The invention is related to the field of cancer therapy. In particular, it relates to novel targeting units and chimeric antigen receptors (CARs) comprising them, nucleic acids encoding the targeting units, nucleic acids encoding the CARs, immune cells expressing CARs and their utility in medicine for treatment of osteosarcoma.

BACKGROUND

Osteosarcoma (OS) is the most common primary bone cancer and the eighth most common form of cancer among children and adolescents.

One clinical trial concerning sarcoma patients including osteosarcoma patients has been reported based on targeting of HER2 by autologous T cells expressing a CAR (Ahmed et al. 2015, J Clin Oncol. 2015 May 20;33(15): 1688-96). In this dose- escalation study, the clinical benefit of HER2-CAR T cell infusion was clearly limited, indicating that further manipulation of the immune system will be essential for worthwhile benefits to be obtained.

In order to achieve a therapeutic CAR T cell, the cell needs to express the CAR in a sufficient amount in the cell membrane, and the antigen binding domain has to convey sufficient affinity and specificity in the CAR construct. It can be expected that only a fraction of CAR T cells with in vitro activity will successfully migrate to tumor metastases in vivo and infiltrate the hostile tumor microenvironment of a solid tumor like OS. Furthermore, the CAR T cells will likely need to sustain their activity over time in order to provide a therapeutic effect in vivo. It is therefore not trivial, but very desirable to obtain novel CARs able to provide a therapeutic effect on OS in vivo when expressed in the cell membrane of immune cells.

Five drug products based on CARs are currently approved:

Kymriah (tisagenlecleucel),

Yescarta (axicabtagene ciloleucel),

Tecartus (brexucabtagene autoleucel),

Breyanzi (lisocabtagene maraleucel) and Abecma (idecabtagene vicleucel)

According to the FDA approved package inserts, all these CARs comprise a murine scFv.

However, murine sequences may trigger unwanted immunogenicity against the therapeutic cells during therapy in humans. Successful humanization of murine antibodies has been reported, but in the context of CAR therapy, retained specificity and affinity for the target antigen is utterly important. SUMMARY OF THE INVENTION

Provided herein are proteins comprising a novel antigen binding domain. Said antigen binding domain, and proteins comprising it, are able to bind to osteosarcoma cells under physiological conditions. The antigen binding domain comprises an antibody light chain variable domain (VL) and an antibody heavy chain variable domain (VH), each comprising three complementarity determining regions (CDRs) flanked by human framework sequences. The human framework sequences reduces the risk of triggering unwanted immunogenic responses against the antigen binding domain. This may be especially important in a therapeutic setting wherein a drug is administered repeatedly to a human osteosarcoma patient. As demonstrated in Example 2 and visualized in Figures 4 and 5, it is found that T cells expressing CARs comprising humanized scFv’s derived from the TP1 Fv, have increased activity in a cytokine-assay involving osteosarcoma cells relative to T cells expressing CARs comprising the murine Fv. It is also found that T cells expressing CARs comprising humanized scFv’s derived from the TP3 Fv, have reatained activity in a cytokine-assay involving osteosarcoma cells relative to T cells expressing CARs comprising the murine Fv. The latter may be beneficial because CARs comprising the murine Fv have already been demonstrated to work well in vivo.

Interestingly, when tested in a cytotoxic assay (Figure 5), the same humanized CARs expressed in primary human T cells were either very similar to the original murine construct (OSCAR-1) or a bit weaker (OSCAR-3). It demonstrates that structure-prediction algorithms are limited in predicting the impact of the modifications on the cellular outcome of the humanized CAR.

In relation to CAR therapy, transient expression of CARs via mRNA may be an attractive option due to the inherent risk of off-target effects and on-target off- tumor effects from CAR T cells. However, the transient expression may require more frequent admin sitration of the therapeuctic cells in order to acheive a therapeutic effect. During such repeated treatment, the CARs herein may be particularly useful

In one aspect the present invention provides a protein for targeting of osteosarcoma cells, comprising a VL and a VH which together form an antigen binding unit, wherein the VL comprises three complementarity determining regions (CDRs); CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 1, 2 and 3 and wherein the VH comprises three CDRs; CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 4, 5 and 6, and wherein all the CDRs are flanked by human framework sequences, and optionally wherein each of the human framework sequences, independently, comprises 0 to 5 amino acid substitutions.

In one embodiment of the first aspect the VL is selected from the group consisting of VL1A (SEQ ID NO: 13), VL1B (SEQ ID NO: 14), VL1C (SEQ ID NO: 15), VL1D (SEQ ID NO: 16), VL1E (SEQ ID NO: 17), VL1F (SEQ ID NO: 18), VL1G (SEQ ID NO: 19) and VL1H (SEQ ID NO: 20), and the VH is selected from the group consisting of VH1A (SEQ ID NO: 21), VH1B (SEQ ID NO: 22) and VH1C (SEQ ID NO: 23), VH1D (SEQ ID NO: 24), VH1E (SEQ ID NO: 25), VH1F (SEQ ID NO: 26), VH1G (SEQ ID NO: 27) and VHIH (SEQ ID NO: 28).

In one embodiment of the first aspect the antigen binding unit is selected from the scFv’s represented by SEQ ID NO: 45 to SEQ ID NO: 78.

In a second aspect the present invention provides a Chimeric Antigen Receptor (CAR) comprising the protein according to the first aspect and any embodiments thereof.

In one embodiment of the second aspect the CAR according comprising a human CD8a hinge.

In a third aspect the present invention provides a nucleic acid encoding the protein according to the first aspect or the CAR according to the second aspects.

In a fourth aspect the present invention provides an immune cell expressing a CAR according according to the second aspect in its cell membrane.

In a fifth aspect the present invention provides a pharmaceutical composition comprising a protein according to the first aspects and embodiments therein.

In one aspect the present invention provides a pharmaceutical composition comprising a cytotoxic immune cell according to the fourth aspect. In a sixth aspect the present invention provides a method of treatment of osteosarcoma in a human patient comprising the step of administering a pharmaceutical composition according to the fifthe aspect.

In a seventh aspect the present invention provides an antibody comprising the protein according to the first aspect and embodiments therein.

In a eighth aspect the present invention provides a pharmaceutical composition comprising the antibody according to the seventh aspect for use in treatment of osteosarcoma.

In a ninth aspect the present invention provides method for treatment of osteosarcoma in a human patient comprising the step of administering a pharmaceutical composition according to the eighth aspect.

In a tenth aspect the present invention provides an antibody according to the seventh aspect for use in an in vitro diagnostic method.

In an eleventh aspect the present invention provides a protein for targeting of osteosarcoma cells, comprising a VL and a VH which together form an antigen binding unit, wherein the VL comprises three complementarity determining regions (CDRs); CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 7, 8 and 9 and wherein the VH comprises three CDRs; CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 10, 11 and 12.

In one embodiment of the eleventh aspect the VL is selected from the group consisting of VL3A (SEQ ID NO: 29), VL3B (SEQ ID NO: 30) and VL3C (SEQ ID NO: 31), VL3D (SEQ ID NO: 32), VL3E (SEQ ID NO: 33), VL3F (SEQ ID NO: 34), VL3G (SEQ ID NO: 35), VL3H (SEQ ID NO: 36) and the VH is selected from the group consisting of VH3A (SEQ ID NO: 37), VH3B (SEQ ID NO: 38) and VH3C (SEQ ID NO: 39), VH3D (SEQ ID NO: 40), VH3E (SEQ ID NO: 41), VH3F (SEQ ID NO: 42), VH3G (SEQ ID NO: 43), VH3H (SEQ ID NO: 44).

In one embodiment of the eleventh aspect the antigen binding unit is selected from the scFv’s represented by SEQ ID NO: 79 to SEQ ID NO: 112. In a twelfth aspect the present invention provides a Chimeric Antigen Receptor (CAR) comprising the protein according to the eleventh aspect and any embodiments thereof.

In one embodiment of the twelfth aspect the CAR according comprising a human CD8a hinge.

In a thirdteenth aspect the present invention provides a nucleic acid encoding the protein according to the eleventh aspect or the CAR according to the twelfth aspects.

In a fourteenth aspect the present invention provides an immune cell expressing a CAR according according to the twelvth aspect in its cell membrane.

In a fifteenth aspect the present invention provides a pharmaceutical composition comprising a protein according to the eleventh aspects and embodiments therein.

In one aspect the pharmaceutical composition comprising a cytotoxic immune cell according to the fourteenth aspect.

In a sixteenth aspect the present invention provides a method of treatment of osteosarcoma in a human patient comprising the step of administering a pharmaceutical composition according to the above aspect.

In a seventeenth aspect the present invention provides an antibody comprising the protein according to the eleventh aspect and embodiments therein.

In a eighteenth aspect the present invention provides a pharmaceutical composition comprising the antibody according to the seventeenth aspect for use in treatment of osteosarcoma.

In a ninteenth aspect the present invention provides method for treatment of osteosarcoma in a human patient comprising the step of administering a pharmaceutical composition according to the eighteenth aspect.

In one more aspect the present invention provides an antibody according to the seventeenth aspect for use in an in vitro diagnostic method. BRIEF DESCRIPTION OF THE DRAWINGS:

Figure 1 visualizes the process of humanization of a second -generation CAR which initially contained a murine scFv. From left to right: murine scFv-CD8a hinge- CD8a transmembrane domain-4- IBB costimulatory άoΐMΐh-Eϋ3z signaling domain.

Figure 2A depicts substitutions in positions relative to the native human framework sequence for VH1A, VH1B and VH1C. Mutations from native are represented by an X above the mutant sequence.

Figure 2B depicts substitutions in positions relative to the native human framework sequence for VF1A, VF1B and VF1C. Mutations from native are represented by an X above the mutant sequence.

Figure 2C depicts substitutions in positions relative to the native human framework sequence for VH3A, VH3B and VH3C. Mutations from native are represented by an X above the mutant sequence.

Figure 2D depicts substitutions in positions relative to the native human framework sequence for VF3A, VF3B and VF3C. Mutations from native are represented by an X above the mutant sequence.

Figure 3A Visualization of the parent scFv comprising SEQ ID NO: 113 and SEQ ID NO: 114 with the murine CDRs coloured dark blue(l), light blue (2), dark green (3), light green (4), yellow(5) and red(6). The murine framework sequences are coloured purple (7).

Figure 3B Overlay of the 25 novel scFv’s (SEQ ID NO: 45 to SEQ ID NO: 78) wherein the human framework sequences are coloured gray (8) and with the murine CDRs coloured dark blue (1), light blue (2), dark green (3), light green (4), yellow(5) and red (6).

Figure 4 In vitro CAR-dependent T cell stimulation by target cell lines was demonstrated by incubating the T cell line, Jurkat-76 constitutively expressing the reporter NFAT-GFP, co-expressing the indicated CAR constructs with osteosarcoma cell line (OSA). Clone 1326 (TP1 scFv-CD8a hinge-CD8a transmembrane domain-4- IBB costimulatory άoh^ΐh^ϋ3z signaling domain = OSCAR-1) and clone 1327 (TP3 scFv-CD8a hinge-CD8a transmembrane domain-4- 1BB costimulatory άoh^ΐh^ϋ3z signaling domain = OSCAR-3) are second generation CARs comprising murine VL and VH sequences. Clone 1668, 1669,

1670 and 1671 are CARs with the same structure, but containing humanized scFv’s derived from TP1, while clone 1672, 1673, 1674 and 1675 are CARs with the same structure, but containing humanized scFv’s derived from TP3. PMA/ionomycin is a positive control for maximal stimulation. No target is J76-NFAT-CAR without target cells which sets the background levels. In this system, when J76-NFAT-GFP are stimulated through the T cell Receptor (TCR) signalling pathway (which is mimicked by the CAR), the NFAT promoter is activated and GFP is produced. The accumulation of GFP in the cytosol can be detected by fluorescent analysis (flow cytometry) and quantified. Here the GFP signal was quantified after 24 hours.

Figure 5a-d: killing assay wherein CAR-expressing human primary T cells or mock T cells (control, without CAR) were co-incubated with ALPL negative (BL41) or ALPL positive (OS A) cells at the indicated E:T ratio for 4 hours (OSCAR- 1) or 8 hours (OSCAR-3). The killing was monitored by BLI assay where the level of luciferase expressed by the target cells was detected and use to quantify the specific killing. As shown BL41 are not killed by T cells expressing OSCAR- 1 or OSCAR-3 (a, b), whereas CAR-expressing T cells are able to lyse OSA cells (c, d). Terminology as in Figure 4.

DETAILED DESCRIPTION

Proteins comprising the novel antigen binding domains may have any suitable format including antibodies, scFv’s, Fab’s, immunotoxins, immunoconjugates, bispecific antibodies, CARs etc. Such proteins, especially antibodies, may be used to target osteosarcoma cells as such, and they may carry a toxic payload in the form of radioactive isotopes like ¹⁷⁷ Lu, ²²⁴ Ra or ²²⁵ Ac .

In some embodiments proteins comprising the novel antigen binding domains described herein have cytotoxic effect on osteosarcoma cells.

As used herein, an antigen binding domain is a protein moiety able to bind an extracellular target epitope under physiological conditions, in particular physiological conditions in a tumor environment. The antigen binding domain comprises an antibody light chain variable domain (VL) and an antibody heavy chain variable domain (VH). Such variable domains are well-known for skilled persons.

The VH and VL may be connected by a disulphide bridges or a peptide linker. Alternatively, the two chains may be embedded in a Fab-fragment of an antibody or an antibody as such. In one embodiment, the antigen binding domain comprises or consists of VL-linker-VH. In another embodiment, the antigen binding domain comprises or consists of VH-linker-VL. Such antigen binding domains are often referred to as single chain Fv-fragments (scFv’s). The linker has to have a certain length in order to allow the VH and VL to form a functional antigen binding domain. In one embodiment, the linker comprises 10 to 30 amino acid residues. In one embodiment, the linker comprises 15 to 25 glycine and/or serine residues.

Each VL and VH herein comprises three complementarity determining regions (CDRs) flanked by human framework sequences. Human framework sequences are structurally conserved regions that normally tend to form a b-sheet structure delicately positioning the CDRs for specific binding to the target antigen under physiological conditions. Many human framework sequences are available from known human antibodies and from the the international ImMunoGeneTics information system (IMGT) online database (see Giudicelli et al, Nucleic Acids Research, 2006, Vol. 34, Database issue D781-D784), but the term also covers human framework sequences comprising amino acid substitutions. Each of the human framework sequences may optionally comprise 0 to 15 amino acid substitutions. An amino acid substitution is a sequence wherein an amino acid residue in a specific position is substituted for a different amino acid residue at the corresponding position, apparent when the sequences are aligned. Each of the human framework sequences may optionally comprise 1 amino acid substitution. Each of the human framework sequences may optionally comprise 2 amino acid substitutions. Each of the human framework sequences may optionally comprise 3 amino acid substitutions. Each of the human framework sequences may optionally comprise 4 amino acid substitutions. Each of the human framework sequences may optionally comprise 5 amino acid substitutions. The substitutions may be conservative substitutions. Even if such framework sequences are not necessarily previously known from human antibodies, they may provide lower immunogenic risk compared to a murine framework sequence.

In one embodiment, 0 to 5 amino acid residues in the human framework sequences are substituted with the corresponding amino acid residue(s) from the murine parent sequences (SEQ ID NO: 113 and SEQ ID NO: 114). In one embodiment, human framework sequences are optionally substituted with the corresponding amino acid residue from the murine parent sequences (SEQ ID NO: 113 and SEQ ID NO: 114) at vernier positions.

In one embodiment, 0 to 5 amino acid residues in the human framework sequences are substituted with the corresponding amino acid residue(s) from the murine parent sequences (SEQ ID NO: 115 and SEQ ID NO: 116). In one embodiment, human framework sequences are optionally substituted with the corresponding amino acid residue from the murine parent sequences (SEQ ID NO: 115 and SEQ ID NO: 116) at vernier positions.

According to Safdari et al .Biotechnology and Genetic Engineering Reviews Volume 29, 2013 - Issue 2:

The vernier positions in VH Frameworkl are residues 2, 27, 28, 29 and 30.

The vernier positions in VH Framework2 are residues 47,48 and 49.

The vernier positions in VH Framework3 are residues 67, 69, 71, 73, 78, 93 and 94. The vernier position in VH Framework4 is residue 103.

The vernier positions in VL Frameworkl are residues 2 and 4.

The vernier positions in VL Framework2 are residues 35, 36, 46, 47, 48 and 49.

The vernier position in VL Framework3 is residue 71.

The vernier position in VL Framework4 is residue 98.

Collectively, scFv’s comprising CDRs from a murine antibody and human framework sequences which each may optionally comprise 0 to 5 substitutions, are referred to as humanized scFv’s. In some embodiments, substitutions may be back to the parent murine amino acid residue (also known as “back mutations”).

Accordingly, each of the framework sequenses in the VLs herein may, independently, contain 0 to 5 substitutions at the same number of vernier positions.

Accordingly, each of the framework sequenses in the VHs herein may, independently, contain 0 to 5 substitutions at the same number of vernier positions.

Accordingly, each of the framework sequenses in the scFv’ herein may, independently, contain 0 to 5 substitutions at the same number of vernier positions.

In one embodiment, the human framework sequences are mature human framework sequences available from known human antibodies. Without being bound by theory, such framework sequences may convey very low risk of triggering unwanted immunogenic responses against the antigen binding domain, and at the same time increase the the likelihood of obtaining stable binding units which are expressed well in cellular systems.

The Frameworkl sequence is N-terminal to the CDR1, Framwork2 sequence is located between CDR1 and CDR2, while Framework3 sequence is located between CDR2 and CDR3. Accordingly, both a VL and VH can be roughly visualized as follows, with the CDRs boxed and the N-terminal indicated as N-:

N- FRAMEWORK1 CDR1 FRAMEWORK2 CDR2 FRAMEWORK3 CDR3 FRAMEWORK4

We provide eight examples of VL TP1 sequences comprising human framework sequences:

VL’s TP1:

VL1A (SEQ ID NO: 13) DIVMTQSPSFLSASVGDRVTITCKASQGVGSAIAWYQQKPGQSPKLLIYWASIRHTGVPD RFTGS GSGTDFTLTISSLQPEDFATYFCQQYSNYPLTFGGGTKVEIKR

VL1B (SEQ ID NO: 14)

DIVMTQSPSSLSASVGDRVT ITCKASQGVGSAIAWYQQKPGQSPKLLIYWAS IRHTG VPDRFTGSGSGTDFTLT I SS LQPEDFATYFCQQYSNYPLT FGGGTKVE IKR

VL1C (SEQ ID NO: 15)

DIVMTQSPSSVSASVGDRVT ITCKASQGVGSAIAWYQQKPGQSPKLLIYWAS IRHTG VPDRFTGSGSGTDFTLT I SS LQPEDFATYFCQQYSNYPLT FGGGTKLTVKR

VL1D (SEQ ID NO: 16)

E IVLTQS PS SVSASVGDRVT I TCKASIQGVGSAIIAWYQQKPGKAPKLL IYIWAS IRHTG

VPSRFSGSGSGTDFTLT I SSLQPEDFATYYCQQYSNYPLT FGGGTKVE IK VL1E (SEQ ID NO: 17)

D IVMTQS PS TLSASVGDRVT I TCKAS|QGVGSAIIAWYQQKPGKAPKLL IY|WAS|IRHTG

VPSRFSGSGSGTEFTLT I SGLQPEDFATYYCQQYSNYPLT FGGGTKLE IK

VL1F (SEQ ID NO: 18) D I QLTQSHRFLSASVGDRVT I TCRAS|QGVGSA|IAWYQQKPGQS PKLL IY|WAS|IRHTG

VPSRFSGSGSGTDFTLT I SSLQPEDVATY FCQQYSNYPLT FGGGTKVE IK

VL1G (SEQ ID NO: 19)

DIQMTQSPSSLSASVGDRIT ITCKAS|QGVGSA|IAWYQQKPGRVPKLLIY|WAS|IRHTG VPSRFSGSGSGTEFTLT I SSLQAEDVATYYC|QQYSNYPLT|FGGGTKVE IK

VL1H (SEQ ID NO: 20)

AIQLTQSHKFLSASVGDRVTITCRASlQGVGSAllAWYQQKPGQSPKLLIYlWASll RHSG

VPSRFSGSGSGTDFTLT I SSLQSEDFADYFCQQYSNYPLT FGGGTKLE IK VL1A comprises substitutions in 12 positions relative to the native human framework sequence as depicted in Figure 2B .

VL1B comprises substitutions in 13 positions relative to the native human framework sequence as depicted in Figure 2B.

VL1C comprises substitutions in 14 positions relative to the native human framework sequence as depicted in Figure 2B. VL1F comprises substitutions in position 8, 9, 10, 24, 42, 43 and 87 (underlined above).

VL1H comprises substitutions in position 8, 9, 10, 24, 42, 43, 56, 80, 85 and 87 (underlined above). We provide eight examples of TP1 VH sequences comprising human framework sequences:

VH’s TP1:

VH1A (SEQ ID NO: 21 EVQLQESGPGLVKPSETLSLTCSVTGDS I TSGYWNWIRQPPGNKLEYMGYI S YSDTT YYNPSLKSRI T I TRDTSKNQYYLKLS SVTAADTATYYCASAYYGSSLSMGNWGQGTT VTVSS

VH1B (SEQ ID NO: 22) EVQLQESGPGLVKPSQTLSLTCSVTGDS I TSGYWNWIRQHPGNKLEYMGYI S YSDTT YYNPSLKSRI T I TRDTSKNQYSLKLS SVTAADTATYY CASAYYGSSLSMGNWGQGTT VTVSS

VH1C (SEQ ID NO: 23) EVQLQESGPGLVKPSQTLSLTCSVTGDS I TSGYWNWIRQHPGNKLEYMGYI S YSDTT YYNPSLKSRI T I TRDTSKNQYSLKLS SVTAADTATYY CASAYYGSSLSMGNWGQGTL VTVSS

VH1D (SEQ ID NO: 24)

QLQLQE SGPGLVKPSQTLSLTCSVSIGDS I TSGYIWNW IRQHPGKGLEWI GYII SYSDT T

YYNPSLKSRI T I SLDTSKNQFSLKLTSVTAADTALYYC|ASAYYGSSLSMGN|WGRGTL VTVSS VH1E (SEQ ID NO: 25)

QVQLQE SGPGLVKPSQTLSLTCTVSIGDS I TSGYIWNW IRQPPGKGLEWI GYII SYSDT T

YYNPSLKSRI S I SLDTSKNQFSLRLS SVTAADTAVYYC[ASAYYGSSLSMGN|WGQGTT VTVSS

VH1F (SEQ ID NO: 26)

EVQLVE SGPGLVKPSQTLSLTCTVSIGDS I TSGYIWNW IRQHPGKGLEWI GYII SYSDT T

YYNPSLKSRVT I SVDTSKNQFSLKLS SVTAADTAVYYC[ASAYYGSSLSMGN|WGQGTL VTVSS IRRFPGNKLEYI GY|I SYSDT T

YYNPSLKSRVT I SRDTSKNQFSLKLS SVTAEDTAS YYC[ASAYYGSSLSMGN|WGQGTL VTVSS

VH1H (SEQ ID NO: 28)

DVQLQESGPGLVKPSQTLSLTCTVS GDS ITSGYWNWIRKFPGNKLEYIGYIS YSDTT

YYNPSLKSRVT I SVDTSKNQFSLKLKSVTAEDTAS YYC[ASAYYGSSLSMGN|WGQGTS ATVSS

VH1A comprises substitutions in 11 positions relative to the native human framework sequence as depicted in Figure 2A. VH1B comprises substitutions in substitutions in 11 positions relative to the native human framework sequence as depicted in Figure 2A.

VH1C comprises substitutions in substitutions in 11 positions relative to the native human framework sequence as depicted in Figure 2A.

VH1G comprises substitutions in position 40, 43, 44, 47, 71, 84 and 89 (underlined above).

VH1H comprises substitutions in position 1, 39, 40, 43, 44, 47, 82, 84 and 89 (underlined above).

We provide eight examples of TP3 VL sequences comprising human framework sequences: VL’s TP3:

VL3A (SEQ ID NO: 29)

DIVLTQSPDSLAVSLGERAT ISCRASKSVSTGYS YLHWYQQKPGQPPKLLIYLASNL ESGVPARFSGSGSGTDFTLT I SSLQAEDVAVYYCQHSRELPLTFGAGTKLE IKR

VL3B (SEQ ID NO: 30)

D I VLTQSPATLSLSPGERATLSCRASKS VS TGYS YLHWYQQKPGQPPKLLIYLASNL ESGVPARFSGSGSGTDFTLT I SSLEPEDFAVYYCQHSRELPLTFGAGTKLE IKR

VL3C (SEQ ID NO: 31)

DIVLTQSPATLSVSPGERATLSCRASKSVS TGYS YLHWYQQKPGQPPKLLIYLASNL ESGVPARFSGSGSGTDFTLT I SSLQSEDFAVYYCQHSRELPLTFGAGTKVE IKR VL3D (SEQ ID NO: 32)

E IVLTQS PS S LSAYVGDRVT I TCRAS|KSVS TGYSY|LHWYQQKPGKAPKLL IY|LAS[NL

ESGVPSRFSGSGSGTDFTLT INSLQPEDFATYYC QHSRELPLT FGGGTKLE IK

VL3E (SEQ ID NO: 33)

DIVLTQSPSSLSASVGDRVT ITCRASlKSVSTGYS YILHWYQQKPGKAPKLLIYILASINL

ESGVPSRFSGSGSGTDFTLT I SSLQPEDFATYYC QHSRELPLT FGQGTKLE IK

VL3F (SEQ ID NO: 34)

E IVLTQSPATLSLSPGERATLSCRAS|KSVSTGYS Y|LHWYQQKPGQAPRLL I Y|LASpSfL ESGVPARFSGSGSGTDFTLT I SSLEPEDFAVYYClQHSRELPLTlFGGGTKVE IK

VL3G (SEQ ID NO: 35)

DIQMTQSPSSLSASVGDRVT ISCRASlKSVSTGYS YILHWYQQKPGQAPKLLIYILASINL

ESGVPSRFSGSGSGTDFTLT I SSLQPEDVATYYC QHSRELPLT FGQGTKVE IK

VL3H (SEQ ID NO: 36)

E ivLTQSPATLSLS I GQRATLSCRASIKSVS TGYS YILHWYQQKPGQPPRLL I YILASINL

E SGI PARFSGSGPGTDFTLT I RPLEDEDFAVYYC QHSRELPLT FGAGTKLELK

VL3A comprises substitutions in 11 positions relative to the native human framework sequence as depicted in Figure 2D.

VL3B comprises substitutions in 14 positions relative to the native human framework sequence as depicted in Figure 2D. VL3C comprises substitutions in 15 positions relative to the native human framework sequence as depicted in Figure 2D.

VL3H comprises substitutions in position 15, 17, 43, 76, 77, 80 (underlined above).

VH’s TP3:

We provide eight examples of VH TP3 sequences comprising human framework sequences:

VH3A (SEQ ID NO: 37) EVQLQQSGAEVKKPGSSVKVSCKAS GYT FTDYNMDWVRQAPGKSLEWIGDINPNYDS TRYNQKFQGRATLTVDKSSSTAYMELSSLRSEDTAVYYCARGDYYVSSYGHDYAMDY WGQGTTVTVSS

VH3B (SEQ ID NO: 38) EVQLQQSGAEVKKPGASVKVSCKAS GYT FTDYNMDWVRQAPGKSLEWIGDINPNYDS TRYNQKFQGRATLTVDKSSS TAYMELSSLRSEDTAVYYCARGDYYVSS YGHDYAMDY WGQGTTVTVSS

VH3C (SEQ ID NO: 39) EVQLQQSGAEVKKPGSSVKVSCKAS GYT FTDYNMDWVRQAPGKSLEWIGDINPNYDS

TRYNQKFQGRATLTVDKSSS TAYMELSSLRSEDTAVYYCARGDYYVSS YGHDYAMDY WGQGTLVTVSS

VH3D (SEQ ID NO: 40) QVQLVQSGAEVTKPGASVKVSCKAS|GYTFTDYN|MDWVRQAPGQGLEWVGD[INPNYDS

[T1RYNQKFKGRVTLTRDTSTTTVYMELSSLRSEDTAVYYC|ARGDYYVSSYGHDYA MDY WGQGTTVTVSS

VH3E (SEQ ID NO: 41) QVQLVESGAEVKKPGASVKLSCKAS|GYTFTDYN[MDWVRQAPGQRLEWMGDINPNYDS

|T]RYNQKFKGRVTITVDKSASTAYMELSSLRSEDTAVYYC|ARGDYYVSSYGHDYA MDY

WGPGTTVTVS S

VH3F (SEQ ID NO: 42) QVQLVQSGAEVKKPGASVKVSCKAS|GYTFTDYN|MDWVRQAPGQGLEWIGD[INPNYDS

[TIRYNQKFKGWVTVTRDTSISTAYMELSRLTSDDTAVYYC|ARGDYYVSSYGHDYA MDY WGQGTMVTVSS VH3G (SEQ ID NO: 43)

QVQLVQSGAEVKKPGASVKVSCKAS|GYT FTDYN[MDWVRQAPGQGLEWMGD[I ^~ NPNYDS |T1RYNQKFKGRVTS TRDT S I S TAYMELSRLRSDDTVVYYC[ARGDYYVS SYGHDYAMDY

WGQGTSVTVS S

VH3H (SEQ ID NO: 44)

EVQLVQSGAELVKPGASVKVSCKAS|GYT FTDYN[MDWVRQSHGRGLEWMGD|INPNYDS

|T]R YNQK FKGRVTS T VDK S SYGHDYAMDY

WGQGTSVTVS S

VH3A comprises substitutions in 12 positions relative to the native human framework sequence as depicted in Figure 2C.

VH3B comprises substitutions in substitutions in 11 positions relative to the native human framework sequence as depicted in Figure 2C.

VH3C comprises substitutions in substitutions in 12 positions relative to the native human framework sequence as depicted in Figure 2C. VH3H comprises substitutions in position 1, 11, 12, 40, 41, 43, 71, 73 and 83 (underlined above).

In antigen binding units, the human framework sequences may tolerate some variation without destroying the specificity and affinity to the target antigen. For example, substitutions of amino acid residues may be tolerated better than deletions or additions of amino acid residues.

However, the CDRs are generally more sensitive for variations, but occationally conservative substitutions may be introduced without destroying the specificity and affinity.

The term "conservative amino acid substitution", as used herein, refers to an amino acid substitution in which one amino acid residue is replaced with another amino acid residue having a similar side chain. Amino acids with similar side chains tend to have similar properties, and thus a conservative substitution of an amino acid important for the structure or function of a polypeptide may be expected to affect polypeptide structure/function less than a non-conservative amino acid substitution at the same position. Families of amino acid residues having similar side chains have been defined in the art, including basic side chains (e.g. lysine, arginine, histidine), acidic side chains (e.g. aspartic acid, glutamic acid), uncharged polar side chains (e.g. asparagine, glutamine, serine, threonine, tyrosine), non-polar side chains (e.g. glycine, cysteine, alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan) and aromatic side chains (e.g. tyrosine, phenylalanine, tryptophan, histidine). Thus, a conservative amino acid substitution may be considered to be a substitution in which a particular amino acid residue is substituted for a different amino acid residue in the same family.

Examples of scFv’s provided herein:

In one particular embodiment, the scFv may be VL1 E-linker- VH1G, VHlG-linker- VL1E, VL 1 D-linker- VH 1 G or VHlG-linker-VLID.

Further examples of scFv’s provided herein:

In one particular embodiment, the scFv may be VL3H-linker-VH3D, VH3D-linker- VL3H, VL3H-linker-VH3F, VH3F-linker-VL3H, VL3H-linker-VH3H or VH3H- linker-VL3H.

Novel chimeric antigen receptors (CARs) are provided. When the CARs herein are expressed on the surface of immune cells, such immune cells may be used in medicine. In particular, said immune cells may be used in treatment of osteosarcoma. In one embodiment, said immune cells be used in treatment of metastatic osteosarcoma.

As used herein, Chimeric Antigen Receptors (CARs) are artificial receptors comprising an extracellular antigen binding domain, a transmembrane domain and an intracellular signaling domain. The antigen binding domain in CARs is usually a scFv.

CARs in the present disclosure may comprise any of the antigen binding domains as mentioned above. For example, CARs in the present disclosure may comprise any of the scFv’s represented by SEQ ID NO: 45 to SEQ ID NO: 78. In particular, the

CARs in the present disclosure may comprise any of the antigen binding domains as mentioned above in the form of a scFv which may be connected to a CD8a hinge or a CD28 hinge. In particular, the CARs in the present disclosure may comprise any of the scFv’s represented by SEQ ID NO: 45 to SEQ ID NO: 78 connected to a CD8a hinge. In particular, the CARs in the present disclosure may comprise any of the scFv’s represented by SEQ ID NO: 45 to SEQ ID NO: 78 connected to a CD8a hinge, wherein the CAR further comprises a transmembrane domain and wherein the intracellular signaling domain comprises a costimulatory domain and a CD3z signaling domain.

Alternatively, CARs in the present disclosure may comprise any of the antigen binding domains as mentioned above. For example, CARs in the present disclosure may comprise any of the scFv’s represented by SEQ ID NO: 88 to SEQ ID NO:

112. In particular, the CARs in the present disclosure may comprise any of the antigen binding domains as mentioned above in the form of a scFv which may be connected to a CD8a hinge. In particular, the CARs in the present disclosure may comprise any of the scFv’s represented by SEQ ID NO: 88 to SEQ ID NO: 112 connected to a CD8a hinge. In particular, the CARs in the present disclosure may comprise any of the scFv’s represented by SEQ ID NO: 88 to SEQ ID NO: 112 connected to a CD8a hinge, wherein the CAR further comprises a transmembrane domain and wherein the intracellular signaling domain comprises a costimulatory domain and a Oϋ3z signaling domain.

Several hinges derived from human CD8a are known and they may be used in CARs according to the present invention. In general, such hinges contain 10 to 60 amino acid residues from the human CD8a-sequence which in full-length CD8a encodes an extracellular domain connected to the transmembrane domain. In general, such hinges can contain 35 to 55 amino acid residues from the human CD8a-sequence. In general, such hinges can contain 45 to 55 amino acid residues from the human CD8a-sequence. Suitable examples include:

FVPVFLPAKPTTTPAPRPPTPAPTIASQPLSLRPEACRPAAGGAVHTRGLDFAC

D (SEQ ID NO: 118) and

TTTPAPRPPTPAPTIASQPFSFRPEACRPAAGGAVHTRGFDFACDISS (SEQ ID NO: 119)

However, the CD8a hinges may comprise 0 to 5 conservative substitutions relative to the human sequence.

The antigen binding domain may be directly attached to a transmembrane domain. However, the CARs may comprise a hinge domain connecting the antigen binding domain to the transmembrane domain. The hinge domain may thus affect the sterical conformation of the antigen binding domain. This may in turn affect the ability of the CAR to bind the target epitope and subsequently trigger signaling into an immune cell. If the target epitope is located too far from the cell membrane of the target cell or if the target epitope is otherwise hidden, the immune cell expressing the CAR may not be efficient. Accordingly, it is preferred that the target epitope is sufficiently accessible for immune cells expressing the CARs.

Four bone-specific isoforms of human tissue-nonspecific alkaline phosphatase (B/I, Blx, Bl, and B2) exist according to Haarhaus et al, NATURE REVIEWS | NEPHROLOGY VOLUME 13 | JULY 2017. Without being bound by theory, the target antigen of the novel binding units herein may be a bone- specific isoform of human tissue-nonspecific alkaline phosphatase (ALPL), in particular an accessible extracellular epitope thereof. Accordingly, the present disclosure provides CARs comprising a human or humanized scFv targeting a bone- specific isoform of human tissue-nonspecific alkaline phosphatase (e.g. the 524 amino acid residue isoform NM_000478). Cytotoxic immune cells expressing said CARs may be used for treatment of OS with lower risk for triggering unwanted immunogenic responses against the therapeutic cells. In effect, there is also provided a method for treatment of OS comprising a step of administering a pharmaceutical composition intravenously or intra-tumorally to a patient diagnosed with OS, wherein the pharmaceutical composition contains cytotoxic cells expressing a CAR with a human or humanized scFv able to specifically bind to a bone-specific isoform of human tissue-nonspecific alkaline phosphatase. There is also provided CARs comprising a human or humanized scFv able to specifically bind to a bone-specific isoform of human tissue-nonspecific alkaline phosphatase. CARs comprising a human or humanized scFv able to specifically bind to a bone -specific isoform B1 of human tissue-nonspecific alkaline phosphatase may be used for treatment of OS. CARs comprising a human or humanized scFv able to specifically bind to a bone- specific isoform B2 of human tissue-nonspecific alkaline phosphatase may be used for treatment of OS .

In one embodiment, there is provided a method of treatment of OS in a human patient comprising the steps: a) transducing T cells or NK cells with mRNA encoding any of the CARs herein b) repeatedly administering an effective dose of a pharmaceutical composition comprising said cells to an OS patient.

The transmembrane domain connects the extracellular domains to an intracellular signaling domain. Both the antigen binding domain and hinge domain are extracellular domains, i.e. that they generally face the extracellular environment when expressed in the cell membrane of an immune cell. As used herein, "transmembrane domain", means the part of the CAR which tend to be embedded in the cell membrane when expressed by an immune effector cell. Suitable transmembrane domains are well known for skilled persons. In particular, transmembrane domains from the human proteins CD8a, CD28 or ICOS may be used. The transmembrane domain is believed to convey a signal into immune cells upon binding of a target by the antigen binding domain.

The “intracellular signaling domain” refers to a part of the CAR located inside the immune cell when the CAR is expressed in the cell membrane. These domains participate in conveying the signal upon binding of the target. A variety of signaling domains are known, and they can be combined and tailored to fit the endogenous signaling machinery in the immune cells. In one embodiment the intracellular signaling domain comprises a "signal 1" domain like the signaling domains obtainable from the human proteins CD3z, FcR-g, CD3e etc. In general, it is believed that "signal F'domains (e.g. CD3z signaling domain) convey a signal upon antigen binding. In another embodiment, the intracellular signaling domain comprises a costimulatory domain. Such domains are well known and often referred to as "signal 2" domains, and they are believed to, subsequently of “signal 1” domains, convey a signal via costimulatory molecules. The "signal 2" is important for the maintenance of the signal and the survival of the cells. If absent, like in first- generation CARs, the CAR T cell may be efficient in killing and in early cytokines release, but it will often become exhausted over time. Examples of such commonly used human "signal 2" domains include 4- IBB signaling domain, CD28 signaling domain, 4- IBB signaling domain and ICOS signaling domain.

The immune cells expressing the CARs herein may be isolated from a patient or a compatible donor by leukapheresis or other suitable methods. Such primary cells may for example be T cells or NK cells. In particular, autologous T cells (both cytotoxic T cells, T helper cells or mixtures of these) may be transduced with nucleic acids encoding the CARs before a pharmaceutical composition comprising the cells is administered back to the patient. The immune cells expressing the CARs may also be cell lines suitable for clinical use like NK-92 cells. Of course, the preferred cells are human when the intended patient is human. However, as OS also affects dogs, canine cells expressing the CARs herein may be used in treatment of OS in dogs.

The pharmaceutical compositions herein can be a composition suitable for administration of therapeutic cells to a patient. The most common administration route for CAR T cells is intravenous administration. Accordingly, said pharmaceutical compositions may for example be sterile aqueous solutions with a neutral pH. Accordingly, said pharmaceutical compositions may for example be sterile aqueous solutions with a physiological pH. For example, a patient’s peripheral blood mononuclear cells may be obtained via a standard leukapheresis procedure. The mononuclear cells may be enriched for T cells, before transducing them with a lentiviral vector or mRNA encoding the CARs. Said cells may then be activated with anti-CD3/CD28 antibody coated beads. The transduced T cells may be expanded in cell culture, washed, and formulated into a sterile suspension, which can be cryopreserved. If so, the product is thawed prior to administration.

One issue related to efficient access of CAR T cells at the target site of osteosarcoma, is to circumvent defenses from solid tumors. In contrast to hematological malignancies, recognition of solid tumors requires egress from the blood into the tumor site, and many malignancies evolve such that T cell infiltration is actively impeded. In situations where the tumor is localized, different administration modes may be used to improve efficacy. For example, regional rather than systemic administration of CAR T cells might enhance efficacy. The pharmaceutical compositions may comprise a pharmaceutically effective dose of the immune cells herein. A pharmaceutically effective dose may for example be in the range of 1 x 10 ⁶ to 1 x 10 ¹⁰ immune cells expressing the CARs. A pharmaceutically effective dose may for example be in the range of 1 x 10 ⁷ to 1 x 10 ⁹ T cells expressing the CARs.

For efficient expression of the claimed CARs in immune cells, a conventional leader peptide may be introduced N-terminally for facilitating location in the cell membrane. One example of a leader peptide is METDTLLLWVLLLWVPGSTG (SEQ ID NO: 117). The leader peptide is believed to be trimmed off and will likely not be present in the functional CAR in the cell membrane.

Accordingly, for expression of a second-generation CAR, nucleic acids encoding the following may be used:

N-LEADER PEPTIDE-VH-LINKER-VL-HINGE-TRANSMEMBRANE DOMAIN- COSTIMULATORY DOMAIN-SIGNALING DOMAIN.

Accordingly, for expression of a second-generation CAR, nucleic acids encoding the following may also be used:

N-LEADER PEPTIDE-VL-LINKER-VH-HINGE-TRANSMEMBRANE DOMAIN- COSTIMULATORY DOMAIN-SIGNALING DOMAIN

The nucleic acids encoding the claimed CARs can be in the form of well-known RNA e.g. mRNA, or DNA expression vectors.

The murine parent VL TP1 (SEQ ID NO: 113) with CDRs boxed: DI VMTQSHKFMSTSVGDRVS ITCKAS QGVGSA|IAWYQQKTGQ S PKLL I Y|WAS1I RHTGVPDRF

TGSGSGTDFTLT ISNVQSEDLADYFC QQY SNYPL T FGGGTKLE IKR

The murine parent VH TP1 (SEQ ID NO: 114) with CDRs boxed:

EVQLQE SGP SLVKP SQTL SLTCSVTII GDS ITSGY ' WNWIRKFPGNKLEYMGY ISYSDTT YYNPS

LKSRIS ITRDTSKNQYYLHLKSVTTEDTATYYC ASAYYGSSLSMGN WGQGTSATVSSD

The murine parent VL TP3 (SEQ ID NO: 115) with CDRs boxed:

DIVLTQS PASLAVSLGQRATISCRAS KSVSTGYSYLHWYQQKPGQPPKLL IYLASNL

ESGVPARFSGSGSGTDFTLNIHPVEEEDAATYYC QHSRELPLT FGAGTKLELK The murine parent VH TP3 (SEQ ID NO: 116) with CDRs boxed:

EVQLQQSGAELVKPGASVKI SCKAS|GYT FTDYN[MDWVKQSHGKSLEWIGD[INPNYDS

|T1RYNQKFKGKATLTVDKS S S TAYMELRSLT SEDTAVYYC[ARGDYYVS SYGHDYAMDY

WGQGTSVTVS S

It has previously been shown in PCT/EP2019/086309 that CARs comprising murine VL TP1 (SEQ ID NO: 113) and murine VH TP1 (SEQ ID NO: 114) provides therapeutic effect and prolonged survival in mice with very agressive OSA lung tumors. It was further shown in PCT/EP2019/086309 that the CARs have cytotoxic effect on tumor cells (OHS cell line).

The murine VL TP1 (SEQ ID NO: 113) and murine VH TP1 (SEQ ID NO: 114) comprises the same 6 CDR’s as provided by the present invention and as represented by SEQ ID NO: 1 to 6 as shown below.

In certain embodiments it is thus expected that a protein for targeting of osteosarcoma cells, comprising a VL and a VH which together form an antigen binding unit according to the present invention, wherein the VL comprises three complementarity determining regions (CDRs); CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 1, 2 and 3 and wherein the VH comprises three CDRs; CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 4, 5 and 6, and wherein all the CDRs are flanked by human framework sequences, and optionally wherein each of the human framework sequences, independently, comprises 0 to 5 amino acid substitutions have a similar effect as described in PCT/EP2019/086309 such as cytotoxic effect on osteosarcoma cells.

In certain embodiments it is further expected that a CAR construct comprising the above antigen binding unit and according to the present invention have a similar effect as described in PCT/EP2019/086309 such as cytotoxic effect on osteosarcoma cells.

It has also previously been shown in PCT/EP2019/086309 that CARs comprising murine VL TP3 (SEQ ID NO: 115) and murine VH TP3 (SEQ ID NO: 116) provides therapeutic effect and prolonged survival in mice with very agressive OSA lung tumors. It was further shown in PCT/EP2019/086309 that the CARs have cytotoxic effect on tumor cells (OHS cell line). The murine VL TP3 (SEQ ID NO: 115) and murine VH TP3 (SEQ ID NO: 116) comprises the same 6 CDR’s as provided by the present invention and as represented by SEQ ID NO: 7 to 12 as shown below. In certain embodiments it is thus expected that a protein for targeting of osteosarcoma cells, comprising a VL and a VH which together form an antigen binding unit according to the present invention, wherein the VL comprises three complementarity determining regions (CDRs); CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 7, 8 and 9 and wherein the VH comprises three CDRs; CDR1, CDR2 and CDR3 which respectively are represented by the amino acid sequences SEQ ID NOs: 10, 11 and 12, and wherein all the CDRs are flanked by human framework sequences, and optionally wherein each of the human framework sequences, independently, comprises 0 to 5 amino acid substitutions have a similar effect as described in PCT/EP2019/086309 such as cytotoxic effect on osteosarcoma cells.

In certain embodiments it is further expected that a CAR construct comprising the above antigen binding unit and according to the present invention have a similar effect as described in PCT/EP2019/086309 such as cytotoxic effect on osteosarcoma cells. SEQUENCES

TP1 VL CDRs

SEQ ID NO: 1 QGVGSA SEQ ID NO: 2 WAS SEQ ID NO: 3 QQYSNYPLT

TP1 VH CDRs

SEQ ID NO: 4 GDSITSGY SEQ ID NO: 5 ISYSDTT SEQ ID NO: 6 ASAYYGSSLSMGN

TP3 VL CDRs

SEQ ID NO: 7 SKSVSTGYSY

SEQ ID NO: 8 LAS SEQ ID NO: 9 QHSRELPLT

TP3 VH CDRs

SEQ ID NO: 10 GYTFTDYN SEQ ID NO: 11 INPNYDST SEQ ID NO: 12 ARGDYYVSSYGHDYAMDY

VL’s TP1

SEQ ID NO: 13 (VL1A) SEQ ID NO: 14 (VL1B)

SEQ ID NO: 15 (VL1C)

SEQ ID NO: 16 (VL1D)

SEQ ID NO: 17 (VL1E) SEQ ID NO: 18 (VL1F)

SEQ ID NO: 19 (VL1G)

SEQ ID NO: 20 (VL1H)

VH’s TP1 SEQ ID NO: 21 (VH1A)

SEQ ID NO: 22 (VH1B)

SEQ ID NO: 23 (VH1C)

SEQ ID NO: 24 (VH1D)

SEQ ID NO: 25 (VH1E) SEQ ID NO: 26 (VH1F)

SEQ ID NO: 27 (VH1G)

SEQ ID NO: 28 (VH1H)

VL’s TP3 SEQ ID NO: 29 (VL3A)

SEQ ID NO: 30 (VL3B)

SEQ ID NO: 31 (VL3C)

SEQ ID NO: 32 (VL3D)

SEQ ID NO: 33 (VL3E) SEQ ID NO: 34 (VL3F)

SEQ ID NO: 35 (VL3G)

SEQ ID NO: 36 (VL3H)

VH’s TP3 SEQ ID NO: 37 (VH3A)

SEQ ID NO: 38 (VH3B)

SEQ ID NO: 39 (VH3C)

SEQ ID NO: 40 (VH3D)

SEQ ID NO: 41 (VH3E) SEQ ID NO: 42 (VH3F)

SEQ ID NO: 43 (VH3G)

SEQ ID NO: 44 (VH3H) scFv’s TP1: SEQ ID NO: 45 VLlA-linker-VHIA SEQ ID NO: 46 VLlA-linker-VHIB SEQ ID NO: 47 VLlA-linker-VHIC SEQ ID NO: 48 VL 1 B -linker- VH 1 A SEQ ID NO: 49 VL IB -linker- VH IB SEQ ID NO: 50 VL IB -linker- VH 1C SEQ ID NO: 51 VL 1 C-linker- VH 1 A SEQ ID NO: 52 VL1 C-linker- VH IB SEQ ID NO: 53 VL1 C-linker- VH 1C SEQ ID NO: 54 VL 1 D -linker- VH 1 D SEQ ID NO: 55 VL1 E-linker- VH ID SEQ ID NO: 56 VL 1 F-linker- VH 1 D SEQ ID NO: 57 VL 1 G-linker- VH 1 D SEQ ID NO: 58 VL 1 H-linker- VH 1 D SEQ ID NO: 59 VL 1 D -linker- VH 1 E SEQ ID NO: 60 VL1 E-linker- VH IE SEQ ID NO: 61 VL1 F-linker- VH IE SEQ ID NO: 62 VL1 G-linker- VH IE SEQ ID NO: 63 VL1 H-linker- VH IE SEQ ID NO: 64 VL 1 D -linker- VH 1 F SEQ ID NO: 65 VL1 E-linker- VH IF SEQ ID NO: 66 VL1 F-linker- VH IF SEQ ID NO: 67 VL1 G-linker- VH IF SEQ ID NO: 68 VL1 H-linker- VH IF SEQ ID NO: 69 VL 1 D -linker- VH 1 G SEQ ID NO: 70 VL1 E-linker- VH1G SEQ ID NO: 71 VL1 F-linker- VH1G SEQ ID NO: 72 VL1 G-linker- VH1G SEQ ID NO: 73 VL1 H-linker- VH1G SEQ ID NO: 74 VL 1 D -linker- VH 1 H SEQ ID NO: 75 VL1 E-linker- VH1H SEQ ID NO: 76 VL1 F-linker- VH1H SEQ ID NO: 77 VL1 G-linker- VH1H SEQ ID NO: 78 VL1 H-linker- VH1H scFv’s TP3: SEQ ID NO: 79 VL3A-linker-VH3A SEQ ID NO: 80 VL3A-linker-VH3B SEQ ID NO: 81 VL3A-linker-VH3C SEQ ID NO: 82 VL3B-linker-VH3A SEQ ID NO: 83 VL3B-linker-VH3B SEQ ID NO: 84 VL3B-linker-VH3C SEQ ID NO: 85 VL3C-linker-VH3A SEQ ID NO: 86 VL3C-linker-VH3B SEQ ID NO: 87 VL3C-linker-VH3C scFv’s:

SEQ ID NO: 88 VL3D-linker-VH3D SEQ ID NO: 89 VL3E-linker-VH3D SEQ ID NO: 90 VL3F-linker-VH3D SEQ ID NO: 91 VL3G-linker-VH3D SEQ ID NO: 92 VL3H-Iinker-VH3D SEQ ID NO: 93 VL3D-linker-VH3E SEQ ID NO: 94 VL3E-linker-VH3E SEQ ID NO: 95 VL3F-linker-VH3E SEQ ID NO: 96 VL3G-linker-VH3E SEQ ID NO: 97 VL3H-linker-VH3E SEQ ID NO: 98 VL3D-linker-VH3F SEQ ID NO: 99 VL3E-linker-VH3F SEQ ID NO: 100 VL3F-linker-VH3F SEQ ID NO: 101 VL3G-linker-VH3F SEQ ID NO: 102 VL3H-linker-VH3F SEQ ID NO: 103 VL3D-linker-VH3G SEQ ID NO: 104 VL3E-linker-VH3G SEQ ID NO: 105 VL3F-linker-VH3G SEQ ID NO: 106 VL3G-linker-VH3G SEQ ID NO: 107 VL3H-linker-VH3G SEQ ID NO: 108 VL3D-linker-VH3H SEQ ID NO: 109 VL3E-linker-VH3H SEQ ID NO: 110 VL3F-linker-VH3H SEQ ID NO: 111 VL3G-linker-VH3H

SEQ ID NO: 112 VL3H-linker-VH3H

Parent:

SEQ ID NO: 113 murine parent VL TP1 SEQ ID NO: 114 murine parent VH TP1

SEQ ID NO: 115 murine parent VL TP3 SEQ ID NO: 116 murine parent VH TP3 Leader peptide

SEQ ID NO: 117 leader peptide

EXAMPLES Example 1

Novel second-generation murine CARs and humanized CARs can be cloned into pENTR vector (Invitrogen) and further subcloned by recombination into Gateway system compatible expression vectors, the retroviral construct pMP71 and the mRNA synthesis construct pCIpA102 following our previous publication (Walchli et al, PLoS One. 2011;6(1 l):e27930).

The T cell line Jurkat-76 consitutively expressing NFAT-GFP reporter gene (Jutz S, Leitner J, Schmetterer K, Doel-Perez I, Majdic O, Grabmeier-Pfistershammer K, Paster W, Huppa JB, Steinberger P. J Immunol Methods. 2016 Jan 15. pii: S0022- 1759(16)30007-2. doi: 10.1016/j.jim.2016.01.007. 10.1016/j.jim.2016.01.007 PubMed 26780292) were transduced by retroviral particles encoding the humanized CARs via well-known methods (Walchli et al, PLoS One. 2011 ;6( 1 l):e27930).

After 48 hours, the presence of the CAR can be analyzed by flow cytometry.

In vitro recognition of target the OSA cell line can be demonstrated by incubating J76-NFAT-GFP-CARs with OSA for 20 hours. If the TCR signalling pathway is stimulated, the NFAT promoter becomes activated and GFP is produced into the cytosol of J76-NFAT-GFP cells. These cells can be analyzed by flowcytometry and the level of GFP can be quantified, it will be proportional to the level of stimulation. This assay can be run with different cell lines, OS cell lines or healthy cells and is sensitive enough to detect if the CAR’s recognition is restricted to OS.

It also informs us about the tonic signal of certain constructs.

In vitro activation by target cell lines was demonstrated by incubating J76-NFAT expressing the humanized CARs with osteosarcoma cell line (OSA) for 24 hours as described in example 1, above. The results are depicted in Figure 4. Clone 1326 (OSCAR-1) and clone 1327 (OSCAR-3) are second generation CARs based on murine sequences. Clone 1668, 1669, 1670 and 1671 are humanized variants of TP1 / OSCAR 1, while clone 1672, 1673, 1674 and 1675 are humanized variants of TP3 / OSCAR 3. PMA/ionomycin is a positive control for normalization of expression data between samples. The results demonstrates that humanization of TP3 scFv does not affect the activity / binding activity of the CAR constructs.

Example 2

In vitro killing of target cells is performed using primary human T cells isolated from healthy donor. The T cells are genetically modified using retrovirus as mentioned above and further expanded. Target cells are also genetically modified with a transgene encoding for a luciferase-GFP fusion protein (Loew, R., Heinz, N., Hampf, M. et al. Improved Tet-responsive promoters with minimized background expression. BMC Biotechnol 10, 81 (2010).). Here, the luciferase will be constitutively produced and its activity detectable upon addition of its substrate, luciferine. As long as the cells are alive they will produce luciferase that will emit detectable light signals, upon cell death, the luciferin will not be metabolized due to the lack of ATP and the signal will be lost. The quantification of the signal will be used to quantify cell death. CAR T cells are co-cultured with target cells positive or negative for the antigen of interest at different Effector-to-Target (E:T) ratio and the luciferase signal will be monitored at different time points and plotted. As shown in Figure 5, cells expressing OSCAR- 1 and OSCAR-3 strictly kill OS A target cells, suggesting that they are specific. In addition, cells expressing the humanized constructs (1669/1671 and 1672/1674) specifically kill OSA target cells with the similar efficiency, suggesting that the functional avidity of the humanized constructs is conserved.

The different clones as described above refer to second generation CARs varying with respect to the scFv only:

The structure of the tested CARs herein can be represented: scFv-CD8a hinge- CD8a transmembrane domain-4- IBB costimulatory OoΐMίh-Oϋ3z signaling domain.

In order to assess CAR efficacy, an animal model can be used in which human osteosarcoma cell lines are engrafted in mice. The engraftment can be performed by intra-peritoneal injection of 10 ⁶ OHS cells expressing the luciferase gene. When the tumor become palpable, the mice can be randomized and treated or not with T cells redirected or not with the novel CARs. Three injections of 10 x 10 ⁶ T cells can be administered and the tumor growth can be monitored by IVIS (luciferase signal detection).