa AGA1::GAL1AGA1::URA3 ura352 trp1 leu2 delta200 his3 delta200 pep4::HIS3 prbd1.6R can1 GAL. The media for yeast growth and induction were described by Angelini et al. All media were supplemented with Penicillin / Streptomycin (Corning 100X Penicillin solution: Streptomycin). The preparation of liquid media and yeast plates without Uracil (Ura) and Tryptophan (Trp) was previously reported by Angelini et al. Yeast media without Uracil (Ura), Tryptophan (Trp) and Leucine (Leu) were prepared using synthetic drop out medium (Sigma Y1171) as an amino acid source. The solution containing non-canonical amino acid (NcAA) at a concentration of 50 mM to be supplemented in the media with a final concentration of 1 mM was prepared fresh each time. Libraries generation Naïve libraries of cysteine-rich macrocyclic peptides displayed on the surface of yeast cell were generated by homologous recombination methods in yeast cells. Disulfide-tethered macrocyclic peptides were displayed on the surface of yeast as N-terminal fusion of a cell- surface anchor protein (CSP). The example given here refers to the CSP protein Aga2, but can be applied to any CSP with a solvent-exposed extracellular N-terminus. Yeast surface display vector was based on a modified version of the pCT-CON backbone (Angelini et al.2015), here renamed pCT-CSP. This vector contains a DNA sequence encoding for a secretory leader sequence, a NheI restriction site, a sequence encoding for a long and flexible spacer (G4S)3, a sequence encoding for the hemagglutinin tag (HA; YPYDVPDYA) followed by a sequence encoding for the CSP protein Aga2 which has been modified to include a silent BamHI restriction site. The synthetic genes were codon optimized for expression in yeast cells and obtained from Integrated DNA Technologies (Coralville, IA, USA). Yeast-encoded macrocyclic peptide naïve libraries were created by inserting the DNA sequences encoding the random peptide sequences C(X)7C, the flexible (G4S)3linker, the HA tag and a N-terminal portion of the CSP protein Aga2 into the pCT-CSP vector. The insert was created by appending the DNA encoding the random macrocyclic peptide sequences to the N-terminus of the (G4S)3 – HA – Aga2 encoding gene in a PCR reaction (30 cycles) using the forward degenerate primer F1-X (SEQ ID NO:47 – 54) and the universal reverse primer R (SEQ ID NO:46). The primers used to construct the libraries were obtained from Integrated DNA Technologies and are shown in Table 5. PCRs of the insert were performed in a 50 µL reaction volume containing the degenerate forward primer F1 (500 nM), a reverse primer R (500 nM), a mixture of dNTPs (200 µM each), 1 ng of pCT-CSP template DNA, 10X DreamTaq buffer (1X), 1.25 U of DreamTaq DNA polymerase (Thermo Fisher) and H2O milli-Q. The PCR programme was set up as follows: i) initial denaturation at 95 °C for 3 min, ii) 30 cycles of 30 sec at 95 °C, 30 sec at 55 °C and 1 min at 72 °C and iii) final elongation at 72 °C for 10 min. The vector backbone was linearised using BamHI-HF (New England Lab) and NheI-HF (New England Lab). Digestion was performed in a 50 µL reaction volume containing 10 µg of pCT-CSP plasmid, 1U of NheI-HF, 1U of BamHI-HF, 5X Cut Smart and H2O milli-Q. The sample was incubated for 24 hours at 37 °C. The linearised plasmid and PCR product were purified and concentrated to approximately 1 µg / µL by ethanol precipitation. The size and quality of the PCR inserts and the linearised plasmid were verified on a 1.2% w / v agarose gel. Primer SEQ ID NO code function sequence (5’- 3’) 46 R libraryGCCAGATGTTGTCGAACTTTCTGATTTAGT generationGG 47 F1-A libraryGGAAGAAGGTGTTCAATTGGACAAGAGA generationGAAGCTTGTNNKTAGNNKNNKNNKNNKN NKTGTTCCGGTGGTGGTGGCTCTGGTGG 48 F1-B libraryGGAAGAAGGTGTTCAATTGGACAAGAGA generationGAAGCTTGTNNKNNKNNKTAGNNKNNKN NKTGTGGTGGTGGTGGCTCTGGTGG 49 F1-C libraryGGAAGAAGGTGTTCAATTGGACAAGAGA generationGAAGCTTGTNNKNNKNNKNNKNNKTAGN NKTGTTCCGGTGGTGGTGGCTCTGGTGG GGAAGAAGGTGTTCAATTGGACAAGAGA 50 F1-D libraryGAAGCTTAGGGTTGTNNKNNKNNKNNKN generationNKNNKNNKTGTTCCGGTGGTGGTGGCTC TGGTGG GGAAGAAGGTGTTCAATTGGACAAGAGA 51 F1-E libraryGAAGCTTGTNNKTAGNNKNNKNNKNNKN generationNKNNKNNKTGTTCCGGTGGTGGTGGCTC TGGTGG GGAAGAAGGTGTTCAATTGGACAAGAGA 52 F1-F libraryGAAGCTTGTNNKNNKNNKNNKTAGNNKN generationNKNNKNNKTGTTCCGGTGGTGGTGGCTC TGGTGG GGAAGAAGGTGTTCAATTGGACAAGAGA 53 F1-G libraryGAAGCTTGTNNKNNKNNKNNKNNKNNK generationNNKTAGNNKTGTTCCGGTGGTGGTGGCTC TGGTGG GGAAGAAGGTGTTCAATTGGACAAGAGA 54 F1-H libraryGAAGCTTAGGGTTGTNNKNNKNNKNNKN generationNKNNKNNKNNKNNKTGTTCCGGTGGTGG TGGCTCTGGTGG Table 5. Primer sequences used for macrocyclic peptide library generation. NGS analysis of naïve libraries To determine the quality of the libraries obtained, NGS analysis was performed as described below. NGS amplicons of 280 base pairs were generated by performing two PCR reactions on DNA plasmid extracted from yeast cells. For each polyclonal library, a target-specific barcode and an Illumina sequencing tag were added during the second PCR. The primers used for the library barcode tag were obtained from Integrated DNA Technologies and are described in Table 6. The first PCR was performed in a 30 μL reaction volume containing 5X HF buffer (1X), DMSO (30% v / v), dNTPs (250 μM each), NGS-R primers (400 nM), NGS-F1 (400 nM), Phusion high-fidelity DNA polymerase (0.6 U; New England Biolabs) and 10 μL of DNA plasmid extracted from yeast. The PCR amplicons were purified using the Zymoprep Clean and Concentrator kit (Zymo Research). The second PCRs were performed in a 50 µL reaction volume containing 5X HF buffer (1X), DMSO (30% v / v), dNTPs (250 μM each), NGS-R2 primers (400 nM), NGS-2F-X primers (400 nM), Phusion high-fidelity DNA polymerase (0.6 U; New England Biolabs) and 10 ng of amplicons from the first PCR. The two PCR programmes were set up as follows: i) initial denaturation at 98 °C for 1 min, ii) 30 cycles of 30 sec at 98 °C, 30 sec at 57 °C, 1 min at 72 °C and iii) final elongation at 72 °C for 5 min. The final PCR products were extracted from an agarose gel (1.5% w / v) using Qiagen Gel Extraction Kit (Qiagen, Hilden, Germany). The final DNA concentration and quality were assessed using a BioPhotometer D30 UV spectrophotometer (Eppendorf, Hamburg, Germany). The amplicons were sequenced using a NovaSeq6000 (IGA Technology Services, Udine, Italy), yielding paired-end reads of 150 bp. Raw data were processed using MatLab scripts that have been adapted to the need of this study (Rebollo et al.2014). SEQ ID NO codesequence(5’- 3’)55 NGS-R1CCAGGCAGTCCGAGAGGGTGAGGATGTTTGAGCGTAATCTGG56 NGS-F1 GGAAGAAGGTGTTCAATTGGACAAGAGAGAAGC 57 NGS-R22GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGCCAGGCAGTCCGAGAGGG58 NGS-F2-ATCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCTGCACATCGATGGAAGAAGGTGTTCAATTGGACAAGAG SEQ ID NO codesequence(5’- 3’)59 NGS-F2-BTCGTCGGCAGCGTCAGATGTGTATAAGAGACAGATGTCCAACTACGGAAGAAGGTGTTCAATTGGACAAGAG60 NGS-F2-CTCGTCGGCAGCGTCAGATGTGTATAAGAGACAGGCGTACACTCACGGAAGAAGGTGTTCAATTGGACAAGAG61 NGS-F2-DTCGTCGGCAGCGTCAGATGTGTATAAGAGACAGTCTGAGGATCCAGGAAGAAGGTGTTCAATTGGACAAGAG62 NGS-F2-ETCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTTACGATGTGGGAAGAAGGTGTTCAATTGGACAAGAG63 NGS-F2-FTCGTCGGCAGCGTCAGATGTGTATAAGAGACAGAGTCAAGATGTTGGAAGAAGGTGTTCAATTGGACAAGAG64 NGS-F2-GTCGTCGGCAGCGTCAGATGTGTATAAGAGACAGGACACTGATGGTGGAAGAAGGTGTTCAATTGGACAAGAG65 NGS-F2-HTCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCACAGAGATACTGGAAGAAGGTGTTCAATTGGACAAGAG66 NGS-F2-ITCGTCGGCAGCGTCAGATGTGTATAAGAGACAGTGCTTGGATTCGGGAAGAAGGTGTTCAATTGGACAAGAG67 NGS-F2-LTCGTCGGCAGCGTCAGATGTGTATAAGAGACAGAGTGTACACTGCGGAAGAAGGTGTTCAATTGGACAAGAGTable 6. Primer sequences used for NGS analysis of yeast-encoded macrocyclic peptide libraries. 7. Generation of yeast libraries with (X)nC(X)mC(X)t structure fused to the C-terminus of a cell surface anchor protein (CSP) To test the procedure for the generation of cysteine-rich macrocyclic peptide libraries fused to the C-terminus of a cell-surface anchor protein (CSP), a library with a (X)nC(X)mC(X)tstructure was designed and constructed. The example given here refers to the CSP protein Aga2, but can be applied to any CSP with a solvent-exposed extracellular C-terminus. The coding vector for the complex expressed on the yeast surface is composed as follows: an inducible promoter, the alpha-mating factor secretion signal, a restriction enzyme, the sequence encoding for the CSP protein Aga2 with a restriction enzyme within the sequence, the influenza haemagglutinin epitope tag (HA-tag;NYPYDVPDYAC), a spacer (G4S)3, and the macrocyclic peptide sequence. Library diversities were calculated as previously reported in literature (Angelini et al. 2015). The determined library diversity per single electroporation is in the range of 107to 5 ^ 107. Yeast strain and media preparation The yeast strain used is this specific study has the following genotype: MATa AGA1::GAL1AGA1::URA3 ura352 trp1 leu2 delta200 his3 delta200 pep4::HIS3 prbd1.6R can1 GAL. The media for yeast growth and induction were described by Angelini et al. All media were supplemented with Penicillin / Streptomycin (Corning 100X Penicillin solution: Streptomycin). Libraries generation The backbone used for library construction was generated by inserting a (G4S)3linker via the Gibson assembly protocol. The gene was optimised for expression in yeast cells by GeneArt Gene Synthesis. Cysteine-rich macrocyclic peptide sequences were incorporated downstream of the cell-surface anchor protein Aga2 via a PCR reaction using the pCT-CSP backbone as template. The DNA encoding the random cysteine-rich macrocyclic peptide sequences was inserted by means of a reverse primer that, starting from the 5'-terminus, includes: the region of paired homology on the plasmid backbone, the random macrocyclic peptide sequence and the pairing region at the (G4S)3linker (common to the pairing region of libraries based on macrocyclic peptides fused to the N-terminus of binding proteins). The primers used to construct the libraries were obtained from Integrated DNA Technologies and are shown in Table 7. The insert PCRs were performed in a 50 µL reaction volume containing a constant LB-F forward primer (500 nM), a degenerate reverse primer R (500 nM), a mixture of dNTPs (200 µM each), 1 ng of pCT-CSP template DNA, 10X DreamTaq buffer (1X), 1.25 U of DreamTaq DNA polymerase (Thermo Fisher) and H2O milli-Q. The PCR programme was set up as follows: i) initial denaturation at 95 °C for 3 min, ii) 30 cycles of 30 sec at 95 °C, 30 sec at 55 °C, 1 min at 72 °C and iii) final elongation at 72 °C for 10 min. The vector backbone was linearised by triple digestion with EcoRI-HF (New England Lab), BamHI-HF (New England Lab) and NheI-HF (New England Lab). Digestion was performed in a 50 µL reaction volume containing 10 µg of plasmid pCT-CSP, 1 U of EcoRI-HF, 1 U of NheI-HF, 1 U of BamH- HF, 5X Cut Smart and H2O milli-Q. The sample was incubated for 24 hrs at 37 °C in a thermocycler. The linearised plasmid and PCR product were purified and concentrated to approximately 1 µg / µL by ethanol precipitation. The size and quality of the PCR inserts and the linearised plasmid were verified on a 1.2% w / v agarose gel. Primer SEQ ID NO code function sequence (5’- 3’) TTTAACGTC 68 LB-F libraryAAGGAGAAAAAACCC generationCGGATCGTTTAACGTCAAGGAGAA AAAACCCCGGATCG GGAAGAAGGTGTTCAATTGGACAA 69 LB-R libraryGAGAGAAGCTGGTTCTGGTGAANN generationKNNKNNKNNKNNKNNKNNKNNKN NKNNKNNKNNKNNKGGTGGTGGT GGCTCTGGTGG Table 7. Primers used for library construction of cysteine-rich macrocyclic peptide sequence C(X)nC fused to the C-terminus of a cell surface anchor protein (CSP)

[0002] Reference Angelini, A., T.F. Chen, S. De Picciotto, N.J. Yang, A. Tzeng, M.S. Santos, J.A. Van Deventer, M.W. Traxlmayr, and K. Dane Wittrup. 2015. “Protein Engineering and Selection Using Yeast Surface Display.” Methods in Molecular Biology 1319: 3–36. https: / / doi.org / 10.1007 / 978-1-4939-2748-7_1. Bacon, Kaitlyn, Abigail Blain, Matthew Burroughs, Nikki McArthur, Balaji M Rao, and Stefano Menegatti.2020. “Isolation of Chemically Cyclized Peptide Binders Using Yeast Surface Display.” ACS Combinatorial Science 22 (10): 519–32. https: / / doi.org / 10.1021 / acscombsci.0c00076. Boder, E T, and K D Wittrup.1997. “Yeast Surface Display for Screening Combinatorial Polypeptide Libraries.” Nature Biotechnology 15 (6): 553–57. https: / / doi.org / 10.1038 / nbt0697-553. Bowen, John, John Schneible, Kaitlyn Bacon, Collin Labar, Stefano Menegatti, and Balaji M Rao. 2021. “Screening of Yeast Display Libraries of Enzymatically Treated Peptides to Discover Macrocyclic Peptide Ligands.” International Journal of Molecular Sciences 22 (4): 1634. https: / / doi.org / 10.3390 / ijms22041634. Chen, Shiyu, Inmaculada Rentero Rebollo, Sergey A. Buth, Julia Morales-Sanfrutos, Jeremy Touati, Petr G. Leiman, and Christian Heinis. 2013. “Bicyclic Peptide Ligands Pulled out of Cysteine-Rich Peptide Libraries.” Journal of the American Chemical Society 135 (17): 6562–69. https: / / doi.org / 10.1021 / ja400461h. Cherf, Gerald M, and Jennifer R Cochran.2015. “Applications of Yeast Surface Display for Protein Engineering.” Methods in Molecular Biology (Clifton, N.J.) 1319: 155–75. https: / / doi.org / 10.1007 / 978-1-4939-2748-7_8. Daly, Rachel, and Milton T W Hearn. 2006. “Expression of the Human Activin Type I and II Receptor Extracellular Domains in Pichia Pastoris.” Protein Expression and Purification 46 (2): 456–67. https: / / doi.org / 10.1016 / j.pep.2005.10.001. Damasceno, Leonardo M, Kyle A Anderson, Gerd Ritter, James M Cregg, Lloyd J Old, and Carl A Batt. 2007. “Cooverexpression of Chaperones for Enhanced Secretion of a Single-Chain Antibody Fragment in Pichia Pastoris.” Applied Microbiology and Biotechnology 74 (2): 381–89. https: / / doi.org / 10.1007 / s00253-006-0652-7. Deventer, James A. Van, and K. Dane Wittrup. 2014. “Yeast Surface Display for Antibody Isolation: Library Construction, Library Screening, and Affinity Maturation.” Methods in Molecular Biology (Clifton, N.J.) 1060: 277–95. https: / / doi.org / 10.1007 / 978- 1-62703-992-5_10. Deyle, Kaycie, Xu-Dong Kong, and Christian Heinis. 2017. “Phage Selection of Cyclic Peptides for Application in Research and Drug Development.” Accounts of Chemical Research 50 (8): 1866–74. https: / / doi.org / 10.1021 / acs.accounts.7b00184. Dotter, Hanna, Melanie Boll, Matthias Eder, and Ann Christin Eder.2021. “Library and Post-Translational Modifications of Peptide-Based Display Systems.” Biotechnology Advances 47 (January): 107699. https: / / doi.org / 10.1016 / j.biotechadv.2021.107699. Felgenhauer, M, J Kohl, and F Rüker. 1990. “Nucleotide Sequences of the CDNAs Encoding the V-Regions of H- and L-Chains of a Human Monoclonal Antibody Specific to HIV-1-Gp41.” Nucleic Acids Research 18 (16): 4927. https: / / doi.org / 10.1093 / nar / 18.16.4927. Fetse, John, Sashi Kandel, Umar-Farouk Mamani, and Kun Cheng. 2023. “Recent Advances in the Development of Therapeutic Peptides.” Trends in Pharmacological Sciences 44 (7): 425–41. https: / / doi.org / 10.1016 / j.tips.2023.04.003. Gellissen, Gerd, Gotthard Kunze, Claude Gaillardin, James M Cregg, Enrico Berardi, Marten Veenhuis, and Ida van der Klei.2005. “New Yeast Expression Platforms Based on Methylotrophic Hansenula Polymorpha and Pichia Pastoris and on Dimorphic Arxula Adeninivorans and Yarrowia Lipolytica - a Comparison.” FEMS Yeast Research 5 (11): 1079–96. https: / / doi.org / 10.1016 / j.femsyr.2005.06.004. Hashimoto, Y, N Koyabu, and T Imoto. 1998. “Effects of Signal Sequences on the Secretion of Hen Lysozyme by Yeast: Construction of Four Secretion Cassette Vectors.” Protein Engineering, Design and Selection 11 (2): 75–77. https: / / doi.org / 10.1093 / protein / 11.2.75. Ishii, Jun, Nobuo Yoshimoto, Kenji Tatematsu, Shun’ichi Kuroda, Chiaki Ogino, Hideki Fukuda, and Akihiko Kondo. 2012. “Cell Wall Trapping of Autocrine Peptides for Human G-Protein-Coupled Receptors on the Yeast Cell Surface.” PLOS ONE 7 (5): e37136. https: / / doi.org / https: / / doi.org / 10.1371 / journal.pone.0037136. Jaroszewicz, Weronika, Joanna Morcinek-Orłowska, Karolina Pierzynowska, Lidia Gaffke, and Grzegorz W^grzyn. 2022. “Phage Display and Other Peptide Display Technologies.” FEMS Microbiology Reviews 46 (2): 1–25. https: / / doi.org / 10.1093 / femsre / fuab052. Ji, Xinjian, Alexander L Nielsen, and Christian Heinis.2024. “Cyclic Peptides for Drug Development.” Angewandte Chemie International Edition 63 (3): e202308251. https: / / doi.org / https: / / doi.org / 10.1002 / anie.202308251. Kingwell, Katie. 2023. “Macrocycle Drugs Serve up New Opportunities.” Nature Reviews. Drug Discovery. England. https: / / doi.org / 10.1038 / d41573-023-00152-3. Koganesawa, Nozomi, Tomoyasu Aizawa, Kazuo Masaki, Atsushi Matsuura, Taisuke Nimori, Hisanori Bando, Keiichi Kawano, and Katsutoshi Nitta.2001. “Construction of an Expression System of Insect Lysozyme Lacking Thermal Stability: The Effect of Selection of Signal Sequence on Level of Expression in the Pichia Pastoris Expression System.” Protein Engineering, Design and Selection 14 (9): 705–10. https: / / doi.org / 10.1093 / protein / 14.9.705. Kong, Xu Dong, Vanessa Carle, Cristina Díaz-Perlas, Kaycie Butler, and Christian Heinis.2020. “Generation of a Large Peptide Phage Display Library by Self-Ligation of Whole-Plasmid PCR Product.” ACS Chemical Biology 15 (11): 2907–15. https: / / doi.org / 10.1021 / acschembio.0c00497. Li, Xinting, Timothy W. Craven, and Paul M. Levine. 2022. “Cyclic Peptide Screening Methods for Preclinical Drug Discovery.” Journal of Medicinal Chemistry 65 (18): 11913–26. https: / / doi.org / 10.1021 / acs.jmedchem.2c01077. Linciano, Sara, Stefano Pluda, Arianna Bacchin, and Alessandro Angelini. 2019. “Molecular Evolution of Peptides by Yeast Surface Display Technology.” MedChemComm 10 (9): 1569–80. https: / / doi.org / 10.1039 / c9md00252a. Listov, Dina, Casper A Goverde, Bruno E Correia, and Sarel Jacob Fleishman. 2024. “Opportunities and Challenges in Design and Optimization of Protein Function.” Nature Reviews Molecular Cell Biology 25 (8): 639–53. https: / / doi.org / 10.1038 / s41580-024- 00718-y. Lopez-Morales, Joanan, Rosario Vanella, Elizabeth A Appelt, Sarah Whillock, Alexandra M Paulk, Eric V Shusta, Benjamin J Hackel, Chang C Liu, and Michael A Nash. 2023. “Protein Engineering and High-Throughput Screening by Yeast Surface Display: Survey of Current Methods.” Small Science 3 (12): 2300095. https: / / doi.org / https: / / doi.org / 10.1002 / smsc.202300095. Lopez-Morales, Joanan, Rosario Vanella, Gordana Kovacevic, Mariana Sá Santos, and Michael A. Nash. 2023. “Titrating Avidity of Yeast-Displayed Proteins Using a Transcriptional Regulator.” ACS Synthetic Biology 12 (2): 419–31. https: / / doi.org / 10.1021 / acssynbio.2c00351. Mullard, Asher. 2023. “Drug Developers Back US$80 Million African Ancestry Genomics Initiative.” Nature Reviews. Drug Discovery. England. https: / / doi.org / 10.1038 / d41573-023-00172-z. Muttenthaler, Markus, Glenn F King, David J Adams, and Paul F Alewood. 2021. “Trends in Peptide Drug Discovery.” Nature Reviews Drug Discovery 20 (4): 309–25. https: / / doi.org / 10.1038 / s41573-020-00135-8. OKA, Chitoshi, Masao TANAKA, Michiro MURAKI, Kazuaki HARATA, Katsunori SUZUKI, and Yoshifumi JIGAMI. 1999. “Human Lysozyme Secretion Increased by Alpha-Factor Pro-Sequence in Pichia Pastoris.” Bioscience, Biotechnology, and Biochemistry 63 (11): 1977–83. https: / / doi.org / 10.1271 / bbb.63.1977. Pan, Xingjie, and Tanja Kortemme.2021. “Recent Advances in de Novo Protein Design: Principles, Methods, and Applications.” Journal of Biological Chemistry 296: 100558. https: / / doi.org / https: / / doi.org / 10.1016 / j.jbc.2021.100558. Passioura, Toby, Takayuki Katoh, Yuki Goto, and Hiroaki Suga.2014. “Selection-Based Discovery of Druglike Macrocyclic Peptides.” Annual Review of Biochemistry 83: 727– 52. https: / / doi.org / 10.1146 / annurev-biochem-060713-035456. Raemaekers, R J, L de Muro, J A Gatehouse, and A P Fordham-Skelton. 1999. “Functional Phytohemagglutinin (PHA) and Galanthus Nivalis Agglutinin (GNA) Expressed in Pichia Pastoris Correct N-Terminal Processing and Secretion of Heterologous Proteins Expressed Using the PHA-E Signal Peptide.” European Journal of Biochemistry 265 (1): 394–403. https: / / doi.org / 10.1046 / j.1432-1327.1999.00749.x. Rebollo, Inmaculada Rentero, Michal Sabisz, Vanessa Baeriswyl, and Christian Heinis. 2014. “Identification of Target-Binding Peptide Motifs by High-Throughput Sequencing of Phage-Selected Peptides.” Nucleic Acids Research 42 (22): e169. https: / / doi.org / 10.1093 / nar / gku940. Rosmalen, Martijn van, Brian M G Janssen, Natalie M Hendrikse, Ardjan J van der Linden, Pascal A Pieters, Dave Wanders, Tom F A de Greef, and Maarten Merkx.2017. “Affinity Maturation of a Cyclic Peptide Handle for Therapeutic Antibodies Using Deep Mutational Scanning.” Journal of Biological Chemistry 292 (4): 1477–89. https: / / doi.org / 10.1074 / jbc.M116.764225. Sharma, Komal, Krishna K. Sharma, Anku Sharma, and Rahul Jain. 2023. “Peptide- Based Drug Discovery: Current Status and Recent Advances.” Drug Discovery Today 28 (2): 103464. https: / / doi.org / 10.1016 / j.drudis.2022.103464. Sloth, Ane Beth, Babak Bakhshinejad, Camilla Stavnsbjerg, Maria Rossing, and Andreas Kjaer.2023. “Depth of Sequencing Plays a Determining Role in the Characterization of Phage Display Peptide Libraries by NGS.” International Journal of Molecular Sciences 24 (6): 5396. https: / / doi.org / 10.3390 / ijms24065396. Smith, George P, and Valery A Petrenko.1997. “Phage Display.” Chemical Reviews 97 (2): 391–410. https: / / doi.org / 10.1021 / cr960065d. Smith, Thomas P., Bhaskar Bhushan, Daniele Granata, Christian S. Kaas, Birgitte Andersen, Klaas W. Decoene, Qiansheng Ren, et al. 2023. “Identification and Engineering of Potent Cyclic Peptides with Selective or Promiscuous Binding through Biochemical Profiling and Bioinformatic Data Analysis.” RSC Chemical Biology 5 (1): 12–18. https: / / doi.org / 10.1039 / d3cb00168g. Stellwagen, Sarah D, Deborah A Sarkes, Bryn L Adams, Mia A Hunt, Rebecca L Renberg, Margaret M Hurley, and Dimitra N Stratis-Cullum.2019. “The next Generation of Biopanning: Next Gen Sequencing Improves Analysis of Bacterial Display Libraries.” BMC Biotechnology 19 (1): 100. https: / / doi.org / 10.1186 / s12896-019-0577-8. Teymennet-Ramírez, Karla V., Fernando Martínez-Morales, and María R. Trejo- Hernández. 2022. “Yeast Surface Display System: Strategies for Improvement and Biotechnological Applications.” Frontiers in Bioengineering and Biotechnology 9 (January): 1–10. https: / / doi.org / 10.3389 / fbioe.2021.794742. Vodnik, Miha, Urska Zager, Borut Strukelj, and Mojca Lunder. 2011. “Phage Display: Selecting Straws Instead of a Needle from a Haystack.” Molecules 16 (1): 790–817. https: / / doi.org / 10.3390 / molecules16010790. Wang, Hui, and Rihe Liu. 2011. “Advantages of MRNA Display Selections over Other Selection Techniques for Investigation of Protein-Protein Interactions.” Expert Review of Proteomics 8 (3): 335–46. https: / / doi.org / 10.1586 / epr.11.15. Zhang, Huiya, and Shiyu Chen.2022. “Cyclic Peptide Drugs Approved in the Last Two Decades (2001-2021).” RSC Chemical Biology 3 (1): 18–31. https: / / doi.org / 10.1039 / d1cb00154j.

Claims

CLAIMS 1. Yeast cell containing an expression vector encoding a fusion protein (I): (NH2- or COOH) – [MP] – [FL] – [IFT] – [CSP] – (COOH or NH2) (I) wherein: (i) (NH2or COOH) indicates the amino or carboxyl terminus of the cysteine-rich macrocyclic peptide [MP] sequence of interest and (COOH or NH2) indicates the respective carboxyl or amino terminus of a protein [CSP] expressed on the surface of yeast cells, wherein the first terminus is located outside the cell; (ii) [MP] indicates a macrocycle peptide (X)nC(X)mC(X)tsequence wherein the two cysteine residues ‘C’ form a disulfide bond; n, m, t are 0 or integers with n = 0-5; m = 3-21; t = 0-5 with the condition that n + m + t = 3-21; or [MP] indicates a macrocycle peptide (X)nC(X)mC(X)tC(X)y sequence wherein two cysteine residues ‘C’ form a disulfide bond; n, m, t and y are 0 or integers, with n = 0- 5; m = 0-22; t = 0-22; y = 0-5 with the conditions that n + m + t + y = 4-22; or [MP] indicates a macrocycle peptide (X)nC(X)mC(X)tC(X)jC(X)y wherein each pair of cysteine residues ‘C’ forms a disulfide bond; n, m, t, j, y are 0 or integers, with n = 0- 5, m = 0-21, t = 0-21, j = 0-21, y = 0-5 with the conditions that n + m + t + j + y = 4- 21.; ‘X’ denotes any natural (cAA) or non-natural (NcAA) amino acid, wherein: - the natural amino acid (cAA) is encoded by any of the 61 naturally occurring codons or it is a random amino acid encoded by a degenerate codon selected from NNN, NNK, NNS, NDT, DBK, NRT, NNC, NWW, RVK; DVT, NVT, VVC, RST, NTT, TDK, with the condition that at least one random amino acid is present in the sequence of the macrocyclic peptide of interest, preferably 20 to 50% of the amino acids in the sequence are random, more preferably > 50% of the amino acids are random; - the non-natural amino acid (NcAA) is encoded by a TAG, TAA or TGA stop codon present within the coding sequence for the peptide of interest in the expression vector; (iii) [FL] indicates a spacer sequence containing 1 to 25 amino acids, preferably selectedfrom the following sequences: (GGGGS)nwhere n ranges from 1 to 5; (G)nwhere n is an integer ranging from 1 to 25; (S)n where n is an integer ranging from 1 to 25; (P)n where n is an integer ranging from 1 to 25; (EAAAK)n where n is an integer ranging from 1 to 5; (A(EAAAK)4ALEA(EAAAK)4A)nwhere n is an integer ranging from 1 to 2; (XP)nwhere n is an integer ranging from 1 to 12 and ‘X’ can be any natural amino acid, preferably an alanine (A), a lysine (K) and a glutamic acid (E); (iv) [IFT] indicates a sequence suitable for the detection and / or isolation of yeast cells expressing the MP-CSP or CSP-MP fusion on cell surface; (v) [CSP] indicates a protein expressed on the surface of yeast cells that can be fused to the N- or C-terminus of the macrocyclic peptide [MP] sequence, containing: - at least one extracellular portion; - a transmembrane portion or a portion binding the membrane and / or wall that is covalently or non-covalently anchored to molecules of the membrane and / or wall.

2. Yeast cell according to claim 1, wherein said [CSP] is Aga1 protein; Aga2 protein; a member of the glycosylphosphotidylinositol (GPI) binding protein family, preferably Cwp1p, Cwp2p, Tip1p, Flop1p, Sed1p, YCR89, Tir1p; or a synthetic, non-naturally occurring protein that is capable of anchoring to the yeast membrane and / or wall in a stable manner via covalent or non-covalent interactions.

3. Yeast cell according to claim 1, wherein said [IFT] is selected from c-myc-tag (SEQ ID NO:1), HA-tag (SEQ ID NO:2), Flag-tag (SEQ ID NO:3), His-tag (SEQ ID NO: 4), Strep-tag (SEQ ID NO:5), TC-tag (SEQ ID NO:6), S-tag (SEQ ID NO:7), AviTag (SEQ ID NO:8), Strep-tag2 (SEQ ID NO:9).

4. Yeast cell according to claim 1, further containing one or more vectors encoding an orthogonal translation system (OTS) consisting of an aminoacyl-tRNA synthetase (aaRS) and the corresponding transfer RNA (tRNA), capable of inserting a non-natural amino acid at the TAG, TAA, TGA stop codons present in the expression vector encoding the fusion protein (I).

5. Expression vector encoding the fusion protein (I) as defined in claims 1-4.

6. Expression vector according to claim 5, comprising a secretion signal at position 5' relative to the fusion protein (I).

7. Expression vector according to claims 5-6, further comprising an inducible promoter functionally linked to the fusion protein (I).

8. A cell population consisting of a plurality of yeast cells, wherein each cell is as defined in claims 1-4 and expresses on its surface multiple copies of a unique macrocycle peptide sequence of interest whose amino acid sequence varies within the cell population.

9. A cell population according to claim 8, wherein the sequence diversity of the cysteine- rich macrocycle peptide of interest is at least 107, preferably >108, more preferably >109.

10. A macrocycle peptide library including a yeast cell population according to claims 8 and 9, wherein different cells express different macrocyclic peptides of interest, preferably a different macrocyclic peptide for each cell, wherein said macrocyclic peptides of interest are selected from the group consisting of (X)nC(X)mC(X)t, (X)nC(X)mC(X)tC(X)y, (X)nC(X)mC(X)tC(X)jC(X)y, wherein X, C, n, m, t, y, j are as defined in claim 1.

11. Use of a macrocyclic peptide library according to claim 10 for screening biologically active macrocyclic peptide ligands.

Images