WO2017149002A1

WO2017149002A1 - Novel fusion proteins for capturing of cells

Info

Publication number: WO2017149002A1
Application number: PCT/EP2017/054744
Authority: WO
Inventors: Ulrich Haupts; Erik Fiedler; Eva BOSSE-DOENECKE; Markus LIEBSCHER; Manja Gloser
Original assignee: Navigo Proteins Gmbh
Priority date: 2016-03-03
Filing date: 2017-03-01
Publication date: 2017-09-08

Abstract

The present invention relates to novel fusion proteins for capturing of cells. The invention features fusion proteins comprising a non-lg protein that is capable of binding a cell-surface target, at least one cleavage site for the release of the cells attached to fusion proteins, at least one peptide linker covalently connecting the non-lg protein and a cleavage site, and a coupling moiety enabling attachment of the fusion protein to a solid support. The invention also relates to the use of these fusion proteins in applications for capturing cells and to methods of capturing cells. The invention further relates to nucleic acid molecules encoding these non-lg fusion proteins, to vectors comprising these nucleic acid molecules, and to host cells comprising these proteins, nucleic acid molecules, or vectors. The invention relates to methods for the production of these fusion proteins.

Description

Novel fusion proteins for capturing of cells

Field of the Invention

The present invention generally relates to novel fusion proteins for isolation or capturing of cells or molecules. More specifically, the invention especially features fusion proteins for capturing of cells expressing a cell-surface target comprising a non-lmmunoglobulin (Ig) protein that is capable of binding a cell-surface target expressed on cells. The fusion protein further comprises at least one cleavage site for the release of the cells bound to the fusion proteins, at least one peptide linker, and a coupling moiety enabling attachment of the fusion protein to a solid support. The invention also relates to the use of non- Ig proteins such as ubiquitin muteins (Affilin^®) for capturing cells from liquids and releasing them from the solid support. The invention further relates to a method of capturing cells, in particular circulating tumor cells (CTC), from a liquid and releasing them from the solid support by cleaving a predefined cleavage site.

Background of the Invention

Cancer remains one of the world's most devastating diseases. Until 2030, the WHO expects more than 21 million new cases every year. Circulating tumor cells (CTCs) are known to seed secondary tumor formation at sites distant from the primary tumors, known as metastasis. CTCs are promising diagnostic and prognostic markers for monitoring cancer progression and anti-cancer treatment responses.

Research efforts on diagnosis and prognosis of metastatic cancer have concentrated on the detection of CTCs in blood samples of patients which is a challenge since CTCs are extremely rare - estimated to be in the range of one tumor cell in the background of 10⁶-10⁹ normal blood cells. Known procedures for capturing cells comprise the coupling of antibodies (immunoglobulins) directed against specific cell- surface markers to a support which serves as a capturing device for CTCs.

However, cells captured via antibody approaches are tightly bound to a solid support of the capturing device which has several disadvantages. One problem is that these cells cannot be gently released from a solid support due to the tight binding of the antibodies to the cells. A disadvantage of cells immobilized to a support is that such cells are not easily accessible for further analysis or purification of cells.

Relatively harsh conditions are used for the unspecific release of captured cells from the support which render the cells unsuitable for further propagation and analysis. A further problem is that the binding of antibodies to cells is often unspecific due to the binding of the constant part of antibodies to undesired cells. An additional problem of antibodies is the possible influence on signaling cascades since receptors for the Fc part of antibodies are found on numerous cells. Upon binding of an antibody to an Fc receptor, cells are activated and signal cascades are generated which might induce severe side effects.

Since current technologies have several disadvantages, there is an urgent need for more effective and gentle methods to capture and release CTCs that are suitable for further uses and applications. More effective methods have the potential to enable the detection and analysis of CTCs from patients with known and unknown metastatic cancer. Due to major limitations in current technologies using antibody binding to capture cells (CTCs), there is an urgent requirement for novel, specific fusion proteins with improved properties to be used in methods for capturing cells.

The present inventors constructed artificial fusion proteins that are particularly well-suited for capturing cells but overcome the disadvantages of the antibody approaches. Thus, it is an objective of the present invention to provide proteins with high specific binding to cells and allowing a gentle and specific release of cells bound by the proteins.

The present invention solves the disadvantages of the antibody approaches by providing novel fusion proteins comprising small engineered non-lg proteins. The fusion proteins comprise non-lg molecules with specificity for surface exposed proteins on cells that are to be captured. It is important that the fusion proteins for affinity purification of cells do not comprise an IgG-Fc part prone to undesired cell interactions. In addition to a specific non-lg targeting moiety, the fusion proteins comprise at least one cleavage site for gentle and specific release of cells bound to the fusion proteins. Preferred fusion proteins comprise at least one peptide linker of sufficient length allowing accessibility of proteases to the cleavage site and thereby allowing a specific and mild release of immobilized cells by proteases. Further, the fusion proteins contain a coupling moiety for the attachment of the fusion protein to a solid support.

The above overview does not necessarily describe all problems solved by the present invention.

Summary of the Invention

In one aspect the present invention relates to a fusion protein for capturing of cells expressing a cell- surface target comprising a non-lmmunoglobulin (Ig) protein that is capable of binding a cell-surface target expressed on cells with a dissociation constant K_D of 1 μΜ or less, and at least one cleavage site for the release of the cells bound to the fusion proteins, and a peptide linker consisting of 5 to 50 amino acids covalently connecting the non-lg protein and a cleavage site, and a coupling moiety enabling attachment of the fusion protein to a solid support comprising a reactive chemical group which specifically forms a covalent bond with a chemical group to enable coupling of the fusion protein to a solid support, selected from an electrophilic group, nucleophilic group, redox-active group, or a group enabling addition reactions, or cyclic addition reactions or click reactions, or combinations thereof. Preferably, the cleavage site is a protease cleavage site, preferably a sequence specific cleavage site, even more preferred a viral protease cleavage site. Preferably, the fusion protein comprises a second peptide linker covalently connecting the protease cleavage site and the coupling moiety, preferably wherein the first and the second peptide linker independently consist of 5 to 50 amino acids, preferably selected from Gly, Ala, Pro, or Ser. Preferably, the coupling moiety is a cysteine.

In another aspect the present invention relates to a non-lg protein wherein the non-lg protein is selected from the group of muteins of ubiquitin, ankyrin repeat protein, lipocalin, Z-domain of staphylococcal protein A, Fyn SH3 domain, tenth domain of human fibronectin, Kunitz domains of various protease inhibitors, Sac7d, multimerized Low Density Lipoprotein Receptor-A, chagasin scaffold or chagasin-like protease inhibitor proteins, fibronectin, FN3 domain, cysteine-knot miniprotein, Armadillo-repeat protein, tetranectin, C-type lectin domain, or CTLA4. In preferred embodiments, the non-lg protein is a ubiquitin mutein which exhibits 80 % to 94 % identity to ubiquitin (SEQ ID NO: 1 ) or 80 % to 94 % identity to bis- ubiquitin of SEQ ID NO: 2.

Another aspect of the present invention relates to the use of the fusion protein in applications for capturing of cells, for example, in diagnostic applications.

A further aspect of the present invention relates to a method of capturing cells comprising the steps of providing a solid support comprising an immobilized fusion protein of any one of the other aspects, providing a liquid sample containing cells, contacting said liquid sample and said solid support, wherein said immobilized fusion protein binds to a specific cell-surface target expressed on cells; and cleaving said immobilized cells from the solid support by protease cleavage thereby obtaining an eluate containing said cells expressing the specific cell-surface target.

In a further aspect, the present invention relates to a nucleic acid molecule encoding the fusion protein, a vector comprising said nucleic acid molecule, and a host cell or a non-human host comprising said fusion protein, said nucleic acid, or said vector.

In another aspect the present invention relates to a method of production of the fusion protein of the invention, comprising the step(s): culturing of the host cell comprising the fusion protein under suitable conditions for the expression of the fusion protein; and optionally isolating said fusion protein.

This summary of the invention does not necessarily describe all features of the present invention. Other embodiments will become apparent from a review of the ensuing detailed description.

Brief Description of the Figures

Figure 1 shows a schematic drawing of a fusion protein of the invention. The arrow illustrates the protease that specifically cleaves at the protease cleavage site.

Figure 2 shows the immobilization of fusion proteins on beads and binding of target to the fusion protein. EGFR (144893) or Her2 (148162) specific fusion proteins immobilized on Dynabeads bind specifically to the corresponding target. The black columns 144893 and 148162 show beads incubated with hEGFR-Fc or hHer2-Fc, respectively, detected with anti-human-lgGFc-HRP or column 148161 (control protein) shows beads incubated with anti-Ubi-Fab-HRP. 148161 : control protein (ubiquitin instead of Affilin as targeting moiety), w/o: only beads without protein. The light grey columns show beads incubated with anti-human-lgGFc-HRP but without target.

Figure 3 shows the binding of EGFR-positive cells A431 and A549 to EGFR-specific fusion protein immobilized on beads and very weak binding to HEK293 cells. Figure 4 shows the cleavage and recovery of cells from Dynabeads. Cells incubated with immobilized EGFR-specific fusion protein on beads, treated with protease (1 ) or without protease (2); cells incubated with immobilized control fusion protein on beads, treated with protease (3) or without protease (4). Figure 4A shows the recovery of cells after TVMV protease treatment. Figure 4B shows beads with cells specifically bound to the fusion proteins before protease treatment confirming specific binding of cells to the immobilized fusion protein (lined bar). Beads after protease treatment are illustrated as black bar. Figure 4C shows the microscopic analysis of binding of EGFR-positive cells to beads coupled with EGFR specific fusion protein and complete release of cells from beads after protease treatment.

Figure 5 shows coupling, cleavage and recovery of cells. Figure 5A shows the coupling of PD1 -specific fusion proteins to beads. Column 1 (151318): coupling of PD1-fusion protein 151318 to beads. Column 2 (148161 ): coupling of control fusion protein 148161 to beads. Column 3 (w/o): beads without coupled fusion protein. Figure 5B shows the recovery of cells after treatment with TVMV protease. 1 , 3 - protease treatment, 2, 4 - no protease treatment, 1 , 2 target specific fusion protein immobilized on beads, 3, 4 beads immobilized with control protein. The suspension after protease treatment containing released cells is shown as dark grey bar. Figure 5C shows beads with cells specifically bound to the fusion proteins before protease treatment confirming specific binding of cells to the immobilized fusion protein (lined bars). Samples after protease treatment are illustrated as black bars, containing cells on beads. Figure 5D shows the microscopic analysis. Shown are EL4-cells (first row) or K562-cells (second row) incubated with fusion protein 151318 coated beads after TVMV protease treatment (first column) or without protease treatment (second column), and incubated with control beads after TVMV protease treatment (third column) or without protease treatment (fourth column).

Figure 6 shows the capturing of EGFR-expressing cells from a mixed cell population using an EGFR- specific fusion protein coupled to beads. (1) beads with immobilized target specific fusion protein and bound cells and (2) supernatant.

Figure 7 shows the coating of plates with different protein concentrations of fusion proteins (144893, referred to as 1 , control 148161 , referred to as 2) for 15 minutes to at least 12 hours (overnight/on). Figure 7A. Coupling of fusion proteins to streptavidin plates was confirmed by incubation with hEGFR-Fc followed by anti-human-lgG-HRP. Figure 7B. Coupling of fusion proteins to streptavidin plates was confirmed by incubation with anti-Ubi-Fab-HRP.

Figure 8. Binding of cells to plates and release of cells from plates. Figure 8A confirms the binding of cells to EGFR-specific fusion proteins immobilized on streptavidin plates and almost complete release of viable cells from the plates after protease digestion. Columns show immobilized fusion protein 144893 or control fusion protein 148161 binding to A549-cells or HEK293-cells (medium grey bars), supernatant of A549-cells and HEK293-cells after protease digestion (black bars), immobilized cells after protease digestion (spotted bars) and immobilized cells without protease treatment (lined bars). Figure 8B confirms the binding of EGFR-specific fusion protein (144893) immobilized on plates to EGFR-positive cells (A549; first column, first row) and no binding of A549-cells on plates immobilized with control protein

148161(second row). The cells are almost completely released from the plates after protease treatment (A549+TVMV; second column, first row) whereas all cells remain bound to the plates after PBS treatment (A549+PBS; third column, first row). Figure 8C shows weak binding of fusion protein 144893 immobilized on streptavidin plates to HEK293 (first column, first row) or no binding of HEK293-cells on plates immobilized control protein (148161)

Figure 9. Binding of cells to plates and release of cells by protease treatment. Figure 9A confirms the binding of Her2-positive cells to Her2-specific fusion protein 148162 immobilized on plates. The columns show immobilized Her2-specific fusion protein 148162 or control protein 148161 binding to SkBr3 cells or HEK293-cells (1 , medium grey bars), supernatant after protease cleavage (2, black bars), immobilized cells after protease digestion (3, spotted bars), and immobilized cells without protease treatment (4, lined bars). Figure 9B confirms the binding of fusion protein 148162 immobilized on streptavidin plates to Her2- positive cells SkBr3 (SkBr3; first column). The cells are nearly completely released from the streptavidin plates after protease treatment (SkBr3+TVMV; second column) whereas all cells are bound to the plates after PBS treatment (SkBr3+PBS; third column). Control fusion protein (148161) with no specificity to Her2 does not bind to SkBr3. HEK293 show no binding to 148162 or 148161 -coupled plates with immobilized fusion proteins (148162 or 148161).

Figure 10. Binding of cells to Maleimide-activated plates. Figure 10A shows coupling of EGFR-specific fusion protein 144893 to Maleimide-activated plates at different concentrations (1 , 0.5, 0.1 , 0.05, and 0 μΜ). Figure 10B shows binding of A549-cells to different concentrations of immobilized fusion protein 144893 on Maleimide-activated plates.

Figure 11 shows the specific isolation of cells by Affilin fusion protein-coupled beads from a blood-cell- mix. Figure 11 A shows the separation of green fluorescent PD1 -positive EL4 cells from blood by 151318- coupled beads in column 1. Only very few PD1-negative K562-cells sporadically bound to 151318-fusion protein-beads (column 2). No binding of cells on 148161-cou pled beads was observed (columns 3 and 4). Figure 11 B shows cells isolated from diluted blood (1 :2). Column 1 shows the specific isolation of PD1- positive cells by using 151318-coupled beads. No binding was observed with control beads (coupled with ubiquitin, referred to as 148161 ; columns 3 and 4) or of PD1-negative K562-cells with 151318-coupled beads (column 2).

Detailed Description of the Invention

Definitions

Before the present invention is described in detail below, it is to be understood that this invention is not limited to the particular methodology, protocols and reagents described herein as these may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention which will be limited only by the appended claims. Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art to which this invention belongs. Preferably, the terms used herein are defined as described in "A multilingual glossary of biotechnological terms: (lUPAC Recommendations)", Leuenberger, H.G.W, Nagel, B. and Kolbl, H. eds. (1995), Helvetica Chimica Acta, CH-4010 Basel, Switzerland).

Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and "comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

Several documents (for example: patents, patent applications, scientific publications, manufacturer's specifications, instructions, GenBank accession number sequence submissions etc.) are cited throughout the text of this specification. Nothing herein is to be construed as an admission that the invention is not entitled to antedate such disclosure by virtue of prior invention. Some of the documents cited herein are characterized as being "incorporated by reference". In the event of a conflict between the definitions or teachings of such incorporated references and definitions or teachings recited in the present specification, the text of the present specification takes precedence.

Sequences: All sequences referred to herein are disclosed in the attached sequence listing that, with its whole content and disclosure, is a part of this specification.

The term "about", as used herein, encompasses the explicitly recited amounts as well as deviations therefrom of ± 15 %. More preferably, a deviation of ±10 %, and most preferably of ± 5 % is encompassed by the term "about".

The term "naturally occurring" as used herein, as applied to an object refers to the fact that an object can be found in nature. For example, a polypeptide or polynucleotide sequence that is present in an organism that can be isolated from a source in nature and which has not been intentionally modified by man in the laboratory is naturally occurring.

In contrast thereto, the terms "non-natural" or "artificial" as used herein interchangeably refer to an object that is not naturally occurring, i.e. the term refers to an object that has been created, produced, or modified by man. For example, a polypeptide or polynucleotide sequence that has been intentionally modified or generated by man in a laboratory is "non-natural".

The terms "protein" and "polypeptide" refer to any chain of two or more amino acids linked by peptide bonds, and do not refer to a specific length of the product. Thus, "peptides", "protein", "amino acid chain," or any other term used to refer to a chain of two or more amino acids, are included within the definition of "polypeptide," and the term "polypeptide" may be used instead of, or interchangeably with any of these terms. The term "polypeptide" is also intended to refer to the products of post-translational modifications of the polypeptide, including without limitation glycosylation, acetylation, phosphorylation, amidation, proteolytic cleavage, modification by non-naturally occurring amino acids and similar modifications which are well known in the art. Thus, fusion proteins comprising two or more protein moieties also fall under the definition of the term "protein" or "polypeptides". The term "fusion protein" relates to a protein comprising at least a first amino acid chain joined genetically to at least a second amino acid chain. Thus, a fusion protein may comprise a multimer of

proteins/peptides which are expressed as a single, linear polypeptide. It may comprise one, two, three, four or even more proteins/peptides. For example, fusion protein can be created through joining of two or more genes that originally coded for separate proteins/peptides.

The term "fused" means that the components are linked by peptide bonds, either directly or via peptide linkers.

Throughout this specification, the term "non-immunoglobulin protein" is often abbreviated as "non-lg protein". Occasionally, both the long form and the abbreviated form are used at the same time, e.g. in the expression "non-immunoglobulin (Ig) protein". It is of importance that the fusion protein of the invention does not comprise an antibody or a fragment thereof, in particular, that the targeting moiety of the fusion protein of the invention does not comprise an immunoglobulin fold as present in antibodies. The non-lg protein of the invention is an artificial protein not existing in nature.

A non-lg protein of the invention is considered to bind specifically to a target protein. As used herein, the terms "bind specifically", "specifically bind", and "specific binding" are understood to mean that the non-lg protein has a selective binding affinity for a particular cell-surface protein with a dissociation constant K_D of 1 μΜ (10^"6 M) or less, more preferably 100 nM (10^"7 M) or less, more preferably 10 nM (10^"8 M) or less, more preferably 1 nM (10^"9 M) or less, more preferably 100 pM (10^"10 M), or more preferably 10 pM (10^"11 M) or less. A high affinity corresponds to a low value of K_D. Appropriate controls as known in the art can be used to distinguish between "specific" and "non-specific" binding.

The term "dissociation constant" or "K_D" defines the specific binding affinity. As used herein, the term "K_D" (usually measured in "mol/L", sometimes abbreviated as "M") is intended to refer to the dissociation equilibrium constant of the particular interaction between a first compound and a second compound. In the context of the present invention, the term K_D is particularly used to describe the binding affinity between a non-immunoglobulin binding protein and a target protein. The dissociation constant K_D can be determined by ELISA or by surface plasmon resonance assays. Typically, the dissociation constant K_D is determined at 20°C, 25°C, or 30°C. If not specifically indicated otherwise, the K_D values recited herein are determined at 25°C by surface plasmon resonance.

The terms "protein capable of binding" or "binding protein" refer to a non-lg protein capable of binding to a surface expressed target protein (e.g. a tumor specific protein; a protein expressed on the surface of a CTC). The binding proteins used in this invention can therefore specifically bind to surface expressed protein of a CTC. Any such binding protein based on a non-lg protein may comprise additional protein domains such as, for example, multimerization moieties, polypeptide tags, polypeptide linkers and/or non- proteinaceous polymer molecules. Some examples of non-proteinaceous polymer molecules are hydroxyethyl starch, polyethylene glycol, polypropylene glycol, or polyoxyalkylene.

The term "Affilin^®" (registered trademark of Scil Proteins GmbH) refers to non-immunoglobulin (Ig) derived binding proteins based on ubiquitin muteins. The terms "Affilin" and„ubiquitin mutein" and ..modified ubiquitin" are all used synonymously and can be exchanged. The terms as used herein refer to derivatives of ubiquitin which differ from unmodified ubiquitin (for example, SEQ ID NO: 1 ) or bis-ubiquitin (for example, SEQ ID NO: 2) by amino acid exchanges, insertions, deletions, or any combination thereof, provided that the Affilin has a specific binding affinity to a target which is at least 10fold lower or absent in unmodified ubiquitin or bis-ubiquitin. This functional property of an Affilin is a de novo created property. An Affilin is not a natural ubiquitin existing in or isolated from nature. An Affilin molecule according to this invention comprises or consists of at least one modified ubiquitin moiety or two modified ubiquitin moieties linked together in a head-to-tail fusion. A "head-to-tail fusion" is to be understood as fusing two ubiquitins together by connecting them in the direction (head) N-C-N-C- (tail), as described for example in EP2379581 B1 which is incorporated herein by reference. Ubiquitin moieties may be connected directly without any linker or with peptide linkers.

The terms "ubiquitin" or ..unmodified ubiquitin" refer to ubiquitin in accordance with SEQ ID NO: 1 (wild type ubiquitin) or to proteins with at least 95 % amino acids identity to SEQ ID NO: 1 (for example, with point mutations F45W, G75A, G76A which do not influence binding to a target, see for example SEQ ID NO: 3). Particularly preferred are ubiquitins from mammals, e.g. humans, primates, pigs, and rodents. On the other hand, it should be noted that the ubiquitin origin is not of high importance since according to the art all eukaryotic ubiquitins are highly conserved and the mammalian ubiquitins examined up to now are even identical with respect to their amino acid sequence. In this sense, ubiquitin from any other eukaryotic source can be used for further modifications to generate a novel binding capability. For instance ubiquitin of yeast differs only in three amino acids from the wild-type human ubiquitin (SEQ ID NO: 1).

The term "bis-ubiquitin" refers to a linear protein wherein two ubiquitin moieties are directly fused to each other in head to tail orientation. The term "bis-ubiquitin" refers to SEQ ID NO: 2 or to proteins with at least 95 % amino acids identity to SEQ ID NO: 2 (for example, with point mutations F45W, G75A, G76A, G151A, G152A).

The term "binding" according to the invention preferably relates to a specific binding. "Specific binding" means herein that a non-lg protein binds stronger to a target such as an epitope for which it is specific, compared to the binding to another molecule. Preferably the dissociation constant (K_D) for the target to which the compound binds specifically is more than 10-fold, preferably more than 20-fold, more preferably more than 50-fold, even more preferably more than 100-fold, 200-fold, 500-fold, or 1000-fold lower than the dissociation constant (K_D) for the molecule to which the binding moiety does not bind specifically.

As used herein, "substitutions" are defined as exchanges of an amino acid by another amino acid. Given the known genetic code, and recombinant and synthetic DNA techniques, the skilled scientist can readily construct DNAs encoding the amino acid variants. The term "insertions" comprises the addition of amino acids to the original amino acid sequence wherein the original amino acid remains stable without significant structural change. The term "deletion" means that one or more amino acids are taken out of the original sequence and the amino acids originally N-terminal and C-terminal of the deleted amino acid are now directly connected and form a continuous amino acid sequence. The term "functional variant" of a protein means herein a variant protein, wherein the function, in relation to the invention defined as affinity, is essentially retained. Thus, one or more amino acids not relevant for said function may have been exchanged.

The term "amino acid sequence identity" refers to a quantitative comparison of the identity (or differences) of the amino acid sequences of two or more proteins. "Percent (%) amino acid sequence identity" with respect to a reference polypeptide sequence is defined as the percentage of amino acid residues in a sequence that are identical with the amino acid residues in the reference polypeptide sequence, after aligning the sequences and introducing gaps, if necessary, to achieve the maximum percent sequence identity. To determine the sequence identity, the sequence of a query protein is aligned to the sequence of a reference protein. Methods for alignment are well known in the art. For example, the SIM Local similarity program is preferably employed (Xiaoquin Huang and Webb Miller (1991), Advances in Applied Mathematics, vol. 12: 337-357), that is freely available (see also: http://www.expasy.org/tools/sim- prot.html). For multiple alignment analysis ClustalW is preferably used (Thompson et al. (1994) Nucleic Acids Res., 22(22): 4673-4680).

Each amino acid of the query sequence that differs from the reference amino acid sequence at a given position is counted as one difference. An insertion or deletion in the query sequence is also counted as one difference. For example, an insertion of a linker between two ubiquitin moieties is counted as one difference compared to the reference sequence. The sum of differences is then related to the length of the reference sequence to yield a percentage of non-identity. The quantitative percentage of identity is calculated as 100 minus the percentage of non-identity. In specific cases of determining the identity of ubiquitin muteins aligned against unmodified ubiquitin, differences in positions 45, 75 and/or 76 are not counted, in particular, because they are not relevant for the novel binding capability of the ubiquitin mutein but are only modifications relevant for biochemical reasons.

As used herein, the term "linker" refers to a moiety that connects a first amino acid chain with at least one further component, preferably a second amino acid chain or a chemical moiety. Preferred embodiments of this invention comprise peptide linkers. For example, a peptide linker is an amino acid sequence that connects a first amino acid chain (protein/peptide) with a second amino acid chain (protein/peptide) via peptide bonds to generate a single, linear polypeptide chain.

As used herein, the term "mixed cell population" refers to a mixture of different cell types in a liquid or semi-solid medium. A mixed cell population comprises more than one (multiple) cell types in a liquid or semi-solid medium. Examples for mixed cell populations are blood or tissue homogenates (for example, a biopsy specimen). For example, tumor tissues may contain a mixture of tumor cells, normal tissue cells, vascular cells supporting tumor growth, immune cells, and possibly other cells.

Circulating tumor cells, also referred to as "CTC", are cells that are detached from tumors and are circulating in the blood of an organism. CTCs have the potential of forming new metastatic tumors.

In the present specification, the terms "target" and "binding partner" are used synonymously and can be exchanged. A target is any protein (e.g. antigen) capable of binding with an affinity as defined herein to the non-lg protein. Preferred target molecules are tumor antigens, such as proteins or epitopes that are present on the outside of a tumor cell but that are absent or minor expressed on non-tumor cells or which are present in tumor tissue but absent or rare on normal tissue.

In the present specification, the terms "immobilization" and "attachment" and "coupling" or "immobilize", "attach", or "couple" are all used synonymously and can be exchanged.

The term "solid support" or "solid carrier" may be any solid or semi-solid insoluble surface suitable for immobilizing the fusion protein of the invention. Attachment of the fusion protein to the solid support can be covalent or non-covalent. Non-covalent interactions can be mediated by electrostatic, π-effects, van der Waals forces, and hydrophobic effects. Covalent attachment is mediated through a chemical reaction between reactive groups on the surface of the solid support and the fusion protein (for example, biotinylated fusion protein) of the invention which form a covalent bond.

The term "coupling moiety" as employed herein means a reactive chemical group that is capable of reacting with other chemical groups to couple the fusion protein to a solid support. Furthermore, the coupling moiety may comprise such a reactive chemical group which forms a covalent or non-covalent bond with chemical groups to enable coupling of the fusion protein to a solid support, amongst other parts. The coupling moiety of the fusion protein reacts only with chemical groups to couple the fusion protein or enable coupling of the fusion protein to a solid support. Reactive chemical groups can be selected from an electrophilic group, nucleophilic group, redox-active group, or a group enabling addition reactions or cyclic addition reactions or click reactions or combinations thereof.

The term "click reaction" is known to those skilled in the art and comprises a number of chemical reactions which are characterized by the following criteria: a) reaction conditions are compatible with biological material, i.e. proteins and even live cells; b) reactions run in aqueous phases c) the reactions proceed to near completion, i.e. they are highly exothermic; d) reactions are highly specific and do not lead to side reactions to a significant extend; they are called "bio-orthogonal" because there is no significant reaction between reactive groups occurring naturally on proteins or cells and the reactants of the click reaction. Thus, click reactions allow the formation of chemical bonds in a highly specific and efficient process under mild and biocompatible conditions.

The term "cleavage site" refers to a part of the fusion protein of the invention that can be cleaved by a specific reagent resulting in the release of the binding moiety from the solid support. Preferably, the cleavage site is located C-terminal of the binding moiety.

In a preferred embodiment, the specific cleavage reagent can be an enzyme such as a protease, in which case the cleavage site refers to a specific peptide bond. Said cleaved peptide bond is located within or near a recognition amino acid sequence. Said recognition sequence can comprise 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, or 15 amino acids or a small protein domain. Said cleavage site is also referred to as "protease cleavage site". Proteases specific for certain amino acid sequences are well known to those skilled in the art.

The terms "cleaving" or "releasing" are used herein interchangeably and are intended to mean physical separation or detachment or dissociation of cells from the solid support. Generally known and practiced methods in the fields of molecular biology, cell biology, protein chemistry and antibody techniques are fully described in the continuously updated publications "Molecular Cloning: A Laboratory Manual", (Sambrook et al., Cold Spring Harbor); Current Protocols in Molecular Biology (F. M. Ausubel et al. Eds., Wiley & Sons); Current Protocols in Protein Science (J. E. Colligan et al. eds., Wiley & Sons); Current Protocols in Cell Biology (J. S. Bonifacino et al., Wiley & Sons) and Current Protocols in Immunology (J. E. Colligan et al., Eds., Wiley & Sons). Known techniques relating to cell culture and media are described in "Large Scale Mammalian Cell Culture (D. Hu et al., Curr. Opin. Biotechnol. 8:148-153, 1997); "Serum free Media" (K. Kitano, Biotechnol. 17:73-106, 1991); and "Suspension Culture of Mammalian Cells" (J.R. Birch et al. Bioprocess Technol. 10:251-270, 1990).

Embodiments of the Invention

The present invention will now be further described. In the following passages different aspects of the invention are defined in more detail. Each aspect defined below may be combined with any other aspect or aspects unless clearly indicated to the contrary. In particular, any feature indicated as being preferred or advantageous may be combined with any other feature or features indicated as being preferred or advantageous.

In a first aspect the present invention is directed to a fusion protein for capturing of cells expressing a cell- surface target comprising a non-lmmunoglobulin (Ig) protein that is capable of binding a cell-surface target expressed on cells with a dissociation constant K_D of 1 μΜ or less, and at least one cleavage site for the release of the cells from the solid support, and a peptide linker connecting the non-lg protein and a cleavage site, and a coupling moiety enabling attachment of the fusion protein to a solid support.

In an embodiment of the first aspect, a fusion protein of the invention comprises a non-lg protein as targeting moiety. A major advantage of non-lg proteins compared to antibodies is the specific binding only to cell surface expressed proteins. The Fc part of the antibody can bind non-specifically to cells and thereby initiate undesired reactions.. Further, non-lg proteins are less complex molecules, small, easy to engineer and can be produced in microorganisms, thus providing technical advantages and thereby eventually lowering costs.

Examples of suitable non-lg proteins are selected from the following molecules: Affilin (ubiquitin muteins), DARPin (ankyrin repeat protein muteins), Anticalin (lipocalin muteins), Affibody (muteins of the Z-domain of staphylococcal protein A), Fynomer (mutein of human Fyn SH3 domain), AdNectin (mutein of the tenth domain of human fibronectin), Kunitz domain peptides (muteins of Kunitz domains of various protease inhibitors), Nanofitins (Sac7d muteinsj, Avimers (muteins of multimerized Low Density Lipoprotein Receptor-A), chagasin scaffold or chagasin-like protease inhibitor proteins, Adnexin (fibronectin muteins), Centryrin (FN3 domain muteins), Knottin (cysteine-knot miniprotein muteins), Armadillo-repeat protein muteins, Atrimers (tetranectin muteins; C-type lectin domain muteins), or CTLA4 based muteins.

Additional information on alternative scaffolds is provided for example in Binz et al., Nat. Biotechnol., 2005 23: 1257 and Skerra, Current Opin. in Biotech., 2007 18: 295-304. The non-lg protein binds with detectable specific binding affinity to a cell surface target. A cell surface expressed target as understood in this invention can be relevant as marker for the detection of a disease. For example, CTC detection markers include, without limitation, any biomarkers for cell surface expressed cancer specific proteins. Cancer-specific biomarkers can include, for example, biomarkers that are specific for a given cancer-type of interest, a clinical cancer-stage of interest, or a cancer cell property of interest. Additionally, cancer-specific biomarkers can include more general cancer markers, such as cancer markers that are present in several cancer-types, but not in normal cells, or cancer markers that generally signal the malignant transformation of a cell. A person of skill will recognize that many specific and general cancer-specific biomarkers are known in the art.

Examples for binding partners for the non-lg proteins are cell surface expressed targets, selected from but by no means limited to Her2, EGFR, PD1 , EpCAM, NCAM, CEA, PDGFR, VEGFR, HGFR, Her3, Her4, EphB4 receptor tyrosine kinase, PSMA, MUC-1 , folate receptor, mesothelin, ALK, androgen receptor, AXL, MET, CD (for example CD31 , CD99, CD117, CD45, CD33, CD44, CD24, CD19, CD20, CD52, CD30), EMA, melan-A/MART-1 , and more. It should be noted, however, that a plurality of other possible targets can be added to this list. See for example Man et al (2011 ) J Clinic Experiment Pathol 1 :102 for a discussion of currently used markers for CTC isolation.

In an embodiment of the first aspect, the non-lg protein, for example an Affilin, has a dissociation constant K_D to a cell surface expressed target between 0.001 nM and 1000 nM, preferably below 100 nM, preferably below 10 nM, more preferably below 1 nM.

Methods for determining binding affinities, i.e. for determining the dissociation constant K_D, are known to a person of ordinary skill in the art and can be selected for instance from the following methods known in the art: surface plasmon resonance (SPR) based technology, Bio-layer interferometry (BLI), enzyme- linked immunosorbent assay (ELISA), flow cytometry, fluorescence spectroscopy techniques, isothermal titration calorimetry (ITC), analytical ultracentrifugation, radioimmunoassay (RIA or IRMA), and enhanced chemiluminescence (ECL). Some of the methods are described in more detail in the Examples below. Typically, the dissociation constant K_D is determined at ambient temperature, or at about 20°C, about 25°C, or about 30°C. If not specifically indicated otherwise, the K_D values recited herein are determined at 25°C by SPR.

The further characterization of the fusion proteins of the invention or of the non-lg-proteins, for example ubiquitin muteins, can be performed in the form of the isolated, soluble proteins. The appropriate methods are known to those skilled in the art or described in the literature. Such methods include the determination of physical, biophysical and functional characteristics of the proteins. The affinity and specificity of the variants isolated can be detected by means of biochemical standard methods such as SPR analysis or ELISA as known to those skilled in the art and as discussed above and in the Examples. For stability analysis, for example spectroscopic or fluorescence-based methods in connection with chemical or physical unfolding are known to those skilled in the art, including e.g. differential scanning fluorimetry (DSF).

For example, the fusion protein comprises a non-lg protein as targeting moiety which is an ubiquitin mutein that exhibits 80 % to 94 % sequence identity to the amino acid sequence of ubiquitin (SEQ ID NO: 1 ) or 80 % to 94 % identity to bis-ubiquitin of SEQ ID NO: 2, provided that the non-lg protein has a specific binding affinity to a cell-surface expressed target. In other words, ubiquitin muteins are modified in 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, or 15 amino acids compared to SEQ ID NO: 1 , preferably modifications in 6, 7, 8, 9, 10, or 11 amino acids, to generate a non-natural protein with newly created measurable binding properties to a target antigen.

The derivatization of a non-lg protein, for example, ubiquitin to generate a mutein that specifically binds to a particular target antigen has been described in the art. For example, a library can be created in which for example the sequence as shown in SEQ ID NO: 1 or SEQ ID NO: 2 has been altered. The step of modification of the selected amino acids is performed according to the invention preferably on the genetic level by random mutagenesis of the selected amino acids. A pre-selection of the amino acids to be modified by substitution, insertion or deletion can be performed based on structural information available for the ubiquitin protein to be modified. Preferably, the modification of the non-lg protein is carried out by means of methods of genetic engineering for the alteration of a DNA belonging to the respective protein. The selection of different sets of amino acids to be randomized leads to different libraries. The gene pool libraries obtained as described above can be combined with appropriate functional genetic elements which enable expression of proteins for selection methods such as display methods. The expressed proteins are contacted with a target molecule to enable binding of the partners to each other if a binding affinity exists. This process enables identification of those proteins which have a binding activity to the target molecule. Contacting according to the invention is preferably performed by means of a suitable presentation and selection method such as the phage display, ribosomal display, mRNA display, cell surface display, yeast surface display, or bacterial surface display methods, preferably by means of the phage display method. See, for example, EP2379581 B1 , EP1626985B1 , EP1567643B1 ; Kay et al., Phage Display of Peptides and Proteins-A Laboratory Manual (1996), Academic Press. The methods mentioned above are known to those skilled in the art. Identified clones with desired binding properties are then sequenced to reveal the amino acid sequences of muteins. The identified binding protein may be subjected to further maturation steps, e.g. by generating additional libraries based on alterations of the identified sequences and repeated phage display, ribosomal display, panning, and screening steps as described above.

In one embodiment of the invention, in order to generate a measurable binding affinity with a K_D of at least e.g. 10^"7 M to a target, a ubiquitin is at least substituted in 5 amino acids selected from region 62-68 (positions 62, 63, 64, 65, 66, 67, 68) of SEQ ID NO: 1. Further amino acids might be modified.

In another aspect of the invention the binding moiety comprises a ubiquitin mutein based on SEQ ID NO: 1 wherein the alteration is carried out at amino acids located in (i) region 2-11 , or (ii) region 62-68 or (iii) in both regions simultaneously. Further positions not comprised by these regions might be altered as well. In one embodiment of the invention, in order to generate a measurable binding affinity with a K_D of at least e.g. 10^"7 M to a target, a ubiquitin moiety is at least substituted in 5 amino acids corresponding to positions 62, 63, 64, 65, 66 of SEQ ID NO: 1 , preferably in combination with an insertion of 2 to 10 amino acids in the loop region corresponding to positions 8 to 11 of SEQ ID NO: 1 , preferably between positions 9 and 10 of SEQ ID NO: 1 (for example, EGFR-specific Affilin-139819; SEQ ID NO: 5). In another embodiment of the invention, two ubiquitin moieties are independently at least substituted in 5 amino acids selected from and corresponding to regions 2 - 11 and 62 - 68, for example selected from positions 2, 4, 6, 8, 62, 63, 64, 65, 66, 68 of SEQ ID NO: 1 , and the two ubiquitin moieties are connected directly or via a peptide linker, preferably directly connected.

In yet another embodiment, the binding moiety of the present invention relates to a binding protein with binding affinity (K_D) of less than 1000 nM for a target, wherein the target binding protein comprises an amino acid sequence wherein at least 12 amino acids selected from positions 42, 44, 68, 70, 72, 73, 74, 82, 138, 139, 140, 141 , and 142 of bis-ubiquitin (SEQ ID NO: 2) are substituted and wherein the binding protein has at least 85 % sequence identity to bis-ubiquitin (SEQ ID NO: 2). A target binding ubiquitin mutein may comprise 1 , 2, 3, 4, 5, or 6 further substitutions, in addition to the at least 12 amino acid substitutions selected from positions 42, 44, 68, 70, 72, 73, 74, 82, 84, 138, 139, 140, 141 , and 142 of bis-ubiquitin (SEQ ID NO: 2), to generate a binding protein for a target with high affinity, for example, a Her2-specific binding protein (Affilin-142628; SEQ ID NO: 7)

In another embodiment, the binding moiety of the present invention relates to a binding protein with binding affinity (K_D) of less than 1000 nM for a target, wherein the target binding protein comprises an amino acid sequence wherein at least 12 amino acids selected from positions 6, 8, 11 , 62, 63, 64, 65, 66, 71 , 82, 84, 138, 139, 140, 141 , 142 of bis-ubiquitin (SEQ ID NO: 2) are substituted and wherein the binding protein has at least 85 % sequence identity to bis-ubiquitin (SEQ ID NO: 2). A target binding ubiquitin mutein may comprise 1 , 2, 3, 4, 5, or 6 further substitutions to generate a binding protein for a target with high affinity, for example, a PD1-specific binding protein (Affilin-128187; SEQ ID NO: 10).

Preferably, the alteration is a substitution, insertion, or deletion as described in the art. The substitution of amino acid residues for the generation of the novel binding proteins derived from ubiquitin can be performed with any desired amino acid. This is described in detail in EP1626985B1 , EP2379581 B1 , and EP2721152, which are incorporated herein by reference.

In one embodiment of the invention, the fusion protein comprises at least one specific cleavage site cleaved by enzymes, for example, proteases (protease cleavage site, PCS) for the release of cells bound to the fusion protein of the invention. A suitable protease for the cleavage of the fusion protein to release cells bound to the fusion protein should have the following requirements: a) not present in the liquid from which the cells are captured in an active form and in significant amounts; b) not naturally occurring in an active form and in significant amounts in body fluids such as blood, urine, saliva, or spinal fluid; c) specificity for an amino acid sequence not commonly present or accessible in proteins presented on the surface of the captured cells.

Proteases that cleave a protein at a particular cleavage site are particularly preferred in this invention. For example, proteases specific for certain amino acid sequences are well known to those skilled in the art and comprise but are not limited to a viral protease including tobacco vein mottling virus (TVMV) protease, Tobacco etch virus (TEV) protease, plum pox virus (PPV) Nla protease, a turnip yellow mosaic virus (TYMV) protease, or to coagulation factor Xa, enterokinase, thrombin, SUMO proteases, ubiquitin proteases, and others. Other proteases with specific cleavage sites may also be used, as known to a person skilled in the art. Sequence specific proteases can be found in a public database in the internet (see for example, MEROPS, the Peptidase Database, http://merops.sanqer.ac.uk ). Proteases may be modified to improve activity, solubility, and/or decrease autolysis. Furthermore, proteases may be engineered to be specific for an amino acid sequence which is not recognized by the unmodified protease or recognized with less efficiency. The use of such modified proteases or active portions thereof is also encompassed within this invention.

The particular cleavage site for a protease is inserted into the fusion protein of the invention. For many proteases, the amino acid sequences for the protease cleavage site (PCS) are known. In one embodiment of the invention, the cleavage site of the fusion protein for the release of the cells bound to the fusion proteins can be for example but not limited to a viral protease cleavage site, for example of the TVMV protease cleavage site (ETVRFQS or ETVRFQG; SEQ ID NO: 21 or SEQ ID NO: 22, respectively), the TEV protease cleavage site (ENLYFQG or ENLYFQS; SEQ ID NO: 23 or SEQ ID NO: 24, respectively). Some variation of amino acids at particular positions in the cleavage site might be compatible with proteolytic cleavage. Preferred embodiments of the invention are protease cleavage sites specific for TVMV.

Exemplary temperatures for the protease reaction include 4 °C, room temperature, and 37 °C, depending on the protease used. Incubation times include 15 minutes to 24 hours.

In other embodiments of the first aspect, the fusion protein comprises at least one peptide linker covalently connecting the non-lg protein and the protease cleavage site. In another embodiment of the first aspect, the fusion protein comprises two peptide linkers, one peptide linker connecting the non-lg protein and the protease cleavage site, and a second peptide linker covalently connecting the protease cleavage site and the coupling moiety.

In preferred embodiments, the order of the parts of the fusion protein from the N-terminus to the C- terminus or the C-terminus to the N-terminus is as follows: (i) targeting moiety - linker 1 - PCS - linker 2 - coupling moiety, or (ii) targeting moiety - linker - PCS - coupling moiety.

In further preferred embodiments of the first aspect, fusion proteins for capturing cells from liquids comprise, essentially consist of or consist of an amino acid sequence selected from the group consisting of SEQ ID NOs: 6, 8, 9, 11 , or functional variants thereof, for example amino acid sequences with 90 % identity.

It is important that a suitable peptide linker should provide enough spatial distance for easy access of the protease. Enough spatial distance refers to the spatial separation between the non-lg binding moiety and the surface attachment site which must allow binding of the protease to its specific cleavage site. Thus, the spatial distance has to be on the order of the molecular dimensions of the chosen protease. For example, a TVMV or TEV protease has a diameter of at least about 25 angstrom. Thus, the linker should be at least of that length in an extended conformation to provide access of the protease to the amino acids of the cleavage site. Similarly, for a different protease with a different diameter the linker should be of the corresponding length. It is preferred that the amino acid sequence of the peptide linker is stable against proteases and/or does not form a secondary structure. The length and composition of a linker may vary between at least five and up to 50 amino acids. More preferably, the peptide linker(s) of the invention have a length of between about 10 and 30 amino acids; e.g., 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, or 30 amino acids. The first and the second peptide linker can be identical or different. Both linkers can independently consist of 5 to 50 amino acids. It is preferred that the first and the second peptide linker each consist of about 10 to 30 amino acids. For example, in one embodiment of the invention the first peptide linker ("linker 1") consists of about 10 amino acids, and the second peptide linker ("linker 2") consists of about 20 amino acids. In different embodiments, the first peptide linker and the second peptide linker each consists of about 15 amino acids and the amino acid sequence of the first and the second linker is identical or different.

In preferred embodiments of the invention, the first and the second peptide linker independently consist of amino acids selected from Gly, Ser, Ala, or Pro. Further, natural or non-natural amino acids with characteristic functional groups can be included in the linker sequence.

Well-known are linkers comprising small amino acids such as glycine and serine. The linkers can be Gly- rich (e.g., more than 50 % of the residues in the linker can be glycine residues). Preferred are Gly-Ser- linkers of variable length consisting of glycine and serine residues only.

Most preferred embodiments of the invention comprise linkers consisting of Ala, Pro, and Ser. Linkers consisting essentially of Ala, Pro, and Ser form unstructured, random coils and are stable in blood and are non-immunogenic. It is preferred that a peptide linker consists of about 40 to 60 % alanine, about 20 to 35 % proline, and about 10 to 30 % serine. In one embodiment of this invention, the linker (SEQ ID NO: 12) consists of about 50 % Ala, about 30 % Pro, and about 20 % Ser. In another embodiment of this invention, the linker (SEQ ID NO: 13) consists of about 50 % Ala, about 25 % Pro, and about 25 % Ser. In yet another embodiment of this invention, the linker consists of about 53% Ala, about 33 % Pro, and about 13 % Ser (SEQ ID NO: 14 and SEQ ID NO: 15). It is further preferred that the amino acids alanine, proline, and serine are evenly distributed throughout the linker amino acid sequence so that not more than a maximum of 2, 3, 4, or 5 identical amino acid residues are adjacent, preferably a maximum of 3 amino acids. Further preferred are linkers essentially consisting of 50 % or more of the amino acids Ala, Pro, or Ser.

Selected but by no means limiting examples for suitable linkers are having the amino acid sequence

SAPASPAPAA (SEQ ID NO: 12, referred to as APS10 herein), SAPAPSSAPAASAPPAAASA (SEQ ID NO: 13, referred to as APS20 herein), ASPAAPAPASPAAPA (SEQ ID NO: 14, referred to as APS15-a herein) or ASPAPAAAPSAPAPA (SEQ ID NO: 15, referred to as APS15-b herein), or of amino acid sequences with 90 % identity. For example, the linkers APS10 or APS20 between the coupling moiety (e.g. Cys) and the protease cleavage site and the Affilin and the protease cleavage site, respectively, provide enough spatial distance for easy access of the protease, as shown in the Examples and illustrated by the Figures of the invention. Any other linker providing enough spatial distance from the surface and the Affilin for easy access of the protease can be used, for example SGGGGSGGGG (SEQ ID NO: 16), GSGGG (SEQ ID NO: 17), GGGGSGGGGSGGGGS (SEQ ID NO: 18), or (GGGGS)_n or (SGGGG (i.e., n repetitions of SEQ ID NO: 19 or SEQ ID NO: 20, wherein n is between 1 and 5 (e.g., n may be 1 , 2, 3, 4, or 5), or other linkers known to those skilled in the art.

In an embodiment of the first aspect, the fusion protein comprises a coupling moiety. A coupling moiety is enabling attachment of the fusion protein to the surface of a solid support. Further, a suitable coupling moiety comprises a reactive chemical group which specifically forms a covalent bond with chemical groups. Preferably, the fusion protein is chemically coupled to an attachment site on a solid support.

It is preferred that the coupling moiety of the fusion protein comprises a reactive group which can undergo a substitution reaction, a redox reaction, an addition reaction, an elimination reaction or a radical reaction.

It is preferred that the coupling moiety is selected from cysteine or maleimide. A preferred fusion protein of the invention comprises one cysteine residue (as illustrated, for example, but not limited to, in SEQ ID NOs: 6, 8, 9, 11 ). Preferably, the cysteine is located at the C-terminus or N-terminus of the fusion protein. Most preferably, the cysteine is located at the C-terminus of the fusion protein. In some embodiments, a short peptide sequence can be fused C-terminally to the fusion protein, e.g. a Strep-tag consisting of 8 amino acids (WSHPQFEK; SEQ ID NO: 25), and allows purification and detection of the fusion protein. Other short-affinity tags known in the art could be fused to the fusion proteins (e.g. His-tag, FLAG-tag).

In a preferred embodiment, the reactive group of the coupling moiety forms a covalent bond with a reactive group on the solid support. For instance, a fusion protein with C-terminal cysteine can be attached to a maleimide group on the solid support.

In another preferred embodiment the cysteine of the fusion protein forms a stable, yet non-covalent complex with reactive groups on the solid support. Such interactions can be mediated, for example, by hydrophobic and/or van der Waals interactions.

In a most preferred embodiment, the fusion protein of the invention is biotinylated, i.e. biotin is covalently attached to the cysteine sulfhydryl groups of fusion protein for binding to the solid support (for example, for binding to a streptavidin or avidin-coated solid support). Biotin refers to biotin (cis-hexahydro-2oxo-1 H- thieno[3,4]imidazole-4-pentanoic acid) and any biotin derivatives and analogs. Such derivatives and analogues are substances which can form a complex with the biotin binding pocket of native or modified streptavidin or avidin. Avidin refers to a native egg-white glycoprotein avidin as well as derivatives or equivalents thereof, such as deglycosylated or recombinant forms of avidin. Streptavidin refers to bacterial streptavidins as well as derivatives or equivalents thereof such as recombinant and truncated streptavidin. Methods for biotinylation of fusion proteins are known in the art and are described in the Examples.

In an embodiment of the invention, the coupling moiety of the fusion protein enables attachment of the fusion protein to a surface of a solid support. A solid support is typically substantially insoluble in aqueous phases. A large number of supports is available and is known to one of ordinary skill in the art. Solid supports that may be employed in accordance with the invention include beads, plates, microtiter plates, filter material, sheets, microchips, wires, microcapillaries, hydrogels, filaments, fibers, to name a few. The form or shape of the solid support may vary, depending on the application. Suitable examples include, but are not limited to, microparticles of various sizes, microtiter plates of any dimension, gels (aerogel, hydrogel, resin), wires, filaments, slides, strips, microchips, wells, fibers, and combinations thereof. The solid material can be magnetic or non-magnetic. Most preferred solid supports are microparticles such as beads (for example, streptavidin beads, thiol-activated beads, avidine beads, or maleimide beads) or plates (for example, streptavidin/avidin coated plates or maleimide activiated plates).

The solid support can be made, for example, from any water-insoluble material such as organic polymers, inorganic polymers or materials, a biological polymer, ceramics, metals (e.g. gold, platin, palladium), alloys, composite materials or hydrogels, an inorganic salt or substance, or glass, to name a few. The surface of the solid may be ordered or non-ordered. More specific examples of solid supports include silica, silica gels, polymeric membranes, derivatized plastic films, glass beads, cotton, plastic beads, alumina gels, poly(acrylate), polystyrene, poly(acrylamide), polyol, agarose, agar, nitrocellulose, diazocellulose, polyvinylchloride, polypropylene, polyethylene (including poly(ethylene glycol)), latex beads, magnetic beads, paramagnetic beads, superparamagnetic beads and the like. In some embodiments, the solid support may include a reactive functional group, including, but not limited to, hydroxyl, carboxyl, amino, thiol, aldehyde, halogen, nitro, nitroso, cyano, amido, urea, carbonate, carbamate, isocyanate, sulfone, sulfonate, sulfonamide, sulfoxide, amine, azide, hydrazide, hydroxylamine, alkene, alkyne, cyclo-octyne, for attaching the fusion proteins of the invention. A suitable solid phase support can be selected on the basis of desired end use and suitability for various synthetic protocols.

In a further aspect of the invention, a fusion protein of the invention is used in applications for capturing of cells, in particular CTCs. Cancer is spread in the body by circulating tumor cells. The cells originate from the primary tumor and are transported to new sites for subsequent growth of tumors (metastases). Thus, the presence of CTCs in the blood of a cancer patient is associated with decreased survival rates of patients with different types of cancer, for example, for breast, colorectal, or prostate cancer. Using the fusion proteins of the invention, CTC levels in patient blood samples can be specifically measured and studied allowing better diagnosis and assessment of patient prognosis. After capturing of cells from a liquid, for example a body fluid, it is desired to release the cells from the surface for further cultivation and analysis. Usually, elution of cells from the capturing moieties requires harsh conditions such as low pH and/or high salt. Both conditions can result in cell damage and are deleterious for cells. Thus, the advantage of releasing captured cells by cleaving the fusion protein as described herein are the gentle conditions which do not harm the cells so that cells remain suitable and compatible for further uses or applications.

In one embodiment of the invention, fusion protein immobilized on solid supports can be used to capture specific cells and to separate the bound cells from the unbound cells in the population. The fusion protein of the invention bound to a solid support is used to segregate a particular target expressing cell from a mixed cell population by specific binding of the cell surface expressed protein to the target-specific fusion protein of the invention. Cells which do not express said protein are not bound by the immobilized fusion protein. Target expressing cells are thereby captured and can be separated from other cells of a mixed cell population. Immobilized fusion proteins specific for particular cell surface proteins can be used to capture cells of interest and positively select the labeled cells to prepare a desired cell population. The advantage of the use of a fusion protein of the invention and fusion protein-coated solid supports is the gentle removal of the cells from the solid support without cell damage or serious alterations of the cells. This is important for further uses and downstream applications.

In one embodiment of the invention, the immobilized fusion protein of the invention is used for isolating a target cell from a mixed population using a fusion protein of the invention which specifically binds the target cell and a protease which cleaves a proteolytic cleavage site in the fusion protein of the invention to release the target cell bound to the immobilized fusion protein. In one aspect of the method, a target cell attached to a fusion protein-coated solid surface is detached from the surface by contacting the target cell/surface complex with a protease, where the protease is specific for a cleavage site in the fusion protein. Contact with the protease results in cleavage of the protease cleavage site in the fusion protein and detaching of the target cell from the support.

In another embodiment of the invention, a method is provided for detaching a target cell attached to a fusion protein-coated solid support, the method comprising contacting the target cell attached to the support with a protease specific for a protease cleavage site in the fusion protein, thereby detaching the target cell from the support.

In a further aspect, the present invention is directed to a method of capturing cells and enrichment of cells from a mixed cells population, for example from a liquid sample, comprising the steps of (a) providing a solid support with an immobilized fusion protein comprising a non-lmmunoglobulin (Ig) protein that is capable of binding a cell-surface target expressed on cells with a dissociation constant K_D of 1 μΜ or less, at least one cleavage site for the release of the cells bound to the fusion proteins, a peptide linker covalently connecting the non-lg protein and a cleavage site, and a coupling moiety enabling attachment of the fusion protein to a solid support, (b) providing a mixed cell population, preferably a liquid sample containing cells, (c) contacting said liquid sample and said solid support under conditions that permit binding of a specific cell-surface target expressed on cells in the liquid sample to the immobilized fusion proteins, and (d) cleaving said immobilized cells from the solid support by protease cleavage, thereby obtaining an eluate containing said cells expressing the specific cell-surface target.

Further steps comprise separating the cells bound to the fusion proteins immobilized to the solid support from cells not bound to the solid support. Optionally, the method further comprises collecting the cells detached from the solid support.

The liquid samples can be any sample suspected to contain cell populations, for example, CTCs, such as but not limited to whole blood, plasma, amniotic fluid, pleural fluid, peritoneal fluid, central spinal fluid, urine, saliva, and bronchial washes. In some embodiments, the liquid sample is a blood sample. As will be appreciated by those skilled in the art, a biological sample can include any fraction or component of blood, without limitation, T-cells, monocytes, neutrophils, erythrocytes, platelets.

The liquid samples of this disclosure can be obtained from any organism, including mammals such as humans, primates, cats, dogs, rabbits, farm animals, and rodents. In particular animals used in the preclinical development of drugs including but not limited to mice and rats. In a further aspect, the present invention is directed to a method of capturing cells from a mixed cells population, for example from a liquid sample, comprising the steps of (a) providing a solid support, (b) providing a fusion protein comprising a non-lmmunoglobulin (Ig) protein that is capable of binding a cell- surface target expressed on cells with a dissociation constant K_D of 1 μΜ or less, at least one cleavage site for the release of the cells bound to the fusion proteins, a peptide linker covalently connecting the non-lg protein and a cleavage site, and a coupling moiety enabling attachment of the fusion protein to a solid support, (c) providing a mixed cell population, preferably a liquid sample containing cells, (d) contacting said liquid sample and the fusion protein under conditions that permit binding of a specific cell- surface target expressed on cells, (e) contacting the cell - fusion protein complex with said solid support to allow the coupling moiety to react with the solid support and thereby immobilize the cell - fusion protein complex, (f) removing the solid support with the bound cells from the liquid and non-bound cells and (g) cleaving said immobilized cells from the solid support by protease cleavage, thereby obtaining an eluate containing said cells expressing the specific cell-surface target.

In some embodiments of the methods, at least about 50 %, 60 %, 70 %, 80 %, 90 %, or 95 % of the target cells in the starting composition are recovered and viable. Table 1 shows recovery rates of cells that were released by protease cleavage from fusion proteins coupled to different solid supports.

Table 1. Recovery of cells after protease cleavage

In another aspect the present invention is directed to a nucleic acid molecule, preferably an isolated nucleic acid molecule, encoding a fusion protein of the first aspect.

The invention further provides an expression vector comprising the polynucleotide of the invention, and a host cell comprising the polynucleotide or the expression vector of the invention. A vector means any molecule or entity (e.g., nucleic acid, plasmid, bacteriophage or virus) that can be used to transfer protein coding information into a host cell.

For example, one or more polynucleotides which encode a fusion protein of the invention may be expressed in a suitable host and the produced fusion protein can be isolated. The present invention furthermore relates to an isolated host cell comprising the nucleic acid molecule of the invention or the vector of the invention. Suitable host cells include prokaryotes or eukaryotes. Various mammalian or insect cell culture systems can be employed to express recombinant proteins.

In another aspect the present invention is directed to a method of production of a fusion protein of the invention, wherein fusion proteins of the invention may be prepared by any of the many conventional and well known techniques such as plain organic synthetic strategies, solid phase-assisted synthesis techniques, fragment ligation techniques or by commercially available automated synthesizers. On the other hand, they may also be prepared by conventional recombinant techniques alone or in combination with conventional synthetic techniques. Furthermore, they may also be prepared by cell-free in vitro transcription/translation.

In another aspect, there is provided a method of producing the fusion protein of the invention, comprising the steps of a) culturing the host cell of the invention under conditions suitable for the expression of the fusion protein and b) isolating the produced fusion protein. Suitable conditions for culturing a prokaryotic or eukaryotic host are well known to the person skilled in the art.

One embodiment of the present invention is directed to a method for the preparation of a fusion protein according to the invention as detailed above, said method comprising the following steps: (a) preparing a nucleic acid encoding a fusion protein as defined above; (b) introducing said nucleic acid into an expression vector; (c) introducing said expression vector into a host cell; (d) cultivating the host cell; (e) subjecting the host cell to culturing conditions under which a fusion protein is expressed, thereby producing a fusion protein as described above; (f) optionally isolating the protein produced in step (e).

Cultivation of cells and protein expression for the purpose of protein production can be performed at any scale, starting from small volume shaker flasks to large fermenters, applying technologies well-known to any skilled in the art.

Following the expression of the fusion protein of the invention, it can be further purified and enriched by methods known in the art. The selected methods depend on several factors known to those skilled in the art, for example the expression vector used, the host organism, the intended field of use, the size of the protein and other factors.

In general, isolation of purified protein from the cultivation mixture can be performed applying conventional methods and technologies well known in the art, such as centrifugation, precipitation, flocculation, different embodiments of chromatography, filtration, ultrafiltration, dialysis, concentration and combinations thereof, and others. Specific examples comprise ion exchange chromatography, gel filtration chromatography (size exclusion chromatography), affinity chromatography, high pressure liquid chromatography (HPLC), reversed phase HPLC, disc gel electrophoresis or immunoprecipitation, see for example in Current Protocols in Protein Science (J. E. Colligan et al. eds., Wiley & Sons).

For simplified purification the fusion protein according to the invention can be fused to other peptide sequences having an increased affinity to separation materials. Preferably, such fusions are selected that do not have a detrimental effect on the functionality of the fusion protein or can be separated after the purification due to the introduction of specific protease cleavage sites. Such methods are also known to

those skilled in the art.

Examples

The following Examples are provided for further illustration of the invention. The invention, however, is not

limited thereto, and the following Examples merely show the practicability of the invention on the basis of

the above description. For a complete disclosure of the invention reference is made also to the literature

cited in the application which is incorporated completely into the application by reference.

Example 1. Generation of expression constructs

A schematic drawing of the fusion proteins is provided in Figure 1. The arrow illustrates the protease.

Fusion proteins were designed having the following structural features:

Targeting moiety: Affilin; linker 1 between Affilin and cleavage site; cleavage site: protease cleavage site;

linker 2 between PCS and coupling moiety; coupling moiety: Cysteine

Exemplary fusion proteins (shown from N-terminus to C-terminus):

. 144893 (EGFR-specific)(SEQ ID NO: 6): Affilin139819-APS₁₀-TVMV-APS₂o-Cys- . 144892 (EGFR-specific)(SEQ ID NO: 9): Affilin139819-APSi₀-TEV-APS₂₀-Cys-

. 148162 (Her2-specific) (SEQ ID NO: 8): Affilin142628-APS₁₀-TVMV-APS₂o-Cys-

. 151318 (PD1-specific) (SEQ ID NO: 1 1 ): Affilin128187-APS₁₀-TVMV-APS₂o-Cys-

Genes were synthesized and cloned into an E. coli expression vector using standard methods known to a

person skilled in the art. For example, a linker consisting of 10 amino acids (SEQ ID NO: 12) was cloned

between the target specific Affilin (SEQ ID NO: 5, 7, or 10) and the protease cleavage site TVMV (SEQ ID

NO: 21 ) or TEV (SEQ ID NO: 23). The amino acid cysteine (Cys) as moiety for enabling coupling to a

solid support was cloned C-terminally of the peptide linker consisting of 20 amino acids (APS20SEQ ID

NO: 13). The second linker was cloned between the protease cleavage site and cysteine. Further, a C- terminal StrepTag (SEQ ID NO: 25) was added to enable standard purifications protocols. DNA

sequencing was used to verify the correct sequence of the fusion proteins.

As target specific ubiquitin muteins, the following proteins were used: Affilin-139819 (SEQ ID NO: 5) is an

EGFR-specific Affilin (ubiquitin mutein) with a K_D of about 20 nM and thermal stability at 73 °C; Affilin- 142628 (SEQ ID NO: 7) is a Her2-specific Affilin with a K_D of about 0.4 nM and thermal stability at 62 °C;

and Affilin-128187 (SEQ ID NO: 17) is a PD1-specific Affilin with K_D of about 0.7 nM and thermal stability

at p5.6 °Cj. Any other target specific functional variants of these Affilin binding proteins or other target- _ _ - j Kommentar [BE1]: Runden? specific non-lg proteins could be used.

As control, fusion protein 148161 (SEQ ID NO: 4) was constructed using wildtype ubiquitin (SEQ ID NO:

1 ) without any binding capability to a surface expressed target: ubiquitin-APS₁₀-TVMV-APS₂₀-Cys. Example 2. Expression of fusion proteins

HMS174 (DE3) competent cells were transformed with either expression plasmid encoding Affilin muteins. Cells were spread onto selective agar plates (Kanamycin) and incubated overnight at 37 °C. Precultures were inoculated from a single colony in 100 ml superrich medium (modified H15 medium 2 % glucose, 5 % Yeast extract, 1 % Casamino acids, 0.76 % glycerol, 0.7 % lactose, 1 % Torula Yeast RNA, 250 mM MOPS, 202 mM TRIS, 10 mg/L RNase A, pH 7.4, Antifoam SE15) and cultured 16 hours at 37 °C at 160 rpm in a conventional orbital shaker in baffled 1 L Erlenmeyer flasks supplemented with 150 μg/ml Kanamycin without lactose and antifoam. Main culture was inoculated from previous overnight culture with an adjusted start-OD₆oo of 0.5 in 400 ml superrich medium in 1 L thick-walled Erlenmeyer flasks that was supplemented with glycerol, glucose, lactose, antifoam agent and 150 μg/ml Kanamycin. Cultures were transferred to a resonant acoustic mixer (RAMbio) and incubated at 37 °C with 20 x g. Aeration was facilitated by Oxy-Pump stoppers. Recombinant protein expression was induced by metabolizing glucose and subsequently allowing lactose to enter the cells. At predefined time points OD₆oo was measured, samples adjusted to 5/OD₆oo were withdrawn, pelleted and frozen at -20 °C. Cells were grown overnight for approximately 24 hours to reach a final OD₆oo of about 45-60. To collect biomass cells were centrifuged at 16000 x g for 10 min at 20 °C. Pellets were weighed (wet weight) and pH was measured in the supernatant. Cells were stored at -20 °C before processing.

Example 3: Analysis of expression and solubility of fusion proteins

Samples taken during fermentation were resuspended in 300 μΙ extraction buffer (PBS supplemented with 0.2 mg/ml Lysozyme, 0.5 x BugBuster, 7.5 mM MgS0₄, 40 U Benzonase) and solubilized by agitation in a thermomixer at 700 rpm at room temperature for 15 min. Soluble fusion proteins were separated from insoluble fusion proteins by centrifugation (16000 x g, 2 min, rt). Supernatant was withdrawn (soluble fraction) and the pellet (insoluble fraction) was resuspended in equivalent amount of urea buffer (8 M urea, 0.2 M Tris, 2 mM EDTA, pH 8.5). From both soluble and insoluble fraction 50 μΙ were taken and 12 μΙ 5 x sample buffer as well as 5 μΙ 0.5 M DTT were added. Samples were boiled at 95 °C for 5 min. Finally 8 μΙ of those samples were applied to a SDS-gel which was then run in accordance to the manufacturer's recommendations. High level expression of all fusion proteins was found under optimized conditions within the chosen period of time. All of the expressed fusion proteins were soluble to more than 95 % (SDS-PAGE).

Example 4. Biochemical characterization of target -specific fusion proteins with protease cleavage site

Purification. Fusion proteins with a C-terminal Strep-tagll (SEQ ID NO: 25) were expressed in the soluble fraction of E. coli. The cells were lysed by sonication and the first purification step was performed with Strep-Tactin columns according to the manufacturer's instructions (IBA lifesciences). To avoid disulfide formation, the buffers were supplemented with 1 mM DTT. The eluted fractions were injected to a HiLoad 16/600 Superdex 75 pg (GE Healthcare) equilibrated with buffer containing 20 mM citrate pH 6.0 and 150 mM NaCI. The peak fractions were pooled and analyzed by SDS-PAGE. Fusion proteins are stable at high temperatures. Thermal stability of the fusion proteins of the invention was determined by Differential Scanning Fluorimetry. Each probe was transferred at a concentration of 0.1 μg μL to a MicroAmp® Optical 384-well plate (ThermoFisher), and SYPRO Orange dye was added at suitable dilution. A temperature ramp from 25 to 95 °C was programmed with a heating rate of 1 °C per minute (ViiA-7 Applied Biosystems). Fluorescence was constantly measured at an excitation wavelength of 520 nm and the emission wavelength at 623 nm (ViiA-7, Applied Biosystems). Similar melting points correlate to related protein structures.

Table 1 : Biochemical characterization

Example 5. Biotinylation of fusion proteins

Fusion proteins were biotinylated with a biotinylation agent via the cysteine sulphydryl (-SH) group of the fusion protein according to the manufacturer's instructions (Thermo Fisher EZ-link HPDP-Biotin, catalog product number 21341 ; https://www.thermofisher.com/order/catalog/product/21341 ). 2 mg of reduced fusion protein was dissolved in 1 ml PBS, 1 mM EDTA. 100 μΙ HPDP-Biotin (dissolved in DMSO) was added to 1 ml of fusion protein solution to result in 0.4 mM Biotin HPDP and incubated for 2 hours at room temperature. For separating the biotin labeled fusion protein from non-reacted HPDP-Biotin, HiTrap Desalting columns (5 ml, GE healthcare) were used. The columns were equilibrated with PBS.

Example 6. Coupling of biotinylated fusion proteins to Dynabeads

Dynabeads are monodisperse polymeric superparamagnetic beads were used in this invention. 25 μΙ with 1 x 10⁷ CELLection™ Biotin Binder Dynabeads^® (life technologies, cat. no. 11533D) were washed twice with 0.1 % BSA in PBS and incubated with 0.2 μ9 biotinylated fusion protein (144893, 148162, or control 148161 ) in PBS for 1 h at room temperature. Beads were washed twice with PBS and once in PBST (PBS with 0.1 % Tween). For further analysis of the immobilization of the fusion proteins to the beads, beads were incubated with 100 nM EGFR-Fc (144893); 100 nM Her2-Fc (148162) or 0.2 g/ml anti-Ubi- Fab-HRP (148161 ) in PBST 30 min at room temperature. Beads with immobilized biotin labeled fusion proteins 144893 and 148162 were incubated 30 min with anti-human-lgG-Fc-HRP in PBST. Beads were washed once with PBST and twice with PBS. As detecting reagent, TMB-Plus (Biotrend, Germany) was used, developed and fixed using 0.2 M H₂S0₄ solution and measured in a plate reader at 450 nm versus 620 nm. Figure 2 shows the immobilization of target-specific fusion proteins on beads. ELISA analysis confirms both the specific coupling of fusion proteins to beads and functionality of the immobilized fusion proteins i.e. specific binding of the target to the immobilized fusion proteins bound to beads.

Example 7. Cell binding to EGFR-specific fusion proteins coupled to beads

Cells of EGFR expressing cell lines A431 and A549 and as control cells of HEK293 (low levels of EGFR expression) were detached from the cell culture flasks by trypsin/EDTA. 2.5 x 10⁵ cells/ml in 10 % FCS/PBS per cell line (100 % value as start concentration) were incubated with 25 μΙ (1x10⁷) 144893- coupled beads 1 h at room temperature. The beads as used herein were magnetic particles that were separated from other particles of the sample by magnetic focusing in a quick, simple, gentle, and efficient way. Beads were separated by application of a magnetic field by using a magnet which was applied to the side of the vessel containing the sample. Beads were aggregated to the wall of the vessel, supernatants removed, and beads washed twice with 10 % FCS/PBS. Beads and the supernatants of the sample (e.g. unbound cells) were analyzed. The cell viability was analyzed by the CellTiter-Glo^® Luminescent Cell Viability Assay (Promega; G7571 ). Figure 3 shows the binding of target-positive cells to beads coupled with target-specific fusion protein. About 40 % of EGFR positive cells are coupled to beads with immobilized fusion protein. The HEK293 cell line expressing only low levels of EGFR shows less than 20 % binding of coated beads. RLU is "relative luminescence unit".

Example 8. Binding of cells on beads coupled with EGFR specific fusion protein and capturing of cells after protease digestion

Cells of EGFR expressing cell lines (A431 , A549, SkBr3) and of HEK293 with low levels of EGFR expression were detached from the cell culture flasks by trypsin/EDTA. Biotinylated EGFR specific fusion protein 144893 or control fusion protein 148161 were immobilized on Dynabeads. Beads were then incubated with 2.5 x 10⁵ cells/ml (100% value as start concentration) of A431 -, A594-, SkBr3-, or HEK293-cells for 30 min at room temperature. Beads were washed and incubated with 200 μg/ml TVMV protease (in 12 % FCS/0.5 mM DTT/PBS or as control with only buffer (PBS with 12 % FCS/0.5 mM DTT/PBS) for 1 h at room temperature. Analysis was by CellTiter-Glo and microscopy as described above. Results for SkBr3 are comparable to the results obtained with A431 and A549.

Figure 4 shows specific binding of target expressing cell lines to beads coupled with target-specific fusion protein and nearly complete recovery of cells after protease treatment. Further, the result confirms that the protease cleavage site is accessible to protease after attachment of the fusion protein to the solid surface. In the figure: cells incubated with immobilized EGFR-specific fusion protein on beads, treated with protease (1 ) or without protease (2); cells incubated with immobilized control fusion protein on beads, treated with protease (3) or without protease (4). Figure 4A shows the recovery of cells after treatment with TVMV protease (more than 50 % of target positive cells are recovered). No cells were recovered from beads with immobilized control fusion protein. Figure 4B shows beads with cells specifically bound to the fusion proteins before protease treatment confirming specific binding of cells to the immobilized fusion protein (lined bars). Beads after protease treatment are illustrated as black bars. EGFR-positive cells show a specific binding to beads coupled with EGFR specific fusion protein. About 10 percent of HEK293-cells show binding to EGFR specific fusion proteins coupled to beads. No binding of cell lines to bead - immobilized control fusion protein without targeting moiety was observed. Figure 4C shows the microscopic analysis of binding of EGFR-positive cells to beads coupled with EGFR specific fusion protein and release of cells from beads by protease treatment. The strong binding of A431- and A549- cells to beads with EGFR specific immobilized fusion protein resulted in cell-bead clusters. Treatment of cell-bead-complexes with TVMV protease dissolves the cell-bead clusters; no or very few clusters are detectable. For HEK293-cells, almost no binding of cells to EGFR-specific immobilized fusion protein is observed. No or only very minor binding of cells to control fusion protein was observed.

Example 9: Binding of cells on beads coupled with PD1 specific fusion protein and capturing of cells after protease digestion

25 μΙ with 1 x1 0⁷ CELLection™ Biotin Binder Dynabeads^® were incubated with 0.2 μg/ml biotinylated 151318 (PD1 fusion protein) or 148161 (control fusion protein) and washed. Protein coupling to beads was analyzed by ELISA using anti-Ubi-Fab. Cells of PD1 positive cell line EL4 and PD1 negative cell line K562 were harvested by centrifugation. PD1 specific fusion protein 151318 or control fusion protein 148161 were immobilized on Dynabeads. Beads were then incubated with 0.2 5x 10⁶ cells/ml (100 % value as start concentration) for 30 min at room temperature. Beads were focused, washed and incubated with 200 g/ml TVMV protease (in 12 % FCS/0.5 mM DTT/PBS or as control with only buffer (PBS in 12 % FCS/0.5 mM DTT/PBS) for 1 h at room temperature. Supernatant and cell-bead-complex were analysed by CellTiter-Glo and microscopy as described above.

Figure 5 shows specific binding of PD1-positive cells to PD1-fusion protein-coupled beads and complete recovery of cells after TVMV treatment. Figure 5A shows the coupling of fusion proteins to beads. Figure 5B shows the recovery of cells after treatment with TVMV protease. No cells were recovered from beads with immobilized control fusion protein. Figure 5C shows beads with cells specifically bound to the fusion proteins before protease treatment confirming specific binding of cells to the immobilized fusion protein About 35 % of PD1-positive EL4-cells show a specific binding to beads coupled with PD1 -specific fusion protein (1 , 2) whereas PD1-negative K562-cells show no binding to immobilized PD1-specific fusion protein. No binding of cells to bead - immobilized control fusion protein without targeting moiety was observed (3, 4). Figure 5D shows the analysis by microscopy. The figure shows that EL4 cells are bound to immobilized fusion protein, but not to immobilized control fusion protein. Further, the figure confirms that TVMV digestion leads to cleavage of EL4-cells. Isolated cells are bead-free. No binding of K562-cells (PD1 negative) on 151318 coupled beads was observed. Example 10. Capturing of EGFR-positive cells from a mixed cell population using EGFR-specific fusion proteins coupled to Dynabeads

2.5 x 10⁵ A549-cells/ml labeled with CellTracker™ Green CMFDA Dye (Life Technologies, cat.-no. C2925) were mixed with 2.5 x 10⁵ HEK293-cells/ml at different ratios. Cell composites were incubated with beads coupled with fusion protein 144893 for 30 min at room temperature. Beads were separated by a magnet as described above and were analyzed by microscopy. Figure 6 confirms that target positive cells can be specifically isolated from a cell mixture by using target specific fusion proteins coupled to beads. Figure 6 shows the microscopic analysis of the capturing of target-positive cells from a mixed cell population. The percentage of green fluorescent cells increased with higher concentration of target-positive cells (A549) in the mixed cell population of A549- and HEK293-cells. Only few remaining unbound cells could be detected in supernatants.

Example 11. Coupling of biotinylated fusion proteins to Streptavidin plates

Streptavidin plate (Pierce® streptavidin Coated High Binding Capacity, Pierce, Thermo Scientific Kit cat. no.15501 Lot#QC216957) was incubated with 10 Mg/ml, 1 Mg/ml, 0.1 Mg/ml, 0.01 Mg ml of biotinylated 144893 or control 148161 in washing buffer (0.1 % BSA/ 0.05 % Tween/ 25 mM Tris-HCL/ 150 mM NaCI pH 7.2) for 15 min, 30 min, 1 h, and 2 h at room temperature or overnight at 4 °C. After washing steps, plates were incubated with 100 nM hEGFR-Fc (target) followed by anti-human-lgG-HRP or with anti-Ubi- Fab-HRP for 30 min at room temperature. Fusion proteins were detected with TMB-substrate, and the reaction stopped with sulfuric acid. Plate was measured in plate reader with 450 nm versus 620 nm. Figure 7 shows that streptavidin-coupled plates can be successfully used as solid support for the coupling of target-specific fusion proteins. ELISA analysis confirms a high specific binding of target after immobilizing a target-specific fusion protein to a concentration of 0.1 μg/ml. No target binding to control protein immobilized on plates was observed. The immobilization of control protein was detected by a ubiquitin specific antibody.

Example 12. Binding of A549-cells to EGFR specific fusion protein coupled to Streptavidin plates and release of A549-cells after TVMV treatment

100 μΙ of 100 nM biotinylated 144893 or control 148161 were immobilized on streptavidin-coated plates for 30 min at room temperature and incubated with 100 μΙ 0.01 % biotin for 20 min. 100 μΙ A549- or HEK293-cells with a concentration of 5 x 10⁵ cells/ml were added to coated plates for 15 min at room temperature. Non-bound cells were removed by washing steps. 200 μg/ml TVMV protease or control buffer were added for 1 h at room temperature. Cell binding and release of cells were analyzed by a cell viability assay using the CellTiterGlo Luminescent Cell Viability Assay or microscopy, as described above. Figure 8A confirms the binding of cells to fusion proteins immobilized on streptavidin plates and the complete release of the cells from the plates after protease digestion by CellTiter-Glo detection. Almost all EGFR-positive cells are bound on wells coated with EGFR-specific fusion protein, but no binding on wells coated with control fusion protein was observed. HEK293-cells, known for a week expression of EGFR, show a weak binding on fusion protein coated wells. After TVMV protease treatment, almost all of the EGFR-specific cells were cleaved from the solid support and recovered in the supernatant. Figure 8B shows from microscopic observation that the cells are almost completely released from the plates after protease treatment whereas all cells are bound to the plates after PBS treatment. Control fusion protein with no specificity to EGFR does not bind to EGFR-positive cells.

Example 13. Binding of SkBr3 cells to Her2 specific fusion protein coupled to plates and isolation of SkBr3 cells after TVMV protease treatment

100 μΙ of 100 nM biotinylated fusion protein (148162 or 148161 , respectively) was coupled to streptavidin- coated plates (Pierce) as described above and incubated with 100 μΙ SkBr3-cells or HEK293-cells with a concentration of 5 x 10⁵cells/ml for 15 min at room temperature followed by incubation with 200 μg/ml TVMV protease for 1 h at room temperature. Analysis of the cell binding and release were done as described above. Figure 9A confirms the binding of target-positive cells to target-specific fusion protein immobilized on plates followed by complete release of viable cells after protease treatment by CellTiter- Glo measurement. Cells are almost completely released from plates after protease treatment. Figure 9B confirms the binding of fusion protein immobilized on plates to Her2-positive cells by microscopy. The cells are nearly completely released from the streptavidin plates after TVMV treatment.

Example 14. Coupling of EGFR-specific fusion protein 144893 on Maleimide-activated plate

Maleimide activated plates provide covalent attachment of the sulfhydryl-containing fusion proteins of the invention (Cysteine at the C-terminus). Different wells of Maleimide-activated plates (Pierce, Thermo Scientific Cat.No.15150 Lot#KH137507) were incubated with 1 μΜ, 0.5 μΜ, 0.1 μΜ, 0.05 μΜ of fusion protein 144893 (not biotinylated) or without fusion protein in binding buffer (0.1 M sodium phosphate, 0.15 M sodium chloride, 10 mM EDTA; pH 7.2) overnight at 4 °C. After washing, plates were blocked with 10 μg/ml Cysteine and coupling efficiency of fusion protein was analyzed by Elisa with anti-Ubi-Fab-HRP as described above. Figure 10A confirms the specific binding of fusion protein 144893 on Maleimide- activated plates in a range of 1 μΜ to 0.05 μΜ.

Example 15. Cell binding on 144893-coupled Maleimide-activated plate

A549-cells were detached from culture flask by trypsin/EDTA and washed in 10 % FCS in PBS. Wells of 144893-coupled Maleimide-activated plate were incubated with 100 μΙ A549-cells in 10 % FCS/PBS with a concentration of 5 x 10⁵ cells/ml for 15 min at rt. Unbound cells were removed by washing steps and cells bound to the immobilized fusion protein were analyzed by microscopy. Strong specific binding of A549-cells was observed with protein concentrations between 1 μΜ and 0.1 μΜ (Figure 10B). Decreased protein concentrations lead to less cell binding. Control wells show no cell binding.

Example 16. Capturing of PD1 -expressing cells from blood

10 ml with 4 x 10⁵ cells/ml PD1-positive EL4-cells and PD1-negative K562-cells were stained with CellTracker™Green CMFDA Dye respectively. 25 μΙ of 1 x 10⁷ CELLection Biotin Binder Dynabeads^® were incubated with 0.2 μg/ml of either biotinylated PD1 -specific fusion protein (151318) or control fusion protein (148161 ). After washing, protein coupling to beads was analyzed by ELISA using anti-Ubi-Fab- POD. 100 μΙ of 1 x 10⁶ cells/ml (stained EL4-cells or K562-cells) were added to 1 ml mouse blood (Centre for Experimental Medicine, Leipzig) or PBS-diluted blood (1 :2) and mixed with 151318-coupled beads or 148161 -coupled beads. The bead-blood-cell mix was incubated 1 h at room temperature with constant rotation. The beads were focused by magnetic particle concentrator Dynabeads^® MPC^®-S and the supernatant was removed. Beads were washed three times with 1 x PBS and analyzed by microscopy with phase contrast and fluorescence settings with an excitation and emission maximum of 494/519 nm. Results are shown in Figure 11. The specific isolation of cells by Affilin fusion protein-coupled beads from blood-cell-mix is confirmed.

Claims

1 . A fusion protein for capturing of cells expressing a cell-surface target comprising

at least one non-lg protein that is capable of binding a cell-surface target expressed on cells with a dissociation constant K_D of 1 μΜ or less, and

at least one cleavage site for the release of the cells bound to the fusion proteins, and a peptide linker consisting of 5 to 50 amino acids covalently connecting the non-lg protein and the cleavage site, and

optionally a second peptide linker of 5 to 50 amino acids covalently connecting the protease cleavage site and the coupling moiety, and

a coupling moiety enabling chemical coupling of the fusion protein to a solid support comprising a reactive chemical group which specifically forms a covalent bond with a chemical group to enable coupling of the fusion protein to a solid support, selected from an electrophilic group, nucleophilic group, redox-active group, or a group enabling addition reactions, or cyclic addition reactions, or click reactions, or combinations thereof.

2. The fusion protein according to claim 1 wherein the order of the parts of the fusion

protein from the N-terminus to the C-terminus or the C-terminus to the N-terminus is (i) non-lg protein capable of binding a cell-surface target - peptide linker 1 - cleavage site - peptide linker 2 - coupling moiety, or (ii) non-lg protein capable of binding a cell- surface target - peptide linker - cleavage site - coupling moiety.

3. The fusion protein according to claim 1 or 2 wherein the non-lg protein capable of

binding a cell-surface target is selected from the group of a ubiquitin mutein, mutein of domains of protein A, ankyrin repeat protein mutein, lipocalin mutein, mutein of human Fyn SH3 domain, mutein of the tenth domain of human fibronectin, mutein of Kunitz domains of various protease inhibitors, Sac7d mutein, chagasin mutein, mutein of multimerized low density lipoprotein receptor-A, mutein of fibronectin, mutein of FN3 domain, mutein of cysteine-knot miniprotein, mutein of Armadillo-repeat protein, mutein of tetranectin, mutein C-type lectin domain or mutein of CTLA4, or an antibody fragment.

4. The fusion protein according to claim 1 to 3 wherein the non-lg protein capable of

binding a cell-surface target is a ubiquitin mutein that exhibits 80 % to 94 % identity to ubiquitin (SEQ ID NO: 1 ) or 80 % to 94 % identity to the ubiquitin-dimer of SEQ ID NO:

2.

5. The fusion protein according to any of the preceding claims, wherein the cleavage site is a protease cleavage site, preferably a sequence specific cleavage site.

6. The fusion protein according to claim 5 wherein the sequence specific cleavage site is a viral protease cleavage site, preferably selected from the group of tobacco vein mottling virus (TVMV) protease, tobacco etch virus (TEV) protease, plum pox virus (PPV) Nla protease, and a turnip yellow mosaic virus (TYMV) protease.

7. The fusion protein according to any of the preceding claims, wherein the first and the second peptide linker consist of 10 to 30 amino acids.

8. The fusion protein according to claim 7, wherein the amino acids of the first and the second peptide linker are selected from Gly, Ser, Ala, or Pro, preferably wherein the first and the second peptide linker consist of the amino acids alanine, proline, and serine.

9. The fusion protein according to any of the preceding claims wherein the first and the second peptide linker consist of amino acid sequences SAPASPAPAA (SEQ ID NO: 12), SAPAPSSAPAASAPPAAASA (SEQ ID NO: 13), ASPAAPAPASPAAPA (SEQ ID NO: 14), or ASPAPAAAPSAPAPA (SEQ ID NO: 15), or of amino acid sequences with 90 % identity.

10. The fusion protein according to any of the preceding claims, wherein the coupling

moiety is selected from cysteine or maleimide.

1 1 . The fusion protein according to any of the preceding claims, wherein the cell-surface target for the ubiquitin mutein is a diagnostically relevant cell-surface protein, preferably selected from cancer-related proteins selected from proteins or epitopes that are present on the outside of a tumor cell but that are absent or less expressed on non-tumor cells or which are present in tumor tissue but absent or rare on normal tissue, preferably wherein the cell-surface target is selected from Her2, EGFR, PD1 , EpCAM, NCAM, CEA, PDGFR, VEGFR, HGFR, Her3, Her4, EphB4 receptor tyrosine kinase, PSMA, MUC-1 , folate receptor, mesothelin, ALK, androgen receptor, AXL, MET, CD (for example CD31 , CD99, CD1 17, CD45, CD33, CD44, CD24, CD19, CD20, CD52, CD30), EMA, melan- A/MART-1 .

12. The fusion protein according to any of the preceding claims, wherein the fusion protein comprises or consists of an amino acid sequence selected from the group consisting of SEQ ID NO: 6, 8, 9, 1 1 , or at least 90 % identical amino acid sequences.

13. Use of the fusion protein of any one of claims 1 to 12 in applications for capturing of cells.

14. A method of capturing cells comprising the steps of (a) providing a solid support

comprising an immobilized fusion protein of any one of claims 1 to 12; (b) providing a liquid sample containing cells, (c) contacting said liquid sample and said solid support, wherein said immobilized fusion protein binds to a specific cell-surface target expressed on cells, (d) optionally separating the cells immobilized to the solid support from unbound cells and the liquid and (e) cleaving said immobilized cells from the solid support by protease cleavage thereby obtaining an eluate containing said cells expressing the specific cell-surface target.

15. A nucleic acid molecule encoding the fusion protein as defined in any one of claims 1 to 12.

16. A vector comprising the nucleic acid molecule of claim 15.

17. A host cell or a non-human host comprising the fusion protein as defined in any one of claims 1 to 12, a nucleic acid as defined in claim 15, or a vector of claim 16.

18. A method of production of the fusion protein of any one of the claims 1 to 12, comprising the step(s): culturing of the host cell of claim 17 under suitable conditions for the expression of the fusion protein; and optionally isolating said fusion protein.