WO2023017115A1

WO2023017115A1 - Fractionation of cells based on a marker protein

Info

Publication number: WO2023017115A1
Application number: PCT/EP2022/072533
Authority: WO
Inventors: Dennis KARTHAUS
Original assignee: Iba Lifesciences Gmbh
Priority date: 2021-08-11
Filing date: 2022-08-11
Publication date: 2023-02-16

Abstract

The present invention relates to method for fractionating cells of a population of cells based on the amount of a marker protein on the cell surface, wherein the population of cells comprises cells comprising a nucleic acid comprising (i) an expression cassette encoding for a marker protein comprising a binding partner B and (ii) a protein of interest, wherein the solid phase comprises a ligand L, wherein the binding partner B is capable of reversibly binding to the ligand L, wherein the reversible bond formed between the binding partner B comprised in the marker protein and the ligand L of the solid phase is displaceable (disruptable) by application of a competitor.

Description

FRACTIONATION OF CELLS BASED ON A MARKER PROTEIN

CROSS-REFERENCE TO RELATED APPLICATIONS

[1] The present application claims the right of priority of European patent application 21 190 744.9 filed with the European Patent Office on 11 August 2021 , the entire content of which is incorporated herein for all purposes.

TECHNICAL FIELD OF THE INVENTION

[2] The present invention relates inter alia to method of fractionating cells of a population of cells based on the amount of a marker protein on the cell surface, wherein the population of cells comprises cells comprising a nucleic acid comprising (i) an expression cassette encoding for a marker protein comprising a binding partner B and (ii) a protein of interest, wherein the solid phase comprises a ligand L, wherein the binding partner B is capable of reversibly binding to the ligand L, wherein the reversible bond formed between the binding partner B comprised in the marker protein and the ligand L of the solid phase is displaceable (disruptable) by application of a competitor.

BACKGROUND

[3] Pure populations of transfected or transduced cells are commonly isolated from mixed samples by co-expression of the gene or shRNA of interest with three sorts of phenotypic marker: an exogenous gene encoding drug or antibiotic resistance; an internal fluorescent protein, such as GFP, enabling Fluorescence-Activated Cell Sorting (FACS); or a cell surface protein combined with antibody labelling. Where antibody labelling of a cell surface marker is used, antibodies may be either conjugated to a fluorochrome for FACS, or to biotin for affinity purification using a solid streptavidin-conjugated matrix, typically magnetic beads.

[4] Matheson et al. (2014), PLOS One, 9(10):e111437 describe methods for isolating cells form a population of cells using a marker protein. However, no method for fractionating cells or enriching highly-producing cells has been described.

[5] WO 2018/134691 A2 discloses cell surface conjugates comprising a cell surface molecule and an agent being capable of binding a streptavidin, streptavidin analog or streptavidin mutein and disclose a method of identifying cells transduced with a cell surface conjugate. However, no method for fractionating cells or enriching highly-producing cells has been described. [6] Thus, there are currently no methods described in prior art, which allow fractionating (sorting) cells of a population of cells, which are characterized by a stable and high expression of a protein of interest. Accordingly, there is still a need for an improved method for fractionating or isolating cells from a population of cells, which are characterized by a stable and high expression of a protein of interest. The present invention aims to address this need.

SUMMARY OF THE INVENTION

[7] This need is solved by the subject-matter as defined in the claims and in the embodiments described herein.

[8] Accordingly, the present invention relates to a method for fractionating cells of a population of cells based on the amount of a marker protein on the cell surface, the method comprising:

(a) contacting the population of cells with a solid phase, wherein the population of cells comprises cells comprising a nucleic acid comprising (i) an expression cassette encoding for a marker protein comprising a binding partner B and (ii) a protein of interest, wherein the solid phase comprises a ligand L, wherein the binding partner B is capable of reversibly binding to the ligand L, wherein the reversible bond formed between the binding partner B comprised in the marker protein and the ligand L of the solid phase is displaceable (disruptable) by application of a competitor;

(b) separating the cells of the population of cells, which are not bound to the solid phase;

(c) adding the competitor to a first concentration, thereby eluting a fraction of the population of cells from the solid phase;

(d) separating the eluted fraction of the population of cells obtained in step (c);

(e) optionally adding the competitor to a second concentration, thereby eluting a fraction of the population of cells from the solid phase;

(f) separating the eluted fraction of the population of cells obtained in step (e);

(g) optionally adding the concentration to a further concentration, thereby eluting a fraction of the population of cells from the solid phase;

(h) optionally separating the eluted fraction of the population of cells eluting in step (g);

(i) optionally repeating steps (g) and (h) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more times. [9] The present invention also relates to a method for enriching or isolating cells comprised of a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells, the method comprising:

(i) optionally repeating steps (g) and (h) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more times.

[10] Accordingly, the present invention further relates to a method for fractionating cells of a population of cells based on the amount of a receptor molecule on the cell surface, the method comprising:

(a) contacting the population of cells with a solid phase and a receptor molecule binding reagent, wherein the receptor molecule binding reagent comprises a binding partner C, which is capable of specifically binding to the receptor molecule on the cell surface, and comprises a binding partner B, wherein the solid phase comprises a ligand L, wherein the binding partner B is capable of reversibly binding to the ligand L, wherein the reversible bond formed between the binding partner B comprised in the marker protein and the ligand L of the solid phase is displaceable (disruptable) by application of a competitor;

[11] Additionally, the method of the invention relates to a method for enriching or isolating cells comprised in a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells, the method comprising:

(e) optionally adding the competitor to a second concentration, thereby eluting a fraction of the population of cells from the solid phase; (f) separating the eluted fraction of the population of cells obtained in step (e);

[12] The population of cells may comprise cells comprising a nucleic acid comprising (i) an expression cassette encoding for a marker protein comprising and (ii) a protein of interest.

[13] The first concentration, the second concentration and the further concentration(s) may be essentially the same.

[14] The second concentration may be higher than the first concentration, the further concentration may be higher than the second concentration and each subsequent further concentration may be higher than the previous further concentration.

[15] The second concentration may be lower than the first concentration, wherein the further concentration is lower than the second concentration and each subsequent further concentration is lower than the previous further concentration.

[16] The amount of the marker protein on the cell surface of a cell of the population of cells may be correlated to the level of expression of the protein of interest.

[17] In the expression cassette, the protein of interest and the marker protein may be operably linked to different promotors.

[18] In the expression cassette, the protein of interest may be operably linked to a promotor and the cell surface signal may be under the control of an internal ribosome entry site (IRES), wherein the protein of interest and the cell surface signal are transcribed on the same RNA.

[19] In the expression cassette, the protein of interest and the marker protein may form a fusion protein.

[20] In the expression cassette, the protein of interest and the marker protein may form a fusion protein, wherein the marker protein and the fusion protein are linked via a self-cleavable peptide such as a 2A peptide.

[21] In the expression cassette, the protein of interest and the marker protein may form a fusion protein, wherein the marker protein and the fusion protein are linked via a cleavable amino acid sequence that can be cleaved by a protease. [22] In the expression cassette, a portion of the protein of interest and the marker protein may form a fusion protein due to a leaky stop codon between the sequence of the protein of interest and the sequence of the marker protein.

[23] In the expression cassette, the protein of interest may be immobilized on the cell surface via methods like cold capture (e.g., as described in Pichler et al. 2009), J Biotechnol, 141 (1- 2):80-3) which preferably stops or reduces secretion of the protein by trapping it temporally on the cell surface e.g., by reducing the temperature of the medium.

[24] The protein of interest may be a membrane protein itself.

[25] The marker protein may be a transmembrane protein or a fragment thereof, wherein the binding site B is comprised in the extracellular domain.

[26] The marker protein may be a peptide fused to a membrane anchor.

[27] The marker protein may comprise a transmembrane domain of a protein selected from the group consisting of EpCAM, VEGFR, integrin, optionally integrins avp3, a4, alip3, a4p7, a5pi , avp3 or an, a member of the TNF receptor superfamily, optionally TRAIL-RI or TRAIL-R2, a member of the epidermal growth factor receptor family, PDGF Receptor, interferon receptor, folate receptor, GPNMB, ICAM-1 , HLA-DR, CEA, CA-125, MUCI, TAG-72, IL-6 receptor, 5T4, GD2, GD3, prostate-specific membrane antigen (PSMA) or a clusters of differentiation cell surface molecule, optionally CD2, CD3, CD4, CD5, CD11 , CDIIa/LFA-1, CD15, CD18/ITGB2, CD19, CD20, CD22, CD23/lgE Receptor, CD25, CD28, CD30, CD33, CD38, CD40, CD41 , CD44, CD51 , CD52, CD62L, CD74, CD80, CD125, CD147/basigin, CD152/CTLA-4, CD154/CD40L, CD195/CCR5 and CD319/SLAMF7, low-affinity nerve growth factor receptor (LNGFR), preferably CD4.

[28] The marker protein may wherein the marker protein comprises or consists of any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25, or a fragment or analog thereof having a sequence identity of 60% or higher compared to any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25.

[29] The the binding partner B and the ligand L may form a binding pair selected from the group of: streptavidin or a streptavidin analog and a ligand binding to streptavidin, a binding pair that binds in the presence of a divalent cation, an oligohistidine peptide and a binding moiety A comprising at least two chelating groups K, wherein each chelating group K is capable of binding to a transition metal ion, thereby rendering binding moiety A capable of binding to the oligohistidine peptide, an antigen and an antibody against said antigen, wherein said binding partner B comprises the antigen and said ligand L comprises the antibody against said antigen.

[30] The fractions of the cells may be eluted by adding biotin.

[31] Biotin may be added to a concentration of at least 3 pM, at least 5 pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin.

[32] The solid phase may be a selected from a bead, a plastic plate, a membrane or a stationary phase suitable for chromatography.

[33] The method may be a batch method. The method may be a batch method a chromatographic method.

[34] The nucleic acid comprised in cells of the population of cells may further comprise a selection marker. The method of the invention may further comprise a step (a’) prior to step (a):

(a’) selecting cells, which express the selection marker.

[35] The host cell may be a prokaryotic cell or a eukaryotic cell.

[36] The protein of interest may be selected from the group consisting of an amylolytic enzyme, a lipolytic enzyme, a proteolytic enzyme, a cellulytic enzyme, an oxidoreductase or a plant cell-wall degrading enzyme, an aminopeptidase, amylase, amyloglucosidase, carbohydrase, carboxypeptidase, catalase, cellulase, chitinase, cutinase, cyclodextrin glycosyltransferase, desoxyribonuclease, esterase, galactosidase, beta-galactosidase, glucoamylase, glucose oxidase, glucosidase, haloperoxidase, hemicellulase, invertase, isomerase, laccase, ligase, lipase, lyase, mannosidase, oxidase, pectinase, peroxidase, phytase, phenoloxidase, polyphenoloxidase, protease, ribonuclease, transferase, transglutaminase, and xylanase, a growth factor, cytokine, receptors, receptor ligands, therapeutic proteins such as interferons, BMPs, GDF proteins, fibroblast growth factors, peptides such as protein inhibitors, membrane proteins, membrane-associated proteins, peptide/protein hormones, peptidic toxins, peptidic antitoxins, antibody or functional fragments thereof such as Fab or F(ab)₂ or derivatives of an antibody such as bispecific antibodies (for example, scFvs), chimeric antibodies, humanized antibodies, single domain antibodies such as Nanobodies or domain antibodies (dAbs) or an anticalin and others. BRIEF DESCRIPTION OF THE DRAWINGS

[37] The invention will be better understood with reference to the detailed description when considered in conjunction with the non-limiting examples and the accompanying drawings, in which:

[38] Figure 1 depicts an exemplary embodiment of the methods for fractionating cells of a population of cells based on the amount of a marker protein on the cell surface of the invention and the methods for enriching or isolating cells comprised in a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells. A cell (1) comprised in the population of cells, which expresses a protein of interest (5) (e.g., a secreted or an intra-cellular protein) and a marker protein as defined herein, e.g., a recombinant fusion protein that comprises a transmembrane domain (2), an extracellular domain, which can be a protein domain and/or an amino acid linker sequence (4), and a binding partner B (3) (e.g., a streptavidin-binding peptide (e.g. Strep-tag®l I peptide) having a sequential arrangement of two or more individual binding modules as described in International Patent Publication WO 02/077018 or U.S. Pat. No. 7,981,632). The population of cells, which comprises the target cells or cells to be fractionated is contacted with a solid phase (e.g. a magnetic bead, an agarose bead or membrane) (8) comprising a ligand L (7), e.g., a streptavidin mutein or a multimerized streptavidin mutein (e.g. Strep-Tactin®) (6) that can bind the streptavidin-peptide, optionally coupled to the solid phase (8) by a linker (7). The binding partner B (3) is capable of reversibly binding to the ligand L, wherein the reversible bond formed between the binding partner B (7) comprised in the marker protein (2,3,4) and the ligand L (6) of the solid phase (8) is displaceable (disruptable) by application of a competitor, e.g. biotin. Not depicted here is an alternative embodiment, in which the binding partner B is comprised in a receptor molecule binding reagent that specifically binds to a receptor molecule (or marker protein) on the cells of the cell populations. Here, the receptor molecule binding reagent acts as some kind of link between the solid phase and its ligand L to the receptor molecule (or marker protein). The underlying principles of the invention are the same.

[39] Figure 2 depicts the exemplary embodiment of the method of the invention after contacting the solid phase (8) with the cell (1). Here, a cell (1) comprising one or more marker proteins (2,3,4) comprising a binding partner B (3) is bound to one or more ligands L (6) comprised by the solid phase (8). Cells that exhibit no expression of the marker protein are not bound to the matrix and can be removed from the sample by consecutive wash steps.

[40] Figure 3 depicts an elution step in the method of the invention. Here, a competitor (e.g. biotin) (9) that can displace (disrupt) the reversible bond between the binding partner B (3) and the ligand (6) is added to the immobilized cells to a first concentration, e.g., to at least 5 pM biotin. The left cell is a cell exhibiting a high expression of the recombinant proteins. The right cell is a cell exhibiting a low expression of the recombinant proteins. The concentration of the competing agent is sufficient to dissociate the majority of low expressing cells from the matrix so they can be removed from the sample by the removal of the liquid phase of the mixture (elution fraction 1). A substantial number of the high expressing cells are still bound to the matrix due more interactions between the binding partner B (3) and the ligand L (6). Therefore, the concentration of the competing reagent is insufficient to break all bindings between the high expressing cell and the matrix or the portion of bound high expressing cells in the dynamic equilibrium leads to a lower elution rate of the cells compared to low producing cells. After elution of at least a fraction of the cells from the solid phase, the cells are separated.

[41] Figure 4 depicts the optional second elution step of the method of the invention. Here, the competitor (9) is added to a second concentration. The second concentration may be essentially the same, e.g., at least 5 pM Biotin, or higher (or lower in some embodiments) than the first concentration. Consecutive incubations with competitors at a second or further concentration(s) can lead to dissociation of high expressing cells from the solid phase by displacement, thereby fractionating the population of cells based on the amount of a marker protein on the cell surface. Dissociated cells in the liquid phase of the mixture can be removed. As described herein, the amount of the marker protein of the cell surface is correlated to the expression of the protein of interest. Thus, fractionating the cells according to the amount of a marker protein on the cell surface also fractionates the cells according to the rate of production of the protein of interest.

[42] Figure 5 depicts the expression cassette of a p2458 vector comprising a cytomegalovirus promoter (CMV, SEQ ID NO 1), enhanced green fluorescence protein (eGFP, SEQ ID NO 2), an internal ribosome entry site (IRES, SEQ ID NO 3), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5), a CD137 protein (CD137, SEQ ID NO 6) and a poly(A) signal sequence (pA, SEQ ID NO 7). The TST and CD137 are linked and form a fusion protein. The elements from BM40 to CD137 are shown in SEQ ID NO 8 and form the marker protein.

[43] Figure 6 depicts the expression cassette of the pCDNA3-hEF-BM40-TST-CD4-eGFP vector comprising a human elongation factor-1 alpha promoter (EF1 a, SEQ ID NO 9), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5), a CD4 protein comprising the transmembrane and extracellular domain (CD4, SEQ ID NO 15) an enhanced green fluorescence protein (eGFP, SEQ ID NO 2) and a poly(A) signal sequence (pA, SEQ ID NO 7). The TST, CD4 and eGFP proteins are linked and form a fusion protein. The elements from BM40t o eGFP are shown in SEQ ID NO 16. [44] Figure 7 depicts the expression cassette of the pCDNA3-hEF-BM40-TST-flex-CD4tm- eGFP vector comprising a human elongation factor-1 alpha promoter (EF1a, SEQ ID NO 9), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5), a CD4 transmembrane domain (SEQ ID NO 12) with a flexible linker (SEQ ID NO 10) to the TST (flex-CD4tm), an enhanced green fluorescence protein (eGFP, SEQ ID NO 2) and a poly(A) signal sequence (pA, SEQ ID NO 7). The TST, flex-CD4tm and eGFP proteins are linked and form a fusion protein. The elements from BM40a to eGFP are shown in SEQ ID NO

17.

[45] Figure 8 depicts the expression cassette of the pCDNA3-hEF-BM40-TST-rigid-CD4tm- eGFP vector comprising a human elongation factor-1 alpha promoter (EF1a, SEQ ID NO 9), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5), a CD4 transmembrane domain (SEQ ID NO 12) with a rigid linker (SEQ ID NO 11) to the TST (rigid-CD4tm), an enhanced green fluorescence protein (eGFP, SEQ ID NO 2) and poly(A) signal sequence (pA, SEQ ID NO 7). The TST, rigid-CD4tm and eGFP proteins are linked and form a fusion protein. The elements from BM40 to eGFP are shown in SEQ ID NO

18.

[46] Figure 9 depicts the expression cassette of the pCDNA3-B18R-PGK-BM40-TST- CD4tm-EGFP-F2A-Puro vector comprising a human elongation factor-1 alpha promoter (EF1a, SEQ ID NO 9), a B18R expression cassette comprising a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5) fused to a B18R protein (SEQ ID NO 24) and a poly(A) signal sequence (SEQ ID NO 7) (B18R). The marker cassette comprises a phosphoglycerate kinase promoter (PGK, SEQ ID NO 19), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5), a CD4 transmembrane domain (SEQ ID NO 12) with a flexible linker (SEQ ID NO 10) to the TST (flex-CD4tm), an enhanced green fluorescence protein (eGFP, SEQ ID NO 2), a F2A selfcleaving sequence (2A, SEQ ID NO 20), a puromycin resistance gene (PuroR, SEQ ID NO 21) and a poly(A) signal sequence (pA, SEQ ID NO 22). The TST, rigid-CD4tm and eGFP proteins are linked and form a fusion protein. The elements from BM40 to PuroR are shown in SEQ ID NO 23.

[47] Figure 10 depicts the expression cassette of the pZSG5-SEAP vector comprising a human elongation factor-1 alpha promoter (EF1a, SEQ ID NO 9), a SEAP expression cassette comprising a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep- tag® (TST, SEQ ID NO 5) fused to the C-terminus of a SEAP protein (SEQ ID NO 28) and a poly(A) signal sequence (SEQ ID NO 22) (SEAP). The marker cassette comprises a phosphoglycerate kinase promoter (PGK, SEQ ID NO 26), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5), a CD4 transmembrane domain (SEQ ID NO 12) with a flexible linker (SEQ ID NO 10) to the TST (flex- CD4tm), an enhanced green fluorescence protein (eGFP, SEQ ID NO 2), a F2A self-cleaving sequence (2A,SEQ ID NO 29), a puromycin resistance gene (PuroR, SEQ ID NO 21) and a poly(A) signal sequence (pA.SEQ ID NO 7). The TST, rigid-CD4tm and eGFP proteins are linked and form a fusion protein. The elements from BM40 to PuroR are shown in SEQ ID NO 27.

[48] Figure 11 depicts the expression cassette of the pZSG4-SEAP vector comprising a human elongation factor-1 alpha promoter (EF1a, SEQ ID NO 9), a SEAP expression cassette comprising a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4) and a Twin-Strep- tag® (TST, SEQ ID NO 5) fused to the C-terminus of a SEAP protein (SEQ ID NO 28) (SEAP). The marker cassette comprises an IRES (SEQ ID NO 25), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep-tag® (TST, SEQ ID NO 5), a CD4 transmembrane domain (SEQ ID NO 12) with a flexible linker (SEQ ID NO 10) to the TST (flex- CD4tm), an enhanced green fluorescence protein (eGFP, SEQ ID NO 2), a F2A self-cleaving sequence (2A, SEQ ID NO 29), a puromycin resistance gene (PuroR, SEQ ID NO 21) and a poly(A) signal sequence (pA,SEQ ID NO 7). The TST, rigid-CD4tm and eGFP proteins are linked and form a fusion protein. The elements from BM40 to PuroR are shown in SEQ ID NO 27.

[49] Figure 12 depicts the expression cassette of the pZSG7-eGFP vector comprising a human elongation factor-1 alpha promoter (EF1a, SEQ ID NO 9), an enhanced green fluorescence protein (eGFP, SEQ ID NO 2) followed by a poly(A) signal sequence (SEQ ID NO 22) (eGFP). The marker cassette comprises a phosphoglycerate kinase promoter (PGK, SEQ ID NO 26), a BM40 secretion signal peptide sequence (BM40, SEQ ID NO 4), a Twin-Strep- tag® (TST, SEQ ID NO 5), a CD4 transmembrane domain that comprises either the SEQ ID NO 12 or SEQ ID NO 13 or SEQ ID NO 14 with a flexible linker (SEQ ID NO 10) to the TST (flex- CD4tm). In SEQ ID NO 13 the last four C-terminal arginine were replaced by alanine. In SEQ ID NO 14 the last nine C-terminal amino acids were removed. The last element of the marker cassette comprises a poly(A) signal sequence (pA, SEQ ID NO 7).

[50] Figure 13 shows HEK-293 cells expressing GFP and co-expressing a truncated CD137 receptor fused to a Twin-Strep-tag® before the selection using the method of invention. Cells were analyzed on an Accuri C6 flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. Viable GFP positive cells were identified by mean fluorescence intensity area (mean Fl-A). [51] Figure 14 shows the results of enriching HEK-293 cells expressing GFP and coexpressing a truncated CD137 receptor fused to a Twin-Strep-tag®. Cultures were selected with the method of invention. Step-Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and elution using a 1 mM biotin Buffer IS solution. Cells were analyzed on an Accuri C6 flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. Viable GFP positive cells were identified by mean fluorescence intensity area (mean Fl-A).

[52] Figure 15 shows the results of enriching HEK-293 cells expressing GFP and coexpressing a truncated CD137 receptor fused to a Twin-Strep-tag®. Cultures were selected with the method of invention. Step-Tactin magnetic microbeads pre-incubated with anti CD137-TST- Fab were used as solid phase and incubated with the cells followed by wash steps and elution using a 1 mM biotin Buffer IS solution. Cells were analyzed on an Accuri C6 flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. Viable GFP positive cells were identified by mean fluorescence intensity area (mean Fl-A).

[53] Figure 16 shows CHO cells expressing GFP and co-expressing a truncated CD137 receptor fused to a Twin-Strep-tag® before the selection using the method of invention. Cells were analyzed on an Accuri C6 flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. Viable GFP positive cells were identified by mean fluorescence intensity area (mean Fl-A).

[54] Figure 17 shows the results of enriching CHO cells expressing GFP and co-expressing a truncated CD137 receptor fused to a Twin-Strep-tag®. Cultures were selected with the method of invention. Step-Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and elution using a 1 mM biotin Buffer IS solution. Cells were analyzed on an Accuri C6 flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. Viable GFP positive cells were identified by mean fluorescence intensity area (mean Fl-A).

[55] Figure 18 shows the results of enriching CHO cells expressing GFP and co-expressing a truncated CD137 receptor fused to a Twin-Strep-tag®. Cultures were selected with the method of invention. Step-Tactin magnetic microbeads pre-incubated with anti CD137-TST-Fab were used as solid phase and incubated with the cells followed by wash steps and elution using a 1 mM biotin Buffer IS solution. Cells were analyzed on an Accuri C6 flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. Viable GFP positive cells were identified by mean fluorescence intensity area (mean Fl-A). [56] Figure 19 shows the fraction of fluorescent cells in three MEXi-293E cultures expressing different variants of the membrane selection marker gene before selection using the method of invention. Left: Cells expressing CD4 transmembrane and extracellular domain fused to GFP (C-terminal) and TST (N-terminal). Middle: Cells expressing CD4 transmembrane domain fused to GFP (C-terminal) and TST (N-terminal) with a rigid linker between CD4 and TST. Right: Cells expressing CD4 transmembrane domain fused to GFP (C-terminal) and TST (N-terminal) with a flexible linker between CD4 and TST.

[57] Figure 20 shows the fraction of fluorescent cells in three MEXi-293E cultures expressing different variants of the membrane selection marker gene after selection using the method of invention. Left: Cells expressing CD4 transmembrane and extracellular domain fused to GFP (C-terminal) and TST (N-terminal). Middle: Cells expressing CD4 transmembrane domain fused to GFP (C-terminal) and TST (N-terminal) with a rigid linker between CD4 and TST. Right: Cells expressing CD4 transmembrane domain fused to GFP (C-terminal) and TST (N-terminal) with a flexible linker between CD4 and TST.

[58] Figure 21 shows the results of experiments for enriching MEXi-293E cells expressing TST-CD4-variants fused to GFP. Cultures were selected with the method of invention. Step- Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and consecutive elution using biotin enriched MEXi-CM. Cells in the different fractions were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. Viable GFP positive cells were identified by forward scatter area and GFP signal. TST-rigid-CD4tm-GFP corresponds to cells expressing a CD4 transmembrane domain fused via a rigid linker to a TST (N-terminal) and to GFP (C-terminal). TST-flexible-CD4tm-GFP corresponds to cells expressing a CD4-transmembrane domain fused via a flexible linker to a TST (N-terminal) and to GFP (C-terminal). TST-CD4-truncated-GFP corresponds to cells expressing a CD4 variant where the intra cellular domain was removed. The transmembrane domain was fused to GFP (C-terminal) and the N-terminus of the protein was fused to a TST.

[59] Figure 22 shows the mean fluorescence intensity of the TST-CD4-GFP positive population in different fractions from the selected populations in Figure 21.

[60] Figure 23 shows the yield of selected cells in each elution fraction expressing different CD4 variants fused to a TST (N-terminal) and GFP (C-terminal) as described in Figure 21.

[61] Figure 24 shows the results of experiments for enriching TST-flexible-CD4tm-GFP (SEQ ID NO 23 positive) expressing CHO cells. CHO-P and CHO-P-48h-l cell cultures were selected with the method of invention. Step-Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and consecutive elutions using biotin enriched Buffer IS. Cells in the different fractions were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[62] Figure 25 shows the mean fluorescence intensity of the selected GFP positive population in different fractions of the CHO cell cultures CHO-P and CHO-P-48h-l in Figure 24. Roman numerals indicate the numbers of selections using the method of invention.

[63] Figure 26 shows the trend of GFP positive cell fraction during the cultivation of TST- flexible-CD4tm-GFP expressing CHO cells that were selected by puromycin only (CHO-P) or in addition with the method of invention (CHO-P-48h-l and CHO-P-48h-ll). CHO-P-48h-l culture was derived by selection with the method of invention 2 days post transfection. CHO-P-48h-ll culture was derived by the cultivation of selected cells from the CHO-P-48h-l selection. Cells were cultivated in CHO-TF medium with puromycin until day 24 (dashed line). After day 24 the cultivation was continued in absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[64] Figure 27 shows the trend of mean fluorescence intensity of the GFP positive cell fraction during the cultivation of TST-flexible-CD4tm-GFP expressing CHO cells that were selected by puromycin only (CHO-P) or in addition with the method of invention (CHO-P-48h-l and CHO-P-48h-ll). CHO-P-48h-l culture was derived by selection with the method of invention 2 days post transfection. CHO-P-48h-ll culture was derived by the cultivation of selected cells from the CHO-P-48h-l selection. Cells were cultivated in CHO-TF medium with puromycin until day 24 (dashed line). After day 24 the cultivation was continued in absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[65] Figure 28 shows the viability during the cultivation of TST-flexible-CD4tm-GFP expressing CHO cells that were selected by puromycin only (CHO-P) or in addition with the method of invention (CHO-P-48h-l and CHO-P-48h-ll). CHO-P-48h-l culture was derived by selection with the method of invention 2 days post transfection. CHO-P-48h-ll culture was derived by the cultivation of selected cells from the CHO-P-48h-l selection. Cells were cultivated in CHO-TF medium with puromycin until day 24 (dashed line). After day 24 the cultivation was continued in absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[66] Figure 29 shows the results of experiments for enriching TST-flexible-CD4tm-GFP expressing MEXi-293E cells. MEXi-P-48h-l and MEXi-P-48h-ll cell cultures were selected with the method of invention. Multimeric Strep-Tactin® mutant 2 coated agarose beads were used as solid phase and incubated with the cells followed by wash steps and consecutive elution using biotin enriched MEXi-TM. Cells in the different elution fractions were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. MEXi-P- 48h-l culture was derived by selection with the method of invention 2 days post transfection. MEXi-P-48h-ll culture was derived by the cultivation of selected cells from the MEXi-P-48h-l selection. Roman numerals indicate the numbers of selections using the method of invention.

[67] Figure 30 shows the mean fluorescence intensity of the GFP positive population in different fractions of the MEXi-293E cell cultures MEXi-P-48h-l and MEXi-P-48h-ll in Figure 29.

[68] Figure 31 shows the results of experiments for enriching TST-flexible-CD4tm-GFP expressing MEXi-293E cells. MEXi-P, MEXi-P-48h-l and MEXi-P-48h-ll cell cultures were selected with the method of invention. Step-Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and consecutive elution using biotin enriched MEXi-CM. Cells in the different fractions were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. MEXi-P-48h-l culture was derived by selection with the method of invention 2 days post transfection. MEXi-P-48h-ll culture was derived by the cultivation of selected cells from the MEXi-P-48h-l selection. Roman numerals indicate the numbers of selections using the method of invention.

[69] Figure 32 shows the mean fluorescence intensity of the GFP positive population in different fractions of the MEXi-293E cell cultures MEXi-P, MEXi-P-48h-l and MEXi-P-48h-ll in Figure 31.

[70] Figure 33 shows the trend of GFP positive cell fraction during the cultivation of TST- flexible-CD4tm-GFP expressing MEXi cells that were selected by puromycin only (MEXi-P) or in addition with the method of invention (MEXi-P-48h-l, MEXi-P-48h-ll and MEXi-P-48h-l II). MEXi- P-48h-l culture was derived by selection with the method of invention 2 days post transfection. MEXi-P-48h-ll culture was derived by the cultivation of selected cells from the MEXi-P-48h-l selection. MEXi-P-48h-l 11 culture was derived by the cultivation of selected cells from the MEXi- P-48h-ll selection. Cells were cultivated in MEXi-CM medium with puromycin until day 21 (second dashed line). After day 21 the cultivation was continued in absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[71] Figure 34 shows the trend of mean fluorescence intensity of the GFP positive cell fraction during the cultivation of TST-flexible-CD4tm-GFP expressing MEXi cells that were selected by puromycin only (MEXi-P) or in addition with the method of invention (MEXi-P-48h-l, MEXi-P-48h-ll and MEXi-P-48h-lll). MEXi-P-48h-l culture was derived by selection with the method of invention 2 days post transfection. MEXi-P-48h-ll culture was derived by the cultivation of selected cells from the MEXi-P-48h-l selection. M EXi-P-48h-l 11 culture was derived by the cultivation of selected cells from the MEXi-P-48h-ll selection. Cells were cultivated in MEXi-CM medium with puromycin until day 21 (second dashed line). After day 21 the cultivation was continued in absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. GFP positive cells were identified by forward scatter area and FITC signal. Roman numerals indicate the numbers of selections using the method of invention.

[72] Figure 35 shows the results of experiments for enriching TST-flexible-CD4tm-GFP expressing CHO cells. The cell cultures were selected with the method of invention. Step-Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and consecutive elution using biotin enriched Buffer IS. Cells in the different fractions were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. CHO-ZSG4-P-I culture was derived by the cultivation of selected cells from the CHO- ZSG4-P selection. CHO-ZSG5-P-I culture was derived by the cultivation of selected cells from the CHO-ZSG5-P selection. Roman numerals indicate the numbers of selections using the method of invention.

[73] Figure 36 shows the mean fluorescence intensity of the GFP positive population in different fractions from the CHO cell cultures in Figure 35.

[74] Figure 37 shows the trend of GFP positive cell fraction during the cultivation TST- flexible-CD4tm-GFP expressing CHO cells that were selected by puromycin only (CHO-ZSG4- P) or in addition with the method of invention (CHO-ZSG4-P-I, CHO-ZSG4-P-II-E2 and CHO- ZSG4-P-II-E3). The expression of the selection marker cassette was controlled by an IRES. CH0-ZSG4-P-I culture was derived by the cultivation of selected cells from the CHO-ZSG4-P selection. CHO-ZSG4-PII cultures were derived by the cultivation of selected cells from the CHO-ZSG4-P-I selection. Cells were cultivated in CHO-TF medium with puromycin until day 21 or until first selection with the method of invention. After this time, the cultivation was continued in the absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[75] Figure 38 shows the trend of mean fluorescence intensity of the GFP positive cell fraction during the cultivation of TST-flexible-CD4tm-GFP expressing CHO cells that were selected by puromycin only (CHO-ZSG4-P) or in addition with the method of invention (CHO- ZSG4-P-I, CHO-ZSG4-P-II-E2 and CHO-ZSG4-P-II-E3). The expression of the selection marker cassette was controlled by an IRES. CHO-ZSG4-P-I culture was derived by the cultivation of selected cells from the CHO-ZSG4-P selection. CHO-ZSG4-PII cultures were derived by the cultivation of selected cells from the CHO-ZSG4-P-I selection. Cells were cultivated in CHO-TF medium with puromycin until day 21 or until first selection with the method of invention. After this time, the cultivation was continued in the absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[76] Figure 39 shows the trend of GFP positive cell fraction during the cultivation of TST- flexible-CD4tm-GFP expressing CHO cells that were selected by puromycin only (CHO-ZSG5- P) or in addition with the method of invention (CHO-ZSG5-P-I, CHO-ZSG5-P-II-E2 and CHO- ZSG5-P-II-E3). The expression of the selection marker cassette was controlled by a PGK promoter. CHO-ZSG5-P-I culture was derived by the cultivation of selected cells from the CHO- ZSG5-P selection. CHO-ZSG5-PII cultures were derived by the cultivation of selected cells from the CHO-ZSG5-P-I selection. Cells were cultivated in CHO-TF medium with puromycin until day 21 or until first selection with the method of invention. After this time, the cultivation was continued in the absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[77] Figure 40 shows the trend of mean fluorescence intensity of the GFP positive cell fraction during the cultivation of TST-flexible-CD4tm-GFP expressing CHO cells that were selected by puromycin only (CHO-ZSG5-P) or in addition with the method of invention (CHO- ZSG5-P-I, CHO-ZSG5-P-II-E2 and CHO-ZSG5-P-II-E3). The expression of the selection marker cassette was controlled by a PGK promoter. CHO-ZSG5-P-I culture was derived by the cultivation of selected cells from the CHO-ZSG5-P selection. CHO-ZSG5-PII cultures were derived by the cultivation of selected cells from the CHO-ZSG5-P-I selection. Cells were cultivated in CHO-TF medium with puromycin until day 21 or until first selection with the method of invention. After this time, the cultivation was continued in the absence of puromycin. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. Roman numerals indicate the numbers of selections using the method of invention.

[78] Figure 41 : Analysis of SEAP-TST concentration in the supernatant of CHO cultures. CHO cultures were selected as described in Example 11e) Marker cassette expression was controlled either by an IRES (ZSG4 cultures, gray) or a PGK promoter (ZSG5 cultures, black). Determined SEAP-TST concentrations are presented in the legend.

[79] Figure 42 yield of SEAP-TST per ml supernatant after purification of CHO supernatants. CHO cultures were selected as described in Example 11f). Marker cassette expression was controlled by a PGK promoter. Cells were cultured as described in Example 11f).

[80] Figure 43 shows the results of experiments for enriching TST-flexible-CD4tm-GFP expressing CHO cells that were not pre-selected with puromycin. The cell cultures were selected with the method of invention only. Step-Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and consecutive elution using biotin enriched Buffer IS. Cells in the different fractions were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. CHO-ZSG4-woP-l culture was derived by the cultivation of selected cells from the CHO-ZSG4-woP selection. CHO-ZSG4- woP-ll culture was derived by the cultivation of selected cells from the CHO-ZSG4-woP-l selection. CHO-ZSG5-woP-l culture was derived by the cultivation of selected cells from the CHO-ZSG5-woP selection. CHO-ZSG5-woP-ll culture was derived by the cultivation of selected cells from the CHO-ZSG5-woP-l selection. CHO-ZSG5-woP-lll culture was derived by the cultivation of selected cells from the CHO-ZSG5-woP-ll selection. Roman numerals indicate the numbers of selections using the method of invention.

[81] Figure 44 shows the mean fluorescence intensity of the GFP positive population in different fractions from the CHO cell cultures in Figure 43.

[82] Figure 45 shows the trend of GFP positive cell fraction during the cultivation of TST- flexible-CD4tm-GFP expressing CHO cells that were selected by the method of invention only. The expression of the selection marker cassette was controlled by a PGK promoter (CHO- ZSG5-woP) or an IRES (CHO-ZSG4-woP). The graph shows the cultivation of cells that were selected as described in Example 12 Day 2 - day 7 show the cultivation of culture CHO-ZSG4- woP-l and CHO-ZSG5-woP-l. Day 7 to day 21 shows the cultivation of CHO-ZSG4-woP-ll and CHO-ZSG5-WOP-II. Day 21 to day 56 show the cultivation of CHO-ZSG5-woP-lll. Day 57 to day 96 show the cultivation of CHO-ZSG5-woP-IV. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal.

[83] Figure 46 shows the trend of mean fluorescence intensity of the GFP positive cell fraction during the cultivation of TST-flexible-CD4tm-GFP expressing CHO cells that were selected by the method of invention only. The expression of the selection marker cassette was controlled by a PGK promoter (CHO-ZSG5-woP) or an IRES (CHO-ZSG4-woP). The graph shows the cultivation of cells that were selected as described in Example 12. Day 2- day 7 show the cultivation of culture CHO-ZSG4-woP-l and CHO-ZSG5-woP-l. Day 7 to day 21 show the cultivation of CHO-ZSG4-woP-ll and CHO-ZSG5-woP-ll. Day 21 to day 56 shows the cultivation of CHO-ZSG5-woP-lll. Day 57 to day 96 show the cultivation of CHO-ZSG5-woP-IV. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter area and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal.

[84] Figure 47 yield of SEAP-TST per ml supernatant after purification of CHO supernatants. CHO cultures were selected as described in Examples 11 and 12. Marker cassette expression was controlled by a PGK promoter. CHO-ZSG5-woP-lll was derived by the consecutive selection of cells with the method of invention only. The other cell cultures were derived from a combination of the method of invention with a pre selection with puromycin.

[85] Figure 48 Analysis of SEAP-TST concentration in the supernatant of MEXi-ZSG-woP-lll cultures using a BLItz system with Strep-Tactin®XT coated BLI biosensors. Binding of SEAP- TST results in an increased signal. Higher SEAP-TST concentration results in s higher (increase of) signal over time. MEXi cultures were selected as described in Example 13. Marker cassette expression was controlled either by an IRES (ZSG4 cultures, grey) or a PGK promoter (ZSG5 cultures, black).

[86] Figure 49 shows the results of experiments for enriching TST-flexible-CD4tm-GFP expressing MEXi-293E cells that were not pre-selected with puromycin. The cell cultures were selected with the method of invention only. Step-Tactin magnetic microbeads were used as solid phase and incubated with the cells followed by wash steps and consecutive elution using biotin enriched MEXi-CM medium. Cells in the different fractions were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal. MEXi-ZSG4-woP-l culture was derived by the cultivation of selected cells from the MEXi-ZSG4-woP selection. MEXi-ZSG4-woP-ll culture was derived by the cultivation of selected cells from the MEXi-ZSG4- woP-l selection. MEXi-ZSG5-woP-l culture was derived by the cultivation of selected cells from the MEXi-ZSG5-woP selection. MEXi-ZSG5-woP-ll culture was derived by the cultivation of selected cells from the MEXi-ZSG5-woP-l selection. Roman numerals indicate the numbers of selections using the method of invention.

[87] Figure 50 shows the mean fluorescence intensity of the GFP positive population in different fractions from the MEXi cell cultures in Figure 49.

[88] Figure 51 shows the trend of GFP positive cell fraction during the cultivation of TST- flexible-CD4tm-GFP expressing H EK-293 cells that were selected by the method of invention only. The expression of the selection marker cassette was controlled by a PGK promoter (MEXi- ZSG5-woP) or an IRES (MEXi-ZSG4-woP). The graph shows the cultivation of cells that were selected as described in Example 13. Day 2 - day 5 show the cultivation of culture MEXi-ZSG4- woP-l and MEXi-ZSG5-woP-l. Day 5 to day 9 shows the cultivation of MEXi-ZSG4-woP-ll and CHO-MEXi-woP-ll. Day 9 to day 15 show the cultivation of MEXi-ZSG4-woP-lll and MEXi- ZSG5-WOP-III. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal.

[89] Figure 52 shows the trend of mean fluorescence intensity of the GFP positive cell fraction during the cultivation of TST-flexible-CD4tm-GFP expressing MEXi cells that were selected by the method of invention only. The expression of the selection marker cassette was controlled by a PGK promoter (MEXi-ZSG5-woP) or an IRES (MEXi-ZSG4-woP). The graph shows the cultivation of cells that were selected as described in Example 13. Day 2 - day 5 show the cultivation of culture MEXi-ZSG4-woP-l and MEXi-ZSG5-woP-l. Day 5 to day 9 shows the cultivation of MEXi-ZSG4-woP-ll and CHO-MEXi-woP-ll. Day 9 to day 15 show the cultivation of MEXi-ZSG4-woP-lll and MEXi-ZSG5-woP-lll. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area. GFP positive cells were identified by forward scatter area and GFP signal.

[90] Figure 53 shows the trend of viable cells during the cultivation of TST-flexible-CD4tm- GFP expressing HEK-293 cells that were selected by the method of invention only. The expression of the selection marker cassette was controlled by a PGK promoter (MEXi-ZSG5- woP) or an IRES (MEXi-ZSG4-woP). The graph shows the cultivation of cells that were selected as described in Example 13 Day 2 - day 5 show the cultivation of culture MEXi-ZSG4-woP-l and MEXi-ZSG5-woP-l. Day 5 to day 9 shows the cultivation of MEXi-ZSG4-woP-ll and CHO- MEXi-woP-ll. Day 9 to day 15 show the cultivation of MEXi-ZSG4-woP-lll and MEXi-ZSG5-woP- III. Cells were analyzed on a Cytoflex flow cytometer. Viable cells were identified via forward scatter width and sideward scatter area.

[91] Figure 54 shows portion of unbound cells in the supernatant after incubation with Strep- Tactin® magnetic Microbeads for 10, 20 or 30 min at cell concentrations of 1 x 10⁵ cells/ml, 1 x 10⁶ cells/ml or 1 x 10⁷ cells/ml.

[92] Figure 55 shows the elution profile of eluted CHO-ZSG5-P-II-E3 cells in the elution fraction 1 to 4 (E1 to E4). Cells were eluted after elution 1 in a new fraction immediately after resuspension in elution medium (no incubation) or were incubated between 10 and 40 min on a roller mixer before removing of supernatant and re-suspension in elution medium.

[93] Figure 56 shows mean fluorescence intensity of eluted GFP positive CHO-ZSG5-P-II-E3 cells in the elution fraction 1 to 4 (E1 to E4). Cells were eluted after elution 1 in a new fraction immediately after re-suspension in elution medium (no incubation) or were incubated between 10 and 40 min on a roller mixer before removing of supernatant and re-suspension in elution medium.

[94] Figure 57 shows the elution profile of eluted cells in elution fractions 1 to 6 (E1 to E6). The first five elutions where conducted in presence of 1 pM biotin. The sixth elution was conducted in the presence of 1000 pM biotin. For each elution the vial containing the cell magnetic microbead mixture was incubated on a magnet once before collection of supernatant that contained dissociated cells. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<20,000). GFP++ corresponds to cells with medium mean Fl (20,000 - 40,000). GFP+++ corresponds to cells with high mean Fl (>40,000).

[95] Figure 58 shows the elution profile of eluted cells in elution fractions 1 to 6 (E1 to E6). The first five elutions where conducted in presence of 1 pM biotin. The sixth elution was conducted in the presence of 1000 pM biotin. For each elution the vial containing the cell magnetic microbead mixture was incubated on a magnet 4 times with suspension steps after each incubation on the magnet before collection of the cells. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<20,000). GFP++ corresponds to cells with medium mean Fl (20,000 - 40,000). GFP+++ corresponds to cells with high mean Fl (>40,000). [96] Figure 59 shows the elution profile of eluted cells in elution fractions 1 to 6 (E1 to E6). The first five elutions where conducted in presence of 10 pM biotin. The sixth elution was conducted in the presence of 1000 pM biotin. For each elution the vial containing the cell magnetic microbead mixture was incubated on a magnet once before collection of supernatant that contained dissociated cells. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<20,000). GFP++ corresponds to cells with medium mean Fl (20,000 - 40,000). GFP+++ corresponds to cells with high mean Fl (>40,000).

[97] Figure 60 shows the elution profile of eluted cells in elution fractions 1 to 6 (E1 to E6). The first five elutions where conducted in presence of 10 pM biotin. The sixth elution was conducted in the presence of 1000 pM biotin. For each elution the vial containing the cell magnetic microbead mixture was incubated on a magnet 4 times with suspension steps after each incubation on the magnet before collection of the cells. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<20,000). GFP++ corresponds to cells with medium mean Fl (20,000 - 40,000). GFP+++ corresponds to cells with high mean Fl (>40,000).

[98] Figure 61 shows the elution profile of eluted cells in elution fractions 1 to 5 (E1 to E5). The five elutions where conducted in presence of 1000 pM biotin. For each elution the vial containing the cell magnetic microbead mixture was incubated on a magnet once before collection of supernatant that contained dissociated. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<20,000). GFP++ corresponds to cells with medium mean Fl (20,000 - 40,000). GFP+++ corresponds to cells with high mean Fl (>40,000).

[99] Figure 62 shows the elution profile of eluted cells in elution fractions 1 to 5 (E1 to E5). The five elutions where conducted in presence of 1000 pM biotin. For each elution the vial containing the cell magnetic microbead mixture was incubated on a magnet 4 times with suspension steps after each incubation on the magnet before collection of the cells. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<20,000). GFP++ corresponds to cells with medium mean Fl (20,000 - 40,000). GFP+++ corresponds to cells with high mean Fl (>40,000).

[100] Figure 63 shows the mean fluorescence intensity of cells in elution fractions 1 to 6 (E1 to E6) from the selection in Example 16. The elutions where conducted in presence of 1pM, 10 pM or 1000 pM biotin. For each elution the vial containing the cell magnetic microbead mixture was either incubated on a magnet 4 times with suspension steps after each incubation on the magnet (magnet +) or the mix was incubated only once on the magnet (magnet -) before collection of the cells.

[101] Figure 64 shows the elution profile of eluted of cells that were eluted from magnetic Microbeads using either a stepwise increased biotin concentration (gradient) or a constant high biotin concentration of 1000 pM. Cells that were eluted with a biotin gradient were eluted with 1 pM, 3 pM, 5 pM, 10 pM and 1000 pM biotin. The concentration of biotin was increased after every four elutions. For all elutions, cells were incubated 4 times on a magnetic separator followed by re-suspension of the cells after each incubation before supernatant was removed.

[102] Figure 65 shows the mean fluorescence intensity of cells that were eluted from magnetic Microbeads using either a stepwise increased biotin concentration (gradient) or a constant high biotin concentration of 1000 pM. Cells that were eluted with a biotin gradient were eluted with 1 pM, 3 pM, 5 pM, 10 pM and 1000 pM biotin. The concentration of biotin was increased after every four elutions. For all elutions, cells were incubated 4 times on a magnetic separator followed by re-suspension of the cells after each incubation before supernatant was removed.

[103] Figure 66 shows the elution profile dependent on the mean Fl of eluted of cells that were eluted from magnetic Microbeads using a stepwise increased biotin concentration (gradient). Cells that were eluted with a biotin gradient were eluted with 1 pM (E1-E4), 3 pM (E5-E8), 5 pM (E9-E12), 10 pM (E13-E16) and 1000 pM biotin (E17-E20). The concentration of biotin was increased after every four elutions. For all elutions, cells were incubated 4 times on a magnetic separator followed by re-suspension of the cells after each incubation before supernatant was removed. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<38,000). GFP++ corresponds to cells with medium mean Fl (38,000 - 70,000). GFP+++ corresponds to cells with high mean Fl (>70,000).

[104] Figure 67 shows the elution profile dependent on the mean Fl of eluted of cells that were eluted from magnetic Microbeads using a constant high biotin concentration of 1000 pM. Cells were incubated 4 times on a magnetic separator followed by re-suspension of the cells after each incubation before supernatant was removed. Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<38,000). GFP++ corresponds to cells with medium mean Fl (38,000 - 70,000). GFP+++ corresponds to cells with high mean Fl (>70,000).

[105] Figure 68 shows the yield of eluted cells after using the method of invention depending on different variants marker proteins. The TST-CD4tm marker was either fused to a eGFP (pZSG5, marker protein comprising SEQ ID NO 5, SEQ ID NO 10, SEQ ID NO 12, SEQ ID NO 2) or was not fused to eGFP. CD4tm variants without eGFP comprised the SEQ ID NO 12 (CD4tm) or SEQ ID NO 13 where the last four C-terminal arginine amino acids of CD4tm were replaced by alanine (CD4tm-A) or SEQ ID NO 14 with deletion of the last nine C-terminal amino acids from CD4tm (CD4tm-delta).

[106] Figure 69 shows the scatter plot of a CHO cell population prior to selection. Cell were transfected with a plasmid comprising the genetic elements according to Figure 12

[107] Figure 12with SEQ ID NO 13. The cells were stained with Strep-Tactin®XT conjugate DY-649 which binds to the TST presented on the cell surface. The scatter plot comprises all events from the flow cytometer analysis. Y-axis corresponds to 649 mean fluorescence intensity and X-axis to GFP mean fluorescence intensity. Mean fluorescence was measured in the PE channel instead of the FITC channel because the Fl of many events in the FITC channel was over the maximum detection level.

[108] Figure 70 shows the scatter plot of a CHO cell population in the first elution fraction after selection using the method of invention. The cells were stained with Strep-Tactin®XT conjugate DY-649 which binds to the TST presented on the cell surface. The scatter plot comprises all events from the flow cytometer analysis, including magnetic microbeads in the sample that are in the GFP negative (GFP-) and 649 negative (649-) gate. Due to the residual microbeads that are GFP negative, the ratio of GFP positive events is low. Y-axis corresponds to 649 mean fluorescence intensity and X-axis to GFP mean fluorescence intensity. Mean fluorescence was measured in the PE channel instead of the FITC channel because the Fl of many events in the FITC channel was over the maximum detection level.

[109] Figure 71 shows the scatter plot of a CHO cell population in the second elution fraction after selection using the method of invention. The cells were stained with Strep-Tactin®XT conjugate DY-649 which binds to the TST presented on the cell surface. The scatter plot comprises all events from the flow cytometer analysis, including magnetic microbeads in the sample that are in the GFP negative (GFP-) and 649 negative (649-) gate. Due to the residual microbeads that are GFP negative, the ratio of GFP positive events is low. Y-axis corresponds to 649 mean fluorescence intensity and X-axis to GFP mean fluorescence intensity. Mean fluorescence was measured in the PE channel instead of the FITC channel because the Fl of many events in the FITC channel was over the maximum detection level.

[110] Figure 72 shows the scatter plot of a CHO cell population in the third elution fraction after selection using the method of invention. The cells were stained with Strep-Tactin®XT conjugate DY-649 which binds to the TST presented on the cell surface. The scatter plot comprises all events from the flow cytometer analysis, including magnetic microbeads in the sample that are in the GFP negative (GFP-) and 649 negative (649-) gate. Due to the residual microbeads that are GFP negative, the ratio of GFP positive events is low. Y-axis corresponds to 649 mean fluorescence intensity and X-axis to GFP mean fluorescence intensity. Mean fluorescence was measured in the PE channel instead of the FITC channel because the Fl of many events in the FITC channel was over the maximum detection level.

[111] Figure 73 shows the mean GFP Fl and mean DY-649 Fl of GFP positive cells before and after selection using the method of invention. Cells were transfected with a plasmid comprising the genetic elements according to Figure 12 with SEQ ID NO 13. The cells were stained with Strep-Tactin®XT conjugate DY-649 which binds to the TST presented on the cell surface.

[112] Figure 74 shows the median GFP Fl of GFP positive cells before and after selection using the method of invention. Cells were transfected with a plasmid comprising the genetic elements according to Figure 12 with SEQ ID NO 13.

DETAILED DESCRIPTION OF THE INVENTION

[113] The present invention is described in detail in the following and will also be further illustrated by the appended examples and figures.

[114] The present invention aims at isolating cells having a high (expressing a protein of interest at a level higher than the mean of the population of cells) and stable expression of a protein of interest from a population of cells or, in other words, fractionating a population of cells into fractions having different capacity for expression of a protein of interest. The method of invention can be used to select a cell population showing an expression of recombinant proteins that exceeds the expression of the original unselected population. Furthermore, the use of the method of invention increases the fraction of cells expressing the recombinant proteins over cells that do not express the recombinant proteins. A consecutive selection using the method of invention can furthermore lead to a cell population, which stably expresses the recombinant proteins. The methods known so far in prior art provide no incentive or guidance how such a method could look like.

[115] E g., WO 2018/134691 A2 discloses a cell surface conjugate, which might include a streptavidin, streptavidin analog or a streptavidin mutein. However, this patent application fails to disclose a method for fractionating cells according to the present invention. In contrast, the disclosure of WO 2018/134691 A2 is limited to a single step isolation of cells expressing the cell surface conjugate and fails to disclose a method, in which a population of cells is fractionated multiple times to fractionate a population of cells into different fractions. Matheson et al, supra, describe a similar one-step method, thus fails to disclose a method of fractionating a cells.

[116] The present inventor has surprisingly found that the method of the present invention can be used to fractionate a population of cells into fractions, wherein generally with each subsequent elution step the fraction of highly and stably producing cells increases. In order to fractionate or isolate high-producing cells from the population of cells, it may however already be sufficient, to carry out only the first elution step. Surprisingly, the present inventor further found that it is not necessary to increase the concentration of the eluant (competitor) with each elution step but that with each elution step the essentially the same concentration can be used. However, increasing the concentration of the competitor with each elution step is of course possible. Since the cells are fractionated based to the amount of a marker protein on the cell surface, whose expression is correlated to the expression of the protein of interest, the cells of a population of cells are also fractionated according to the expression rate of the protein of interest.

[117] The population of cells may arise from a genetic modification of a population of cells, e.g. by transfection or transduction. Such a population of cells typically comprises a variety of cells with many having little or no expression of the protein of interest, many having an intermediate expression of the protein of interest while only a few have a high expression of the protein of interest. The present invention thus provides a method to isolate these few high expressing cells as shown in the Examples. The inventor could further show that the cells isolated with the method of the invention show a stable and high expression over an extended period of time - in contrast to selecting cells with prior art techniques based on e.g., antibiotic resistance. Furthermore, the present invention avoids the production of a significant amount of resistance proteins because isolation of cells can depend on a much lower amount of surface proteins. Thereby, the metabolic costs for producing genes that enable resistance, e.g. to an antibiotic, can be avoided.

[118] The basic principle of the invention is described briefly in Figs. 1 to 4. In short, a marker protein is co-expressed with the protein of interest. Thus, the expression of the marker protein correlates with the expression of the protein of interest. The marker protein is a cell surface marker comprising a binding partner B, which can reversible bind to a ligand L present on a solid phase. In some embodiments, the marker protein does not comprise the binding partner B but the binding partner B is included in a receptor molecule binding reagent, which in turn specifically binds to the marker protein (receptor molecule) and comprises the binding partner B. After contacting the cells with the solid phase, unbound cells are removed, e.g. by washing the solid phase. Then, cells are eluted by adding a competitor, which can disrupt the reversibly bond between the binding partner B and the ligand L, to a first concentration. Thus, a first fraction is eluted from the solid phase, which can be removed. In some embodiments, a first elution is sufficient to isolate high and stable-producing cells. Optionally, in a second step, the competitor is added to a second concentration and the eluted cells are removed again, thereby eluting a second fraction of cells. This can be repeated multiple times. As surprisingly shown by the inventor, it is not necessary (but possible) to increase the concentration of the competitor with each repetition. Increasing the concentration with each repetition may be useful to fractionate cells according to the expression of a receptor molecule (or marker protein) on their cell surface, in particular in combination with the embodiment, in which the marker protein does not comprise the binding partner B but the binding partner B is included in a receptor molecule binding reagent, which in turn specifically binds to the marker protein (receptor molecule) and comprises the binding partner B. The concentrations may also essentially be the same. In the 2^nd, 3^rd, 4^th or further elutions, the fraction of high expressing cells can be even higher.

[119] A further exemplary embodiment of the method of the invention is described in the following: A population of cells comprising cells expressing the recombinant proteins (protein of interest (POI) and marker protein (MP)) is contacted with a solid phase which is linked to a streptavidin mutein (e.g. Strep-Tactin®) or a multimerized streptavidin mutein (ligand L). The solid phase may be ferromagnetic beads, agarose beads, a membrane or another solid surface. A streptavidin mutein (ligand L) is bound or linked to the solid phase by using an appropriate method. The cell-solid phase-mixture is incubated. Thus, cells that express the recombinant proteins (POI and MP) are immobilized on the solid phase via the binding of the streptavidin- binding peptide of the MP (binding partner B) to the streptavidin mutein (ligand L) bound to the solid phase. Cells that exhibit no MP expression and thus no POI expression are removed from the mixture by removal of the liquid phase (wash procedure). Consecutive wash steps with a washing buffer can be used to remove cells not expressing the POI and MP. After the removal of non-producing cells, cells expressing the POI and the MP can be selected by removing the cells from the solid phase (elution). The addition of biotin (competitor) to the wash buffer blocks the biotin/streptavidin-binding peptide binding site of the streptavidin mutein once the streptavidin-binding peptide dissociates from the binding site. This finally leads to a complete dissociation of the cells from the matrix. Cells exhibiting a low POI and thus a low MP expression have less streptavidin-binding peptide bounds to the matrix than cells with a high POI and MP expression. Therefore, low expressing cells dissociate earlier in the elution. Due to this effect low expressing cells represents the majority in the liquid phase of the first elution. By consecutive elution steps the mean POI and MP expression of the eluted populations in each elution fraction are increased with highest expressions in the latest elution fractions. Thus, the cells are fractionated into fractions based on the amount of a marker protein of the cell surface. [120] Accordingly, the present invention relates to a method for fractionating cells of a population of cells based on the amount of a marker protein on the cell surface, the method comprising:

(f) separating the eluted fraction of the population of cells obtained in step (e).

[121] The present invention also relates to a method for enriching or isolating cells comprised of a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells, the method comprising:

(c) adding the competitor to a first concentration, thereby eluting a fraction of the population of cells from the solid phase; (d) separating the eluted fraction of the population of cells obtained in step (c);

(h) optionally separating the eluted fraction of the population of cells eluting in step

(g);

[122] The method of the invention is not limited to one or two elution steps but may include 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more consecutive elution steps. These further elutions might allow fractionation of cells having an even higher expression of the marker protein on the cell surface. Accordingly, the method of the present invention may further comprise

[123] Steps (g) and (h) may be repeated once. Steps (g) and (h) may be repeated twice. Steps

(g) and (h) may be repeated 3 times. Steps (g) and (h) may be repeated 4 times. Steps (g) and

(h) may be repeated 5 times. Steps (g) and (h) may be repeated 6 times. Steps (g) and (h) may be repeated 7 times. Steps (g) and (h) may be repeated 38 times. Steps (g) and (h) may be repeated 9 times. Steps (g) and (h) may be repeated 10 times. Steps (g) and (h) may be repeated 11 or more times.

[124] The method of fractionating cells of a population of cells based on the amount of a marker protein on the cell surface of the present invention may also be seen as a method for enriching cells comprised in a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells. In other words, the method of the present invention can be used to isolate cells of a cell population, which are high producers of a protein of interest, since the amount of the marker protein on the cell surface of a cell of the population of cells is correlated to the level of expression of the protein of interest. Such cells expressing a protein of interest at a level higher than the mean of the population of cells preferably are eluted with the second or third fraction. [125] The present invention generally allows (at least) two different modes of fractionating the of the population the cells: (i) the first, the second and each of the first concentration of the competitor are essentially the same or (ii) the second concentration is higher than the first concentration, wherein the further concentration is higher than the second concentration and each subsequent further concentration is higher than the previous further concentration. Option (i) is of particular interest, since the Inventor surprisingly found that it is not essential to use increasing concentrations of the competitor to elute or isolate cells expressing the protein of interest at a high level.

[126] “Essentially the same concentration” as used herein may relate to a concentration (of the competitor) that results in a comparable amount of eluted cells in comparison to the first concentration or a reference concentration. Comparable in this context relates to a concentration that results in the elution of ± 15%, more preferably ± 10%, more preferably ± 5%, more preferably ± 3%, more preferably ± 2%, or more preferably ± 1 % of the cells using the first concentration or a reference concentration. “Essentially the same concentration” as used herein may relate to a concentration (of the competitor), which is ± 200%, more preferably ± 100%, more preferably ± 50%, more preferably ± 20%, more preferably ± 15%, more preferably ± 10%, more preferably ± 5%, more preferably ± 3%, more preferably ± 2%, or more preferably ±1 % of the first or a reference concentration. “Reference concentration” describes any concentration, to which the concentration should be similar (same concentration). Accordingly, the first concentration, the second concentration and the further concentration(s) may essentially be the same.

[127] Exceptionally stable cells can be obtained by a consecutive use of the method of invention (see e.g., Example 10). Transient expressing cells lack an expression after a few days while stable expressing cells maintain POI and MP expression. With each selection/elution/fractionation round (i.e., steps ((c) & (d), steps (e) & (f) or steps (g) & (h)) using the method of invention non-expressing cells are removed from the culture leaving only stable expressing cells after a few days. Between these selection/elution/fractionation rounds, i.e. between each of the steps, i.e. steps ((c) & (d), steps (e) & (f) or steps (g) & (h), there might be a defined period of time, such as about 1 day, about 2 days, about 3 days, about four days, about five days, or more than about 5 days. The defined period of time may also be between about 1 and about 30 days, about 2 and about 25 days, about 3 and about 10 days or between about 4 and 8 days.

[128] Alternatively, cells can be selected using a conventional method (e.g. antibiotic selection, glutamine synthetase method, DHFR, FACS sorting) to establish a stable expressing population, preferably prior to carrying out the method of the invention. Here, the method of invention may be carried out after a stable population is generated to select high expressing cells which increase the productivity of the cultured cells compared to the population of cells that are not selected by the method of invention.

[129] “Fractionating” as used herein relates to dividing a population of cells into different populations or fractions. These fractions distinguish themselves in their expression of the marker protein and thus their expression of the protein of interest.

[130] The term “cell” as used herein encompasses all biological entities/vesicles in which a membrane, which can also be a lipid bilayer, separates the interior from the outside environment (ambience). Virtually any cell that contains at least one marker protein can be separated or fractionated from other components included in a sample. In order to achieve an avidity effect, as discussed below, for a method as described herein, the marker protein is typically present in two or more copies on the surface of the (target) cell.

[131] In some embodiments the cell may be a prokaryotic cell, such as a bacterial cell. The cell may in some embodiments be an archaeon. The cell may in some embodiments be a virus or an organelle such as a mitochondrion, a chloroplast, a microsome, a lysosome, a Golgi apparatus or a nucleus. In some embodiments the cell may be a eukaryotic cell, such as a plant cell, a fungal cell, a yeast cell, a protozoon or an animal cell. The cell includes in some embodiments a cell nucleus. In some embodiments the cell is a mammalian cell, including but not limited to cells obtained from a human, mouse, rabbit, guinea pig, squirrel, hamster, cat, dog, lemur, goat, pig, horse, rhesus monkey, macaque, or a chimpanzee. Examples of a mammalian cell include, but are not limited to, a blood cell or a tissue cell, e.g. a hepatocyte or a stem cell, e.g. CD34-positive peripheral stem cells or Nanog or Oct-4 expressing stem cells derived from a suitable source. A blood cell may for instance be a leukocyte or an erythrocyte. A leukocyte may for example be a neutrophil, an eosinophil, a basophil, a monocyte, a lymphocyte, a macrophage or a dendritic cell. A respective lymphocyte may for example be a T cell - including a CM -specific CD8+ T-lymphocyte, a cytotoxic T-cell a, memory T-cell (an illustrative example of memory T-cells are CD62L+CD8+ specific central memory T-cells) or a regulatory T-cell (an illustrative example of Treg are CD4+CD25+CD45RA+ Treg cells), a T- helper cell (for example, a CD4+ T-helper cell), a B cell or a natural killer cell, to mention only a few illustrative examples. The cell may be selected from the group consisting of CHO cells, CHO-S, ExpiCHO, Freestyle CHO-S, CHO-GS, CHO-K1 , CHO-DXB11, CHO-DG44, CHO duk, CHO-DP12, CHOZN, GS-CHOK1SV, HEK-293 cells, HEK-293T cells, HEK-293-6E, HEK-293- EBNA, HEK 293SF-3F6, 293 c18, Expi293, 293-F, insect cells, SF9, ExpiSf9, Hi-5, Sf21 , human amniocytes and CAP®. [132] The cell is typically a cell or, as mentioned above, any other biological entity, in which a membrane, which may in some embodiments be a lipid bilayer, separates the interior from the ambience. The cell is further characterized by having a marker protein on the surface. Such a cell can be purified, enriched or fractionated by the methods described herein, under subsequent removal of any used purification reagent, such as a competitor and/or a solid phase. A respective method or use offers - beyond the advantage that, if the target is a cell or an organelle, the physiological status is not altered - the regulatory advantage that the purification reagents are not administered to a subject such as a patient during the use of such purified biological entities as medicaments.

[133] The cell may, for instance, be a cell of a tissue, such as an organ or a portion thereof. Examples of a respective organ include, without being limited thereto, adrenal tissue, bone, blood, bladder, brain, cartilage, colon, eye, heart, kidney, liver, lung, muscle, nerve, ovary, pancreas, prostate, skin, small intestine, spleen, stomach, testicular, thymus, tumor, vascular or uterus tissue, or connective tissue. In some embodiments the cell is a stem cell, a lymphocyte or a cancer cell.

[134] A sample from which the cell is to be isolated or fractionated may be of any origin. It may for instance, but not limited to, be derived from humans, animals, plants, bacteria, fungi, or protozoae. Accordingly, any of the following samples selected from, but not limited to, the group consisting of a soil sample, an air sample, an environmental sample, a cell culture sample, a bone marrow sample, a food sample, a blood sample, a serum sample, a plasma sample, an urine sample, a stool sample, a semen sample, a lymphatic fluid sample, a cerebrospinal fluid sample, a nasopharyngeal wash sample, a sputum sample, a mouth swab sample, a throat swab sample, a nasal swab sample, a bronchoalveolar lavage sample, a bronchial secretion sample, a milk sample, an amniotic fluid sample, a biopsy sample, a cancer sample, a tumour sample, a tissue sample, a cell sample, a cell culture sample, a cell lysate sample, a virus culture sample, a nail sample, a hair sample, a skin sample, a forensic sample, an infection sample, a nosocomial infection sample, a space sample or any combination thereof. Where desired, a respective sample may have been preprocessed to any degree. As an illustrative example, a tissue sample may have been digested, homogenized or centrifuged prior to being used in a method described herein. In another illustrative example, a sample of a body fluid such as blood might be obtained by standard isolation of blood cells. If an isolation method described here is used in basic research, the sample might be cells of in vitro cell culture experiments. The sample will typically have been prepared in form of a fluid, such as a solution or dispersion. [135] In case the cell is a prokaryotic cell, it preferably is of a species selected from the group consisting of Lactobacillus spp., Yersinia spp., Escherichia spp., Klebsiella spp., Bordetella spp., Neisseria spp., Aeromonas spp., Franciesella spp., Corynebacterium spp., Citrobacter spp., Chlamydia spp., Hemophilus spp., Brucella spp., Mycobacterium spp., Legionella spp., Rhodococcus spp., Pseudomonas spp., Helicobacter spp., Salmonella spp., Vibrio spp., Bacillus spp., Leishmania spp. and Erysipelothrix spp., Shigella spp., Listeria spp., Rickettsia spp., Acetoanaerobium spp., Aerococcaceae spp., Carnobacteriaceae spp., Enterococcace spp., Leuconostocacease spp., Streptococcaceae spp., and bacteria with GRAS status, more preferably E. coli.

[136] After contacting the population of cells of step (a) of the method of the invention, the solid phase may subsequently be washed with a mobile phase, such as an aqueous medium, e.g. a buffer, in order to remove any matter, including, e.g., cells not expressing the marker protein that has not been immobilized on the chromatography matrix. Such washing may be carried out on any solid phase employed in the context of a method or use described herein. Such washing may be carried out after each step of the method of the invention. Accordingly, the washing may be carried out after step (a) of the invention. Accordingly, the washing may be carried out after step (b) of the invention. Accordingly, the washing may be carried out after step (c) of the invention. Accordingly, the washing may be carried out after step (d) of the invention. Accordingly, the washing may be carried out after step (e) of the invention. Accordingly, the washing may be carried out after step (f) of the invention. Accordingly, the washing may be carried out after step (g) of the invention.

Receptor molecule binding reagents

[137] As described in Examples 2 and 4, it is furthermore not essential that the binding partner B is comprised in the marker protein within the methods of the invention. Alternatively, the binding partner B may be provided with a receptor binding molecule, which in turn specifically binds to the marker protein or any other receptor molecule on the cell surface.

[138] Accordingly, the present invention further relates to a method for fractionating cells of a population of cells based on the amount of a receptor molecule on the cell surface, the method comprising:

[139] Additionally, the method of the invention relates to a method for enriching cells comprised in a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells, the method comprising:

(d) separating the eluted fraction of the population of cells obtained in step (c); (e) optionally adding the competitor to a second concentration, thereby eluting a fraction of the population of cells from the solid phase;

[140] The receptor molecule may be a marker protein as described herein. Accordingly, the population of cells may comprise cells comprising a nucleic acid comprising (i) an expression cassette encoding for a marker protein comprising and (ii) a protein of interest.

[141] The receptor molecule that is located on the cell surface, including located on an accessible surface of a biological entity, may be any molecule present on the cell surface during a separation process in a method according to the invention. The receptor molecule is a molecule against which a receptor molecule binding reagent is directed. In some embodiments the receptor is a peptide or a protein, such as a membrane receptor protein. In some embodiments the receptor is a lipid or a polysaccharide. A receptor molecule that is a protein may be a peripheral membrane protein or an integral membrane protein. It may in some embodiments have one or more domains that span the membrane. As a few illustrative examples, the membrane protein may be a CD molecule (cluster of differentiation) such as CD3, CD4, CD8, CD247 (T cell markers), CD8, CD62L, CD45RA (marker for memory T cells), CD4, CD25, CD45RA (markers for regulatory T cells), CD56 (marker for natural killer cells), CD19 (B cell marker) and CD34, Oct-4, Nanog (stem cell markers), to name only a few illustrative example. Accordingly, the target cell may, for instance, a population or subpopulation of blood cells; e.g. lymphocytes such as T cells, T-helper cells, for example, CD4⁺ T-helper cells, B cells or natural killer cells; monocytes; or stem cells, e.g. CD34-positive peripheral stem cells or Nanog or Oct-4 expressing stem cells. Most T cells that have CD8 on their surface are cytotoxic T cells. The target cell may thus be CD8⁺ a cytotoxic T-cell. The receptor may also be a marker for a tumour cell. The membrane protein may also be a G-protein coupled receptor, such as an odorant receptors, a rhodopsin receptor, a rhodopsin pheromone receptor, a peptide hormone receptor, a taste receptor, a GABA receptor, an opiate receptor, a serotonin receptor, a Ca²⁺ receptor, melanopsin, a neurotransmitter receptor, a receptor kinase such as a serin/threonin kinase, a tyrosine kinase, a porin/channel such as a chloride channel, a potassium channel, or a cell adhesion receptor such as metallo protease, an integrin or a cadherin. [142] In some embodiments the receptor molecule binding reagent has a single (monovalent) binding site C capable of specifically binding to the receptor molecule. Examples of such monovalent receptor molecule binding reagents are soluble MHC I peptides (which are being complexed with an cell specific antigen presenting peptides (such MHC molecules are, for example, described in International Patent Application, WO 02/054065 or, Schmidt, J. et al., J. Biol. Chem. [2011] 286, 48, 41723-41735 and are commercially available from IBA GmbH orTC Metrix S.A., for example), mononovalent antibody fragments such as, for example, Fab fragments, Fv fragments or single chain Fv fragments (scFv) or a monovalent artificial binding molecule (proteinaceous or other) such as a mutein based on a polypeptide of the lipocalin family (also known as “Anticalin®). Alternatively, the receptor molecule binding reagent may also have two or more binding sites C. Examples of such receptor molecule binding reagents are intact (bivalent) antibody molecules or an antibody fragment in which both binding sites are retained such as an F(ab')₂ fragment. The receptor molecule binding reagent may be a multivalent molecule such as a pentameric IgE molecule.

[143] In the method of the invention the one or more binding sites C of the receptor molecule binding reagent, which specifically binds to the receptor molecule, may for instance be an antibody, a fragment thereof and a proteinaceous binding molecule with antibody-like functions. Examples of (recombinant) antibody fragments are Fab fragments, Fv fragments, single-chain Fv fragments (scFv), a divalent antibody fragment such as an (Fab)2'-fragment, diabodies, triabodies (lliades, R, et al., FEBS Lett (1997) 409, 437-441), decabodies (Stone, E., et al., Journal of Immunological Methods (2007) 318, 88-94) and other domain antibodies (Holt, L.J., et al., Trends Biotechnol. (2003), 21, 11, 484-490). In some embodiments one or more binding sites of the receptor molecule binding reagent may be a bivalent proteinaceous artificial binding molecule such as a dimeric lipocalin mutein that is also known as "duocalin". In some embodiments the receptor binding reagent may have a single second binding site, i.e. , it may be monovalent. Examples of monovalent receptor binding reagents include, but are not limited to, a monovalent antibody fragment, a proteinaceous binding molecule with antibody-like binding properties or an MHC molecule. Examples of monovalent antibody fragments include, but are not limited to a Fab fragment, a Fv fragment, and a single-chain Fv fragment (scFv), including a divalent single-chain Fv fragment.

[144] As mentioned above, an example of a proteinaceous binding molecule with antibody-like functions is a mutein based on a polypeptide of the lipocalin family (see for example, WO 03/029462, Beste et al., Proc. Natl. Acad. Sci. U.S.A. (1999) 96, 1898-1903). Lipocalins, such as the bilin binding protein, the human neutrophil gelatinase-associated lipocalin, human Apolipoprotein D or human tear lipocalin possess natural ligand-binding sites that can be modified so that they bind a given target. Further examples of a proteinaceous binding molecule with antibody-like binding properties that can be used as a receptor binding reagent that specifically binds to the receptor molecule include, but are not limited to, the so-called glubodies (see e.g. international patent application WO 96/23879), proteins based on the ankyrin scaffold (Mosavi, L.K., et al., Protein Science (2004) 13, 6, 1435-1448) or crystalline scaffold (e.g. international patent application WO 01/04144) the proteins described in Skerra, J. Mol. Recognit. (2000) 13, 167-187, AdNectins, tetranectins and avimers. Avimers, including multivalent avimer proteins evolved by exon shuffling of a family of human receptor domains, contain so called A-domains that occur as strings of multiple domains in several cell surface receptors (Silverman, J., et al., Nature Biotechnology (2005) 23, 1556-1561). Adnectins, derived from a domain of human fibronectin, contain three loops that can be engineered for immunoglobulin-like binding to targets (Gill, D.S. & Damle, N.K., Current Opinion in Biotechnology (2006) 17, 653-658). Tetranectins, derived from the respective human homotrimeric protein, likewise contain loop regions in a C-type lectin domain that can be engineered for desired binding (ibid.).

[145] Yet further examples of suitable proteinaceous binding molecules are an EGF-like domain, a Kringle-domain, a fibronectin type I domain, a fibronectin type II domain, a fibronectin type III domain, a PAN domain, a G1a domain, a SRCR domain, a Kunitz/Bovine pancreatic trypsin Inhibitor domain, tendamistat, a Kazal-type serine protease inhibitor domain, a Trefoil (P- type) domain, a von Willebrand factor type C domain, an Anaphylatoxin-like domain, a CUB domain, a thyroglobulin type I repeat, LDL-receptor class A domain, a Sushi domain, a Link domain, a Thrombospondin type I domain, an immunoglobulin domain or a an immunoglobulin- like domain (for example, domain antibodies or camel heavy chain antibodies), a C-type lectin domain, a MAM domain, a von Willebrand factor type A domain, a Somatomedin B domain, a WAP-type four disulfide core domain, a F5/8 type C domain, a Hemopexin domain, an SH2 domain, an SH3 domain, a Laminin-type EGF-like domain, a C2 domain, "Kappabodies" (cf. III. et al., Protein Eng (1997) 10, 949-57, a so called "minibody" (Martin et al., EMBO J (1994) 13, 5303-5309), a diabody (cf. Holliger et al., PNAS USA (1993)90, 6444-6448), a so called "Janusis" (cf. Traunecker et al., EMBO J (1991) 10, 3655-3659, or Traunecker et al., Int J Cancer (1992) Suppl 7, 51-52), a nanobody, a microbody, an affilin, an affibody, a knottin, ubiquitin, a zinc-finger protein, an autofluorescent protein or a leucine-rich repeat protein. An example of a nucleic acid molecule with antibody-like functions is an aptamer. An aptamer folds into a defined three-dimensional motif and shows high affinity for a given target structure.

Marker protein

[146] The method of the present invention is based on the interaction of the ligand L comprised in the solid phase with the binding partner B comprised in the marker protein or the receptor molecule binding reagent. The term “marker protein” as used herein relates to a protein, which is present on the cell surface of the cell and preferably comprises a ligand binding partner B. Since the marker protein and the protein of interest preferably are expressed from the same nucleic acid, the amount of the marker protein on the cell surface preferably is correlated to the expression rate of the protein of interest. In some embodiments, the marker protein and the protein of interest are identical.

[147] The mode how the marker protein is anchored to the cell surface is not particularly limited. The modes include, but are not limited to, one or more transmembrane domain(s) or a membrane anchor that anchor the marker protein to the cell surface.

[148] Accordingly, the marker protein may be a transmembrane protein or a fragment thereof, wherein the binding site B is comprised in the extracellular domain. In some embodiments, the marker protein contains a transmembrane domain. In some aspects, the marker protein is a type I, type II, type III or type IV membrane protein. In some aspects, type I proteins have a single transmembrane stretch of hydrophobic residues, with the portion of the polypeptide on the amino (N)-terminal side of the transmembrane domain exposed on the exterior side of the membrane and the carboxy (C)-terminal portion exposed on the cytoplasmic side. In some aspects, type I membrane proteins are subdivided into types la (with cleavable signal sequences) and lb (without cleavable signal sequence). In some aspects, type II membrane proteins span the membrane only once, but they have their amino terminus on the cytoplasmic side of the cell and the carboxy terminus on the exterior. In some aspects, type III membrane proteins have multiple transmembrane domains in a single polypeptide chain and can be subdivided into type Ilia proteins (with cleavable signal sequences) and type lllb (with amino termini exposed on the exterior surface of the membrane, but without cleavable signal sequences). In some aspects, type IV proteins have multiple homologous domains which make up an assembly that spans the membrane multiple times, with the domains present on a single polypeptide chain or one or more different polypeptide chains.

[149] The marker protein may comprise a transmembrane domain of a protein selected from the group consisting of EpCAM, VEGFR, integrin, optionally integrins avp3, a4, alip3, a4p7, a5pi, avp3 or an, a member of the TNF receptor superfamily, optionally TRAIL-RI or TRAIL-R2, a member of the epidermal growth factor receptor family, PDGF Receptor, interferon receptor, folate receptor, GPNMB, ICAM-1, HLA-DR, CEA, CA-125, MUCI, TAG-72, IL-6 receptor, 5T4, GD2, GD3, prostate-specific membrane antigen (PSMA) or a clusters of differentiation cell surface molecule, optionally CD2, CD3, CD4, CD5, CD11 , CDIIa/LFA-1, CD15, CD18/ITGB2, CD19, CD20, CD22, CD23/lgE Receptor, CD25, CD28, CD30, CD33, CD38, CD40, CD41 , CD44, CD51, CD52, CD62L, CD74, CD80, CD125, CD147/basigin, CD152/CTLA-4, CD154/CD40L, CD195/CCR5 and CD319/SLAMF7, low-affinity nerve growth factor receptor (LNGFR). Preferably, the marker protein comprises a transmembrane domain of CD4, e.g. as depicted in SEQ ID NO: 12.

[150] The transmembrane domain, e.g. that of CD4, may be fused to an extracellular domain, e.g. as shown in SEQ ID NO: 13, which is however not necessary. It is sufficient, to fuse the binding partner B, e.g., a Twin-StrepTag® (TST) as described herein, to the transmembrane domain, preferably using a linker. Thus, the CD4 transmembrane domain may be fused/coupled to the binding partner B, e.g., the TST, with a flexible linker, e.g., the linker depicted in SEQ ID NO: 10. The CD4 transmembrane domain may be fused/coupled to the binding partner B, e.g., the TST, with a rigid linker, e.g., the linker depicted in SEQ ID NO: 11. However, further suitable linker are known in the art and are e.g., described in Chen et al. (2013), Adv Drug Deliv Rev., 65(10):1357-1369, which is hereby incorporated by reference in total, while the sequences of linkers shown in Table 2 are incorporated explicitly by reference.

[151] The marker protein may comprise or consist of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25), or a fragment or analog thereof having a sequence identity of 60% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25). The marker protein may comprise or consist of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23- 25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25) or a fragment or analog thereof having a sequence identity of 70% or higher compared to any one of SEQ ID NO: -8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25). The marker protein may comprise or consist of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25), or a fragment or analog thereof having a sequence identity of 80% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23- 25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25). The marker protein may comprise or consist of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25), or a fragment or analog thereof having a sequence identity of 90% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25). The marker protein may comprise or consist of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25), or a fragment or analog thereof having a sequence identity of 95% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25). The marker protein may comprise or consist of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25), or a fragment or analog thereof having a sequence identity of 97.5% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25). [152] The marker protein may comprise or consist of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25), or a fragment or analog thereof having a sequence identity of 99% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28 (or any one of SEQ ID NO: 2, 6, 8, 12-17, 21 or 25).

[153] The marker protein may comprise or consist of SEQ ID NO: 1, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 1. Preferably, the marker protein comprises SEQ ID NO: 1. Preferably, the marker protein consists of SEQ ID NO: 1.

[154] The marker protein may comprise or consist of SEQ ID NO: 2, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 2. Preferably, the marker protein comprises SEQ ID NO: 2. Preferably, the marker protein consists of SEQ ID NO: 2.

[155] The marker protein may comprise or consist of SEQ ID NO: 3, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 3. Preferably, the marker protein comprises SEQ ID NO: 3. Preferably, the marker protein consists of SEQ ID NO: 3.

[156] The marker protein may comprise or consist of SEQ ID NO: 4, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 4. Preferably, the marker protein comprises SEQ ID NO: 4. Preferably, the marker protein consists of SEQ ID NO: 4.

[157] The marker protein may comprise or consist of SEQ ID NO: 5, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 5. Preferably, the marker protein comprises SEQ ID NO: 5. Preferably, the marker protein consists of SEQ ID NO: 5.

[158] The marker protein may comprise or consist of SEQ ID NO: 6, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 6. Preferably, the marker protein comprises SEQ ID NO: 6. Preferably, the marker protein consists of SEQ ID NO: 6.

[159] The marker protein may comprise or consist of SEQ ID NO: 7, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 7. Preferably, the marker protein comprises SEQ ID NO: 7. Preferably, the marker protein consists of SEQ ID NO: 7.

[160] The marker protein may comprise or consist of SEQ ID NO: 8, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 8. Preferably, the marker protein comprises SEQ ID NO: 8. Preferably, the marker protein consists of SEQ ID NO: 8.

[161] The marker protein may comprise or consist of SEQ ID NO: 9, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 9. Preferably, the marker protein comprises SEQ ID NO: 9. Preferably, the marker protein consists of SEQ ID NO: 9.

[162] The marker protein may comprise or consist of SEQ ID NO: 10, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 10. Preferably, the marker protein comprises SEQ ID NO: 10. Preferably, the marker protein consists of SEQ ID NO: 10.

[163] The marker protein may comprise or consist of SEQ ID NO: 11 , or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 11. Preferably, the marker protein comprises SEQ ID NO: 11. Preferably, the marker protein consists of SEQ ID NO: 11.

[164] The marker protein may comprise or consist of SEQ ID NO: 12, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 12. Preferably, the marker protein comprises SEQ ID NO: 12. Preferably, the marker protein consists of SEQ ID NO: 12. [165] The marker protein may comprise or consist of SEQ ID NO: 13, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 13. Preferably, the marker protein comprises SEQ ID NO: 13. Preferably, the marker protein consists of SEQ ID NO: 13.

[166] The marker protein may comprise or consist of SEQ ID NO: 14, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 14. Preferably, the marker protein comprises SEQ ID NO: 14. Preferably, the marker protein consists of SEQ ID NO: 14.

[167] The marker protein may comprise or consist of SEQ ID NO: 15, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 15. Preferably, the marker protein comprises SEQ ID NO: 15. Preferably, the marker protein consists of SEQ ID NO: 15.

[168] The marker protein may comprise or consist of SEQ ID NO: 16, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 16. Preferably, the marker protein comprises SEQ ID NO: 16. Preferably, the marker protein consists of SEQ ID NO: 16.

[169] The marker protein may comprise or consist of SEQ ID NO: 17, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 17. Preferably, the marker protein comprises SEQ ID NO: 17. Preferably, the marker protein consists of SEQ ID NO: 17.

[170] The marker protein may comprise or consist of SEQ ID NO: 18, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 18. Preferably, the marker protein comprises SEQ ID NO: 18. Preferably, the marker protein consists of SEQ ID NO: 18.

[171] The marker protein may comprise or consist of SEQ ID NO: 19, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 19. Preferably, the marker protein comprises SEQ ID NO: 19. Preferably, the marker protein consists of SEQ ID NO: 19.

[172] The marker protein may comprise or consist of SEQ ID NO: 20, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 20. Preferably, the marker protein comprises SEQ ID NO: 20. Preferably, the marker protein consists of SEQ ID NO: 20.

[173] The marker protein may comprise or consist of SEQ ID NO: 21 , or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 21. Preferably, the marker protein comprises SEQ ID NO: 21. Preferably, the marker protein consists of SEQ ID NO: 21.

[174] The marker protein may comprise or consist of SEQ ID NO: 22, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 22. Preferably, the marker protein comprises SEQ ID NO: 22. Preferably, the marker protein consists of SEQ ID NO: 22.

[175] The marker protein may comprise or consist of SEQ ID NO: 23, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 23. Preferably, the marker protein comprises SEQ ID NO: 23. Preferably, the marker protein consists of SEQ ID NO: 23.

[176] The marker protein may comprise or consist of SEQ ID NO: 24, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 24. Preferably, the marker protein comprises SEQ ID NO: 24. Preferably, the marker protein consists of SEQ ID NO: 24.

[177] The marker protein may comprise or consist of SEQ ID NO: 25, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 25. Preferably, the marker protein comprises SEQ ID NO: 25. Preferably, the marker protein consists of SEQ ID NO: 25. [178] The marker protein may comprise or consist of SEQ ID NO: 26, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 26. Preferably, the marker protein comprises SEQ ID NO: 26. Preferably, the marker protein consists of SEQ ID NO: 26.

[179] The marker protein may comprise or consist of SEQ ID NO: 27, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 27. Preferably, the marker protein comprises SEQ ID NO: 27. Preferably, the marker protein consists of SEQ ID NO: 27.

[180] The marker protein may comprise or consist of SEQ ID NO: 29, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 29. Preferably, the marker protein comprises SEQ ID NO: 29. Preferably, the marker protein consists of SEQ ID NO: 29.

[181] The marker protein may comprise or consist of SEQ ID NO: 30, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 30. Preferably, the marker protein comprises SEQ ID NO: 30. Preferably, the marker protein consists of SEQ ID NO: 30.

[182] The marker protein may comprise or consist of SEQ ID NO: 31 , or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 31. Preferably, the marker protein comprises SEQ ID NO: 312. Preferably, the marker protein consists of SEQ ID NO: 31.

[183] The marker protein may comprise or consist of SEQ ID NO: 32, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 32. Preferably, the marker protein comprises SEQ ID NO: 32. Preferably, the marker protein consists of SEQ ID NO: 32.

[184] The marker protein may comprise or consist of SEQ ID NO: 33, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 33. Preferably, the marker protein comprises SEQ ID NO: 33. Preferably, the marker protein consists of SEQ ID NO: 33.

[185] The marker protein may comprise or consist of SEQ ID NO: 34, or a fragment or analog thereof having a sequence identity of 60% or higher, 70% or higher, 80% or higher, 90% or higher, 95% or higher, 97.5% or higher, or 99% or higher compared to SEQ ID NO: 34. Preferably, the marker protein comprises SEQ ID NO: 34. Preferably, the marker protein consists of SEQ ID NO: 34.

[186] The marker protein may also be a peptide fused to a membrane anchor. A “membrane anchor” may be a lipid anchor yielding, e.g., prenylated proteins, fatty acylated proteins or glycosylphosphatidylinositol-linked proteins (GPI). The marker protein may comprise one or more membrane anchors.

[187] Prenylated proteins are proteins with covalently attached hydrophobic isoprene polymers (i.e. branched five-carbon hydrocarbon) at cysteine residues of the protein. More specifically, these isoprenoid groups, usually farnesyl (15-carbon) and geranylgeranyl (20-carbon) can be attached to the protein via thioether linkages at cysteine residues near the C terminal of the protein. This prenylation of lipid chains to proteins facilitate their interaction with the cell membrane. The prenylation motif “CAAX box” (SEQ ID NO: 56) is the most common prenylation site in proteins, that is, the site where farnesyl or geranylgeranyl covalently attach. In the CAAX box sequence, the C represents the cysteine that is prenylated, the A represents any aliphatic amino acid and the X determines the type of prenylation that will occur. If the X is an Ala, Met, Ser or Gin the protein will be farnesylated via the farnesyltransferase enzyme and if the X is a Leu then the protein will be geranylgeranylated via the geranylgeranyltransferase I enzyme.

[188] Fatty acylated proteins are proteins that have been (post-translationally) modified to include the covalent attachment of fatty acids at certain amino acid residues. The most common fatty acids that are covalently attached to the protein are the saturated myristic (14-carbon) acid and palmitic acid (16-carbon). Proteins can be modified to contain either one or both of these fatty acids. Fatty acylated proteins are proteins that have been (post-translationally) modified to include the covalent attachment of fatty acids at certain amino acid residues. The most common fatty acids that are covalently attached to the protein are the saturated myristic (14-carbon) acid and palmitic acid (16-carbon). Proteins can be modified to contain either one or both of these fatty acids. N-myristoylation (i.e. attachment of myristic acid) is generally an irreversible protein modification that typically occurs during protein synthesis in which the myrisitc acid is attached to the a-amino group of an N-terminal glycine residue through an amide linkage. This reaction can be facilitated by N-myristoyltransferase. These proteins usually begin with a Met-Gly sequence and with either a serine or threonine at position 5. S-palmitoylation (i.e. attachment of palmitic acid) is a reversible protein modification in which a palmitic acid is attached to a specific cysteine residue via thioester linkage. The palmitoyl group can be removed by palmitoyl thioesterases.

[189] Glycosylphosphatidylinositol-anchored proteins (GPI-anchored proteins) are attached to a GPI complex molecular group via an amide linkage to the protein's C-terminal carboxyl group. This GPI complex may comprise of several main components that are all interconnected: a phosphoethanolamine, a linear tetrasaccharide (composed of three mannoses and a glucosaminyl) and a phosphatidylinositol. The phosphatidylinositol group typically is glycosidically linked to the non-N-acetylated glucosamine of the tetrasaccharide. A phosphodiester bond is then formed between the mannose at the non-reducing end (of the tetrasaccaride) and the phosphoethanolamine. The phosphoethanolamine is then amide linked to the C-terminal of the carboxyl group of the respective protein. The GPI attachment typically occurs through the action of GPI-transamidase complex. The fatty acid chains of the phosphatidylinositol are inserted into the membrane and thus are what anchor the protein to the membrane. These proteins preferably are only located on the exterior surface of the plasma membrane.

[190] The ligand binding partner B comprised in the marker protein preferably is presented on the outside of the cell (to be isolated or fractionated).

[191] In some embodiments, the marker protein contains at least one extracellular domain and a transmembrane domain. In some embodiments, the marker protein is capable of being expressed on the surface of the cell. In some embodiments, the marker protein is a cell surface receptor, ligand, glycoprotein, cell adhesion molecule, antigen, integrin, or cluster of differentiation (CD) or is a modified form thereof.

[192] In some embodiments, the marker protein is not a chimeric antigen receptor. In some embodiments, the marker protein is a modified cell surface molecule that is altered compared to a reference cell surface molecule. In some cases, the modified cell surface molecule lacks a functional intracellular signaling domain and/or is not capable of mediating intracellular signaling.

[193] In some embodiments, the marker protein contains a modified cell surface molecule that is altered compared to a reference cell surface molecule. In some embodiments, the reference cell surface molecule is a cell surface receptor, ligand, glycoprotein, cell adhesion molecule, antigen, integrin, or cluster of differentiation (CD). In some embodiments, the reference cell surface molecule is a cell surface receptor. In some embodiments, the reference cell surface molecule is a native mammalian cell surface molecule, such as a native mammalian cell surface receptor. In some cases, the marker protein is a native human membrane protein.

[194] In some embodiments, the reference cell surface molecule can be one that contains an extracellular domain or regions containing one or more epitope(s) recognized by an antibody or an antigen-binding fragment thereof. The antibody or antigen-binding fragment can include polyclonal and monoclonal antibodies, including intact antibodies and functional (antigen binding) antibody fragments, including fragment antigen binding (Fab) fragments, F(ab') 2 fragments, Fab' fragments, Fv fragments, recombinant IgG (rlgG) fragments, variable heavy chain (VH) regions capable of specifically binding the antigen, single chain antibody fragments, including single chain variable fragments (scFv), and single domain antibodies (e.g., sdAb, sdFv, nanobody) fragments. Antibodies or antigen-binding fragment thereof can include intact or full-length antibodies, including antibodies of any class or sub-class, including IgG and sub classes thereof, IgM, IgE, IgA, and IgD, or portion or fragments of a full-length antibody. In some aspects, the antibody is an antibody or antigen-binding fragment thereof that is clinically approved. In some aspects, the one or more epitopes can contain contiguous or non-contiguous sequences of a molecule or protein. In some aspects, the one or more epitope(s) is present in the extracellular portion or region of the reference cell surface molecule, such that the reference cell surface molecule can be recognized, identified or detected by the antibody or antigenbinding fragment.

[195] In some embodiments, the marker protein also contains a binding domain capable of specifically binding to a binding partner, an antigen, a substrate or a ligand. In such embodiments, among the provided the marker protein are modified cell surface molecules in which such a binding domain is modified or altered, e.g. is mutated or deleted, such that the ability of the modified cell surface molecule to bind to its normal cognate binding partner, antigen, substrate or ligand is reduced compared to the binding of the reference cell surface molecule to the binding partner, antigen, substrate or ligand. In some cases, the altered binding is reduced by greater than or greater than about 40%, greater than or greater than about 50%, greater than or greater than about 60%, greater than or greater than about 70%, greater than or greater than about 80%, greater than or greater than about 90% or more.

[196] In some embodiments, the reference cell surface molecule further contains an intracellular (or cytoplasmic) region or domain, i.e., a region of one or more contiguous amino acids present inside the cell and/or in the cytoplasmic side of the cell. In some cases, the intracellular region of a reference cell surface molecule contains an intracellular signaling domain and/or is capable of mediating intracellular signaling by directly or indirectly modulating cellular signal transduction pathways, and/or downstream responses, functions or activities, such as gene and protein expression, changes in subcellular localization of molecules, intracellular trafficking, changes in protein-protein interaction, receptor internalization, cellular differentiation, proliferation and/or survival.

[197] In some embodiments, the intracellular signaling region or domain, e.g. present in or containing a cytoplasmic tail of the reference cell surface molecule, contains one or more motifs or residues that are capable of being phosphorylated and/or interacting with one or more adaptor proteins in a signal transduction pathway or downstream process in the cell upon a molecular or cellular signal, e.g., when activated or exposed to its cognate antigen or ligand. In some embodiments, the motif is or contains a tyrosine-based motif, or a dileucine based motif (e.g. LL). In some aspects, the intracellular signaling domain of a reference cell surface molecule can be present at or near the C-terminus of type I membrane proteins or at or near the N-terminus of type II membrane proteins. In such embodiments, among the provided cell surface molecules are modified cell surface molecules in which amino acid residues of such an intracellular region or domain is modified or altered, such as mutated, e.g., by one or more substitution, deletion, truncation and/or insertion, such that the ability of the modified cell surface molecule to modulate cellular signal transduction pathways, and/or downstream responses, functions or activities is reduced or prevented. In some cases, the altered signaling and/or downstream responses, functions or activities is reduced by greater than or greater than about 40%, greater than or greater than about 50%, greater than or greater than about 60%, greater than or greater than about 70%, greater than or greater than about 80%, greater than or greater than about 90% or more compared to such signaling and/or downstream responses, functions or activities of a reference cell surface molecule.

[198] In some embodiments, the reference cell surface molecule is different from and/or not identical to the antigen, e.g., a cell surface-expressed antigen, targeted by the recombinant receptor, e.g., chimeric antigen receptor (CAR). In some embodiments, the reference cell surface molecule or modified form thereof, is not specifically bound and/or recognized by the ligand- or antigen-binding domain of the recombinant receptor, e.g., chimeric antigen receptor (CAR).

[199] In some embodiments, the reference cell surface molecule is or includes a cell surface protein and/or a receptor. In some embodiments, the reference cell surface molecule is EpCAM, VEGFR, integrins (e.g., integrins anb3, a4, allbp3, a4b7, a5b1 , anb3, av), TNF receptor superfamily (e.g., TRAIL-R1 , TRAIL-R2), PDGF Receptor, interferon receptor, folate receptor, GPNMB, ICAM-1 , HLA-DR, CEA, CA-125, MUC1 , TAG-72, IL-6 receptor, 5T4, GD2, GD3, prostate-specific membrane antigen (PSMA), or clusters of differentiation (e.g., CD2, CD3, CD4, CD5, CD11 , CDIIa/LFA-1 , CD15, CD18/ITGB2, CD19, CD20, CD22, CD23/lgE Receptor, CD25, CD28, CD30, CD33, CD38, CD40, CD41, CD44, CD51 , CD52, CD62L, CD74, CD80, CD125, CD147/basigin, CD152/CTLA-4, CD154/CD40L, CD195/CCR5, CD3 19/SLAMF7).

The marker protein may comprise a leader sequence or a secretion signal such as the BM40 secretion signal, e.g. as depicted in SEQ ID NO: 4. Further examples of secretion signals are listed in the following (number in brackets refers to an exemplary corresponding UniProt database accession number (v1 of the sequence) and denote after the “|” the sequence positions of the secretion signal in said sequence): INTERLEUKIN-2 (P60568|1-20), CD5 ANTIGEN-LIKE (O43866|1-19), IMMUNOGLOBULIN KAPPA VARIABLE 1-39 (P01597| 1 -22), ALBUMIN (P02768| 1 -18), CD4 human (P01730| 1-25), CD4 mouse (P06332|1-26), CD137 human (Q0701111 -23), ALKALINE PHOSPHATASE, PLACENTAL TYPE (P05187| 1 -22), TUMOR NECROSIS FACTOR RECEPTOR SUPERFAMILY MEMBER 16 (P08138| 1 -28), VON WILLEBRAND FACTOR (P04275|1-22), ERYTHROPOIETIN RECEPTOR (P19235|1-24). These sequences are herewith incorporated by reference.

Expression cassettes

[200] As already described herein, the amount of the marker protein on the cell surface of a cell of the population of cells advantageously is correlated to the level of expression of the protein of interest in said cell. The amount the marker protein on the cell surface is thus an indicator for the ability of a cell (fractionated/isolated by the method of the invention) to produce the protein of interest. Accordingly, a high amount of the marker protein on the cell surface is indicative for a cell producing high amounts of the protein of interest. The inventor found that this correlation can be achieved by expressing the marker protein and the protein of interest from a single nucleic acid or, in other words, using an expression cassette encoding for a marker protein optionally comprising a binding partner B and (ii) a protein of interest.

[201] As used herein, the term "expression cassette" refers to a nucleic acid molecule capable of directing expression of a particular nucleotide sequence in an appropriate host cell. An expression cassette preferably includes a promoter operatively linked to the nucleotide sequence of interest, which is operatively linked to one or more termination signals. It may also include sequences required for proper translation of the nucleotide sequence. The coding region can encode a protein of interest and can also encode a functional RNA of interest, including but not limited to, antisense RNA or a non-translated RNA, in the sense or antisense direction. The expression cassette comprising the nucleotide sequence of interest can be chimeric, meaning that at least one of its components is heterologous with respect to at least one of its other components. The expression cassette can also be one that is naturally occurring but has been obtained in a recombinant form useful for heterologous expression. In some embodiments, however, the expression cassette is heterologous with respect to the host; i.e. , the particular nucleic acid sequence of the expression cassette does not occur naturally in the host cell and was introduced into the host cell or an ancestor of the host cell by a transformation event. The expression of the nucleotide sequence in the expression cassette can be under the control of a constitutive promoter or of an inducible promoter that initiates transcription only when the host cell is exposed to some particular external stimulus. In the case of a multicellular organism such as a plant or an animal, the promoter can also be specific to a particular tissue, organ, or stage of development.

[202] The terms “expression” and “expressed”, as used herein means that a sequence included in a nucleic acid molecule and encoding a peptide/protein is converted into its peptide/protein product. Thus, where the nucleic acid is DNA, expression refers to the transcription of a sequence of the DNA into RNA and the translation of the RNA into protein. Where the nucleic acid is RNA, expression may include the replication of this RNA into further RNA copies and/or the reverse transcription of the RNA into DNA and optionally the transcription of this DNA into further RNA molecule(s). In any case expression of RNA includes the translation of any of the RNA species provided/produced into protein. Hence, expression is performed by translation and includes one or more processes selected from the group consisting of transcription, reverse transcription and replication. Expression of the protein or peptide of the member of the plurality of peptides and/ or proteins may be carried out using an in vitro expression system. Many suitable systems are commercially available. A nucleic acid molecule, such as DNA, is said to be "capable of expressing" a peptide/protein if it contains nucleotide sequences which contain transcriptional and translational regulatory information and such sequences are operably linked to nucleotide sequences which encode the polypeptide. A suitable embodiment for expression purposes is the use of a vector, in particular an expression vector. Thus, provided is also a host cell transformed/ transfected with an expression vector. A nucleic acid or an expression cassette encoding for a marker protein and a protein of interest as described herein is capable of expressing said marker protein and said protein of interest.

[203] The term "nucleic acid” as used herein may refer to any nucleic acid molecule in any possible configuration, such as single stranded, double stranded or a combination thereof. Nucleic acids may include for instance DNA molecules, RNA molecules, analogues of the DNA or RNA generated using nucleotide analogues or using nucleic acid chemistry, locked nucleic acid molecules (LNA), protein nucleic acids molecules (PNA) and tecto-RNA molecules (e.g. Liu, B., et al., J. Am. Chem. Soc. (2004) 126, 4076-4077). A PNA molecule is a nucleic acid molecule in which the backbone is a pseudopeptide rather than a sugar. Accordingly, PNA generally has a charge neutral backbone, in contrast to for example DNA or RNA. Nevertheless, PNA is capable of hybridising at least complementary and substantially complementary nucleic acid strands, just as e.g. DNA or RNA (to which PNA is considered a structural mimic). An LNA molecule has a modified RNA backbone with a methylene bridge between C4' and 02', which locks the furanose ring in a N-type configuration, providing the respective molecule with a higher duplex stability and nuclease resistance. Unlike a PNA molecule an LNA molecule has a charged backbone. DNA or RNA may be of genomic or synthetic origin and may be single or double stranded. Such nucleic acid can be e.g. mRNA, cRNA, synthetic RNA, genomic DNA, cDNA, synthetic DNA, a copolymer of DNA and RNA, oligonucleotides, etc. A respective nucleic acid may furthermore contain non-natural nucleotide analogues and/or be linked to an affinity tag or a label. Preferably, the nucleic acid is DNA or RNA, more preferably DNA.

[204] By “gene” is meant a unit of inheritance that occupies a specific locus on a chromosome and that is a segment of nucleic acid associated with a biological function. A gene encompasses transcriptional and/or translational regulatory sequences as well as a coding region. Besides a coding sequence a gene may include a promoter region, a cis-regulatory sequence, a non-expressed DNA segment that is a specific recognition sequence for regulatory proteins, a non-expressed DNA segment that contributes to gene expression, a DNA segment designed to have desired parameters, or combinations thereof. A gene can be obtained by a variety of methods, including cloning from a biological sample, synthesis based on known or predicted sequence information, and recombinant derivation of an existing sequence.

[205] The term "promoter" as used throughout this document, refers to a nucleic acid sequence needed for gene sequence expression. Promoter regions vary from organism to organism, but are well known to those skilled in the art for different organisms. For example, in prokaryotes, the promoter region contains both the promoter (which directs the initiation of RNA transcription) as well as the DNA sequences which, when transcribed into RNA, will signal synthesis initiation. Such regions will normally include those 5'-non-coding sequences involved with initiation of transcription and translation, such as the TATA box, capping sequence, CAAT sequence, and the like. Both constitutive and inducible promoters can be used in the context of the present invention, in accordance with the needs of a particular embodiment. A large number of promoters recognized by a variety of potential host cells are well known. The selected promoter can be operably linked to cistron DNA encoding a polypeptide described herein by removing the promoter from the source DNA via restriction enzyme digestion and inserting the isolated promoter sequence into the vector of choice. Both the native promoter sequence and many heterologous promoters may be used to direct amplification and/or expression of a selected nucleic acid sequence. In some embodiments, the promoter can be a non-viral promoter or a viral promoter, such as a cytomegalovirus (CMV) promoter such as the promoter depicted in SEQ ID NO: 1 , an SV40 promoter, an RSV promoter, and a promoter found in the long-terminal repeat of the murine stem cell virus. The promoter may be the human elongation factor-1 alpha promoter (EF1a) such as the promoter depicted in SEQ ID NO 9). The promoter may be a phosphoglycerate kinase promoter (PGK) such as the promoter depicted in SEQ ID NO: 26 or 19, preferably 26. Other promoters known to a skilled artisan also are contemplated.

[206] In a method disclosed herein an expression cassette may be introduced into a host cells by any suitable technique of nucleic acid delivery for transformation of a cell available in the art. Examples of suitable techniques include, but are not limited to, direct delivery of DNA, e.g. via transfection, injection, including microinjection, electroporation, calcium phosphate precipitation, by using DEAE-dextran followed by polyethylene glycol, direct sonic loading, liposome mediated transfection, receptor-mediated transfection, microprojectile bombardment, agitation with silicon carbide fibers, Agrobacterium-mediated transformation, desiccation/ inhibition-mediated DNA uptake or any combination thereof.

[207] In the expression cassette used in the context of the invention, the protein of interest and the marker protein may be operably linked to different promotors. Preferably, the marker protein is operably linked to a weak promotor, thereby reducing the burden of the cell caused by expression of the marker protein, while the protein of interest advantageously is operably linked to a strong promotor. The promoter of the marker protein may be inducible and the promoter of the protein of interest may be constitutive. The promoter of the marker protein may be inducible and the promoter of the protein of interest may be inducible. The promoter of the marker protein may be constitutive and the promoter of the protein of interest may be inducible.

[208] An alternative to using two promoters - one for each of the marker protein and the protein of interest - is the use of multicistronic (bicistronic or tricistronic, see e.g., U.S. Patent No. 6,060,273) expression cassettes. Expression cassettes can be engineered as a bicistronic unit containing an IRES (internal ribosome entry site), which allows coexpression of gene products ((e.g. encoding the marker protein and the protein of interest) by a message RNA (mRNA) from a single promoter. Accordingly, the protein of interest may be operably linked to a promotor and the marker protein may be under the control of an internal ribosome entry site (IRES), wherein the protein of interest and the marker protein are transcribed on the same RNA. Exemplary IRES sequences are shown in SEQ ID NO: 3 or 25.

[209] Alternatively, in some cases, a single promoter may direct expression of an expression cassette that contains, in a single open reading frame (ORF), two or three genes (e.g. the marker protein and the protein of interest) separated from one another by sequences encoding a self cleavage peptide (e.g., 2 sequences) or a protease recognition site (e.g., furin). The ORF thus encodes a single polypeptide, which, either during (in the case of 2A) or after translation, is processed into the individual proteins. In some cases, the peptide, such as T2A, can cause the ribosome to skip (ribosome skipping) synthesis of a peptide bond at the C-terminus of a 2A element, leading to separation between the end of the 2A sequence and the next peptide downstream (see, for example, de Felipe. Genetic Vaccines and Ther. 2:13 (2004) and deFelipe et al. Traffic 5:616-626 (2004)). Many 2A elements are known in the art. Examples of 2A sequences that can be used in the methods and system disclosed herein, without limitation, 2A sequences from the foot-and-mouth disease virus (F2A, e.g., SEQ ID NO: 20 or 29, preferably SEQ ID NO: 29), equine rhinitis A virus (E2A, e.g., SEQ ID NO: 30), Thosea asigna virus (T2A, e.g., SEQ ID NO: 31 or SEQ ID NO:32), and porcine teschovirus-1 (P2A, e.g., SEQ ID NO: 33 or 34) as described in U.S. Patent Publication No. 20070116690. Accordingly, the protein of interest and the marker protein may form a fusion protein, wherein the marker protein and the fusion protein can be linked via a self-cleavable peptide such as a 2A peptide.

[210] Exemplary expression cassettes are further depicted in SEQ ID NOs:

Binding pairs L and B

[211] As described herein, the binding partner B comprised in the marker protein or the receptor molecule binding reagent and the ligand L comprised in the solid phase are capable of reversibly binding to each other. Disruption of the binding between the binding partner B and the ligand L results in the elution of a population of cells from the solid phase. In the following, exemplary binding pairs L and B will be described.

[212] The non-covalent bond that is formed between the binding partner B that is included in the marker protein or the receptor molecule binding reagent and the ligand L of the solid phase may be of any desired strength and affinity, as long as it is disruptable or reversible under the conditions under which the method of the invention is performed. The dissociation constant (K_D) of the binding between the binding partner B that is included in the marker protein or the receptor molecule binding reagent and the ligand L of the solid phase may have a value in the range from about 10’² M to about 10'¹³ M. Thus, this reversible bond can, for example, have a K_D from about 10'² M to about 10'¹³ M, or from about 10'³ M to about 10'¹² M or from about 10'⁴ M to about 10'¹¹M, or from about 10'⁵ M to about 10'¹°M. The K_D of this bond as well as the K_D, k_off and k_on rate of the bond formed between the binding site B of the marker protein or the receptor molecule binding reagent and the ligand L of the solid phase can be determined by any suitable means, for example, by fluorescence titration, equilibrium dialysis or surface plasmon resonance. The marker protein may include at least one, including two, three or more, second binding partners B and the solid phase may include at least one, at least two, such as three, four, five, six, seven, eight or more binding sites for the binding partner that is included in the marker protein. As described in US patent 7,776,562 or International Patent applications WO 2002/054065 or WO 2015/166049, any combination of a binding partner B and an solid phase with one or more corresponding ligands L can be chosen, as long as the binding partner B and the ligand L of the are able to reversibly bind each other.

[213] As already described herein, the binding of the binding partner B to the ligand L is reversible. Disrupting (displacing) the reversible binding of the binding partner B to the ligand L can be achieved by contacting the cells with a composition comprising a substance capable of reversing the bond between the binding partner B to the ligand L (“competitor” as used herein) . For example, the competitor is a free binding partner and/or is a competition agent (e.g. a biotin, a biotin analog, a biologically active fragment thereof). In some embodiments, the methods of the invention include after contacting cells in the sample to the solid phase containing the binding molecule bound thereto, applying a competitor to disrupt (displace) the bond between the binding partner B to the ligand L, thereby recovering the selected cells from the solid phase or in other words, eluting a fraction of the cells from the solid phase. The choice of the competitor depends on the particular binding partner B to ligand L. In some embodiments, the binding partner B is a streptavidin mutein (e.g. Strep-Tactin®) for recognition of a streptavidin binding peptide (e.g. Strep-tag® or a Twin-Strep-tag®) comprised in the ligand L and the competitor is biotin or a biotin analog.

[214] The term “competitor” or “competition reagent” - both terms can be used interchangeably - as used herein refers to any reagent or condition that is able to reduce, interfere with or abrogate the formation of a complex between a pair of binding agents or moieties, such as a binding partner B and a ligand L. The term “competition” is meant to refer any interference with binding, regardless of the nature of such interference. Such interference may in some embodiments also be a non-competitive binding to a certain binding site. An example of such a competition mechanism is the metal chelation by a chelating reagent such as EDTA or EGTA, when the reversibly bond is mediated by complexed metal ions such as Ca²⁺, Ni²⁺, Co²⁺, or Zn²⁺. This mechanism applies for binding pairs such as calmodulin and calmodulin binding peptides that bind in the presence of Ca²⁺ or for binding pairs that are used in Immobilized Metal-chelate Affinity Chromatography (IMAC). In some embodiments, a competition reagent may have a binding site that is capable of specifically binding to the binding site included on one of the binding partners, e.g. binding site B or ligand L. It is also possible that the entire competition reagent is capable of specifically binding to the binding site included on one of these binding partners. In some embodiments competition is provided by a change in pH or the salt strength of a buffer and the competition reagent is then either an increased or decreased pH or salt strength. A change in pH can, for example, be used for displacing/disrupting the binding of streptavidin to a streptavidin binding peptide or for displacing/disrupting the binding between protein A or protein G and an antibody Fc domain. [215] Preferably, the competitor is Biotin or a derivative thereof, more preferably Biotin. Biotin may be added to a concentration of at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin.

[216] In some embodiments, the competitor is or contains biotin, a biotin analog or a biologically active fragment thereof. Accordingly, the fractions may be eluted by adding biotin. Among such embodiments are particularly those in which the binding partner B is a streptavidin binding peptide and the ligand L is a streptavidin or a streptavidin mutein or analog, including any as described. In these embodiments, biotin may be added to a concentration of at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin. Accordingly, the first concentration may be at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin. The second concentration may be at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin. The third concentration at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin. Any further concentration may be at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin. The first, the second, the third and each subsequent (further) concentration may be at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin. Preferably, the concentration, i.e. the first, the second, the third and each subsequent concentration, of biotin is between 0,9 and 1,1 mM Biotin, more preferably about 1 mM Biotin.

[217] The first concentration may be at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM. The second concentration may be at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM. The third concentration at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM. Any further concentration may be at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM.

[218] The binding partner B included in the marker protein or the receptor molecule binding reagent may be an oligopeptide, a polypeptide, a protein, a nucleic acid, a lipid, a saccharide, an oligosaccharide, or a polysaccharide. Such a binding partner has a higher affinity to the binding site of the solid phase than to other matter. Examples of a respective binding partner include, but are not limited to, an immunoglobulin molecule, a fragment thereof and a proteinaceous binding molecule with antibody-like functions.

[219] In some embodiments the binding partner B that is included in the marker protein or the receptor molecule binding reagent includes biotin and the ligand L comprised in the solid phase includes a streptavidin analog or an avidin analog that reversibly binds to biotin. In some embodiments the binding partner B that is included in the marker protein or the receptor molecule binding reagent includes a biotin analog that reversibly binds to streptavidin or avidin, and the ligand L comprised in the solid phase includes streptavidin, avidin, a streptavidin analog or an avidin analog that reversibly binds to the respective biotin analog. In some embodiments the binding partner B that is included in the marker protein or the receptor molecule binding reagent includes a streptavidin or avidin binding peptide and the ligand L comprised in the solid phase includes streptavidin, avidin, a streptavidin analog or an avidin analog that reversibly binds to the respective streptavidin or avidin binding peptide.

[220] In some embodiments the binding partner B that is included in the marker protein or the receptor molecule binding reagent may include a streptavidin-binding peptide that comprises or consists of one of the following sequences: a) -Trp-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO: 35), wherein Xaa is any amino acid and Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg , b) -Trp-Arg-His-Pro-GIn-Phe-Gly-Gly- (SEQ ID NO: 36), c) -Trp-Ser-His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO: 37), d) a sequential arrangement of at least two streptavidin binding peptides, wherein each peptide binds streptavidin, wherein the distance between two peptides is at least 0 and not greater than 50 amino acids and wherein each of the at least two peptides comprises the amino acid sequence -His-Pro-Baa- in which Baa is selected from the group consisting of glutamine, asparagine and methionine, e) a sequential arrangement as recited in d), wherein one of the at least two peptides comprises the sequence -His-Pro-GIn- f) a sequential arrangement as recited in d), wherein one of the peptides comprises an amino acid sequence -His-Pro-GIn-Phe- (SEQ ID NO: 38), g) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino sequence -Oaa-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO: 39), where Oaa is Trp, Lys or Arg, Xaa is any amino acid and where either Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg, h) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino acid sequence -Trp-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO: 40) where Xaa is any amino acid and where either Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg, i) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino acid sequence -Trp-Ser-His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO: 41), j) the amino acid sequence -Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(Xaa)n-Trp-Ser- His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO: 42) wherein Xaa is any amino acid and n is an integer from 0 to 12. k) an amino acid sequence selected from the group consisting of Trp-Arg-His-Pro-GIn- Phe-Gly-Gly (SEQ ID NO: 41), Trp-Ser-His-Pro-GIn-Phe-Glu-Lys (SEQ ID NO: 43), Trp- Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₃-Trp-Ser-His-Pro-Gln-Phe-Glu-Lys (SEQ ID NO: 44), Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₂-Trp-Ser-His-Pro-Gln- Phe-Glu-Lys (SEQ ID NO: 45) or Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₂-Gly- Gly-Ser-Ala-Trp-Ser-His-Pro-GIn-Phe-Glu-Lys (SEQ ID NO: 46).

In these cases, the solid phase may include the streptavidin mutein (analog) Val⁴⁴-Thr⁴⁵-Ala⁴⁶- Arg⁴⁷ (SEQ ID NO: 54) or the streptavidin mutein (analog) lle⁴⁴-Gly⁴⁵-Ala⁴⁶-Arg⁴⁷ (SEQ ID NO: 55), both of which are described in US patent 6,103,493, for example, and are commercially available under the trademark Strep-Tactin®. Such multimeric streptavidin muteins may also be referred to as multimerized Strep-Tactin.

[221] In some embodiments, the binding partner B of the marker protein or the receptor molecule binding reagent includes a moiety known to the skilled artisan as an affinity tag. In such an embodiment the solid phase includes a corresponding binding partner, for example, an antibody or an antibody fragment, known to bind to the affinity tag. As a few illustrative examples of known affinity tags, the binding partner B that is included in the marker protein or the receptor molecule binding reagent may include an oligohistidine, maltose-binding protein, glutathione-S-transferase (GST), chitin binding protein (CBP) or thioredoxin, calmodulin binding peptide (CBP), FLAG-peptide, the HA-tag (sequence: Tyr-Pro-Tyr-Asp-Val-Pro-Asp-Tyr-Ala, SEQ ID NO: 48,), the VSV-G-tag (sequence: Tyr-Thr-Asp-lle-Glu-Met-Asn-Arg-Leu-Gly-Lys, SEQ ID NO: 49), the HSV-tag or HSV epitope of the herpes simplex virus glycoprotein D (sequence: Gln-Pro-Glu-Leu-Ala-Pro-Glu-Asp-Pro-Glu-Asp, SEQ ID NO: 50), the T7 epitope (Ala-Ser-Met-Thr-Gly-Gly-GIn-GIn-Met-Gly, SEQ ID NO: 51), maltose binding protein (MBP), the “myc” epitope of the transcription factor c-myc of the sequence Glu-GIn-Lys-Leu-lle-Ser-Glu- Glu-Asp-Leu (SEQ ID NO: 52), the V5-tag (sequence: Gly-Lys-Pro-lle-Pro-Asn-Pro-Leu-Leu- Gly-Leu-Asp-Ser-Thr, SEQ ID NO: 53), or glutathione-S-transferase (GST). In such an embodiment the complex formed between the one or more ligands L of the solid phase, in this case an antibody or antibody fragment, and the antigen can be disrupted competitively by adding the free antigen, i.e. the free peptide (epitope tag) or the free protein (such as MBP or CBP). The affinity tag might also be an oligonucleotide tag. Such an oligonucleotide tag may, for instance, be used to hybridize to an oligonucleotide with a complementary sequence, linked to or included in the solid phase.

[222] Further examples of a suitable binding pair include using an immunoglobulin domain such as antibody Fc domain as binding partner B in the marker protein or the receptor molecule binding reagent and protein A, protein G or protein L comprised as ligand L in the solid phase. Protein A, protein G and protein L are all able to reversibly bind an antibody Fc domain. The binding can be disrupted by a change in the buffer conditions, for example, by increasing the salt strength of the buffer or by reducing the pH from, for example a neutral pH of about 7.0 to a pH of about 3.0 to about 2.5. These changes of the buffer conditions can be seen as adding a competitor to a certain concentration.

[223] In some embodiments the binding between the binding partner B that is included in the marker protein or the receptor molecule binding reagent and one or more ligands L of the solid phase occurs in the presence of a divalent, a trivalent or a tetravalent cation. In this regard in some embodiments the solid phase includes a divalent, a trivalent or a tetravalent cation, typically held, e.g. complexed, by means of a suitable chelator. The binding partner that is included in the marker protein or the receptor molecule binding reagent may in such an embodiment include a moiety that includes, e.g. complexes, a divalent, a trivalent or a tetravalent cation. Examples of a respective metal chelator, include, but are not limited to, ethylenediamine, ethylenediaminetetraacetic acid (EDTA), ethylene glycol tetraacetic acid (EGTA), diethylenetriaminepentaacetic acid (DTPA), N,N-bis(carboxymethyl)glycine (also called nitrilotriacetic acid, NTA), 1 ,2-bis(o-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA), 2,3-dimercapto-1-propanol (dimercaprol), porphine and heme. As an example, EDTA forms a complex with most divalent, trivalent and tetravalent metal ions, such as e.g. calcium (Ca²⁺), manganese (Mn²⁺), copper (Cu²⁺), iron (Fe²⁺), cobalt (Co³⁺) and zirconium (Zr⁴⁺), while BAPTA is specific for Ca²⁺. As an illustrative example, a standard method used in the art is the formation of a complex between an oligohistidine tag and copper (Cu²⁺), nickel (Ni²⁺), cobalt (Co²⁺), or zinc (Zn²⁺) ions, which are presented by means of the chelator nitrilotriacetic acid (NTA).

[224] In some embodiments the binding partner B that is included in the marker protein or the receptor molecule binding reagent includes a calmodulin binding peptide and the solid phase’s ligand L includes multimeric calmodulin as described in US Patent 5,985,658, for example. In some embodiments the binding partner B that is included in the marker protein or the receptor molecule binding reagent includes a FLAG peptide and the solid phase includes an antibody that binds to the FLAG peptide, e.g. the FLAG peptide, which binds to the monoclonal antibody 4E11 as described in US Patent 4,851 ,341. In some embodiments the antibody that binds to the FLAG peptide may be the commercially available monoclonal antibody M1. In one embodiment the binding partner B that is included in the marker protein or the receptor molecule binding reagent includes an oligohistidine tag and the solid phase includes chelating groups K that bind a transition metal ion and thereby are also able of binding an oligohistidine tag. The disruption of all these binding complexes may be accomplished by metal ion chelation, e.g. calcium chelation, for instance by adding EDTA or EGTA (supra). Calmodulin, antibodies such as 4E11 or chelated metal ions or free chelators may be multimerized by conventional methods, e.g. by biotinylation and complexation with streptavidin or avidin or multimers thereof or by the introduction of carboxyl residues into a polysaccharide, e.g. dextran, essentially as described in Noguchi, A, et al. Bioconjugate Chemistry (1992) 3, 132-137 in a first step and linking calmodulin or antibodies or chelated metal ions or free chelators via primary amino groups to the carboxyl groups in the polysaccharide, e.g. dextran, backbone using conventional carbodiimide chemistry in a second step. In such embodiments the binding between the binding partner B that is included in the marker protein or the receptor molecule binding reagent and the one or more ligands L of the solid phase can be disrupted by metal ion chelation. The metal chelation may, for example, be accomplished by addition of EGTA or EDTA, which can be seen as competitors.

[225] In some embodiments the solid phase comprises an oligomer or a polymer of streptavidin or avidin or of any analog of streptavidin or Avidin as ligand L. Here, the ligand L is the natural biotin binding of avidin or streptavidin. A respective oligomer or polymer may be obtained from a corresponding monomeric streptavidin or avidin, or analog thereof. In addition, a variety of techniques for forming an oligomer or polymer are known in the art. The respective oligomer or polymer may for instance be crosslinked by a polysaccharide. In one embodiment oligomers or polymers of streptavidin or of avidin or of analogs of streptavidin or of avidin are prepared by the introduction of carboxyl residues into a polysaccharide, e. g. dextran, essentially as described in Noguchi, A, et al., Bioconjugate Chemistry (1992) 3,132-137 in a first step. Then streptavidin or avidin or analogs thereof may be linked via primary amino groups of internal lysine residue and/or the free N-terminus to the carboxyl groups in the dextran backbone using conventional carbodiimide chemistry in a second step. Cross-linked oligomers or polymers of streptavidin or avidin or of any analog of streptavidin or avidin may also be obtained by crosslinking via bifunctional molecules, serving as a linker, such as glutardialdehyde or by other methods described in the art. The use of iminothiolan/SMCC, NHS activated carboxydextran or dendrimers are further examples of crosslinking techniques established in the art.

[226] As an illustrative example, an oligomer or a polymer of streptavidin or of avidin or of an analog of streptavidin or of avidin may be prepared by the introduction of carboxyl residue into a polysaccharide such as dextran, essentially as described by Noguchi et al. (Bioconjugate Chemistry [1992] 3, 132-137) in a first step. Then streptavidin or avidin or an analog thereof can be coupled via primary amino groups of internal lysine residues and/or the free N -terminus to the carboxyl groups in the dextran backbone using conventional carbodiimide chemistry in a second step. In one embodiment the coupling reaction is performed at a molar ratio of about 60 moles streptavidin or streptavidin mutein per mole of dextran.

[227] In some embodiments, streptavidin muteins are those streptavidin muteins which are described in US Patent 6,103,493 and also in DE 196 41 876.3. These streptavidin muteins have at least one mutation within the region of amino acid positions 44 to 53, based on the amino acid sequence of wild-type streptavidin. In some embodiments a mutein of a minimal streptavidin is used. A mutein of a minimal streptavidin starts N-terminally in the region of amino acids 10 to 16 of wild-type streptavidin and ends C-terminally in the region of amino acids 133 to 142 of wild-type streptavidin. Examples of such streptavidin muteins have a hydrophobic aliphatic amino acid instead of Glu at position 44, any amino acid at position 45, a hydrophobic aliphatic amino acid at position 46 or/and a basic amino acid instead of Vai at position 47. The streptavidin mutein may be the mutein Val⁴⁴-Thr⁴⁵-Ala⁴⁶-Arg⁴⁷ (SEQ ID NO: 54) or the streptavidin mutein (analog) lle⁴⁴-Gly⁴⁵-Ala⁴⁶-Arg⁴⁷ (SEQ ID NO: 55), both of which are described in US patent 6,103,493, for example, and which are commercially available under the trademark Strep-Tactin®.

[228] As wild-type streptavidin (wt-streptavidin), the amino acid sequence disclosed by Argarana et al., Nucleic Acids Res. 14 (1986) 1871-1882 is referred to. Streptavidin muteins are polypeptides which are distinguished from the sequence of wild-type streptavidin by one or more amino acid substitutions, deletions or additions and which retain the binding properties of wt-streptavidin. Streptavidin-like polypeptides and streptavidin muteins are polypeptides which essentially are immunologically equivalent to wild-type streptavidin and are in particular capable of binding biotin, biotin derivative or biotin analogues with the same or different affinity as wt- streptavidin. Streptavidin-like polypeptides or streptavidin muteins may contain amino acids which are not part of wild-type streptavidin or they may include only a part of wild-type streptavidin. Streptavidin-like polypeptides are also polypeptides which are not identical to wildtype streptavidin, since the host does not have the enzymes which are required in order to transform the host-produced polypeptide into the structure of wild-type streptavidin. The term “streptavidin” also includes streptavidin tetramers and streptavidin dimers, in particular streptavidin homotetramers, streptavidin homodimers, streptavidin heterotetramers and streptavidin heterodimers. Each subunit normally has a binding site for biotin or biotin analogues or for streptavidin-binding peptides. Examples of streptavidins or streptavidin muteins are mentioned, for example, in WO 86/02077, DE 19641876 Al, US 6,022,951 , WO 98/40396 or WO 96/24606.

[229] In illustrative examples, the ligand L and the ligand binding partner B form a binding pair selected from the group of

- streptavidin or a streptavidin analog as ligand L and a ligand binding partner B (molecule) binding to streptavidin,

- a binding pair that binds in the presence of a divalent cation,

- an oligohistidine peptide as binding partner B and a binding moiety A comprising at least two chelating groups K as ligand L, wherein each chelating group is capable of binding to a transition metal ion, thereby rendering moiety A capable of binding to the oligohistidine peptide,

- an antigen and an antibody against said antigen, wherein said binding partner B comprises the antigen and said ligand L comprises the antibody against said antigen,

- a protein selected from the group of glutathione-S-transferase, maltose binding protein (MBP), a chitin binding domain, and a cellulose binding domain as ligand binding partner B and a glutathione, maltose, chitin, or cellulose, respectively as ligand L,

- an antibody Fc domain as ligand binding partner B and an immunoglobulin binding protein such as protein A, protein G or protein L as ligand L.

[230] In further illustrative examples, the binding partner B and the ligand L form a binding pair selected from the group of streptavidin or a streptavidin analog and a ligand binding to streptavidin, a binding pair that binds in the presence of a divalent cation, an oligohistidine peptide and a binding moiety A comprising at least two chelating groups K, wherein each chelating group K is capable of binding to a transition metal ion, thereby rendering binding moiety A capable of binding to the oligohistidine peptide, an antigen and an antibody against said antigen, wherein said binding partner B comprises the antigen and said ligand L comprises the antibody against said antigen.

[231] In further illustrative examples for a binding pair that binds in the presence of a divalent cation, said binding partner B comprises a calmodulin binding peptide and the said ligand L comprises calmodulin, or wherein said binding partner B comprises a FLAG peptide and said multimerization reagent comprises an antibody binding the FLAG peptide, or wherein said binding partner B comprises an oligohistidine tag and said ligand L comprises a chelated transition metal.

[232] The divalent cation may be selected from the group consisting of Ca²⁺, Ni ²⁺or Co²⁺. The binding between said binding partner B and said ligand L may be disrupted by metal ion chelation, preferably wherein the metal chelation is accomplished by addition of EDTA or EGTA, which here can be seen as the competitor.

[233] A method as disclosed herein may be carried out at any temperature at which the viability of the target cell is at least essentially uncompromised. When reference is made herein to conditions that are at least essentially not harmful, not detrimental or at least essentially not compromising viability, conditions are referred to, under which the percentage of target cells that can be recovered with full viability, is at least 70 %, including at least 75 %, at least 80 %, at least 85 %, at least 90 %, at least 92 %, at least 95 %, at least 97 %, at least 98 %, at least 99 % or at least 99.5 %. In some embodiments a method according to the invention is carried out at a temperature of about 20 °C or below, such as about 14 °C or below, about 9 °C or below or about 6 °C or below. Depending on the target cell to be isolated a suitable temperature range may for instance be from about 2 °C to about 45 °C, including from about 2 °C to about 40 °C, from about 3 °C to about 35 °C, or from about 4 °C to about 30 °C if an aqueous medium is used to encompass the target cell. In some embodiments a method according to the invention is carried out at a constant temperature value, or at a selected temperature value ± about 5 °C, ± about 4 °C, ± about 3 °C, ± about 2 °C, ± about 1 °C or ± about 0.5 °C. The temperature may, for example, be selected to have a value of about 5 °C, about 10 °C, about 15 °C, about 20 °C or about 25 °C. In some embodiments the temperature is altered, i.e. increased, decreased or varied by combinations thereof, during a method according to the present invention. The temperature may for example be altered within a range as defined above, e.g. in the range from about 2 °C to about 40 °C or within the range from about 3 °C to about 35 °C. The person skilled in the art is able to empirically determine a suitable temperature, taking into account the nature of the cells and the isolation conditions. For example, temperature insensitive cells such as cancer cells might be isolated at room temperature or even elevated temperature such as 37 °C. Solid phases

[234] As described herein, the method of the invention makes use of a solid phase comprising a ligand L, wherein the ligand L is capable of reversibly binding to the binding partner B comprised in the marker protein or the receptor molecule binding reagent on the cell surface. The solid phase may be used in a batch method or a chromatographic method. Accordingly, the method of the present invention may be a batch method. The method of the present invention may also be a chromatographic method, the solid phase may comprise one or more ligands L.

[235] As already mentioned, the method described herein may be practiced as part of fluid chromatography, typically a liquid chromatography. Any material may be employed as chromatography matrix in the context of the invention, as long as the material is suitable for the chromatographic isolation of the chosen biological entity such as cells. The chromatography matrix corresponds to the solid phase in the context of the method of the invention. A suitable chromatography material is at least essentially innocuous, i.e. not detrimental to cell viability (or the viability or stability of the biological entity), when used in a packed chromatography column under desired conditions for cell isolation and/or cell separation. A chromatography matrix as used in a method described herein typically remains in a predefined location, typically in a predefined position, whereas the location of the sample to be separated and of components included therein is being altered, i.e. the chromatography matrix can also be seen as a stationary phase. As an illustrative example, if packed-bed chromatography columns are employed, the solid phase is generally confined between the bottom of the column and the flow adapter. Where chromatography is carried out as expanded bed adsorption, the resin becomes fluidized in use, and beads employed arrange in the form of a concentration gradient, individual beads taking a position where their sedimentation velocity matches the upward liquid flow velocity. The chromatography matrix is thus a “stationary phase” (corresponding to the “solid phase” used in the context of the present invention) in line with the common understanding of the person skilled in the art in that the stationary phase is that part of a chromatographic system through which the mobile phase flows and where components included in the liquid phase are being disseminated between the phases.

[236] If beads are employed, in column chromatography beads are commonly rather uniform in size, whereas in expanded bed adsorption beads are variable in size, typically ranging from about 50 to about 400 pm. In this regard, it is noted that particles such as freely moveable magnetic beads that are added to a liquid sample, mixed with the sample and are then removed from the sample, for example, by discarding the supernatant (liquid) while holding the beads temporarily in place (for example, by an external magnetic or by centrifugation) are in one embodiment not a solid phase as used herein. However, the method of the invention can also be practiced in a batch mode. In such a method (magnetic) beads can be added to a sample containing the target cells for immobilization of the cells (via a complex formed between the binding partner B comprised in the marker protein or the receptor molecule binding reagent and the ligand L of the solid phase) on such beads, and the beads are then separated from the sample, for example by temporarily holding the beads in place, while discarding the supernatant. Such a batch method is also a method according to the invention.

[237] Typically, the respective chromatography matrix has the form of a solid or semi-solid phase, whereas the sample that contains the cell to be fractionated/isolated/separated is a fluid phase. The mobile phase used to achieve separation is likewise a fluid phase. The chromatography matrix can be a particulate material (of any suitable size and shape) or a monolithic chromatography material, including a paper substrate or membrane. Thus, the chromatography can for example be column chromatography. In some embodiments the chromatography may be planar chromatography. In some embodiments the chromatography may be expanded bed chromatography. If a particulate matrix material is used in column chromatography, the particulate matrix material may, for example, have a mean particle size of about 5 pm to about 200 pm, or from about 5 pm to about 400 pm, or from about 5 pm to about 600 pm. As explained in detail the following, the chromatography matrix may, for example, be or include a polymeric resin or a metal oxide or a metalloid oxide. If planar chromatography is used, the matrix material may be any material suitable for planar chromatography, such as conventional cellulose-based or organic polymer based membranes (for example, a paper membrane, a nitrocellulose membrane or a polyvinylidene difluoride (PVDF) membrane) or silica coated glass plates. In one embodiment, the chromatography matrix/solid phase is a nonmagnetic material or non-magnetisable material.

[238] Non-magnetic or non-magnetisable chromatography solid phases that are used in the art, and that are also suitable in a method described herein, include derivatized silica or a crosslinked gel. A crosslinked gel (which is typically manufactured in a bead form) may be based on a natural polymer, i.e. on a polymer class that occurs in nature. For example, a natural polymer on which a chromatography solid phase is based is a polysaccharide. A respective polysaccharide is generally crosslinked. An example of a polysaccharide matrix is an agarose gel (for example, Superflow™ agarose or a Sepharose® material such as Superflow™ Sepharose® that are commercially available in different bead and pore sizes) or a gel of crosslinked dextran(s). A further illustrative example is a particulate cross-linked agarose matrix, to which dextran is covalently bonded, that is commercially available (in various bead sizes and with various pore sizes) as Sephadex® or Superdex®, both available from GE Healthcare. Another illustrative example of such a chromatography material is Sephacryl® which is also available in different bead and pore sizes from GE Healthcare. Solid phases including beads made from Agarose preferably are used in batch mode, more preferably Agarose beads having a diameter of 200 pm or larger.

[239] A crosslinked gel may also be based on a synthetic polymer, i.e. on a polymer class that does not occur in nature. Usually such a synthetic polymer on which a chromatography solid phase for cell separation is based is a polymer that has polar monomer units, and which is therefore in itself polar. Such a polar polymer is hydrophilic. Hydrophilic ("water-loving") molecules, also termed lipophobic ("fat hating"), contain moieties that can form dipole-dipole interactions with water molecules. Hydrophobic ("water hating") molecules, also termed lipophilic, have a tendency to separate from water.

[240] Illustrative examples of suitable synthetic polymers are polyacrylamide(s), a styrene- divinylbenzene gel and a copolymer of an acrylate and a diol or of an acrylamide and a diol. An illustrative example is a polymethacrylate gel, commercially available as a Fractogel®. A further example is a copolymer of ethylene glycol and methacrylate, commercially available as a Toyopearl®. In some embodiments a chromatography solid phase may also include natural and synthetic polymer components, such as a composite matrix or a composite or a co-polymer of a polysaccharide and agarose, e.g. a polyacrylamide/agarose composite, or of a polysaccharide and N,N'-methylenebisacrylamide. An illustrative example of a copolymer of a dextran and N,N'- methylenebisacrylamide is the above-mentioned Sephacryl® series of material. A derivatized silica may include silica particles that are coupled to a synthetic or to a natural polymer. Examples of such embodiments include, but are not limited to, polysaccharide grafted silica, polyvinylpyrrolidone grafted silica, polyethylene oxide grafted silica, poly(2- hydroxyethylaspartamide) silica and poly(N-isopropylacrylamide) grafted silica.

[241] A solid phase such as a chromatography matrix employed in a method described herein may also include magnetically attractable particles. Also such respective magnetically attractable particles may include a ligand L that is capable of binding a binding partner B comprised in the solid phase. Magnetically attractable particles may contain diamagnetic, ferromagnetic, paramagnetic or superparamagnetic material. Superparamagnetic material responds to a magnetic field with an induced magnetic field without a resulting permanent magnetization. Magnetic particles based on iron oxide are for example commercially available as Dynabeads® from Dynal Biotech, as magnetic MicroBeads from Miltenyi Biotec, as magnetic porous glass beads from CPG Inc., as well as from various other sources, such as Roche Applied Science, BIOCLON, BioSource International Inc., micromod, AMBION, Merck, Bangs Laboratories, Polysciences, or Novagen Inc., to name only a few. Magnetic nanoparticles based on superparamagnetic Co and FeCo, as well as ferromagnetic Co nanocrystals have been described, for example by Hutten, A. et al. (J. Biotech. (2004), 112, 47-63). In some embodiments a chromatography matrix employed in a method disclosed herein is void of any magnetically attractable matter.

[242] In a method of fractionating cells of a population of cells based on the amount of a marker protein on the cell surface, a chromatography matrix - here corresponding to the solid phase - may be employed as an affinity chromatography matrix. An affinity chromatography matrix may itself include permanently bonded (usually covalently bonded) moieties that are capable to specifically bind a selected target. For example, a conventional affinity chromatography matrix may include an antibody that binds a particular given target. Alternatively, a chromatography matrix that is used for Immobilized Metal-chelate Affinity Chromatography (IMAC) is modified with a chelating ligand agent such as tridentate iminodiacetic acid to be able to form coordination bonds between metal ions and certain exposed side chains of a protein or with oligohistidine tags, for example. Thus, in the art an affinity chromatography matrix is generally designed such that it, by itself, is able to specifically bind the target that is to be isolated. In some embodiments of a method as disclosed herein, a solid phase is used as a replacement for a “selection cartridge” as described in International Patent Application WO 2013/124474.

[243] A chromatography matrix, such as a first or, if employed, a second stationary (solid) phase, is in some embodiments included in a chromatography column, for example packed therein. In some embodiments a first and a second solid phase is employed. The first solid phase corresponds to the solid phase described above; it includes for instance a ligand L. The second solid phase may be used to deplete the eluate of the first solid phase from reagents used such as the receptor binding reagent, a competition reagent and/or a multimerization reagent. Such a second solid phase can thus be a “removal cartridge” as described in International Patent Application WO 2013/124474. In some embodiments the second solid phase includes an affinity reagent, typically covalently attached thereto. The affinity reagent may be able to bind the competitor. Such a chromatography matrix may be an affinity chromatography matrix. It may also be a gel filtration matrix, to which the affinity reagent has been coupled. By means of the immobilized affinity reagent the chromatography matrix can deplete a mobile phase of the receptor molecule binding reagent. A sample that is contacted with the chromatography matrix, for example, loaded onto a column packed therewith, can likewise be depleted of the receptor binding reagent. As an illustrative example, in case of using biotin as competitor, the affinity reagent can be streptavidin coupled to a chromatography matrix such as Sephararose™ (see International Patent Application WO 2013/124474 for a detailed description of such a “removal cartridge”). Thus, the present invention provides the possibility of automated isolation of the target cells under removal of all reagents used in the method of the invention. [244] After applying a sample that contains the population of cells, a chromatography matrix as used herein may subsequently be washed with a mobile phase, such as an aqueous medium, e.g. a buffer, in order to remove any matter that has not been immobilized on the chromatography matrix. Such washing may be carried out on any solid phase employed in the context of a method or use described herein. The respective chromatography matrix may be used as a first solid phase or as a secondary solid phase.

[245] Dissociation of the above-described cells may then be induced, for example, by a change in conditions. Such a change in conditions may for instance be a change in the pH or ionic strength of an aqueous mobile phase. In some embodiments, a competitor is employed in order to induce dissociation of the reversible non-covalent complex between the marker protein or the receptor molecule binding reagent and the solid phase. The competitor is able to associate to the solid phase by occupying or blocking the binding site of the ligand L of the solid phase for the binding partner B included in the marker protein or the receptor molecule binding reagent. By using a competitor with a particularly high affinity for the solid phase or by using an excess of the competitor relative to concentration of the marker protein or the receptor molecule binding reagent, the non-covalent bonding between the marker protein or the receptor molecule binding reagent and the solid phase may be disrupted. The (target) cell is allowed to elute from the chromatography matrix, e.g. from the column into which the chromatography matrix is packed. The eluate, comprising a fraction of cells, is collected and the target cell thereby collected and/or fractionated.

[246] The fluid phase used as the mobile phase in chromatography may be any fluid suitable for preserving the biological activity of the (target) cell or cell to be fractionated. Typically, the fluid is a liquid. In some embodiments the respective liquid is or includes water, for example in the form of an aqueous solution. Further components may be included in a respective aqueous solution, for example dissolved or suspended therein. As an illustrative example an aqueous solution may include one or more buffer compounds. Numerous buffer compounds are used in the art and may be used to carry out the various processes described herein. Examples of buffers include, but are not limited to, solutions of salts of phosphate such as phosphate buffered saline (PBS), carbonate, succinate, carbonate, citrate, acetate, formate, borate, N-(2- acetamido)-2-amino-ethanesulfonate (also called (ACES), N-(2-hydroxyethyl)-piperazine-N'-2- ethanesulfonic acid (also called HEPES), 4-(2-hydroxyethyl)-1-piperazine-propanesulfonic acid (also called HEPPS), piperazine-1,4-bis(2-ethanesulfonic acid) (also called PIPES), (2- [Tris(hydroxymethyl)-methylamino]-1-ethansulfonic acid (also called TES), 2-cyclohexylamino- ethanesulfonic acid (also called CHES) and N-(2-acetamido)-iminodiacetate (also called ADA). Any counter ion may be used in these salts; ammonium, sodium, and potassium may serve as illustrative examples. Further examples of buffers include, but are not limited to, tri- ethanolamine, diethanolamine, zwitter-ionic buffers such as betaine, ethylamine, triethylamine, glycine, glycylglycine, histidine, tris-(hydroxymethyl)aminomethane (also called TRIS), bis-(2- hydroxyethyl)-imino-tris(hydroxymethyl)-methane (also called BIS-TRIS), and N- [Tris(hydroxymethyl)-methyl]-glycine (also called TRICINE), to name only a few. The buffer may further include components that stabilize the target cell to be isolated, for example proteins such as (serum) albumin, growth factors, trace elements and the like. The choice of the suitable mobile phase is within the knowledge of the person of average skill in the art and can be carried out empirically.

Proteins of Interest

[247] The term "protein of interest” (POI) as used herein refers to a protein that is produced by means of recombinant technology in a host cell. More specifically, the protein may either be a polypeptide not naturally occurring in the host cell, i.e. a heterologous protein e.g. an artificial protein such as a protein not naturally produced by wild-type cells, or else may be native to the host cell, i.e. a homologous protein to the host cell, but is produced, for example, by transformation with a self-replicating vector containing the nucleic acid sequence encoding the POI, or upon integration by recombinant techniques of one or more copies of the nucleic acid sequence encoding the POI into the genome of the host cell, or by recombinant modification of one or more regulatory sequences controlling the expression of the gene encoding the POI, e.g. of the promoter sequence. In general, the proteins of interest referred to herein may be produced by methods of recombinant expression well known to a person skilled in the art. The protein of interest may be a recombinant protein.

[248] There is no limitation with respect to the protein of interest (POI). The POI is usually a eukaryotic or prokaryotic polypeptide, variant or derivative thereof, or an artificial polypeptide, such as a polypeptide not naturally produced by wild-type cells. The POI can be any eukaryotic or prokaryotic protein. The protein can be a naturally secreted protein or an intracellular protein, i.e. a protein which is not naturally secreted or a membrane protein. The present invention also includes biologically active fragments of proteins. In another embodiment, a POI may be an amino acid chain or present in a complex, such as a dimer, trimer, hetero-dimer, multimer or oligomer.

[249] The protein of interest may be selected from the group consisting of an amylolytic enzyme, a lipolytic enzyme, a proteolytic enzyme, a cellulytic enzyme, an oxidoreductase or a plant cell-wall degrading enzyme, an aminopeptidase, amylase, amyloglucosidase, carbohydrase, carboxypeptidase, catalase, cellulase, chitinase, cutinase, cyclodextrin glycosyltransferase, desoxyribonuclease, esterase, galactosidase, beta-galactosidase, glucoamylase, glucose oxidase, glucosidase, haloperoxidase, hemicellulase, invertase, isomerase, laccase, ligase, lipase, lyase, mannosidase, oxidase, pectinase, peroxidase, phytase, phenoloxidase, polyphenoloxidase, protease, ribonuclease, transferase, transglutaminase, and xylanase, a growth factor, cytokine, receptors, receptor ligands, therapeutic proteins such as interferons, BMPs, GDF proteins, fibroblast growth factors, peptides such as protein inhibitors, membrane proteins, membrane-associated proteins, peptide/protein hormones, peptidic toxins, peptidic antitoxins, antibody or functional fragments thereof such as Fab or F(ab)₂ or derivatives of an antibody such as bispecific antibodies (for example, scFvs), chimeric antibodies, humanized antibodies, single domain antibodies such as Nanobodies or domain antibodies (dAbs) or an anticalin and others.

Selection marker

[250] One advantage of the present invention is that it is not necessary to initially and/or continuously select the cells for producing a protein of interest fractionated/isolated by the method of the present invention using a selection marker. In contrast, the metabolic burden to the cells caused by expressing the selection marker can be avoided (see also Examples 7 and 11). However, such a (pre-)selection step may be nevertheless a step of the method of the invention. Accordingly, the method of the invention may further comprise a step (a’) prior to step (a):

(a’) selecting cells, which express the selection marker.

[251] A “selection marker” as used herein generally relates to a gene that is introduced into the modified organism, e.g. a cell comprised in a population of cells, as a marker, preferably together with the "gene of interest" encoding the protein of interest, i.e. the gene that is actually desired, in order to be able to recognize individuals with successful genetic modification. Frequently used selection markers are antibiotic resistance, auxotrophies or herbicide resistance. Successfully genetically modified organisms can then survive on a selection medium containing the corresponding substance. In some cases, selection markers create a burden on the metabolism of the transgenic organism and can potentially be transferred to other organisms through horizontal gene transfer, which is why methods have been developed to remove the selection marker after selection, e.g., with TALENs or zinc finger nucleases. As described herein, the method of the invention also is such a method to avoid selection markers. Positive selection occurs, for example, by using antibiotic or herbicide resistance (transgenic organisms can grow), and negative selection by using toxic genes (non-transgenic organisms can grow). Auxotrophs can be used for both negative and positive selection.

[252] An exemplary selection marker is a puromycin resistance gene, e.g., encoding the protein depicted in SEQ ID NO: 21. The invention further relates to the following items:

1. A method for fractionating cells of a population of cells based on the amount of a marker protein on the cell surface, the method comprising:

2. A method for fractionating cells of a population of cells based on the amount of a receptor molecule on the cell surface, the method comprising:

(i) optionally repeating steps (g) and (h) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more times. A method for enriching or isolating cells comprised in a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells, the method comprising:

(e) optionally adding the competitor to a second concentration, thereby eluting a fraction of the population of cells from the solid phase; (f) optionally separating the eluted fraction of the population of cells obtained in step

(e);

(i) optionally repeating steps (g) and (h) 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10 or more times. The method of any one of the preceding items, wherein the first concentration, the second concentration and the further concentration(s) are essentially the same. 6. The method of any one of items 1-4, wherein the second concentration is higher than the first concentration, wherein the further concentration is higher than the second concentration and each subsequent further concentration is higher than the previous further concentration.

7. The method of any one of items 2, 4, 5 or 6, wherein the population of cells comprises cells comprising a nucleic acid comprising (i) an expression cassette encoding for a marker protein comprising and (ii) a protein of interest.

8. The method of any one of the preceding items, wherein the amount of the marker protein or the receptor molecule on the cell surface of a cell of the population of cells is correlated to the level of expression of the protein of interest.

9. The method of any one of the preceding items, wherein in the expression cassette

(i) the protein of interest and the marker protein are operably linked to different promotors;

(ii) the protein of interest is operably linked to a promotor and the marker protein is under the control of an internal ribosome entry site (IRES), wherein the protein of interest and the marker protein are transcribed on the same RNA;

(iii) the protein of interest and the marker protein form a fusion protein, wherein the marker protein and the fusion protein are linked via a self-cleavable peptide such as a 2A peptide, are linked via a cleavable amino acid sequence that can be cleaved by a protease or form a fusion protein due to a leaky stop codon between the sequence of the protein of interest and the sequence of the marker protein; or

(iv) the protein of interest is immobilized on the cell surface via methods like cold capture which preferably stops or reduces secretion of the protein by trapping it temporally on the cell surface e.g., by reducing the temperature of the medium.

10. The method of any one of the preceding items, wherein the marker protein

(i) is a transmembrane protein or a fragment thereof, wherein the binding site B is comprised in the extracellular domain; or

(ii) is a peptide fused to a membrane anchor.

11 . The method of any one of the preceding items, wherein the marker protein comprises a transmembrane domain of a protein selected from the group consisting of EpCAM, VEGFR, integrin, optionally integrins avp3, a4, alip3, a4p7, a5pi , avp3 or an, a member of the TNF receptor superfamily, optionally TRAIL-RI or TRAIL-R2, a member of the epidermal growth factor receptor family, PDGF Receptor, interferon receptor, folate receptor, GPNMB, ICAM-1 , HLA-DR, CEA, CA-125, MUCI, TAG-72, IL-6 receptor, 5T4, GD2, GD3, prostate-specific membrane antigen (PSMA) or a clusters of differentiation cell surface molecule, optionally CD2, CD3, CD4, CD5, CD11 , CDIIa/LFA-1 , CD15, CD18/ITGB2, CD19, CD20, CD22, CD23/lgE Receptor, CD25, CD28, CD30, CD33, CD38, CD40, CD41 , CD44, CD51 , CD52, CD62L, CD74, CD80, CD125, CD147/basigin, CD152/CTLA-4, CD154/CD40L, CD195/CCR5 and CD319/SLAMF7, low-affinity nerve growth factor receptor (LNGFR), preferably CD4. The method of any one of the preceding items, wherein the marker protein comprises or consists of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28, or a fragment or analog thereof having a sequence identity of 60% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28. The method of any one of the preceding items, wherein the binding partner B and the ligand L form a binding pair selected from the group of streptavidin or a streptavidin analog and a ligand binding to streptavidin, a binding pair that binds in the presence of a divalent cation, an oligohistidine peptide and a binding moiety A comprising at least two chelating groups K, wherein each chelating group K is capable of binding to a transition metal ion, thereby rendering binding moiety A capable of binding to the oligohistidine peptide, an antigen and an antibody against said antigen, wherein said binding partner B comprises the antigen and said ligand L comprises the antibody against said antigen. The method of item 13, wherein

(a) said binding partner B comprises biotin and said ligand L comprises a streptavidin analog or an Avidin analog that reversibly binds to biotin,

(b) said binding partner B comprises a biotin analog that reversibly binds to streptavidin or Avidin and said ligand L comprises streptavidin, or Avidin, or a streptavidin analog, or an Avidin analog that reversibly binds to said biotin analog, or (c) said binding partner B comprises a streptavidin or Avidin binding peptide and said ligand L comprises streptavidin, or Avidin, or a streptavidin analog, or an Avidin analog that reversibly binds to said streptavidin or Avidin binding peptide.

15. The method of item 14, wherein said ligand L comprises a streptavidin mutein comprising the amino acid sequence Val⁴⁴-Thr⁴⁵-Ala⁴⁶-Arg⁴⁷ (SEQ ID NO: 54) at sequence positions 44 to 47 of wild-type streptavidin or a streptavidin mutein comprising the amino acid sequence lle⁴⁴-Gly⁴⁵-Ala⁴⁶-Arg⁴⁷ (SEQ ID NO: 55) at sequence positions 44 to 47 of wild-type streptavidin and wherein said binding partner B comprises the streptavidin-binding peptide that comprises or consists of one of the following sequences: a) -Trp-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO: 35), wherein Xaa is any amino acid and Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg , b) -Trp-Arg-His-Pro-GIn-Phe-Gly-Gly- (SEQ ID NO: 36), c) -Trp-Ser-His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO: 37), d) a sequential arrangement of at least two streptavidin binding peptides, wherein each peptide binds streptavidin, wherein the distance between two peptides is at least 0 and not greater than 50 amino acids and wherein each of the at least two peptides comprises the amino acid sequence -His-Pro-Baa- in which Baa is selected from the group consisting of glutamine, asparagine and methionine, e) a sequential arrangement as recited in d), wherein one of the at least two peptides comprises the sequence -His-Pro-GIn- f) a sequential arrangement as recited in d), wherein one of the peptides comprises an amino acid sequence -His-Pro-GIn-Phe- (SEQ ID NO: 38), g) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino sequence -Oaa-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO: 39), where Oaa is Trp, Lys or Arg, Xaa is any amino acid and where either Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg, h) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino acid sequence -Trp-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO:

40) where Xaa is any amino acid and where either Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg, i) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino acid sequence -Trp-Ser-His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO:

41), j) the amino acid sequence -Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(Xaa)n-Trp-Ser- His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO: 42) wherein Xaa is any amino acid and n is an integer from 0 to 12. k) an amino acid sequence selected from the group consisting of Trp-Arg-His-Pro-GIn- Phe-Gly-Gly (SEQ ID NO: 41), Trp-Ser-His-Pro-GIn-Phe-Glu-Lys (SEQ ID NO: 43), Trp- Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₃-Trp-Ser-His-Pro-Gln-Phe-Glu-Lys (SEQ ID NO: 44), Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₂-Trp-Ser-His-Pro-Gln- Phe-Glu-Lys (SEQ ID NO: 46) or Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₂-Gly- Gly-Ser-Ala-Trp-Ser-His-Pro-GIn-Phe-Glu-Lys (SEQ ID NO: 47).

16. The method of item 13, wherein for a binding pair that binds in the presence of a divalent cation, said binding partner B comprises a calmodulin binding peptide and the said ligand L comprises calmodulin, or wherein said binding partner B comprises a FLAG peptide and said multimerization reagent comprises an antibody binding the FLAG peptide, or wherein said binding partner B comprises an oligohistidine tag and said ligand L comprises a chelated transition metal.

17. The method of item 16, wherein the divalent cation is selected from the group consisting of Ca²⁺, Ni ²⁺or Co²⁺.

18. The method of item 16, wherein the binding between said binding partner B and said ligand L is disrupted by metal ion chelation, preferably wherein the metal chelation is accomplished by addition of EDTA or EGTA.

19. The method of item 13, wherein the antigen comprised in said binding partner B is an epitope tag.

20. The method of item 19, where the epitope tag is selected from the group consisting of the Myc-tag (sequence: EQKLISEEDL, SEQ ID NO: 52), the HA-tag (sequence: YPYDVPDYA, SEQ ID NO: 48), the VSV-G-tag (sequence: YTDIEMNRLGK, SEQ ID NO: 49), the HSV-tag (sequence: QPELAPEDPED, SEQ ID NO: 50), and the V5-tag (sequence: GKPIPNPLLGLDST, SEQ ID NO: 53).

21. The method of item 13, wherein the antigen comprised in said binding partner B is a protein. 22. The method of item 21 , wherein the protein is selected from the group of glutathione-S- transferase, maltose binding protein (MBP), chitin binding protein (CBP) and thioredoxin.

23. The method of item 15, wherein the fractions of the cells are eluted by adding biotin.

24. The method of item 23, wherein biotin is added to a concentration of at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin.

25. The method of any one of the preceding items, wherein the solid phase is a selected from a bead, a plastic plate, a membrane or a solid phase suitable for chromatography.

26. The method of any one of the preceding items, wherein the method is a batch method or a chromatographic method.

27. The method of any one of the preceding items, wherein the nucleic acid comprised in cells of the population of cells further comprises a selection marker.

28. The method of item 27, wherein the method further comprises a step (a’) prior to step (a):

(a’) selecting cells, which express the selection marker.

29. The method of any one of the preceding items, wherein the cell is a prokaryotic cell or a eukaryotic cell.

30. The method of item 29, wherein the cell is a prokaryotic cell, preferably of the species selected from the group consisting of Lactobacillus spp., Yersinia spp., Escherichia spp., Klebsiella spp., Bordetella spp., Neisseria spp., Aeromonas spp., Franciesella spp., Corynebacterium spp., Citrobacter spp., Chlamydia spp., Hemophilus spp., Brucella spp., Mycobacterium spp., Legionella spp., Rhodococcus spp., Pseudomonas spp., Helicobacter spp., Salmonella spp., Vibrio spp., Bacillus spp., Leish mania spp. and Erysipelothrix spp., Shigella spp., Listeria spp., Rickettsia spp., Acetoanaerobium spp., Aerococcaceae spp., Carnobacteriaceae spp., Enterococcace spp., Leuconostocacease spp., Streptococcaceae spp., and bacteria with GRAS status, preferably E. coli. 31. The method of item 29, wherein the host cell is a eukaryotic cell, preferably a cell selected from the group consisting of CHO cells, CHO-S, ExpiCHO, Freestyle CHO-S, CHO-GS, CHO-K1, CHO-DXB11, CHO-DG44, CHO duk, CHO-DP12, CHOZN, GS- CHOK1SV, HEK-293 cells, HEK-293T cells, HEK-293-6E, HEK-293-EBNA, HEK 293SF-3F6, 293 c18, Expi293, 293-F, insect cells, SF9, ExpiSf9, Hi-5, Sf21, human amniocytes and CAP®.

32. The method of any one of the preceding items, wherein the protein of interest is selected from the group consisting of an amylolytic enzyme, a lipolytic enzyme, a proteolytic enzyme, a cellulytic enzyme, an oxidoreductase or a plant cell-wall degrading enzyme, an aminopeptidase, amylase, amyloglucosidase, carbohydrase, carboxypeptidase, catalase, cellulase, chitinase, cutinase, cyclodextrin glycosyltransferase, desoxyribonuclease, esterase, galactosidase, beta-galactosidase, glucoamylase, glucose oxidase, glucosidase, haloperoxidase, hemicellulase, invertase, isomerase, laccase, ligase, lipase, lyase, mannosidase, oxidase, pectinase, peroxidase, phytase, phenoloxidase, polyphenoloxidase, protease, ribonuclease, transferase, transglutaminase, and xylanase, a growth factor, cytokine, receptors, receptor ligands, therapeutic proteins such as interferons, BMPs, GDF proteins, fibroblast growth factors, peptides such as protein inhibitors, membrane proteins, membrane-associated proteins, peptide/protein hormones, peptidic toxins, peptidic antitoxins, antibody or functional fragments thereof such as Fab or F(ab)₂ or derivatives of an antibody such as bispecific antibodies (for example, scFvs), chimeric antibodies, humanized antibodies, single domain antibodies such as Nanobodies or domain antibodies (dAbs) or an anticalin and others.

33. The method of any one of items 1-4 and 7-32, wherein the second concentration is lower than the first concentration, wherein the further concentration is lower than the second concentration and each subsequent further concentration is lower than the previous further concentration.

34. The method of any one of the preceding items, wherein the marker protein and the protein of interest are identical.

****

[253] It is noted that as used herein, the singular forms “a”, “an”, and “the”, include plural references unless the context clearly indicates otherwise. Thus, for example, reference to “a reagent” includes one or more of such different reagents and reference to “the method” includes reference to equivalent steps and methods known to those of ordinary skill in the art that could be modified or substituted for the methods described herein.

[254] Unless otherwise indicated, the term "at least" preceding a series of elements is to be understood to refer to every element in the series. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the present invention.

[255] The term "and/or" wherever used herein includes the meaning of "and", "or" and "all or any other combination of the elements connected by said term".

[256] The term “less than” or in turn “more than” does not include the concrete number.

[257] For example, less than 20 means less than the number indicated. Similarly, more than or greater than means more than or greater than the indicated number, e.g. more than 80 % means more than or greater than the indicated number of 80 %.

[258] 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 integer or step. When used herein the term “comprising” can be substituted with the term “containing” or “including” or sometimes when used herein with the term “having”. When used herein “consisting of" excludes any element, step, or ingredient not specified.

[259] The term “including” means “including but not limited to”. “Including” and “including but not limited to” are used interchangeably.

[260] As used herein the terms "about", "approximately" or “essentially” mean within 20%, preferably within 15%, preferably within 10%, and more preferably within 5% of a given value or range. It also includes the concrete number, i.e. “about 20” includes the number of 20.

[261] It should be understood that this invention is not limited to the particular methodology, protocols, material, reagents, and substances, etc., described herein and as such can vary. 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 is defined solely by the claims.

[262] All publications cited throughout the text of this specification (including all patents, patent application, scientific publications, instructions, etc.), whether supra or infra, are hereby incorporated by reference in their entirety. Nothing herein is to be construed as an admission that the invention is not entitled to antedate such disclosure by virtue of prior invention. To the extent the material incorporated by reference contradicts or is inconsistent with this specification, the specification will supersede any such material. [263] The content of all documents and patent documents cited herein is incorporated by reference in their entirety.

Overview of Sequences

EXAMPLES

[264] An even better understanding of the present invention and of its advantages will be evident from the following examples, offered for illustrative purposes only. The examples are not intended to limit the scope of the present invention in any way. Sequences of Vector elements

[265] Figures 5 to 11 provide an overview of the constructs used (see description of the drawings for further explanation).

Example 1 - Selection of GFP expressing HEK-293 cells that present a truncated CD137- TST receptor using Strep-Tactin® magnetic microbeads

[266] 4 x 10⁷ cells of a HEK-293HEK-293 suspension cell culture transfected with a DNA plasmid, comprising the genetic elements comprised in Figure 5 comprising the expression cassette of SEQ ID NO 8 which expression is controlled by an IRES sequence (SEQ ID NO 3), were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in 40 ml of Buffer IS (8,1 mM Na₂HPO₄, 1 ,5 mM KH₂PO₄, 137 mM NaCI, 5 g/l BSA, pH 7.4) containing 1 mM EDTA. The cells were centrifuged again at 100 x g for 5 min and the cell pellet was suspended in 2 ml TrypLE solution (Thermo Fisher, Cat. No. 12604013). The mixture was incubated for 2 min at room temperature (20°C). Then 40 ml Buffer IS containing 1 mM EDTA was added. 50% of the cell mixture was centrifuged at 100 x g for 5 min. The pellet was suspended in 250 pl Buffer IS containing 1 mM EDTA.

[267] 125 pl of the suspension were added to 60 pl Strep-Tactin® magnetic microbeads (IBA, Cat. No. 6-5510-050) that were prepared as follows. Strep-Tactin® magnetic microbeads were washed with 1 ml Buffer IS containing 1 mM EDTA on a magnet before use. The cell magnetic microbead solution was incubated for 20 min at 2-8°C on a roller mixer. Afterwards CD137-TST positive cells were separated on StrepMan magnet (IBA, Cat. No. 6-5650-065) as follows: After incubation on the roller mixer, the cell-microbead-mixture was transferred to a 15 ml centrifugation tube containing 5 ml Buffer IS containing 1 mM EDTA .

[268] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads (CD137-TST positive cell fraction) were carefully washed off the tube wall by suspending in 5 ml Buffer IS containing 1 mM EDTA .

[269] This was repeated twice. Then the tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet. Positive cells on Strep-Tactin® magnetic microbeads were suspended in 500 pl of a 1 mM D-biotin working solution. The solution was prepared by mixing Buffer IS with 100 mM D-biotin stock solution (IBA, Cat. No. 6-0219-001). [270] The cell bead mixture with the D-biotin working solution were mixed thoroughly by pipetting and incubated for 10 min at 2-8°C on a roller mixer. The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the positive cell fraction was transferred to fresh tube by careful pipetting while keeping the reaction tube on the StrepMan Magnet.

[271] Selected cells in elution were centrifuged at 100 x g for 5 min at 4°C and suspended in 300 pl Buffer IS containing 1 mM EDTA. Cells were analyzed on a Accuri C6 flow cytometer

[272] By the use of the method of invention, the fraction of GFP positive cells was increased compared to the fraction of GFP positive cells of the cultures before selection (Figure 12 and Figure 14). The fraction of GFP positive cells was 48% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 93%. The mean fluorescence intensity of the eluted population was increased compared to the original population (Figure 14). The mean fluorescence intensity was 2345714 before selection with the method of invention. Mean fluorescence intensity was increased to 4622371 in the eluted cells fraction.

Example 2 - Selection of GFP expressing HEK-293 cells that present truncated CD137- TST receptor using Strep-Tactin® magnetic microbeads and anti CD137 TST-Fab

[273] 4 x 10⁷ cells of a HEK-293 suspension cell culture transfected with a DNA plasmid comprising, comprising the genetic elements comprised in Figure 5 comprising the expression cassette of SEQ ID NO 8 which expression is controlled by an IRES sequence (SEQ ID NO 3), were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in 40 ml of Buffer IS containing 1 mM EDTA. The cells were centrifuged again at 100 x g for 5 min and the cell pellet was suspended in 2 ml TrypLE solution (Thermo Fisher, Cat. No. 12604013). The mixture was incubated for 2 min at room temperature (20°C). Then 40 ml Buffer IS containing 1 mM EDTA was added. 50% of the cell mixture was centrifuged at 100 x g for 5 min. The pellet was suspended in 250 pl Buffer IS containing 1 mM EDTA.

[274] 16 pl of anti CD137-Fab-TST solution (250 pg/ml) were added to 60 pl of Strep- Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) that were prepared as follows. Strep-Tactin® magnetic microbeads were washed with 1 ml Buffer IS containing 1 mM EDTA on a magnet before use. The Fab-magnetic microbead mixture was incubated under gentle constant agitation on a roller mixer at 2-8°C for 30 min.

[275] 125 pl of the cell suspension were added to the mixture. Afterwards, CD137-Twin-Strep- tag® positive cells were separated on StrepMan magnet (I BA, Cat. No. 6-5650-065) as follows: After 20 min incubation of the cell-Fab- microbead-mixture on a roller mixer at 2-8°C, the cell- Fab- microbead-mixture was transferred to a 15 ml centrifugation tube containing 5 ml Buffer IS containing 1 mM EDTA. CD137 positive cells were selected as described in Example 1.

[276] By the use of the method of invention, the fraction of GFP positive cells was increased compared to the fraction of GFP positive cells of the cultures before selection (Figure 13 and Figure 15). The fraction of GFP positive cells was 48% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 98%. The mean fluorescence intensity of the eluted population was increased compared to the original population (Figure 15). The mean fluorescence intensity was 2345714 before selection with the method of invention. Mean fluorescence intensity was increased to 3934904 in the eluted cells fraction.

Example 3 - Selection of GFP expressing CHO cells that present truncated CD137-TST receptor using Strep-Tactin® magnetic microbeads

[277] 4 x 10⁷ cells of a CHO suspension cell culture transfected with a DNA plasmid comprising, comprising the genetic elements comprised in Figure 5 comprising the expression cassette of SEQ ID NO 8, which expression is controlled by an IRES sequence (SEQ ID NO 3), were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in 40 ml of Buffer IS containing 1 mM EDTA. The cells were centrifuged again at 100 x g for 5 min. The pellet was suspended in 300 pl Buffer IS containing 1 mM EDTA .

[278] 150 pl of the suspension was added to 60 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) that were prepared as follows. Strep-Tactin® magnetic microbeads were washed with 1 ml Buffer IS containing 1 mM EDTA on a magnet before use. Afterwards, CD137-Twin-Strep-tag® positive cells were separated on StrepMan magnet (I BA, Cat. No. 6- 5650-065) as follows: After 20 min incubation on the roller mixer at 2-8 °C, the cell bead mixture was transferred to a 15 ml centrifugation tube containing 5 ml Buffer IS containing 1 mM EDTA.

[279] CD137 positive cells were selected as described in Example 1.

[280] By the use of the method of invention, the fraction of GFP positive cells was increased compared to the fraction of GFP positive cells of the cultures before selection (Figure 16 and Figure 17). The fraction of GFP positive was 20% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 77%. The mean fluorescence intensity was 781694 before selection with the method of invention. Mean fluorescence intensity was 712424 in the eluted cells fraction (Figure 17). Example 4 - Selection of GFP expressing CHO cells that present truncated CD137-TST receptor using Strep-Tactin® magnetic microbeads and anti CD137 TST-Fab

[281] 4 x 10⁷ cells of a CHO suspension cell culture transfected with a DNA plasmid comprising, comprising the genetic elements comprised in Figure 5 comprising the expression cassette of SEQ ID NO 8, which expression is controlled by an IRES (SEQ ID NO 3), were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in 40 ml of Buffer IS containing 1 mM EDTA. The cells were centrifuged again at 100 x g for 5 min. The pellet was suspended in 300 pl Buffer IS containing 1 mM EDTA.

[282] 16 pl of anti CD137-Fab-TST solution (250 pg/ml) were added to 60 pl of Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) that were prepared as follows. Strep-Tactin® magnetic microbeads were washed with 1 ml Buffer IS containing 1 mM EDTA on a magnet before use. The Fab-magnetic bead mixture was incubated under gentle constant agitation on a roller mixer at 2-8°C for 30 min.

[283] 150 pl of the cell suspension from para. [281] was added to the mix in [282], Afterwards, CD137-Twin-Strep-tag® positive cells were separated on StrepMan magnet (I BA, Cat. No. 6- 5650-065) as follows: After 20 min incubation of the cell-Fab-microbead-mixture on a roller mixer at 2-8°C, the cell-Fab-microbead-mixture was transferred to a 15 ml centrifugation tube containing 5 ml Buffer IS containing 1 mM EDTA.

[284] CD137 positive cells were selected as described in Example 1.

[285] By the use of the method of invention, the fraction of GFP positive cells was increased compared to the fraction of GFP positive cells of the cultures before selection (Figure 16 and Figure 18). The fraction of GFP positive cells was 20% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 81%. The mean fluorescence intensity was 781694 before selection with the method of invention. Mean fluorescence intensity was 598718 in the eluted cells fraction (Figure 18).

Example 5 - Selection of HEK-293 cells displaying different CD4 surface protein variants fused to a Twin-Strep-tag® using a Strep-Tactin® coated nitro cellulose mebrane

[286] Strep-Tactin® was bound on a nitro cellulose membrane.

[287] 1 ,05 x 10⁷ cells of a HEK-293 suspension cell culture transfected with DNA plasmids either comprising the genetic elements comprised in Figure 6 comprising the expression cassette of SEQ ID NO 16 or genetic elements comprised in Figure 7 comprising the expression cassette of SEQ ID NO 17 or the genetic elements comprised in Figure 8 comprising the expression cassette of SEQ ID NO 18, which expression is controlled in all plasmids by an EF1a promoter (SEQ ID NO 9) were centrifuged at 100 x g for 5 min at room temperature. The supernatant was discarded and the cell pellet was suspended in 10 ml PBS (Capricorn). The cells were centrifuged again at 100 x g for 5 min at room temperature and the cell pellet was suspended in 1 ml Dissociation buffer (Gibco, Cat. No. 1851598). The mixture was incubated for 5 min at room temperature (20°C). Then 9 ml MEXi-TM (IBA GmbH, Cat. No. 2-6011-010) (37°C) were added to the mixture. Cell aggregates were dispensed by pipetting and the suspension was centrifuged at 100 x g for 5 min at room temperature. The pellet was suspended in 400 pl MEXi-TM. The cells were added to the membrane and incubated for 20 min in a well of microwell-plate at 37°C and 5% CO₂ in an incubator. Every 5 minutes the well plate was shaken gently.

[288] After incubation the membrane was transferred to a well with fresh MEXi-TM. The well plate was shaken gently to wash unbound negative cells from the membrane. The wash step was repeated 5 times.

[289] Then, the membrane was transferred into a petri dish containing 10 ml of MEXi-CM elution solution. The MEXi-CM elution solution was prepared by mixing MEXi-CM with a 100 mM D-biotin stock solution to increase the biotin concentration of MEXi-CM by 1mM D- Biotin. The membrane was incubated for 10 min to elute positive cells from the membrane. Afterwards, the medium with the eluted cells was centrifuged at 200 x g for 5 min. The pellet was suspended in 500 pl MEXi-CM.

[290] Cells expressing the transmembrane domain of CD4 (CD4tm) fused to TST and GFP were isolated successfully using the Strep-Tactin® coated nitro cellulose membrane (Figure 19 and Figure 20. 98% of selected cells expressed the TST-CD4-transmembrane-GFP fusion protein (Figure 20). If a rigid linker between the CD4 transmembrane domain and the TST was used (TST-rigid-CD4tm-GFP), the yield of selected cells was higher compared to cells expressing a CD4 transmembrane domain with a flexible linker (TST-flexible-CD4tm-GFP). The mean fluorescence intensity of the isolated cell population was 1.4 fold (TST-flexible-CD4tm- GFP) and 1.2 fold (TST-rigid-CD4tm-GFP) increased due to the selection.

Example 6 - Selection of HEK-293 cells displaying different CD4 surface protein variants fused to a Twin-Strep-tag® using Strep-Tactin® magnetic microbeads

[291] 1.0 x 10⁷ cells of a HEK-293 suspension cell culture transfected with DNA plasmids either comprising the genetic elements comprised in Figure 6 comprising the expression cassette of SEQ ID NO 16 or genetic elements comprised in Figure 7 comprising the expression cassette of SEQ ID NO 17 or the genetic elements comprised in Figure 8 comprising the expression cassette of SEQ ID NO 18 which expression is controlled in all plasmids by an EF1a promoter (SEQ ID NO 9) were centrifuged at 100 x g for 5 min at room temperature. The supernatant was discarded and the cell pellet was suspended in 10 ml PBS (Capricorn). The cells were centrifuged again at 100 x g for 5 min at room temperature and the cell pellet was suspended in 1.5 ml Dissociation buffer (Gibco, Cat. No. 1851598). The mixture was incubated for 5 min at room temperature (20°C). Then 9 ml MEXi-TM (37°C) were added to the mixture. Cell aggregates were dispensed by pipetting and the suspension was centrifuged at 100 x g for 5 min at room temperature. The pellet was suspended in 100 pl MEXi-TM.

[292] 100 pl of the suspension was added to 30 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) that were prepared as follows. Strep-Tactin® magnetic microbeads were washed with 1 ml MEXi-TM on a magnet before use. Afterwards, positive cells (expressing CD4-variants fused to Twin-Strep-tag®) were separated on StrepMan magnet (I BA, Cat. No. 6- 5650-065) as follows: After incubation for 46 minutes at 2-8°C on a roller mixer, the cell bead mixture was transferred to a 15 ml centrifugation tube containing 10 ml MEXi-TM.

[293] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads (CD4-TST-Variants, positive cell fraction) were carefully washed off the tube wall by suspending in 10 ml MEXi-TM. This procedure was repeated twice.

[294] Then the tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet. Positive cells on Strep-Tactin® magnetic microbeads were suspended in 10 ml D-biotin-MEXi working solution. The solution was prepared by mixing MEXi-CM with a 100 mM D-biotin stock solution to increase the biotin concentration of MEXi-CM by 1 mM.

[295] The cell bead mixture with the D-biotin-MEXi working solution were mixed thoroughly by pipetting and incubated for 88 min at 2-8°C on a roller mixer. The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the positive cell fraction was transferred to fresh tube by careful pipetting while keeping the reaction tube on the StrepMan Magnet. This procedure was repeated two times with 5 ml D-biotin-MEXi working solution and a 10 min incubation on the roller mixer at 2-8°C.

[296] All three tested TST-CD4-GFP variants allow the selection of TST-CD4-GFP expressing cells leading to a significant higher portion of TST-CD4-GFP positive cells in the selected cell populations compared to the cell population before selection (Figure 21). The fraction of GFP positive cells expressing only the transmembrane domain of CD4 linked by a rigid linker to TST (TST-rigid-CD4tm-GFP) was 54% before selection. After selection with the method of invention, the fraction of TST-rigid-CD4tm-GFP positive cells was 95% in elution 1 and 100% in elution 2. The fraction of GFP positive cells expressing only the transmembrane domain of CD4 linked by a flexible linker to TST (TST-flexible-CD4tm-GFP) was 52% before selection. After the selection with the method of invention, the fraction of TST-flexible-CD4tm-GFP positive cells was 98% in elution 1 and 100% in elution 2. The fraction of GFP positive cells expressing the transmembrane and the extracellular domain of CD4 fused to TST (TST-CD4-truncated-GFP) was 42% before selection. After the selection with the method of invention, the fraction of TST- CD4-truncated-GFP positive cells was 86% in elution 1 and 95% in elution 2. Cells expressing only the transmembrane domain CD4 variants exhibit higher mean fluorescence intensity levels and thus higher TST-CD4-GFP marker expressions than cells expressing the transmembrane and the extracellular domain of CD4 (Figure 22). Fluorescence Intensity was higher in elution fraction 2 compared to elution fraction 1.

[297] Cells expressing only the CD4 transmembrane domain exhibit higher yields after selection compared to cells expressing the transmembrane and extra cellular domain of CD4 (Figure 23). The yield of TST-flexible-CD4tm-GFP cells was 1.48 fold higher and the yield of TST-rigid-CD4tm-GFP cells was 1.74 fold higher than the yield of TST-CD4-truncated-GFP cells. These higher yields demonstrate that a short extracellular domain is sufficient for the selection of the cells and also increase the yield of cells due to a higher concentration of TST on the cells’ surface. In addition expression of the transmembrane domain only reduces the cost for the cell to produce the selection marker compared to protein of larger size. These reduced costs can be an explanation for the higher expression of marker protein if only the transmembrane domain is expressed. In addition the orientation of the TST at the N-terminus of the CD4 extracellular domain might be suboptimal for binding given to the structure of the protein. The transmembrane domain with an artificial linker might place the tag in more beneficial orientation for binding to the Strep-Tactin®. All three CD4 marker variants allowed the effective selection of marker expressing cells. However, the expression of TST-rigid-CD4tm- GFP and TST-flexible-CD4tm-GFP allowed a more efficient selection with higher yields.

Example 7 - Selection of CHO cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using puromycin combined with the method of invention

[298] For this experiment a CHO cell suspension culture was transfected with a DNA plasmid comprising the genetic elements comprised in Figure 9 comprising the expression cassette of SEQ ID NO 23 which expression is controlled by a PGK promoter (SEQ ID NO 19). Cells of the culture were selected with different methods as described in the following examples of this section. Cells in the examples were analyzed by a Cytoflex flow cytometer (Beckman Coulter, Inc., Model No. A00-1-1102) to detect viable GFP positive cells and their mean fluorescence intensity.

Example 7.1- Selection of CHO cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using puromycin

[299] 2 days after transfection, 6 x 10⁶ cells of a CHO suspension cell culture transfected with a DNA plasmid comprising the genetic elements comprised in Figure 9 comprising the expression cassette of SEQ ID NO 23 which expression is controlled by a PGK promoter (SEQ ID NO 19) were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in 10 ml CHO-TF medium (Xell AG, Cat. No. 886-0001) containing 6 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). The cells were subcultured every 3 to 4 days in CHO-TF medium with 6 pg/ml puromycin. For subculturing, the cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in CHO-TF medium (Xell AG, Cat. No. 886-0001) containing 6 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). The cell concentration was adjusted to 6 x 10⁵ cells/ml at each subculturing. Cells were cultured at 37°C and 5% CO₂. The cultivation was terminated at day 15 because the viability of the culture was 21% and cell density decreased from 4.6 x 10⁵ cells/ml to 1.5 x 10⁵ cells/ml within 4 days. The selection with puromycin thus constitutes a metabolic burden to the cells.

Example 7.2- Selection of CHO cells displaying truncated a CD4 surface protein fused to a Twin-Strep-tag® using puromycin and Strep-Tactin® magnetic microbeads

[300] 2 days after transfection, 2 x 10⁷ cells of a CHO suspension cell culture transfected with a DNA plasmid comprising the genetic elements comprised in Figure 9 comprising the expression cassette of SEQ ID NO 23 which expression is controlled by a PGK promoter (SEQ ID NO 19) were centrifuged at 100 x g for 5 min at 4°C. The supernatant was discarded and the cell pellet was suspended in 15 ml Buffer IS. Cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 15 ml 2-8°C cold Buffer IS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 300 pl 2-8°C cold Buffer IS.

[301] 300 pl of the suspension was added to 60 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) that were as follows. Strep-Tactin® magnetic microbeads were washed 2 times with 1 ml Buffer IS on a magnet before use. SEQ ID NO 23 positive cells were separated on StrepMan magnet (I BA, Cat. No. 6-5650-065) as follows: After incubation for 30 min on the roller mixer at 2-8 °C, the cell bead mixture was transferred to a 15 ml centrifugation tube containing 10 ml cold Buffer IS. [302] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads (SEQ ID NO 23 positive cell fraction) were carefully washed off the tube wall by suspending in 10 ml cold Buffer IS. This procedure was repeated twice.

[303] Then the tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the

StrepMan Magnet.

[304] Positive cells on Strep-Tactin® magnetic microbeads were eluted in 5 ml cold D-biotin- Buffer IS working solution. The solution was prepared by mixing Buffer IS with a 100 mM D- biotin stock solution. The cell bead mixture with the D-biotin-Buffer IS working solution was mixed thoroughly by pipetting and incubated for 40 min at 2-8°C on a roller mixer. The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing a positive cell fraction was transferred to a fresh tube by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet. Steps of this paragraph were repeated 3 more times. Incubation times on the roller mixer at 2-8°C were 15 min, 30 min and 60 min for elution 2, elution 3 and elution 4.

[305] Elution 1 was performed with a D-biotin-Buffer IS working solution containing 1 mM

Biotin. Elution 2 was performed with a D-biotin-Buffer IS working solution containing 2 mM

Biotin. Elution 3 was performed with a D-biotin-Buffer IS working solution containing 3 mM

Biotin. Elution 4 was performed with a D-biotin-Buffer IS working solution containing 1 mM

Biotin.

[306] Cells from elution 2 and elution 3 were pooled and centrifuged at 100 x g for 5 min at 4°C. Then the Cells (named CHO-P-48h-l) were seeded at a cell density of 3.47 x 10⁵cells/ml in 1 ml fresh CHO-TF medium (37°C).

[307] Next day, cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in CHO-TF medium (Xell AG, Cat. No. 886-0001) containing 6 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). Cell concentration was adjusted to 5.6 x 10⁵ cells/ml. The cells were subcultured every 3 to 4 days in CHO-TF medium with 6 pg/ml puromycin. For subculturing, the cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in CHO-TF medium (Xell AG, Cat. No. 886-0001) containing 6 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). The cell concentration was adjusted to 6 x 10⁵ cells/ml at each subculturing. Cells were cultured at 37°C and 5% CO₂.

[308] 22 days after the magnetic microbead selection, cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in CHO-TF medium. Cells were subcultured every 3 to 4 days in CHO-TF medium without puromycin. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing.

Example 7.3- Second selection of puromycin pre-selected CHO cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using Strep-Tactin® magnetic microbeads

[309] 7.2 x 10⁶ cells of the CHO culture CHO-P-48h-l (from the section above, Example 7.2) were selected again 22 days after the first selection of the original culture. The selection was performed as described in Example 7.2 except of 30 pl magnetic microbead solution was used and incubation time on the roller mixer for all elution steps was 10 min.

[310] Cells from elution 3 and elution 4 were pooled and centrifuged at 100 x g for 5 min at 4°C. Then cells (named CHO-P-48h-ll) were seeded in fresh CHO-TF medium (37°C) without puromycin. Cells were further subcultured every 3-4 days in CHO-TF medium. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing. Cells were cultured at 37°C and 5% CO₂.

Results of selection in examples 7.1 to 7.3

[311] By use of the method of invention, the fraction of GFP positive cells was increased in all elution fractions compared to the fraction of GFP positive cells of the cultures before selection (Figure 24). The fraction of GFP positive cells in the culture CHO-P was 24% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 72% in elution 1 , 82% in elution 2 and 77% in elution 3. The fraction of GFP positive cells in the culture CHO-P-48h-l was 58% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 69% in elution 1 , 84% in elution 2, 97% in elution 3 and 99% in elution 4. The mean fluorescence intensity of the eluted populations was increased compared to the original population. An increase of the mean fluorescence intensity correlates with an increase of the elution fraction number (Figure 25). For CHO-P cells the mean fluorescence intensity was 28995 before selection with the method of invention. Mean fluorescence intensity was 34012 in elution 1 , 63740 in elution 2 and 124938 in elution 3. For CHO-P-48h-l cells the mean fluorescence intensity was 118067 before selection with the method of invention. Mean fluorescence intensity was 99150 in elution 1 , 115923 in elution 2, 142044 in elution 3 and 163622 in elution 4. [312] CHO cells that were selected using the method of invention and further cultivated afterwards, show an increased fraction of TST-flexible-CD4tm-GFP positive cells compared to their corresponding cultures prior to the selection with the method of invention. Fraction of GFP positive cells range from 43% to 58% in the CHO-P-48h-l culture over the last 24 days of cultivation. Cells selected from CHO-P-48-I (named CHO-P-48h-ll) using the method of invention comprise a GFP positive cell fraction from 87% to 94% over the last 20 days of cultivation (Figure 26). In addition, cultures selected using the methods of invention have an increased expression of the TST-flexible-CD4tm-GFP fusion protein during their cultivation compared to the corresponding cultures prior to selection with the method of invention. Cells selected with puromycin only (CHO-P) did not exceed a mean fluorescence intensity of 40527. CHO-P-48h-l exceeded a mean fluorescence intensity of 100000 in the same time. The mean fluorescence intensity of CHO-P-48h-l decreased in the absence of puromycin after day 24 and varied between 47122 and 71877 at the end of the cultivation. CHO-P-48h-ll exceeded a mean fluorescence intensity of 137347 after selection. The mean fluorescence intensity decreased to 82328 at the end of the cultivation. Another finding was a higher viability of CHO-P-48h-l compared to CHO-P. CHO-P culture viability decreased to 21% on day 17 while CHO-P-48-I showed a viability of 80% at this time (Figure 27).

Example 8 - Selection of HEK-293 cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using puromycin combined with the method of invention

[313] For this experiment a MEXi-293E cell suspension culture was transfected with a DNA plasmid comprising the genetic elements comprised in Figure 9 comprising the expression cassette of SEQ ID NO 23 which expression is controlled by a PGK promoter (SEQ ID NO 19). Cells of the culture were selected with different methods as described in the following examples of this section. Cells in the examples were analyzed by a Cytoflex flow cytometer (Beckman Coulter, Inc., Model No. A00-1-1102) to detect viable GFP positive cells and their mean fluorescence intensity.

Example 8.1- Selection of HEK-293 cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using puromycin

[314] 2 days after transfection, 6 x 10⁶ cells of a MEXi-293E suspension cell culture transfected with a DNA plasmid comprising the genetic elements comprised in Figure 9 comprising the expression cassette of SEQ ID NO 23 which expression is controlled by a PGK promoter (SEQ ID NO 19) were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in 10 ml MEXi-CM medium (IBA, Cat. No. 2-6010- 010) containing 1.7 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). The cells were subcultured every 3 to 4 days in MEXi-CM medium with 1 pg/ml puromycin (name MEXi-P). For subculturing, the cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in MEXi-CM medium containing 1 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). The cell concentration was adjusted to 6 x 10⁵ at each subculturing. Cells were cultured at 37°C and 5% CO₂.

Example 8.2- Selection of HEK-293 cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using puromycin and Strep-Tactin® magnetic microbeads

[315] 2 days after transfection, 2 x 10⁷ cells of a MEXi-293E suspension cell culture transfected with a DNA plasmid comprising the genetic elements comprised in Figure 9 comprising the expression cassette of SEQ ID NO 23 which expression is controlled by a PGK promoter (SEQ ID NO 19) were centrifuged at 100 x g for 5 min at 4°C. The supernatant was discarded and the cell pellet was suspended in 10 ml PBS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 5 ml 37°C warm Dissociation buffer (Gibco, Cat. No. 1851598). The mixture was incubated for 3 min at room temperature (20°C). Then 30 ml MEXi-TM (2-8°C) were added to the mixture. Cell aggregates were dispensed by pipetting and the suspension was centrifuged at 100 x g for 5 min at 4°C. The pellet was suspended in 300 pl MEXi-TM.

[316] 300 pl of the suspension was added to 60 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) that were prepared as follows. Strep-Tactin® magnetic microbeads were washed 2 times with 1 ml MEXi-TM on a magnet before use. SEQ ID NO 23 positive cells were separated on StrepMan magnet (I BA, Cat. No. 6-5650-065) as follows: After incubation for 23 min on the roller mixer at 2-8 °C, the cell microbeads mixture was transferred to a 15 ml centrifugation tube containing 10 ml cold MEXi-TM.

[317] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads (SEQ ID NO 23 positive cell fraction) were carefully washed off the tube wall by suspending in 10 ml cold MEXi-TM. This procedure was repeated twice.

[318] Then the tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet.

[319] Positive cells on Strep-Tactin® magnetic microbeads were eluted in 5 ml cold D-biotin- MEXi-CM working solution. The solution was prepared by mixing MEXi-CM with a 100 mM D- biotin stock solution to increase the biotin concentration of MEXi-CM by 1 mM. The cell bead mixture with the D-biotin-MEXi-CM working solution was mixed thoroughly by pipetting and incubated for 10 min at 2-8°C on a roller mixer. The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing a positive cell fraction was transferred to a fresh tube by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet. Steps of this paragraph were repeated 2 more times. Incubation times on the roller mixer were 23 min for elution 3.

[320] Cells from elution 2 and elution 3 were pooled and centrifuged at 100 x g for 5 min at 4°C. Then the Cells (named MEXi-P-48h-l) were seeded at a cell density of 4.7 x 10⁵ cells/ml in fresh MEXi-CM medium (37°C).

[321] 2 days after selection, cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in MEXi-CM medium (IBA, Cat. No. 2-6010-010) containing 1.0 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). Cell concentration was adjusted to 6 x 10⁵ cells/ml. The cells were subcultured every 3 to 4 days in MEXi-CM medium with 1 pg/ml puromycin. For subculturing, the cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in MEXi-CM medium (IBA GmbH) containing 1 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). The cell concentration was adjusted to 6 x 10⁵ cells/ml at each subculturing. Cells were cultured at 37°C and 5% CO₂.

[322] 19 days after the magnetic microbead selection, cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in MEXi-CM medium. Cells were subcultured every 3 to 4 days in MEXi-CM medium without puromycin. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing. Cells were cultured at 37°C and 5% CO₂.

Example 8.3 - Second selection of puromycin pre-selected HEK-293 cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using Strep-Tactin® magnetic microbeads

[323] 2 x 10⁷ cells of the MEXi-293E culture MEXi-P-48h-l (from para. [321]) were selected again 19 days after the first selection of the original culture. The selection was performed as described in Example 8.2 except of the following changes: The incubation time on the roller was 75 min during the incubation of the cells and the beads for binding of positive cell fraction. The incubation time on the roller was 20 min for elution 1 and 15 min for elution 2 and elution 3. The biotin concentration in the D-biotin-MEXi-CM working solution was increased by 2 mM in elution 2 and elution 3. [324] Cells from elution 2 and elution 3 were pooled and centrifuged at 100 x g for 5 min at 4°C. Then cells (named MEXi-P-48h-ll) were seeded at a cell density of 3 x 10⁵ cells/ml in 2.6 ml fresh MEXi-CM medium (37°C) without puromycin. Cells were further subcultured every 3-4 days in MEXi-CM medium. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing.

Example 8.4 - Third selection of HEK-293 cells displaying truncated CD4 surface protein fused to a Twin-Strep-tag® using Strep-Tactin® magnetic microbeads

[325] 1.22 x 10⁷ cells of the MEXi-293E culture MEXi-P-48h-ll (from para. [324]) were selected again 37 days after the first selection of the original culture. The selection was performed as described in Example 8.3 except of the following changes: The incubation time on the roller was 10 min for all elutions. The biotin concentration in the D-biotin-MEXi-CM working solution was increased by 2 mM in elution 2 and elution 3.

[326] Cells from elution 3 were centrifuged at 100 x g for 5 min at 4°C. Then cells (named MEXi-P-48h-ll I) were seeded at a cell density of 3 x 10⁵ cells/ml in 1. 5ml fresh MEXi-CM medium (37°C) without puromycin. Cells were further subcultured every 3-4 days in MEXi-CM medium. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing.

Example 8.5 - Selection of HEK-293 cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using large Strep-Tactin® multimer Agarose beads

[327] In this experiment Strep-Tactin® multimer (“mutant 2” of WO 2017/186669) was immobilized on CellThru agarose beads Resin (Sterogene Bioseparations Inc., average diameter of the beads: 200-300 pm).

[328] 900 pl of a 50% Strep-Tactin® multimer agarose bead suspension was washed 4 times with 5 ml MEXi-TM medium: The beads were settled by gravity at the bottom of a 15 ml centrifugation tube. Supernatant was removed by a pipette and beads were suspended in 5 ml MEXi-TM. After the wash process Beads were suspended in 900 pl MEXi-TM.

[329] 30 days after the first bead based selection of the original culture (MEXi-P), 2.00 - 2.20 x 10⁷ cells of the MEXi-293E cultures MEXi-P-48h-l and MEXi-P-48h-ll were centrifuged at 100 x g at 20°C for 5 min. The pellets were suspended in 10 ml MEXi-TM medium and the wash procedure was repeated one more time. Cells were centrifuged again at 100 x g at 20°C for 5 min and the pellet was suspended in 500 pl MEXi-TM. Cells were added to 300 pl of the Strep-Tactin® multimer agarose bead suspensions prepared in para. [328] and incubated at 2- 8°C on a roller mixer for at least 10 min. [330] The cell-bead mixture was washed 5 times with 1.5 ml MEXi-TM by gravity settling of the bead bound cells, aspiration of supernatant and suspending the bead bound cells in the medium. Cells were eluted by repeating this procedure 3 times with 1.5 ml D-biotin-MEXi-TM working solution. The solution was prepared by mixing MEXi-TM with a 100 mM D-biotin stock solution to adjust the biotin concentration of MEXi-TM to 1 mM.

[331] Multimeric Strep-Tactin® mutant 2 (WO 2017/186669) coated agarose beads can be used instead of Strep-Tactin® magnetic microbeads for selection of cells using the method of invention. The fraction of GFP positive cells was increased in all elution fractions compared to the fraction of GFP positive cells of the cultures before selection (Figure 29). The fraction of GFP positive cells in the culture MEXi-P-48h-l was 18% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 72% in elution 1 and 77% in elution 2. The fraction of GFP positive cells in the culture MEXi-P-48h-ll was 70% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 90% in elution 1 and 96% in elution 2.

[332] The mean fluorescence intensity of the eluted populations was increased compared to the original population. An increase of the mean fluorescence intensity correlates with an increase of the elution fraction number (Figure 30). For MEXi-P-48h-l cells the mean fluorescence intensity was 21250 before selection with the method of invention. Mean fluorescence intensity was 62874 in elution 1 and 86373 in elution 2. For MEXi-P-48h-ll cells the mean fluorescence intensity was 39303 before selection with the method of invention. Mean fluorescence intensity was 66457 in elution 1 and 75477 in elution 2.

Results of selection in examples 8.1 to 8.4

[333] By the use of the method of invention, the fraction of GFP positive cells was increased in all elution fractions compared to the fraction of GFP positive cells of the cultures before selection (Figure 31). The fraction of GFP positive cells in the culture MEXi-P was 70% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 95% in elution 1 , 98% in elution 2 and 98% in elution 3. The fraction of GFP positive cells in the culture MEXi-P-48h-l was 24% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 71 % in elution 1 , 79% in elution 2 and 71% in elution 3. The fraction of GFP positive cells in the culture MEXi-P-48h-ll was 70% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 94% in elution 1 , 99% in elution 2 and 99% in elution 3. The mean fluorescence intensity of the eluted populations was increased compared to the original population. An increase of the mean fluorescence intensity correlates with an increase of the elution fraction number (Figure 32). For MEXi-P cells the mean fluorescence intensity was 44563 before selection with the method of invention. Mean fluorescence intensity was 39509 in elution 1 , 76537 in elution 2 and 103296 in elution 3. For MEXi-P-48h-l cells the mean fluorescence intensity was 20043 before selection with the method of invention. Mean fluorescence intensity was 20543 in elution 1 , 30543 in elution 2 and 30956 in elution 3. For MEXi-P-48h-ll cells the mean fluorescence intensity was 50451 before selection with the method of invention. Mean fluorescence intensity was 49725 in elution 1 , 66454 in elution 2 and 93095 in elution 3.

[334] MEXi-293E cells that were selected using the method of invention and further cultivated afterwards, show an increased fraction of TST-flexible-CD4tm-GFP positive cells compared to the corresponding cultures prior to the selection with the method of invention (Figure 33). Fraction of GFP positive cells range from 18% to 26% in the MEXi-P-48h-l culture during the last 25 days of cultivation. Cells selected from MEXi-P-48-l (named MEXi-P-48h-ll) using the method of invention comprise a GFP positive cell fraction between 68% and 73% during the last 17 days of cultivation. Cells selected from MEXi-P-48-ll (named MEXi-P-48h-l II) using the method of invention comprise a GFP positive cell fraction between 94% and 99% during the last 7 days of cultivation. In addition, cultures that were selected using the method of invention have an increased expression of the TST-flexible-CD4tm-GFP fusion protein during their cultivation compared to their corresponding cultures prior to the selection with the method of invention (Figure 34). Cells selected with puromycin only (MEXi-P) showed a mean fluorescence intensity between 10646 and 20043 (value of 23240 on day 17 was an outlier due to a low number of GFP positive cells in the analysis) between day 5 and day 22. In contrast MEXi-P-48h-l showed a mean fluorescence intensity between 20324 and 25986 in the same time. The mean fluorescence intensity of MEXI-P-48h-l increased in the absence of puromycin after day 21 from 17083 to 25210 at the end of the cultivation. Mean fluorescence intensity of MEXi-P-48h-ll varied between 34890 and 50451 after selection. Mean fluorescence intensity of M EXi-P-48h-l 11 varied between 88568 and 109285 after selection.

Example 10- Comparison of experiments in examples 7 and 8

[335] A comparison of the results of the experiments of Examples 7 and 8 shows that the method of invention can be used to select high producing stable HEK-293 (MEXi-293E) and CHO (CHO-S) cells. The use of the method of invention increased the fraction of GFP positive cells in the selected elution fractions for both cell lines. If cells from the selection process were further cultivated, both HEK-293 and CHO cells, maintained a higher portion of GFP positive cells compared to the original culture. The same observation was made for both cell lines regarding the mean fluorescence intensity, representing the expression of TST-flexible-CD4tm- GFP, which was increased due to the selection using the method of invention. Thus it can be summarized that the method of invention is robust and efficient in increasing the portion of expressing cells and selecting cell populations that exhibit higher mean expression levels. Furthermore, cells selected with the method of invention showed higher expression levels and a larger portion of TST-flexible-CD4tm-GFP expressing cells compared to the already existing method which is based on a selection with an antibiotic agent (puromycin) only. Therefore, the method of invention provides a more efficient way to establish a stable cell culture with high producing cells.

Example 11 Selection of CHO cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using puromycin combined with the method of invention to increase the production of secreted alkaline phosphatase (SEAP)

[336] For this experiment CHO cell suspension cultures were transfected with a DNA plasmid either comprising the genetic elements comprised in Figure 11 (ZSG4) or Figure 10 (ZSG5) comprising the expression cassette of SEQ ID NO 27 controlled by an IRES (SEQ ID NO 25, ZSG4) or a PGK promoter (SEQ ID NO 26, ZSG5). Cells of the cultures were selected with different methods as described in the following examples of this section. Cells in the examples were analyzed by a Cytoflex flow cytometer (Beckman Coulter, Inc., Model No. A00-1-1102) to detect viable GFP positive cells and their mean fluorescence intensity. a) Selection of CHO cells displaying a truncated CD4 surface protein fused to a Twin- Strep-tag® using puromycin

[337] 2 days after transfection, 12 x 10⁶ cells of CHO suspension cell cultures transfected with a DNA plasmid either comprising the genetic elements comprised in Figure 11 (ZSG4) or Figure 10 (ZSG5) comprising the expression cassette of SEQ ID NO 27 controlled by an IRES (SEQ ID NO 25, ZSG4) or a PGK promoter (SEQ ID NO 26, ZSG5) were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in 20 ml CHO-TF medium (Xell AG, Cat. No. 886-0001) containing 7 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240) The cells were subcultured every 3 to 4 days in CHO-TF medium with 7 pg/ml puromycin. For subculturing, the cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in CHO-TF medium (Xell AG, Cat. No. 886- 0001) containing 7 pg/ml puromycin (Carl Roth GmbH Co. KG, Cat. No. 0240). The cell concentration was adjusted to 6 x 10⁵ at each subculturing. 21 days after transfection, cells were centrifuged at 100 x g for 5 min. The supernatant was discarded and the cell pellet was suspended in CHO-TF medium. Cells were subcultured every 3 to 4 days in CHO-TF medium without puromycin. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing. b) Selection of puromycin pre-selected CHO cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using Strep-Tactin® magnetic microbeads

[338] 18 days after transfection of the cells with an IRES and PGK controlled marker gene expression that were cultivated in the presence of puromycin, 2 x 10⁷ cells were centrifuged at 100 x g for 5 min at 4°C. The supernatant was discarded and the cell pellet was suspended in 10 ml cold Buffer IS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 10 ml cold Buffer IS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 300 pl cold Buffer IS.

[339] 300 pl of the suspension was added to 60 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) which were prepared as follows. Strep-Tactin® magnetic microbeads were washed 2 times with 1 ml Buffer IS on a magnet before use. TST-flexible-CD4tm-GFP positive cells were separated on StrepMan magnet (I BA, Cat. No. 6-5650-065) as follows: After incubation for 45 min on the roller mixer at 2-8°C, the cell-microbead-mixture was transferred to a 15 ml centrifugation tube containing 10 ml cold Buffer IS.

[340] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads (TST-flexible-CD4tm-GFP positive cell fraction) were carefully washed off the tube wall by suspending in 10 ml cold Buffer IS. This procedure was repeated twice.

[341] Then the tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the

StrepMan Magnet.

[342] Positive cells on Strep-Tactin® magnetic microbeads were eluted in 5 ml cold D-biotin- Buffer IS working solution. The solution was prepared by mixing Buffer IS with a 100 mM D- biotin stock solution. The cell bead mixture with the D-biotin-Buffer IS working solution was mixed thoroughly by pipetting and incubated for 10 min at 2-8°C on a roller mixer. The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing a positive cell fraction was transferred to a fresh tube by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet. The steps of this paragraph were repeated 3 more times. Incubation times on the roller mixer at 2-8°C were 10 min, 23 min and 20 min for elution 2, elution 3 and elution 4. [343] Elution 1 was performed with a D-biotin-Buffer IS working solution containing 1 mM

Biotin.

[344] Cells from elution 3 and elution 4 were pooled and centrifuged at 100 x g for 5 min at 4°C. Then the cells (CHO-ZSG4-P-I (IRES) and CHO-ZSG5-P-I (PGK)) were seeded at a cell density of 4 x 10⁵ cells/ml in fresh CHO-TF medium (37°C) without puromycin. Cells were subcultured every 3 to 4 days in CHO-TF medium without puromycin. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing. c) Second selection of CHO cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using Strep-Tactin® magnetic microbeads

[345] 2 x 10⁷ cells of the CHO cultures CHO-ZSG4-P-I and CHO-ZSG5-P-I (from step [344]) were selected again 10 days after the first selection of the original cultures CHO-ZSG4-P and CHO-ZSG5-P. The selection was performed as described in Example 11b) with the exception that incubation times were 19 min for elution 1 , 23 min for elution 2 and 10 min for elution 3. In addition, no fourth elution was conducted.

[346] Elutions were performed with a D-biotin-CHO-TF working solution. The solution was prepared by mixing CHO-TF medium with a 100 mM D-biotin stock solution to increase the biotin concentration of CHO-TF by 3 mM biotin (elution 1), 5 mM biotin (elution 2) and 0.55 mM biotin (elution 3).

[347] Cells from elution 2 and elution 3 were centrifuged at 100 x g for 5 min at 4°C. Then the cells (CHO-ZSG4-P-II-E2, CHO-ZSG5-P-II-E2 and CHO-ZSG5-P-II-E3) were seeded at a cell density of 4 x 10⁵ cells/ml in fresh CHO-TF medium (37°C) without puromycin. CHO-ZSG4-P-II- E3 cells were seeded at a cell density of 2.7 x 10⁵ cells/ml in fresh CHO-TF medium (37°C) without puromycin. Cells were subcultured every 3 to 4 days in CHO-TF medium without puromycin. The cell concentration was adjusted to 3-6 x 10⁵ cells/ml at each subculturing. d) Results

[348] By the use of the method of invention, the fraction of GFP positive cells was increased in all elution fractions compared to the fraction of GFP positive cells of the cultures before selection (Figure 35). The fraction of GFP positive cells in the culture CHO-ZSG4-P was 72% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 99% in elution 1 , elution 2, elution 3 and elution 4. The fraction of GFP positive cells in the culture CHO-ZSG5-P was 73% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 97% in elution 1 and 99% in elution 2, elution 3 and elution 4. The fraction of GFP positive cells in the culture CHO-ZSG4-P-I was 95% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 99% in elution 1, elution 2 and elution 3. The fraction of GFP positive cells in the culture CHO- ZSG5-P-I was 92% before selection. After the selection with the method of invention, the fraction of GFP positive cells was 99% in elution 1 and 100% in elution 2 and elution 3.

[349] The mean fluorescence intensity of the selected populations was increased compared to the original population during the selection process. An increase of the mean fluorescence intensity correlates with an increase of the elution fraction number (Figure 36). For CHO-ZSG4- P cells the mean fluorescence intensity was 53357 before selection with the method of invention. Mean fluorescence intensity was 49220 in elution 1, 60032 in elution 2, 70474 in elution 3 and 74834 in elution 4. For CHO-ZSG5-P cells the mean fluorescence intensity was 40547 before selection with the method of invention. Mean fluorescence intensity was 31517 in elution 1 , 40559 in elution 2, 47929 in elution 3 and 52186 in elution 4. For CHO-ZSG4-P-I cells the mean fluorescence intensity was 39140 before selection with the method of invention. Mean fluorescence intensity was 50570 in elution 1, 60128 in elution 2 and 75828 in elution 3. For CHO-ZSG5-P-I cells the mean fluorescence intensity was 37004 before selection with the method of invention. Mean fluorescence intensity was 45826 in elution 1, 52478 in elution 2, 60915 in elution 3.

[350] CHO cells that were selected using the method of invention and further cultivated afterwards, show an increased fraction of TST-flexible-CD4tm-GFP positive cells compared to their corresponding cultures prior to the selection with the method of invention. The fraction of GFP positive cells range from 67% to 76% in the CHO-ZSG4-P culture over the last 21 days of cultivation. Cells selected from CHO-ZSG4-P (named CHO-ZSG4-P-I) using the method of invention comprise a GFP positive cell fraction between 91 % and 96% in a period of 10 days. Cells selected from CHO-ZSG4-P-I (named CHO-ZSG4-P-II-E2 and CHO-ZSG4-P-II-E3) using the method of invention comprise a GFP positive cell fraction between 96% and 99% (CHO- ZSG4-P-II-E2) or between 91 % and 99% (CHO-ZSG4-P-II-E3) in a period of 7 days (Figure 37).

Fraction of GFP positive cells decrease from 75% to 48% in the CHO-ZSG5-P culture over the last 25 days of cultivation. Fraction of GFP positive cells decrease from 95% to 82% in the CHO-ZSG5-P-I culture, which was derived from CHO-ZSG5-P by using the method of invention, in 22 days. Cells selected from CHO-ZSG5-P-I (named CHO-ZSG5-P-II-E2 and CHO-ZSG5-P- II-E3) using the method of invention comprise a GFP positive cell fraction between 95% and 100% (CHO-ZSG5-P-II-E2) or between 81 % and 100% (CHO-ZSG4-P-II-E3) in a period of 7 days or 70 days (Figure 39). [351] In addition, cultures that were selected using the method of invention have an increased expression of the TST-flexible-CD4tm-GFP fusion protein during their cultivation in the absence of puromycin compared to their corresponding cultures prior to the selection with the method of invention. Cells selected with puromycin only (CHO-ZSG4-P and CHO-ZSG5-P) show a mean fluorescence intensity between 23839 and 28444 (CHO-ZSG4-P) or 14962 and 22422 (CHO- ZSG5-P) after the removal of puromycin from the cultivation (Figure 38 and Figure 40). The Mean fluorescence intensity of CHO-ZSG4-P-I varied between 31783 and 37638. CHO-ZSG4- P-II-E2 mean fluorescence intensity varied between 44584 and 60128 and CHO-ZSG4-P-II-E3 mean fluorescence intensity varied between 57236 and 75828 (Figure 38). The Mean fluorescence intensity of CHO-ZSG5-P-I varied between 30085 and 50589. CHO-ZSG5-P-II-E2 mean fluorescence intensity varied between 48308 and 57899 and CHO-ZSG5-P-II-E3 mean fluorescence intensity varied between 53217 and 82905 (Figure 40).

[352] Both tested strategies for the expression of the marker protein (IRES and PGK controlled) were compatible with the method of invention. The use of the method of invention increased the fraction of cells expressing the marker protein in case the marker protein expression was controlled by an IRES or by a PGK promoter. Furthermore, the marker protein expression was increased by consecutive selection using the method of invention for IRES and PGK controlled expression. Compared to cells that were selected with puromycin only, the cells selected using the method of invention exhibit higher marker gene expression levels and a larger fraction of marker gene expressing cells over longer times of cultivation. Thus, the method of invention outperformed the traditional method of antibiotic selection. e) Determination of secreted alkaline phosphatase (SEAP) expression of CHO cultures that were cultured for 4 days

[353] The method of invention can be used to select cells via a marker gene that codes for a transmembrane domain of a protein fused to a Twin-Strep-tag® that is presented on the surface of cells expressing a protein of interest. Here SEAP fused to a C-terminal Twin-Strep-tag® (SEQ ID NO 28) was expressed as protein of interest (POI). The example describes the determination of the SEAP productivity in cells from cultures described in Example 11 31 days after transfection of the CHO cultures. The following cultures from the examples in Example 11 are included into this experiment: CHO-ZSG4-P, CHO-ZSG4-P-I, CHO-ZSG4-P-II-E2, CHO- ZSG4-P-II-E3, CHO-ZSG5-P, CHO-ZSG5-P-I, CHO-ZSG5-P-II-E2, CHO-ZSG5-P-II-E3.

[354] The cells were centrifuged at 100 x g for 5 min at room temperature and suspended at 3 x 10⁵ cells/ml in fresh CHO-TF medium 31 days after the transfection of the initial CHO culture. Cells were cultivated at 37°C and 5% CO₂ in 50 ml TubeSpin® Bioreaktor 50 tubes (TPP, Cat No. 87050) at 300 rpm. After 4 days, cell culture supernatant was harvested by centrifugation of cells at 100 x g for 5 min. 5 l BioLock solution (IBA GmbH, Cat. No. 2-0205- 050) and 100 pl 10x Buffer W (IBA GmbH, Cat. No. 2-1003-100) were added per 1 ml supernatant. The supernatant was centrifuged again for 20 min at 3345 x g at 4°C and supernatant was purified via 200 pl Strep-Tactin®XT Superflow high capacity gravity columns (IBA GmbH, Germany, Cat. No. 2-4031-001) according to the manufactures protocol. Eluted protein was determined by photometric measurement at 280 nm in a Nanodrop 2000 (Thermo Scientifc, Cat. No, ND-2000). In Addition, the SEAP concentration in the supernatant was measured by a BLItz system (Sartorius AG) equipped with Strep-Tactin®XT coated biosensors. The sensors were prepared by coupling of Strep-Tactin®XT (IBA GbmH, Cat. No. 2-4202-001) on Amine Reactive (AR2G) sensors (Sartorius AG, Cat. No. 18-5092) using the Amine Reactive 2nd Generation (AR2G) Reagent Kit (Sartorius AG, Cat. No. 18-5095).

[355] Analysis of the cell culture supernatant with the BLItz system showed that cultures where the marker cassette expression was controlled by a PGK promoter exhibit higher protein concentrations compared to cultures where the marker cassette was controlled by an IRES (Figure 41). Furthermore, ZSG5 cultures that were selected using the method of invention (CHO-ZSG5-P-I, CHO-ZSG5-P-II-E2, CHO-ZSG5-P-II-E3) exhibit higher SEAP-TST concentrations in the supernatant and thus higher SEAP-TST expressions than the culture that was selected only with puromycin (CHO-ZSG5-P). In addition, the CHO-ZSG5 cultures that were selected a second time using the method of invention (CHO-ZSG5-P-II) exhibit higher SEAP-TST concentrations in the supernatant than CHO-ZSG5-P-I that was selected with the method of invention just one time.

[356] Determination of the protein concentration in the elution of the purified samples was only possible for CHO-ZSG5-P-I and CHO-ZSG5-P-II cultures. In the elutions of the other samples, the measured absorption signal at 280 nm was below the quantification limit. The results from the purification agree with the finding from the BLItz analysis that CHO-ZSG5-P-I and CHO- ZSG5-P-II cultures exhibit the highest protein concentrations in the experiment and therefore that the method of invention increases the protein expression compared to antibiotic selected cultures. f) Determination of secreted alkaline phosphatase (SEAP) expression of CHO cultures cultured for 7 days

[357] The experiment described in the section above was repeated with the cultures CHO- ZSG5-P, CHO-ZSG5-P-I and CHO-ZSG5-P-II-E3 38 days after the transfection of the CHO cultures. The cell density at the time of inoculation was adjusted to 3.8 to 4.8 x 10⁵ cells/ml and the cells were cultivated for 7 days before the cell culture supernatant was harvested. Instead of cultivation in 50ml TubeSpin® Bioreaktor 50, cells were cultured in 250 ml flasks at 37°C, 5% CO₂ and 125 rpm.

[358] The use of the method of invention increased the volumetric SEAP-TST (SEQ ID NO 28) yield in the selected cultures (Figure 42). The selection of CHO-ZSG5-P using the method of invention increased the volumetric yield by factor 2. Selection of CHO-ZSG5-P-I using the method of invention increased the volumetric SEAP-TST yield by factor 1.6 compared to CHO- ZSG5-P-I and by factor 3.3 compared to CHO-ZSG5-P.

Example 12 - Selection of CHO cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using only the method of invention to increase the production of secreted alkaline phosphatase (SEAP)

[359] For this experiment CHO cell suspension cultures were transfected with a DNA plasmid either comprising the genetic elements comprised in Figure 11 (ZSG4) or Figure 10 (ZSG5) comprising the expression cassette of SEQ ID NO 27 controlled by an IRES (SEQ ID NO 25, ZSG4) or a PGK promoter (SEQ ID NO 26, ZSG5). Two days post transfection the cells were selected using the method of invention as described in Example 11b with following exceptions: 4.1 x 10⁷ cells were used for selection, elution 3 was incubated for 23 min and a fourth elution step with 1 mM biotin in Buffer IS and 20 min incubation time was performed. Elution fraction 2 elution fraction 3 and elution fraction 4 were pooled and cells were seeded to 4 x 10⁵ cells/ml in CHO-TF Medium (name: CHO-ZSG4-woP-l and CHO-ZSG5-woP-l). Cells were cultivated at 37°C, 5% CO₂ in a shaker incubator.

[360] Five days after the first selection a second selection using the method of invention was performed as described in para. [359] with following exceptions: 1 ,76 x 10⁷ (CHO-ZSG5-woP-l and 6.9 x 10⁶ (CHO-ZSG4-woP-l) cells were used for selection. CHO-ZSG5-woP-l cells were incubated with 60 pl Microbeads and CHO-ZSG4-woP-l cells were incubated with 30 pl microbeads. Only two elution steps were performed: Elution 1 with 2 mM biotin in Buffer IS and an incubation for 13 min (E1) and elution 2 with 6 mM biotin in Buffer IS for 45 min (E2). Cells in elution fraction 2 were seeded to 3 x 10⁴ cells/ml (CHO-ZSG4-woP-ll) or 1.2 x 10⁵ cells/ml (CHO-ZSG5-woP-ll) in CHO-TF Medium.

[361] 14 days after the second selection a third selection using the method of invention was performed as described in para. [359] with following exceptions: 2 x 10⁷ (CHO-ZSG5-woP-ll and CHO-ZSG4-woP-ll) cells were used for selection. CHO-ZSG5-woP-ll cells were incubated with 60 pl microbeads and CHO-ZSG4-woP-ll cells were incubated with 30 pl Microbeads. Three elution steps were performed: Elution 1 with 3 mM biotin in Buffer IS and an incubation for 15 min (E1), elution 2 with 5 mM biotin in Buffer IS for 17 min (E2) and elution 3 with 1 mM biotin in Buffer IS for 17 min (E3). CHO-ZSG5-woP-PII cells in elution fraction 3 were seeded at 2.5 x 10⁵ cells/ml (named CHO-ZSG5-woP-lll) in CHO-TF Medium. The amount of eluted cells from sample CHO-ZSG4-woP-ll was too low for further cultivation.

[362] 35 days after the third selection a fourth selection using the method of invention was performed as described in para. [359] with following exceptions: 2 x 10⁷ (CHO-ZSG5-woP-lll cells were used for selection. CHO-ZSG5-woP-lll cells were incubated with 60 pl microbeads for 75 min. Three elution steps were performed: Elution 1 with 3 mM biotin in Buffer IS and an incubation for 30 min (E1), elution 2 with 5 mM biotin in Buffer IS for 16 min (E2) and elution 3 with 1 mM biotin in Buffer IS for 16 min (E3). CHO-ZSG5-woP-PIII cells in elution fraction 3 were seeded at 3 x 10⁵ cells/ml (named CHO-ZSG5-woP-IV) in CHO-TF Medium.

[363] The experiment described in Example 11f was performed with the culture CHO-ZSG5- woP-lll 12 days after the selection. The cell density at the time of inoculation was adjusted to 4.0 x 10⁵ cells/ml and the cells were cultivated for 7 days before the cell culture supernatant was harvested.

[364] As shown in the previous examples, the use of the method of invention increased the fraction of GFP positive cells as well as the mean fluorescence intensity of the cells and thus the expression of the marker gene cassette (Figure 43 and Figure 44) for ZSG4 and ZSG5 cultures. In all selections the mean fluorescence intensity of the cells increased with higher elution fraction and exceeded the mean fluorescence intensity of the original population.

[365] The use of the method of invention only, without a pre-selection process with puromycin, resulted in stable expressing cells. The mean fluorescence intensity and the fraction of GFP positive cells of the selected ZSG5 are stable after the third selection but the levels of both indicators are increased by the fourth selection (Figure 45 and Figure 46). The SEAP-TST protein expression in the CHO-ZSG5-woP-lll culture was 12.4 pg/ml and thus in the range of the SEAP-TST expression of CHO-ZSG5-P-II-E3 which was pre-selected with puromycin (Figure 47). Therefore, the method of invention can be used alone to select cells with a productivity which is comparable of the productivity from cells selected with the method of invention and puromycin. In addition, the use of the method of invention alone allows the selection of a culture with a three-fold higher expression compared to a standard antibiotic selection (Figure 47). Furthermore, the mean Fl of the culture selected only with the method of invention was comparable with the mean Fl of the puromycin selected culture CHO-ZSG5-P. Therefore, the ratio of expressed SEAP-TST protein to marker protein was higher in CHO- ZSG5-woP-lll cells than on CHO-ZSG5-P cells. This demonstrates that the method of invention favors cells with a high POI expression while the antibiotic selection favors the selection with a high marker expression but a lower expression of protein of interest. Thus, cells selected with the method of invention only use their resources more efficient regarding the expression of the protein of interest. Obviously, this effect is given to the fact that the method of invention exposes the cells to the selection pressure only for a short time during the binding and elution of the cells form the beads. In contrast an antibiotic selection exposes the cells to a continuous selection pressure over several days.

Example 13 - Selection of HEK-293 cells displaying a truncated CD4 surface protein fused to a Twin-Strep-tag® using only the method of invention to increase the production of secreted alkaline phosphatase (SEAP)

[366] 6 x 10⁷ cells of a MEXi-293E suspension cell culture transfected with a DNA plasmid either comprising the genetic elements comprised in Figure 11 (ZSG4) or Figure 10 (ZSG5) comprising the expression cassette of SEQ ID NO 27 controlled by an IRES (SEQ ID NO 25, ZSG4) or a PGK promoter (SEQ ID NO 26, ZSG5) were centrifuged at 100 x g for 5 min at room temperature (20°C) two days post transfection. The supernatant was discarded and the cell pellet was suspended in 10 ml PBS. The cells were centrifuged again at 100 x g for 5 min at 20°C. The pellet was suspended in 3 ml 37°C warm Dissociation buffer (Gibco, Cat. No. 1851598). The mixture was incubated for 10 min at 37°C. Then 30 ml MEXi-TM (2-8°C) were added to the mixture. Cell aggregates were dispensed by pipetting and the suspension was centrifuged at 100 x g for 5 min at 4°C.

[367] The cells were mixed with 500 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6- 5510-050) which were prepared as follows. Strep-Tactin® magnetic microbeads were washed 2 times with 1 ml MEXi-TM on a magnet before use. TST-flexible-CD4tm-GFP positive cells were separated on StrepMan magnet (I BA, Cat. No. 6-5650-065) as follows: After incubation for 47 min on the roller mixer at 2-8 °C, the cell-microbead-mixture was transferred to a 15 ml centrifugation tube containing 10 ml cold MEXi-TM.

[368] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads were carefully washed off the tube wall by suspending in 10 ml cold MEXi-TM. This procedure was repeated twice.

[369] Then the tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet. [370] Positive cells on Strep-Tactin® magnetic microbeads were eluted in 5 ml cold D-biotin- MEXi-CM working solution. The solution was prepared by mixing MEXi-CM with a 100 mM D- biotin stock solution to increase the biotin concentration of MEXi-CM by 1 mM.

[371] The cell bead mixture with the D-biotin-MEXi-CM working solution was mixed thoroughly by pipetting and incubated for 37 min at 2-8°C on a roller mixer. The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing a positive cell fraction was transferred to fresh tube by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet. Steps of this paragraph were repeated 3 more times except of omitting the incubation on the roller mixer. Instead, the further elution steps were performed immediately after mixing the cell-bead mixture with D-biotin-MEXi- CM working solution.

[372] Cells from elution 1 , 2, 3 and 4 were pooled and centrifuged at 100 x g for 5 min at 4°C. Then the cells (named MEXi-ZSG4-woP-l and MEXi-ZSG5-woP-l) were seeded at a cell density of 6 x 10⁵ cells/ml in fresh MEXi-CM medium (37°C).

[373] Three days after the first selection a second selection using the method of invention was performed. 7.6 x 10⁷ (MEXi-ZSG5-woP-l and MEXi-ZSG4-woP-l) cells were centrifuged at 100 x g for 5 min at room temperature (20°C). The supernatant was discarded and the cell pellet was suspended in 10 ml 2-8°C cold MEXi-TM. The cells were centrifuged again at 100 x g for 5 min at 20°C. The pellet was suspended 20 ml 2-8°C cold MEXi-TM. The suspension was centrifuged at 100 x g for 5 min at 4°C and the selection was performed as described in paras. [367] to [372], with following exceptions: Cells were incubated with Strep-Tactin® magnetic Microbeads for 26 min on a roller mixer at 2-8°C. Elution 1 was incubated 17 min on a roller mixer. Cells from elution 1 , 2, 3 and 4 were pooled and centrifuged at 100 x g for 5 min at 4°C. Then the cells (named MEXi-ZSG4-woP-ll and MEXi-ZSG5-woP-ll) were seeded at a cell density of 4 x 10⁵ cells/ml in fresh MEXi-CM medium (37°C).

[374] Four days after the second selection a third selection using the method of invention was performed as described in para. [373] with following exceptions: 1.79 x 10⁷ cells (MEXi-ZSG4- woP-ll) or 2.1 x 10⁷ cells (MEXi-ZSG5-woP-ll) were used for selection. The cells were mixed with 300 pl MEXi-TM before they were incubated with 140 pl Strep-Tactin® magnetic microbeads on the roller mixer. Cells from elution 1 , 2, 3 and 4 (MEXi-ZSG4-woP-ll) or 1 , 2, and 3 respectively (MEXi-ZSG5-woP-ll) were pooled and centrifuged at 100 x g for 5 min at 4°C. The cells (named MEXi-ZSG4-woP-lll and MEXi-ZSG5-woP-lll) were seeded at a cell density of 4 x 10⁵ cells/ml in fresh MEXi-CM medium (37°C). [375] The cells were cultured for 4 days in 50 ml TubeSpin® Bioreaktor 50 tubes (TPP, Cat No. 87050) at 300 rpm (throw of shaker: 10 mm), 37°C and 5% CO₂. On day 4 a sample of each culture was centrifuged at 100 x g for 5 min at 25°C. 80 pl 10x Buffer W and 4 pl BioLock solution (IBA GmbH, Cat. No. 2-205-050) were added to 800 pl supernatant. Samples were centrifuged again at 10.000 x g for 10 min and the SEAP concentration in the supernatant was measured by a BLItz system (Sartorius AG) equipped with Strep-Tactin®XT coated biosensors. The sensors were prepared by coupling of Strep-Tactin®XT (IBA GbmH, Cat. No. 2-4202-001) on Amine Reactive (AR2G) sensors (Sartorius AG, Cat. No. 18-5092) using the Amine Reactive 2nd Generation (AR2G) Reagent Kit (Sartorius AG, Cat. No. 18-5095).

[376] Analysis of the cell culture supernatant with the BLItz system showed that the SEAP- TST (SEQ ID NO 28) expression was higher if the marker cassette expression was controlled by a PGK promoter (MEXi-ZSG5-woP-lll) compared to the culture where the marker cassette was controlled by an IRES (MEXi-ZSG4-woP-lll) (Figure 48).

[377] SEAP Twin-Strep-tag® fusion protein expression was also measured by purification of cell culture supernatant. The cells (MEXi-ZSG4-woP-lll and MEXi-ZSG5-woP-lll) were centrifuged at 100 x g for 5 min at room temperature and suspended at 3 x 10⁵ cells/ml in fresh MEXi-CM medium. Cells were cultivated at 37°C and 5% CO₂ in 250 ml shake flasks at 125 rpm. After 8 days, cell culture supernatant was harvested by centrifugation of cells at 100 x g for 5 min. The supernatant was centrifuged again for 20 min at 3345 x g at 4°C and supernatant was purified via 200 pl Strep-Tactin®XT Superflow high capacity gravity columns (IBA GmbH, Germany, Catn No. 2-4031-001) according to the manufactures protocol. Eluted protein was determined by photometric measurement at 280 nm in a Nanodrop 2000 (Thermo Scientifc, Cat. No, ND-2000).

[378] The volumetric SEAP-TST protein expression in the MEXi-ZSG5-woP-lll culture was 1.6 fold higher (8.2 pg/ml) than the expression of MEXi-ZSG4-woP-lll (5.0 pg/ml).

[379] As shown in previous Examples, the use of the method of invention increased the fraction of GFP positive cells as well as the mean fluorescence intensity of the cells and thus the expression of the marker gene cassette for ZSG4 and ZSG5 cultures (Figure 49 and Figure 50). In all selections the mean fluorescence intensity of the cells increased with higher elution fraction and exceeded the mean fluorescence intensity of the original population. The use of the method of invention only, without a pre-selection process with puromycin, resulted in stable expressing cells. The mean fluorescence intensity and the fraction of GFP positive cells of the selected ZSG5 are stabilized after the third selection (Figure 51 , Figure 52). The selection with the method of invention did not influenced the cell cultures viability negatively (Figure 53).

Example 14 - Optimization of cell concentration during binding step of cells to Strep- Tactin® magnetic microbeads

[380] 8 x 10⁵ CHO-ZSG5-P-II-E3 cells were centrifuged at 100 x g for 5 min at 4°C. The supernatant was discarded and the cell pellet was suspended in 3 ml cold Buffer IS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 3 ml cold Buffer IS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 80 pl cold Buffer IS.

[381] Cells were seeded together with 10 pl of Strep-Tactin® magnetic microbeads to a final concentration of 1 x 10⁵ cells/ml, 1 x 10⁶ cells/ml and 1 x 10⁷ cells/ml in 1.5 ml reaction tubes. Each tube contained 1 x 10⁷ cells. Magnetic microbeads were prepared as follows. Strep- Tactin® magnetic microbeads were washed 2 times with 1 ml Buffer IS on a magnet before use. Cells were incubated on a roller mixer at 2-8°C. One tube with a cell concentration of 1 x 10⁵ cells/ml and one tube with a cell concentration of 1 x 10⁷ cells/ml were incubated for 10 min. One tube with a cell concentration of 1 x 10⁵ cells/ml and one tube with a cell concentration of 1 x 10⁷ cells/ml were incubated for 30 min. One tube with a cell concentration of 1 x 10⁶ cells/ml was incubated for 20 min. After incubation on the roller mixer, the tubes were placed for at least 3 min on a Magnetic Separator (IBA, Cat. No. 2-1602-000). Supernatants containing the unbound cells were removed and transferred to new reaction tubes and analyzed on Cytoflex flow cytometer.

[382] The higher the cell concentration was adjusted the lower was the portion of unbound cells found in the flow through. At a cell concentration, of 1 x 10⁷ cells/ml only 2.8% to 5.6% of the cells in the sample were found in the supernatant. Therefore, cell concentration can be adjusted to at least 1 x 10⁷ cells/ml for optimal binding conditions. The incubation time had no significant effect to the portion of bound cell. Thus, 10 min incubation is sufficient for effective cell binding (Figure 54). Because it is not possible to keep the same total volume in all experiments if the same number of total cells per tube is kept constant while changing the cell concentration, the volumes in the experiments were different between the different cell concentrations. Therefore, experiments with 1 x 10⁷ cells/ml exhibit the lowest total volume and consequently the highest microbead concentration. Hence, the combination of high microbead and cell concentration contribute to the optimal binding conditions. Example 15 - Optimization of the incubation time in the elution steps

[383] 1 x 10⁷ CHO-ZSG5-P-II-E3 cells were selected using the method of invention as described in Example 11 with following exceptions: After the wash procedure the cells were suspended in 3.5 ml of CHO-TF medium containing additional 10 pM biotin. The suspension was distributed to 6 reaction tubes with 0.5 ml suspension per tube. All tubes were incubated on a roller mixer at 2 - 8°C for 30 min. The tubes were placed for at least 3 min on a Magnetic Separator (IBA, Cat. No. 2-1602-000).

[384] Supernatant containing a positive cell fraction was transferred to fresh tube by careful pipetting while keeping the reaction tube on the Magnetic Separator. The remaining cell bead mixture was suspended in 0.5 ml elution medium. 3 of the 6 samples were mixed briefly and placed immediately on the Magnetic Separator again. For these three samples the procedure of this paragarph was repeated without incubation between the elution steps until all cells were detached from the beads.

[385] The other three tubes were incubated for 10 min (elution 2), 40 min (elution 3) and 10 min (elution 4) between the elution steps.

[386] The eluted cells were analyzed on Cytoflex flow cytometer. No significant difference was observed between the elution of samples with longer incubation times between the elution steps and samples that were placed immediately after mixing on the Magnetic separator (Figure 55 and Figure 56). The amount of eluted cells in the elution fractions and the mean fluorescence intensity of GFP positive cells showed the same trend in both experiments. Most cells were eluted in elution fraction 2 and the mean fluorescence intensity of the cells increased with higher elution fraction.

Example 16 - Influence of magnet incubation on the elution profile during magnetic separation

[387] 4.59 x 10⁷ MEXI-293E cells of a MEXi-293E suspension cell culture transfected with a DNA plasmid comprising the genetic elements comprised in Figure 10 comprising the expression cassette of SEQ ID NO 27 which expression is controlled by a PGK promoter (SEQ ID NO 26) were selected using the method of invention as described in Example 13, paras. [366] to [368] with following exceptions: The cells were mixed with 100 pl Strep-Tactin® magnetic microbeads (IBA, Cat. No. 6-5510-050) and incubated for 20 min on a roller mixer.

[388] After the wash procedure cells were suspended in 3.1 ml MEXi-TM medium and 300 pl of the suspension were distributed to each of six 2 ml reaction tubes. The tubes were placed for at least 3 min on a Magnetic Separator (IBA, Cat. No. 2-1602-000). [389] Supernatants containing the negative cell fraction were removed by careful pipetting while keeping the reaction tubes on the Magnetic Separator. Then reaction tubes were removed from the Magnetic Separator.

[390] Three MEXi-TM working solutions with D-biotin concentrations of 1 pM, 10 pM and 1000 pM were prepared. The solutions were prepared by mixing MEXi-TM with a 100 mM D- biotin stock. Cells in two tubes each were suspended with 500 pl of 1 pM, 10 pM or 1000 pM biotin MEXi-TM working solution. For the following elutions the tubes of each biotin concentration were separated into two groups.

[391] Elution in group 1 was conducted as follows: After 22 min incubation on a roller mixer at 2-8°C the tubes were placed in the Magnetic Separator and the supernatant with eluted cells was transferred to new tube (elution fraction 1, E1). The cell-magnet bead mixture was suspended immediately with 500 pl of the appropriate MEXi-TM working solution (1 pM, 10 pM or 1000 pM) on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s. The procedure of this paragraph was repeated three times without incubation on the roller mixer between the elutions.

[392] Then, tubes were placed in the Magnetic Separator and the supernatant with unbound cells was transferred to new tube (E5). The cell-magnet bead mixture in tubes that were eluted with 1 pM or 10 pM MEXi-TM working solution were suspended immediately with 500 pl of 1000 pM MEXi-TM working solution on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s.

[393] Then, tubes were placed in the Magnetic Separator and the supernatant with eluted cells was transferred to new tube (E6).

[394] Elution fractions 1 to 6 (E1-E6) were analyzed on a Cytoflex flow cytometer.

[395] Elution in group 2 was conducted as follows: After a 9 min incubation on a roller mixer at 2-8°C the tubes were removed from the roller mixer.

[396] Tubes were placed for 3 min in the Magnetic Separator without removal of supernatant. Then the tubes with the cell-magnet bead mixture were suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s.

[397] The procedure of paras. [396] was repeated three times.

[398] Then the tubes with the cell-bead mixture were incubated for further 2 min on the roller mixer. [399] The procedure of paras. [396]was repeated.

[400] Then the tubes with the cell-bead mixture were incubated for further 4 min on the roller mixer.

[401] The tubes were placed in the Magnetic Separator and the supernatant with eluted cells was transferred to new tube (elution fraction 1, E1). The cell-magnet bead mixture was suspended immediately with 500 pl of the appropriate MEXi-TM working solution (1 pM, 10 pM or 1000 pM).

[402] Then the tubes with the cell-magnet bead mixture were suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s. Tubes were placed for 3 min in the Magnetic Separator without removal of supernatant. The procedure of this paragraph was repeated three times.

[403] The elution process described in paras. [401] to [402] was repeated three times to get elution fraction E2, E3 and E4 in para. [401],

[404] Then tubes were placed in the Magnetic Separator and the supernatant with unbound cells was transferred to new tube (E5). The cell-magnet bead mixture in tubes that were eluted with 1 pM or 10 pM MEXi-TM working solution were suspended immediately with 500 pl of 1000 pM MEXi-TM working solution.

[405] Then the tubes with the cell-magnet bead mixture were suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s.

[406] Tubes were placed for 3 min in the Magnetic Separator without removal of supernatant.

[407] The procedure in paras. [405] and [406] was repeated three times.

[408] Then tubes were placed in the Magnetic Separator and the supernatant with eluted cells was transferred to new tube (E6).

[409] Elution fractions 1 to 6 (E1-E6) were analyzed on a Cytoflex flow cytometer.

[410] The method of several repeated incubations on the magnet followed by re-suspension of the cells before transferring of the supernatant of an elution fraction changed the elution profile of the cells significantly. If 1 pM biotin was used for elution, repeated magnet incubations increased the number of eluted cells in each elution fraction (Figure 58) compared to the method with only one incubation on the magnet before collection of eluted cells (Figure 57). [411] If biotin concentrations of 10 pM or 1000 pM were used, the majority of cells were eluted in the first elution fraction and the amount of eluted cells decreased with every consecutive elution fraction in case the method of repeated magnet incubation was used (Figure 60 and Figure 62). In contrast, the number of eluted cells increased from elution fraction 1 to elution fraction 3 and decreased after the third elution with every consecutive elution if the sample was incubated only once on the magnet before each elution step (Figure 59 and Figure 61).

The total amount of eluted cells in the first five elution fractions at a specific biotin concentration was always higher for the method of repeated magnet incubations. In addition, the total number of eluted cells in the first five elution fractions increased at higher biotin concentrations.

[412] The mean fluorescence intensity of the cells in elution fractions at a biotin concentration of 1 pM increased only between by 28% from elution fraction 1 to elution fraction 5. In contrast the Fl increased between 205% and 219% from elution fraction 1 to elution fraction 5 in experiments with 10 pM or 1000 pM biotin (Figure 63). To further analyze the elution profile, eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<20,000). GFP++ corresponds to cells with medium mean Fl (20,000 - 40,000). GFP+++ corresponds to cells with high mean Fl (>40,000). If the sample was not incubated on the magnet and suspended several times before elution, the increase in the mean Fl is mainly based on an increased number of medium and high producer cells (GFP++ and GFP+++) in the first three elution fractions. Although, the number of low producer (GFP+) also increased this higher number of low producers cannot compensate the effect on mean Fl due to the increased number of medium and high producer cells (Figure 59 and Figure 61). After the third elution, the mean Fl increased mainly due to the removal of low producer cells in the former elution fractions and thus a higher portion of GFP++ and GFP+++ cells. In contrast, if the sample was incubated on the magnet and suspended several times before elution, the increase in the mean Fl is mainly based on removal of low producers (GFP+) in the first elution fraction (Figure 60 and Figure 62). More medium and high producers were already eluted in elution 1 , especially for elutions with 1000 pM Biotin, but most notably the number of eluted GFP+ cells was much higher in the first two elutions compared to experiments with one incubation on the magnet. Due to this strong removal of GFP+ cells in the first elutions the portion of GFP++ and GFP+++ cells was high in the following fractions leading to an increase in mean Fl.

[413] Taken together, the repeated incubation of the samples on a magnet followed by resuspension prior to each elution step increased the Fl and cell yield compared to experiments where cells were incubated only once on the magnet before elution. Therefore, repeated incubation of cells on a magnet followed by re-suspension can increase the efficiency of the method to select high producing cells.

Example 17 - Selection of cells with different expression levels by using a biotin gradient and repeated magnetic incubation

[414] 6.51 x 10⁶ GFP positive MEXI-293E cells of a MEXi-293E suspension cell culture transfected with a DNA plasmid comprising the genetic elements comprised in Figure 10 comprising the expression cassette of SEQ ID NO 27 which expression is controlled by a PGK promoter (SEQ ID NO 26) were selected using the method of invention as described in Example 16, paras. [387] to [389] with following exceptions: The cells were mixed with 45 pl Strep- Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) and incubated for 20 min on a roller mixer. After the wash procedure cells were suspended in 1.3 ml MEXi-TM medium and 300 pl of the suspension were distributed to each of four 2 ml reaction tubes.

[415] Five MEXi-TM working solutions with D-biotin concentrations of 1 pM, 3 pM, 5 pM, 10 pM and 1000 pM were prepared. The solutions were prepared by mixing MEXi-TM with a 100 mM D-biotin stock. For the following elutions the four tubes were separated into two groups with two tubes in each group.

[416] Elution in both groups were conducted as follows: Cells were suspended with 500 pl of MEXi-TM working solution. After 11 min incubation on a roller mixer at 2-8°C the tubes were removed from the roller mixer.

[417] Tubes were placed for 30 s in the Magnetic Separator without removal of supernatant. Then the tubes with the cell-magnet bead mixture were suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s and the tubes were incubated on the roller mixer again for 2 min.

[418] Tubes were placed for 30 s in the Magnetic Separator without removal of supernatant. Then the tubes with the cell-magnet bead mixture were suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s and the tubes were incubated on the roller mixer again for 10 min.

[419] The tubes were placed in the Magnetic Separator and the supernatant with eluted cells was transferred to new tube (elution fraction 1 , E1). The cell-magnet bead mixture was suspended immediately with 500 pl of MEXi-TM working solution.

[420] Then the tubes with the cell-magnet bead mixture were suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s. [421] Tubes were placed for 30 s in the Magnetic Separator without removal of supernatant.

[422] The procedure of paras. [420] and [421] was repeated three times.

[423] The elution process described in paras. [419] to [422] was repeated 3 times in group 1 to get elution fraction E2 to E4 and it was repeated 19 times to in group 2 to get elution fraction E2 to E20.

[424] The two tubes in group 1 were eluted with 1000 pM biotin MEXi-TM working solution in all elution fractions

[425] The two tubes in group 2 were eluted with 1 pM biotin MEXi-TM working solution in elution 1 to elution 4, with 3 pM biotin MEXi-TM working solution in elution 5 to elution 8, with 5 pM biotin MEXi-TM working solution in elution 9 to elution 12, with 10 pM biotin MEXi-TM working solution in elution 13 to elution 16 and with 1000 pM biotin MEXi-TM working solution in elution 17 to elution 20.

[426] Elution fractions were analyzed on a Cytoflex flow cytometer.

[427] Eluted cells were separated in three categories based on their mean Fl and thus on their marker protein expression. GFP+ corresponds to cells with low mean Fl (<38,000). GFP++ corresponds to cells with medium mean Fl (38,000 - 70,000). GFP+++ corresponds to cells with high mean Fl (>70,000).

[428] The majority of cells in the gradient experiment (group 2) were eluted in the elution fractions that were treated with 1 pM and 3 pM biotin. The number of cells that was eluted at biotin concentrations >3 pM was significant lower. In the experiment using 1000 pM biotin only, (group 1) the number of eluted cells decreased with every consecutive elution fraction. Thus, in both experiments the majority of cells was eluted in the early elution fractions Figure 64. The mean fluorescence intensity of the eluted cells in both experiments (group 1 and group 2) increased continuously with every consecutive elution fraction (Figure 65). Because the mean Fl in the fourth elution at 1000 pM was comparable to the mean Fl in elution fraction 20 in the gradient elution experiment, both strategies allow the selection of high producing cells. Therefore, if only high expressing cells should be selected, the procedure can be simplified by adding a significant excess of biotin to the elution buffer right at the beginning of elution process (here 1000 pM biotin). This reduced the number of elution fractions to 4 instead of 20 in the described experiment. By using a high initial biotin concentration, the majority of cells in the first two elution fractions are low expressing cells (GFP+) (Figure 67). Because these cells are removed from the beads, their portion in the following elution fractions is reduced which results in a relative excess of medium (GFP++) and high expression cells (GFP+++) in the last two elution fractions and an increase in mean Fl (Figure 67). In contrast a gradient elution allows a more accurate fractionation of cells with different expression levels (Figure 66). The number of medium (GFP++) and high expressing cells (GFP+++) is low in the first four elutions at a biotin concentration of 1 pM. The portion of GFP++ and GFP+++ cells in the elution fractions after E4 increased and the number of GFP+ cells decreased which results in an average increase of the mean Fl (Figure 66). The high number of eluted cells in the first elution fractions is expected because the mean fluorescence intensity of the unselected original culture was 22580. The mean fluorescence intensity of cells in all elution fractions from elution 1 to elution 8 was 22064. This explains why the majority of cells was eluted in the elution fractions where 1 pM and 3 pM biotin was used for elution. The majority of cells within the original population had a low expression level of marker protein and thus these cells eluted at low biotin concentrations. These results further fortifies the principal of the biotin gradient elution that populations of cells can be separated and selected into subpopulations based on the concentration of marker protein on the cell surface by using different biotin concentrations in the elution. This effect can be used to enrich cells specifically within a defined range of expression level. Obviously, this method is not limited to the selection of genetic modified cells expressing a recombinant cell surface protein. It can also be used for any kind of selection where cells in a population are selected in subpopulations based on the density of a surface protein that interacts either directly with a ligand on a matrix or mediated by other proteins like antibodies or antigens that are fused to a binding partner that binds a ligand on a matrix. It is important that the binding to the matrix is reversible. The dissociation can be induced by the addition of a competing agent that binds to the binding site of the ligand which is otherwise associated with the binding peptide or protein on the cell surface or the binding partner of a cell surface protein like a receptor or antibody.

Example 18 - Increasing cell productivity by further optimization of marker cassette

[429] CHO cells were transfected in CHO-TF medium (Xell AG, Bielefeld, Germany) with a DNA plasmid comprising the genetic elements comprised in Figure 10 (ZSG5) or comprised in

[430] Figure 12 comprising the marker protein variant SEQ ID NO 12 and instead of eGFP a SEAP-TST fusion protein comprising SEQ ID NO 4, SEQ ID NO 28 and SEQ ID NO 5 was expressed as protein of interest. Cells were cultivated in a shaker incubator at 37°C, 5% CO₂ for 48 hours. After 48 hours temperature was lowered to 32°C. Cells were cultivated until viability was 75% or lower. At the end of cultivation cells were separated from cell culture supernatant by centrifugation. 25 pl BioLock solution and 100 pl 10-fold concentrated Buffer W were added per ml supernatant. SEAP-TST in the supernatant was purified via 1 ml Strep-Tactin®XT gravity flow columns (IBA Lifesciences GmbH, Gottingen, Germany). Elution fractions were analyzed with a Nanodrop 2000 (Thermo Scientifc, Cat. No, ND-2000). [431] 740 g SEAP-TST was purified from the culture transfected with pZSG5. In contrast 2016 pg SEAP-TST was purified from the culture that was transfected with the plasmid comprising the marker cassette without GFP and puromycin. Therefore, the protein of interest expression can be increased by removing the GFP gene and the resistance gene from the marker cassette.

Example 19 - optimizing marker protein by size reduction due to removal of eGFP and puromycin resistance gene

[432] 6 x 10⁶ GFP positive MEXi-293E cells of suspension a cell culture that was transfected with a DNA plasmid comprising

(1) the genetic elements comprised in Figure 10 comprising a marker protein comprising SEQ ID NO 27(ZSG5) or

[433] the genetic elements comprised in

(2) Figure 12 comprising a marker protein comprising SEQ ID NO 12 (CD4tm) or

[434] the genetic elements comprised in

(3) Figure 12 comprising a marker protein comprising SEQ ID NO 14 (CD4tm-delta) or

[435] the genetic elements comprised in

(4) Figure 12 comprising a marker protein comprising SEQ ID NO 13 (CD4tm-A) was centrifuged at 100 x g for 5 min at room temperature (20°C). The supernatant was discarded and the cell pellet was suspended in 10 ml PBS (37°C). The cells were centrifuged again at 100 x g for 5 min at room temperature (20°C). The pellet was suspended in 15 ml cold MEXi-TM medium. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 360 pl cold MEXi-TM.

[436] The suspension was added to 40 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) which were prepared as follows. Strep-Tactin® magnetic microbeads were washed 2 times with 1 ml MEXi-TM medium on a magnet before use. Positive cells were separated on StrepMan magnet (I BA, Cat. No. 6-5650-065) as follows: After incubation for 20 min on the roller mixer at 2-8°C, the cell-microbead-mixture was transferred to a 15 ml centrifugation tube containing 10 ml cold MEXi-TM medium.

[437] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads were carefully washed off the tube wall by suspending in 10 ml MEXi-TM medium. This procedure was repeated twice. [438] Then cells were suspended in 700 pl cold MEXi-TM medium and divided in 2 x 300 pl which were transferred into 2 ml reaction tubes. The tubes were put on a Magnetic Separator and the supernatant was discarded.

[439] Cells were eluted from the beads using MEXi-TM medium supplemented with 1 mM biotin. Cells from paras. [438] were suspended in 500 pl elution buffer and incubated at 2-8°C on a roller mixer for 10 min. Then the tube was removed from the mixer.

[440] The tube was placed for 30 s in the Magnetic Separator without removal of supernatant. Then the tube with the cell-magnet bead mixture was suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s.

[441] The procedure of paras. [440] was repeated two times.

[442] The tubes were placed in the Magnetic Separator and the supernatant with eluted cells was transferred to new tube (elution fraction 1 , E1). The cell-magnet bead mixture was suspended immediately with 500 pl of elution buffer.

[443] The elution process described in paras. [440] to [442] was repeated 3 times in group 1 to get elution fraction E2 to E4.

[444] The results in Example 18 shows that the removal of proteins and elements like eGFP, 2A, and the Puromycin resistance gene increases the productivity of a transient expressing cell culture compared to cells that express all these elements. Cells that expressed only the TST- CD4tm protein showed a 8.2 fold reduced cell yield compared to cells that expressed a TST- CD4-tm-eGFP fusion protein (Figure 68). Staining of the cells with Strep-Tactin conjugate DY- 649 that binds to the TST of the TST-CD4tm fusion protein before selection showed now significant difference in the mean Fl of stained cells compared to cells that express TST-CD4tm- eGFP. Therefore, the functional marker concentration on the cell surface was comparable and the reduced yield is not a result of a lower marker concentration. Popik and Alee (2004), JBC, 279(1):704-712, verified that CD4 is localized in rafts in its native context and that the RHRRR (SEQ ID NO: 57) amino acid sequence is responsible for the localization in rafts. By replacement of the four arginine with alanine CD4 localization in rafts was prevented resulting in more uniformly distribution over the cell membrane. We assumed that the fusion of eGFP close to this amino acid motif results in a masking of this motif leading to a uniform distribution of the maker on the cell surface and that such a uniform distribution is beneficial for the selection. Thus, we replaced the four arginine by alanine (CD4tm-A) or removed the last nine amino acids from the C-terminus of our marker protein (CD4tm-delta). By replacing the arginine, the yield of selected cells was increased by 4.4-fold so the yield was only 1.8 fold lower compared to cells that express the TST-CD4tm-eGFP variant. Removal of the 9 C-terminal amino acids resulted in a 2.7-fold higher yield compared to TST-CD4tm variant (Figure 68). In the CD4tm-delta variant the palmitoylated cystines were also deleted. These palmitoylated cystines anchor the protein in the membrane. Therefore, the removal of the palmitoylated cystines might decrease the binding efficiency leading to lower yields. Summarized, the expression of TST-CD4tm-A is the optimal expression marker as it provided surprisingly high cells yield and reduces the resources the cells need to invest to express the marker protein cassette compared to ZSG5 resulting in higher expression levels of the protein of interest.

Example 20 - Expression of POI from cells that were selected by the method of invention using the optimizing marker protein from Example 19

[445] 6 x 10⁶ CHO cells of a suspension cell culture transfected with a DNA plasmid comprising the genetic elements comprised in

[446] Figure 12 comprising a marker protein comprising SEQ ID NO 13 were centrifuged at 100 x g for 5 min at 4°C. The supernatant was discarded and the cell pellet was suspended in 5 ml cold Buffer IS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 5 ml cold Buffer IS. The cells were centrifuged again at 100 x g for 5 min at 4°C. The pellet was suspended in 400 pl cold Buffer IS.

[447] 400 pl of the suspension was added to 80 pl Strep-Tactin® magnetic microbeads (I BA, Cat. No. 6-5510-050) which were prepared as follows. Strep-Tactin® magnetic microbeads were washed 2 times with 1 ml Buffer IS on a magnet before use. Positive cells were separated on StrepMan magnet (I BA, Cat. No. 6-5650-065) as follows: After incubation for 5 min on the roller mixer at 2-8°C, the cell-microbead-mixture was transferred to a 15 ml centrifugation tube containing 10 ml cold Buffer IS.

[448] The tube was placed for at least 3 min firmly onto the StrepMan Magnet. Supernatant containing the negative cell fraction was removed by careful pipetting while keeping the reaction tube on the StrepMan Magnet. Reaction tube was removed from the StrepMan Magnet and cells bound to Strep-Tactin® magnetic microbeads (eGFP and TST-flexible-CD4tm-A positive cell fraction) were carefully washed off the tube wall by suspending in 10 ml cold Buffer IS. This procedure was repeated twice.

[449] Then cells were suspended in 700 pl cold Buffer IS and 350 pl suspension were transferred into a 2 ml reaction tube. The tube was put on a Magnetic Separator and the supernatant was discarded. [450] Cells were eluted from the beads using Buffer IS supplemented with 1 mM biotin. Cells from paras. [449] were suspended in 500 pl elution buffer and incubated at 2-8°C on a roller mixer for 10 min. Then the tube was removed from the mixer.

[451] The tube was placed for 30 s in the Magnetic Separator without removal of supernatant. Then the tube with the cell-magnet bead mixture was suspended on a vortex mixer (Scientific Industries, Inc., G-560E) at level 7 for 10 s.

[452] The procedure of paras. [451] was repeated two times.

[453] The tubes were placed in the Magnetic Separator and the supernatant with eluted cells was transferred to new tube (elution fraction 1, E1). The cell-magnet bead mixture was suspended immediately with 500 pl of elution buffer.

[454] The elution process described in paras. [451] to [453] was repeated 2 times in group 1 to get elution fraction E2 to E3.

[455] Elution fractions were stained with Strep-Tactin®XT conjugate DY-649 and analyzed on a Cytoflex flow cytometer. GFP fluorescence intensity was measured in the PE channel because of high values in the FITC channel that exceed the maximum detection level.

[456] The marker protein staining of the unselected population with Strep-Tacitin®XT DY-649 conjugate in Figure 69 shows a correlation of GFP expression (POI) and marker protein (CD4tm-A, SEQ ID NO 13) expression in cells transfected with a plasmid comprising the elements in

[457] Figure 12. Even if, the expression of GFP at a given marker expression varies, cells exhibit higher GFP expression also expressed more marker protein. Almost all cells with a GFP mean fluorescence intensity above 10⁵ were stained while a significant portion of cells with an GFP mean fluorescence intensity below 10⁵ were not stained. Actually, by using the method of invention, the vast majority of the cells with a low GFP expression corresponding to a mean Fl below 10⁵ was removed (Figure 70 to Figure 72). Mean GFP Fl of positive cells increased by factor 8.2 from 199342 in the unselected population to 1638926 in elution 1. Elution 2 exhibit a mean GFP Fl of 1677330 and elution 3 of 1485621 (Figure 73). The increase in mean Fl is important because it would result in an increased total GFP production capacity of the selected population. The median GFP Fl was increased 173-fold from 5749 in the unselected population to 994216 in elution 1. Elution 2 exhibit a median GFP Fl of 1027707 and elution 3 of 616239. Figure 73 shows the mean GFP Fl and mean DY-649 Fl of GFP positive cells before and after selection using the method of invention. Cells were transfected with a plasmid comprising the genetic elements according to Figure 12 with SEQ ID NO 13. The cells were stained with Strep- TactinOXT conjugate DY-649 which binds to the TST presented on the cell surface.

[458] Figure 74)

[459] The increased median shows the enrichment of high producing clones in selected populations. Thus, less clones must be picked if a clonal cell line with high productivity should be generated from such a pool. Due to the selection, the mean DY-649 signal of GFP positive and DY-649 positive cells increased 1.8-fold from 36202 in the unselected population to 65422 in elution 1. With every further elution fraction the 649 mean Fl was further increased as expected based on the results of the other examples in this patent. Elution 2 exhibit a mean 649 Fl of 107122 and elution 3 of 153038 (Figure 73). The Fl in elution 3 corresponds to an increase of mean 649 Fl by 4.2-fold and agrees well with the increase of expression marker protein in selected populations in the other examples. Thus, it can be concluded that the optimized marker protein shows the same robust selection pattern as the other constructs that were used in this study. Hence, cells with a lower marker expression elute primarily in the first elution which leads to an increased portion of cells with high marker expression in later elution fractions. In this experiment the POI expression did not increase in elution fractions after elution 1. However, it is well known that GFP as protein of interest does not correspond well on a quantitative level to the expression of secreted or membrane proteins due to its cytosolic location and expression. Furthermore, a variance of expressed GFP at a given expression level of marker protein can be expected due to the heterogenous nature of a polyclonal population.

Claims

CLAIMS A method of fractionating cells of a population of cells based on the amount of a marker protein on the cell surface, the method comprising:

(i) optionally repeating steps (g) and (h) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more times. A method of fractionating cells of a population of cells based on the amount of a receptor molecule on the cell surface, the method comprising:

(i) optionally repeating steps (g) and (h) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more times. A method of enriching or isolating cells comprised in a population of cells, wherein the enriched cells express a protein of interest at a level higher than the mean of the population of cells, the method comprising:

(e) adding the competitor to a second concentration, thereby eluting a fraction of the population of cells from the solid phase;

(f) optionally separating the eluted fraction of the population of cells obtained in step (e); (g) optionally adding the concentration to a further concentration, thereby eluting a fraction of the population of cells from the solid phase;

(g);

(i) optionally repeating steps (g) and (h) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more times. The method of any one of claims 1 to 4, wherein the first concentration, the second concentration and the further concentration(s) are essentially the same.

The method of any one of claims 1 to 4, wherein the second concentration is higher than the first concentration, wherein the further concentration is higher than the second concentration and each subsequent further concentration is higher than the previous further concentration. The method of any one of claims 2, 4, 5 or 6, wherein the population of cells comprises cells comprising a nucleic acid comprising (i) an expression cassette encoding for a marker protein comprising and (ii) a protein of interest. The method of any one of the preceding claims 1 to 7, wherein the amount of the marker protein or the receptor molecule on the cell surface of a cell of the population of cells is correlated to the level of expression of the protein of interest. The method of any one of the preceding claims 1 to 8, wherein in the expression cassette

(iv) the protein of interest is immobilized on the cell surface via methods like cold capture which preferably stops or reduces secretion of the protein by trapping it temporally on the cell surface e.g., by reducing the temperature of the medium. The method of any one of the preceding claims 1 to 9, wherein the marker protein

(ii) is a peptide fused to a membrane anchor. The method of any one of the preceding claims 1 to 10, wherein the marker protein comprises a transmembrane domain of a protein selected from the group consisting of EpCAM, VEGFR, integrin, optionally integrins avp3, a4, alip3, a4p7, a5pi, avp3 or an, a member of the TNF receptor superfamily, optionally TRAIL-RI or TRAIL-R2, a member of the epidermal growth factor receptor family, PDGF Receptor, interferon receptor, folate receptor, GPNMB, ICAM-1 , HLA-DR, CEA, CA-125, MUCI, TAG-72, IL-6 receptor, 5T4, GD2, GD3, prostate-specific membrane antigen (PSMA) or a clusters of differentiation cell surface molecule, optionally CD2, CD3, CD4, CD5, CD11 , CDIIa/LFA- 1, CD15, CD18/ITGB2, CD19, CD20, CD22, CD23/lgE Receptor, CD25, CD28, CD30, CD33, CD38, CD40, CD41, CD44, CD51, CD52, CD62L, CD74, CD80, CD125, CD147/basigin, CD152/CTLA-4, CD154/CD40L, CD195/CCR5 and CD319/SLAMF7, and low-affinity nerve growth factor receptor (LNGFR), wherein the protein is preferably CD4. The method of any one of the preceding claims 1 to 11 , wherein the marker protein comprises or consists of any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28, or a fragment or analog thereof having a sequence identity of 60% or higher compared to any one of SEQ ID NO: 2-8, 10-18, 20, 21 , 23-25, 27, or 28. The method of any one of the preceding claims 1 to 12, wherein the binding partner B and the ligand L form a binding pair selected from the group of streptavidin or a streptavidin analog and a ligand binding to streptavidin, a binding pair that binds in the presence of a divalent cation, an oligohistidine peptide and a binding moiety A comprising at least two chelating groups K, wherein each chelating group K is capable of binding to a transition metal ion, thereby rendering binding moiety A capable of binding to the oligohistidine peptide, an antigen and an antibody against said antigen, wherein said binding partner B comprises the antigen and said ligand L comprises the antibody against said antigen. The method of claim 13, wherein

(b) said binding partner B comprises a biotin analog that reversibly binds to streptavidin or Avidin and said ligand L comprises streptavidin, or Avidin, or a streptavidin analog, or an Avidin analog that reversibly binds to said biotin analog, or

(c) said binding partner B comprises a streptavidin or Avidin binding peptide and said ligand L comprises streptavidin, or Avidin, or a streptavidin analog, or an Avidin analog that reversibly binds to said streptavidin or Avidin binding peptide.

133 The method of claim 14, wherein said ligand L comprises a streptavidin mutein comprising the amino acid sequence Val⁴⁴-Thr⁴⁵-Ala⁴⁶-Arg⁴⁷ (SEQ ID NO: 54) at sequence positions 44 to 47 of wild-type streptavidin or a streptavidin mutein comprising the amino acid sequence lle⁴⁴-Gly⁴⁵-Ala⁴⁶-Arg⁴⁷ (SEQ ID NO: 55) at sequence positions 44 to 47 of wild-type streptavidin and wherein said binding partner B comprises the streptavidin-binding peptide that comprises or consists of one of the following sequences: a) -Trp-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO: 35), wherein Xaa is any amino acid and Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg , b) -Trp-Arg-His-Pro-GIn-Phe-Gly-Gly- (SEQ ID NO: 36), c) -Trp-Ser-His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO: 37), d) a sequential arrangement of at least two streptavidin binding peptides, wherein each peptide binds streptavidin, wherein the distance between two peptides is at least 0 and not greater than 50 amino acids and wherein each of the at least two peptides comprises the amino acid sequence -His-Pro-Baa- in which Baa is selected from the group consisting of glutamine, asparagine and methionine, e) a sequential arrangement as recited in d), wherein one of the at least two peptides comprises the sequence -His-Pro-GIn- f) a sequential arrangement as recited in d), wherein one of the peptides comprises an amino acid sequence -His-Pro-GIn-Phe- (SEQ ID NO: 38), g) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino sequence -Oaa-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO: 39), where Oaa is Trp, Lys or Arg, Xaa is any amino acid and where either Yaa and Zaa are both Gly or Yaa is Glu and Zaa is Lys or Arg, h) a sequential arrangement as recited in d) wherein at least one peptide includes at least the amino acid sequence -Trp-Xaa-His-Pro-GIn-Phe-Yaa-Zaa- (SEQ ID NO:

41), j) the amino acid sequence -Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(Xaa)n-Trp-Ser- His-Pro-GIn-Phe-Glu-Lys- (SEQ ID NO: 42) wherein Xaa is any amino acid and n is an integer from 0 to 12. k) an amino acid sequence selected from the group consisting of Trp-Arg-His-Pro-GIn- Phe-Gly-Gly (SEQ ID NO: 41), Trp-Ser-His-Pro-GIn-Phe-Glu-Lys (SEQ ID NO: 43), Trp- Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₃-Trp-Ser-His-Pro-Gln-Phe-Glu-Lys (SEQ ID NO: 44), Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₂-Trp-Ser-His-Pro-Gln-

134 Phe-Glu-Lys (SEQ ID NO: 46) or Trp-Ser-His-Pro-Gln-Phe-Glu-Lys-(GlyGlyGlySer)₂-Gly- Gly-Ser-Ala-Trp-Ser-His-Pro-GIn-Phe-Glu-Lys (SEQ ID NO: 47). The method of claim 13, wherein for a binding pair that binds in the presence of a divalent cation, said binding partner B comprises a calmodulin binding peptide and the said ligand L comprises calmodulin, or wherein said binding partner B comprises a FLAG peptide and said multimerization reagent comprises an antibody binding the FLAG peptide, or wherein said binding partner B comprises an oligohistidine tag and said ligand L comprises a chelated transition metal. The method of claim 16, wherein the divalent cation is selected from the group consisting of Ca²⁺, Ni ²⁺or Co²⁺. The method of claim 16, wherein the binding between said binding partner B and said ligand L is disrupted by metal ion chelation, preferably wherein the metal chelation is accomplished by addition of EDTA or EGTA. The method of claim 13, wherein the antigen comprised in said binding partner B is an epitope tag. The method of claim 19, where the epitope tag is selected from the group consisting of the Myc-tag (sequence: EQKLISEEDL, SEQ ID NO: 52), the HA-tag (sequence: YPYDVPDYA, SEQ ID NO: 48), the VSV-G-tag (sequence: YTDIEMNRLGK, SEQ ID NO: 49), the HSV-tag (sequence: QPELAPEDPED, SEQ I D NO: 50), and the V5-tag (sequence: GKPIPNPLLGLDST, SEQ ID NO: 53). The method of claim 13, wherein the antigen comprised in said binding partner B is a protein. The method of claim 21 , wherein the protein is selected from the group consisting of glutathione-S-transferase, maltose binding protein (MBP), chitin binding protein (CBP) and thioredoxin. The method of claim 15, wherein the fractions of the cells are eluted by adding biotin.

135 The method of claim 23, wherein biotin is added to a concentration of at least 3 pM, at least 5pM, at least 10 pM, at least 100 pM, at least 250 pM, at least 500 pM, at least 1 mM, at least 2.5 mM, at least 5 mM or at least 10 mM Biotin. The method of any one of the preceding claims, wherein the solid phase is a selected from a bead, a plastic plate, a membrane or a solid phase suitable for chromatography. The method of any one of the preceding claims, wherein the method is a batch method or a chromatographic method. The method of any one of the preceding claims, wherein the nucleic acid comprised in cells of the population of cells further comprises a selection marker. The method of claim 27, wherein the method further comprises a step (a’) prior to step (a):

(a’) selecting cells, which express the selection marker. The method of any one of the preceding claims 1 to 28, wherein the cell is a prokaryotic cell or a eukaryotic cell. The method of claim 29, wherein the cell is a prokaryotic cell, preferably of the species selected from the group consisting of Lactobacillus spp., Yersinia spp., Escherichia spp., Klebsiella spp., Bordetella spp., Neisseria spp., Aeromonas spp., Franciesella spp., Corynebacterium spp., Citrobacter spp., Chlamydia spp., Hemophilus spp., Brucella spp., Mycobacterium spp., Legionella spp., Rhodococcus spp., Pseudomonas spp., Helicobacter spp., Salmonella spp., Vibrio spp., Bacillus spp., Leish mania spp. and Erysipelothrix spp., Shigella spp., Listeria spp., Rickettsia spp., Acetoanaerobium spp., Aerococcaceae spp., Carnobacteriaceae spp., Enterococcace spp., Leuconostocacease spp., Streptococcaceae spp., and bacteria with GRAS status, preferably E. coli. The method of claim 29, wherein the host cell is a eukaryotic cell, preferably a cell selected from the group consisting of CHO cells, CHO-S, ExpiCHO, Freestyle CHO-S, CHO-GS, CHO-K1, CHO-DXB11, CHO-DG44, CHO duk, CHO-DP12, CHOZN, GS- CHOK1SV, HEK-293 cells, HEK-293T cells, HEK-293-6E, HEK-293-EBNA, HEK 293SF-3F6, 293 c18, Expi293, 293-F, insect cells, SF9, ExpiSf9, Hi-5, Sf21, human amniocytes and CAP®.

136 The method of any one of the preceding claims 1 to 31 , wherein the protein of interest is selected from the group consisting of an amylolytic enzyme, a lipolytic enzyme, a proteolytic enzyme, a cellulytic enzyme, an oxidoreductase or a plant cell-wall degrading enzyme, an aminopeptidase, amylase, amyloglucosidase, carbohydrase, carboxypeptidase, catalase, cellulase, chitinase, cutinase, cyclodextrin glycosyltransferase, desoxyribonuclease, esterase, galactosidase, beta-galactosidase, glucoamylase, glucose oxidase, glucosidase, haloperoxidase, hemicellulase, invertase, isomerase, laccase, ligase, lipase, lyase, mannosidase, oxidase, pectinase, peroxidase, phytase, phenoloxidase, polyphenoloxidase, protease, ribonuclease, transferase, transglutaminase, and xylanase, a growth factor, cytokine, receptors, receptor ligands, therapeutic proteins such as interferons, BMPs, GDF proteins, fibroblast growth factors, peptides such as protein inhibitors, membrane proteins, membrane-associated proteins, peptide/protein hormones, peptidic toxins, peptidic antitoxins, antibody or functional fragments thereof such as Fab or F(ab)₂ or derivatives of an antibody such as bispecific antibodies (for example, scFvs), chimeric antibodies, humanized antibodies, single domain antibodies such as Nanobodies or domain antibodies (dAbs) or an anticalin and others. The method of any one of claims 1 to 4 and 7 to 32, wherein the second concentration is lower than the first concentration, wherein the further concentration is lower than the second concentration and each subsequent further concentration is lower than the previous further concentration. The method of any one of the preceding claims 1 to 34, wherein the marker protein and the protein of interest are identical.

137