A GROUP OF CHIMERIC ANTIGEN RECEPTORS (CARS)

Abstract

A group of chimeric antigen receptors (CARs) having two, three or four CAR molecules, wherein the members of the group of CARs can be different in their amino acid sequences, and wherein each of the CAR molecules of the group includes at least a transmembrane domain and an ectodomain comprising either an antigen binding moiety or a binding site to which another polypeptide is able to bind, wherein the polypeptide comprises an antigen binding moiety; wherein each CAR molecule of the group includes at least one dimerization domain, wherein this dimerization of a pair of dimerization domains is either induced by a regulating molecule and optionally reduced by another regulating molecule, or occurs in the absence of a regulating molecule and is reduced by a regulating molecule, and wherein the antigen binding moieties of the CAR molecules of the group specific for one target antigen.

Claims

1. A group of chimeric antigen receptors (CARs) consisting of two, three or four CAR molecules, wherein the members of the group of CARs can be different or identical in their amino acid sequences to one another; wherein each of the CAR molecules of the group comprise at least a transmembrane domain and an ectodomain comprising either an antigen binding moiety or a binding site to which another polypeptide is able to bind, wherein the another polypeptide comprises an antigen binding moiety; wherein at least one CAR molecule of the group additionally comprises an endodomain, which comprises at least a signalling region which can transduce a signal via at least one immunoreceptor tyrosine-based activation motif (ITAM) or at least one immunoreceptor tyrosine-based inhibitory motif (ITIM); wherein the endodomain of each CAR molecule of the group, in case the respective CAR molecule comprises an endodomain, is located on the intracellular side of a cell membrane, if expressed in a cell, wherein the ectodomain of each CAR molecule of the group translocates to the extracellular side of a cell membrane, if expressed in a cell, and wherein the transmembrane domain of each CAR molecule of the group is located in a cell membrane, if expressed in a cell; wherein each CAR molecule of the group comprises at least one dimerization domain, which can mediate homo- or heterodimerization with other CAR molecules of the group, wherein this dimerization of a pair of dimerization domains is either induced by a regulating molecule and optionally reduced by another regulating molecule, or occurs in the absence of a regulating molecule and is reduced by a regulating molecule, wherein a regulating molecule is able to bind under physiological conditions to at least one member of a pair of dimerization domains and by inducing or reducing dimerization either induces or reduces the formation of a non-covalently complexed group of CARs consisting of two, three or four CAR molecules; wherein the ectodomain of each CAR molecule of the group in its prevalent conformation is free of cysteine amino acid moieties which are able to form intermolecular disulphide bonds with other CAR molecules of the group, respectively; wherein the antigen binding moieties of the CAR molecules of the group and of the other polypeptides being able to bind to the CAR molecules of the group are either specific for one target antigen or for a non-covalent or a covalent complex of different target antigens; wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 1 mM and 100 nM; and wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 1 mM and 100 nM.

2. The group of CARs according to claim 1, wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 1 mM and 150 nM; and wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 1 mM and 150 nM.

3. The group of CARs according to claim 1, wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 500 μM and 100 nM; and wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 500 μM and 100 nM.

4. The group of CARs according to claim 1, wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 500 μM and 150′ nM; and wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 500 μM and 150 nM.

5. The group of CARs according to claim 1, wherein each target antigen specifically recognized by the antigen binding moieties of the group of CARs or of other polypeptides being able to bind to CAR molecules of the group is a naturally occurring cellular surface antigen or a polypeptide, carbohydrate or lipid bound to a naturally occurring cellular surface antigen.

6. The group of CARs according to claim 1, wherein the antigen binding moieties of the group of CARs and of other polypeptides being able to bind to CAR molecules of the group bind to one or more target antigens present on a cell, preferably one or more target antigens of a cell, on a solid surface, or a lipid bilayer, especially wherein at least one target antigen comprises a molecule preferably selected from the group consisting of CD19, CD20, CD22, CD23, CD28, CD30, CD33, CD35, CD38, CD40, CD42c, CD43, CD44, CD44v6, CD47, CD49D, CD52, CD53, CD56, CD70, CD72, CD73, CD74, CD79A, CD79B, CD80, CD82, CD85A, CD85B, CD85D, CD85H, CD85K, CD96, CD107a, CD112, CD115, CD117, CD120b, CD123, CD146, CD148, CD155, CD185, CD200, CD204, CD221, CD271, CD276, CD279, CD280, CD281, CD301, CD312, CD353, CD362, BCMA, CD16V, CLL-1, Ig kappa, TRBC1, TRBC2, CKLF, CLEC2D, EMC10, EphA2, FR-a, FLT3LG, FLT3, Lewis-Y, HLA-G, ICAM5, IGHA1/IgA1, IL-1RAP, IL-17RE, IL-27RA, MILR1, MR1, PSCA, PTCRA, PODXL2, PTPRCAP, ULBP2, AJAP1, ASGR1, CADM1, CADM4, CDH15, CDH23, CDHR5, CELSR3, CSPG4, FAT4, GJA3, GJB2, GPC2, GPC3, IGSF9, LRFN4, LRRN6A/LINGO1, LRRC15, LRRC8E, LRIG1, LGR4, LYPD1, MARVELD2, MEGF10, MPZLI1, MTDH, PANX3, PCDHB6, PCDHB10, PCDHB12, PCDHB13, PCDHB18, PCDHGA3, PEP, SGCB, vezatin, DAGLB, SYT11, WFDC10A, ACVR2A, ACVR2B, anaplastic lymphoma kinase, cadherin 24, DLK1, GFRA2, GFRA3, EPHB2, EPHB3, EPHB4, EFNB1, EPOR, FGFR2, FGFR4, GALR2, GLG1, GLP1R, HBEGF, IGF2R, UNC5C, VASN, DLL3, FZD10, KREMEN2, TMEM169, TMEM198, NRG1, TMEFF1, ADRA2C, CHRNA1, CHRNB4, CHRNA3, CHRNG, DRD4, GABRB3, GRIN3A, GRIN2C, GRIK4, HTR7, APT8B2, NKAIN1, NKAIN4, CACNA1A, CACNA1B, CACNA1I, CACNG8, CACNG4, CLCN7, KCN§ A4, KCNG2, KCNN3, KCNQ2, KCNU1, PKD1L2, PKD2L1, SLC5A8, SLC6A2, SLC6A6, SLC6A11, SLC6A15, SLC7A1, SLC7A5P1, SLC7A6, SLC9A1, SLC10A3, SLC10A4, SLC13A5, SLC16A8, SLC18A1, SLC18A3, SLC19A1, SLC26A10, SLC29A4, SLC30A1, SLC30A5, SLC35E2, SLC38A6, SLC38A9, SLC39A7, SLC39A8, SLC43A3, TRPM4, TRPV4, TMEM16J, TMEM142B, ADORA2B, BAI1, EDG6, GPR1, GPR26, GPR34, GPR44, GPR56, GPR68, GPR173, GPR175, LGR4, MMD, NTSR2, OPN3, OR2L2, OSTM1, P2RX3, P2RY8, P2RY11, P2RY13, PTGE3, SSTR5, TBXA2R, ADAM22, ADAMTS7, CST11, MMP14, LPPR1, LPPR3, LPPR5, SEMA4A, SEMA6B, ALS2CR4, LEPROTL1, MS4A4A, ROM1, TM4SF5, VANGL1, VANGL2, C18orf1, GSGL1, ITM2A, KIAA1715, LDLRAD3, OZD3, STEAP1, MCAM, CHRNA1, CHRNA3, CHRNA5, CHRNA7, CHRNB4, KIAA1524, NRM.3, RPRM, GRM8, KCNH4, Melanocortin 1 receptor, PTPRH, SDK1, SCN9A, SORCS1, CLSTN2, Endothelin converting enzyme like-1, Lysophosphatic acid receptor 2, LTB4R, TLR2, Neurotropic tyrosine kinase 1, MUC16, B7-H4, epidermal growth factor receptor (EGFR), ERBB2, HER3, EGFR variant III (EGFRvIII), HGFR, FOLR1, MSLN, CA-125, MUC-1, prostate-specific membrane antigen (PSMA), mesothelin, epithelial cell adhesion molecule (EpCAM), L1-CAM, CEACAM1, CEACAM5, CEACAM6, VEGFR1, VEGFR2, high molecular weight-melanoma associated antigen (HMW-MAA), MAGE-A 1, IL-13R-α2, disialogangliosides (GD2 and GD3), tumour-associated carbohydrate antigens (CA-125, CA-242, Tn and sialyl-Tn), 4-1BB, 5T4, BAFF, carbonic anhydrase 9 (CA-IX), C-MET, CCR1, CCR4, FAP, fibronectin extra domain-B (ED-B), GPNMB, IGF-1 receptor, integrin α5β1, integrin αvβ3, ITB5, ITGAX, embigin, PDGF-Rα, ROR1, Syndecan 1, TAG-72, tenascin C, TRAIL-R1, TRAIL-R2, NKG2D-Ligands, a major histocompatibility complex (MHC) molecule presenting a tumour-specific peptide epitope, preferably PR1/HLA-A2, a lineage-specific or tissue-specific tissue antigen, preferably CD3, CD4, CD5, CD7, CD8, CD24, CD25, CD34, CD80, CD86, CD133, CD138, CD152, CD319, endoglin, and an MHC molecule.

7. A nucleic acid molecule comprising nucleotide sequences encoding the individual CAR molecules of a group of CARs according to claim 1, wherein the nucleic acid is selected from DNA, RNA, or in vitro transcribed RNA.

8. A kit of nucleic acid molecules comprising nucleotide sequences encoding the individual CAR molecules of a group of CARs according to claim 1, wherein the nucleic acid is selected from DNA, RNA, or in vitro transcribed RNA.

9. A vector or a kit of vectors comprising nucleotide sequences encoding the individual CAR molecules of a group of CARs according to claim 1, wherein the nucleic acid is DNA or RNA.

10. A cell modified in vitro or ex vivo with a nucleic acid molecule or a kit of nucleic acid molecules wherein the nucleic acid is selected from DNA, RNA or in vitro transcribed RNA to produce the individual CAR molecules of a group of CARs according to claim 1, or a kit comprising two or more of said modified cells.

11. A pharmaceutical preparation comprising a nucleic acid or a kit of nucleic acids according to claim 7.

12. The group of CARs according to claim 1 for use in a method of treatment of a cancer in an individual, wherein the method comprises: i) genetically modifying NK cells or preferably T lymphocytes obtained from the individual with at least one nucleic acid molecule comprising sequences encoding the respective CAR molecules of the group of CARs, wherein each antigen binding moiety of the group of CARs is specific for a target antigen on a cancer cell in the individual, and wherein said genetic modification is carried out in vitro or ex vivo; ii) introducing the genetically modified cells into the individual; and iii) administering to the individual an effective amount of at least one regulating molecule for either inducing or reducing dimerization of the respective CAR molecules of the group, preferably inducing dimerization of the respective CAR molecules of the group, thereby either inducing or reducing non-covalent complexation of the group of CARs, preferably inducing non-covalent complexation of the group of CARs, wherein the non-covalently complexed group of CARs upon contact with a cancer cell expressing the respective target antigen or the respective covalent or non-covalent complex of different target antigens mediates activation of the genetically modified cell, which leads to killing of the cancer cell and thereby enables treating the cancer.

13. The cell according to claim 10 for use in a method of treatment of a cancer in an individual, wherein each antigen binding moiety of the group of CARs is specific for a target antigen on a cancer cell in the individual, and wherein the method comprises: i) introducing the cell into the individual; and ii) administering to the individual an effective amount of at least one regulating molecule for either inducing or alternatively reducing, preferably inducing, the formation of a non-covalent complex comprising two, three or four CAR molecules of the group, wherein the non-covalently complexed group of CARs upon contact with a cancer cell expressing the respective target antigen or the respective covalent or non-covalent complex of different target antigens-mediates activation of the genetically modified cell, which leads to killing of the cancer cell and thereby enables treating the cancer.

14. A kit comprising: a group of CARs according to claim 1; and one, two or three regulating molecules, preferably two, even more preferably one regulating molecule.

15. The group of CARs according to claim 1, for use in the treatment of a disease which is characterised by the need to bind a T lymphocyte or an NK cell to a target antigen on a cell, preferably for use in the treatment of a tumour patient, especially a tumour patient with a tumour selected from Ewing's sarcoma, rhabdomyosarcoma, osteosarcoma, osteogenic sarcoma, mesothelioma, fibrosarcoma, myxosarcoma, liposarcoma, chondrosarcoma, chordoma, angiosarcoma, endotheliosarcoma, lymphangiosarcoma, lymphangioendotheliosarcoma, synovioma, leiomyosarcoma, melanoma, glioma, astrocytoma, medulloblastoma, neuroblastoma, retinoblastoma, oligodendroglioma, menangioma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, chronic myeloproliferative syndromes, acute myelogenous leukemias, chronic lymphocytic leukemias (CLL) including B-cell CLL, T-cell CLL, prolymphocytic leukemia and hairy cell leukemia, acute lymphoblastic leukemias, B-cell lymphomas, Hodgkin's lymphoma, non-Hodgkin's lymphoma, esophageal carcinoma, hepatocellular carcinoma, basal cell carcinoma, squamous cell carcinoma, bladder carcinoma, transitional cell carcinoma, bronchogenic carcinoma, colon carcinoma, colorectal carcinoma, gastric carcinoma, lung carcinoma, including small cell carcinoma and non-small cell carcinoma of the lung, adrenocortical carcinoma, thyroid carcinoma, pancreatic carcinoma, breast carcinoma, ovarian carcinoma, prostate carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinoma, cystadenocarcinoma, medullary carcinoma, renal cell carcinoma, ductal carcinoma, bile duct carcinoma, choriocarcinoma, seminoma, embryonal carcinoma, Wilm's tumor, cervical carcinoma, uterine carcinoma, testicular carcinoma, osteogenic carcinoma, epithelial carcinoma, and nasopharyngeal carcinoma, atypical meningioma, islet cell carcinoma, medullary carcinoma, mesenchymoma, hepatocellular carcinoma, hepatoblastoma, clear cell carcinoma, and neurofibroma mediastinum.

Description

SHORT DESCRIPTION OF THE FIGURES

[0291] FIG. 1 shows the schematics of exemplary architectures of the group of CARs.

[0292] FIG. 2 shows the K.sub.d values of different rcSso7d-based antigen binding moieties towards human EGFR.

[0293] FIG. 3 shows that extracellular disulphide bond forming cysteines can prevent the exploitation of the avidity effect for regulation of CAR function.

[0294] FIG. 4 shows that dimerization/oligomerization of scFv-comprising CAR molecules prevents the exploitation of the avidity effect for regulation of CAR function.

[0295] FIG. 5 shows the regulation of CAR function by regulating the avidity of a group of CARs by conditional homodimerization.

[0296] FIG. 6 shows the regulation of the function of stably transduced T cells expressing a group of CARs in vivo.

[0297] FIG. 7 shows the functional in vitro characterization of the CAR modified T cells used for the in vivo experiments.

[0298] FIG. 8 shows the regulation of the avidity of a group of CARs by heterodimerization and the sensitivity of CAR T cells depending on whether the target antigen is monomeric or dimeric.

[0299] FIG. 9 shows the regulation of the avidity of a group of CARs by VEGF.

[0300] FIG. 10 shows the generation and function of an affibody-based group of CARs directed against HER2.

[0301] FIG. 11 shows groups of CARs consisting of three or four CAR molecules.

[0302] FIG. 12 shows groups of CARs comprising different co-stimulatory domains.

[0303] FIG. 13 shows the expression of CAR molecules comprising different rcSso7d and affibody based binding moieties fused to different CAR signalling backbones.

[0304] FIG. 14 shows the schematics of the design of different CAR molecules.

[0305] FIG. 15 shows the amino acid sequences of different CAR molecules.

EXAMPLES

[0306] FIG. 1: Schematics of exemplary architectures of the group of CARs. FIG. 1A schematically shows the basic architecture of a CAR molecule of a group of CARs in the version in which the antigen binding moiety is integrated into the CAR molecule (left) and in the version in which the CAR molecule comprises a binding site which binds to a binding site in another polypeptide comprising an antigen binding moiety (right). Low affinity interaction occurs either between the antigen binding moiety and the target antigen or between the binding site in the CAR molecule and a binding site in the other polypeptide binding to the binding site in the CAR molecule. At least one of the CAR molecules of a group must comprise at least one signalling region comprising either at least one ITAM or at least one ITIM. In the shown example of a CAR molecule, the endodomain exemplarily comprises a single signalling region. The lines between each component of the shown CAR molecules indicate optional linkers. The dimerization domains (of which at least one is mandatory for each CAR molecule of the group) and optional additional domains or components are not indicated.

[0307] FIG. 1B schematically illustrates how many CAR molecules in groups of CARs consisting of two, three or four CAR molecules can comprise at least one signalling region (totality of the signalling regions of a given CAR molecule is symbolized by a white box). Of the CAR molecules comprising at least one signalling region either all or only some, but at least one CAR molecule comprise at least one ITAM or alternatively at least one ITIM. For simplicity, the ectodomains and dimerization domains and optional additional domains or components are not shown. The lines between each component of the shown CAR molecules indicate optional linkers.

[0308] FIG. 1C schematically illustrates the arrangement of signalling regions with the example of a group of CARs consisting of two CAR molecules. The depicted examples show only some of the possible combinations of the different arrangements. For example, a CAR molecule may comprise two or more ITAM-comprising signalling regions or two or more co-stimulatory signalling regions. In inhibitory groups of CARs (not shown) the ITAM-comprising signalling regions are substituted by ITIM-comprising signalling regions and co-stimulatory signalling regions are either omitted or substituted by other inhibitory domains. For simplicity, the ectodomains, the dimerization domains and optional additional domains or components are not shown. The lines between each component of the shown CAR molecules indicate optional linkers.

[0309] FIGS. 1D-1N schematically exemplify how dimerization domains can be used for non-covalent complexation of groups of CARs. The depicted examples show only some of the possible arrangements. In the depicted examples different pairs of homo- and heterodimerization domains are shown at different positions in the CAR molecules each exemplarily comprising one or two signalling regions. In a preferred example (FIG. 1E, left) a group of CARs consisting of three CAR molecules comprises only the two members of a single pair of heterodimerization domains and correspondingly can be regulated by a single kind of regulating molecule. For simplicity, the illustrations show only the signalling regions, the transmembrane domain and dimerization domains. The lines between some of the components of the shown CAR molecules indicate optional linkers. Similar arrangements of dimerization domains can also be incorporated extracellularly.

[0310] FIG. 1O exemplifies non-covalent complexation of a group of CARs by an extracellular soluble factor acting as a regulating molecule. The regulating molecule is schematically exemplified with a monomeric protein and with proteins which natively either homo- or heterodimerize. The shown CAR molecules exemplarily comprise only one signalling region. Optional additional dimerization domains and optional additional domains or components are not indicated. The order of the antigen binding domains (or binding sites) and the dimerization domains may also be inverted. That is, the antigen binding domains (or binding sites) may also be more proximal to the plasma membrane than the dimerization domains. The lines between each component of the shown CAR molecules indicate optional linkers.

[0311] FIG. 2 shows the K.sub.d values of different rcSso7d-based antigen binding moieties (fused to superfolder GFP (sfGFP)) towards human EGFR as determined by three complementary methods: (a) Flow cytometric quantification of the amount of the different sfGFP-fusion proteins bound to Jurkat T cells expressing high levels of truncated EGFR (tEGFR), (b) and (c) surface plasmon resonance (SPR) analysis by using a matrix coated with a chimeric EGFR protein comprising the extracellular domain of EGFR fused to the Fc domain of IgG1. Affinities were either determined in a kinetic (b) or a steady-state analysis mode (c).

[0312] FIG. 3: Extracellular disulphide bond forming cysteines prevent the exploitation of the avidity effect for regulation of CAR function. (A) Schematic representation of the architecture of the CAR signalling backbones “Cys” for S-8cys-BB-3z (“Cys”) and “Ser” for S-8ser-BB-3z (“Ser”), which are capable for disulphide bond formation or not, respectively. These signalling backbones were fused to the different rcSso7d-based antigen binding moieties and expressed for functional testing in primary human T cells. (B) Typical CAR expression (shown with rcSso7d variant “E11.4.1-WT” fused to either the “Cys” or “Ser” CAR backbone) 20 hours after electroporation of 5 μg of mRNA in primary human T cells. T cells expressing no transgene (“no CAR”) served as negative controls. (C) Expression of tEGFR in Jurkat cells, which were used as target cells, 20 hours after electroporation of 3 μg tEGFR-encoding mRNA. Jurkat T cells without construct (“no construct”) and a respective isotype control (“isotype control”) served as negative controls. The function of the CARs was tested by determining the capacity of primary human T cells modified with the different CARs for target cell killing (D) and IFN-γ production (E). The T cells of four different donors (indicated by different symbols) were electroporated with 5 μg mRNA of the indicated CAR construct and co-cultured on the next day with Jurkat T cells (electroporated with 3 μg of tEGFR-mRNA) for 4 hours at 37° C. at an effector:target (E:T) ratio of 2:1. T cells without CAR (“no CAR”) served as negative controls.

[0313] FIG. 4: Dimerization/multimerization of scFv-comprising CAR molecules prevents the exploitation of the avidity effect for regulation of CAR function. (A) Schematic representation of the CARs used in the experiments. (B, C and D) Expression of CAR constructs in human primary T cells 20 hours after electroporation of 5 μg the respective mRNAs. T cells without a CAR (“no CAR”) served as a negative control. (E) Expression of tHER2 in Jurkat cells, which were used as target cells, 20 hours after electroporation of 5 μg tEGFR-encoding mRNA. Jurkat T cells without construct (“no construct”) served as negative controls. CAR T cells were co-cultured with Jurkat-tHER2 cells for 4 hours at 37° C. at an E:T ratio of 2:1. FIGS. 4F and 4G show the capacity of the CARs in which the scFv 4D5-5 was fused to the two different CAR signalling backbones (capable for disulphide bond formation or not, i.e., “Cys” for 4D5-5-8cys-BB-3z (SEQ ID NO: 59) and “Ser” for 4D5-5-8ser-BB-3z (SEQ ID NO: 60)) to trigger cytotoxicity (G) and IFN-γ production (F). The function of CARs in which V.sub.H and V.sub.L were fused onto two separate polypeptides (“V.sub.H and V.sub.L”; 4D5-5(split)-8ser-BB-FKBP(36V)-3z (SEQ ID NO: 56 and SEQ ID NO: 57)) is shown in (H). Primary T cells expressing the CAR 4D5-5-8ser-BB-3z (SEQ ID NO: 60), in which the monomeric signalling backbone was fused to the 4D5-5-scFv, served as a positive control. CAR T cells of three different donors (indicated by different symbols) were co-cultured with a mixture of tHER2-transfected and non-transfected Jurkat T cells for 4 hours at 37° C. at an E:T ratio of 4:1:1 (T cells:tHER2.sup.pos Jurkat cells:tHER2.sup.neg Jurkat cells) and the ratio of viable tHER2.sup.pos and tHER2.sup.neg target cells was determined by flow cytometry. To induce dimerization, the T cells were pre-treated with the dimerization agent AP20187 (10 nM, 30 minutes, 37° C.). Treatment with the same concentration of DMSO served as a control condition. T cells without CAR (“no CAR”) and T cells expressing only the V.sub.H chain (“4D5-5(V.sub.H)-8ser-BB-FKBP(36V)-3z” (SEQ ID NO: 57)) served as negative controls.

[0314] FIG. 5: Regulating CAR avidity by homodimerization of CAR molecules. (A) Schematic representation of the group of CARs complexed by the regulating molecule AP20187. In the shown example an rcSso7d-based antigen binding moiety directed against EGFR was used as an antigen binding moiety. (B) Expression of CAR molecules comprising an antigen binding moiety with either high or low affinity to EGFR (i.e., S(WT)-8ser-BB-FKBP(36V)-3z “E11.4.1-WT” (SEQ ID NO: 49) and S(G32A)-8ser-BB-FKBP(36V)-3z “E11.4.1-G32A” (SEQ ID NO: 46), respectively) in primary human T cells 14 days after activation by anti-CD3/CD28-antibody coated beads (i.e., 13 days after stable transduction with the respective CAR constructs). T cells without a CAR (“no CAR”) served as negative controls. (C) Expression of EGFR in the target cell line “A431-fLuc”. Unstained cells (“unstained”) and a respective isotype control (“isotype control”) served as negative controls. (D) Capacity of primary human T cells transduced with different CAR constructs for killing of the luciferase-expressing target cell line A431-fLuc. Primary T cells of three donors (indicated by different symbols) were stably transduced with vectors coding for either the rcSso7d-based CARs S(WT)-8ser-BB-FKBP(36V)-3z (“E11.4.1-WT”) and S(32)-8ser-BB-FKBP(36V)-3z (“E11.4.1-G32A”)) or the scFv-based CAR directed against CD19 (CD19-8cys-BB-3z “CD19-BBz” (SEQ ID NO: 58)), which served as a negative control. T cells without CAR (“no CAR”) served as an additional negative control. AP20187 served as regulating molecule for non-covalent dimerization of “E11.4.1-WT” (S(WT)-8ser-BB-FKBP(36V)-3z) and “E11.4.1-G32A” (S(G32A)-8ser-BB-FKBP(36V)-3z). Dimerization was induced by pre-treatment of the T cells with 10 nM AP20187 for 30 minutes at 37° C. Treatment with the same concentration of DMSO served as a control condition. Cytotoxicity of the modified T cells was determined by quantifying viable, luciferase expressing A431-fLuc target cells after co-culture for 4 hours at 37° C. at an E:T ratio of 10:1.

[0315] FIG. 6: Regulation of the function of stably transduced CAR T cells in vivo in an NSG mouse model. (A) Efficacy of dimerization induced activation of CARs in an in vivo NSG mouse model. 9-16 weeks old NSG mice were intravenously injected with 0.5×10.sup.6 Nalm6 cells that were stably transduced with a vector coding for luciferase and tEGFR (“Nalm6-tEGFR-fLuc”). Three days later, the NSG mice were treated by intravenous administration of 10×10.sup.6 CAR T cells (14 days after activation by anti-CD3/CD28-antibody coated beads, i.e., 13 days after stable transduction). An injection with phosphate buffered saline (PBS) served as a control condition. Five NSG mice were used for each treatment group, except for the group treated with S(WT)-8ser-BB-FKBP(36V)-3z (SEQ ID NO: 49), which consisted of four mice. Dimerization was induced by intraperitoneal administration of 2 mg/kg of the homodimerization agent AP20187 over a period of 11 days (days of injection indicated by arrows). Groups that did not receive a dimerization agent, received the respective vehicle solution intraperitoneally as a control condition. Tumour size (mean of each treatment group) is shown as total photon flux determined within the region of interest that encompassed the entire body of the NSG mice. (B) Shows that postponed administration of the dimerizer in mice, which were treated with the S(32)-8ser-BB-FKBP(36V)-3z (SEQ ID NO: 46) CAR T cells but did not receive the dimerizer AP20187 until day 11, results in efficient CAR activation and control of tumour growth.

[0316] FIG. 7: Regulation of the function of stably transduced CAR T cells in vitro. (A) Expression of CAR molecules comprising an antigen binding moiety with either high or low affinity to EGFR (i.e., S(WT)-8ser-BB-FKBP(36V)-3z “E11.4.1-WT” (SEQ ID NO: 49) and S(G32A)-8ser-BB-FKBP(36V)-3z “E11.4.1-G32A” (SEQ ID NO: 46), respectively) in primary human T cells 14 days after activation by anti-CD3/CD28-antibody coated beads (i.e., 13 days after stable transduction with the respective CAR constructs). T cells without a CAR (“no CAR”) served as negative controls. (B) Expression of the anti-CD19 CAR CD19-8cys-BB-3z “CD19-BBz” (SEQ ID NO: 58) in primary human T cells 14 days after activation by anti-CD3/CD28-antibody coated beads (i.e., 13 days after stable transduction). T cells without a CAR (“no CAR”) served as negative controls. (C and D) Expression of tEGFR in Nalm6-fLuc and Nalm6-tEGFR-fLuc cells, respectively. Unstained cells and a respective isotype control served as negative controls. (E and F) Lysis of Nalm6-fLuc and Nalm6-tEGFR-fLuc cells by primary human T cells expressing different CAR molecules. Primary T cells of three donors (indicated by different symbols) were stably transduced with vectors coding for either the rcSso7d-based CARs (S(WT)-8ser-BB-FKBP(36V)-3z “E11.4.1-WT” and S(G32A)-8ser-BB-FKBP(36V)-3z “E11.4.1-G32A”) or the scFv-based CAR directed against CD19 (CD19-8cys-BB-3z “CD19-BBz”), which served as a positive control. T cells without CAR (“no CAR”) served as a negative control. AP20187 served as regulating molecule for non-covalent dimerization of “E11.4.1-WT” (S(WT)-8ser-BB-FKBP(36V)-3z) and “E11.4.1-G32A” (S(G32A)-8ser-BB-FKBP(36V)-3z). Dimerization was induced by pre-treatment of the T cells with 10 nM AP20187 for 30 minutes at 37° C. Treatment with the same concentration of DMSO served as a control condition. Cytotoxicity of the modified T cells was determined by quantifying viable, luciferase expressing target cells after co-culture for 4 hours at 37° C. at an E:T ratio of 10:1.

[0317] FIG. 8: Regulating the avidity of a group of CARs by heterodimerization of CAR molecules and influence of the colocalization of target antigens on the sensitivity of the CAR T cells. Depicted in (A) is the expression of a tEGFR fusion protein (“FKBP F36V”, for conditional homodimerization) in Jurkat cells, which served as target cells, 20 hours after electroporation of 5 μg of the respective mRNA. Unstained cells (“mock”) and a respective isotype control (“isotype”) served as negative controls. (B and C) Correlation of the expression levels of tEGFR in Jurkat T cells electroporated with different amounts of the respective mRNAs. (D and E) Expression of the different CARs in primary human T cells 20 hours after electroporation of 5 μg of the respective mRNA. The rcSso7d variants with high and low affinity (“E11.4.1-WT” and “E11.4.1.-G32A, respectively) were fused to both a 8ser-BB-FKBP-3z and a 8ser-BB-FRB-3z backbone for conditional heterodimerization by the regulating molecule AP21967. The expression of the resulting constructs (S(G32A)-8ser-BB-FKBP-3z (SEQ ID NO: 48), S(G32A)-8ser-BB-FRB-3z (SEQ ID NO: 47), S(WT)-8ser-BB-FKBP-3z (SEQ ID NO: 50), S(WT)-8ser-FRB-3z (SEQ ID NO: 51)) was detected via their integrated Tags (Strep II for 8ser-BB-FRB-3z and FLAG for 8ser-BB-FKBP-3z). (F) Molecular architecture of the experimental system used for determining the sensitivity of dimerization-induced activation of CAR T cells in response to target cells expressing either monomeric or dimeric target antigen tEGFR. In this system, the CAR molecules were conditionally heterodimerized by the regulating molecule AP21967 and the target antigen tEGFR was conditionally homodimerized by AP20187. Depicted in (G) and (H) is the capacity of the CARs for triggering cytotoxicity and IFN-γ production, respectively, (mean±standard deviation, n=3, three different donors) in primary human T cells in dependence of both the number of tEGFR molecules expressed in the target cells and the presence or absence of AP21967 and AP20187. Primary T cells from three donors were electroporated with 5 μg mRNA of each CAR molecule (i.e., either the low-affinity constructs S(G32A)-8ser-BB-FKBP-3z plus S(G32A)-ser8-BB-FRB-3z or the high-affinity constructs S(WT)-8ser-BB-FKBP-3z plus S(WT)-8ser-BB-FRB-3z, as indicated) and co-cultured with Jurkat T cells expressing varying amounts of tEGFR. Co-culture was done for 4 hours at 37° C. at an E:T ratio of 4:1:1 (T cells:tEGFR.sup.pos Jurkat cells:tEGFR.sup.neg Jurkat cells). The cytotoxicity of the CAR T cells was determined by flow cytometric quantification of the ratio of viable tEGFR.sup.pos and tEGFR.sup.neg target cells. To induce CAR-dimerization, the primary human T cells were pre-treated with AP21967 (500 nM, 30 minutes, 37° C.) and to induce EGFR-dimerization, Jurkat T cells were pre-treated with AP20187 (10 nM, 30 minutes, 37° C.). Treatment with the same concentration of ethanol (in the case of AP21967) or DMSO (in the case of AP20187) served as a control condition.

[0318] FIG. 9: Generation of a CAR which can be regulated by VEGF as example of an extracellular factor accumulating in the tumour microenvironment. In the shown example the soluble factor VEGF was used as a regulating molecule and the EGFR-specific rcSso7d-based binder E11.4.1-G32A was used as antigen binding moiety. The schematic of the architecture of a respective CAR is shown in (A) and (B). Expression of the target antigen tEGFR in Jurkat T cells 20 hours after electroporation of 5 μg of mRNA is depicted in (C). Expression of the two polypeptides in primary human T cells was detected using an anti-IgG1-antibody (D). The anti-Strep II-tag antibody was additionally used for detection of the CAR molecule containing the VEGF binding site (Janus-CT6-Fc domain without transmembrane domain) and for the control CAR without IgG-Fc domains. Cytotoxicity triggered by the different CARs in primary T cells was determined in a FACS-based cytotoxicity assay (E). CAR T cells from three different donors (indicated by different symbols) were co-cultured with tEGFR-transfected Jurkat cells for 4 hours at 37° C. at an E:T ratio of 4:1:1 (T cells:tEGFR.sup.pos Jurkat cells:tEGFR.sup.neg Jurkat cells). Dimerization of the CARs was induced by pre-treatment of the T cells with VEGF (concentration as indicated, 30 minutes, 37° C.). T cells without a CAR (“no CAR”) served as a negative control and T cells with the CAR S(WT)-8ser-BB-FKBP(36V)-3z served as a positive control.

[0319] FIG. 10: Screening of affibody-based binding moieties suited for use in a group of CARs according to the present invention. The affinity of the affibody zHER2-WT was reduced by replacing all amino acids that are potentially involved in epitope binding by alanine. The different variants of zHER2-WT were fused to the CAR signalling backbone 8ser-BB-FKBP(36V)-3z containing the intracellular homodimerization domain FKBP(F36V) for conditional dimerization. The architecture of these CARs is shown in (A). (B and C) Expression of the target antigen tHER2 in Jurkat T cells and the expression of affibody-based CARs in primary T cells, respectively. Primary T cells and Jurkat T cells were electroporated with 5 μg mRNA coding for the respective construct and expression was determined 20 hours after electroporation. Primary T cells and Jurkat T cells expressing no construct (“no CAR” and “no construct”, respectively) were used as negative controls. (D) Expression of 13 different affibody-based CARs in primary T cells (“L9A”, “R10A”, “Q11A”, “Y13A”, “W14A”, “Q17A”, “W24A”, “T25A”, “S27A”, “R28A”, “R32A”, “Y35A” and “zHER2-WT”, respectively) (the sequences of the affibody based antigen binding moieties fused to the CAR signalling backbone are depicted in SEQ ID NO:26 to SEQ ID NO:38). Primary T cells were electroporated with 5 μg mRNA coding for the respective construct and expression was determined 20 hours after electroporation. T cells expressing no construct (“no CAR”) were used as a negative control. (E) Dimerization-induced activation of affibody-based CARs. Primary T cells from two donors (indicated by different symbols) were electroporated with 5 μg for each construct. Specific lysis of target cells was determined in a luciferase-based cytotoxicity assay after co-incubation of the different CAR T cells with tHER2-transfected (4 hours at 37° C. at an E:T ratio of 2:1). Dimerization of the CARs was induced by pre-treatment of the T cells with AP20187 (10 nM, 30 minutes, 37° C.). Treatment with the same concentration of DMSO served as a control condition. T cells without a CAR (“no CAR”) served as a negative control. FIG. 10F shows the K.sub.d values of respective Affibody-based binding moietiess (fused to superfolder GFP (sfGFP)) towards human HER2 as determined by SPR analysis by using a matrix coated with a chimeric HER2 protein comprising the extracellular domain of HER2 fused to the Fc domain of IgG1. Affinities were determined in a kinetic mode.

[0320] FIG. 11: Functional characterization of groups of CARs consisting of three and four CAR molecules. (A and B) Schematic representation of groups of CARs consisting of three or four CAR molecules, respectively. (C and D) Expression of trimeric and tetrameric groups of CARs in Jurkat T cells 20 hours after electroporation of 5 μg mRNA of each construct. Jurkat T cells expressing no construct (“no CAR”) were used as a negative control.

[0321] FIG. 12: Expression and function of groups of CARs comprising different co-stimulatory molecules. (A) Schematic representation of a group of CARs that consists of two CAR molecules each containing either the co-stimulatory domain of CD28 or ICOS or OX40 in its co-stimulatory signalling region. (B and C) Expression of the CAR molecules (red histograms) 20 hours after electroporation of 5 μg mRNA in Jurkat T cells or primary human T cells, respectively. Jurkat T cells or primary human T cells, respectively, expressing no construct (“no CAR”) were used as a negative control (filled blue histograms). (D) Induction of NF-κB and NF-AT promoters in Jurkat T cells electroporated with 5 μg of mRNA encoding S(G32A)-8ser-OX40-FKBP(36V)-3z. These cells were either co-cultured or not for further 20 hours in the presence or absence of AP20187 with Jurkat T cells electroporated with 5 μg of tEGFR encoding mRNA. Induction of NF-κB and NF-AT promoters in the CAR expressing reporter cells was detected by flow cytometric analysis of the expression of enhanced green fluorescent protein (eGFP) and cyan fluorescent variant of GFP (CFP), respectively. The cytotoxicity of primary human T cells expressing the indicated CAR molecules is shown in (E). 20 hours after electroporation of 5 μg mRNA encoding the respective CAR molecules the T cells were co-cocultured with target cells for further 4 or 20 hours at 37° C. at an E:T ratio of 4:1:1 (T cells:tEGFR.sup.pos Jurkat cells:tEGFR.sup.neg Jurkat cells).

[0322] Dimerization of the CAR molecules of the group was induced by pre-treatment of the Jurkat cells (D) and primary human T cells (E) with 10 nM AP20187 for 30 minutes at 37° C. Treatment with the same concentration of DMSO served as a control condition.

[0323] FIG. 13: Expression of CAR molecules comprising rcSso7d and affibody based binding moieties fused to different CAR signalling backbones. (A) Expression of CAR constructs “Myc-S(18.4.2)-8cys-BB-3z” (SEQ ID NO: 39), “S(18.4.2)-8cys-BB-3z” (SEQ ID NO: 40) and “S(18.4.2)-G4S-8cys-BB-3z” (SEQ ID NO: 41) in primary human T cells 20 hours after electroporation of 5 μg of the respective mRNAs. Primary T cells without a CAR served as negative controls (filled histogram). Expression was detected using a fusion protein consisting of the extracellular domain of human EGFR and the Fc domain of IgG1 and an anti-human-IgG1-antibody. (B) Expression of CAR constructs “S(G32A)-G4S-myc-8cys-BB-3z” (SEQ ID NO: 42), “S(G32A)-G4S-StrepII-8cys-BB-3z” (SEQ ID NO: 43) and “S(G32A)-G4S-his-8cys-BB-3z” (SEQ ID NO: 45) in primary human T cells 20 hours after electroporation of 5 μg of the respective mRNAs. Primary T cells without a CAR served as negative controls (filled histogram). CAR expression was detected using either an anti-c-myc antibody, an anti-Strep II antibody, or an anti-hexahistidine antibody, respectively. (C) Expression of CAR constructs “S(WT)-8cys-BB-3z” (SEQ ID NO: 74), “S(G25A)-8cys-BB-3z” (SEQ ID NO: 72), “S(G32A)-8cys-BB-3z” (SEQ ID NO: 43), “S(WT)-8ser-BB-3z” (SEQ ID NO: 75), “S(G25A)-8ser-BB-3z” (SEQ ID NO: 73), “S(G32A)-8ser-BB-3z” (SEQ ID NO: 44) in human primary T cells 20 hours after electroporation of 5 μg of the respective mRNAs. Primary T cells without a CAR served as negative controls (filled histogram). Expression was detected using an anti-Strep II antibody. (D) Expression of CAR constructs “S(G32A)-8ser-BB-FKBP(36V)-3z” (SEQ ID NO: 46), “S(G32A)-8ser-BB-FRB-3z” (SEQ ID NO: 47), “S(G32A)-8ser-BB-FKBP-3z” (SEQ ID NO: 48), “A(WT)-8ser-BB-FKBP(36V)-3z” (SEQ ID NO: 52) in primary human T cells 20 hours after electroporation of 5 μg of the respective mRNAs. Primary T cells without a CAR served as negative controls (filled histogram). Expression was detected using either an anti-FLAG antibody, an anti-Strep II antibody, or an anti-hexahistidine antibody, as indicated.

[0324] FIG. 14 shows the schematics of the design of different CAR molecules. The corresponding amino acid sequences are shown in FIG. 15.

[0325] FIG. 15 shows the amino acid sequences of different CAR molecules.

Example 1: Generation of a Low-Affinity Single Domain Binding Moiety Based on rcSso7d for Use in a Group of CARs According to the Present Invention

[0326] The first example shows a strategy for generating an antigen binding moiety with low affinity that is suited for use as an antigen binding moiety in a group of CARs, according to the present invention. Reduced charge Sso7d (rcSso7d) is a charge-reduced version of a small (˜7 kDa) DNA-binding protein from the archaeon Sulfolobus solfataricus. Charge-reduction minimizes unspecific binding due to reduced electrostatic interactions. rcSso7d is a single-domain protein antigen binding moiety with high thermal stability and monomeric behaviour and therefore is an example of a suited binding scaffold. Starting from the well characterized antigen binding moiety rcSso7d E11.4.1, which binds to human EGFR with a K.sub.d of 19 nM (Traxlmayr et al., J Biol Chem. 2016; 291(43):22496-22508), we generated low affinity mutants by performing an alanine scan in which we replaced all amino acids potentially involved in epitope binding by alanine, i.e. in each mutant one position involved in antigen binding was mutated to alanine. Mutants of rcSso7d E11.4.1 were fused to sfGFP and expressed as soluble proteins in a bacterial expression system. The schematics of the architecture of the fusion proteins are shown in FIG. 14G. Binding affinities were determined (i) by performing titration experiments of the soluble fusion proteins of the binding moieties with sfGFP on Jurkat T cells that were engineered by mRNA electroporation to express high levels of the respective target antigen EGFR, and (ii) by performing SPR experiments on protein A chips loaded with the extracellular domain of EGFR fused to IgG-Fc. The result of the alanine scan and the obtained affinities of the antigen binding moieties are shown in FIG. 2.

Alanine-Scanning of Protein Antigen Binding Moieties:

[0327] Site-directed mutagenesis of all amino acids involved in epitope binding was performed using the QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent Genomics), according to the manufacturer's instructions. Primers were designed using the QuikChange Primer Design software (Agilent Genomics) and oligonucleotides were synthesized by Biomers.

Expression and Purification of rcSso7d-Based Antigen Binding Moieties:

[0328] Binding scaffolds were expressed as sfGFP fusion proteins (consisting of an N-terminal hexahistidine tag followed by either rcSso7d or the Affibody and sfGFP) using the pE-SUMO vector (Life Sensors). The nucleotide sequence that encodes the sfGFP reporter protein was obtained from Addgene (plasmid #54737). Briefly, Escherichia coli cells (Tuner DE3) were transformed with sequence-verified plasmids using heat shock transformation. After overnight cultivation at 37° C., cultures were diluted 1:100 in terrific broth (TB) medium (12 g/L tryptone, 24 g/L yeast extract, 4% glycerol, 2.31 g/L KH.sub.2PO.sub.4 and 16.43 g/L K.sub.2HPO.sub.4*3H.sub.2O) supplemented with kanamycin (50 μg/mL) and incubated at 37° C. while shaking. When cultures reached an A.sub.600 of roughly 2, expression of the transgene was induced by addition of 1 mM of isopropyl β-D-1-thiogalactopyranoside (IPTG) and cells were further cultured overnight at 20° C. Cells were harvested by centrifugation (5000 g, 20 minutes, 4° C.), resuspended in sonication buffer (50 mM sodium phosphate, 300 mM NaCl, 3% glycerol, 1% Triton X-100, pH 8.0), sonicated (2×90 seconds, duty cycle 50%, amplitude set to 5) and centrifuged again to remove cell debris. Hexahistidine-tagged fusion proteins were purified from crude cell extracts using TALON metal affinity resin (Clontech Laboratories). After addition of 10 mM imidazole, the sonicated supernatants were applied onto the resin twice, followed by washing step with equilibration buffer (50 mM sodium phosphate, 300 mM NaCl, pH 8.0) with increasing amounts of imidazole (5-15 mM). Binding scaffolds were eluted by applying equilibration buffer supplemented with 250 mM imidazole. After buffer exchange to PBS using Amicon Ultra-15 10K centrifugal filters (Merck Millipore), concentrations were determined by measuring the absorbance at 280 nm using the respective molar absorption coefficient and finally proteins were directly frozen at −80° C.

Maintenance of Human Cell Lines:

[0329] Jurkat T cells were a gift from Dr. Sabine Strehl at the Children's Cancer Research Institute (CCRI) and were maintained in RPMI-1640 (Thermo Scientific) supplemented with 10% FCS (Sigma Aldrich) and 1% penicillin-streptomycin (Thermo Scientific). Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads (Thermo Scientific), a flow cytometer-based cell counting platform.

In Vitro Transcription and Electroporation of mRNA:

[0330] In vitro transcription was performed using the mMessage mMachine T7 Ultra Kit (Ambion) according to the manufacturer's instructions. 50-200 ng of column purified PCR product was used as a reaction template. The resulting mRNA was purified with an adapted protocol using the RNeasy column purification kit (Qiagen). Briefly, the mRNA solution was diluted with a mixture of RLT buffer (Qiagen), ethanol (Merck) and 2-mercaptoethanol (Merck). The mixture was loaded onto an RNeasy column and purification was performed according to the manufacturer's instructions. Elution was performed with nuclease-free water (Thermo Scientific) and purified mRNAs were frozen at −80° C. until electroporation. For transient transgene expression, Jurkat T cells were electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocols were used for the respective cell types: Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

Antibodies and Flow Cytometry:

[0331] Jurkat T cells were resuspended in FACS buffer (PBS (Thermo Scientific), 0.2% human albumin (CSL Behring) and 0.02% sodium azide) and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then directly processed with a BD LSRFortessa. Expression of the engineered target antigen tEGFR was detected either with a PE- or APC-conjugated anti-EGFR antibody (clone AY13, BioLegend). Analysis was done by the FlowJo software.

Determination of Binding Affinities on Cell Membranes:

[0332] Jurkat T cells were engineered to express high levels of the respective tumour antigen. Hence, 3 μg of mRNA coding for tEGFR were electroporated into Jurkat T cells one day prior to the co-culture with the effector cells. After washing in PBS, cells were resuspended in PBS containing 0.1% BSA (Sigma Aldrich) and incubated with varying concentrations of the binder proteins fused to sfGFP in order to determine the affinities of the antigen binding moieties towards the respective tumour antigen. After incubation for 1 hour at 4° C. while shaking, plates were centrifuged (450 g, 7 minutes, 4° C.), the supernatant was discarded and cells were acquired with a BD LSRFortessa. Cells were kept on ice to avoid endocytosis. The K.sub.d was obtained by curve fitting using Microsoft Excel (Microsoft Corporation).

Determination of Binding Affinities Using Surface Plasmon Resonance (SPR):

[0333] SPR experiments were performed with a Biacore T200 instrument (GE Healthcare). All experiments were conducted in degassed and filtered PBS, pH 7.4, containing 0.1% BSA and 0.05% Tween-20 (Merck Millipore), at 25° C. hEGFR-Fc (R&D) was immobilized on a Protein A sensor chip (GE Healthcare) at a flow rate of 10 μL/min for 60 seconds at a concentration 6.67 μg/mL. To determine the affinity of rcSso7d-based antigen binding moieties, five concentrations (depending on the expected K.sub.d of the antigen binding moiety) of the respective protein were injected at a flow rate of 30 μL/min for 15 seconds in the single-cycle kinetic mode, followed by a dissociation step (30 seconds). Regeneration was performed using 10 mM Glycine-HCl, pH 1.7 at a flow rate of 30 μL/min for 30 seconds. The K.sub.d was obtained by curve fitting using the Biacore T200 Evaluation Software (GE Healthcare).

Example 2: Extracellular Disulphide Bond-Forming Cysteines Prevent the Full Exploitation of Avidity Effects for Reversible Control of CAR Function

[0334] Extracellular disulphide-bond forming cysteines in extracellular hinge regions as e.g. CD8α can prevent the exploitation of the avidity effect according to present invention. This is demonstrated in example 2, in which the low affinity mutant of the binding moiety “E11.4.1 G32A” of example 1 was fused to CAR signalling backbones in which the two extracellular cysteine residues in the hinge region of CD8α (UniProt ID P01732, positions C164 and C181) were substituted by serine residues or not, respectively. Whereas the cysteine-containing CAR-variant (“Cys”) efficiently triggered T cell activation in response to target cells, the serine-containing variant (“Ser”) did not or only poorly trigger the T cells. This example thus illustrates the importance of preventing disulphide-bond formation for generating CAR molecules that are suited for use in a group of CARs according to the present invention. The schematics in FIG. 3A illustrate the design of the tested constructs. FIGS. 3B and 3C show the expression of the CARs and target antigens. Primary human T cells were electroporated with 5 μg mRNA for each construct and CAR expression was detected 20 hours after electroporation via a Strep II Tag. Jurkat T cells were electroporated with 3 μg mRNA encoding a truncated version of EGFR (tEGFR). Full length EGFR was truncated both N- and C-terminally to create a functionally inert human polypeptide that has diminished dimerization properties due to the inability to bind its natural ligand EGF and the absence of the kinase domain (Wang et al., Blood. 2011; 118(5):1255-1263). The resulting transgene consisted of the leader sequence of the granulocyte-macrophage colony-stimulating factor 2 receptor alpha subunit (GM-CSF-Ra) and amino acids 334 to 675 (Uniprot P00533) of human EGFR, comprising two extracellular membrane-proximal domains and the transmembrane domain. Transgene expression was detected 20 hours after electroporation via antibodies directed against EGFR. 20 hours after electroporation of the T cells, the function of the CARs was determined by using a luciferase-based cytotoxicity assay (FIG. 3D) and by quantifying cytokines released from the T cells using ELISA (FIG. 3E). FIGS. 3D and 3E show that the antigen binding moiety with the lowest tested affinity (“E11.4.1-G32A”) could trigger the T cells only upon bivalent interaction with the target cells, i.e, when fused to the CAR containing the cysteines in the CD8α hinge (“Cys”), but not or only poorly when fused to the CAR in which those cysteines were replaced by serine (“Ser”). Contrary, antigen binding moieties with increased affinities (E11.4.1-WT and E11.4.1-G25A) triggered potent cytotoxicity also upon monovalent interaction, i.e., when fused to the “Ser” CAR backbone.

Maintenance of Human Cell Lines:

[0335] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos s T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail (STEMCELL Technologies). Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO (Sigma Aldrich) until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads (Thermo Scientific) according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2 (Peprotech). Primary T cells were cultivated for at least 14 days before experiments were conducted. Jurkat T cells were a gift from Dr. Sabine Strehl at the CCRI and were maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

In Vitro Transcription and Electroporation of mRNA:

[0336] In vitro transcription was performed using the mMessage mMachine T7 Ultra Kit according to the manufacturer's instructions. 50-200 ng of column purified PCR product was used as a reaction template. The resulting mRNA was purified with an adapted protocol using the RNeasy column purification kit. Briefly, the mRNA solution was diluted with a mixture of RLT buffer, ethanol and 2-mercaptoethanol. The mixture was loaded onto an RNeasy column and purification was performed according to the manufacturer's instructions. Elution was performed with nuclease-free water and purified mRNAs were frozen at −80° C. until electroporation. For transient transgene expression, primary T cells or Jurkat T cells were electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocols were used for the respective cell types: primary T cells (square wave protocol, 500 V, 5 ms and 4 mm cuvettes), Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

Antibodies and Flow Cytometry:

[0337] Primary human T cells or tumour cell lines were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide) and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) as a primary antibody and a PE- or APC-conjugated secondary antibody (eBioscience). Expression of the engineered target antigen tEGFR was detected either with a PE- or APC-conjugated anti-EGFR antibody (clone AY13, BioLegend). Analysis was done by the FlowJo software.

Construction of Transgene Constructs:

[0338] The nucleotide sequences encoding the signal peptide CD33, the human CD8α hinge, human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain and the CD3ζ ITAM signalling domain were synthesized by GenScript. Sequences encoding the extracellular and transmembrane domain of EGFR was obtained from Addgene (plasmid #11011). Insertion of a Strep II tag (NWSHPQFEK) and flexible linkers was performed by PCR. Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix (New England BioLabs), according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

Luciferase-Based Cytotoxicity Assay:

[0339] Luciferase-expressing tumour cells were co-cultured with CAR T cells at an E:T cell ratio of 2:1 with 10,000 target cells/well in white round-bottom 96 well plates (Sigma Aldrich) for 4 hours at 37° C. in cytotoxicity assay medium, consisting of phenol-free RPMI (Thermo Scientific), 10% FCS, 1% L-Glutamine (Thermo Scientific) and 1% penicillin-streptomycin. Finally, remaining living cells were quantified by determination of the residual luciferase activity of the co-culture. After equilibration to room temperature for 10 minutes, luciferin was added to the cell suspension (150 μg/mL final concentration; Perkin Elmer) and luciferase activity was measured 20 minutes later using the ENSPIRE Multimode plate reader. The percentage of specific lysis was determined with the following formula:

% specific lysis=100−((RLU from well with effector and target cell co-culture)/(RLU from well with target cells only)×100)).

Cytokine Release by CAR T Cells:

[0340] Cytokine secretion of primary CAR T cells was assessed by co-cultivation with target cells at E:T ratios of 1:1 or 2:1 in flat-bottom 96 well plates for 4 hours or 24 hours at 37° C. In some experiments, released cytokines were quantified in the supernatants from the co-culture experiments for determining cytotoxicity. The supernatants were centrifuged (1600 rpm, 7 minutes, 4° C.) to remove remaining cells and debris and were subsequently frozen at −80° C. For analysis of secreted IFN-γ, ELISA was performed using the Human IFN gamma ELISA Ready-SET-Go!® kit (eBioscience) according to the manufacturer's instructions. Measurements were conducted using the ENSPIRE Multimode plate reader.

Example 3: Single-Chain Variable Fragments (scFv) can Trigger CAR Clustering in Cell Membranes and Thereby Prevent the Exploitation of the Avidity Effect for Reversible Control of CAR Function

[0341] The third example demonstrates that the integration of scFv-based binding moieties in CAR molecules can prevent the exploitation of the avidity effect for specific recognition of antigen combinations. The schematics of the CAR constructs shown in FIG. 4A illustrate the design of the tested CAR variants (4D5-5-8cys-BB-3z, 4D5-5-8ser-BB-3z, 4D5-5(split)-8ser-BB-FKBP(36V)-3z). In the shown example the scFv 4D5-5 directed against HER2 was used as an antigen binding moiety and incorporated into either a monomeric (“Ser”) or a dimeric (“Cys”) CAR signalling backbone. FIG. 4B shows the expression of the CARs in primary T cells. The effective binding affinity for the scFv 4D5-5 was reported to be 1.1 μM (Liu et al., Cancer Res. 2015; 75(17):3596-3607), which is comparable to the affinity of E11.4.1-G32A. Jurkat T cells expressing a truncated form of HER2 (tHER2) served as a target cell line (FIG. 3E). IFN-γ secretion by the CAR T cells (FIG. 3F) and lysis of target cells (FIG. 3G) triggered by the serine-containing CAR “4D5-5-8ser-BB-3z” is only slightly diminished compared to the cysteine-containing CAR “4D5-5-8cys-BB-3z”, despite the low-affinity scFv, which is in stark contrast to the observations with the low-affinity rcSso7d-based antigen binding moiety in the CAR S(G32A)-8ser-BB-3z (SEQ ID NO: 44)) of Example 2 (FIG. 3). As observed in diabody formation, V.sub.H and V.sub.L linked together in an scFv at the surface of T cells can not only dimerize within the same single-chain molecule (i.e. within the same CAR molecule), but can also form intermolecular links (i.e. between different CAR molecules). This would mediate dimerization or even oligomerization as reported for purified scFv proteins (Atwell et al., Protein Eng. 1999; 12(7):597-604) and would explain the previously observed CAR clustering (Long et al., Nat Med. 2015; 21(6):581-590). Ultimately, this dimerization or oligomerization of the scFvs would result in the formation of bivalent or multivalent CARs, even when based on a monomeric CAR backbone. Accordingly, cutting the linker between V.sub.H and V.sub.L would prevent oligomerization and therefore activation of low-affinity CARs based on monomeric CAR backbones. We further assumed that V.sub.H and V.sub.L domains without linker still can at least partially heterodimerize on the surface of T cells to form functional V.sub.H/V.sub.L-heterodimers (i.e., Fvs). Provided that there is no unspecific stickiness between Fvs, these Fvs due to their low affinity (Kd 1.1 μM in the case of 4D5-5) should be able to trigger T cell activation only upon controlled dimerization of two Fvs (via an FKBP F36V-domain intracellularly fused to the V.sub.Hcarrying chain). Indeed, FIG. 4H illustrates that this is the case, as could be shown by separating V.sub.H and V.sub.L of the low-affinity scFv 4D5-5 onto two separate membrane-anchored molecules (SEQ ID NO: 56 and SEQ ID NO: 57). As predicted, CAR T cells expressing these constructs (FIGS. 4C and 4D) were not activated by tHER2.sup.pos target cells. However, the T cells were activated by those target cells when the V.sub.H construct was homodimerized by AP20187 (FIG. 4H). This T cell activation in the presence of dimerizer confirmed that the two separate constructs indeed formed functional Fvs on the T cell surface and that the lack of activation in the absence of dimerizer was due to the monovalent nature of the Fvs. This agrees with the low affinity of 4D5-5 and with the observations obtained with the rcSso7d-based antigen binding moiety in Example 2. For comparison we roughly adjusted the expression of the scFv-version (i.e. V.sub.H and V.sub.L connected by the linker “218”) to the expression levels obtained with the V.sub.H construct (FIG. 4C), which despite the low expression still resulted in strong activation of the CAR T cells (FIG. 4H). Together, the data in FIG. 4 strongly suggest that at least certain versions of scFvs partly dimerize (or oligomerize) by intermolecular heterodimerization of V.sub.H and V.sub.L between neighbouring molecules on the T cell surface. Similar to dimerization or oligomerization caused by cysteine amino acid residues (discussed above), uncontrolled dimerization or oligomerization of CAR molecules mediated by at least certain scFvs variants can cause homodimerization or homooligomerization independent of a regulating molecule. Therefore, in preferred embodiments, the antigen binding moieties of the CAR molecules of the group of CARs, or the antigen binding moieties of the other polypeptides binding to the CAR molecules of the group, according to the present invention are not scFvs.

Maintenance of Human Cell Lines:

[0342] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail. Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. Jurkat T cells were a gift from Dr. Sabine Strehl at the CCRI and were maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

In Vitro Transcription and Electroporation of mRNA:

[0343] In vitro transcription was performed using the mMessage mMachine T7 Ultra Kit according to the manufacturer's instructions. 50-200 ng of column purified PCR product was used as a reaction template. The resulting mRNA was purified with an adapted protocol using the RNeasy column purification kit. Briefly, the mRNA solution was diluted with a mixture of RLT buffer, ethanol and 2-mercaptoethanol. The mixture was loaded onto an RNeasy column and purification was performed according to the manufacturer's instructions. Elution was performed with nuclease-free water and purified mRNAs were frozen at −80° C. until electroporation. For transient transgene expression, primary T cells or Jurkat T cells were electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocols were used for the respective cell types: primary T cells (square wave protocol, 500 V, 5 ms and 4 mm cuvettes), Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

Antibodies and Flow Cytometry:

[0344] Primary human T cells or tumour cell lines were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide) and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) or via the FLAG tag using an anti-FLAG tag antibody (clone L5, BioLegend) as a primary antibody and a PE- or APC-conjugated secondary antibody in case of the Strep II tag antibody. Expression of the engineered target antigen tHER2 was detected with a PE-conjugated anti-HER2 antibody (clone 24D2, BioLegend). Analysis was done by the FlowJo software.

Construction of Transgene Constructs:

[0345] The nucleotide sequences encoding the GM-CSF-Ra signal peptide, the anti-human CD19 scFv FMC63, the human CD8α hinge, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain and the CD3ζ ITAM signalling domain were synthesized by GenScript. The nucleotide sequences encoding the signal peptide IgGk, the anti-human HER2 scFv 4D5-5 and the dimerization domain FKBP F36V were synthesized by GeneArt (Thermo Scientific). Sequences encoding the extracellular and transmembrane domain of HER2 were obtained from Addgene (plasmid #16257). Insertion of a Strep II tag (NWSHPQFEK) or FLAG tag (DYKDDDDK) and flexible linkers was performed by PCR. Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

Luciferase-Based Cytotoxicity Assay:

[0346] Luciferase-expressing tumour cells were co-cultured with CAR T cells at an E:T cell ratio of 2:1 with 10,000 target cells/well in white round-bottom 96 well plates for 4 hours at 37° C. in cytotoxicity assay medium, consisting of phenol-free RPMI, 10% FCS, 1% L-Glutamine and 1% penicillin-streptomycin. Finally, remaining living cells were quantified by determination of the residual luciferase activity of the co-culture. After equilibration to room temperature for 10 minutes, luciferin was added to the cell suspension (150 μg/mL final concentration) and luciferase activity was measured 20 minutes later using the ENSPIRE Multimode plate reader. The percentage of specific lysis was determined with the following formula:

% specific lysis=100−((RLU from well with effector and target cell co-culture)/(RLU from well with target cells only)×100)).

FACS-Based Cytotoxicity Assay:

[0347] For the FACS-based cytotoxicity assay two populations of target cells were generated: (i) Jurkat cells electroporated with mRNA encoding eGFP and RNA encoding the respective target antigen, and (ii) Jurkat cells only electroporated with mRNA encoding mCherry. These two populations were mixed at a 1:1 ratio and co-cultured with CAR T cells at an E:T cell ratio of 4:1:1 with 20,000 target cells/well in round-bottom 96 well plates for 4 hours at 37° C. Target cells without the addition of CAR T cells served as a control condition (“targets only”). After the incubation period, the co-cultures were centrifuged (5 minutes, 1600 rpm, 4° C.), supernatants were collected for subsequent cytokine measurements and the remaining cells were resuspended in 100 μL of FACS buffer, consisting of PBS, 0.2% human albumin and 0.02% sodium azide. The viability of target antigen.sup.pos and target antigen.sup.neg cell populations was determined using a BD LSRFortessa flow cytometer and specific lysis was calculated with the following formula:

% specific lysis=(1−(((% eGFP.sup.pos cells of the sample)/(% mCherry.sup.pos cells of the sample)/(% eGFP.sup.pos cells of the “targets only” control)/(% mCherry.sup.pos cells of the “targets only” control))))*100.

Cytokine Release by CAR T Cells:

[0348] Cytokine secretion of primary CAR T cells was assessed by co-cultivation with target cells at E:T ratios of 1:1 or 2:1 in flat-bottom 96 well plates for 4 hours or 24 hours at 37° C. In some experiments, released cytokines were quantified in the supernatants from the co-culture experiments for determining cytotoxicity. The supernatants were centrifuged (1600 rpm, 7 minutes, 4° C.) to remove remaining cells and debris and were subsequently frozen at −80° C. For analysis of secreted IFN-γ, ELISA was performed using the Human IFN gamma ELISA Ready-SET-Go!® kit (eBioscience) according to the manufacturer's instructions. Measurements were conducted using the ENSPIRE Multimode plate reader.

In Vitro Dimerization of Transgenes:

[0349] Dimerization of transgenes was induced prior to co-cultivation experiments. Primary T cells were diluted to the final cell concentration in the respective cell culture medium. The homodimerization agent AP20187 (MedChemExpress) was diluted in cell culture medium and was added at 10 nM final concentration. Addition of the respective vehicle control DMSO at the same concentration served as control. Cells were incubated at 37° C. for 30 minutes to ensure efficient dimerization of transgenes and subsequently used for in vitro experiments.

Example 4: Generation of Switchable CARs by Regulating Avidity Via Homodimerization of the Domain FKBP F36V

[0350] The fourth example illustrates the regulation of CAR function by regulating the avidity of a group of CARs by conditional homodimerization. For this purpose, we integrated into a monomeric CAR backbone a homodimerization domain based on the FKBP F36V variant (shown in FIG. 5A) and fused this backbone to a binding moiety with high or with low affinity towards EGFR, i.e., S(WT)-8ser-BB-FKBP(36V)-3z “E11.4.1-WT” (SEQ ID NO: 49) and S(G32A)-8ser-BB-FKBP(36V)-3z “E11.4.1-G32A” (SEQ ID NO: 46), respectively. Primary human T cells were transduced with lentiviral vectors encoding the resulting CAR molecules. Expression of the CARs was detected via the Strep II tag and is shown in FIG. 5B. For functional analysis, CAR T cells were co-cultured for 4 hours at 37° C. at an E:T ratio of 10:1 with the target cell line A431-fLuc, which expresses high levels of EGFR (FIG. 5C). Cytotoxicity of target cells was determined in a luciferase-based cytotoxicity assay. T cells expressing a CD19-specific reference CAR CD19-8cys-BB-3z “CD19-BBz” (SEQ ID NO: 58) and T cells without any CAR (“no CAR”) were used as control conditions. CAR dimerization was induced by addition of 10 nM AP20187 to the T cells prior to co-cultivation with the target cells. Addition of the vehicle control DMSO served as control. FIG. 5D shows that the function of a CAR with the low affinity-binding moiety E11.4.1-G32A (S(G32A)-8ser-BB-FKBP(36V)-3z (SEQ ID NO: 46)) but not with the high affinity-binding moiety E11.4.1-WT (S(WT)-8ser-BB-FKBP(36V)-3z (SEQ ID NO: 49)) strongly depended on the presence of the dimerization agent AP20187. The dimerization agent itself had no effects on the lysis of target cells, as can be seen in the control conditions with T cells without CAR expression.

Maintenance of Human Cell Lines

[0351] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail. Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. A431 epidermoid carcinoma cells were maintained in DMEM (Thermo Scientific) supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

Transduction of T Cells and Cell Lines

[0352] Virus production of pantropic VSV-G pseudotyped lentivirus was performed from Lenti-X 293T cells (Clontech Laboratories) with a third-generation Puromycin-selectable pCDH transgene vector (System Biosciences) and second-generation viral packaging plasmids pMD2.G and psPAX2 (both obtained from Addgene, plasmids #12259 and #12260 respectively). Co-transfection was performed using Purefection Transfection Reagent (System Biosciences) according to the manufacturer's instructions. Supernatants were collected one and two days after transfection and were concentrated using the Lenti-X Concentrator (Clontech Laboratories) according to the manufacturer's instructions.

[0353] Twenty-four hours prior to the lentiviral transduction, primary T cells were activated using anti-CD3/28 beads, according to the manufacturer's instructions. Cell culture plates were coated with RetroNectin (Clontech Laboratories), according to the manufacturer's instructions, to promote co-localization of lentivirus and primary T cells. Cells were exposed to concentrated lentiviral supernatants for one day, followed by removal of the virus particles. After three days, T cells were treated with 1 μg/mL Puromycin (Sigma Aldrich) to ensure high and uniform expression of the transgene. T cells were expanded in T cell transduction medium, consisting of AIM-V (Life Technologies) supplemented with 2% Octaplas (Octapharma), 1% L-Glutamine, 2.5% HEPES (Thermo Scientific) and 200 IU/mL recombinant human IL-2.

[0354] Cell lines were split 24 hours before lentiviral transduction to ensure exponential cell growth at the timepoint of transduction. Cells were exposed to varying concentrations of lentiviral supernatants for one day. Puromycin selection was performed three days after transduction with concentrations varying from 1 to 8 μg/mL to exclude non-transduced cells.

Antibodies and Flow Cytometry

[0355] Primary human T cells or tumour cell lines were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) or with Protein L in case of the CD19-BBz CAR as primary antibodies and a PE- or APC-conjugated secondary antibody. Expression EGFR was detected with a PE-conjugated anti-EGFR antibody (clone AY13, BioLegend). Analysis was done by the FlowJo software.

Construction of Transgene Constructs

[0356] The nucleotide sequences encoding the CD33 signal peptide, the low affinity rcSso7d variant E11.4.1-G32A, the Strep II tag (NWSHPQFEK), a flexible G.sub.4S linker, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain, the dimerization domain FKBP F36V and the CD3ζ ITAM signalling domain were synthesized by GeneArt (Thermo Scientific). Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

Luciferase-Based Cytotoxicity Assay

[0357] Luciferase-expressing tumour cells were co-cultured with CAR T cells at an E:T cell ratio of 2:1 with 10,000 target cells/well in white round-bottom 96 well plates for 4 hours at 37° C. in cytotoxicity assay medium, consisting of phenol-free RPMI, 10% FCS, 1% L-Glutamine and 1% penicillin-streptomycin. Finally, remaining living cells were quantified by determination of the residual luciferase activity of the co-culture. After equilibration to room temperature for 10 minutes, luciferin was added to the cell suspension (150 μg/mL final concentration) and luciferase activity was measured 20 minutes later using the ENSPIRE Multimode plate reader. The percentage of specific lysis was determined with the following formula:

% specific lysis=100−((RLU from well with effector and target cell co-culture)/(RLU from well with target cells only)×100)).

In Vitro Dimerization of Transgenes

[0358] Dimerization of transgenes was induced prior to co-cultivation experiments. Primary T cells were diluted to the final cell concentration in the respective cell culture medium. The homodimerization agent AP20187 was diluted in cell culture medium and was added at 10 nM final concentration. Addition of the respective vehicle control DMSO at the same concentration served as control. Cells were incubated at 37° C. for 30 minutes to ensure efficient dimerization of transgenes and subsequently used for in vitro experiments.

Example 5: Treatment of Tumour Bearing Mice with Stably Transduced T Cells Expressing a Group of CARs Whose Avidity can be Controlled by Drug Administration

[0359] In example 5 we show in a leukaemia model with immunodeficient NOD.Cg-Prkdc.sup.scid Il2rg.sup.tm1WJI/SzJ (NSG) mice that tumour growth can efficiently be inhibited by lentivirally transduced T cells expressing the low-affinity CAR “S(G32A)-8ser-BB-FKBP(36V)-3z” (SEQ ID NO: 46) in the presence but not the absence of a regulating molecule.

[0360] For the in vivo model we used the B-ALL cell line Nalm6 which was transduced with a vector for high expression of tEGFR (approx. 1×10.sup.6 tEGFR molecules per cell) and firefly luciferase for in vivo quantification of tumour growth by using bioluminescence imaging. Intravenous (i.v.) injection of 0.5×10.sup.6 Nalm6-tEGFR-fLuc cells into NSG mice resulted in exponential tumour growth in untreated mice. FIG. 6A shows that the growth of this cell line in NSG mice, however, was efficiently inhibited when 10×10.sup.6 T cells expressing either an anti-CD19 CAR CD19-8cys-BB-3z (SEQ ID NO: 58) or the high-affinity anti-EGFR CAR “S(WT)-8ser-BB-FKBP(36V)-3z” (SEQ ID NO: 49) were intravenously injected three days after injection of the tumour cells. Importantly, also T cells with the low-affinity EGFR-CAR “S(G32A)-8ser-BB-FKBP(36V)-3z” (SEQ ID NO: 46) inhibited leukaemia outgrowth, but only if the homodimerizer AP20187, i.e., the regulating molecule, was regularly administered (FIG. 6A), demonstrating that the CAR S(G32A)-8ser-BB-FKBP(36V)-3z is only fully active in the dimeric state (i.e., complexed group of CARs=ON-state). In the absence of the dimerizer (i.e., non-complexed CARs=OFF-state) we observed only a moderate growth inhibition (FIG. 6A). When the dimerizer, i.e., the regulating molecule, was administered to this latter group of mice 11 days after T cell injection, then the triggered complexation of the CAR molecules resulted in a strong reduction of tumour burden in 2 mice and moderate inhibition in the 3 other mice (FIG. 6B). Together, these in vivo experiments confirm our in vitro data, demonstrating that cells, which express CAR molecules with low affinity antigen binding moieties, are only efficiently triggered, if the CAR molecules are complexed (i.e. assembled) into a group of CARs, thereby resulting in avidity.

[0361] Expression of the CARs was detected with an anti-Strep II tag antibody in case of rcSso7d-based CARs and with Protein L in case of the CD19-specific CAR (FIGS. 7A and 7B, respectively). For the functional characterization in vitro, the CAR T cells were co-cultured with Nalm6 cells that were either expressing no EGFR (“Nalm6-fLuc”) or high levels of EGFR (“Nalm6-tEGFR-fLuc”) (FIGS. 7E and 7F) for 4 hours at 37° C. at an E:T ratio of 10:1. CAR homodimerization was induced by addition of 10 nM AP20187 to the T cells prior to co-cultivation with the target cells. Addition of the vehicle control DMSO served as control. FIGS. 7C and 7D show the cytolytic capacity of CAR T cells co-cultured with either Nalm6-fLuc or Nalm6-EGFR-fLuc cells, respectively.

Maintenance of Human Cell Lines:

[0362] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail. Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

[0363] Transduction of T cells and cell lines: Virus production of pantropic VSV-G pseudotyped lentivirus was performed from Lenti-X 293T cells with a third-generation Puromycin-selectable pCDH transgene vector and second-generation viral packaging plasmids pMD2.G and psPAX2 (both obtained from Addgene, plasmids #12259 and #12260 respectively). Co-transfection was performed using Purefection Transfection Reagent according to the manufacturer's instructions. Supernatants were collected one and two days after transfection and were concentrated using the Lenti-X Concentrator according to the manufacturer's instructions.

[0364] Twenty-four hours prior to the lentiviral transduction, primary T cells were activated using anti-CD3/28 beads, according to the manufacturer's instructions. Cell culture plates were coated with RetroNectin, according to the manufacturer's instructions, to promote co-localization of lentivirus and primary T cells. Cells were exposed to concentrated lentiviral supernatants for one day, followed by removal of the virus particles. After three days, T cells were treated with 1 μg/mL Puromycin to ensure high and uniform expression of the transgene. T cells were expanded in T cell transduction medium, consisting of AIM-V supplemented with 2% Octaplas, 1% L-Glutamine, 2.5% HEPES and 200 IU/mL recombinant human IL-2.

[0365] Cell lines were split 24 hours before lentiviral transduction to ensure exponential cell growth at the time point of transduction. Cells were exposed to varying concentrations of lentiviral supernatants for one day. Puromycin selection was performed three days after transduction with concentrations varying from 1 to 8 μg/mL in order to exclude non-transduced cells.

Construction of Transgene Constructs:

[0366] The nucleotide sequences encoding the CD33 signal peptide, the low affinity rcSso7d variant E11.4.1-G32A, the Strep II tag (NWSHPQFEK), a flexible G.sub.4S linker, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain, the dimerization domain FKBP F36V and the CD3ζ ITAM signalling domain were synthesized by GeneArt (Thermo Scientific). The nucleotide sequences encoding the GM-CSF-Ra signal peptide and the anti-human CD19 scFv FMC63 were synthesized by GenScript. The nucleotide sequence encoding the extracellular and transmembrane domain of EGFR was obtained from Addgene (plasmid #11011). Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

Antibodies and Flow Cytometry:

[0367] Primary human T cells or tumour cell lines were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) or Protein L in case of the CD19-BBz CAR as primary antibodies and a PE- or APC-conjugated secondary antibody. Expression EGFR was detected with a PE-conjugated anti-EGFR antibody (clone AY13, BioLegend). Analysis was done by the FlowJo software.

Luciferase-Based Cytotoxicity Assay:

[0368] Luciferase-expressing tumour cells were co-cultured with CAR T cells at an E:T cell ratio of 2:1 with 10,000 target cells/well in white round-bottom 96 well plates for 4 hours at 37° C. in cytotoxicity assay medium, consisting of phenol-free RPMI, 10% FCS, 1% L-Glutamine and 1% penicillin-streptomycin. Finally, remaining living cells were quantified by determination of the residual luciferase activity of the co-culture. After equilibration to room temperature for 10 minutes, luciferin was added to the cell suspension (150 μg/mL final concentration) and luciferase activity was measured 20 minutes later using the ENSPIRE Multimode plate reader. The percentage of specific lysis was calculated with the following formula:

% specific lysis=100−((RLU from well with effector and target cell co-culture)/(RLU from well with target cells only)×100)).

In Vitro Dimerization of Transgenes:

[0369] Dimerization of transgenes was induced prior to co-cultivation experiments. Primary T cells were diluted to the final cell concentration in the respective cell culture medium. The homodimerization agent AP20187 was diluted in cell culture medium and was added at 10 nM final concentration. Addition of the respective vehicle control DMSO at the same concentration served as control. Cells were incubated at 37° C. for 30 minutes to ensure efficient dimerization of transgenes and subsequently used for in vitro experiments.

In Vivo Target Cell Killing:

[0370] NOD.Cg-Prkdc.sup.scid Il2rg.sup.tm1WJI/SzJ (NSG) mice were housed in the Anna Spiegel facility for animal breeding. For subsequent experiments, mice were transferred to the preclinical research laboratories (PIL) of the Medical University of Vienna. All procedures were performed as approved (GZ: 813267/2015/24) by the Magistratsabteilung 58, Vienna.

[0371] Primary T cells were engineered to express the CD19-specific control CAR (CD19-BBz), the EGFR-specific high and low affinity CARs (“E11.4.1-WT” and “E11.4.1-G32A”, respectively) using the protocol depicted in “Transduction of T cells and cell lines”. After transduction, the CAR T cells were expanded for over 14 days prior to in vivo experiments to generate sufficient cell numbers.

[0372] The homodimerization agent AP20187 (Clontech Laboratories) was dissolved in vehicle solution according to manufacturer's instructions. Briefly, AP20187 was initially dissolved to a concentration of 12.5 mg/mL in ethanol with rigorous vortexing. The compound was then diluted to the final working concentration of 0.5 mg/mL using an appropriate mixture of PEG-400 (Sigma Aldrich) and Tween-80 (Sigma Aldrich) in water. The resulting vehicle solution consisted of 4% ethanol, 10% PEG-400 and 1.7% Tween-80 in water for injection. The working stock of AP20187 was prepared immediately prior to injection, sterile-filtered and was used within 30 minutes.

[0373] Nalm6 cells engineered to express high levels of tEGFR-FKBP and fLuc (“Nalm6-tEGFR-fLuc”) were resuspended in PBS, filtered through a 35 μm cell strainer (Corning Falcon) and set to a final concentration of 5×10.sup.6/mL. 0.5×10.sup.6 cells were injected intravenously (i.v.) into the tail-vein of each NSG mouse (male and female mice; Medical University of Vienna, Department of Biomedical Research). Three days later, mice were treated with respective CAR T cells (10×10.sup.6 CAR T cells i.v. into the tail-vein), followed by injection of the homodimerization agent AP20187 (2 mg/kg dosage) or vehicle control using intraperitoneal (i.p.) injections. The dimerization agent AP20187 (2 mg/kg) or the vehicle control was administered on day 0 (right after T cell injection), day 1, day 2, day 4, day 7, day 9 and day 11 where indicated. All control conditions were treated with the respective vehicle control (4% ethanol, 10% PEG-400 and 1.7% Tween-80 in water for injection). Tumour growth and control was monitored by bioluminescence imaging (BLI). Mice were sacrificed by cervical dislocation at the end of the experiment.

Bioluminescence Imaging (BLI):

[0374] BLI imaging of tumour growth was performed at the preclinical research laboratory (PIL) of the Medical University of Vienna using an IVIS Spectrum In Vivo Imaging System (Perkin Elmer). D-Luciferin substrate (Perkin Elmer) was dissolved in PBS to a final concentration of 15 mg/mL and sterile filtered. Mice were anesthetized with isoflurane and received i.p. injections of the luciferin working stock (final dosage 150 mg/kg body weight). After 15-20 minutes, 1-3 mice were transferred to the IVIS Imaging System, bioluminescence was measured in medium binning mode at an acquisition time of 1 seconds to 2 minutes to obtain unsaturated images. Luciferase activity was analysed with the Living Image Software (Caliper) and photon flux was determined within the region of interest that encompassed the entire body of the mouse.

Example 6: Regulating CAR Avidity by Heterodimerization of CAR Molecules and Sensitivity of the CAR T Cells Depending on Whether the Target Antigen is Monomeric or Dimeric

[0375] In Example 6 the avidity of CAR molecules was regulated by heterodimerization. Further, the example shows the sensitivity of T cells expressing a group of CARs (S(G32A)-8ser-BB-FKBP-3z plus S(G32A)-8ser-BB-FRB-3z) comprising the low affinity binding moiety “E11.4.1-G32A” in comparison to T cells expressing a group of CARs (S(WT)-8ser-BB-FKBP-3z plus S(WT)-8ser-FRB-3z) comprising the high affinity binding moiety “E11.4.1-WT” plus/minus regulating molecule and the impact on whether the target antigens are monomeric or associated.

[0376] The target cells (Jurkat T cells) were electroporated with different amounts of mRNA (0.02 μg-10 μg) coding for tEGFR which yielded expression levels of tEGFR ranging from few hundred molecules/cell up to ˜300,000 molecules/cell. The expression of tEGFR (MFI and molecules/cell) in Jurkat T cells is shown in FIGS. 8A-C. To minimize cross-talk between dimerization domains, orthogonal dimerization systems were used for target and effector cells. For this purpose, we integrated the low-affinity antigen binding moiety E11.4.1-G32A and the high-affinity antigen binding moiety E11.4.1-WT into the monomeric CAR backbones 8ser-BB-FKBP-3z and 8ser-BB-FRB-3z (S(G32A)-8ser-BB-FKBP-3z+S(G32A)-8ser-BB-FRB-3z for E11.4.1-G32A and S(WT)-8ser-BB-FKBP-3z+S(WT)-8ser-BB-FRB-3z for E11.4.1-WT) comprising the heterodimerization domains FKBP or FRB, respectively, for conditional heterodimerization of the resulting CAR molecules by the rapalogue AP21967 (shown in FIG. 8F). Primary human T cells were electroporated with 5 μg mRNA for each construct and expression was detected 20 hours after electroporation with Strep II tag and FLAG tag (FIGS. 8D and 8E). FIGS. 8G and 8H show the capacity of T cells expressing the CARs with low and high affinity binding domains to trigger cytotoxicity and secrete cytokines depending on the number of antigen molecules present on surface of Jurkat T cells in presence or absence of AP21967 (for heterodimerization of the CARs) and AP20187 (for homodimerization of tEGFR). The group of CARs comprising the high-affinity CAR molecules S(WT)-8ser-BB-FKBP-3z and S(WT)-8ser-BB-FRB-3z triggered efficient effector functions at low antigen doses independent of dimerization of the antigen or the group of CARs. The group of CARs comprising the low-affinity CAR molecules S(G32A)-8ser-BB-FKBP-3z and S(G32A)-8ser-BB-FRB-3z was highly dependent on the presence of the regulating molecule AP21967 and did not or only poorly trigger effector functions in the absence of AP21967 even at high antigen doses. Potent cytolytic functions and secretion of cytokines by the T cells was only observed, when both the CAR molecules and the antigen molecules were dimerized.

Maintenance of Human Cell Lines

[0377] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail. Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. Jurkat T cells were a gift from Dr. Sabine Strehl at the CCRI and were maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

Antibodies and Flow Cytometry

[0378] Primary human T cells or tumour cell lines were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected either via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) or via the FLAG tag using an anti-FLAG tag antibody (clone L5, BioLegend) as a primary antibody and a PE- or APC-conjugated secondary antibody, in the case of anti-Strep II tag antibodies. Expression of tEGFR was detected via a PE- or APC-conjugated anti-EGFR antibody (clone AY13, BioLegend). Analysis was done by the FlowJo software.

Determination of Transgene Surface Densities

[0379] The number of molecules of tEGFR and the different CARs on the surface of the different cells was quantified using the QuantiBRITE Phycoerythrin Fluorescence Quantitation Kit (Becton Dickinson) according to the manufacturer's instructions. Transgene-expressing cells and non-transfected control cells were stained with saturating concentrations of the respective PE-labelled antibodies using the protocol described in “Antibodies and flow cytometry”. The geometric mean of the fluorescence intensity was determined, corrected for unspecific binding using control cells and used to estimate the number of antibodies bound per cell (ABC).

Construction of Transgene Constructs

[0380] The nucleotide sequences encoding the CD33 signal peptide, low affinity rcSso7d variant E11.4.1-G32A, the Strep II tag (NWSHPQFEK), a flexible G.sub.4S linker, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain, the dimerization domains FKBP F36V and FRB and the CD3ζ ITAM signalling domain were synthesized by GeneArt (Thermo Scientific). The nucleotide sequences encoding the dimerization domain FKBP was synthesized by Genscript. The sequence encoding the extracellular and transmembrane domain of EGFR was obtained from Addgene (plasmid #11011). Insertion of flexible linkers was performed by PCR. Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

In Vitro Transcription and Electroporation of mRNA

[0381] In vitro transcription was performed using the mMessage mMachine T7 Ultra Kit according to the manufacturer's instructions. 50-200 ng of column purified PCR product was used as a reaction template. The resulting mRNA was purified with an adapted protocol using the RNeasy column purification kit. Briefly, the mRNA solution was diluted with a mixture of RLT buffer, ethanol (Merck) and 2-mercaptoethanol. The mixture was loaded onto an RNeasy column and purification was performed according to the manufacturer's instructions. Elution was performed with nuclease-free water and purified mRNAs were frozen at −80° C. until electroporation. For transient transgene expression, primary T cells or Jurkat T cells were electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocols were used for the respective cell types: primary T cells (square wave protocol, 500 V, 5 ms and 4 mm cuvettes), Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

FACS-Based Cytotoxicity Assay

[0382] For the FACS-based cytotoxicity assay two populations of target cells were generated: (i) Jurkat cells electroporated with mRNA encoding eGFP and RNA encoding the respective target antigen, and (ii) Jurkat cells only electroporated with mRNA encoding mCherry. These two populations were mixed at a 1:1 ratio and co-cultured with CAR T cells at an E:T cell ratio of 4:1:1 with 20,000 target cells/well in round-bottom 96 well plates for 4 hours at 37° C. Target cells without the addition of CAR T cells served as a control condition (“targets only”). After the incubation period, the co-cultures were centrifuged (5 minutes, 1600 rpm, 4° C.), supernatants were collected for subsequent cytokine measurements and the remaining cells were resuspended in 100 μL of FACS buffer, consisting of PBS, 0.2% human albumin and 0.02% sodium azide. The viability of target antigen.sup.pos and target antigen.sup.neg cell populations was determined using a BD LSRFortessa flow cytometer and specific lysis was calculated with the following formula:

% specific lysis=(1-(((% eGFP.sup.pos cells of the sample)/(% mCherry.sup.pos cells of the sample)/(% eGFP.sup.pos cells of the “targets only” control)/(% mCherry.sup.pos cells of the “targets only” control))))*100.

Cytokine Release by CAR T Cells

[0383] Cytokine secretion of primary CAR T cells was assessed by co-cultivation with target cells at E:T ratios of 1:1 or 2:1 in flat-bottom 96 well plates for 4 hours or 24 hours at 37° C. In some experiments, released cytokines were quantified in the supernatants from the co-culture experiments for determining cytotoxicity. The supernatants were centrifuged (1600 rpm, 7 minutes, 4° C.) to remove remaining cells and debris and were subsequently frozen at −80° C. For analysis of secreted IFN-γ, ELISA was performed using the Human IFN gamma ELISA Ready-SET-Go!® kit (eBioscience) according to the manufacturer's instructions. Measurements were conducted using the ENSPIRE Multimode plate reader.

In Vitro Dimerization of Transgenes

[0384] Dimerization of transgenes was induced prior to co-cultivation experiments. Primary T cells and Jurkat T cells were diluted to the final cell concentration in the respective cell culture medium. The homodimerization agent AP20187 and the heterodimerization agent AP21967 (Clontech Laboratories) were diluted in cell culture medium and were added at 10 nM and 500 nM final concentration, respectively. Addition of the vehicle control DMSO and ethanol at the same concentration, respectively, served as control. Cells were incubated at 37° C. for 30 minutes to ensure efficient dimerization of transgenes and subsequently used for in vitro experiments.

Example 7: Regulation of the Avidity of a Group of CARs by VEGF

[0385] The seventh example demonstrates a strategy to create a group of CARs that can be complexed by an extracellular soluble factor, which in this case serves as the regulating molecule according to the present invention. In the shown example VEGF was used as a potential dimerization agent (i.e. regulating molecule) for homodimerization of the CAR S(G32A)-J.CT6-8ser-BB-3z (SEQ ID NO: 64 and SEQ ID NO: 65). For this purpose, an engineered CH2-CH3-IgG1-Fc domain was integrated into the ectodomain of a CAR molecule (SEQ ID NO: 65) and co-expressed with a soluble construct comprising the CH2-CH3-Fc domain “Janus CT6” (SEQ ID NO: 64), that was engineered for high affinity binding to VEGF (Lobner et al., MAbs. 2017; 9(7):1088-1104) and that covalently heterodimerizes via formation of disulphide bridges with the CH2-CH3-Fc domain in the ectodomain of the CAR molecule. The CH2-CH3 domains of both constructs were engineered for minimizing homodimerization (Lobner et al., MAbs. 2017; 9(7):1088-1104). In the given example E11.4.1-G32A was used as antigen binding moiety due to its dependence on cooperative binding. FIG. 9A shows a schematic representation of a VEGF-dependent EGFR-specific CAR comprising the two constructs (SEQ ID NO: 64 and SEQ ID NO: 65) and FIG. 9B shows the effect of addition of VEGF. Jurkat T cells were electroporated with 5 μg mRNA for tEGFR and expression was detected 20 hours after electroporation (FIG. 9C). Primary human T cells were electroporated with 5 μg mRNA for each construct. The T cells then expressed the CAR molecule comprising a monomeric second-generation CAR signalling backbone (SEQ ID NO: 65) that associates with the construct comprising the low-affinity E11.4.1-G32A binding moiety which was fused to the Janus-CT6-Fc domain and did not contain a transmembrane domain (SEQ ID NO: 64). CAR T cells expressing the two constructs (FIG. 9D) were treated with varying amounts of VEGF and were co-cultivated with Jurkat T cells expressing high levels of tEGFR (FIG. 9C). FIG. 9E shows the capacity to trigger cytotoxicity as determined by using a FACS-based cytotoxicity assay. FIG. 9E shows that the group of CARs triggered T cell cytotoxicity against target cells in a VEGF (i.e. regulating molecule)-dependent manner. This demonstrates that extracellular soluble factors such as VEGF can serve as regulating molecules, thereby promoting complexation of the group of CARs according to the present invention.

Maintenance of Human Cell Lines:

[0386] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail. Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS (Sigma Aldrich) and 10% DMSO until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. Jurkat T cells were a gift from Dr. Sabine Strehl at the CCRI and were maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

Antibodies and Flow Cytometry:

[0387] Primary human T cells or tumour cell lines were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected either via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) or via the Fc domain by using a biotinylated anti-human-IgG1-antibody (clone JDC-10, Biozol) as a primary antibody and a PE-conjugated streptavidin as secondary staining reagent. Expression of the engineered target antigen tEGFR was detected either with a PE- or APC-conjugated anti-EGFR antibody (clone AY13, BioLegend) Analysis was done by the FlowJo software.

Construction of Transgene Constructs:

[0388] The nucleotide sequences encoding the CD33 signal peptide, low affinity rcSso7d variant E11.4.1-G32A, the Strep II tag (NWSHPQFEK), a flexible G.sub.4S linker, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain and the CD3ζ ITAM signalling domain were synthesized by GeneArt (Thermo Scientific). Plasmids comprising the CH2-CH3-Fc domain “Janus CT6” and the mutated “WT” CH2-CH3-Fc domain were a kind gift from Elisabeth Lobner at the University of Natural Resources and Life Sciences in Vienna. Sequences encoding the extracellular and transmembrane domains of EGFR were obtained from Addgene (plasmid #11011). Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

In Vitro Transcription and Electroporation of mRNA:

[0389] In vitro transcription was performed using the mMessage mMachine T7 Ultra Kit according to the manufacturer's instructions. 50-200 ng of column purified PCR product was used as a reaction template. The resulting mRNA was purified with an adapted protocol using the RNeasy column purification kit. Briefly, the mRNA solution was diluted with a mixture of RLT buffer, ethanol and 2-mercaptoethanol. The mixture was loaded onto an RNeasy column and purification was performed according to the manufacturer's instructions. Elution was performed with nuclease-free water and purified mRNAs were frozen at −80° C. until electroporation. For transient transgene expression, primary T cells or Jurkat T cells were electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocols were used for the respective cell types: primary T cells (square wave protocol, 500 V, 5 ms and 4 mm cuvettes), Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

FACS-Based Cytotoxicity Assay:

[0390] For the FACS-based cytotoxicity assay two populations of target cells were generated: (i) Jurkat cells electroporated with mRNA encoding eGFP and RNA encoding the respective target antigen, and (ii) Jurkat cells only electroporated with mRNA encoding mCherry. These two populations were mixed at a 1:1 ratio and co-cultured with CAR T cells at an E:T cell ratio of 4:1:1 with 20,000 target cells/well in round-bottom 96 well plates for 4 hours at 37° C. Target cells without the addition of CAR T cells served as a control condition (“targets only”). After the incubation period, the co-cultures were centrifuged (5 minutes, 1600 rpm, 4° C.), supernatants were collected for subsequent cytokine measurements and the remaining cells were resuspended in 100 μL of FACS buffer, consisting of PBS, 0.2% human albumin and 0.02% sodium azide. The viability of target antigen.sup.pos and target antigen.sup.neg cell populations was determined using a BD LSRFortessa flow cytometer and specific lysis was calculated with the following formula:

% specific lysis=(1-(((% eGFP.sup.pos cells of the sample)/(% mCherry.sup.pos cells of the sample)/(% eGFP.sup.pos cells of the “targets only” control)/(% mCherry.sup.pos cells of the “targets only” control))))*100.

Recombinant Expression and Purification of VEGF:

[0391] Recombinant expression of a truncated form of human VEGF (residues 14-108) was previously described by (Lobner et al., MAbs. 2017; 9(7):1088-1104).

Example 8: Generation of an Affibody-Based Group of CARs Directed Against HER2

[0392] In the eighth example, we generated and identified an affibody based binding moiety directed against HER2 which is suitable for use in a group of CARs according to the present invention. Again, we started from a well characterized existent antigen binding moiety that was engineered for high affinity binding to human HER2 (Wikman et al., Protein Eng Des Sel. 2004; 17(5):455-462). To eliminate a potential N-glycosylation site and to reduce IgG binding, two point mutations (N23A and S33K) were introduced into the framework region of the binding scaffold (Feldwisch et al., J. Mol. Biol. 2010; 398(2):232-47), which resulted in the antigen binding moiety “zHER2-WT”. Low affinity mutants were generated by performing an alanine-scan of “zHER2-WT” by mutating all amino acids involved in antigen binding consecutively to alanine, resulting in various mutants containing one alanine-mutation each. Instead of expressing the 13 mutants in E. coli and determining their affinities for selection of appropriate antigen binding moieties, we performed a functional screening by directly integrating all mutants into a CAR backbone (exemplified with binder zHER2-WT (WT) in SEQ ID NO: 52) that could be conditionally homodimerized (FIG. 10A). By this way, T cells expressing the different CARs could be directly screened for activation in presence and absence of the homodimerizer in co-culture with Jurkat T cells that were electroporated with 5 μg mRNA encoding for tHER2. FIG. 10B shows the expression of tHER2 in the Jurkat cells. Primary human T cells were electroporated with 5 μg mRNA for each CAR construct and expression was detected 20 hours after electroporation via the hexahistidine tag (FIG. 10C). Expression of all 13 different affibody-based CARs was comparable (FIG. 10D). Primary T cells expressing no construct were used as a negative control (“no CAR”). For the functional screening, dimerization of CAR molecules was induced by treatment of the CAR T cells with 10 nM of AP20187 for 30 minutes at 37° C. prior to co-cultivation with Jurkat T cells. Addition of the vehicle control DMSO served as a control. After co-cultivation for 4 hours at 37° C. at an E:T ratio of 2:1, the capacity of the CARs to trigger cytotoxicity was determined by performing a luciferase-based cytotoxicity assay. The capacity of the different CARs to trigger cytotoxicity in T cells in presence or absence of 10 nM AP20187 is shown in FIG. 10E. The high-affinity affibody zHER2-WT triggered efficient target cell lysis independent of presence of AP20187. Similarly, CARs comprising the mutants Q11A, Q17A, W24A, T25A, S27A and R28A displayed no significant dependence on the presence of the dimerizer. No cytotoxicity was triggered by CARs comprising the affibody antigen binding moieties with substitutions Y13A and W14A, whereas Y35A triggered cytotoxicity at low levels. Dimerization-induced activation was observed with the mutants L9A-, R10A- and R32A, which therefore represent binding moieties suited for integration into CAR molecules according to the present invention.

Maintenance of Human Cell Lines:

[0393] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail (STEMCELL Technologies). Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. Jurkat T cells were a gift from Dr. Sabine Strehl at the CCRI and were maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

Antibodies and Flow Cytometry:

[0394] Primary human T cells or tumour cell lines were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected via the hexahistidine tag using an AF647-conjugated anti-pentahistidine tag antibody (Qiagen). Expression of the engineered target antigen tHER2 was detected with a PE-conjugated anti-HER2 antibody (clone 24D2, BioLegend). Analysis was done by the FlowJo software.

In Vitro Transcription and Electroporation of mRNA:

[0395] In vitro transcription was performed using the mMessage mMachine T7 Ultra Kit according to the manufacturer's instructions. 50-200 ng of column purified PCR product was used as a reaction template. The resulting mRNA was purified with an adapted protocol using the RNeasy column purification kit. Briefly, the mRNA solution was diluted with a mixture of RLT buffer, ethanol and 2-mercaptoethanol. The mixture was loaded onto an RNeasy column and purification was performed according to the manufacturer's instructions. Elution was performed with nuclease-free water and purified mRNAs were frozen at −80° C. until electroporation. For transient transgene expression, primary T cells or Jurkat T cells were electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocols were used for the respective cell types: primary T cells (square wave protocol, 500 V, 5 ms and 4 mm cuvettes), Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

Construction of Transgene Constructs:

[0396] The nucleotide sequences encoding the CD33 signal peptide, the affibody zHER2-WT, the hexahistidine tag, a flexible G.sub.4S linker, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain, the dimerization domain FKBP F36V and the CD3ζ ITAM signalling domain were synthesized by GeneArt. Sequences encoding the extracellular and transmembrane domain of HER2 were obtained from Addgene (plasmid #16257). Insertion of flexible linkers was performed by PCR. Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

Alanine-Scanning of Protein Antigen Binding Moieties:

[0397] Site-directed mutagenesis of all amino acids involved in epitope binding was performed using the QuikChange Lightning Site-Directed Mutagenesis Kit, according to the manufacturer's instructions. Primers were designed using the QuikChange Primer Design software (Agilent Genomics) and oligonucleotides were synthesized by Biomers.

Luciferase-Based Cytotoxicity Assay:

[0398] Luciferase-expressing tumour cells were co-cultured with CAR T cells at an E:T cell ratio of 2:1 with 10,000 target cells/well in white round-bottom 96 well plates for 4 hours at 37° C. in cytotoxicity assay medium, consisting of phenol-free RPMI, 10% FCS, 1% L-Glutamine and 1% penicillin-streptomycin. Finally, remaining living cells were quantified by determination of the residual luciferase activity of the co-culture. After equilibration to room temperature for 10 minutes, luciferin was added to the cell suspension (150 μg/mL final concentration) and luciferase activity was measured 20 minutes later using the ENSPIRE Multimode plate reader. The percentage of specific lysis was calculated with the following formula:

% specific lysis=100−((RLU from well with effector and target cell co-culture)/(RLU from well with target cells only)×100)).

In Vitro Dimerization of Transgenes:

[0399] Dimerization of transgenes was induced prior to co-cultivation experiments. Primary T cells were diluted to the final cell concentration in the respective cell culture medium. The homodimerization agent AP20187 was diluted in cell culture medium and was added at 10 nM final concentration. Addition of the vehicle control DMSO at the same concentration served as control. Cells were incubated at 37° C. for 30 minutes to ensure efficient dimerization of transgenes and subsequently used for in vitro experiments.

Expression and Purification of Affibody-Based Antigen Binding Moieties:

[0400] Binding scaffolds were expressed as sfGFP fusion proteins (consisting of an N-terminal hexahistidine tag followed by either rcSso7d or the Affibody and sfGFP) using the pE-SUMO vector. The schematics of the architecture of the fusion proteins are shown in FIG. 14G. Different mutants of the Affibody-based binder zHER2 were fused to sfGFP in the same way as indicated in FIG. 14G. The nucleotide sequence that encodes the sfGFP reporter protein was obtained from Addgene (plasmid #54737). Briefly, Escherichia coli cells (Tuner DE3) were transformed with sequence-verified plasmids using heat shock transformation. After overnight cultivation at 37° C., cultures were diluted 1:100 in TB medium (12 g/L tryptone, 24 g/L yeast extract, 4% glycerol, 2.31 g/L KH.sub.2PO.sub.4 and 16.43 g/L K.sub.2HPO.sub.4*3H.sub.2O) supplemented with kanamycin (50 μg/mL) and incubated at 37° C. while shaking. When cultures reached an A.sub.600 of roughly 2, expression of the transgene was induced by addition of 1 mM of IPTG and cells were further cultured overnight at 20° C. Cells were harvested by centrifugation (5000 g, 20 minutes, 4° C.), resuspended in sonication buffer (50 mM sodium phosphate, 300 mM NaCl, 3% glycerol, 1% Triton X-100, pH 8.0), sonicated (2×90 seconds, duty cycle 50%, amplitude set to 5) and centrifuged again to remove cell debris. Hexahistidine-tagged fusion proteins were purified from crude cell extracts using TALON metal affinity resin. After addition of 10 mM imidazole, the sonicated supernatants were applied onto the resin twice, followed by washing step with equilibration buffer (50 mM sodium phosphate, 300 mM NaCl, pH 8.0) with increasing amounts of imidazole (5-15 mM). Binding scaffolds were eluted by applying equilibration buffer supplemented with 250 mM imidazole. After buffer exchange to PBS using Amicon Ultra-15 10K centrifugal filters, concentrations were determined by measuring the absorbance at 280 nm using the respective molar absorption coefficient and finally proteins were directly frozen at −80° C.

Determination of Binding Affinities Using Surface Plasmon Resonance (SPR):

[0401] SPR experiments were performed with a Biacore T200 instrument. All experiments were conducted in degassed and filtered PBS, pH 7.4, containing 0.1% BSA and 0.05% Tween-20 (Merck Millipore), at 25° C. hHER2-Fc (R&D) was immobilized on a Protein A sensor chip at a flow rate of 10 μL/min for 60 seconds at a concentration 4 μg/mL. To determine the affinity of Affibody-based antigen binding moieties, five concentrations (depending on the expected K.sub.d of the antigen binding moiety) of the respective protein were injected at a flow rate of 30 μL/min for 15 seconds (zHER2-R10A and zHER2-R32A) or 60 seconds (zHER2-WT) in the single-cycle kinetic mode, followed by a dissociation step (60 seconds for zHER2-R10A and zHER2-R32A and 180 seconds for zHER2-WT). Regeneration was performed using 10 mM Glycine-HCl, pH 1.5 at a flow rate of 30 μL/min for 30 seconds. The K.sub.dwas obtained by curve fitting using the Biacore T200 Evaluation Software (GE Healthcare).

Example 10: Generation of Groups of CARs Comprising Three or Four CAR Molecules

[0402] By using two orthogonal dimerization platforms (FKBP/FRB using AP21967; FKBP F36V/FKBP F36V using AP20187) and the low affinity binding moiety E11.4.1-G32A, we exemplify a strategy to create conditionally active groups of CARs comprising three (comprising the two constructs (SEQ ID NO: 48) and (SEQ ID NO: 67)) or four CAR molecules (comprising the two constructs (SEQ ID NO: 48) and (SEQ ID NO: 68)) in the complexed state. Complexing three or four CAR molecules into a group of CARs increases the avidity effect which can increase the sensitivity towards the respective antigen. FIG. 11A and FIG. 11B show a schematic representation of a trimeric and a tetrameric CAR, respectively. Jurkat T cells were electroporated with 5 μg of mRNA coding for the two separate chains of either the trimeric or the tetrameric group of CARs and CAR expression was detected 20 hours after electroporation via Strep II tag or FLAG tag, depending on the respective signalling chain. FIG. 11C shows the expression of the trimeric and tetrameric group of CARs.

Maintenance of Human Cell Lines

[0403] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3.sup.pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail (STEMCELL Technologies). Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO until use. CD3.sup.pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. A Jurkat T reporter cell line engineered with an NFκB-dependent eGFP gene and a NFAT-dependent CFP gene is a kind gift from Dr. Peter Steinberger at the Medical University of Vienna and is maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines are regularly tested for mycoplasma contamination and authentication is performed at Multiplexion, Germany. Cell densities are monitored with AccuCheck counting beads.

In Vitro Transcription and Electroporation of mRNA

[0404] In vitro transcription is performed using the mMessage mMachine T7 Ultra Kit according to the manufacturer's instructions. 50-200 ng of column purified PCR product are used as a reaction template. The resulting mRNA is purified with an adapted protocol using the RNeasy column purification kit. Briefly, the mRNA solution is diluted with a mixture of RLT buffer, ethanol and 2-mercaptoethanol. The mixture is loaded onto an RNeasy column and purification is performed according to the manufacturer's instructions. Elution is performed with nuclease-free water and purified mRNAs are frozen at −80° C. until electroporation. For transient transgene expression, Jurkat T cells are electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocol is used: primary T cells (square wave protocol, 500 V, 5 ms and 4 mm cuvettes), Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

Antibodies and Flow Cytometry

[0405] Primary T cells and Jurkat T cells are resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide and treated for 10 minutes at 4° C. with 10% human serum. Cells are stained with the respective primary antibody for 25 minutes at 4° C. Stained cells are washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs is detected either via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) or via the FLAG tag using an anti-FLAG tag antibody (clone L5, BioLegend) as a primary antibody and a PE- or APC-conjugated secondary antibody, in the case of anti-Strep II tag antibodies. Expression of the engineered target antigen tEGFR was detected either with a PE- or APC-conjugated anti-EGFR antibody (clone AY13, BioLegend). Analysis was done by the FlowJo software.

Construction of Transgene Constructs

[0406] The nucleotide sequences encoding the CD33 signal peptide, the low affinity rcSso7d variant E11.4.1-G32A, the Strep II tag, a flexible G.sub.4S linker, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain, the dimerization domains FKBP F36V and FRB and the CD3ζ ITAM signalling domain are synthesized by GeneArt (Thermo Scientific). The nucleotide sequence encoding the dimerization domain FKBP is synthesized by Genscript. Sequences encoding the extracellular and transmembrane domain of EGFR are obtained from Addgene (plasmid #11011). Flexible linkers and the FLAG tag are inserted by using respective PCR primers. Assembly of nucleotide sequences into functional transgenes is performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs are amplified by PCR and subsequently used for in vitro transcription.

Measuring Transcription Factor Activity by Jurkat T Reporter Cell Line

[0407] The Jurkat T reporter cell line is differentially labelled with distinct fluorescent proteins to enable efficient differentiation of Jurkat T reporter cells and Jurkat T target cells expressing the respective tumour antigen within the same well. Suitable fluorescent protein can be dKeima (Addgene #54618), mAmetrine (Addgene #54505) or similar proteins that have minimal cross-talk to the reporter proteins. Activity of the transcription factors NF-AT and NF-κB in Jurkat T reporter cells expressing the respective CAR is assessed by co-cultivation with target cells at an E:T ratios of 0.25:1, 0.5:1, 1:1 or 2:1 in round-bottom 96 well plates for 4 hours, 8 hours, 16 hours or 24 hours at 37° C. Cells are acquired using a BD LSRFortessa and activation of Jurkat T reporter cells is determined by measuring the geometric mean of the fluorescence intensity of the respective reporter protein or the percentage of reporter protein positive cells.

In Vitro Dimerization of Transgenes

[0408] Dimerization of transgenes is induced prior to co-cultivation experiments. Primary T cells and Jurkat T cells are diluted to the final cell concentration in the respective cell culture medium. The homodimerization agent AP20187 and the heterodimerization agent AP21967 are diluted in cell culture medium and are added at 10 nM and 500 nM final concentration, respectively. Addition of the respective vehicle control DMSO or ethanol, respectively, at the same concentration served as control. Cells are incubated at 37° C. for 30 minutes to ensure efficient dimerization of transgenes and subsequently used for in vitro experiments.

FACS-Based Cytotoxicity Assay:

[0409] For the FACS-based cytotoxicity assay two populations of target cells are generated: (i) Jurkat cells electroporated with mRNA encoding eGFP and RNA encoding the respective target antigen, and (ii) Jurkat cells only electroporated with mRNA encoding mCherry. These two populations are mixed at a 1:1 ratio and co-cultured with CAR T cells at an E:T cell ratio of 4:1:1 with 20,000 target cells/well in round-bottom 96 well plates for 4 hours or 24 hours at 37° C. Target cells without the addition of CAR T cells serve as a control condition (“targets only”). After the incubation period, the co-cultures are centrifuged (5 minutes, 1600 rpm, 4° C.), supernatants are collected for subsequent cytokine measurements and the remaining cells are resuspended in 100 μL of FACS buffer, consisting of PBS, 0.2% human albumin and 0.02% sodium azide. The viability of target antigen.sup.pos and target antigen.sup.neg cell populations is determined using a BD LSRFortessa flow cytometer and specific lysis is calculated with the following formula:

% specific lysis=(1−(((% eGFP.sup.pos cells of the sample)/(% mCherry.sup.pos cells of the sample)/(% eGFP.sup.pos cells of the “targets only” control)/(% mCherry.sup.pos cells of the “targets only” control))))*100.

Example 11: Generation of Groups of CARs Comprising Different Co-Stimulatory Domains in the Co-Stimulatory Signalling Region of the CAR Molecules

[0410] Over the last two decades it has been demonstrated that CAR molecules of the 2.sup.nd generation containing different co-stimulatory domains can efficiently activate T cells. Example 11 shows that CAR molecules comprising different co-stimulatory domains in their signalling region also work in the context of a group of CARs of the present invention for exploitation of the aviditiy effect for controlling T cell function by regulating molecules. FIG. 12A shows the architecture of EGFR-specific CAR molecules S(G32A)-8ser-28-FKBP(36V)-3z (SEQ-ID NO: 69), S(G32A)-8ser-ICOS-FKBP(36V)-3z (SEQ-ID NO: 70) and S(G32A)-8ser-OX40-FKBP(36V)-3z (SEQ-ID NO:71) containing either CD28 or ICOS or OX40 in the co-stimulatory signalling region. The expression of those CAR molecules in Jurkat cells and primary human T cells, respectively, is illustrated in FIGS. 12B and C. Expression was analyzed by flow cytometry via the integrated Strep II tag 20 hours after electroporation of 5 μg of the respective mRNA. The capacity of those CAR molecules in the non-complexed (i.e. monovalent) and complexed (i.e. bivalent) state to activate the promoters NF-κB and NF-AT in Jurkat cells and to trigger cytotoxic effector functions in primary human T cells is shown in FIGS. 12D and E, respectively. FIG. 12D illustrates that S(G32A)-8ser-OX40-FKBP(36V)-3z, when expressed in Jurkat cells stably transduced with an NF-κB and NF-AT reporter, could efficiently trigger NF-κB and NF-AT when complexed by the regulating molecule AP20187 into a bivalent group of CARs but not in the uncomplexed monovalent state. Similarly, specific lysis was efficiently triggered in primary human T cells by the CAR molecules S(G32A)-8ser-ICOS-FKBP(36V)-3z and S(G32A)-8ser-OX40-FKBP(36V)-3z only if the CAR molecules were complexed by AP20187 into a group of CARs, respectively (FIG. 12E). Together, this confirms that the type of co-stimulatory domain in the co-stimulatory signalling region of CAR molecules of a group of CARs of the present invention can be varied.

Maintenance of Human Cell Lines

[0411] Primary human T cells were obtained from de-identified healthy donor's blood after apheresis (Buffy coats from the Austrian Red Cross, Vienna, Austria). CD3pos T cells were enriched by negative selection using RosetteSep Human T cell Enrichment Cocktail (STEMCELL Technologies). Isolated and purified T cells were cryopreserved in RPMI-1640 medium supplemented with 20% FCS and 10% DMSO until use. CD3pos T Cells were activated with anti-CD3/CD28 beads according to the manufacturer's instructions and were expanded in human T cell medium, consisting of RPMI-1640 supplemented with 10% FCS, 1% penicillin-streptomycin and 200 IU/mL recombinant human IL-2. Primary T cells were cultivated for at least 14 days before experiments were conducted. Jurkat T cells were a gift from Dr. Sabine Strehl at the CCRI and were maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. A Jurkat T reporter cell line engineered with an NFκB-dependent eGFP gene and a NFAT-dependent CFP gene is a kind gift from Dr. Peter Steinberger at the Medical University of Vienna and is maintained in RPMI-1640 supplemented with 10% FCS and 1% penicillin-streptomycin. Cell lines were regularly tested for mycoplasma contamination and authentication was performed at Multiplexion, Germany. Cell densities were monitored with AccuCheck counting beads.

[0412] In vitro transcription and electroporation of mRNA In vitro transcription was performed using the mMessage mMachine T7 Ultra Kit according to the manufacturer's instructions. 50-200 ng of column purified PCR product were used as a reaction template. The resulting mRNA was purified with an adapted protocol using the RNeasy column purification kit. Briefly, the mRNA solution was diluted with a mixture of RLT buffer, ethanol and 2-mercaptoethanol. The mixture was loaded onto an RNeasy column and purification was performed according to the manufacturer's instructions. Elution was performed with nuclease-free water and purified mRNAs were frozen at −80° C. until electroporation. For transient transgene expression, primary T cells were electroporated with varying amounts of the respective mRNA using the Gene Pulser (Biorad). Following protocol is used: primary T cells (square wave protocol, 500 V, 5 ms and 4 mm cuvettes), Jurkat T cells (square wave protocol, 500 V, 3 ms and 4 mm cuvettes).

Antibodies and Flow Cytometry

[0413] Primary T cells were resuspended in FACS buffer (PBS, 0.2% human albumin and 0.02% sodium azide and treated for 10 minutes at 4° C. with 10% human serum. Cells were stained with the respective primary antibody for 25 minutes at 4° C. Stained cells were washed two times in FACS buffer and then either stained with the secondary antibody for 25 minutes at 4° C. or processed directly with a BD LSRFortessa. Expression of CAR constructs was detected via the Strep II tag using an anti-Strep II tag antibody (clone 5A9F9, Genscript) as a primary antibody and a PE- or APC-conjugated secondary antibody. Expression of the engineered target antigens tEGFR was detected with a PE- or APC-conjugated anti-EGFR antibody (clone AY13, BioLegend). Analysis was done by the FlowJo software.

Construction of Transgene Constructs

[0414] The nucleotide sequences encoding the CD33 signal peptide, the low affinity rcSso7d variant E11.4.1-G32A, the Strep II tag, a flexible G4S linker, the human monomeric CD8α hinge (UniProt ID P01732, C164S and C181S) and CD8α transmembrane domain, the 4-1BB co-stimulatory domain, the dimerization domain FKBP F36V and the CD3ζ ITAM signalling domain were synthesized by GeneArt (Thermo Scientific). The nucleotide sequences encoding the intracellular domains of CD28, ICOS and OX40 were derived from cDNA clones (Sino Biological). Sequences encoding the extracellular and transmembrane domain of EGFR were obtained from Addgene (plasmid #11011). Flexible linkers were inserted by using respective PCR primers. Assembly of nucleotide sequences into functional transgenes was performed by using the Gibson Assembly Master Mix, according to the manufacturer's instructions. The schematics and sequences are shown in FIGS. 14 and 15, respectively. The resulting constructs were amplified by PCR and subsequently used for in vitro transcription.

Measuring Transcription Factor Activity by Jurkat T Reporter Cell Line

[0415] The Jurkat T reporter cell line was differentially labelled with distinct fluorescent proteins to enable efficient differentiation of Jurkat T reporter cells and Jurkat T target cells expressing the respective tumour antigen within the same well. Suitable fluorescent proteins were dKeima (Addgene #54618), mAmetrine (Addgene #54505) or similar proteins that have minimal cross-talk to the reporter proteins. Activity of the transcription factors NF-AT and NF-κB in Jurkat T reporter cells expressing the respective CAR was assessed by co-cultivation with target cells at an E:T ratios of 0.25:1, 0.5:1, 1:1 or 2:1 in round-bottom 96 well plates for 24 hours at 37° C. Cells were acquired using a BD LSRFortessa and activation of Jurkat T reporter cells was determined by measuring the geometric mean of the fluorescence intensity of the respective reporter protein or the percentage of reporter protein positive cells.

FACS-Based Cytotoxicity Assay:

[0416] For the FACS-based cytotoxicity assay two populations of target cells were generated: (i) Jurkat cells electroporated with mRNA encoding eGFP and RNA encoding the respective target antigen, and (ii) Jurkat cells only electroporated with mRNA encoding mCherry. These two populations were mixed at a 1:1 ratio and co-cultured with CAR T cells at an E:T cell ratio of 4:1:1 with 20,000 target cells/well in round-bottom 96 well plates for 4 hours or 24 hours at 37° C. Target cells without the addition of CAR T cells served as a control condition (“targets only”). After the incubation period, the co-cultures were centrifuged (5 minutes, 1600 rpm, 4° C.), supernatants were collected for subsequent cytokine measurements and the remaining cells were resuspended in 100 μL of FACS buffer, consisting of PBS, 0.2% human albumin and 0.02% sodium azide. The viability of target antigen.sup.pos and target antigen.sup.neg cell populations was determined using a BD LSRFortessa flow cytometer and specific lysis was calculated with the following formula:

% specific lysis=(1−(((% eGFP.sup.pos cells of the sample)/(% mCherry.sup.pos cells of the sample)/(% eGFP.sup.pos cells of the “targets only” control)/(% mCherry.sup.pos cells of the “targets only” control))))*100.

[0417] The present invention therefore discloses the following preferred embodiments:

[0418] 1. A group of chimeric antigen receptors (CARs) consisting of two, three or four CAR molecules,

wherein the members of the group of CARs can be different or identical in their amino acid sequences to one another, and
wherein each of the CAR molecules of the group comprise at least a transmembrane domain and an ectodomain comprising either an antigen binding moiety or a binding site to which another polypeptide is able to bind, wherein the another polypeptide comprises an antigen binding moiety, and
wherein at least one CAR molecule of the group additionally comprises an endodomain, which comprises at least a signalling region which can transduce a signal via at least one immunoreceptor tyrosine-based activation motif (ITAM) or at least one immunoreceptor tyrosine-based inhibitory motif (ITIM), and
wherein the endodomain of each CAR molecule of the group, in case the respective CAR molecule comprises an endodomain, is located on the intracellular side of a cell membrane, if expressed in a cell, wherein the ectodomain of each CAR molecule of the group translocates to the extracellular side of a cell membrane, if expressed in a cell, and wherein the transmembrane domain of each CAR molecule of the group is located in a cell membrane, if expressed in a cell;
wherein each CAR molecule of the group comprises at least one dimerization domain, which can mediate homo- or heterodimerization with other CAR molecules of the group,
wherein this dimerization of a pair of dimerization domains is either induced by a regulating molecule and optionally reduced by another regulating molecule, or occurs in the absence of a regulating molecule and is reduced by a regulating molecule,
wherein a regulating molecule is able to bind under physiological conditions to at least one member of a pair of dimerization domains and by inducing or reducing dimerization either induces or reduces the formation of a non-covalently complexed group of CARs consisting of two, three or four CAR molecules, and
wherein the ectodomain of each CAR molecule of the group in its prevalent conformation is free of cysteine amino acid moieties which are able to form intermolecular disulphide bonds with other CAR molecules of the group, respectively, and
wherein the antigen binding moieties of the CAR molecules of the group and of the other polypeptides being able to bind to the CAR molecules of the group are either specific for one target antigen or for a non-covalent or a covalent complex of different target antigens, and
wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 1 mM and 100 nM, and
wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 1 mM and 100 nM.
2. A group of CARs according to embodiment 1, wherein the antigen binding moiety comprises only one protein domain.
3. A group of CARs according to embodiment 1 or 2, wherein the antigen binding moiety comprises only one protein domain and does not cause dimerization or oligomerization of CAR molecules of the group when expressed on the surface of a human cell, and
wherein said protein domain preferably is selected from the group consisting of a human or non-human VH or VL single domain antibody (nanobody) or an engineered antigen binding moiety based on the Z-domain of staphylococcal Protein A, lipocalins, SH3 domains, fibronectin type III (FN3) domains, knottins, Sso7d, rcSso7d, Sac7d, Gp2, DARPins, ubiquitin, a receptor, a ligand of a receptor, or a co-receptor.
4. A group of CARs according to any one of embodiments 1 to 3, wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 1 mM and 150 nM, preferably between 1 mM and 200 nM, more preferably between 1 mM and 300 nM, especially between 1 mM and 400 nM, and
wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 1 mM and 150 nM, preferably between 1 mM and 200 nM, more preferably between 1 mM and 300 nM, especially between 1 mM and 400 nM.
5. A group of CARs according to any one of embodiments 1 to 3, wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 500 μM and 100 nM, preferably between 250 μM and 100 nM, more preferably between 125 μM and 100 nM, especially between 50 μM and 100 nM, and
wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 500 μM and 100 nM, preferably between 250 μM and 100 nM, more preferably between 125 μM and 100 nM, especially between 50 μM and 100 nM.
6. A group of CARs according to any one of embodiments 1 to 3, wherein the affinity of each individual antigen binding moiety of a CAR molecule of the group to its target antigen is between 500 μM and 150 nM, preferably between 250 μM and 200 nM, more preferably between 125 μM and 300 nM, especially between 50 μM and 400 nM, and
wherein the affinity of each individual antigen binding moiety of another polypeptide to its target antigen or alternatively the affinity of this other polypeptide to the binding site of its respective CAR molecule is between 500 μM and 150 nM, preferably between 250 μM and 200 nM, more preferably between 125 μM and 300 nM, especially between 50 μM and 400 nM.
7. A group of CARs according to any one of embodiments 1 to 6, wherein each target antigen specifically recognized by the antigen binding moieties of the group of CARs or of other polypeptides being able to bind to CAR molecules of the group is a naturally occurring cellular surface antigen or a polypeptide, carbohydrate or lipid bound to a naturally occurring cellular surface antigen.
8. A group of CARs according to any one of embodiments 1 to 7, wherein the antigen binding moieties of the group of CARs and of other polypeptides being able to bind to CAR molecules of the group bind to one or more target antigens present on a cell, preferably one or more target antigens of a cell, on a solid surface, or a lipid bilayer.
9. A group of CARs according to any one of embodiments 1 to 8, wherein at least one target antigen, to which at least one antigen binding moiety of the group of CARs, or of another polypeptide being able to bind to a CAR molecule of the group, specifically binds, comprises a molecule preferably selected from the group consisting of CD19, CD20, CD22, CD23, CD28, CD30, CD33, CD35, CD38, CD40, CD42c, CD43, CD44, CD44v6, CD47, CD49D, CD52, CD53, CD56, CD70, CD72, CD73, CD74, CD79A, CD79B, CD80, CD82, CD85A, CD85B, CD85D, CD85H, CD85K, CD96, CD107a, CD112, CD115, CD117, CD120b, CD123, CD146, CD148, CD155, CD185, CD200, CD204, CD221, CD271, CD276, CD279, CD280, CD281, CD301, CD312, CD353, CD362, BCMA, CD16V, CLL-1, Ig kappa, TRBC1, TRBC2, CKLF, CLEC2D, EMC10, EphA2, FR-a, FLT3LG, FLT3, Lewis-Y, HLA-G, ICAM5, IGHA1/IgA1, IL-1RAP, IL-17RE, IL-27RA, MILR1, MR1, PSCA, PTCRA, PODXL2, PTPRCAP, ULBP2, AJAP1, ASGR1, CADM1, CADM4, CDH15, CDH23, CDHR5, CELSR3, CSPG4, FAT4, GJA3, GJB2, GPC2, GPC3, IGSF9, LRFN4, LRRN6A/LINGO1, LRRC15, LRRC8E, LRIG1, LGR4, LYPD1, MARVELD2, MEGF10, MPZLI1, MTDH, PANX3, PCDHB6, PCDHB10, PCDHB12, PCDHB13, PCDHB18, PCDHGA3, PEP, SGCB, vezatin, DAGLB, SYT11, WFDC10A, ACVR2A, ACVR2B, anaplastic lymphoma kinase, cadherin 24, DLK1, GFRA2, GFRA3, EPHB2, EPHB3, EPHB4, EFNB1, EPOR, FGFR2, FGFR4, GALR2, GLG1, GLP1R, HBEGF, IGF2R, UNC5C, VASN, DLL3, FZD10, KREMEN2, TMEM169, TMEM198, NRG1, TMEFF1, ADRA2C, CHRNA1, CHRNB4, CHRNA3, CHRNG, DRD4, GABRB3, GRIN3A, GRIN2C, GRIK4, HTR7, APT8B2, NKAIN1, NKAIN4, CACNA1A, CACNA1B, CACNA1I, CACNG8, CACNG4, CLCN7, KCN§ A4, KCNG2, KCNN3, KCNQ2, KCNU1, PKD1L2, PKD2L1, SLC5A8, SLC6A2, SLC6A6, SLC6A11, SLC6A15, SLC7A1, SLC7A5P1, SLC7A6, SLC9A1, SLC10A3, SLC10A4, SLC13A5, SLC16A8, SLC18A1, SLC18A3, SLC19A1, SLC26A10, SLC29A4, SLC30A1, SLC30A5, SLC35E2, SLC38A6, SLC38A9, SLC39A7, SLC39A8, SLC43A3, TRPM4, TRPV4, TMEM16J, TMEM142B, ADORA2B, BAI1, EDG6, GPR1, GPR26, GPR34, GPR44, GPR56, GPR68, GPR173, GPR175, LGR4, MMD, NTSR2, OPN3, OR2L2, OSTM1, P2RX3, P2RY8, P2RY11, P2RY13, PTGE3, SSTR5, TBXA2R, ADAM22, ADAMTS7, CST11, MMP14, LPPR1, LPPR3, LPPR5, SEMA4A, SEMA6B, ALS2CR4, LEPROTL1, MS4A4A, ROM1, TM4SF5, VANGL1, VANGL2, C18orf1, GSGL1, ITM2A, KIAA1715, LDLRAD3, OZD3, STEAP1, MCAM, CHRNA1, CHRNA3, CHRNA5, CHRNA7, CHRNB4, KIAA1524, NRM.3, RPRM, GRM8, KCNH4, Melanocortin 1 receptor, PTPRH, SDK1, SCN9A, SORCS1, CLSTN2, Endothelin converting enzyme like-1, Lysophosphatic acid receptor 2, LTB4R, TLR2, Neurotropic tyrosine kinase 1, MUC16, B7-H4, epidermal growth factor receptor (EGFR), ERBB2, HER3, EGFR variant III (EGFRvIII), HGFR, FOLR1, MSLN, CA-125, MUC-1, prostate-specific membrane antigen (PSMA), mesothelin, epithelial cell adhesion molecule (EpCAM), L1-CAM, CEACAM1, CEACAM5, CEACAM6, VEGFR1, VEGFR2, high molecular weight-melanoma associated antigen (HMW-MAA), MAGE-A 1, IL-13R-α2, disialogangliosides (GD2 and GD3), tumour-associated carbohydrate antigens (CA-125, CA-242, Tn and sialyl-Tn), 4-1BB, 5T4, BAFF, carbonic anhydrase 9 (CA-IX), C-MET, CCR1, CCR4, FAP, fibronectin extra domain-B (ED-B), GPNMB, IGF-1 receptor, integrin α5β1, integrin αvβ3, ITB5, ITGAX, embigin, PDGF-Rα, ROR1, Syndecan 1, TAG-72, tenascin C, TRAIL-R1, TRAIL-R2, NKG2D-Ligands, a major histocompatibility complex (MHC) molecule presenting a tumour-specific peptide epitope, preferably PR1/HLA-A2, a lineage-specific or tissue-specific tissue antigen, preferably CD3, CD4, CD5, CD7, CD8, CD24, CD25, CD34, CD80, CD86, CD133, CD138, CD152, CD319, endoglin, and an MHC molecule.
10. A group of CARs according to any one of embodiments 1 to 9, wherein the ectodomain of the CAR molecules of the group comprises a structurally flexible hinge region interposed between the antigen binding moiety and the transmembrane domain, preferably a hinge region derived from CD8 alpha (amino acid sequence position 138-182 according to UniProtKB/Swiss-Prot P01732-1), or CD28 (amino acid sequence position 114-152 according to UniProtKB/Swiss-Prot P10747), or PD-1 (amino acid sequence position 146-170 according to UniProtKB/Swiss-Prot Q15116), wherein the sequences derived from CD8 alpha, CD28 or PD-1 can be N-terminally and/or C-terminally truncated and can have any length within the borders of the said sequence region, and wherein the cysteine residues in the said hinges derived from CD8 alpha and CD28 are deleted or replaced by other amino acid residues.
11. A group of CARs according to any one of embodiments 1 to 10, wherein the domains of the CAR molecules are derived from different proteins, wherein at least two of these domains are connected via amino acid linker sequences, wherein the linker preferably comprises 1 to 40 amino acids in length.
12. A group of CARs according to any one of embodiments 1 to 11, wherein dimerization of the dimerization domains of at least two CAR molecules of the group, preferably of all CAR molecules of the group, is enhanced by binding of a regulating molecule.
13. A group of CARs according to any one of embodiments 1 to 12, wherein in the case of at least two heterodimerization domains within a CAR molecule the heterodimerization domains of that CAR molecule are separated by the cell membrane, and/or are each the same member of a pair of heterodimerization domains, and/or are members of different pairs of heterodimerization domains and therefore not able to bind to each other in the presence and absence of a regulating molecule in order to prevent the formation of complexes comprising an uncontrolled number of CAR molecules of the group.
14. A group of CARs according to any one of embodiments 1 to 13, wherein the dimerization domains of at least two CAR molecules of the group are selected from: calcineurin catalytic subunit A (CnA), cyclophilin, dihydrofolate reductase (DHFR), gyrase B (GyrB), GAI, GID1, PYL, ABI, preferably from FK506 binding protein 12 (FKBP12, FKBP), FKBP mutant F36V, and FKBP-rapamycin associated protein (FRB) mutant T82L.
15. A group of CARs according to any one of embodiments 1 to 14, wherein dimerization of at least two CAR molecules of the group is mediated by a pair of heterodimerization domains comprising a ligand binding domain from a nuclear receptor and a co-regulator peptide.
16. A group of CARs according to any one of embodiments 1 to 15, wherein dimerization of at least two CAR molecules of the group is mediated by a pair of heterodimerization domains comprising a lipocalin-fold molecule and a lipocalin-fold binding interaction partner.
17. A group of CARs according to any one of embodiments 1 to 16, wherein dimerization of at least two CAR molecules of the group is mediated by a pair of heterodimerization domains comprising a ligand binding domain from a nuclear receptor and a co-regulator peptide, and wherein the ligand binding domain from a nuclear receptor is selected from an estrogen receptor, an ecdysone receptor, a glucocorticoid receptor, an androgen receptor, a thyroid hormone receptor, a mineralocorticoid receptor, a progesterone receptor, a vitamin D receptor, a PPARγ receptor, a PPARβ receptor, a PPARα receptor, a pregnane X receptor, a liver X receptor, a farnesoid X receptor, a retinoid X receptor, a RAR-related orphan receptor, a retinoic acid receptor, and the respective compatible co-regulators of the nuclear receptors selected from SRC1, GRIP1, AIB1, PGC1a, PGC1b, PRC, TRAP220, ASC2, ASC2-1, ASC2-2, CBP, CBP-1, CBP-2, P300, CIA, ARA70, ARA70-1, ARA70-2, NSD1, SMAP, Tip60, ERAP140, Nix1, LCoR, N-CoR, SMRT, RIP140, RIP140-1, RIP140-2, RIP140-3, RIP140-4, RIP140-5, RIP140-6, RIP140-7, RIP140-8, RIP140-9, PRIC285, PRIC285-1, PRIC285-2, PRIC285-3, PRIC285-4, PRIC285-5, SRC1-1, SRC1-2, SRC1-3, SRC1-4a, SRC1-4b, SRC2, SRC3, SRC3-1, PGC1, TRAP220-1, TRAP220-2, NR0B1, NRIP1, TIF1, TIF2, CoRNR Box, CoRNR1, CoRNR2, αβV, EA2, TA1, EAB1, GRIP1-1, GRIP1-2, GRIP1-3, AIB1-1, AIB1-2, AIB1-3, PGC1a, PGC1b.
18. A group of CARs according to any one of embodiments 1 to 17, wherein dimerization of at least two CAR molecules of the group is mediated by a pair of heterodimerization domains comprising a lipocalin-fold molecule and a lipocalin-fold binding interaction partner, and wherein the lipocalin-fold molecule is a derivative of a naturally occurring lipocalin or iLBP with up to 15, up to 30, or up to 50 amino acid deletions and/or up to 15, up to 30, or up to 50 amino acid insertions outside of the structurally conserved β-barrel structure, preferably corresponding structurally to the regions of amino acid residues selected from

[0419] amino acid residues 1-20, 31-40, 48-51, 59-70, 79-84, 89-101, 110-113, 121-131 and 139-183 in human RBP4, which define the regions adjoining the structurally conserved β-strands in human RBP4 according to the amino acid residue numbering scheme in the PDB entry 1RBP;

[0420] amino acid residues 1-13, 24-36, 44-47, 55-61, 70-75, 80-83, 92-95, 103-110 and 118-158 in human TLC (according to the amino acid residue numbering scheme in Schiefner et al., Acc Chem Res. 2015; 48(4):976-985), which define the regions adjoining the structurally conserved β-strands in human TLC;

[0421] amino acid residues 1-43, 54-68, 76-80, 88-95, 104-109, 114-118, 127-130, 138-141 and 149-188 in human ApoM (according to the amino acid residue numbering scheme in Schiefner et al., Acc Chem Res. 2015; 48(4):976-985), which define the regions adjoining the structurally conserved β-strands in human ApoM;

[0422] amino acid residues 1-4, 13-40, 46-49, 55-60, 66-70, 74-80, 88-92, 97-107, 113-118, 125-128 and 136-137 in human CRABPII (according to the amino acid residue numbering scheme in PDB entry 2FS6), which define the regions adjoining the structurally conserved β-strands in human CRABPII;

[0423] amino acid residues 1-4, 13-38, 44-47, 53-58, 64-68, 72-78, 86-90, 95-98, 104-108, 115-118 and 126-127 in human FABP1 (according to the amino acid residue numbering scheme in PDB entry 2F73), which define the regions adjoining the structurally conserved β-strands in human FABP1.

19. A group of CARs according to any one of embodiments 1 to 18, wherein dimerization of at least two CAR molecules of the group is mediated by a pair of heterodimerization domains comprising a lipocalin-fold molecule and a lipocalin-fold binding interaction partner, and wherein the lipocalin-fold molecule is a derivative of a naturally occurring lipocalin or iLBP with at least 70%, preferably at least 80%, especially at least 90% sequence identity in the β-barrel structure, whereby this β-barrel structure is defined as the regions preferably corresponding structurally to the regions of amino acid residues selected from

[0424] amino acid residues 21-30, 41-47, 52-58, 71-78, 85-88, 102-109, 114-120 and 132-138 in human RBP4 (according to the amino acid residue numbering scheme in the PDB entry 1RBP), which define the structurally conserved β-strands in human RBP4;

[0425] amino acid residues 14-23, 37-43, 48-54, 62-69, 76-79, 84-91, 96-102 and 111-117 in human tear lipocalin (TLC; as defined by Schiefner et al., Acc Chem Res. 2015; 48(4):976-985), which define the structurally conserved β-strands in human TLC;

[0426] amino acid residues 44-53, 69-75, 81-87, 96-103, 110-113, 119-126, 131-137 and 142-148 in human apolipoprotein M (ApoM; as defined by Schiefner et al., Acc Chem Res. 2015; 48(4):976-985), which define the structurally conserved β-strands in human ApoM;

[0427] amino acid residues 5-12, 41-45, 50-54, 61-65, 71-73, 81-87, 93-96, 108-112, 119-124 and 129-135 in human cellular retinoic acid binding protein II (CRABPII; according to the amino acid residue numbering scheme in PDB entry 2FS6), which define the structurally conserved β-strands in human CRABPII;

[0428] amino acid residues 5-12, 39-43, 48-52, 59-63, 69-71, 79-85, 91-94, 99-103, 109-114 and 119-125 in human fatty acid binding protein 1 (FABP1; according to the amino acid residue numbering scheme in PDB entry 2F73), which define the structurally conserved β-strands in human FABP1;

20. A group of CARs according to any one of embodiments 1 to 19, wherein dimerization of at least two CAR molecules of the group is mediated by a pair of heterodimerization domains comprising a lipocalin-fold molecule and a lipocalin-fold binding interaction partner, and wherein the lipocalin-fold molecule is a fragment of a naturally occurring lipocalin or a derivative thereof with a length of at least 80, preferably at least 100, especially at least 120, amino acids covering at least the structurally conserved β-barrel structure of the lipocalin-fold, or wherein the lipocalin-fold molecule is a fragment of a naturally occurring iLBP or a derivative thereof with a length of at least 80, preferably at least 85, especially at least 90, amino acids covering at least the structurally conserved β-barrel structure of the lipocalin-fold,
wherein the structurally conserved β-barrel structure comprises or consists of amino acid positions preferably corresponding structurally to the regions of amino acid residues selected from

[0429] amino acid residues 21-30, 41-47, 52-58, 71-78, 85-88, 102-109, 114-120 and 132-138 in human RBP4 (according to the amino acid residue numbering scheme in the PDB entry 1RBP), which define the structurally conserved β-strands in human RBP4;

[0430] amino acid residues 14-23, 37-43, 48-54, 62-69, 76-79, 84-91, 96-102 and 111-117 in human tear lipocalin (TLC; as defined by Schiefner et al., Acc Chem Res. 2015; 48(4):976-985), which define the structurally conserved β-strands in human TLC;

[0431] amino acid residues 44-53, 69-75, 81-87, 96-103, 110-113, 119-126, 131-137 and 142-148 in human apolipoprotein M (ApoM; as defined by Schiefner et al., Acc Chem Res. 2015; 48(4):976-985), which define the structurally conserved β-strands in human ApoM;

[0432] amino acid residues 5-12, 41-45, 50-54, 61-65, 71-73, 81-87, 93-96, 108-112, 119-124 and 129-135 in human cellular retinoic acid binding protein II (CRABPII; according to the amino acid residue numbering scheme in PDB entry 2FS6), which define the structurally conserved β-strands in human CRABPII;

[0433] amino acid residues 5-12, 39-43, 48-52, 59-63, 69-71, 79-85, 91-94, 99-103, 109-114 and 119-125 in human fatty acid binding protein 1 (FABP1; according to the amino acid residue numbering scheme in PDB entry 2F73), which define the structurally conserved β-strands in human FABP1;

21. A group of CARs according to any one of embodiments 1 to 20, wherein a regulating molecule is a molecule which is soluble at the concentrations that can be achieved in the physiological environment in the human body, or under physiological conditions within a cell, at the surface of a cell or under standardised physiological conditions, preferably at PBS conditions, wherein PBS conditions are 137 mM NaCl, 2.7 mM KCl, 10 mM Na.sub.2HPO.sub.4 and 18 mM KH.sub.2PO.sub.4).
22. A group of CARs according to any one of embodiments 1 to 21, wherein at least one regulating molecule is selected from rapamycin, a rapamycin-analog, abscisic acid, gibberellin, or gibberellin-analog GA3-AM, coumermycin, bis-methotrexate, AP20187, or AP1903.
23. A group of CARs according to any one of embodiments 1 to 22, wherein at least one regulating molecule binds to the ligand binding domain from a nuclear receptor and is selected from corticosterone (11beta,21-dihydroxy-4-pregnene-3,20-dione); deoxycorticosterone (21-hydroxy-4-pregnene-3,20-dione); cortisol (11beta,17,21-trihydroxy-4-pregnene-3,20-dione); 11-deoxycortisol (17,21-dihydroxy-4-pregnene-3,20-dione); cortisone (17,21-dihydroxy-4-pregnene-3,11,20-trione); 18-hydroxycorticosterone (11beta,18,21-trihydroxy-4-pregnene-3,20-dione); 1.alpha.-hydroxycorticosterone (1 alpha,11beta,21-trihydroxy-4-pregnene-3,20-dione); aldosterone 18,11-hemiacetal of 11beta,21-dihydroxy-3,20-dioxo-4-pregnen-18-a1; androstenedione (4-androstene-3,17-dione); 4-hydroxy-androstenedione; 11β-hydroxyandrostenedione (11 beta-4-androstene-3,17-dione); androstanediol (3-beta,17-beta-Androstanediol); androsterone (3alpha-hydroxy-5alpha-androstan-17-one); epiandrosterone (3beta-hydroxy-5alpha-androstan-17-one); adrenosterone (4-androstene-3,11,17-trione); dehydroepiandrosterone (3beta-hydroxy-5-androsten-17-one); dehydroepiandrosterone sulphate (3beta-sulfoxy-5-androsten-17-one); testosterone (17beta-hydroxy-4-androsten-3-one); epitestosterone (17alpha-hydroxy-4-androsten-3-one); 5α-dihydrotestosterone (17beta-hydroxy-5alpha-androstan-3-one 5β-dihydrotestosterone; 5-beta-dihydroxy testosterone (17beta-hydroxy-5beta-androstan-3-one); 11β-hydroxytestosterone (11beta,17beta-dihydroxy-4-androsten-3-one); 11-ketotestosterone (17beta-hydroxy-4-androsten-3,17-dione); estrone (3-hydroxy-1,3,5(10)-estratrien-17-one); estradiol (1,3,5(10)-estratriene-3,17beta-diol); estriol 1,3,5(10)-estratriene-3,16alpha,17beta-triol; pregnenolone (3-beta-hydroxy-5-pregnen-20-one); 17-hydroxypregnenolone (3-beta,17-dihydroxy-5-pregnen-20-one); progesterone (4-pregnene-3,20-dione); 17-hydroxyprogesterone (17-hydroxy-4-pregnene-3,20-dione); progesterone (pregn-4-ene-3,20-dione); T3; T4; spironolactone; eplerenone; cyproterone acetate, hydroxyflutamide; enzalutamide; ARN-509; 3,3′-diindolylmethane (DIM); bexlosteride; bicalutamide; N-butylbenzene-sulfonamide (NBBS); dutasteride; epristeride; finasteride; flutamide; izonsteride; ketoconazole; N-butylbenzene-sulfonamide; nilutamide; megestrol; turosteride; mifepristone (RU-486; 11β-[4 N,N-dimethylaminophenyl]-17β-hydroxy-17-(1-propynyl)-estra-4,9-dien-3-one); Lilopristone (11β-(4 N,N-dimethylaminophenyl)-17-hydroxy-17-((Z)-3-hydroxypropenyl)estra-4,9-dien-3-one); onapristone (11β-(4 N,N-dimethylaminophenyl)-17α-hydroxy-17-(3-hydroxypropyl)-13α-estra-4,9-dien-3-one); asoprisnil (benzaldehyde, 4-[(11β,17β)-17-methoxy-17-(methoxymethyl)-3-oxoestra-4,9-dien-11-yl]-1-(E)-oxim; J867); J912 (4-[17β-Hydroxy-17α-(methoxymethyl)-3-oxoestra-4,9-dien-11β-yl]benzaldehyd-(1E)-oxim); CDB-2914 (17α-acetoxy-11β-(4-N,N-dimethylaminophenyl)-19-norpregna-4,9-dien-3,20-dione); JNJ-1250132 (6α,11β,17β)-11-(4-dimethylaminophenyl)-6-methyl-4′,5′-dihydrospiro[estra-4,9-diene-17,2′(3′H)-furan]-3-one (ORG-31710); (11β,17α)-11-(4-acetylphenyl)-17,23-epoxy-19,24-dinorchola-4,9-,20-trien-3-one (ORG-33628); (7β,11β,17β)-11-(4-dimethylaminophenyl-7-methyl]-4′,5′-dihydrospiro[estra-4,9-diene-17,2′(3′H)-furan]-3-one (ORG-31806); ZK-112993; ORG-31376; ORG-33245; ORG-31167; ORG-31343; RU-2992; RU-1479; RU-25056; RU-49295; RU-46556; RU-26819; LG1127; LG120753; LG120830; LG1447; LG121046; CGP-19984A; RTI-3021-012; RTI-3021-022; RTI-3021-020; RWJ-25333; ZK-136796; ZK-114043; ZK-230211; ZK-136798; ZK-98229; ZK-98734; ZK-137316; 4-[17β-Methoxy-17α-(methoxymethyl)-3-oxoestra-4,9-dien-11-yl]benzaldehyde-1-(E)-oxime; 4-[17β-Methoxy-17α-(methoxymethyl)-3-oxoestra-4,9-dien-11β-yl]benzaldehyde-1-(E)-[O-(ethylamino)carbonyl]oxime; 4-[17β-Methoxy-17α-(methoxymethyl)-3-oxoestra-4,9-dien-11β-yl]benzaldehyde-1-(E)-[O-(ethylthio)carbonyl]oxime; (Z)-6′-(4-cyanophenyl)-9,11α-dihydro-17-hydroxy-17α-[4-(1-oxo-3-methylbutoxy)-1-butenyl]4′H-naphtho[3′,2′,1′;10,9,11]estr-4-en-3-one; 11β-(4-acetylphenyl)-17β-hydroxy-17α-(1,1,2,2,2-penta-fluoroethyl)estra-4,9-dien-3-one; 11β-(4-Acetylphenyl)-19,24-dinor-17,23-epoxy-17alpha-chola-4,9,20-trie-n-3-one; (Z)-11beta,19-[4-(3-Pyridinyl)-o-phenylene]-17beta-hydroxy-17α-[3-hydroxy-1-propenyl]-4-androsten-3-one; 11beta-[4-(1-methylethenyl)phenyl]-17α-hydroxy-17beta-R-hydroxypropyl)-13α-estra-4,9-dien-3-one; 4′,5′-Dihydro-11beta-[4-(dimethylamino)phenyl]-6beta-methylspiro[estra-4,-9-dien-17beta,2′(3′H)-furan]-3-one; drospirenone; T3 (3,5,3′-triiodo-L-thyronine); KB-141 (3,5-dichloro-4-(4-hydroxy-3-isopropylphenoxy)phenylacetic acid); sobetirome (3,5-dimethyl-4-(4′-hydroxy-3′-isopropylbenzyl)-phenoxy acetic acid); GC-24 (3,5-dimethyl-4-(4′-hydroxy-3′-benzyl)benzylphenoxyacetic acid); 4-OH-PCB106 (4-OH-2′,3,3′,4′,5′-pentachlorobiphenyl); eprotirome; MB07811 ((2R,4S)-4-(3-chlorophenyl)-2-[(3,5-dimethyl-4-(4′-hydroxy-3′-isopropylbenzyl)phenoxy)methyl]-2-oxido-[1,3,2]-dioxaphosphonane); QH2; (3,5-dimethyl-4-(4′-hydroxy-3′-isopropylbenzyl)phenoxy)methylphosphonic acid (MB07344); tamoxifen; 4-OH-tamoxifen; raloxifene; lasofoxifene, bazedoxifene; falsodex; clomifene; femarelle; ormeloxifene; toremifiene; ospemifene; estradiol (17-beta-estradiol); ethinyl estradiol; a thiazolidinedione (preferably rosiglitazone, pioglitazone, lobeglitazone, troglitazone); farglitazar; aleglitazar; fenofibric acid; benzopyranoquinoline A 276575; Mapracorat; ZK 216348; 55D1E1; dexamethasone; prednisolone; prednisone; methylprednisolone; fluticasone propionate; beclomethasone-17-monopropionate; betamethasone; rimexolone; paramethasone; hydrocortisone; 1,25-dihydroxyvitamin D3 (calcitriol); paricalitol; doxercalciferol; 25-hydroxyvitamin D3 (calcifediol); cholecalciferol; ergocalciferol; tacalciol; 22-dihydroergocalciferol; (6Z)-Tacalciol; 2-methylene-19-nor-20(S)-1α-hydroxy-bishomopregnacalciferol; 19-nor-26,27-dimethylene-20(S)-2-methylene-1α,25-dihydroxyvitamin D3; 2-methylene-1α,25-dihydroxy-(17E)-17(20)-dehydro-19-nor-vitamin D3; 2-methylene-19-nor-(24R)-1α,25-dihydroxyvitamin D2; 2-methylene-(20R,25S)-19,26-dinor-1α,25-dihydroxyvitamin D3; 2-methylene-19-nor-1α-hydroxy-pregnacalciferol; 1α-hydroxy-2-methylene-19-nor-homopregnacalciferol; (20R)-1α-hydroxy-2-methylene-19-nor-bishomopregnacalciferol; 2-methylene-19-nor-(20S)-1α-hydroxy-trishomopregnacalciferol; 2-methylene-23,23-difluoro-1α-hydroxy-19-nor-bishomopregnacalcifero-1; 2-methylene-(20S)-23,23-difluoro-1α-hydroxy-19-nor-bishomopregnan-calciferol; (2-(3′ hydroxypropyl-1′,2′-idene)-19,23,24-trinor-(20S)-1α-hydroxyvitamin D3; 2-methylene-18,19-dinor-(20S)-1α,25-dihydroxyvitamin D3; retinoic acid; all-trans-retinoic acid; 9-cis-retinoic acid; tamibarotene; 13-cis-retinoic acid; (2E,4E,6Z,8E)-3,7-dimethyl-9-(2,6,6-trimethyl-1-cyclohexeneyl)nona-2,4,6,-8-tetraenoic acid; 9-(4-methoxy-2,3,6-trimethyl-phenyl)-3,7-dimethyl-nona-2,4,6,8-tetraenoic acid; 6-[3-(1-adamantyl)-4-methoxyphenyl]-2-napthoic acid; 4-[l-(3,5,5,8,8-pentamethyl-tetralin-2-yl)ethenyl]benzoic acid; retinobenzoic acid; ethyl 6-[2-(4,4-dimethylthiochroman-6-yl)ethynyl]pyridine-3-carboxylate; retinoyl t-butyrate; retinoyl pinacol; retinoyl cholesterol; obeticholic acid; LY2562175 (6-(4-((5-cyclopropyl-3-(2,6-dichlorophenyl)isoxazol-4-yl)methoxy)piperidin-1-yl)-1-methyl-1H-indole-3-carboxylic acid); GW4064 (3-[2-[2-Chloro-4-[[3-(2,6-dichlorophenyl)-5-(1-methylethyl)-4-isoxazolyl]methoxy]phenyl]ethenyl]benzoic acid); T0901317 (N-(2,2,2-Trifluoroethyl)-N-[4-[2,2,2-trifluoro-1-hydroxy-1-(trifluoromethyl)ethyl]phenyl]benzenesulfonamide); GW3965 (3-[3-[[[2-Chloro-3-(trifluoromethyl)phenyl]methyl] (2,2-diphenylethyl)amino]propoxy]benzeneacetic acid hydrochloride); LXR-623; GNE-3500 (27, 1-{4-[3-fluoro-4-((3S,6R)-3-methyl-1,1-dioxo-6-phenyl-[1,2]thiazinan-2-ylmethyl)-phenyl]-piperazin-1-yl}-ethanone); 73, 27-dihydroxycholesterol; 7α, 27-dihydroxycholesterol; 9-cis retinoic acid; LGD100268; CD3254 (3-[4-Hydroxy-3-(5,6,7,8-tetrahydro-3,5,5,8,8-pentamethyl-2-naphthalenyl)phenyl]-2-propenoic acid); CD2915 (Sorensen et al. (1997) Skin Pharmacol. 10:144); rifampicin; chlotrimazole; and lovastatin.
24. A group of CARs according to any one of embodiments 1 to 23, wherein at least one regulating molecule is Tamoxifen and binds to the ligand binding domain from a nuclear receptor, preferably from estrogen receptor alpha or from estrogen receptor beta.
25. A group of CARs according to any one of embodiments 1 to 24, wherein at least one regulating molecule binds to a lipocalin-fold molecule and is selected from fenretinide (Pubchem CID 5288209), N-Ethylretinamide (Pubchem CID 5288173), all-trans retinoic acid (Pubchem CID 444795), axerophthene (Pubchem CID 5287722), A1120 (Pubchem CID 25138295) and derivatives thereof (Cioffi et al., J Med Chem. 2014; 57(18):7731-7757; Cioffi et al., J Med Chem. 2015; 58(15):5863-5888), 1,4-butanediol (Pubchem CID 8064), sphingosine-1-phosphate (Pubchem CID 5283560), tetradecanoic acid (Pubchem CID 11005), indicaxanthin (Pubchem CID 6096870 and 12310796), vulgaxanthin I (Pubchem CID 5281217), Montelukast (Pubchem CID 5281040), Cyclandelate (Pubchem CID 2893), Oxolamine (Pubchem CID 13738), Mazaticol (Pubchem CID 4019), Butoctamid (Pubchem CID 65780), Tonabersat (Pubchem CID 6918324), Novazin (Pubchem CID 65734), Diphenidol (Pubchem CID 3055), Alloclamide (Pubchem CID 71837), Diacetolol (Pubchem CID 50894), Acotiamide (Pubchem CID 5282338), Acoziborole (Pubchem CID 44178354), Acumapimod (Pubchem CID 11338127), Apalutamide (Pubchem CID 24872560), ASP3026 (Pubchem CID 25134326), AZD1480 (Pubchem CID 16659841), BIIB021 (Pubchem CID 16736529), Branaplam (Pubchem CID 89971189), Brequinar (Pubchem CID 57030), Chlorproguanil (Pubchem CID 9571037), Clindamycin (Pubchem CID 446598), Emricasan (Pubchem CID 12000240), Enasidenib (Pubchem CID 89683805), Enolicam (Pubchem CID 54679203), Flurazepam (Pubchem CID 3393), ILX-295501 (Pubchem CID 127737), Indibulin (Pubchem CID 2929), Metoclopramide (Pubchem CID 4168), Mevastatin (Pubchem CID 64715), MGGBYMDAPCCKCT-UHFFFAOYSA-N (Pubchem CID 25134326), MK0686 (Pubchem CID 16102897), Navarixin (Pubchem CID 11281445), Nefazodone (Pubchem CID 4449), Pantoprazole (Pubchem CID 4679), Pavinetant (Pubchem CID 23649245), Proxazole (Pubchem CID 8590), SCYX-7158 (Pubchem CID 44178354), Siccanin (Pubchem CID 71902), Sulfaguanole (Pubchem CID 9571041), Sunitinib (Pubchem CID 5329102), Suvorexant (Pubchem CID 24965990), Tiapride (Pubchem CID 5467), Tonabersat (Pubchem CID 6918324), VNBRGSXVFBYQNN-UHFFFAOYSA-N(Pubchem CID 24794418), YUHNXUAATAMVKD-PZJWPPBQSA-N(Pubchem CID 44548240), Ulimorelin (Pubchem CID 11526696), Xipamide (Pubchem CID 26618), Tropesin (Pubchem CID 47530), Triclabendazole (Pubchem CID 50248), Triclabendazole sulfoxide (Pubchem CID 127657), Triclabendazole sulfone (Pubchem CID 10340439) and Trametinib (Pubchem CID 11707110).
26. A group of CARs according to any one of embodiments 1 to 25, wherein at least one regulating molecule is selected from rapamycin, a rapamycin-analog, AP20187, AP1903, Tamoxifen, Emricasan and A1120.
27. A group of CARs according to any one of embodiments 1 to 26, wherein the order of the domains in the CAR molecules of the group from the extracellular to the intracellular side on the surface of a cell is: an antigen binding moiety or a binding site to which another polypeptide comprising an antigen binding moiety is able to bind, preferably a hinge region for spatial optimization, and a transmembrane domain,
wherein in the preferred case of an ITAM-containing group of CARs the transmembrane domain is preferably followed in at least one CAR molecule by a signalling region comprising a co-stimulatory domain, wherein preferably this co-stimulatory signalling region, or optionally the transmembrane domain, is followed by at least one dimerization domain, and further, in at least one CAR molecule, by a signalling region comprising at least one ITAM, wherein the order of the co-stimulatory and the ITAM-containing signalling region can be inverted, and wherein CAR molecules not comprising an ITAM either lack a co-stimulatory signalling region, or comprise one co-stimulatory signalling region, or two co-stimulatory signalling regions, or even more co-stimulatory signalling regions, but preferably not more than two co-stimulatory signalling regions, or even more preferably only one co-stimulatory signalling region, and
wherein in the case of an ITIM comprising group of CARs the transmembrane domain is preferably followed in at least one CAR molecule by an ITIM-comprising inhibitory signalling region, wherein preferably this inhibitory signalling region, or optionally the transmembrane domain, is followed by at least one dimerization domain, and optionally a second inhibitory signalling region, and
wherein in an ITAM- and ITIM-comprising group of CARs any two adjacent components of a CAR molecule can optionally be separated by a linker, and
wherein in an ITAM- and ITIM-comprising group of CARs the dimerization domains, of which at least one is mandatory for each CAR molecule of the group, can be located alternatively or additionally in the ectodomain or the transmembrane domain, however, preferably between the transmembrane domain and a signalling region, and/or especially between two signalling regions and/or especially at the intracellular end of the CAR molecules.
28. A group of CARs according to any one of embodiments 1 to 27, wherein each CAR molecule of the group comprises at least a signalling region which can transduce a signal via at least one immunoreceptor tyrosine-based activation motif (ITAM) or at least one immunoreceptor tyrosine-based inhibitory motif (ITIM).
29. A group of CARs according to any one of embodiments 1 to 28, wherein the dimerization domains are located in the endodomains and/or the transmembrane domains, preferably in the endodomains, of the CAR molecules of the group.
30. A group of CARs according to any one of embodiments 1 to 28, wherein the ectodomains of at least two of the CAR molecules of the group comprise a dimerization domain which preferably comprises only one protein domain, and wherein the regulating molecule is a molecule secreted from a cell and induces dimerization of said dimerization domains.
31. A group of CARs according to any one of embodiments 1 to 30, wherein at least one CAR molecule of the group comprises an endodomain containing a signalling region which can transduce a signal via at least one ITAM, and wherein the group of CARs preferably comprises at least three ITAMs.
32. A group of CARs according to any one of embodiments 1 to 31, wherein the endodomain of at least one CAR molecule of the group comprises at least one ITAM, said ITAM is selected from CD3 zeta, DAP12, Fc-epsilon receptor 1 gamma chain, CD3 delta, CD3 epsilon, CD3 gamma, and CD79A (antigen receptor complex-associated protein alpha chain), preferably CD3 zeta.
33. A group of CARs according to any one of embodiments 1 to 32, wherein the co-stimulatory domain of a co-stimulatory signalling region in the endodomain of a CAR molecule of the group is derived from 4-1BB (CD137), CD28, ICOS, BTLA, OX-40, CD2, CD6, CD27, CD30, CD40, GITR, and HVEM, preferably 4-1BB and ICOS.
34. A group of CARs according to any one of embodiments 1 to 30, wherein at least one CAR molecule of the group comprises an endodomain containing a cytoplasmic inhibitory domain derived from PD-1, CD85A, CD85C, CD85D, CD85J, CD85K, LAIR1, TIGIT, CEACAM1, CD96, KIR2DL, KIR3DL, SLAM family members, CD300/LMIR family members, CD22 and other Siglec family members.
35. A group of CARs according to any one of embodiments 1 to 30 and 34, wherein at least one CAR molecule of the group comprises an endodomain containing a signalling region which can transduce a signal via at least one ITIM which is derived from PD-1, CD85A, CD85C, CD85D, CD85J, CD85K, LAIR1, TIGIT, CEACAM1, CD96, KIR2DL, KIR3DL, SLAM family members, ITIM comprising CD300/LMIR family members, CD22 and other ITIM comprising Siglec family members.
36. A group of CARs according to any one of embodiments 1 to 35, wherein the ectodomain of each CAR molecule comprises an antigen binding moiety.
37. A group of CARs according to any one of embodiments 1 to 36, wherein the group consists of two or three CAR molecules, preferably two CAR molecules.
38. A group of CARs according to any one of embodiments 1 to 14, 21, 22, 26 to 29 or 31 to 37, wherein the group consists of three CAR molecules, and wherein the endodomain of one of the three CAR molecules comprises two copies of the same member of a pair of dimerization domains selected from either FKBP12 or alternatively FRB (mutant T82L), and wherein the endodomains of each of the other two CAR molecules comprise one copy of the other member of the pair of dimerization domains, respectively, and wherein the regulating molecule is a rapalog.
39. A group of CARs according to any one of embodiments 1 to 13, 15, 17, 23, 24, 26 to 29 or 31 to 37, wherein the group consists of three CAR molecules, and wherein the endodomain of one of the three CAR molecules comprises two copies of the same member of a pair of dimerization domains selected from either the ligand binding domain of a nuclear receptor or alternatively a compatible co-regulator of the same nuclear receptor, and wherein the endodomains of each of the other two CAR molecules comprise one copy of the other member of the pair of dimerization domains, respectively, and wherein the regulating molecule is Tamoxifen.
40. A group of CARs according to any one of embodiments 1 to 13, 16, 18 to 21, 25 to 29 or 31 to 37, wherein the group consists of three CAR molecules, and wherein the endodomain of one of the three CAR molecules comprises two copies of the same member of a pair of dimerization domains selected from either a lipocalin-fold molecule or a lipocalin-fold binding interaction partner, and wherein the endodomains of each of the other two CAR molecules comprise one copy of the other member of the pair of dimerization domains, respectively, and
wherein the regulating molecule is selected from fenretinide (15-[(4-hydroxyphenyl)amino]retinal), N-Ethylretinamide, all-trans retinoic acid, axerophthene, A1120 (PubChem CID 25138295) and derivatives thereof, 1,4-butanediol, sphingosine-1-phosphate, tetradecanoic acid, indicaxanthin, vulgaxanthin, Montelukast, Cyclandelate, Oxolamine, Mazaticol, Butoctamid, Tonabersat, Novazin, Diphenidol, Neobornyval, Alloclamide, Diacetolol, Acotiamide, Acoziborole, Acumapimod, Apalutamide, ASP3026, AZD1480, BIIB021, Branaplam, Brequinar, Chlorproguanil, Clindamycin, Emricasan, Enasidenib, Enolicam, Flurazepam, ILX-295501, Indibulin, Metoclopramide, Mevastatin, MGGBYMDAPCCKCT-UHFFFAOYSA-N, MK0686, Navarixin, Nefazodone, Pantoprazole, Pavinetant, Proxazole, SCYX-7158, Siccanin, Sulfaguanole, Sunitinib, Suvorexant, Tiapride, Tonabersat, VNBRGSXVFBYQNN-UHFFFAOYSA-N, YUHNXUAATAMVKD-PZJWPPBQSA-N, Ulimorelin, Xipamide, Tropesin, Triclabendazole, Triclabendazole sulfoxide, Triclabendazole sulfone and Trametinib.
41. A nucleic acid molecule comprising nucleotide sequences encoding the individual CAR molecules of a group of CARs according to any one of embodiments 1 to 40, 64 or 65, wherein the nucleic acid is selected from DNA, RNA, or in vitro transcribed RNA.
42. A kit of nucleic acid molecules comprising nucleotide sequences encoding the individual CAR molecules of a group of CARs according to any one of embodiments 1 to 40, 64 or 65, wherein the nucleic acid is selected from DNA, RNA, or in vitro transcribed RNA.
43. A kit of nucleic acid molecules according to embodiment 42, wherein the nucleic acid molecules are present in a vector and preferably packaged as DNA or RNA into an infectious virus particle.
44. A nucleic acid molecule or a kit of nucleic acid molecules according to any one of embodiments 41 to 43, wherein the nucleic acid sequences are linked to a sequence mediating strong and stable transgene expression in lymphocytes, wherein such a sequence preferably comprises the 5′-LTR of a gamma retrovirus or subelements R and U3 of a 5′-LTR of the Moloney murine leukaemia virus (MMLV) or the promoter of the murine stem cell virus (MSCV) or the promoter of phosphoglycerate kinase (PGK) or even more preferably the human elongation factor 1 (EF-1) alpha promoter.
45. A kit of nucleic acid molecules according to any one of embodiments 42 to 44, wherein the first nucleic acid comprises nucleotide sequences encoding a first CAR molecule of the group and wherein the second nucleic acid comprises nucleotide sequences encoding a second CAR molecule of the group and, optionally, wherein the kit further comprises a third nucleic acid, said third nucleic acid comprising nucleotide sequences encoding a third CAR molecule of the group if the group of CARs consists of at least three different CAR molecules and, optionally, wherein the kit further comprises a fourth nucleic acid, said fourth nucleic acid comprising nucleotide sequences encoding a fourth CAR molecule of the group if the group of CARs consists of four different CAR molecules.
46. A vector or a kit of vectors comprising nucleotide sequences encoding the individual CAR molecules of a group of CARs according to any one of embodiments 1 to 40, 64 or 65, wherein the nucleic acid is DNA or RNA.
47. A vector or a kit of vectors according to embodiment 46, wherein the vector is a recombinant adeno-associated virus (rAAV) vector or a transposon vector, preferably a Sleeping Beauty transposon vector or PiggyBac transposon vector, or wherein a vector is a retroviral vector, preferably a gamma-retroviral vector or a lentiviral vector.
48. A vector or a kit of vectors according to embodiments 46 or 47, wherein the vector is an expression vector, preferably an expression vector in which the nucleotide sequences are operably linked to a sequence mediating strong and stable transgene expression in lymphocytes, wherein such a sequence preferably comprises the 5′-LTR of a gamma retrovirus or subelements R and U3 of a 5′-LTR of the Moloney murine leukaemia virus (MMLV) or the promoter of the murine stem cell virus (MSCV) or the promoter of phosphoglycerate kinase (PGK) or even more preferably the human elongation factor 1 (EF-1) alpha promoter.
49. A kit of vectors according to any one of embodiments 46 to 48, wherein the first vector comprises a nucleotide sequence encoding a first CAR molecule of the group and wherein the second vector comprises a nucleotide sequence encoding a second CAR molecule of the group and, optionally, wherein the kit further comprises a third vector, said third vector comprising a nucleotide sequence encoding a third CAR molecule of the group if the group of CARs consists of at least three different CAR molecules and, optionally, wherein the kit further comprises a fourth vector, said fourth vector comprising a nucleotide sequence encoding a fourth CAR molecule of the group if the group of CARs consists of four different CAR molecules.
50. A cell modified in vitro or ex vivo with a nucleic acid molecule or a kit of nucleic acid molecules according to any one of embodiments 41 to 45 or with a vector or a kit of vectors according to any one of embodiments 46 to 49 to produce the individual CAR molecules of a group of CARs according to any one of embodiments 1 to 40, 64 or 65, or a kit comprising two or more of said modified cells.
51. A cell or kit of cells according to embodiment 50, wherein the cell is a mammalian cell, preferably a hematopoietic stem cell (HSC), or a cell derived from a HSC, more preferably an NK cell or a T cell, especially a T cell.
52. A cell or kit of cells according to embodiments 50 or 51, wherein the cell is transfected or transduced with a vector or with a kit of vectors according to any one of embodiments 46 to 49.
53. A cell or kit of cells according to any one of embodiments 50 to 52, wherein the cell has stably integrated the nucleotide sequences encoding a group of CARs according to any one of embodiments 1 to 40, 64 or 65 into its genome.
54. A cell or kit of cells according to any one of embodiments 50 to 52, wherein the cell has stably integrated the nucleotide sequences encoding a group of CARs according to any one of embodiments 1 to 40, 64 or 65 into its genome by the use of site directed nuclease technology, preferably by the use of zinc finger nucleases or TALENs, or even more preferably CRISPR/Cas technology.
55. A pharmaceutical preparation comprising a nucleic acid or a kit of nucleic acids according to any one of embodiments 41 to 45, a vector or a kit of vectors according to any one of embodiments 46 to 49, or a cell or a kit of cells according to any one of embodiments 50 to 54.
56. A pharmaceutical preparation according to embodiment 55, wherein the viral vectors are preferably contained in infectious virus particles.
57. A method of making a cell according to any one of embodiments 50 to 54, the method comprising introducing into the cell, preferably stably integrating into the genome of the cell, in vitro or ex vivo a nucleic acid molecule or a kit of nucleic acid molecules according to any one of embodiments 41 to 45, or a vector or a kit of vectors according to any one of embodiments 46 to 49.
58. A group of CARs according to any one of embodiments 1 to 40, 64 or 65 for use in a method of treatment of a cancer in an individual, wherein the method comprises:
i) genetically modifying NK cells or preferably T lymphocytes obtained from the individual with at least one nucleic acid molecule comprising sequences encoding the respective CAR molecules of the group of CARs, wherein each antigen binding moiety of the group of CARs is specific for a target antigen on a cancer cell in the individual, and wherein said genetic modification is carried out in vitro or ex vivo;
ii) introducing the genetically modified cells into the individual; and
iii) administering to the individual an effective amount of at least one regulating molecule for either inducing or reducing dimerization of the respective CAR molecules of the group, preferably inducing dimerization of the respective CAR molecules of the group, thereby either inducing or reducing non-covalent complexation of the group of CARs, preferably inducing non-covalent complexation of the group of CARs, wherein the non-covalently complexed group of CARs upon contact with a cancer cell expressing the respective target antigen or the respective covalent or non-covalent complex of different target antigens mediates activation of the genetically modified cell, which leads to killing of the cancer cell and thereby enables treating the cancer.
59. A cell according to any one of embodiments 50 to 54 for use in a method of treatment of a cancer in an individual, wherein each antigen binding moiety of the group of CARs is specific for a target antigen on a cancer cell in the individual, and wherein the method comprises:
i) introducing the cell into the individual; and
ii) administering to the individual an effective amount of at least one regulating molecule for either inducing or alternatively reducing, preferably inducing, the formation of a non-covalent complex comprising two, three or four CAR molecules of the group, preferably three CAR molecules of the group, even more preferably two CAR molecules of the group, wherein the non-covalently complexed group of CARs upon contact with a cancer cell expressing the respective target antigen or the respective covalent or non-covalent complex of different target antigens mediates activation of the genetically modified cell, which leads to killing of the cancer cell and thereby enables treating the cancer.
60. A kit comprising:

[0434] a group of CARs according to any one of embodiments 1 to 40, 64 or 65, a vector or kit of vectors according to any one of embodiments 46 to 49, or a cell or a kit of cells according to any one of embodiments 50 to 54, and

[0435] one, two or three regulating molecules, preferably two, even more preferably one regulating molecule.

61. A group of CARs according to any one of embodiments 1 to 40, 64 or 65, a vector or kit of vectors according to any one of embodiments 46 to 49, a cell or kit of cells according to any one of embodiments 50 to 54, especially a T lymphocyte or NK cell, or a kit according to any one of embodiments 42 to 49 or 60 for use in the treatment of a disease which is characterised by the need to bind a T lymphocyte or an NK cell to a target antigen on a cell, preferably for use in the treatment of a tumour patient, especially a tumour patient with a tumour selected from Ewing's sarcoma, rhabdomyosarcoma, osteosarcoma, osteogenic sarcoma, mesothelioma, fibrosarcoma, myxosarcoma, liposarcoma, chondrosarcoma, chordoma, angiosarcoma, endotheliosarcoma, lymphangiosarcoma, lymphangioendotheliosarcoma, synovioma, leiomyosarcoma, melanoma, glioma, astrocytoma, medulloblastoma, neuroblastoma, retinoblastoma, oligodendroglioma, menangioma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, chronic myeloproliferative syndromes, acute myelogenous leukemias, chronic lymphocytic leukemias (CLL) including B-cell CLL, T-cell CLL, prolymphocytic leukemia and hairy cell leukemia, acute lymphoblastic leukemias, B-cell lymphomas, Hodgkin's lymphoma, non-Hodgkin's lymphoma, esophageal carcinoma, hepatocellular carcinoma, basal cell carcinoma, squamous cell carcinoma, bladder carcinoma, transitional cell carcinoma, bronchogenic carcinoma, colon carcinoma, colorectal carcinoma, gastric carcinoma, lung carcinoma, including small cell carcinoma and non-small cell carcinoma of the lung, adrenocortical carcinoma, thyroid carcinoma, pancreatic carcinoma, breast carcinoma, ovarian carcinoma, prostate carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinoma, cystadenocarcinoma, medullary carcinoma, renal cell carcinoma, ductal carcinoma, bile duct carcinoma, choriocarcinoma, seminoma, embryonal carcinoma, Wilm's tumor, cervical carcinoma, uterine carcinoma, testicular carcinoma, osteogenic carcinoma, epithelial carcinoma, and nasopharyngeal carcinoma, atypical meningioma, islet cell carcinoma, medullary carcinoma, mesenchymoma, hepatocellular carcinoma, hepatoblastoma, clear cell carcinoma, and neurofibroma mediastinum.
62. A method of modulating the activity of a T lymphocyte or an NK cell, the method comprising contacting the T lymphocyte or NK cell in vivo or ex vivo in the presence of at least one regulating molecule with a target antigen, or a non-covalent or a covalent complex of different target antigens, on a cell or solid surface, wherein the T lymphocyte or NK cell is genetically modified to produce the group of CARs according to any one of embodiments 1 to 40, 64 or 65, and wherein the presence of the regulating molecule(s) either induces or alternatively reduces, preferably induces the formation of a non-covalently complexed group of CARs and thereby modulates at least one activity of the genetically modified cell.
63. A method according to embodiment 62, wherein the activity of the genetically modified cell is proliferation, cell survival, apoptosis, gene expression, or immune activation.
64. A group of CARs according to any one of embodiments 1 to 40, wherein the non-covalent complexation of the group of CARs is induced in the presence of an effective concentration of one or more regulating molecules, and wherein the group of CARs, when expressed in an NK cell or preferably a T lymphocyte, elicits in its non-covalently complexed state a response in the host cell upon contact with a target antigen expressing target cell, wherein this response is defined by the excretion of interferon-gamma, and/or Macrophage inflammatory protein-1 (MIP-1) alpha, and/or MIP-1 beta, and/or granzyme B, and/or IL-2, and/or TNF, and/or IL-10, and/or IL-4, and/or by cell degranulation, wherein cell degranulation is preferably detected by the percentage of CD107a positive effector cells, after contact with a target cell expressing at least 100,000 molecules of each target antigen recognized by the group of CARs, wherein the response elicited in the presence of an effective concentration of all regulating molecules required for inducing the non-covalent complexation of the group of CARs is at least 20% higher, preferably at least 50% higher, and even more preferably at least 100% higher than the response elicited in the absence of any regulating molecule, wherein the effective concentration of each required regulating molecule is the concentration achieved by administration of an effective amount of each required regulating molecule in one or more doses to an individual in need thereof.
65. A group of CARs according to any one of embodiments 1 to 11 or 13 to 40, wherein the non-covalent complexation of the group of CARs is reduced by the presence of an effective concentration of one or more regulating molecules, and wherein the group of CARs, when expressed in an NK cell or preferably a T lymphocyte, elicits in its non-covalently complexed state a response in the host cell upon contact with a target antigen expressing target cell, wherein this response is defined by the excretion of interferon-gamma, and/or Macrophage inflammatory protein-1 (MIP-1) alpha, and/or MIP-1 beta, and/or granzyme B, and/or IL-2, and/or TNF, and/or IL-10, and/or IL-4, and/or by cell degranulation, wherein cell degranulation is preferably detected by the percentage of CD107a positive effector cells, after contact with a target cell expressing at least 100,000 molecules of each target antigen recognized by the group of CARs, wherein the response elicited by the group of CARs in the absence of any regulating molecule is at least 20% higher, preferably at least 50% higher, and even more preferably at least 100% higher than the response elicited by the group of CARs in the presence of an effective concentration of each regulating molecule required for reducing dimerization of the dimerization domains comprised by the group of CARs, wherein the effective concentration of each required regulating molecule is the concentration achieved by administration of an effective amount of each required regulating molecule in one or more doses to an individual in need thereof.