Methods of protein engineering and functional screening

Abstract

The present invention relates to modifying effector proteins (e.g. E3 ligases) in order to change their activity (e.g. to induce a gain of function). The invention relates to methods of modifying effector proteins by inserting a peptide motif or library of peptide motifs into the effector protein, libraries, and methods of screening with the modified effector proteins to identify new protein function i.e. new protein-protein interactions. For example, the invention relates to a method of identifying a modified effector protein (e.g. E3 ligase) with an induced change in substrate recognition activity, identified by measuring a change to substrate binding activity and / or target protein degradation. For example, the invention enables identification of new E3 ligase: Protein of interest (POI) interfaces, which can be used for drug discovery and development e.g. of monovalent degrader drugs such as molecular glues.

Description

[0001] Methods of protein engineering and functional screening

[0002] TECHNICAL FIELD

[0003] This invention relates to modifying effector proteins in order to change their activity (e.g. to induce a gain of function). The invention relates to methods of modifying effector proteins, and methods of screening with the mutant effector proteins to identify new protein function i.e. new protein-protein interactions. For example, the invention relates to a method of identifying a modified effector protein (e.g. E3 ligase) with an induced change of activity. For example, the invention relates to modifying E3 ligases in order to change their substrate recognition activity e.g. to induce degradation of a target protein. For example, the invention enables identification of new E3 ligase: Protein of interest (POI) interfaces, which can be used for drug discovery and development e.g. of monovalent degrader drugs such as molecular glues.

[0004] BACKGROUND

[0005] Targeted protein degradation

[0006] Despite enormous efforts to advance traditional pharmacology approaches, more than three quarters of all human proteins remain beyond the reach of therapeutic development. The challenges of developing therapeutics against the ‘difficult to drug’ or ‘yet to be drugged’ proteome include less defined ligand-binding pockets, non-catalytic protein-protein interaction functional modes and less-investigated 3D structures. Targeted protein degradation is a new approach that could overcome this and other limitations, and thus represents a promising therapeutic strategy. Targeted protein degradation can be used to induce degradation of proteins commonly considered 'undruggable'. It can also be used to remove unwanted peptides and / or proteins that have accumulated due to abnormal proteolysis (e.g. as associated with specific diseases e.g. Alzheimer’s).

[0007] Proteolysis of proteins is crucial in regulation of cellular processes. One of the mechanisms of achieving proteolysis is via the ubiquitin proteolytic system. Proteins and / or peptides are targeted for degradation by adding ubiquitin to the target proteins and / or peptide which removes unwanted proteins.

[0008] Chimeric constructs (e.g. proteolysis targeting chimeras (PROTACs)) and monovalent molecular glue compounds enable recruitment of target proteins to ubiquitin protein ligases for degradation, removing them from cells and tissues to resolve the disease state. Molecular glues and bivalent degraders

[0009] Molecular glues are an important new class of small molecule drug that exhibit a therapeutic effect by enhancing the affinity between proteins, inducing new interactions or stabilising a molecular complex. Similar to bivalent degraders, like PROTACs, molecular glues are able to exploit the cell’s proteostasis mechanisms, and this method to induce close proximity between target and effector proteins has great potential in the treatment of diseases from cancer to neurodegeneration and beyond.

[0010] Development of novel bivalent degraders (e.g. PROTACs) follows a logical path and can be based on modular medicinal chemistry process. Monovalent degraders (such as molecular glues) are hard to rationally design, especially beyond the immunomodulatory imide drug (IMiD)-like molecules via Cereblon (CRBN). Properties of both modalities present major promise, but face technical challenges in the route for discovery. Monovalent degraders have a small footprint with more predictable translation and traditional medicinal chemistry features better known. The mechanism of action (MOA) relies on inducing the close proximity between E3 ligase and the substrate client often including direct neo-PPI (neo-protein-protein interaction). Industry reliance is on serendipitous screening or de novo prediction of novel ternary complexes ahead of drug design - this is slow and high risk.

[0011] Therefore, a platform to robustly screen for functional neo-substrate:E3 ligase interactions is needed to unblock molecular glue discovery. The inventors have developed methods of modifying E3 ligases by insertion of peptide motifs (considered as functional mimics of molecular glues) and methods to screen for their changed activity (i.e. gain of function), and thereby the identification of new induced and neomorphic E3 ligase: Protein of interest (POI) interfaces. The present application provides a systematic library-based approach. The approach combines insertional mutagenesis and screening for an induced change in activity (gain of function), optionally coupled with next generation sequencing to identify active mutation sequences. This approach has broad applicability to other effector proteins besides E3 ligases.

[0012] E3 ligase engineering

[0013] Recombinant E3 ligases have previously been described, e.g. in US2010015116AA entitled: ‘Designer Ubiquitin Ligases For Regulation Of Intracellular Pathogenic Proteins’. This application relates to a recombinant ubiquitin ligase molecule that includes a large, antibodybased toxin binding domain that has an affinity for an enzymatically active fragment of one or more toxins or toxin serotypes; and an E3-ligase domain that comprises an E3-ligase or polypeptide that facilitates E2-mediated ubiquitination of the toxin enzymatically active fragment. In the exemplified embodiment, the toxin binding domain is a non-cleavable SNARE polypeptide or fragment thereof that binds to the enzymatically active fragment of a botulinum neurotoxin (BoNT). The invention in US2010015116AA couples two large extant function proteins in a linear genetic fusion. These two features limit its application as peptidic mimics for rational molecular glues discovery. Moreover, there is no disclosure in US2010015116AA of a systematic screening platform for neo-substrate:E3 ligase interactions based on engineered E3 ligases.

[0014] Chen et al (Chen et al, (2022) Cell Death & Differentiation volume 29, pagesl 955-1969: Disease-associated KBTBD4 mutations in medulloblastoma elicit neomorphic ubiquitylation activity to promote CoREST degradation) describes a novel mechanism in cancer pathogenesis whereby an insertion mutation in the E3 ubiquitin ligase adaptor Kelch Repeat and BTB Domain Containing 4 (KBTBD4) drives its recognition of neo-substrates for degradation. They observe that KBTBD4 mutants promote the recruitment and ubiquitylation of the REST Corepressor (CoREST), which forms a complex to modulate chromatin accessibility and transcriptional programmes. The degradation of CoREST promoted by KBTBD4 mutation diverts epigenetic programmes inducing significant alterations in transcription to promote increased sternness of cancer cells. This paper discloses that a disease-associated in-frame insertion of three amino acids (i.e. ‘PRR’) in the Kelch domain can change the substrate binding activity of KBTBD4. They show that medulloblastoma- associated hotspot mutations in the Kelch domain of KBTBD4 promote the recruitment of CoREST as a neo-substrate for ubiquitylation and degradation. This may be the first prototypical example of neomorphic mutation in an E3 ubiquitin ligase, however this discovery is limited to a single neosubstrate (CoREST) with limited application in the design of new therapies. Moreover, there is no suggestion in the paper that KBTBD4 could be used to screen for neomorphic insertions in KBTBD4 or indeed any of the many other E3 ligases.

[0015] Winter et al (Winter et al, (2023) Nat Chem Biol. 2023 Mar; 19(3): 323-333: Functional E3 ligase hotspots and resistance mechanisms to small-molecule degraders) looks at how amino acid substitutions affect drug-induced formation of a ternary complex (between: i) drug i.e. small-molecule degraders, ii) E3 ligase and iii) substrate / POl) and thereby affect the degree of substrate / POl degradation by a particular degrader. In a mutational scanning approach to identify functional E3 ligase hotspots, they focus on CRBN and VHL and on mutating residues that are proximal to the degrader binding site. Some loss of function mutations are identified. They do not insert peptide motifs or look for any gain of function mutations that change the actual substrate / POl to be degraded. Scott et al (Scott et al, (2023) Molecular Cell, Volume 83, Issue 5, 770 - 786. e9: E3 ligase autoinhibition by C-degron mimicry maintains C-degron substrate fidelity) looks at oligomeric status of KLHDC2 in the presence or absence of endogenous substrates containing KLHDC2- specific degron motifs and proposes that KLHDC2 possesses auto inhibitory tetramers via its C-terminus mimicking a KLHDC2 degron. This work demonstrates that bona fide peptidic degrons can induce disassembly of auto inhibitory KLHDC2 tetramers and identified a KLHDC2 mutant (-KK) which renders KLHDC2 monomeric. However, there is no disclosure in Scott et al of any gain of neosubstrates by KLHDC2 induced by insertion of peptidic sequences within the coding region.

[0016] SUMMARY OF INVENTION

[0017] The present inventors have developed a method of engineering proteins to change their activity, and a method of screening for the change in activity. For example, the invention relates to engineering effector proteins which have existing known and defined functional activities, such as enzyme functions, post-translational modifications or other defined biological controls, in order to change their activity towards a specific biological outcome (for example changes to effector proteins in the class E3 ligase, deubiquitinase, kinase etc). Including to engineer the substrate recognition regions or domains of said effector proteins by peptide sequence or library insertion to change their client or substrate repertoire or activity. In some embodiments, the engineered effector protein is an E3 ligase and the modified biological activity resulting from the engineering of said effector protein is a change in or induction of degradation of a target protein. In some embodiments the engineered E3 ligase is cereblon (CRBN). In some embodiments the engineered E3 ligase is KLHDC2.

[0018] For example, the present invention relates to a systematic screening platform based on engineered E3 ligases comprising intra-protein insertions of peptide motifs, to identify neosubstrate: E3 ligase interactions. Identification of new E3 ligase: Protein of interest (POI) interfaces is useful for example for discovery and development of monovalent degrader drugs such as molecular glues.

[0019] Methods of screening

[0020] In a first aspect, the invention provides a method of identifying a modified effector protein having an induced change in activity, wherein the method comprises modifying the effector protein by inserting a peptide motif or library of peptide motifs into the effector protein and identifying a change in activity, wherein the activity is substrate recognition activity, identified by measuring a change to substrate binding activity and / or target protein degradation. Suitably the insertion is within or proximal to a substrate binding interface of the effector protein. By ‘proximal to’ we mean, up to 5 angstrom, up to 10 angstrom, or up to 20 angstrom. The present specification teaches how to determine suitable locations for peptide motif insertions in an effector protein for induced neomorphic function according to the invention. The present inventors have shown how to model the interaction interface between the effector protein and potential neosubstrates and identify flexible regions within and proximal to the protein-protein interface as a starting point. The most flexible, conserved and exposed loops will likely present the most opportunity to induce neomorphic interactions and these are therefore suitable locations for insertion of a peptide motif or library of peptide motifs into effector proteins. All of these features together can be used to determine the likely position of inserted peptides that could induce new protein-protein interactions. Examples are provided herein, using the effector proteins CRBN and KLHDC2, and show successful gain of function using this positional selection criteria for peptide motif insertions.

[0021] In some embodiments, the effector protein is an E3 ligase. For example, the invention provides a method of identifying an E3 ligase having an induced change in activity, wherein the method comprises modifying the effector protein by inserting a peptide motif or library of peptide motifs into the effector protein and identifying a change in activity, wherein the activity is substrate recognition activity, identified by measuring a change to substrate binding activity and / or target protein degradation. Suitably the insertion is within or proximal to a substrate binding interface of the E3 ligase. Substrate binding interfaces of E3 ligases are discussed in Hanzl et al 2023 Nat Chem Biol. 2023 Mar;19(3):323-333 and Pallavi et al 2022, ACS Cent Sci. 2022 Apr 27; 8(4): 417-429.

[0022] The substrate is a protein of interest (POI), and maybe a target protein for degradation (or an accessory protein that binds to a target protein for degradation). A peptide motif is a peptide sequence of 1 or more amino acids. Suitably, each peptide motif comprises a polypeptide of length 1 to 110 amino acids.

[0023] Suitably, measuring a change to substrate binding activity or target protein degradation is by biochemical assays (e.g. immunoprecipitation), protein abundance (e.g. by flow cytometry), cell survival, fitness or measurement of biomarkers that represent different cellular states (e.g. biomarkers associated with the activity of the E3 ligase (e.g. protection from CSN5i-3- mediated degradation)).

[0024] By ‘inserting within or proximal to a substrate-binding interface’ of the effector protein e.g. an E3 ligase, we mean near to the substrate binding interface of an effector protein e.g. an E3 ligase. The substrate-binding interface is the domain(s) of an effector protein e.g. E3 ligase in which there is a protein-protein interaction between the effector protein (e.g. E3 ligase protein) and the substrate.

[0025] An example of an E3 ligase is Cereblon (CRBN). In some embodiments, there is a method of identifying a modified CRBN with an induced change of activity, wherein the method comprises modifying the CRBN by inserting a peptide motif or library of peptide motifs into CRBN and identifying a change in activity, wherein the change in activity is a change in substrate recognition activity, identified by measuring a change to substrate binding activity and / or target protein degradation.

[0026] Another example of an E3 ligase is KLHDC2. In some embodiments, there is a method of identifying a modified KLHDC2 with an induced change of activity, wherein the method comprises modifying the KLHDC2 by inserting a peptide motif or library of peptide motifs into KLHDC2 and identifying a change in activity, wherein the change in activity is a change in substrate recognition activity, identified by measuring a change to substrate binding activity and / or target protein degradation.

[0027] This insertional mutagenesis and screening approach could be applied to any effector protein, including an E3 ligase or other protein classes besides E3 ligases: e.g. deubiquitinases, chaperones, kinases, phosphatases, transcription factors, and other enzymes causing post- translational modifications to proteins. In any aspect of the invention, a peptide motif or library of peptide motifs can be inserted into an effector protein e.g. an E3 ligase or protein other than an E3 ligase, i.e. inserted into effector proteins such as: deubiquitinases, chaperones, kinases, phosphatases, transcription factors and other enzymes causing post-translational modifications to proteins. Suitably, the insertion is within or proximal to a substrate binding interface. A suitable change in activity (e.g. change in substrate binding activity or downstream effect e.g. target protein degradation) can then be screened for and active mutants and insertion sequences identified.

[0028] For example: A method of identifying a modified deubiquitinase having an induced change in activity, wherein the method comprises inserting a peptide motif or library of peptide motifs into a deubiquitinase protein and identifying a change in activity, wherein the activity is substrate recognition activity, including to induce stabilisation of a target protein. Suitably, the change in substrate recognition activity of the deubiquitinase is identified by measuring a change to substrate binding activity, or target protein stabilisation. Suitably, the insertion is within or proximal to a substrate binding interface of the deubiquitinase.

[0029] The term ‘active mutant’ is used to mean an active insertion mutant, i.e. an effector protein modified to comprise a peptide motif insertion, and which has been identified as having a change of activity as a result of that insertion. The active mutant is a modified effector protein identified as having an induced change in activity. In some embodiments, the effector protein, e.g. an E3 ligase, may also have been modified in other ways e.g. optimised to make the effector protein more suitable for the screening methods. For example, codon optimised, e.g. CRBN-codon optimised (e.g. CRBN-201 optimised according to SEQ ID NOs 34 and 35). For example, KLHDC2-KK mutant is a modified version of KLHDC2, which lacks the native protein’s C-terminal degron motif and which prevents it from self-oligomerising and so is well suited to substrate binding (Scott et al, (2023) Molecular Cell, Volume 83, Issue 5, 770 - 786. e9: E3 ligase autoinhibition by C-degron mimicry maintains C-degron substrate fidelity). Only some KLHDC2-KK mutants will be identified as active insertion mutants. Other E3 ligases / effector proteins may be similarly suitably modified in ways beyond the peptide motif insertions e.g. to facilitate the screening methods.

[0030] In some embodiments of the first aspect of the invention, the method comprises inserting a library of peptide motifs into a plurality of E3 ligase proteins, creating a library of E3 ligase mutants, wherein the plurality of E3 ligase proteins comprises one or more types of E3 ligase protein (e.g. CRBN, VHL, KBTBD4, KLHDC2), and the insertion is at one or more sites in the E3 ligase protein. Suitably, the library of E3 ligase mutants comprises the same E3 ligase protein with lots of different peptide motif insertions at one site. Alternatively, the library may comprise several different E3 ligase proteins each containing one or more peptide motif insertions at one or more sites (in a multiplex approach). One or more of these E3 ligase mutants may be identified as active mutants by a method according to the invention.

[0031] In some embodiments, the E3 ligase protein is further modified to optimise it for the method. For example, the KLHDC2-KK mutant is a mutant version of KLHDC2 which prevents itself oligomerising so it can be used for recruitment of a substrate.

[0032] In some embodiments, the change in substrate recognition activity is a change in the interaction of the E3 ligase and a target protein, optionally via an intermediate accessory protein, by measuring target protein degradation. For example, a change in target protein degradation may be identified by measuring the change in E3 ligase:target protein interactions or the change in abundance of a defined protein of interest or a biomarker functionally linked to the protein of interest or by another phenotypic effect associated with loss of the protein, such as cell death.

[0033] In some embodiments, the change in substrate recognition activity of the E3 ligase induces degradation of a target protein. This may be an increase in degradation level or degradation of a new target protein. In some embodiments, the induced change in activity is new substrate binding by the modified protein either at the modified site or distal from the modified site. The precise site of insertion and composition of the insertion library will also dictate the optimal length of insertion which results in proper folding and maintaining of the native function of the enzyme. By new substrate binding, we mean the modified effector protein (e.g. E3 ligase) shows functional affinity for an alternative substrate. In this case, a new protein-protein interaction (PPI) is identified. The new PPI can be a new E3 ligase:Target protein interaction or a new E3 ligase:Accessory protein interaction. A new E3 ligase: POI interface may be identified.

[0034] In some embodiments, the induced change in activity is improved binding for one or more existing or known substrate proteins (e.g. GSPT1 , SALL4, CK1a, LISP1 , SELENOK, SELENOS, etc.).

[0035] In some embodiments, the induced change in activity is new and / or improved substrate binding activity in the presence of one or more small molecule ligands or drugs (e.g. known monovalent degraders or molecular glues e.g. pomalidamide, IMiDs).

[0036] In some embodiments, where the E3 ligase is KLHDC2 or KLHDC2-KK, the induced change in activity is measured by a change to substrate binding activity, target protein degradation and / or phenotypic effects, for example, changes of cell proliferation. Suitably, some embodiments of the method use KLHDC2 ligands in cooperation with the peptide insertions to cause degradation and / or phenotypic effects. In some embodiments, the method of screening comprises a cell assay, and wherein the peptide motif insertion results in cellular phenotypic response such as cell death, caused by an unanticipated binding and degradation of a new substrate protein, or neosubstrate.

[0037] In some embodiments, the one or more peptide motifs are inserted into an internal domain of the protein and not at the N or C terminal. By not at the N or C-terminal, we mean up to 5 angstrom, up to 10 angstrom, up to 20 angstrom or up to 30 angstrom from the C or N terminal. Preferably, by not at the N or C-terminal we mean up to 10 angstrom from the C or N terminal. That is an in-frame intra-protein insertion. The peptide motif may occupy a pocket in the E3 ligase or the peptide motif may be placed at an interface between the E3 ligase and a known substrate. The latter might tolerate insertion of a longer peptide motif (e.g. 10 or more amino acids). For example, the peptide motif may be a loop presented on the surface of the E3 ligase, presenting a new neomorphic interface. Some insertions will be more buried / exposed than others and this would also be determined by the sequence and length of the insertion.

[0038] In some embodiments, the peptide motif comprises a polypeptide of length 1 to 110 amino acids. In some embodiments the peptide motif comprises a polypeptide of length less than 100 amino acids. In some embodiments, the peptide motif comprises a polypeptide length of 3 to 15 or 4 to 15 amino acids. In some embodiments, the peptide motif is an insertion of 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14 or 15 amino acids. In some embodiments, the peptide motif comprises a polypeptide length of at least 1 or at least 2 or at least 3 or at least 4 or at least 5 or at least 6 or at least 7 or at least 8 or at least 9 or at least 10 or at least 11 or at least 12 or at least 13 or at least 14 or at least 15 amino acids. In some embodiments, the peptide motif comprises a polypeptide length of 3 to 110 or 4 to 110 or 6 to 110 or 9 to 110 or 10 to 110 or 11 to 110 or 15 to 110 or 3 to 15 or 4 to 15 or 6 to 15 or 9 to 15 or 10 to 15 or 11 to 15 or 3 to 9 or 4 to 9 or 6 to 9 or 3 to 10 or 4 to 10 or 6 to 10 or 3 to 11 or 4 to 11 or 6 to 11 or 10 to 11 amino acids.

[0039] In some embodiments, the target protein is selected from the list of target proteins in Table 1 or Table 2 or the target protein is associated with a disease, disorder or condition when mutated (e.g. to cause cell or organismal death or cancer), expressed or over-expressed in a eukaryotic cell. In other embodiments, the target protein is a therapeutic target protein or a gain of function target protein. Target proteins can be selected from human disease areas including but not limited to oncology, central nervous system (CNS), inflammation and many neurodegenerative diseases.

[0040] In some embodiments, the E3 ligase is cereblon (CRBN) (SEQ ID NO. 1). Suitably, the location of the insertion of the peptide motif is at region P1 (defined by codons 125 - 178 (SEQ ID NO. 2)) or at region P2 (defined by codons 328 - 379 (SEQ ID NO. 3)) in the wildtype cereblon protein. In some embodiments, the location of the insertion of the peptide motif at P1 is at a subregion defined by codons 138-162 (SEQ ID NO. 4) or codons 148-152 (SEQ ID NO. 5) (for example, the peptide motif insertion replaces codons 149 -151 or the peptide motif is inserted between codons 150 and 151) In some embodiments, the location of the insertion of the peptide motif at P2 is at a subregion defined by codons 351-354 (SEQ ID NO. 6) or 352- 353. In some embodiments, the peptide motif is inserted at up to 5 angstroms or up to 10 angstroms from a said specific P1 or P2 region / subregion (a said specific codon range) in the wildtype cereblon protein. In some embodiments, the peptide motif is inserted into a region defined by a codon range which has an end point up to 1 , 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids away from an endpoint of said codon ranges.

[0041] In some embodiments, the E3 ligase is KBTBD4 (SEQ ID NO. 7). The location of the insertion is suitably in the Kelch domain of the protein KBTBD4. The location of the insertion can be in the KBTBD4 Kelch motif region defined by codons 308-313 (SEQ ID NO. 8). Suitably, the peptide motif inserted into this specific location of KBTBD4 comprises a polypeptide length of 1 to 110 amino acids. In some embodiments, the E3 ligase is KLHDC2 (SEQ ID NO. 9) (or a related mutant e.g. KLHDC2-KK (SEQ ID NO. 10)). The location of the insertion is suitably in the Kelch domain (amino acids 42 to 376 (SEQ ID NO. 11)) of the wild-type protein KLHDC2. Suitably, the location of the insertion of the peptide motif is at Loop A, B, C, D, E or F of the wild-type protein KLHDC2, identified and named by the present inventors as suitable locations for insertion. Suitably, the location of the insertion of the peptide motif is within Loop A (Loop A is defined by codons 46-66 (SEQ ID NO. 12)) or within Loop B (Loop B is defined by codons 105-117 (SEQ ID NO. 13)) or within Loop C (Loop C is defined by codons 159-194 (SEQ ID NO. 14)) or within Loop D (Loop D is defined by codons 232-244 (SEQ ID NO. 15)) or within Loop E (Loop E is defined by codons 283-296 (SEQ ID NO. 16)) or within Loop F (Loop F is defined by codons 334-353 (SEQ ID NO. 17)) in the wildtype KLHDC2 protein. For example, the peptide motif is suitably inserted centrally within loop A, B, C, D, E or F. In some embodiments, a peptide motif is inserted in Loop A at codons 52-58 (SEQ ID NO. 18) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop B at codons IOS- 111 (SEQ ID NO. 19) or up to 5 angstroms or up to 10 angstroms of that region, or a peptide motif is inserted in Loop C at codons 177-186 (SEQ ID NO. 20) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop D at codons 236-239 (SEQ ID NO. 21) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop E at codons 289- 292 (SEQ ID NO. 22) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop F at codons 343-346 (SEQ ID NO. 23) or up to 5 angstroms or up to 10 angstroms from that region. In some embodiments, a peptide motif insertion within Loop A replaces codons 52-58, and a peptide motif insertion within Loop B replaces codons 108-111 , and a peptide motif insertion within Loop C replaces codons 177-186, and a peptide motif insertion within Loop D replaces codons 236-239, and a peptide motif insertion within Loop E replaces codons 289- 292, and a peptide motif insertion within Loop F replaces codons 343-346. Suitably, the peptide motif inserted into these specific locations of KLHDC2 comprises a polypeptide length of 1 to 110 amino acids. Optionally, wherein the peptide motif inserted is 3-15 amino acids long, e.g. 11 amino acids long. For example, Figure 18C and 18D show HiBiT (VSGWRLFKKIS) (SEQ ID NO. 24), which is 11 amino acids long, inserted into KLHDC2 centrally within Loops A (SEQ ID NO. 25), C (SEQ ID NO. 26), D (SEQ ID NO. 27), and F (SEQ ID NO. 28).

[0042] In some embodiments, phenotypic measurements are used to identify active mutant proteins and / or active mutant sequences (i.e. peptide and / or nucleic acid sequences). This provides information about new protein-protein interactions e.g. identification of an E3:POI interface for monovalent drug development. The identification of new and induced, functional protein: protein interactions as described in the present application provides the required starting information for drug design, since novel induced molecular pockets are presented into which small molecule drug can be docked or designed. Suitably the phenotypic measurements for functional events are protein abundance, cell survival, cell fitness (e.g. cell abundance “dropout” screening), or measurement of biomarkers that represent different cellular states.

[0043] In some embodiments, the method comprises screening a population of mammalian cells comprising a library of effector proteins (e.g. E3 ligase proteins) with peptide motif insertions. In some embodiments of the method wherein a modified effector protein (e.g.E3 ligase protein) with an induced change of activity is identified, a phenotypic measurement and next generation sequencing is used to identify an active peptide motif insertion mutation sequence (hit). By ‘active peptide motif insertion mutation sequence’ / ‘active insertion mutation sequence’, we mean a peptide insertion sequence of an identified active mutant, i.e. causing the induced change in activity. Suitably, the method comprises; a. exposing a population of in vitro cultured mammalian cells capable of displaying said phenotype to a library of modified effector proteins (e.g. E3 ligases) modified according to a method described herein, b. identifying in said cell population an alteration in said phenotype following said exposure, c. selection of said cells undergoing the phenotypic change, to obtain a harvested cell population (for example by cell harvesting or fluorescence associated cell sorting), d. sequencing of said harvested cell population by next generation sequencing to identify an enriched or depleted insertion (hit) in said harvested population to identify an active insertion mutation sequence.

[0044] In some embodiments of the method, a modified CRBN having an induced change in activity is identified. Suitably, next generation sequencing (NGS) is used to identify active peptide motif insertion mutation sequences (hits) in a phenotypic measurement. Suitably, the method of screening comprises; a. exposing a population of in vitro cultured mammalian cells capable of displaying said phenotype to a library of modified CRBN according to a method described herein, b. identifying in said cell population an alteration in said phenotype following said exposure, c. selection of said cells undergoing the phenotypic change, to obtain a harvested cell population (for example by cell harvesting or fluorescence associated cell sorting), d. sequencing of said harvested cell population by next generation sequencing to identify an enriched or depleted insertion (hit) in said harvested population to identify an active insertion mutation sequence.

[0045] Suitably, this method comprises a library of modified cereblon (CRBN) proteins, each modified protein comprising a peptide motif at region P1 (defined by codons 125 - 178 (SEQ ID NO. 2)) or at region P2 (defined by codons 328 - 379 (SEQ ID NO. 3)) of the wildtype cereblon protein, and active peptide motif insertions (hits) are identified by the method. In some embodiments, the location of the insertion of the peptide motif at P1 is at a subregion defined by codons 138-162 (SEQ ID NO. 4) or codons 148-152 (SEQ ID NO. 5) (for example, the peptide motif insertion replaces codons 149 -151 or the peptide motif is inserted between codons 150 and 151). In some embodiments, the location of the insertion of the peptide motif at P2 is at a subregion defined by codons 351-354 (SEQ ID NO. 6) or 352-353. In some embodiments, the peptide motif is inserted at up to 5 angstroms or up to 10 angstroms from a said specific P1 or P2 region / subregion (a said specific codon range) in the wildtype cereblon protein. In some embodiments, the peptide motif is inserted into a region defined by a codon range which has an end point up to 1 , 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids away from an endpoint of said codon ranges.

[0046] Suitably, a high diversity CRBN insertion library is cloned at the P1 site or P2 site (e.g. at a P1 or P2 subregion). The P1 or P2 library is then inserted into a vector via recombination. A lentiviral vector encoding CRBN library is transduced into a mammalian cell line e.g. a CRBNk0cell line. Suitable vectors other than a lentiviral vector known in the art can be used to transduce or transfect mammalian cell lines. Library expressing cells are then selected and expanded. This library insertion approach can be applied to other insertion sites in other effector proteins.

[0047] Alternatively, this method suitably comprises a library of modified KLHDC2 I KLHDC2-KK proteins, each modified protein comprising a peptide motif at Loop A, B, C, D, E or F of the wildtype KLHDC2 protein (suitable insertion regions as defined herein), and active peptide motif insertions (hits) are identified by the method.

[0048] Suitably, a high diversity KLHDC2 insertion library is cloned at the A, B, C, D, E or F site. The A, B C, D, E or F library is then inserted into a vector via recombination. A lentiviral vector encoding KLHDC2 library is transduced into a mammalian cell line e.g. a KLHDC2' / _cell line. Suitable vectors other than a lentiviral vector known in the art can be used to transduce mammalian cell lines. Library expressing cells are then selected and expanded. This library insertion approach can be applied to other insertion sites in other effector proteins.

[0049] A peptide motif library according to the invention comprises more than one peptide motif. Suitably, a peptide motif library for insertion into an effector protein comprises at least 5,000 different amino acid sequences or nucleotide sequences encoding the amino acid sequences. Suitably, a peptide motif library in accordance with any aspect or embodiment of the invention comprises at least approximately 6,000, 10,000, 15,000, 20,000, 50,000, 100,000, 200,000, 250,000, 300,000, 475,000 or 500,000 distinct peptide motifs or distinct nucleotide sequences encoding the peptide motifs. In some embodiments, a library of 24,000 distinct nucleotide sequences is generated for insertion into CRBN e.g. at the P1 or P2 position. In some embodiments, a library of 150,000 sequences is inserted into CRBN e.g. at the P1 or P2 position. Suitably a peptide motif library for insertion comprises or encodes peptide motif amino acid sequences of the same or different lengths.

[0050] Similarly, a library of effector proteins modified by peptide insertion according to the invention (e.g. a library of mutant cereblon proteins, a library of mutant KLHDC2 proteins, or a library of mutant KBTBD4 proteins), suitably comprises at least approximately 5,000, 6,000, 10,000, 15,000, 20,000, 50,000, 100,000, 200,000, 250,000, 300,000, 475,000 or 500,000 modified effector proteins.

[0051] Other methods based on screening results

[0052] In another aspect, the invention provides a method of identifying an E3 ligase:POI interface suitable for monovalent degrader discovery / development, comprising a method of screening for a change in activity in a E3 ligase according to the first aspect of the invention. The identification of new protein-protein interaction is used to identify suitable protein interfaces to enable drug design (e.g. of a monovalent degrader or molecular glue)

[0053] New interfaces, active mutants and active peptide sequences identified can be used to design and screen for a monovalent degrader or molecular glue.

[0054] In another aspect, the invention provides a method of identifying a biological therapeutic agent that can reduce the amount of a target protein in a cell, the method comprising a method of screening according to the first aspect of the invention and identifying active mutants that induce target protein degradation for use as a therapeutic agent by contact with the cell; optionally wherein the therapeutic agent reduces the amount of target protein in a cell in the presence of a small molecule degrader. The invention also provides a biological therapeutic identified by such a method.

[0055] Engineering Effector proteins e.g. Cereblon or KLHDC2 In another aspect, the invention provides a method of changing the substrate recognition activity of an effector protein, the method comprising:

[0056] Making a mutant effector protein by inserting a peptide sequence of 1 to 110 amino acids into a location within or proximal to a substrate recognition interface of the wildtype effector protein; screening the mutant effector protein for a change in substrate recognition activity; optionally wherein the change induces degradation of a target protein; selecting an active mutant with changed activity.

[0057] Cereblon

[0058] In some embodiments, the effector protein is an E3 ligase protein e.g. cereblon. In some embodiments, this method comprises making a mutant cereblon protein by inserting a peptide sequence of 1 to 110 amino acids into the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wildtype cereblon protein. In some embodiments, the location of the insertion of the peptide motif at P1 is at a subregion defined by codons 138-162 (SEQ ID NO. 4) or codons 148-152 (SEQ ID NO. 5) (for example, the peptide motif insertion replaces codons 149 -151 or the peptide motif is inserted between codons 150 and 151). In some embodiments, the location of the insertion of the peptide motif at P2 is at a subregion defined by codons 351-354 (SEQ ID NO. 6) or 352- 353. In some embodiments, the peptide motif is inserted at up to 5 angstroms or up to 10 angstroms from a said specific P1 or P2 region / subregion (a said specific codon range) in the wildtype cereblon protein. In some embodiments, the peptide motif is inserted into a region defined by a codon range which has an end point up to 1 , 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids away from an endpoint of said codon ranges.

[0059] In another aspect, the invention provides a method of engineering cereblon to change its substrate recognition activity, wherein the method comprises inserting a peptide sequence of 1 to 110 amino acids at the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wildtype cereblon protein. In some embodiments, the location of the insertion of the peptide motif at P1 is at a subregion defined by codons 138-162 (SEQ ID NO. 4) or codons 148-152 (SEQ ID NO. 5) (for example, the peptide motif insertion replaces codons 149 -151 or the peptide motif is inserted between codons 150 and 151). In some embodiments, the location of the insertion of the peptide motif at P2 is at a subregion defined by codons 351-354 (SEQ ID NO. 6) or 352-353. In some embodiments, the peptide motif is inserted at up to 5 angstroms or up to 10 angstroms from a said specific P1 or P2 region / subregion (a said specific codon range) in the wildtype cereblon protein. In some embodiments, the peptide motif is inserted into a region defined by a codon range which has an end point up to 1 , 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids away from an endpoint of said codon ranges.

[0060] The change in substrate recognition activity may induce degradation of a target protein.

[0061] In another aspect, the invention provides a mutant cereblon protein, wherein the mutant protein comprises a peptide sequence of 1 to 110 amino acids inserted into the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wild-type protein cereblon. Insertional mutagenesis at either of these two sites maintains the E3 ligase stability and function or provides intra-E3 locations suitable for engineering new E3 ligase: POI interface for future molecular glue development. In some embodiments, the location of the insertion of the peptide motif at P1 is at a subregion defined by codons 138-162 (SEQ ID NO. 4) or codons 148-152 (SEQ ID NO. 5) (for example, the peptide motif insertion replaces codons 149 -151 or the peptide motif is inserted between codons 150 and 151). In some embodiments, the location of the insertion of the peptide motif at P2 is at a subregion defined by codons 351-354 (SEQ ID NO. 6) or 352-353. In some embodiments, the peptide motif is inserted at up to 5 angstroms or up to 10 angstroms from a said specific P1 or P2 region / subregion (a said specific codon range) in the wildtype cereblon protein. In some embodiments, the peptide motif is inserted into a region defined by a codon range which has an end point up to 1 , 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids away from an endpoint of said codon ranges.

[0062] In another aspect, the invention provides a library of mutant cereblon proteins as defined above. For example, a library of mutant cereblon proteins, wherein each mutant protein comprises a peptide sequence of 1 to 110 amino acids inserted into the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wild-type protein cereblon. In some embodiments, the location of the insertion of the peptide motif at P1 is at a subregion defined by codons 138-162 (SEQ ID NO. 4) or codons 148-152 (SEQ ID NO. 5) (for example, the peptide motif insertion replaces codons 149-151 or the peptide motif is inserted between codons 150 and 151). In some embodiments, the location of the insertion of the peptide motif at P2 is at a subregion defined by codons 351- 354 (SEQ ID NO. 6) or 352-353. In some embodiments, the peptide motif is inserted at up to 5 angstroms or up to 10 angstroms from a said specific P1 or P2 region / subregion (a said specific codon range) in the wildtype cereblon protein. In some embodiments, the peptide motif is inserted into a region defined by a codon range which has an end point up to 1 , 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids away from an endpoint of said codon ranges.

[0063] In another aspect, the invention provides a population of mammalian cells comprising a library of cereblon mutant proteins, wherein the library is as defined above. In another aspect, the invention provides a method of screening for a change in activity in a cereblon protein, wherein the method comprises using a population of mammalian cells comprising a library of cereblon mutant proteins as described herein or a library of cereblon mutant proteins as described herein. Suitably the change in activity is new cereblon substrate recognition activity (e.g. new or improved substrate binding), and the new substrate recognition activity induces target protein degradation, which can be measured either directly with protein abundance tools or by monitoring the cellular phenotypic effects that result from the change in function. The method enables active mutant cereblon proteins to be identified and isolated.

[0064] In another aspect, the invention provides a method of identifying an E3 ligase: Protein of Interest (POI) interface suitable for monovalent degrader discovery / development, comprising a method of screening as defined above wherein the identification of new protein-protein interaction is used to identify suitable protein interfaces to enable drug design.

[0065] KLHDC2

[0066] In some embodiments, the effector protein is an E3 ligase protein e.g. KLHDC2. In some embodiments, this method comprises making a mutant KLHDC2 protein by inserting a peptide sequence of 1 to 110 amino acids within Loop A, B, C, D, E or F of the wildtype KLHDC2 protein, for example centrally within these loops. Suitably, the location of the insertion of the peptide motif is within Loop A (Loop A is defined by codons 46-66 (SEQ ID NO. 12)) or within Loop B (Loop B is defined by codons 105-117 (SEQ ID NO. 13)) or within Loop C (Loop C is defined by codons 159-194 (SEQ ID NO. 14)) or within Loop D (Loop D is defined by codons 232-244 (SEQ ID NO. 15)) or within Loop E (Loop E is defined by codons 283-296 (SEQ ID NO. 16)) or within Loop F (Loop F is defined by codons 334-353 (SEQ ID NO. 17)) in the wildtype KLHDC2 protein. For example, the peptide motif is inserted centrally within loop A, B, C, D, E or F. In some embodiments, a peptide motif is inserted in Loop A at codons 52-58 (SEQ ID NO. 18) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop B at codons 108-111 (SEQ ID NO. 19) or up to 5 angstroms or up to 10 angstroms of that region, or a peptide motif is inserted in Loop C at codons 177-186 (SEQ ID NO. 20) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop D at codons 236-239 (SEQ ID NO. 21) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop E at codons 289- 292 (SEQ ID NO. 22) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop F at codons 343-346 (SEQ ID NO. 23) or up to 5 angstroms or up to 10 angstroms from that region. In some embodiments, a peptide motif insertion within Loop A replaces codons 52-58, and a peptide motif insertion within Loop B replaces codons 108-111 , and a peptide motif insertion within Loop C replaces codons 177-186, and a peptide motif insertion within Loop D replaces codons 236-239, and a peptide motif insertion within Loop E replaces codons 289- 292, and a peptide motif insertion within Loop F replaces codons 343- 346. Suitably, the peptide motif inserted into these specific locations of KLHDC2 comprises a polypeptide length of 1 to 110 amino acids. Optionally, wherein the peptide motif inserted is 3-15 amino acids long, e.g. 11 amino acids long. For example, Figures 18C and 18D show HiBiT (VSGWRLFKKIS) (SEQ ID NO. 24), which is 11 amino acids long, inserted into KLHDC2 centrally within Loops A (SEQ ID NO. 25), C (SEQ ID NO. 26), D (SEQ ID NO. 27), and F (SEQ ID NO. 28).

[0067] In another aspect, the invention provides a method of engineering KLHDC2 to change its substrate recognition activity, wherein the method comprises inserting a peptide sequence of 1 to 110 amino acids (e.g. centrally) within Loop A, B, C, D, E or F of the wildtype KLHDC2 protein, as defined above. The change in substrate recognition activity may induce degradation of a target protein.

[0068] In another aspect, the invention provides a mutant KLHDC2 protein, wherein the mutant protein comprises a peptide sequence of 1 to 110 amino acids inserted (e.g. centrally) within Loop A, B, C, D, E or F of the wildtype KLHDC2 protein, as defined above. Examples show that insertional mutagenesis at any of the four loop sites, A, C, D or F, maintains the E3 ligase stability and function or provides intra-E3 locations suitable for engineering new E3 ligase: POI interface for future molecular glue development.

[0069] In another aspect, the invention provides a library of mutant KLHDC2 proteins as defined above.

[0070] In another aspect, the invention provides a population of mammalian cells comprising a library of KLHDC2 mutant proteins, wherein the library is as defined above.

[0071] In another aspect, the invention provides a method of screening for a change in activity in a KLHDC2 protein, wherein the method comprises using a population of mammalian cells comprising a library of KLHDC2 mutant proteins as described herein or a library of KLHDC2 mutant proteins as described herein. Suitably the change in activity is new KLHDC2 substrate recognition activity (e.g. new or improved substrate binding), and the new substrate recognition activity induces target protein degradation, which can be measured either directly with protein abundance tools or by monitoring the cellular phenotypic effects that result from the change in function. The method enables active mutant KLHDC2 proteins to be identified and isolated.

[0072] In another aspect, the invention provides a method of identifying an E3 ligase: Protein of Interest (POI) interface suitable for monovalent degrader discovery / development, comprising a method of screening as defined above wherein the identification of new protein-protein interaction is used to identify suitable protein interfaces to enable drug design.

[0073] DETAILED DESCRIPTION

[0074] Certain aspects and embodiments of the invention will now be illustrated by way of example and with reference to the following Figures and Examples.

[0075] FIGURES

[0076] Figure 1. Expression and stability of the E3 ligase Cereblon (CRBN) with varying length intrasequence insertions. Fig. 1A: the expression level of a mutant insertion into CRBN is measured by flow cytometry staining, and compared to the wild-type form (CRBNm), either reconstituted in CRBNKOcells or the non-deleted wild-type cells. Fig. 1 B: western blotting to examine CRBN expression and stability from three different length insertions (4, 9 or 15 amino acids) in a defined position within the peptide sequence of CRBN.

[0077] Figure 2. The cereblon E3 ligase remains functionally active when remodelled with intramolecular peptide loop insertions. Fig. 2A is a graphic depiction of the remodelling of the E3 ligase cereblon, where oligopeptide motifs are introduced into the sequence of the protein, such that the loop is presented on the surface of the E3 ligase, presenting a new neomorphic interface. Fig. 2B is a second cartoon illustrating the expected effect of the dBET6 on BRD4. Fig. 2C are the results of an experiment in which a peptide motif sequence comprising the epitope tag “HA” was inserted into two positions, either P1 or P2 within CRBN and cells expressing these novel variants were then subsequently exposed to the bivalent degrader drug, dBET6, for 18 hours. BRD4 was then probed on the cell lysate membranes as well as CRBN and a control protein (GAPDH).

[0078] Figure 3. Demonstration that neomorphic insertion E3 ligase mutants can engage in new, stable protein-protein interactions consistent with a new function. Fig. 3A is a graphic of the experimental strategy. Intra-molecularly HA-inserted CRBN protein was co-expressed in CRBNk0cells with a GFP-tagged intracellular single-chain antibody harbouring specific affinity for the HA epitope tag (HA-15F11). Fig. 3B is the input data that shows the expression of the GFP-tagged nanobody on western blotting, and the various CRBN insertion mutants can also be stained with antibodies targeting CRBN. When the GFP-tagged nanobody is used as the target for immunoprecipitations (using a GFP-trap bead; IP panels), CRBN can then only be retrieved with high affinity following the introduction of the HA tag into sites P1 or P2, and not with wild-type (WT) or negative controls.

[0079] Figure 4. Confirmation of different physiological states found for P1 and P2 insertion sites. Fig. 4A: a graphic describing experimental setup. When cells are treated with the signalosome inhibitor CSN5i-3, the deNEDDylation process becomes inhibited, trapping activated cullin-ring ligases (CRL) complexes in the NEDDylated and activated form. Fig. 4B: different insertion site mutants of CRBN are tested for CRBN abundance after CSN5i-3 treatment. Fig. 4C: the results shown in Fig. 4B are corroborated, in the presence of HA- targeting nanobodies.

[0080] Figure 5. Demonstration of neomorphic screens of CRBN. Fig. 5A: cloning a high diversity CRBN insertion library at P1 site. Variable sequences at P1 site flanked by overlapping regions of a CRBN plasmid is amplified via PCR. The rest of a codon-optimised CRBN expression plasmid lacking P1 site is also amplified as vector. P1 library is then inserted into vector via recombination to replace wild type P1 sequence. Fig. 5B: illustration of a PROTEINi® screen. CRBN library from Fig. 5A is packaged into recombinant lentiviruses, and transduced into a mammalian cell line. Library-expressing cells are selected and expanded, with or without small molecule treatment. Cells are then harvested and immunostained if needed, and sorted using fluorescence-activated cell sorting. The relative distribution of each sequence of the library in each population is determined by next generation sequencing. Fig. 5C: library-specific amplification of CRBN ORF over endogenous coding sequences. After genomic DNA extraction, primers complimentary to library-encoded CRBN ORF flanking P1 site are used to amplify P1 amplicon. Amplification of endogenous CRBN ORF is prevented due to insufficient overlap with primer sequences.

[0081] Figure 6. Examples of data analysis for phenotypic screens. Figs. 6A-C: gain-of-function screens at P1 site. Fig. 6A: library-expressing cells are treated with Pomalidomide and sorted on POI1 intensity. POI1 is a substrate of CRBN_wt in the presence of CC-90009 but not Pomalidomide. Fig. 6B: library-expressing cells are treated with Pomalidomide and expanded. Cells harvested before and after Pomalidomide treatment are analysed to identify mutations depleted post Pomalidomide treatment. Treatment of CC-90009 lead to reduced proliferation of cells expressing CRBN_wt. Fig. 6C: correlation of Fig. 6A and Fig. 6B. POI1 degradation-related mutations should correlate with depletion and clustered at the top right of figure, similar to CRBN_wt treated with CC-90009. CRBN_wt treated with Pomalidomide results in no POI1 degradation nor depletion, located to the bottom left of the figure, together with most of the library. Mutations associated with degradation of neosubstrates other than PO11 , and leading to reduced proliferation, are clustered at the bottom right of the figure. Fig. 6D-F: gain-of-function screen at P2 site. Fig. 6D: library-expressing cells are stained with antibody against POI2 and sorted. siRNA targeting POI2 serves as a positive control. Fig. 6E: library-expressing cells are treated with signalosome inhibitor CSN5i-3 and stained with CRBN antibody. Cells are sorted on CRBN intensity to identify mutations associated with high CRBN intensity. A previously-characterised CRBN_P2 serves as non-catalytically active control (purple) against activated CRBN (CRBN_wt treated with Pomalidomide, red). Fig. 6F: correlation of Fig. 6D and Fig. 6E. Mutations associated with POI2 degradation should cluster at the top right of the figure.

[0082] Figure 7. Graphic describing the concept of using an MG-PROTEINi® library to screen for E3 ligases having an induced change of activity, and to model a hit-induced ternary complex to design a small molecule monovalent glue.

[0083] Figure 8. Schematic illustration of the proof of concept degradation by engineered CRBN. Fig. 8A illustrates successful degradation of an engineered CRBN. Fig. 8B shows the flow cytometry expression analysis of the CRBN cell lines with different integration positions of the HiBiT tag. Fig. 8C shows the comparison of the luminescence generated from the different CRBN-HiBiT variants. Fig.8D shows the comparison of the luminescence generated by stable cell variants after treatment with either a proteasome inhibitor (MG132), or an inhibitor of the cullin ligases (MLN4924) or DMSO (control)

[0084] Figure 9. Screen-scale library sequence integration into CRBN and detection by nextgeneration sequencing. Fig. 9A illustrates the library insertion for position P2 in CRBN. Fig. 9B illustrates a gel electrophoresis plot with different concentrations of DNA from cells that were subjected to PCR. Fig. 9C is a histogram showing the distribution of sequences found with different counts.

[0085] Figure 10. Molecular dynamics simulations of the insertion space for remodelling sequences in CRBN site P2. Fig. 10A. is an illustration of the CRBN loop insertion strategy of the invention. Fig. 10B. shows molecular dynamics simulation overlays and comparison to the CRBN:GSPT1 :CC885 crystal structure (PDB: 5HXB). Four alanine residues were modelled into the CRBN sequence and protein structural simulations (AlphaFold2) used to reconstitute the expected three-dimensional properties of the protein interaction followed by further refinement with molecular dynamics simulations. Fig. 10C. shows additional sequences were designed of various compositions and lengths and predictions were made for relative energetic properties of the interactions. Comparison to the endogenous protein interactions allowed a ddG calculation, where negative kcal / mol is indicative of improved binding affinity compared to the non-inserted sequence variant of CRBN.

[0086] Figure 11. Screen data for protein engineering and functional screening for neomorphic sequence changes in CRBN. Fig. 11 A shows flow cytometry measurements for three proteins. HAP1 cells were collected and subjected to staining with antibodies specific to the named proteins. Fig. 11 B shows waterfall plots for multiple different samples where thousands of library sequences have been identified by next-generation sequencing. Sequences are designed and synthesised by chip array synthesis before being cloned into a reference plasmid at the nominated location and analysed after cells have been transduced with the new sequences. Plots show cumulative library representation where measurement of over 1000 sequences for each library is preferred for confidence measurements. Fig. 11C shows the resulting statistical analysis data where each library sequence is measured in multiple screen samples or flow cytometry gated populations (e.g. high fluorescence or low fluorescence). DeSEQ2 analysis packages are used to calculate statistical significance and changes in abundance and hits are referenced on the plots for multiple different screening paradigms.

[0087] Figure 12. Schematic for screening. Cells are first transduced with a lentivirus pool containing all library variants as encoded sequences ahead of selection by antibiotic. Cell sorting by cytometry is then completed or cells are grown out for evaluation of cell fitness by abundance measurements. Deep sequencing identifies the active sequences in each population.

[0088] Figure 13. Demonstration of screening for induced neomorphic activity (degradation of GSPT1). Fig. 13A shows schematic for CRBN with intra-sequence insertions to induce new activity. Fig. 13B. A library was inserted into CRBN P1 site and screening completed for degradation of GSPT1, deep sequencing data was clustered for each of the samples using principle component analysis. II indicates unsorted cells; L indicates low expression population; H indicates high expression and P indicates the library plasmid before it is introduced into cells. Each with three independent replicates, clustered by similarity. In Fig. 13C, samples are compared using DeSEQ2 analysis of each individual sequence variant and plotted as an MA plot to identify active sequences. Control populations are marked and perform as expected and the indicated robust and deep hit space is compared in either cells treated with DMSO or pomalidomide. In Fig. 13D a Z-score analysis is used to aggregate sequences of similar or identical sequences by assimilation and comparison between samples. Hit space for active sequences is indicated in either DMSO or pom-treated samples.

[0089] Figure 14. Active sequence insertion analysis of physiochemical properties following a screen for GSPT 1 degrading variants of CRBN. In Fig. 14A the active degrading sequences for GSPT1 are compared by length of insert whereas in Fig. 14B the overall volume is computed for residues in the active insertions. In Fig. 14C the overall charge of the insert sequences is calculated and compared for each hit population, either in DMSO or pom- treated samples and in comparison to the entire library. In Fig. 14D an example of a cluster of very similar 3 residue hits is shown in an MA plot with the marked sequences.

[0090] Figure 15. Further screening approaches to identify neomorphic activity of CRBN. A library was inserted in CRBN P1 site and cells were stained for abundance of the protein target STAT3. Loss of STAT3 protein is indicative of degradation, and measured by flow cytometry. Low STAT3 staining cells were collected and deep sequenced. Active degrading sequences are found by deep sequencing to identify them. In Fig. 15A a volcano plot indicates the performance of the screen after analysis and the hit spaces are shown. In Fig. 15B a similar screen for total abundance of insert sequences over time is shown. If CRBN variants are lost over time in a controlled manner it is indicative of loss of the cells that house them, and therefore of cell death in the screen. These hits are shown in the noted space on the volcano plot. In Fig. 15C the cell lethal hit space is directly compared to the GSPT1 degraded hit space from other screens, where overlaps and unique degrading sequences are highlighted.

[0091] Figure 16. Functional re-constitution of KLHDC2 in HAP1 cells. Isogenic KLHDC2 knock-out HAP1 cells (HAP1KLHDC2-) were transduced with lentivirus encoding HA-tagged KLHDC2 (HA- KLHDC2) or -KK mutant (HA-KLHDC2-KK). Stable cell lines are then transduced with lentivirus encoding GFP reporters fused to different prospective KLHDC2 degrons. Fig. 16A shows reconstitution of KLHDC2 (WT or the -KK mutant) using flow cytometry and anti-HA staining. Figs. 16B-D shows functional validation of reconstituted HA-KLHDC2 and the -KK mutant using: (B) GFP fused to SelenoK degron on the C-terminus; (C) GFP fused to a mutant SelenoK degron where the last glycine residue is mutated to Leucine; and (D) GFP fused to a prospective degron sequence (Hit 528). In each experiment, reduction of normalised GFP / miRFP observed when treated with DMSO, and this is rescued by a proteasomal inhibitor (MG132) or a NEDDylation inhibitor (MLN4924). The observed GFP degradation by HA- KLHDC2 is dose-dependent, where HA-KLHDC2 (WT or KK mutant) introduced at high MOI (MOM) results in degradation similar to levels seen in parental cells expressing endogenous KLHDC2. Figure 17. Identification of potential regions in KLHDC2 substrate recognition motif to engineer insertions. Three examples are given to demonstrate the method: Fig. 17A shows key amino acid residues involved in KLHDC2 ligand binding. Fig. 17B shows predicted residues on KLHDC2 surface which could induce neo-PPIs. In Fig. 17C the solvent accessible, aromatic side-chains of the protein are shown.

[0092] Figure 18. Intra-KLHDC2 editing via modification of surface-exposed flexible regions. After selection exemplified in Figure 17, four potential regions (Loops A, C, D and F) within KLHDC2 kelch domain were selected to incorporate HiBiT coding sequences. Figs. 18A and 18B show the relevant loop sequences in relation to: (A) structure of the kelch domain and (B) coding sequence of KLHDC2 (SEQ ID NO. 9). Fig. 18C shows comparison between original (WT) sequences of Loop A (SEQ ID NO. 12) Loop C (SEQ ID NO. 14), Loop D (SEQ ID NO. 15) and Loop F (SEQ ID NO. 17) and engineered (with insertion) sequences (SEQ ID NOs. 25, 26, 27 & 28) with HiBiT sequence (SEQ ID NO. 24) underlined. Fig. 18D illustrates the predicted structures with HiBiT sequences incorporated.

[0093] Figure 19. Evaluation of engineered KLHDC2 mutants to induce neo-PPIs. Following insertion of the HiBiT sequence (SEQ ID NO. 24) into each of Loops A, C, D and F of KLHDC2, LgBiT is stably expressed in cells by lentivirus transduction. HiBiT is an engineered short peptidic sequence which compliments LgBiT to a functional luciferase. The induction of novel proteinprotein interactions is assessed by measuring luminescence from intracellularly assembled HiBiT and LgBiT. The luminescence from KLHDC2 mutants are compared to background (No HiBiT) and N-terminal insertion of HiBiT where little to no structural hinderance is expected (Poirson et al., Nature 2024 volume 628, pages 878-886: Proteome-scale discovery of protein degradation and stabilization effectors). Two sites (Loops A&D) show luminescence significantly higher than No HiBiT control, indicating neo-PPI enabled by HiBiT insertions. Comparison to MLN4924 inhibition indicate that LgBiT interaction via HiBiT insertions in Loops A&D are resistant to proteasomal degradation via CRL.

[0094] Figure 20. Evaluation of engineered KLHDC2 mutants to degrade a GFP-degron fusion reporter. Following HiBiT insertion in N-terminus and Loops A / C / D / F of KLHDC2, stable cells are transduced with GFP fused to SelenoK degron. New stable cells are treated with DMSO / MG132 / MLN4924 then GFP intensity measured by flow cytometry. This result demonstrates that compared to HAP1KLHDC2' cells, re-constituted KLHDC2 mutants with HiBiT insertions in all regions possess different ability to reduce GFP-SelK reporter in a proteasomal and CRL-dependent manner. Figure 21. Evaluation of engineered KLHDC2 mutants to degrade an endogenous target BRD4 via KLHDC2-dependent PROTACs in HTRF assay. Fig. 21A demonstrates the experiment set up where all stable KLHDC2 mutants-expressing cells display similar dose response on BRD4 degradation via a CRBN-dependent PROTAC dBET6. Figs. 21 B and 21 C show that HiBiT insertions at different regions have different impacts on BRD4 degradation with different KLHDC2-dependent PROTACs. Note that HiBiT in Loops D and F consistently degrade BRD4 better than HiBiT in A and C, and that N-terminal insertion of HiBiT has a negative impact on small molecule-mediated degradation of BRD4.

[0095] Figure 22. Molecular modelling of an induced interaction between GSPT1 and CRBN following the introduction and identification of a hit peptide sequence. Fig. 22A shows these two proteins in cartoon form modelled using molecular dynamics and protein structural prediction tools, and based on X-ray crystallography data. The CRBN has been modelled with an inserted 3 amino acid loop at position P1 which was found to robustly induce the degradation of GSPT1 in the presence of pomalidomide. The loop is modelled in space-filling representation. In Fig.22B the space occupied by the inserted peptide loop is represented by a pharmacophore to register the corresponding chemical properties of each residue. In Fig. 22C predictions of electrostatic interactions were made using conventional computational chemistry tools and are presented. Interactions from the inserted loop which forming hydrogen bonds with GSPT1 residues were identified. In Fig. 22D further property delineation was conducted including hydrophobicity, charge and special features and these are presented. In Fig. 22E a table of different insertion features (SEQ ID NOs. 29-32) is presented for P1 in CRBN for screening and for the induction of neomorphic interactions.

[0096] Definitions

[0097] Targeted protein degradation

[0098] Despite enormous efforts to advance traditional pharmacology approaches, more than three quarters of all human proteins remain beyond the reach of therapeutic development. Targeted protein degradation is a new approach that could overcome this and other limitations, and thus represents a promising therapeutic strategy. Targeted protein degradation can be used to induce degradation of proteins commonly considered 'undruggable'. It can also be used to remove unwanted peptides and / or proteins that have accumulated due to abnormal proteolysis (e.g. as associated with specific diseases e.g. Alzheimer’s).

[0099] Proteolysis of proteins is crucial in regulation of cellular processes. One of the mechanisms of achieving proteolysis is via the ubiquitin proteolytic system. Proteins and / or peptides are targeted for degradation by adding ubiquitin to the target proteins and / or peptide which removes unwanted proteins.

[0100] Chimeric constructs and molecular glue compounds enable recruitment of target proteins to ubiquitin protein ligases for degradation.

[0101] Proteolysis Targeting Chimeras

[0102] Chimeric constructs have been developed as a novel way of specifically targeting proteins and / or peptides for proteolysis. A Proteolysis Targeting Chimera is a chimeric construct having a target protein binding moiety and an E3 (ubiquitin) ligase binding moiety, bringing the target protein in close proximity to the E3 ligase for ubiquitination (W00220740 A2).

[0103] Proteolysis Targeting Chimeras are heterobifunctional molecules that can bring a target protein into proximity to an E3 ubiquitin ligase and thus induce the ubiquitination and subsequent degradation of the target protein. Over the last few years, many target proteins that have been successfully degraded by Proteolysis Targeting Chimeras, however, Proteolysis Targeting Chimeras are proving difficult to develop as therapeutics.

[0104] Molecular glues

[0105] An alternative way to recruit target proteins to ubiquitin protein ligases for degradation is to use a molecular glue. A molecular glue is a small molecule that stabilizes the interaction between two proteins that do not normally interact.

[0106] The most commonly employed molecular glue induces a novel interaction between a substrate receptor of an E3 ubiquitin ligase and a target protein leading to proteolysis of the target. Examples of molecular glues that induce degradation of protein targets include the immunomodulatory imide drugs (also known as IMiD; e.g., thalidomide, lenalidomide, pomalidomide), which generate a novel interaction between a substrate (e.g., IKZF1 / 3, also known as Ikaros / Aiolos) and cereblon, a substrate receptor for Cullin-RING ubiquitin ligase 4 (CRL4). They work in a similar manner to Proteolysis Targeting Chimera molecules, bringing about targeted protein degradation. Distinct from Proteolysis Targeting Chimera molecules, known molecular glues can induce interaction between two proteins but without high affinity for either protein independently. In some known examples these small molecules can insert into a naturally occurring PPI (Protein-protein interaction) interface, with contacts optimized for both the substrate and ligase within the same small molecule entity causing a stabilisation of the induced ternary complex, and therefore enhancing the functional consequences of the PPI. To date, molecular glues have been identified serendipitously which has hampered their broad application. Proteins engineered

[0107] For example, in some embodiments, the invention relates to engineering E3 ligases in order to change their substrate recognition activity, including to cause a change in or induction of degradation of a target protein. Suitably the E3 ligase is a human E3 ligase. In some embodiments, the E3 ligase is cereblon (CRBN). Suitably, CRBN is a human CRBN. In some embodiments, the E3 ligase is KLHDC2 or a mutant of KLHDC2, e.g. KLHDC2-KK. Suitably, KLHDC2 is human KLHDC2.

[0108] As E3 ubiquitin ligases play a crucial role in the induction of degradation and therefore regulation of cellular processes, they are an attractive therapeutic target. There are approximately 500-1000 human E3 ligases as estimated by human genome sequence analysis. These can be divided into four families: HECT-domain E3s (HECT E3s; homologous to E6AP C terminus E3s), ll-box E3s, monomeric RING E3s (RING-finger E3s) and multisubunit E3s (WO2018064589). Currently only a very limited number of known E3 ligases have been demonstrated to be linkable in a targeted degradation approach (e.g. Mouse double minute 2 homolog (MDM2), Von Hippel-Lindau (VHL), Cereblon (CRBN), F-box / WD repeatcontaining protein 1A (FBXW1A; beta-TrCP1) and Cellular Inhibitor of Apoptosis Protein 1 (c- IAP1 ; BIRC2)) (see WO2016118666, WO2019043217, WO2018226542 and

[0109] WG2017011371 ; An S, Fu L. EBioMedicine (2018), Volume 36, 553 - 562; Liu Y, Mallampalli RK. Semin Cancer Biol. (2015) 36:105-119). MDM2, VHL, CRBN, FBXW1A and C-IAP1 ligases all belong to the RING E3 ligase family, the largest family of E3s (Bulatov E and Ciulli A. Targeting Cullin-RING E3 ubiquitin ligases for drug discovery: structure, assembly and small-molecule modulation. Biochem J. 2015). Despite the crucial role played by E3 ligases in cellular regulation, to date the majority of small molecule degrader drugs are designed to recruit either VHL or CRBN to induce ubiquitination (Salama et al, Int. J. Mol. Sci. 2022, 23(23), 15440; Targeted protein degradation: Clinical Advances in the Field of Oncology)).

[0110] In some embodiments, the E3 ligase is cereblon (CRBN). Cereblon forms an E3 ubiquitin ligase complex with damaged DNA binding protein 1 (DDB1), Cullin-4A (CUL4A), and regulator of cullins 1 (ROC1). This complex ubiquitinates a number of other proteins and was first uncovered as the target of the thalidomide analogue drugs. The molecular glue and IMiD activities elicited by these drugs are exploited therapeutically in multiple myeloma treatment and overall CRBN is amongst the most well-studied E3 ligase.

[0111] In some embodiments, the E3 ligase is KBTBD4. KBTBD4 is an E3 ligase of the class BBK, engaging CUL3 and containing a BTB-BACK-Kelch domain for substrate binding. Cullin RING ubiquitin Ligase 3 (CRL3) adaptor KBTBD4 forms a multi-subunit E3 using a BTB domain to engage Cul3 and a Kelch domain to recruit substrates. In some embodiments, the E3 ligase is KLHDC2 (Kelch-like homology domaincontaining protein 2; also known as Kelch domain-containing protein 2). KLHDC2 is a substrate adapter protein and member of the cullin-2 (CLIL2) ubiquitin ligase complex. KLHDC2 is the substrate-recognition component of a Cul2-RING (CRL2) E3 ubiquitin-protein ligase complex of the DesCEND (destruction via C-end degrons) pathway, which recognizes a C-degron located at the extreme C terminus of target proteins, leading to their ubiquitination and degradation.

[0112] A similar approach of insertional mutagenesis to induce a change in activity could be used to modify other protein classes besides E3 ligases and to screen for gain of function mutations. For example, the invention relates to modifying effector proteins which have existing known and defined functional activities, such as enzyme functions, post-translational modifications or other defined biological controls, in order to change their activity towards a specific biological outcome (for example modifying effector proteins in the class: E3 ligase, deubiquitinase, kinase etc, and screening for a change in their activity). Including to engineer the substrate recognition regions or domains of said effector proteins by peptide sequence or library insertion to change their substrate recognition activity e.g. substrate repertoire or activity. In some embodiments, the engineered effector protein is an E3 ligase and the modified biological activity resulting from the engineering of said effector protein is a change in or induction of degradation of a target protein. In some embodiments the engineered effector protein is cereblon (CRBN). In some embodiments the engineered effector protein is KLHDC2.

[0113] For example, effector proteins include: E3 ligases, deubiquitinases, chaperones, kinases, phosphatases, transcription factors, and other enzymes causing post-translational modifications to proteins. Suitably the effector protein is a mammalian effector protein e.g. a human effector protein. In any aspect of the invention, a peptide motif or library of peptide motifs can be inserted into an E3 ligase or into an effector protein other than an E3 ligase, i.e. inserted into effector proteins such as: deubiquitinases, chaperones, kinases, phosphatases, transcription factors and other enzymes causing post-translational modifications to proteins. Suitably, the insertion is within or proximal to a substrate binding interface.

[0114] For example, deubiquitinases could be engineered to comprise insertion mutations in order to change their substrate recognition activity, including to cause a change in or induction of stabilisation of a target protein.

[0115] Deubiquitinating enzymes (DllBs), also known as deubiquitinating peptidases, deubiquitinating isopeptidases, deubiquitinases, ubiquitin proteases, ubiquitin hydrolases, ubiquitin isopeptidases, are a large group of proteases that cleave ubiquitin from proteins. Ubiquitin is attached to proteins in order to regulate the degradation of proteins via the proteasome and lysosome; coordinate the cellular localisation of proteins; activate and inactivate proteins; and modulate protein-protein interactions. DUBs can reverse these effects by cleaving the peptide or isopeptide bond between ubiquitin and its substrate protein. In humans there are nearly 100 DUB genes, which can be classified into two main classes: cysteine proteases and metalloproteases. The cysteine proteases comprise ubiquitin-specific proteases (USPs), ubiquitin C-terminal hydrolases (UCHs), Machado- Josephin domain proteases (MJDs) and ovarian tumour proteases (OTU). The metalloprotease group contains only the Jab1 / Mov34 / Mpr1 Pad1 N-terminal+ (MPN+) (JAMM) domain proteases. DUBs have also been targeted for drug development, both using inhibitors and induced proximity ligands to harness their protein stabilisation properties (Henning et al, Nat Chem Biol. 2022 Apr; 18(4): 412-421 ; Deubiquitinase-Targeting Chimeras for Targeted Protein Stabilization.

[0116] Other suitable effector proteins include: chaperones, kinases, phosphatases, transcription factors, and other enzymes causing post-translational modifications to proteins.

[0117] The terms: engineered protein, modified protein and mutant protein are used interchangeably herein to mean proteins with a peptide motif insertion. Active mutants are such modified proteins showing a desired change in activity.

[0118] In one aspect, the present invention relates to screening modified proteins for insertion mutations that result in a gain of function. In some embodiments, the method of screening couples insertional mutagenesis with next generation sequencing (NGS) (e.g. deep sequencing) to identify active insertion mutation sequences.

[0119] Substrate / Protein of Interest (POI) / Target protein

[0120] Effector proteins are modified, by engineering the substrate recognition regions or domains of said effector proteins by insertion of a peptide motif (or library of peptide motifs) to change their substrate recognition activity i.e. their substrate repertoire or activity (i.e. to change which substrate(s) they bind, or the degree to which they bind a particular substrate). Active mutants can be identified by measuring a change in substrate binding activity and / or target protein degradation activity. So the substrate can be any protein to be bound by an effector protein or a modified effector protein. In some embodiments, the substrate can be a target protein to be degraded directly or the substrate can be an intermediate protein involved in target protein degradation. Other terms for ‘substrate protein’ used herein include: client protein / client, protein of interest (POI), neosubstrate.

[0121] The protein of interest (POI) can be any protein substrate. It can be a target protein for degradation or it can be an intermediate accessory protein that binds to a target protein for degradation. Suitably the accessory protein is a regulatory protein. Alternatively, the POI can be a target protein for stabilisation or an intermediate accessory protein that binds to a target protein for stabilisation. Suitably the accessory protein is a regulatory protein. So a molecular glue could be designed to bind to the target protein and / or the accessory protein.

[0122] In some embodiments, the target protein is a protein identified as therapeutically desirable to perturb, for example by degrading. In general, these are proteins that are associated with a gain of function toxicity in disease.

[0123] In some embodiments, the target protein is selected from the list of target proteins in Table 1 or Table 2 herein.

[0124] In some embodiments, the target protein is associated with a disease, disorder or condition when mutated, expressed or over-expressed in a eukaryotic cell. Suitably, the target protein is one selected from a class of proteins selected from the group consisting of: members of an oncogenic pathway; viral host factors; viral proteins; mis-folded proteins; aggregating proteins; toxic proteins; proteins involved in immune recognition, immune response or auto-immunity; shuttle proteins.

[0125] Tables 1 and 2 provide examples of some suitable target proteins.

[0126] In some embodiments, the target protein is selected from Table 1:

[0127]

[0128]

[0129]

[0130]

[0131] In some embodiments, the target protein is a fusion target protein. In some embodiments, the fusion target protein is selected from Table 2 (e.g. RUNX1 in fusion with any of fusion partners ETV6, MECOM, or RUNX1TI; ABL1 in fusion with any of fusion partners BCR, NLIP214, ENILI or ETV6; and so on). Table 2. Exemplary Fusion Target Proteins

[0132] Peptide motifs, structure and locations.

[0133] The present invention relates to a method of inserting a peptide motif or a library of peptide motifs into one or more proteins in order to change their activity. The peptide motif can be one or more amino acids long. So by insertion, we mean the addition and / or replacement of one or more amino acids. In some embodiments, the peptide motif is a polypeptide sequence of 1 to 110 amino acids in length. This length range is suitable for synthesis in high throughput. The new inserted sequence of one or more amino acids is referred to herein as an inserted peptide sequence / peptide sequence insertion or an inserted peptide motif / peptide motif insertion or the insertion.

[0134] In some embodiments, the inserted peptide sequences are 3, 4, 6, 9, 10, 11 or 15 amino acids long. In some embodiments, the peptide motif comprises a polypeptide length of at least 1 or at least 2 or at least 3 or at least 4 or at least 5 or at least 6 or at least 7 or at least 8 or at least 9 or at least 10 or at least 11 or at least 12 or at least 13 or at least 14 or at least 15 amino acids. In some embodiments, the peptide motif comprises a polypeptide length of 3 to 110 or 4 to 110 or 6 to 110 or 9 to 110 or 10 to 110 or 11 to 110 or 15 to 110 or 3 to 15 or 4 to 15 or 6 to 15 or 9 to 15 or 10 to 15 or 11 to 15 or 3 to 9 or 4 to 9 or 6 to 9 or 3 to 10 or 4 to 10 or 6 to 10 or 3 to 11 or 4 to 11 or 6 to 11 or 10 to 11 amino acids.

[0135] Suitably, the peptide motif is inserted as an intra-protein insertion / in-frame insertion i.e. within an internal domain of the protein, not at the C- or N-terminus. The peptide motifs can be inserted into the protein (e.g. E3 ligase; whose activity is aimed to be changed) at different locations (e.g. within or proximal to a substrate binding interface of the protein). In some embodiments, the insertion occupies a pocket in the protein (e.g. an invaginated, druggable pocket). In some embodiments, the insertion occupies an interface between known and unknown substrates and tolerates a bigger insertion. A longer insertion allows for more diversity, more surface area to bind and a longer extension out of the protein to “hook” the protein of interest. The peptide insertions into E3 ligases are such that the E3 ligase remains active as a competent E3 ligase.

[0136] In some embodiments, the inserted peptide sequences form loops, e.g. such that the loop is presented on the surface of the protein (e.g. E3 ligase) presenting a new neomorphic interface.

[0137] Peptide motif insertion positions are given in codon ranges, i.e. using amino acid residue numbering.

[0138] In some embodiments, the inserted peptide sequences (e.g. 3 to 15 amino acids long, or 3, 4, 6, 9, 10, 11or 15 amino acids long) are placed in the native protein sequence (e.g. CRBN insertion or KLHDC2 insertion). Suitably the insertion is presented as a loop on the surface of the E3 ligase, presenting a new neomorphic interface.

[0139] In some embodiments, the insertion is within or adjacent to the LON domain or the sensor loop of the cereblon protein. Suitably, the location of the insertion of the peptide motif is at region P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wild-type protein cereblon. In some embodiments, the location of the insertion of the peptide motif at P1 is at a subregion defined by codons 138- 162 (SEQ ID NO. 4) or codons 148-152 (SEQ ID NO. 5) (for example, the peptide motif insertion replaces codons 149-151 or the peptide motif is inserted between codons 150 and 151). In some embodiments, the location of the insertion of the peptide motif at P2 is at a subregion defined by codons 351-354 (SEQ ID NO. 6) or 352-353. In some embodiments, the peptide motif is inserted at up to 5 angstroms or up to 10 angstroms from one of the said specific P1 or P2 regions / subregions (said specific codon ranges) in the wildtype cereblon protein. In some embodiments, the peptide motif is inserted into a region defined by a codon range which has an end point up to 1 , 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids away from an endpoint of said codon ranges. For example, as set out in Figure 22E, there is provided a library of peptide motif insertions (3 to 10 amino acids in length) at a P1 subregion defined by codons 148-152 (SEQ ID NO. 5). For example, as shown in Figure 22E, the library may include peptide motif insertions of 6 or 10 amino acids replacing codons 149 -151 , and peptide motif insertions of 3 amino acids inserted between codons 150 and 151 of wild-type CRBN. For example (as in example 10), the library may include peptide motif insertions of 3 or 4 amino acids inserted at a P2 subregion defined by codons 351-354 (SEQ ID NO. 6) or 352-353.

[0140] In some embodiments, the protein is KBTBD4 (SEQ ID NO. 7) and the peptide motif insertion is an in-frame insertion of one or more amino acids resulting in a change in the activity of KBTBD4 (e.g. promote the recruitment of CoREST as a neo-substrate for ubiquitylation and degradation). In some embodiments, the inserted peptide sequences are placed into the Kelch domain of KBTBD4. In some embodiments, the insertion is within the KBTBD4 Kelch motif region defined by codons 308-313 (SEQ ID NO. 8).

[0141] In some embodiments, the inserted peptide sequences (e.g. 3-15 amino acids long, or 11 amino acids long) are placed in the native protein sequence (e.g. KLHDC2 insertion). Suitably the insertion is presented as a loop (e.g. a new loop or within an existing loop) on the surface of the E3 ligase, presenting a new neomorphic interface. In some embodiments, the insertion is at a location (e.g. centrally) within Loop A (SEQ ID NO. 12), B (SEQ ID NO. 13), C (SEQ ID NO. 14), D (SEQ ID NO. 15), E (SEQ ID NO. 16) or F (SEQ ID NO. 17) of the KLHDC2 protein. Suitably, the location of the insertion of the peptide motif is within Loop A (Loop A is defined by codons 46-66 (SEQ ID NO. 12)) or within Loop B (Loop B is defined by codons 105-117 (SEQ ID NO. 13)) or within Loop C (Loop C is defined by codons 159-194 (SEQ ID NO. 14)) or within Loop D (Loop D is defined by codons 232-244 (SEQ ID NO. 15)) or within Loop E (Loop E is defined by codons 283-296 (SEQ ID NO. 16)) or within Loop F (Loop F is defined by codons 334-353 (SEQ ID NO. 17)) in the wildtype KLHDC2 protein. For example, the peptide motif is inserted centrally within loop A, B, C, D, E or F. In some embodiments, a peptide motif is inserted in Loop A at codons 52-58 (SEQ ID NO. 18) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop B at codons IOS- 111 (SEQ ID NO. 19) or up to 5 angstroms or up to 10 angstroms of that region, or a peptide motif is inserted in Loop C at codons 177-186 (SEQ ID NO. 20) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop D at codons 236-239 (SEQ ID NO. 21) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop E at codons 289- 292 (SEQ ID NO. 22) or up to 5 angstroms or up to 10 angstroms from that region, or a peptide motif is inserted in Loop F at codons 343-346 (SEQ ID NO. 23) or up to 5 angstroms or up to 10 angstroms from that region. In some embodiments, a peptide motif insertion within Loop A replaces codons 52-58, and a peptide motif insertion within Loop B replaces codons 108-111 , and a peptide motif insertion within Loop C replaces codons 177-186, and a peptide motif insertion within Loop D replaces codons 236-239, and a peptide motif insertion within Loop E replaces codons 289- 292, and a peptide motif insertion within Loop F replaces codons 343-346. Suitably, the peptide motif inserted into these specific locations of KLHDC2 comprises a polypeptide length of 1 to 110 amino acids. Optionally, wherein the peptide motif inserted is 3-15 amino acids long, e.g. 11 amino acids long. Optionally, wherein the peptide motif is inserted (e.g. centrally) within Loop A, C, D or F. For example, Figure 18C and D show HiBiT (VSGWRLFKKIS) (SEQ ID NO. 24), which is 11 amino acids long, inserted into KLHDC2 centrally within Loops A (SEQ ID NO. 25), C (SEQ ID NO. 26), D (SEQ ID NO. 27), and F (SEQ ID NO. 28). Figure 18C shows the 11 amino acid Hi BiT insertion at the following positions: in Loop A replacing codons 52-58 (SEQ ID NO. 18), in Loop C replacing codons 177-186 (SEQ ID NO. 20), in Loop D replacing codons 236-239 (SEQ ID NO. 21) and in Loop F replacing codons 343-346 (SEQ ID NO. 23).

[0142] Libraries

[0143] Protein fragment expression libraries that can be screened in high throughput in phenotypic assays are known and often dubbed ‘Protein-interference’ (Protein-i). For example, W02013 / 116903A1 describes randomly fragmented bacterial phylomer libraries, and using these in screening methods for characterising an interaction site on a target protein that modulates the phenotype of a mammalian cell. Another example, is as set out in W02020 / 074891 , where the present inventors screen millions of carefully designed miniprotein fragments, called “PROTEINi®” for phenotypic effects in disease relevant cellular models.

[0144] The present invention relates to a method of inserting peptide motifs and libraries of peptide motifs into proteins in order to change their activity. For example, a library of designed MG- PROTEINi® is introduced into E3 ligase to ablate POI (Protein of Interest) abundance (e.g. Target proteins such as GSPT1 , KRAS, STAT3 or any target protein for example in Table 1 or Table 2). In some embodiments, methods according to the invention use computationally designed intramolecular libraries to create a vast diversity of modified effector proteins, e.g. surface-edited E3 ligases. Phenotypic screening of the modified effector proteins can then be deployed to identify specific sites and precise changes resulting in an induced change of activity, e.g. induced degradation activity.

[0145] Screening for change of activity

[0146] The present invention relates to screening modified effector proteins (with insertion mutations) for a change in activity. The change in activity can be any new activity e.g. a gain of function induced by the modification. The invention enables identification of mutant effector proteins with new protein function i.e. new protein-protein interactions. For example, the invention relates to screening modified effector proteins (e.g. E3 ligases) for a change in substrate recognition activity i.e. a change in substrate binding activity and / or target protein degradation.

[0147] In some embodiments, the invention relates to induced effector protein (e.g. E3 ligase):POI interactions to inform drug design. In one embodiment, a library of designed MG- PROTEINi® is introduced into E3 ligase to ablate POI (Protein of Interest) abundance (e.g. GSPT1 , KRAS, STAT3 etc). Hits induce productive neomorphic PPI (Protein-protein interaction) between POI and E3 ligase and are used for either in silico or structural data acquisition. The pocket identification provides a novel experimentally validated locus for monovalent drug discovery since a new neomorphic interaction is necessarily described by the discovery.

[0148] The method enables identification of new protein function (or a “neomorphic” mutation), resulting from a peptide motif insertion, e.g. identification of a new protein- protein interaction (PPI). The method enables identification of ‘gain-of-function’ insertion mutations.

[0149] For example, a nominated E3 ligase (e.g. CRBN or KLHDC2) is mutated to insert an intramolecular sequence or library of sequences which are then screened to identify neomorphic activity.

[0150] For example, screening for gain of function insertion mutations in ubiquitin ligases (e.g. neosubstrate recruitment).

[0151] The new function is a change in substrate recognition activity which is detected by a change in substrate binding and / or a change in target protein (POI) degradation. Suitably, POI degradation is monitored by measuring destabilisation of a nominated POI directly in a high- throughput cell assay, such as by western blotting, fluorescence activated cell sorting to monitor the change in abundance of a defined protein of interest; or a biomarker functionally linked to the protein of interest or by another phenotypic effect associated with loss of the protein, such as cell death.

[0152] Once a new activity has been identified, next generation sequencing can be used to identify active sequences in a phenotypic measurement. Active mutant proteins can be identified and isolated.

[0153] An identified peptide motif could be described as mimicking a monovalent degrader, and is therefore useful for monovalent degrader discovery. For example, the screening output provides a new way to induce PPI between an E3 ligase and a POI, and this stable and functional interaction can be recapitulated by small molecules. In one aspect, the inserted mutation sequence provides exact molecular design and modelling input for drug design using equivalent chemical groups and side chain interactions. In another aspect, a de novo pocket between the E3 ligase and POI can be targeted for small molecule drug design resulting in monovalent degrader drugs.

[0154] All references, patents, and publications cited herein are hereby incorporated by reference in their entirety.

[0155] The following examples relate to this new approach of neomorphic enzyme remodelling by intra-molecular E3 ligase library insertional mutagenesis and gain of function assays. The examples focus on the following E3 ligases: CRBN and KLHDC2 / KLHDC2-KK, and demonstrate that the invention is generalisable to other effector proteins including other E3 ligases (e.g. KBTBD4, and VHL).

[0156] EXAMPLES

[0157] Example 1 - Expression and stability of the E3 ligase Cereblon (CRBN) with varying length intra-sequence insertions.

[0158] In order to determine whether CRBN could be modified with varying lengths of peptide motifs, sequences that are either 4, 9 or 15 amino acids long were inserted in the internal domain and not at the C- or N-terminal of the native CRBN sequence. Mutagenised CRBN was then expressed from a plasmid or lentiviral integration in a cell line from which endogenous CRBN has been genetically deleted (CRBNk0). Cells were harvested, fixed and permeabilised, stained with anti-CRBN antibody (CST#71810). The expression level of a mutant insertion into CRBN was measured by flow cytometry staining, and compared to the wild-type form (CRBNm), either reconstituted in CRBNKOcells or the non-deleted wild-type cells (Fig. 1A). The traces in Fig. 1A shows overlap indicating that the insertion in CRBN is expressed to an approximately equal level in the cells to the native form. A similar experiment using western blotting was undertaken to examine CRBN expression and stability from three different length insertions in a defined position within the peptide sequence of CRBN (Fig. 1 B). Membranes were probed with primary antibodies against CRBN (CST#71810) or GAPDH (CST#97166) as a loading control. After the addition of secondary antibodies, protein signals were detected by LI-COR. Here, each CRBN form can be clearly seen by western blotting and the increased length can be seen by a retarded migration on the gel. In conclusion, lengths or oligopeptide insertion were introduced into CRBN at a defined location without compromising the overall expression or stability of the proteins as compared to the endogenous or wild-type forms. This demonstrated that a peptide motif of varying lengths can be inserted into an effector protein (CRBN) and a stable modified CRBN can be generated as compared to a wild-type CRBN.

[0159] Example 2 - The cereblon E3 ligase remains functionally active when remodelled with intra-molecular peptide loop insertions

[0160] In order to assess whether the insertion of a peptide motif at region P1 or at P2 generates a functional modified CRBN, peptide motifs were inserted at each of these sites and the enzyme activity was measured. A graphic depiction of the remodelling of the E3 ligase CRBN is illustrated in Fig. 2A, where oligopeptide motifs (4, 9 or 15 amino acids long) are introduced into the sequence of the protein, such that the loop is presented on the surface of the E3 ligase, presenting a new-neomorphic interface. Two different positions are approximately indicated, denoted P1 and P2. On the introduction of the new sequences, this example presents a near-complete or partial expected activity of the E3 ligase, whereby it remains conducive to small molecule-induced interactions and degradation via degrader drugs, such as dBET6, which causes the degradation of BRD4 via CRBN ligase activity. A second graphic depiction illustrating the expected effect of the dBET6 on BRD4 is shown in Fig. 2B. CRBN expression is reconstituted in CRBNk0cells (WT, P1 and P2) and cells treated with DMSO or 0.1 pM dBET6 or 0.1 pM dBET6 with 1 pM MLN4924 for 18 hours (Fig. 2C). Cells are lysed and analysed by western blot using antibodies against BRD4 (CST#13440), CRBN (CST#71810) and GAPDH (CST#97166). A peptide motif sequence comprising the epitope tag “HA” was inserted into two positions, either P1 or P2 within CRBN. Wild-type (WT) CRBN when reconstituted into CRBNk0cells shows almost equivalent performance to the parental cells (Par) in which endogenous CRBN is active. Importantly, insertions in position P1 and to a lesser extent P2 show activity similar to the WT CRBN, indicating that the remodelled enzyme remains active as a competent E3 ligase. This activity is also reversed when cells of any variant are treated with the NEDDylation inhibitor MLN4924, which blocks cullin ligase function. Therefore, the insertion of a peptide motif in position P1 or P2 generates a functionally active modified E3 ligase.

[0161] Example 3 - Demonstration that neomorphic insertion E3 ligase mutants can engage in new, stable protein-protein interactions consistent with a new function

[0162] The following experiment was undertaken in order to assess whether a modified E3 ligase with a peptide motif inserted within or proximal to its substrate binding interface results in a change of activity such as a change to substrate binding activity and / or protein degradation. A graphic of the experimental strategy is illustrated in Fig. 3A. Intra-molecularly HA-inserted CRBN protein was co-expressed in CRBNk0cells with a GFP-tagged intracellular single-chain antibody harbouring specific affinity for the HA epitope tag (HA-15F11). CRBNk0cells are reconstituted with CRBN_wt or CRBN_P1 / P2 expression, co-expressed with a nanobody-GFP fusion targeting P1 or P2 (Fig. 3B). Cells are treated with DMSO or 1 pM Pomalidomide (1 pM Pom, a molecular glue) or 1 pM Pomalidomide with 1 pM proteasome inhibitor MG132 for 4 hours. Cells are then lysed and GFPTrap is used to enrich GFP from total cell lysates. Whole cell lysates (Input) and enriched fractions (IP) are then analysed via western blot with different antibodies (anti-GFP CST#2956, anti-CRBN CST#71810, anti-DDB1 CST#5428, anti-Cul4A Abeam ab92554 and anti-Tubulin CST#86298). The input data shows the expression of the GFP-tagged nanobody on western blotting, and the various CRBN insertion mutants can also be stained with antibodies targeting CRBN. When the GFP-tagged nanobody is used as the target for immunoprecipitations (IP), CRBN can then only be retrieved with high affinity following the introduction of the HA tag into sites P1 or P2, and not with wild-type (WT) or negative controls. This indicates a specific and stable interaction is formed by the two components when co-expressed in cells together, and confirms the activity and recruitment of the inserted tag for induced proximity. The interaction is not dependent on CRBN activation status, since it is not affected by the introduction of the IMiD degrader drug pomalidomide (Pom) nor is it affected by blockade of the proteasome through MG 132 treatment. Thus, both sites P1 and P2 are permissive for de novo induced proximity between a control epitope tag and a complementation counterpart, nanobody HA-15F11 and CRBN insertion mutants are therefore primed for neomorphic induced proximity and degradation discovery methods via these sites and others. In conclusion, this demonstrates that a modified E3 ligase with a peptide motif inserted within or proximal to its substrate binding interface (e.g. at P1 and P2), does indeed result in a change of activity of the E3 ligase. In this specific example, the change of activity was to induce close proximity between the modified E3 ligase and the GFP tagged nanobody for targeted protein degradation.

[0163] Example 4 - Confirmation of different physiological states found for P1 and P2 insertion sites.

[0164] A graphic illustrating the experimental setup is shown in Fig. 4A. When cells are treated with the signalosome inhibitor CSN5i-3, the deNEDDylation process becomes inhibited, trapping activated cullin-ring ligases (CRL) complexes in the NEDDylated and activated form. For CRL ligases like CRBN-containing ones, this results in hyperactivated state and the autodegradation of CRBN from the cells. Substrate binding from Protein-of-interest (POI) or endogenous clients protects CRBN from autodegradation, as it maintains competition from alternative lysine sites on these clients for ubiquitylation. Different insertion site mutants of CRBN are tested for CRBN abundance after CSN5i-3 treatment (Fig. 4B). Cells are treated with DMSO or 1 pM CSN5i-3 for 4 hours and CRBN expression detected using anti-CRBN antibody (CST#71810) and analysed by flow cytometry. Wild-type (WT) reconstituted CRBN (orange) is partially protected from degradation and so is the P1 site insertion mutant. P2 insertion site is autodegraded, indicating this variant is found in a distinct physiological form to the P1 mutants and may be either less active, or less bound by native clients than the WT or P1 mutants. The same experiment as Example 3 was undertaken. WT / P1 / P2 expressing cells were treated with 1 pM CSN5i-3 for 4 hours and GFPTrap was used to enrich CRBN interacting with nanobody-GFP fusion (Fig. 4C). It is shown that the presence or binding of the nanobody (robust for both sites) is insufficient to protect the CRBN from autodegradation. Thus, sites P1 and P2 show alternative physiological properties and offer distinct and discrete opportunities for induced neomorphic activity of CRBN. Importantly, site P1 has apparently high native activity and presumably high endogenous client binding and offers a potentially highly conservative insertion site.

[0165] Example 5 - Demonstration of neomorphic screens of CRBN.

[0166] In order to assess whether a library of peptide motifs could be cloned into one of CRBN’s substrate binding interfaces, a high diversity library was cloned into the P1 site (Fig. 5A). A CRBN expression plasmid with P1 and P2 sites is illustrated. For example, a library was inserted at the P1 position with length of variable sequence composition of 3AA to 10AA, and comprised a variety of insertions to and replacements of the native CRBN sequence (e.g. as in Figure 22E). P1 site library, depicted by different patterns, is amplified by PCR. The rest of a codon-optimised CRBN (SEQ ID NOs. 34 and 35) expression plasmid lacking P1 site is also amplified as vector. P1 library is then inserted into vector via recombination (e.g. using NEBuilder® HiFi DNA Assembly Master Mix E2621X). A PROTEINi® screen is illustrated in Fig. 5B. Lentivirus encoding CRBN library from Fig. 5A is produced using second / third generation recombinant lentivirus technology, and transduced into a mammalian CRBNk0cell line. Library-expressing cells are selected and expanded, with or without small molecule treatment. Cells are then harvested, fixed and permeabilised, and immunostained if needed, before sorted using fluorescence-activated cell sorting. The relative distribution of each sequence of the library in each population is determined by next generation sequencing. PCR amplification of CRBN library over endogenous CRBN loci is illustrated in Fig. 5C. Primers designed to be complimentary to library-encoded CRBN ORF flanking P1 site (P1_Fwd and P1_Rv) are used to amplify P1 amplicon via PCR. P1_Fwd and P1_Rv do not bind to endogenous CRBN loci. This example demonstrates ectopic CRBN loci can be incorporated into cell models and distinguished from endogenous loci.

[0167] Example 6 - Examples of data analysis for phenotypic screens.

[0168] In order to identify whether cloning a variable length library of peptide motifs (e.g. 3AA to 10AA) into the P1 or P2 position results in a change in activity, gain of function screens were performed at the P1 site (illustrated in Figs.6A-6C) and at the P2 site (illustrated in Figs. 6D- 6F). Cells expressing a library of peptide motifs at the P1 position of CRBN’s substrate binding interface are treated with Pomalidomide and sorted on POI1 intensity as shown in Fig. 6A. POI1 is a substrate of CRBN_wt in the presence of CC-90009 but not Pomalidomide (Fig. 6A.). Library-expressing cells are treated with Pomalidomide and expanded in Fig. 6B. Cells harvested before and after Pomalidomide treatment are analysed to identify mutations depleted upon Pomalidomide treatment. Cells expressing CRBN_wt treated with CC-90009 is used as positive control for reduced proliferation. Correlation between screens from Fig. 6A and Fig. 6B are shown in Fig. 6C. POI 1 degradation-related insertions from Fig. 6A should correlate with depletion from Fig. 6B and clustered at the top right of figure, similar to CRBN_wt treated with CC-90009. CRBN_wt treated with Pomalidomide results in no POI1 degradation or depletion. Mutations associated with degradation of neosubstrates other than PO11 , and leading to reduced proliferation, are clustered at the bottom right of the figure. Cells expressing a library of peptide motifs at the P2 position of CRBN’s substrate binding interface are stained with antibody against POI2 and sorted by FACS (Fig. 6D). siRNA targeting POI2 in CRBNk0cells serves as a positive control for POI2 reduction. Library-expressing cells are treated with signalosome inhibitor CSN5i-3 and stained with anti-CRBN antibody (CST#71810) (Fig. 6E). Cells are sorted on CRBN intensity to identify mutations associated with high CRBN intensity. A previously-characterised CRBN_P2 serves as non-catalytically active control against activated CRBN (CRBN_wt treated with Pomalidomide). Correlation of Fig. 6D and Fig. 6E are shown in Fig. 6F. Mutations associated with POI2 degradation should correlate with shielded CRBN degradation upon CSN5i-3 treatment, clustering at the top right of the figure. CRBN_wt treated with Pomalidomide shows no POI2 degradation, where siRNA targeting POI2 leads to POI2 degradation but no CRBN shielding upon CSN5i-3 treatment. This example demonstrates the methodology for identifying and scoring active peptide insertions which degrade a POI or alter the function of the nominated effector protein, for example CRBN.

[0169] Example 7 - Proof-of-concept degradation by engineered CRBN

[0170] In order to illustrate target protein degradation by a modified CRBN, a mutated form of the CRBN ligase was generated using the molecular biology insertion strategy of previous examples, but where the inserted sequence was a HiBiT peptide sequence (SEQ ID NO. 24). Hi BiT is a fragment of a luminescent protein NanoLuciferase which can bind with high affinity to the remainder of the NanoLuciferase enzyme, which is termed LgBiT. When LgBiT and HiBiT are bound, the resulting complementation allows the protein to metabolise an exogenously provided substrate and create visible light. When LgBiT and HiBiT are not bound, no luminescence or light can be generated by the enzyme. CRBN-KO cell lines were produced which stably expressed mutated forms of CRBN, where the HiBiT sequence was inserted either into the P1 position in CRBN or at the N or C terminus of the protein. The LgBiT protein is then transiently expressed in cells through a plasmid transfection of varying DNA concentrations ahead of lysing of the cells and assessment of luminescence. Figure 8A illustrates the degradation process. Figure 8B illustrates the expression analysis through flow cytometry of the stable CRBN cell lines with different integration positions of the HiBiT tag. Note that all variants are measurably and stably expressed to approximately the same degree. Figure 8C shows the comparison of the luminescence resulting from the different CRBN-HiBiT variants following increasing titrations of LgBiT plasmid DNA. CRBN with HiBiT integrated in the P1 position (HiBiT-P1) shows a markedly diminished light emission at all concentrations of LgBiT added as compared to the C-terminal domain (CTD) position variant (HiBiT-CTD). Figure 8D shows the transfection of 100 nM of LgBiT into the stable cell variant housing the P1 position insertion of HiBiT prior to treatment with either a proteasome inhibitor (MG132) or an inhibitor of the cullin ligases (which includes CRBN complex; MLN4924). Luminescent signal is rescued by these compound treatment, indicating that the LgBiT is likely degraded by the HiBiT-CRBN in the P1 position. Note that this degradation is induced only by the HiBiT loop insertion, and occurs without the requirement for I M i D treatment. Therefore, this example demonstrates that a mutant version of the CRBN ligase, wherein a peptide motif is inserted into the P1 position can successfully induce a change in activity e.g. target protein degradation, as illustrated with the diminished luminescence signal generated when HiBit was integrated in the P1 position.

[0171] Example 8 - library-scale integration and sequence detection for screening

[0172] To assess the methodology for creating new protein-protein interactions on the surface of CRBN, a library of 24,000 distinct nucleotide sequences was generated for insertion into the position P2 in CRBN. These insertions create a new mini-PPI domain in CRBN, and their precise behaviour is analysed in the context of CRBN function, for instance in degrading known or new protein substrates in cells. Figure 9A shows the schematic of the library insertion for position P2 in CRBN and the subsequent amplification of the fragment from genomic DNA in cells. Figure 9B shows the varying concentrations of DNA from cells were subjected to PCR as indicated in Figure 9A and analysed by gel electrophoresis. Specific bands of amplified product are shown which are found only when the library-containing cells are present, and not the parental cell (HAP1). Figure 9C shows that the DNA amplicons from the PCR reactions were sequenced by next-generation sequencing (Illumina NovaSeq) and sequences mapped back to the designed library. The histogram shows the distribution of sequences found with different counts. All 24,000 sequences were identified with a very small and consistent distribution, clearly demonstrating that hit sequences from this can be separated and identified by this technique. This example shows that insertion of new peptide sequences in a defined location within a chosen effector protein can be conducted at a library scale. As illustrated by the example, the new sequences and library composition can be analysed for phenotypic effects using next-generation sequencing.

[0173] Example 9 - Molecular modelling and predictions for a P2 library screen and molecular glue design.

[0174] In order to demonstrate the selection principles for peptide insertion position and the translation of inserted sequences into small molecule molecular glues, a computational analysis of the insertion site P2 in CRBN was completed. Comparison of the molecular architecture of an inserted sequence in CRBN at the P2 site and a known molecular glue, CC- 885 (Fig. 10). In Fig. 10A a schematic is shows the concept of one aspect of the invention where sequences of defined amino acid composition and order are inserted into CRBN at defined sites, for instance P2. An example of the molecular detail of which are shown in Fig. 10B where left panel shows the crystallographic data for the CRBN protein in complex with GSPT1 and the GSPT1 molecular glue CC-885, whereas the right panel shows the same domain and interface but derived from protein structural modelling with a 4 amino acids (4AA) inserted sequence at the site P2 and highlights the equivalent space occupancy to the molecular glue drug with inserted protein loops. Multiple sequence inserts were designed for the site P2 and using molecular dynamics simulations a calculation of the free energy was derived for each one (ddG; Fig. 10C). More negative ddG sequences are more likely to show high affinity interactions between CRBN and GSTP1 . This example shows the similarity at the atomic level of the inserted sequences to the small molecule molecular glue and the use of computational tools to nominate, evaluate and prioritise the insertion positions, length and sequence compositions in order to recapitulate the effect of the molecular glue.

[0175] Example 10 - Screen evaluation analysis for neomorphic activity screens with CRBN.

[0176] In order to demonstrate the screening analysis conducted for the discovery of neomorphic sequence insertions, a screen with a library at position P2 of CRBN was conducted using several parallel phenotypic readouts. Screening was completed with a library of 24,000 sequences inserted at the P2 position (e.g. at a P2 subregion defined by codons 351-354 (SEQ ID NO. 6) or 352-353) in CRBN (Fig. 11). The use of antibodies to endogenous proteins allowed the monitoring of protein abundance by flow cytometry. Three proteins are shown with staining properties suitable for pooled-based screening for neomorphic activity and loss of abundance for these proteins (Fig. 11 A). After introduction of the library by lentivirus, transduced cells are isolated by antibiotic selection and subsequently collected by flow cytometry ahead of deep sequencing. The deep sequence analysis is shown to yield high library recovery, where waterfall plots of samples show the percentage of the library found with a defined number of sequencing reads and which is sufficient to robustly determine the most active sequences (Fig. 11 B). After further statistical analysis, hit sequences can be identified by volcano plots for each biomarker protein, including staining for CRBN protein itself following treatment with deNEDDylation inhibitors such as CSN5i - this traps the CRBN in an active form and causes autodegradation as described in Example 4. Hit space is indicated by the corresponding markers on the scatter plots (Fig. 11C). Therefore, the example demonstrates the use of a screening method according to the invention to find the inserted sequences in CRBN which induced a new and non-native (neomorphic) degradation of target therapeutic proteins.

[0177] Example 11 - Screening demonstration using co-treatment of cells with IMiD scaffolds such as pomalidomide.

[0178] In order to illustrate the use of existing small molecules and phenotypic screening outputs to combine with the sequence insertion invention, a new screen was completed with a library at the P1 position. In this case the library was of approximately 150,000 sequences of variable length from 3AA to 10AA (e.g. as in Fig. 22E). With the library in this position, CRBN is able to maintain binding activity to IMiD-like drugs such as pomalidomide which can be used in combination with the insertion to find neomorphic insertions. The use of cell fitness as a phenotypic approach is also shown to illustrate the use of the invention for neosubstrate target discovery. Screening schematic with libraries as demonstrated in Example 10 but where cells can also be co-treated with molecular glue scaffolds such as IMiDs (Fig. 12). Staining of cells for target proteins (such as GSPT 1 and others) or measurements of cell abundance (so-called “dropout” screening) allows the measurement of cell fitness as a result of CRBN or other E3 ligase library insertions.

[0179] Example 12 - Demonstration of screening with CRBN libraries at the P1 position for neomorphic activity resulting from inserted sequences.

[0180] In Fig. 13A a graphic of CRBN protein with an illustrative loop insertion depicting the peptide sequence insertions at P1 according to the invention. Screening data collected as in Example 10 here showing comparisons of whole-library analysis evaluated using PCA analysis based on a library of 150,000 insertion sequences (Fig. 13B). Sequences cluster together by the origin of their collection for each replicate of the experiment indicating high comparison and expected results for robust neomorphic sequence discovery (where II = unsorted cells; L = low expression population; H = high expression and P = the library plasmid before it is introduced into cells). New sequences that cause the robust and rapid degradation of GSPT 1 are shown in Fig. 13C in MA plots, where positive and negative controls are clearly detected as expected and the hit space is indicated. For example, for the hits which are found to be active in the presence of the molecule there were approximately 3000 robust hits. Sequence convergence and consensus analysis subsequently allowed for the clustering of these hits to several representative active motifs. Here cells are treated either with DMSO or an inactive degrader of GSPT1 , pomalidomide (pom). Library inserts are found in both cases and with substantial enrichment in the pom-treated cells - these hits represent sequences that endorse CRBN with sufficient new affinity for GSPT 1 that the inactive molecular glue pom is rescued and is now found to be able to robustly degrade GSPT1 , illustrative of the application of the insertion sequences to molecular glue drug design principles based on new protein-protein interactions. The same data is normalised to further validate hits and aggregate multiple sequence insertions with similar / identical side-chain properties using Z-score analysis and corresponding robust hit space is shown (Fig. 13D). Therefore, examples 11 and 12 demonstrate the use of small molecules in combination with inserted sequences at the P1 position to find neomorphic activity. It also shows the approach to identify the best sequences by comparison of different measurements.

[0181] Example 13 - Features and characteristics of hit space from CRBN inserted sequences for the induction of neomorphic degradation of GSPT1.

[0182] In order to illustrate the properties of the new and unique sequences identified as neomorphic from screening activities in examples 10 and 11 , a computational analysis of the sequence composition and chemical properties was completed. Neomorphic-inducing sequences from screens such as that shown in Example 12 are evaluated for chemical and geometric properties to define the features which cause recruitment of CRBN to GSPT1 (Fig. 14). The length of the active sequences was found to differ depending on the presence or absence of proximal drug-binding, consistent with the requirement of CRBN to bind both drug and GSPT 1 for hits with Pom treatment, but where longer sequences with presumably increased affinity endorsements were required in the absence of the drug (Fig. 14A). Similarly, the overall volume of the amino acid side chains was calculated for active neomorphic sequences and shown to vary similarly, whereby larger insertions are required where cells are not treated with small molecule molecular glue precursors such as pomalidomide (Fig. 14B). Chemical properties of each sequence type (requires or does not require IMiD scaffold) is also strikingly different (Fig. 14C) consistent with the definition of the protein sequence properties required for inducing activity in the system, determining the small molecule glue dependencies and the precision of the invention for discovering new active PPI sequences. Although there is a preference for certain lengths for the active sequences found in either the presence or absence of the small molecule, where there are active short sequences in the absence of pom these were of strikingly similar sequence composition. This sequence motif overlap for active sequences is shown in Fig. 14D, where highlighted marks in the MA plot indicate short (three amino acid) insertions with substantial sequence similarity, all showing robust degradation of GSPT1 (in the absence of Pom). Therefore, this example shows the distinct chemical properties of the active sequence, demonstrating the consistency and robustness of the invention and principles for screening. It also shows the properties for the translation of the peptide sequences into small molecules which can be designed to accommodate the same physical features as the identified sequences.

[0183] Example 14 - Demonstrations of neomorphic induction of degradation of additional neosubstrates for CRBN.

[0184] In order to further illustrate the application of the invention for other phenotypic readouts, for other protein degradation and for neosubstrate target discovery, a further set of analyses were performed using a library of 150,000 sequences of variable length from 3AA to 10AA of several types (see Fig. 22E) in the P1 position of CRBN. This is further demonstration for neomorphic CRBN discovery as in Example 11 and Example 12, but with a measure instead of the protein STAT3 or cell fitness is shown (Fig. 15). In Fig. 15A a volcano plot shows the hit sequences obtained for the degradation of STAT3 by direct staining of library-containing cells with antibodies to STAT3 followed by deep sequencing and statistical analysis (Fig. 15A), with hit active sequences highlighted including sequences with substantial overlapping amino acid properties. Cells were deep sequenced at increasing time intervals from library introduction as in Example 11 and analysed for library component abundance by deep sequencing. Sequences which are lost over time (drop-out) are indicative of loss of cell fitness, likely as a result of the induced neomorphic degradation of proteins which are essential to cell survival (Fig. 15B). Since the protein GSPT1 is an essential protein for cells, the overlay of the cell lethal sequences and the sequences that cause loss of GSPT 1 directly can provide further robust confirmation of the neomorphic activity induced by the modified CRBN (bottom right quadrant sequences, Fig. 15C). Therefore, this example illustrates the breadth of the invention’s application to the therapeutic targets that would be suitable for degradation by neomorphic sequence insertion.

[0185] Example 15 - A suitable model system to explore induced KLHDC2-based neosubstrate discovery In order to illustrate the application of the invention to effector proteins of multiple identities and to show the design principles for the selection of insertion positions, a further set of screens is shown using an additional E3 ligase. KLHDC2 is a further suitable effector protein for screening that can be used to induce and discover neomorphic interactions including those that change the substrate binding reactions of the effector protein and their subsequent degradation. In Fig. 16A a HAP1 cell line was created in which the endogenous KLHDC2 had been deleted by CRISPR editing and then variants of KLHDC2 protein were subsequently reintroduced and analysed for expression by flow cytometry staining of an integrated epitope tag, HA. Two different multiplicity of infections (MOI) of lentivirus were used to simulate low (~0.3 MOI) and high (»1.0) expression levels of the protein, and cells were selected for integration by antibiotics. Both the wild-type (WT) and a mutational variant of KLHDC2 which lacks a C-terminal degron motif (KLHDC2-KK) were found to express similarly at either the high or low abundance points. KLHDC2-KK is expected to be found as a monomer (Scott et al, (2023) Molecular Cell, Volume 83, Issue 5, 770 - 786. e9: E3 ligase autoinhibition by C- degron mimicry maintains C-degron substrate fidelity)) and might therefore be more available for induced recruitment to neosubstrates by the present invention. In Figs. 16B-D the native function of the introduced variants of the enzyme KLHDC2 are determined. SelK (Rusnac et al., Mol Cell 2018, Dec 6;72(5):813-822. e4: Recognition of the Diglycine C-End Degron by CRL2KLHDC2Ubiquitin Ligase; Koren et al., Cell 2018, Jun 14;173(7):1622-1635.e14: The Eukaryotic Proteome Is Shaped by E3 Ubiquitin Ligases Targeting C-Terminal Degrons) and peptide Hit 528 are expected to be substrates of the KLHDC2 ligase, whereas SelK-G-1 L (C- terminal G of the degron replaced with L) is not. These peptides were fused to GFP for reporter-based tracking of abundance and co-expressed with KLHDC2 variants from Fig. 16A. In each case, all variants of KLHDC2 caused degradation of the correct substrates to a degree comparable to the unmodified HAP1 cells (parental) and this degradation was rescued by inhibitors of the cullins and proteasome pathway. This example shows that these variants of KLHDC2 are therefore functional in cells and are suitable for induced neomorphic function screening according to the present invention.

[0186] Example 16 - Modelling and identification for sites of peptide loop insertion for a new E3 ligase

[0187] In order to illustrate the features required for nomination of suitable insertion positions, molecular modelling of the KLHDC2 enzyme was completed and is shown in Fig. 17. These are example methods which lead to selection of positions within KLHDC2 suitable for induced neomorphic functional discovery. In Fig. 17A the structure of KLHDC2 is shown with a ligand binding domain indicated (see for examples Hickey et al., Nat Struct Mol Biol 2024 Feb;31(2):311-322: Co-opting the E3 ligase KLHDC2 for targeted protein degradation by small molecules) in hashed depiction. These are residues on the KLHDC2 protein that are known to form electrostatic interactions with a ligand to the enzyme. In cases where the ligand could be used in combination with the inserted sequences, these positions in the protein should be avoided for insertions to prevent diminished ligand binding. Alternatively, where the peptide insertion is designed to replace the ligand binding then these positions could be targeted for peptide insertions. In Fig. 17B the predicted sites on KLHDC2 which could interact with other proteins are highlighted as individual residues. These are determined from existing and predicted protein-protein interactions. Where the peptide insertions are designed to further strengthen and enhance existing interactions, these sequences can be targeted for insertion or the adjacent strands of the protein used to provide additional interactions and recruitment or stabilisation of the complex. In Fig. 17C the solvent accessible, aromatic side-chains of the protein are highlighted with hashing. This analysis allows the flexibility and accessibility of the protein strands to be evaluated and the optimal position of peptide insertion positions to be determined. The most flexible, conserved and exposed loops will present the most opportunity to induce neomorphic interactions and these are therefore suitable for library positions. All of these features together can be used to determine the likely position of inserted peptides that could induce new protein-protein interactions. Therefore, this example demonstrates the positional selection criteria for library insertion for induced neomorphic function according to the present invention.

[0188] Example 17 - Insertion loops in KLHDC2 suitable for neomorphic function induction

[0189] In order to further demonstrate the selection and the position of the library insertions in effector proteins such as CRBN, KLHDC2 and others, molecular modelling of the KLHDC2 protein is presented in Fig. 18. The position of inserted peptide sequences in the enzyme KLHDC2 are highlighted and further simulated for the discovery of induced neomorphic function. In Fig. 18A the crystal structure of the enzyme KLHDC2 is shown with the ligand-binding domain labelled in addition to six positions on the Kelch domain of the protein into which peptide insertions are created and structurally modelled (Loops A, B, C, D, E and F). In Fig. 18B the amino acid sequence of the enzyme is shown (SEQ ID NO. 9) with four of these positions (Loops A (SEQ ID NO. 12) , C (SEQ ID NO. 14), D (SEQ ID NO. 15) and F (SEQ ID NO. 17)) illustrated and in table in Fig. 18C four separate HiBiT sequences are integrated into the native sequence of KLHDC2 at the named positions, either with or without some replacement of the endogenous sequence (SEQ ID NOs. 25-28) in order to accommodate the properties of the protein. In Fig. 18D a protein structural simulation of these HiBiT sequences in the context of the insertion is shown for demonstration of modification of KLHDC2 for functional remodelling with libraries of new peptide sequences. Therefore, this example demonstrates the position criteria of the inserted libraries for effector proteins such as CRBN, KLHDC2 and others using molecular modelling and simulation.

[0190] Example 18 - Modified KLHDC2 has neomorphic function

[0191] To demonstrate that changes to native KLHDC2 and other effector protein function can be induced by library or peptide sequence insertion, KLHDC2 was modified using a demonstration peptide sequence and monitored for neomorphic activity. Introduction of the HiBiT sequence into defined positions in the enzyme KLHDC2 is expected to yield a new binding function consistent with the accessibility of the epitope to the complementary LgBiT fragment. In Fig. 19 this is measured by the luminescence created with the HiBiT and LgBiT when they interact in the presence of ATP and the NanoLuciferase substrate molecule. In the case of all peptide insertion positions, a luminescence signal is measured that is either modestly or significantly higher than the baseline (“No HiBiT”) control, where KLHDC2 lacks any ability to bind LgBiT. In some cases where a binding event between these two proteins (a neomorphic event) also results in a stable and productive ternary complex suitable for ubiquitin transfer then degradation might also ensue. This can be monitored by the increase in luminescence when the cullin complex is inhibited by the drug MLN4924. For the peptide insertions in the body of KLHDC2 (at Loop A, Loop, C, Loop D and Loop F) a degree of increased luminescence is measured, and for the N-terminal sequences (as controls) there is a robust degree of degradation, consistent with published observations of the flexibility of this N-terminal site (see Poirson et al., Nature 2024 volume 628, pages 878-886: Proteome-scale discovery of protein degradation and stabilization effectors). Therefore, this example demonstrates that for nominated effector proteins, such as KLHDC2, the position of inserted peptide sequences can be determined and evaluated by neomorphic function analysis and a new activity can be induced by peptide insertion according to the present invention.

[0192] Example 19 - Modified KLHDC2 is a degradation-competent enzyme

[0193] To further illustrate the suitability of inserted sequences and libraries according to the present invention in different effector proteins, such as KLHDC2, examination of the degradation function of the effector protein was determined. Induced neomorphic function of the enzyme KLHDC2 most likely requires that the enzyme is able to maintain some or all of its native function in addition to the gain-of-function induced by peptide motif insertions according to the present invention. In Fig. 20, KLHDC2 enzyme with new non-native peptide sequences inserted at the indicated locations in the protein were tested for their ability to degrade a natural substrate via a reporter protein tool (GFP-SelK). In each case degradation was observed by the decrease in the abundance of the reporter proteins and the subsequent rescue of the protein abundance by inhibitors of the proteasome and cullin pathways, such as MG 132 and MLN4924. Therefore, this example demonstrates that the effector protein’s native enzymatic capability is maintained to the extent that gain-of-substrate or neomorphic activity can be induced by peptide insertion or peptide library insertion according to the present invention.

[0194] Example 20 - Modified KLHDC2 is recruited by bivalent degraders for degradation

[0195] To illustrate the use of small molecules in combination with the inserted peptide libraries for the discovery of neomorphic function, the effector protein KLHDC2 was used in the presence of bifunctional degraders which recruit this enzyme and the degradation of a target protein assessed. Neomorphic function of a modified KLHDC2 can be used to design and develop small molecule molecular glue degraders which exploit known ligands of the enzyme in the screening process and edit the molecular properties of these ligands to mimic the induced protein-protein interactions created by the inserted peptides. Demonstration that PROTACs based on ligands of the KLHDC2 enzyme are able to functionally bind the enzyme are shown in Fig. 21. The cellular target BRD4 is degraded by the molecule dBET6 in a dose dependent way and is not affected by changes to KLHDC2, as expected. In contrast, the KLHDC2-based PROTACs PMC-0017429 and PMC-0017430 cause the degradation of BRD4 only in presence of KLHDC2. Insertions at the referenced KLHDC2 loop positions or at the N-terminal domain of the KLHDC2 alter the ability of the degrader molecules to function by varying degrees. This informs on the ability of the KLHDC2 to be induced into a functional complex with BRD4 and the impact of the peptide insertion on this phenomenon and therefore demonstrates the ability of the enzyme to be recruited into a functional neomorphic remodelling screen. Therefore, this example demonstrates that the selected positions for insertion maintain the ligand-binding function of the effector protein. It also shows that inserted peptide libraries according to the present invention can be used in combination with small molecule ligands to induce neomorphic function.

[0196] Example 21 - Inserted sequences in effector proteins can be used to model molecular glue degrader drugs

[0197] To illustrate the design of small molecule degraders from the active (hit) peptide sequence insertions according to the present invention, a molecular modelling analysis is shown in Fig. 22. Molecular glue chemical architecture arising from a functional neomorphic degradationinducing peptide insert on CRBN is described. In Fig. 22A an active degrading hit (3 amino acid peptide insertion sequence) is shown as a space filling graphic and modelled in the P1 position in CRBN (Type short - see Fig. 22E). This active hit was found to robustly induce the degradation of GSPT1 in the presence of pomalidomide as shown in Figure 13. GSPT1 protein is shown on the right as a modelled protein-protein interaction which was found to be induced by the inserted sequence (Example 11 and 12). In Fig. 22B detail on a pharmacophore to define the binding site used for virtual screening is shown. In Fig. 22C pharmacophore points based on the hydrogen bonding between the peptide insert and target protein GSPT1 is shown. In Fig. 22D a mesh-based inverse binding site model shows positions of hydrogen bond donors, hydrogen bond acceptors and hydrophobic sites are shown as hashed spheres in the graphic. In Fig. 22E a table shows the position and context in the CRBN protein for the library insertions at position P1 (SEQ ID NOs. 29-32). Hits were identified from all variants of the library composition. Therefore, this example demonstrates the use of chemical properties from the inserted peptide sequence of an identified active mutant that are used to design small molecule molecular glues.

[0198] Sequences

[0199] CRBN wild-type protein sequence (SEQ ID NO. 1)

[0200] MAGEGDQQDAAHNMGNHLPLLPAESEEEDEMEVEDQDSKEAKKPNIINFDTSLPTSHTYL GADMEEFHGRTLHDDDSCQVIPVLPQVMMILIPGQTLPLQLFHPQEVSMVRNLIQKDRTFA VLAYSNVQEREAQFGTTAEIYAYREEQDFGIEIVKVKAIGRQRFKVLELRTQSDGIQQAKVQI LPECVLPSTMSAVQLESLNKCQIFPSKPVSREDQCSYKWWQKYQKRKFHCANLTSWPRW LYSLYDAETLMDRIKKQLREWDENLKDDSLPSNPIDFSYRVAACLPIDDVLRIQLLKIGSAIQR LRCELDIMNKCTSLCCKQCQETEITTKNEIFSLSLCGPMAAYVNPHGYVHETLTVYKACNLN

[0201] LIGRPSTEHSWFPGYAWTVAQCKICASHIGWKFTATKKDMSPQKFWGLTRSALLPTIPDTE DEISPDKVILCL

[0202] CRBN wild-type nucleic acid sequence (SEQ ID NO. 33)

[0203] ATGGCCGGCGAAGGAGATCAGCAGGACGCTGCGCACAACATGGGCAACCACCTGCCG

[0204] CTCCTGCCTGCAGAGAGTGAGGAAGAAGATGAAATGGAAGTTGAAGACCAGGATAGTA

[0205] AAGAAGCCAAAAAACCAAACATCATAAATTTTGACACCAGTCTGCCGACATCACATACAT

[0206] ACCTAGGTGCTGATATGGAAGAATTTCATGGCAGGACTTTGCACGATGACGACAGCTGT

[0207] CAGGTGATTCCAGTTCTTCCACAAGTGATGATGATCCTGATTCCCGGACAGACATTACC

[0208] TCTTCAGCTTTTTCACCCTCAAGAAGTCAGTATGGTGCGGAATTTAATTCAGAAAGATAG

[0209] AACCTTTGCTGTTCTTGCATACAGCAATGTACAGGAAAGGGAAGCACAGTTTGGAACAA

[0210] CAGCAGAGATATATGCCTATCGAGAAGAACAGGATTTTGGAATTGAGATAGTGAAAGTG

[0211] AAAGCAATTGGAAGACAAAGGTTCAAAGTCCTTGAGCTAAGAACACAGTCAGATGGAAT CCAGCAAGCTAAAGTGCAAATTCTTCCCGAATGTGTGTTGCCTTCAACCATGTCTGCAG

[0212] TTCAATTAGAATCCCTCAATAAGTGCCAGATATTTCCTTCAAAACCTGTCTCAAGAGAAG

[0213] ACCAATGTTCATATAAATGGTGGCAGAAATACCAGAAGAGAAAGTTTCATTGTGCAAATC

[0214] TAACTTCATGGCCTCGCTGGCTGTATTCCTTATATGATGCTGAGACCTTAATGGACAGA

[0215] ATCAAGAAACAGCTACGTGAATGGGATGAAAATCTAAAAGATGATTCTCTTCCTTCAAAT

[0216] CCAATAGATTTTTCTTACAGAGTAGCTGCTTGTCTTCCTATTGATGATGTATTGAGAATT

[0217] CAGCTCCTTAAAATTGGCAGTGCTATCCAGCGACTTCGCTGTGAATTAGACATTATGAA

[0218] TAAATGTACTTCCCTTTGCTGTAAACAATGTCAAGAAACAGAAATAACAACCAAAAATGA

[0219] AATATTCAGTTTATCCTTATGTGGGCCGATGGCAGCTTATGTGAATCCTCATGGATATGT

[0220] GCATGAGACACTTACTGTGTATAAGGCTTGCAACTTGAATCTGATAGGCCGGCCTTCTA

[0221] CAGAACACAGCTGGTTTCCTGGGTATGCCTGGACTGTTGCCCAGTGTAAGATCTGTGC

[0222] AAGCCATATTGGATGGAAGTTTACGGCCACCAAAAAAGACATGTCACCTCAAAAATTTT

[0223] GGGGCTTAACGCGATCTGCTCTGTTGCCCACGATCCCAGACACTGAAGATGAAATAAG

[0224] TCCAGACAAAGTAATACTTTGCTTGTAA

[0225] CRBN-201 optimised protein sequence (SEQ ID NO. 34)

[0226] MAGEGDQQDAAHNMGNHLPLLPAESEEEDEMEVEDQDSKEAKKPNIINFDTSLPTSHTYL GADMEEFHGRTLHDDDSCQVIPVLPQVMMILIPGQTLPLQLFHPQEVSMVRNLIQKDRTFA VLAYSNVQEREAQFGTTAEIYAYREEQDFGIEIVKVKAIGRQRFKVLELRTQSDGIQQAKVQI LPECVLPSTMSAVQLESLNKCQIFPSKPVSREDQCSYKWWQKYQKRKFHCANLTSWPRW LYSLYDAETLMDRIKKQLREWDENLKDDSLPSNPIDFSYRVAACLPIDDVLRIQLLKIGSAIQR LRCELDIMNKCTSLCCKQCQETEITTKNEIFSLSLCGPMAAYVNPHGYVHETLTVYKACNLN

[0227] LIGRPSTEHSWFPGYAWTVAQCKICASHIGWKFTATKKDMSPQKFWGLTRSALLPTIPDTE DEISPDKVILCL-

[0228] CRBN-201 optimised nucleic acid sequence (SEQ ID NO. 35)

[0229] ATGGCAGGCGAAGGTGACCAGCAAGACGCCGCTCACAACATGGGGAATCACCTCCCT

[0230] CTGCTGCCCGCGGAGTCTGAAGAAGAGGATGAGATGGAAGTGGAAGACCAAGATTCTA

[0231] AAGAAGCTAAAAAACCCAACATCATAAACTTTGACACGTCTTTGCCCACTTCACATACTT

[0232] ACCTGGGGGCCGACATGGAGGAATTCCATGGCCGAACGCTTCATGATGATGACTCCTG

[0233] TCAGGTGATCCCGGTTCTCCCTCAGGTAATGATGATTCTGATCCCCGGGCAGACATTG

[0234] CCACTTCAACTGTTCCATCCACAGGAGGTAAGCATGGTCAGGAATTTGATTCAGAAAGA

[0235] TAGAACCTTTGCGGTGCTTGCATATTCTAATGTACAAGAGCGCGAAGCCCAATTTGGTA

[0236] CGACGGCGGAGATTTATGCATATAGAGAAGAACAAGATTTCGGGATTGAAATAGTTAAG

[0237] GTCAAGGCGATAGGAAGACAAAGATTTAAAGTTCTGGAGCTGCGAACGCAGAGTGACG

[0238] GCATTCAGCAGGCCAAAGTACAGATCCTTCCTGAATGCGTATTGCCATCTACGATGAGT

[0239] GCTGTGCAGCTGGAGAGTTTGAACAAGTGCCAAATCTTTCCTTCTAAACCCGTGTCACG

[0240] AGAGGACCAGTGCAGCTATAAGTGGTGGCAGAAGTATCAGAAGAGAAAATTCCACTGC

[0241] GCCAACTTGACTTCATGGCCTAGATGGCTCTACTCACTCTACGACGCTGAGACGCTGAT

[0242] GGATCGCATTAAAAAGCAACTTCGAGAATGGGATGAAAATCTGAAAGATGATTCCTTGC

[0243] CTAGTAATCCAATAGATTTTAGTTATCGAGTTGCGGCCTGTCTCCCAATTGACGACGTTC

[0244] TTAGAATCCAGCTCTTGAAAATAGGATCTGCTATACAAAGGCTTCGATGTGAGCTTGATA

[0245] TCATGAACAAGTGCACATCTTTGTGTTGTAAGCAGTGCCAAGAAACCGAAATCACAACC AAAAACGAAATCTTTAGTCTGAGTCTTTGCGGGCCAATGGCCGCATATGTAAACCCTCA

[0246] TGGCTACGTGCATGAAACGCTCACCGTCTATAAGGCATGTAATCTGAATCTCATAGGGA

[0247] GGCCTTCAACTGAGCATTCTTGGTTCCCCGGTTACGCTTGGACAGTAGCGCAGTGTAA

[0248] AATTTGCGCGTCTCACATAGGCTGGAAATTCACTGCTACGAAAAAGGATATGTCCCCGC

[0249] AGAAGTTCTGGGGCTTGACTAGATCAGCTCTTCTGCCGACGATCCCGGATACTGAGGA

[0250] TGAGATATCACCCGACAAAGTCATCTTGTGTTTGTAA

[0251] KBTBD4 wild-type protein sequence (SEQ ID NO. 7)

[0252] MKGGNADSWQREKLASMESPEEPGASMDENYFVNYTFKDRSHSGRVAQGIMKLCLEEEL

[0253] FADVTISVEGREFQLHRLVLSAQSCFFRSMFTSNLKEAHNRVIVLQDVSESVFQLLVDYIYH

[0254] GTVKLRAEELQEIYEVSDMYQLTSLFEECSRFLARTVQVGNCLQVMWLADRHSDPELYTAA

[0255] KHCAKTHLAQLQNTEEFLHLPHRLLTDIISDGVPCSQNPTEAIEAWINFNKEEREAFAESLRT

[0256] SLKEIGENVHIYLIGKESSRTHSLAVSLHCAEDDSISVSGQNSLCHQITAACKHGGDLYVVG

[0257] GSIPRRMWKCNNATVDWEWCAPLPRDRLQHTLVSVPGKDAIYSLGGKTLQDTLSNAVIYY

[0258] RVGDNVWTETTQLEVAVSGAAGANLNGIIYLLGGEENDLDFFTKPSRLIQCFDTETDKCHVK

[0259] PYVLPFAGRMHAAVHKDLVFIVAEGDSLVCYNPLLDSFTRLCLPEAWSSAPSLWKIASCNG SIYVFRDRYKKGDANTYKLDPATSAVTVTRGIKVLLTNLQFVLA

[0260] KLHDC2 wild-type protein sequence (SEQ ID NO. 9)

[0261] ADGNEDLRADDLPGPAFESYESMELACPAERSGHVAVSDGRHMFVWGGYKSNQVRGLY

[0262] DFYLPREELWIYNMETGRWKKINTEGDVPPSMSGSCAVCVDRVLYLFGGHHSRGNTNKFY

[0263] MLDSRSTDRVLQWERIDCQGIPPSSKDKLGVWVYKNKLIFFGGYGYLPEDKVLGTFEFDET

[0264] SFWNSSHPRGWNDHVHILDTETFTWSQPITTGKAPSPRAAHACATVGNRGFVFGGRYRD

[0265] ARMNDLHYLNLDTWEWNELIPQGICPVGRSWHSLTPVSSDHLFLFGGFTTDKQPLSDAWT

[0266] YCISKNEWIQFNHPYTEKPRLWHTACASDEGEVIVFGGCANNLLVHHRAAHSNEILIFSVQP KSLVRLSLEAVICFKEMLANSWNCLPKHLLHSVNQRFGSNNTSGS

[0267] KLHDC2 optimised nucleic acid sequence (SEQ ID NO. 36)

[0268] GCTGATGGCAACGAGGATCTGCGGGCTGACGACTTGCCTGGGCCAGCCTTCGAGAGC

[0269] TATGAGTCCATGGAGCTTGCCTGCCCCGCTGAGCGCAGCGGCCACGTAGCCGTCAGC

[0270] GACGGGCGCCACATGTTCGTCTGGGGCGGCTACAAGAGTAATCAAGTCAGAGGATTAT

[0271] ATGACTTTTATCTGCCTAGAGAAGAACTATGGATCTACAACATGGAGACTGGAAGATGG

[0272] AAAAAAATCAACACTGAAGGTGATGTTCCTCCTTCTATGTCAGGAAGCTGTGCTGTGTG

[0273] TGTAGACAGGGTGCTGTACTTGTTTGGAGGACACCATTCAAGAGGCAATACCAATAAGT

[0274] TCTACATGCTGGATTCAAGGTCTACAGACAGAGTGTTACAGTGGGAAAGAATTGATTGC

[0275] CAAGGAATTCCTCCATCATCAAAGGACAAACTTGGTGTCTGGGTATATAAAAACAAGTTA

[0276] ATATTTTTTGGAGGGTATGGATATTTGCCTGAAGATAAAGTATTGGGAACTTTTGAATTC

[0277] GATGAAACATCTTTTTGGAATTCAAGTCATCCAAGAGGATGGAATGATCATGTACATATT TTAGATACTGAAACATTTACCTGGAGCCAGCCTATAACTACTGGTAAAGCACCTTCACCT CGTGCTGCCCATGCCTGTGCAACTGTCGGAAATAGAGGCTTCGTGTTTGGAGGCAGAT ATCGAGATGCTAGAATGAATGATCTTCACTATCTTAATCTGGATACATGGGAGTGGAAT

[0278] GAATTAATTCCACAAGGCATATGCCCAGTTGGTCGATCTTGGCACTCACTAACACCAGT

[0279] TTCTTCAGATCATCTTTTTCTCTTTGGAGGATTTACCACTGATAAACAGCCACTAAGTGA

[0280] TGCCTGGACTTACTGCATCAGTAAAAATGAATGGATACAATTTAATCATCCATATACCGA

[0281] AAAACCAAGGTTATGGCACACAGCTTGTGCCAGCGATGAAGGAGAAGTAATTGTTTTTG

[0282] GTGGATGTGCCAACAACTTGCTTGTCCATCACAGAGCTGCACACAGTAATGAAATACTA

[0283] ATATTTTCAGTTCAACCAAAATCTCTTGTACGGCTAAGCTTAGAAGCAGTCATTTGCTTT

[0284] AAAGAAATGTTAGCCAACTCATGGAACTGCCTTCCAAAACACTTACTTCACAGTGTTAAT

[0285] CAGAGGTTTGGTAGTAACAACACTTCTGGATCT

[0286] KLHDC2-KK mutant protein sequence (SEQ ID NO. 10)

[0287] ADGNEDLRADDLPGPAFESYESMELACPAERSGHVAVSDGRHMFVWGGYKSNQVRGLY

[0288] DFYLPREELWIYNMETGRWKKINTEGDVPPSMSGSCAVCVDRVLYLFGGHHSRGNTNKFY

[0289] MLDSRSTDRVLQWERIDCQGIPPSSKDKLGVWVYKNKLIFFGGYGYLPEDKVLGTFEFDET

[0290] SFWNSSHPRGWNDHVHILDTETFTWSQPITTGKAPSPRAAHACATVGNRGFVFGGRYRD

[0291] ARMNDLHYLNLDTWEWNELIPQGICPVGRSWHSLTPVSSDHLFLFGGFTTDKQPLSDAWT YCISKNEWIQFNHPYTEKPRLWHTACASDEGEVIVFGGCANNLLVHHRAAHSNEILIFSVQP KSLVRLSLEAVICFKEMLANSWNCLPKHLLHSVNQRFGSNNTSKK

[0292] KLHDC2-KK optimised nucleic acid sequence (SEQ ID NO. 37)

[0293] GCTGATGGCAACGAGGATCTGCGGGCTGACGACTTGCCTGGGCCAGCCTTCGAGAGC

[0294] TATGAGTCCATGGAGCTTGCCTGCCCCGCTGAGCGCAGCGGCCACGTAGCCGTCAGC

[0295] GACGGGCGCCACATGTTCGTCTGGGGCGGCTACAAGAGTAATCAAGTCAGAGGATTAT

[0296] ATGACTTTTATCTGCCTAGAGAAGAACTATGGATCTACAACATGGAGACTGGAAGATGG

[0297] AAAAAAATCAACACTGAAGGTGATGTTCCTCCTTCTATGTCAGGAAGCTGTGCTGTGTG

[0298] TGTAGACAGGGTGCTGTACTTGTTTGGAGGACACCATTCAAGAGGCAATACCAATAAGT

[0299] TCTACATGCTGGATTCAAGGTCTACAGACAGAGTGTTACAGTGGGAAAGAATTGATTGC

[0300] CAAGGAATTCCTCCATCATCAAAGGACAAACTTGGTGTCTGGGTATATAAAAACAAGTTA

[0301] ATATTTTTTGGAGGGTATGGATATTTGCCTGAAGATAAAGTATTGGGAACTTTTGAATTC

[0302] GATGAAACATCTTTTTGGAATTCAAGTCATCCAAGAGGATGGAATGATCATGTACATATT TTAGATACTGAAACATTTACCTGGAGCCAGCCTATAACTACTGGTAAAGCACCTTCACCT CGTGCTGCCCATGCCTGTGCAACTGTCGGAAATAGAGGCTTCGTGTTTGGAGGCAGAT ATCGAGATGCTAGAATGAATGATCTTCACTATCTTAATCTGGATACATGGGAGTGGAAT

[0303] GAATTAATTCCACAAGGCATATGCCCAGTTGGTCGATCTTGGCACTCACTAACACCAGT

[0304] TTCTTCAGATCATCTTTTTCTCTTTGGAGGATTTACCACTGATAAACAGCCACTAAGTGA

[0305] TGCCTGGACTTACTGCATCAGTAAAAATGAATGGATACAATTTAATCATCCATATACCGA

[0306] AAAACCAAGGTTATGGCACACAGCTTGTGCCAGCGATGAAGGAGAAGTAATTGTTTTTG

[0307] GTGGATGTGCCAACAACTTGCTTGTCCATCACAGAGCTGCACACAGTAATGAAATACTA

[0308] ATATTTTCAGTTCAACCAAAATCTCTTGTACGGCTAAGCTTAGAAGCAGTCATTTGCTTT AAAGAAATGTTAGCCAACTCATGGAACTGCCTTCCAAAACACTTACTTCACAGTGTTAAT CAGAGGTTTGGTAGTAACAACACTTCTAAGAAG

Claims

CLAIMS1. A method of identifying a modified effector protein having an induced change in activity, wherein the method comprises modifying the effector protein by inserting a peptide motif or library of peptide motifs into the effector protein and identifying a change in activity, wherein the change in activity is a change in substrate recognition activity, identified by measuring a change to substrate binding activity and / or target protein degradation.

2. A method according to claim 1 , wherein each peptide motif comprises a polypeptide of length 1 to 110 amino acids.

3. A method according to claim 1 or 2, wherein the effector protein is an E3 ligase.

4. A method according to claim 3, wherein the method comprises inserting a library of peptide motifs into a plurality of E3 ligase proteins, creating a library of E3 ligase mutants, wherein the plurality of E3 ligase proteins comprises one or more types of E3 ligase protein (e.g. CRBN, VHL, KBTBD4, KLHDC2), and the insertion is at one or more sites in the E3 ligase protein.

5. A method according to claim 3 or 4, wherein the E3 ligase protein is further modified to optimise it for the method (e.g. KLHDC2-KK mutant).

6. A method according to any preceding claim, wherein the change in activity is binding of a new substrate.

7. A method according to any of claims 1 to 5, wherein the change in activity is improved binding for one or more existing or known substrate proteins.

8. A method according to any preceding claim, wherein the change in activity is measured in the presence of one or more small molecule ligands or drugs.

9. A method according to any preceding claim, wherein the method comprises a cell assay, and wherein the change in activity is identified by a cellular phenotypic response caused by unanticipated substrate binding and / or degradation of a new target protein.

10. A method according to any of claims 3 to 9, wherein the change in substrate recognition activity is a change in the interaction of the E3 ligase and a target protein, optionally via an intermediate accessory protein, by measuring target protein degradation.

11. A method according to any of claims 3 to 10, wherein the change in substrate recognition activity of the E3 ligase induces degradation of a target protein.

12. A method according to any preceding claim, wherein the peptide motif(s) are inserted into an internal domain of the effector protein and not at the N or C terminal.

13. A method according to any preceding claim, wherein the target protein is selected from the list of target proteins in Table 1 or Table 2 or the target protein is associated with a disease, disorder or condition when mutated, expressed or overexpressed in a eukaryotic cell.

14. A method according to any preceding claim, wherein the effector protein is cereblon (CRBN).

15. A method according claim 14, wherein the location of the insertion of the peptide motif is within the region P1 (defined by codons 125-178 (SEQ ID NO. 2)) or region P2 (defined by codons 328 - 379 (SEQ ID NO. 3)) of the wildtype cereblon protein (according to SEQ ID NO. 1).

16. A method according to any of claims 3 to 12 wherein the E3 ligase is KBTBD4; optionally wherein the location of the insertion is in the Kelch domain of the protein KBTBD4; further optionally wherein the location of the insertion is in the KBTBD4 Kelch motif region defined by codons 308-313 (SEQ ID NO. 8).

17. A method according to any of claims 3 to 12, wherein the E3 ligase is KLHDC2 or KLHDC2-KK; optionally wherein the location of the insertion is in the Kelch domain of the wild-type protein KLHDC2 (SEQ ID NO. 11); further optionally, wherein the location of the insertion of the peptide motif is within Loop A (Loop A is defined by codons 46-66 (SEQ ID NO. 12) or within Loop C (Loop C is defined codons 159- 194 (SEQ ID NO. 14)) or within Loop D (Loop D is defined by codons 232-244 (SEQ ID NO. 15)) or within Loop F (Loop F is defined by codons 334-353 (SEQ ID NO. 17) of the wildtype KLHDC2 protein (according to SEQ ID NO. 9).

18. A method according to any preceding claim, wherein the method comprises screening a population of mammalian cells comprising a library of effector proteins (e.g. E3 ligases) with peptide motif insertions.

19. A method according to any of claims 3 to 18, wherein a modified E3 ligase with an induced change of activity is identified as an active mutant, and optionally the active mutant is isolated.

20. A method according to any preceding claim, wherein a modified effector protein with an induced change of activity is identified, and wherein a phenotypic measurement and next generation sequencing is used to identify an active insertion mutation sequence; optionally wherein the method comprises; a. exposing a population of in vitro cultured mammalian cells capable of displaying said phenotype to a library of modified effector proteins modified according to any of the preceding claims,b. identifying in said cell population an alteration in said phenotype following said exposure, c. selection of said cells undergoing the phenotypic change to obtain a harvested cell population, d. sequencing of the harvested cell population by next generation sequencing to identify an enriched or depleted peptide motif insertion in the harvested population and to identify an active insertion mutation sequence.

21. A method of identifying an E3 ligase:Protein of Interest (POI) interface suitable for monovalent degrader discovery / development, comprising a method according to any preceding claim, wherein the effector protein is an E3 ligase; optionally wherein the E3 ligase is CRBN or KLHDC2.

22. A method of identifying a biological therapeutic agent that can reduce the amount of a target protein in a cell, the method comprising a method of screening according to claim 19 and identifying active mutants that induce target protein degradation for use as a therapeutic agent; optionally wherein the therapeutic agent reduces the amount of target protein in a cell in the presence of a small molecule degrader; and optionally wherein the active mutant is an active cereblon mutant or an active KLHDC2 mutant.

23. A method of changing the substrate recognition activity of an effector protein, the method comprising: making a mutant effector protein by inserting a peptide motif sequence of 1 to 110 amino acids into a location within or proximal to a substrate recognition interface of the wildtype effector protein; screening the mutant effector protein for a change in substrate recognition activity; optionally wherein the change induces degradation of a target protein; selecting an active mutant with changed activity.

24. A method according to claim 23, wherein the method comprises making a mutant cereblon protein by inserting a peptide sequence of 1 to 110 amino acids into the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wildtype cereblon protein.

25. A method of engineering cereblon to change its substrate recognition activity, wherein the method comprises inserting a peptide sequence of 1 to 110 amino acids at the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wildtype cereblon protein.

26. A mutant cereblon protein, wherein the mutant protein comprises a peptide sequence of 1 to 110 amino acids inserted into the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wild-type protein cereblon.

27. A library of mutant cereblon proteins, wherein each mutant protein comprises a peptide sequence of 1 to 110 amino acids inserted into the location P1 (region defined by codons 125-178 (SEQ ID NO. 2)) or P2 (region defined by codons 328 - 379 (SEQ ID NO. 3)) of the wild-type protein cereblon.

28. A population of mammalian cells comprising a library of cereblon mutants, wherein the library is as defined in claim 27.

29. A method of screening for a change in activity in a cereblon protein, wherein the method comprises using the population of cells of claim 28 or a library of cereblon mutant proteins according to claim 27; and identifying active cereblon mutants.

30. A method of screening according to claim 29, wherein the change in activity is new cereblon substrate recognition activity, and the new substrate recognition activity induces target protein degradation, optionally wherein the change in activity is identified by measuring resulting cellular phenotypic effects.31 . A method of identifying an Cereblon: Protein of Interest (POI) interface suitable for monovalent degrader discovery / development, comprising a method of screening according to claim 29 or 30.

32. A method according to claim 23, wherein the method comprises making a mutant KLHDC2 protein by inserting a peptide sequence of 1 to 110 amino acids within Loop A, C, D or F of the wildtype KLHDC2 protein; wherein Loop A is defined by codons 46-66 (SEQ ID NO. 12), Loop C is defined by codons 159-194 (SEQ ID NO. 14), Loop D is defined by codons 232-244 (SEQ ID NO. 15) and Loop F is defined by codons 334-353 (SEQ ID NO. 17) of the wildtype KLHDC2 protein (according to SEQ ID NO. 9).

33. A method of engineering KLHDC2 to change its substrate recognition activity, wherein the method comprises inserting a peptide sequence of 1 to 110 amino acids within Loop A, C, D or F of the wildtype KLHDC2 protein; wherein Loop A is defined by codons 46-66 (SEQ ID NO. 12), Loop C is defined by codons 159-194 (SEQ ID NO. 14), Loop D is defined by codons 232-244 (SEQ ID NO. 15) and Loop F is defined by codons 334-353 (SEQ ID NO. 17) of the wildtype KLHDC2 protein (according to SEQ ID NO. 9).

34. A mutant KLHDC2 protein, wherein the mutant protein comprises a peptide sequence of 1 to 110 amino acids inserted within Loop A, C, D or F of the wildtype KLHDC2 protein; wherein Loop A is defined by codons 46-66 (SEQ ID NO. 12),Loop C is defined by codons 159-194 (SEQ ID NO. 14), Loop D is defined by codons 232-244 (SEQ ID NO. 15) and Loop F is defined by codons 334-353 (SEQ ID NO. 17) of the wildtype KLHDC2 protein (according to SEQ ID NO. 9).

35. A library of mutant KLHDC2 proteins, wherein each mutant protein comprises a peptide sequence of 1 to 110 amino acids inserted within Loop A, C, D or F of the wildtype KLHDC2 protein; wherein Loop A is defined by codons 46-66 (SEQ ID NO. 12), Loop C is defined by codons 159-194 (SEQ ID NO. 14), Loop D is defined by codons 232-244 (SEQ ID NO. 15) and Loop F is defined by codons 334-353 (SEQ ID NO. 17) of the wildtype KLHDC2 protein (according to SEQ ID NO. 9)..

36. A population of mammalian cells comprising a library of KLHDC2 mutants, wherein the library is as defined in claim 35.

37. A method of screening for a change in activity in a KLHDC2 protein, wherein the method comprises using the population of cells of claim 36 or a library of KLHDC2 mutant proteins according to claim 35; and identifying active KLHDC2 mutants.

38. A method of screening according to claim 37, wherein the change in activity is new KLHDC2 substrate recognition activity, and the new substrate recognition activity induces target protein degradation, optionally wherein the change in activity is identified by measuring resulting cellular phenotypic effects.

39. A method of identifying an KLHDC2:Protein of Interest (POI) interface suitable for monovalent degrader discovery / development, comprising a method of screening according to claim 37 or 38.

Images