Mesothelin and MUC16 bispecific chimeric antigen receptor (CAR) t cells

Abstract

The present disclosure relates to chimeric antigen receptors (CARs) that bind mesothelin and MUC16 (mucin 16) and methods of use thereof.

Description

[0001] MESOTHELIN AND MUC16 BISPECIFIC CHIMERIC ANTIGEN RECEPTOR (CAR) T CELLS

[0002] RELATED APPLICATIONS

[0003] This application claims benefit under 35 U.S.C. § 119(e) to U.S. Provisional Application No. 63 / 568,354, filed March 21, 2024, entitled “Mesothelin and MUC16 Bispecific Chimeric Antigen Receptor (CAR) T cells”, the entire contents of which is incorporated herein by reference.

[0004] REFERENCE TO AN ELECTRONIC SEQUENCE LISTING

[0005] The contents of the electronic sequence listing (M105370053WO00-SEQ-ARM.xml; Size: 40,826 bytes; and Date of Creation: March 18, 2025) are herein incorporated by reference in its entirety.

[0006] BACKGROUND

[0007] Cell-based gene therapies have shown tremendous promise in treating diseases like hematological cancers. For example, chimeric antigen receptor-T cells (CAR-T cells) are cellbased gene therapies for cancer that have had great success in the treatment of hematologic malignancies. Autologous CAR-T cells are made by collecting a patient’s T cells and genetically modifying them to express a chimeric antigen receptor (a CAR), which confers a novel specificity to the T cells: recognition of a tumor surface antigen through the CAR activates the T cell and initiates tumor killing and expansion of the CAR-T cells.

[0008] SUMMARY

[0009] Prior studies have shown that chimeric antigen receptor T cells (CAR-T cells) are effective in treating hematological malignancies, but treatment of solid tumors with CAR-T cell therapy has been more challenging. This disparity is likely due to a combination of several differences between solid tumors and blood cancers, from identification of target antigens to how CAR-T cells interact with and kill different kinds of tumor cells. While much is known about the expression of surface markers that could serve as CAR targets for blood cancers, solid tumors are not generally diagnosed or characterized by their expression of surface markers, and solid tumors tend to have more heterogeneity, which could lead to outgrowth of tumors with low or no target antigen expression, in particular when only one candidate antigen is targeted by a CAR in CAR-T therapy. Accordingly, in some aspects, this disclosure describes a bispecific chimeric antigen receptor (CAR; also referred to as a tandem CAR or tanCAR) comprising: (i) a first antigenbinding domain that binds to mesothelin; and (ii) a second antigen-binding domain that binds to mucin 16 (MUC16). In some embodiments, the first antigen-binding domain comprises a mesothelin-binding antibody.

[0010] In some embodiments, the mesothelin-binding antibody is a mesothelin-binding antibody fragment. In some embodiments, the mesothelin-binding antibody fragment is an antigenbinding fragment (Fab), a Fab', or a F(ab')2, a fragment variable (fv), or a single chain variable fragment (scFv).

[0011] In some embodiments, the mesothelin-binding antibody comprises: (a) a heavy chain variable domain (VH) comprising three complementarity determining regions CDR-H1, CDR- H2, and CDR-H3, wherein the CDR-H1 comprises an amino acid sequence of SEQ ID NO: 3; the CDR-H2 comprises an amino acid sequence of SEQ ID NO: 4; and the CDR-H3 comprises an amino acid sequence of SEQ ID NO: 5, and (b) a light chain variable domain (VL) comprising three complementarity determining regions CDR-L1, CDR-L2, and CDR-L3, wherein the CDR-L1 comprises an amino acid sequence of SEQ ID NO: 6; the CDR-L2 comprises an amino acid sequence of SEQ ID NO: 7; and the CDR-L3 comprises an amino acid sequence of SEQ ID NO: 8.

[0012] In some embodiments, the VH comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 1 and the VL comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 2. In some embodiments, the VH comprises an amino acid sequence of SEQ ID NO: 1 and the VL comprises an amino acid sequence of SEQ ID NO: 2. In some embodiments, the bispecific CAR comprises an amino acid sequence of any one of SEQ ID NOs: 19-21 between the VH and the VL. In some embodiments, the mesothelin-binding antibody fragment comprises an amino acid sequence of SEQ ID NO: 9.

[0013] In some embodiments, the second antigen-binding domain comprises a MUC16-binding antibody. In some embodiments, the MUC16-binding antibody binds to the ectodomain of MUC16. In some embodiments, the MUC16-binding antibody is a MUC16-binding antibody fragment. In some embodiments, the MUC16-binding antibody fragment is an antigen-binding fragment (Fab), a Fab', or a F(ab')2, a fragment variable (fv), or a single chain variable fragment (scFv).

[0014] In some embodiments, the MUC16-binding antibody comprises: (a) a heavy chain variable domain (VH) comprising three complementarity determining regions CDR-H1, CDR- H2, and CDR-H3, wherein the CDR-H1 comprises an amino acid sequence of SEQ ID NO: 12; the CDR-H2 comprises an amino acid sequence of SEQ ID NO: 13; and the CDR-H3 comprises an amino acid sequence of SEQ ID NO: 14, and (b) a light chain variable domain (VL) comprising three complementarity determining regions CDR-L1, CDR-L2, and CDR-L3, wherein the CDR-L1 comprises an amino acid sequence of SEQ ID NO: 15; the CDR-L2 comprises an amino acid sequence of SEQ ID NO: 16; and the CDR-L3 comprises an amino acid sequence of SEQ ID NO: 17.

[0015] In some embodiments, the VH comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 10 and the VL comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 11. In some embodiments, the VH comprises an amino acid sequence of SEQ ID NO: 10 and the VL comprises an amino acid sequence of SEQ ID NO: 11. In some embodiments, the bispecific CAR comprises an amino acid sequence of SEQ ID NO: 21 between the VH and the VL. In some embodiments, the MUC16-binding antibody fragment comprises an amino acid sequence of SEQ ID NO: 18.

[0016] In some embodiments, the bispecific CAR further comprises a peptide linker between the first antigen-binding domain that binds to mesothelin and the second antigen-binding domain that binds to MUC16. In some embodiments, the peptide linker comprises SEQ ID NO: 19. In some embodiments, the peptide linker comprises SEQ ID NO: 20. In some embodiments, the peptide linker comprises SEQ ID NO: 21.

[0017] In some embodiments, the bispecific CAR comprises, from N-terminal to C-terminal: (i) the first antigen-binding domain that binds to mesothelin; (ii) the peptide linker; and (iii) the second antigen-binding domain that binds to MUC16.

[0018] In some embodiments, the bispecific CAR comprises, from N-terminal to C-terminal: (i) the first antigen-binding domain that binds to MUC16; (ii) the peptide linker; and (iii) the second antigen-binding domain that binds to mesothelin.

[0019] In some embodiments, the bispecific CAR comprises: (a) a transmembrane domain; (b) a co-stimulatory domain; and (c) an intracellular signaling domain.

[0020] In some embodiments, the transmembrane domain is selected from the group consisting of alpha (a), beta (P) or zeta (Q chain of a T cell receptor, CD28, CD3 epsilon (a), CD45, CD4, CD5, CD8, CD9, CD16, CD22, CD33, CD37, CD64, CD80, CD86, CD134, CD137, CD154, KIRDS2, 0X40, CD2, CD27, LFA-1 (CDl la, CD18), ICOS (CD278), 4-1BB (CD137), 4- 1BBL, GITR, CD40, BAFFR, HVEM (LIGHTR), SLAMF7, NKp80 (KLRFI), CD 160, CD 19, IL2R beta (P), IL2R gamma (y), IL7Ra, ITGA1, VLA1, CD49a, ITGA4, IA4, CD49D, ITGA6, VLA-6, CD49f, ITGAD, CDl ld, ITGAE, CD103, ITGAL, CDl la, LFA-1, ITGAM, CDl lb, ITGAX, CDl lcJTGBl, CD29, ITGB2, CD18, LFA-1, ITGB7, TNFR2, DNAM1 (CD226), SLAMF4 (CD244, 2B4), CD84, CD96 (Tactile), CEACAM1, CRT AM, Ly9 (CD229), CD 160 (BY55), PSGL1, CD100 (SEMA4D), SLAMF6 (NTB-A, LylO8), SLAM (SLAMF1, CD150, IPO-3), BLAME (SLAMF8), SELPLG (CD162), LTBR, PAG / Cbp, NKp44, NKp30, NKp46, NKG2D, and / or NKG2C transmembrane domains. In some embodiments, the transmembrane domain comprises a CD8 transmembrane domain. In some embodiments, CD8 transmembrane domain comprises a CD8 transmembrane domain and a CD8 hinge domain (CD8 hinge / TM). In some embodiments, the CD8 hinge / TM comprises an amino acid sequence of SEQ ID NO: 22.

[0021] In some embodiments, the co-stimulatory domain comprises a 4-1BB, CD27, CD28, 0X40, CARD11, CD2, CD7, CD27, CD28, CD30, CD40, CD54 (ICAM), CD83, CD134 (0X40), CD137 (4-1BB), CD150 (SLAMF1), CD152 (CTLA4), CD223 (LAG3), CD270 (HVEM), CD273 (PD-L2), CD274 (PD-L1), CD278 (ICOS), DAP10, LAT, NKD2C SLP76, TRIM, or ZAP70 co-stimulatory domain. In some embodiments, the co-stimulatory domain comprises a 4- IBB co-stimulatory domain. In some embodiments, the co-stimulatory domain comprises a 4-1BB co-stimulatory domain comprising an amino acid sequence of SEQ ID NO: 23.

[0022] In some embodiments, the intracellular signaling domain comprises a CD28, 4- IBB, CD27, TCRi FcRy, FcRp, CD3 gamma (y), CD3 theta (6), CD3 sigma (o), CD3 eta (q), CD3s, CD3(^, CD22, CD79a, CD79b, or CD66d intracellular signaling domain. In some embodiments, the intracellular signaling domain comprises a CD3(^ signaling domain. In some embodiments, the CD3(^ signaling domain comprises an amino acid sequence of SEQ ID NO: 24.

[0023] In some embodiments, the bispecific CAR further comprises a leader sequence. In some embodiments, the leader sequence comprising a CD8 leader sequence. In some embodiments, the CD8 leader sequence comprises SEQ ID NO: 25.

[0024] In some embodiments, the bispecific CAR further comprises a CD8 hinge / transmembrane domain, a 4-1BB co-stimulatory domain, and a CD3(^ intracellular signaling domain.

[0025] In some embodiments, the bispecific CAR comprises an amino acid sequence having at least 90% identity to any one of SEQ ID NOs: 26-31. In some embodiments, the bispecific CAR comprises an amino acid sequence of any one of SEQ ID NOs: 26-31. In some embodiments, the bispecific CAR comprises an amino acid sequence of SEQ ID NO: 26. In some embodiments, the bispecific CAR comprises an amino acid sequence of SEQ ID NO: 28. In some embodiments, this disclosure provides a polynucleotide comprising a nucleic acid encoding the bispecific CAR. In some embodiments, this disclosure provides a vector comprising the polynucleotide In some embodiments, the vector is a viral vector. In some embodiments, the vector is a lentiviral vector.

[0026] In some embodiments, this disclosure provides a cell comprising the CAR, the polynucleotide, or the vector. In some embodiments, the cell is an immune cell. In some embodiments, the cell is a T cell.

[0027] In some embodiments, this disclosure provides a method of treating a subject having a mesothelin and / or MUC16 expressing cancer, the method comprising administering the cell to the subject. In some embodiments, the method comprises administering the cell to the subject. In some embodiments, the cancer is ovarian cancer. In some embodiments, the cancer is pancreatic cancer. In some embodiments, the cell in an autologous T cell. In some embodiments, the cell in an allogeneic T cell.

[0028] BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The following drawings form part of the present specification and are included to further demonstrate certain aspects of the present disclosure, which can be better understood by reference to one or more of these drawings in combination with the detailed description of specific embodiments presented herein. The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

[0030] FIGS. 1A-1B relate to screening for scFv configuration for bispecific chimeric antigen receptors (CARs) (“TanCARs”). FIG. 1A is a schematic representation of transgenes encoding different mono-specific and tandem scFv, and workflow for the in vitro screening of tandem scFvs FIG. IB is a multivariable graph summarizing mean intensity fluorescence (MFI) values of green fluorescent protein (GFP) in CAR-Jurkat NFAT-eGFP cells after exposure to different tumors cells (ASPC1 and 0VCAR3), and MFI values of soluble antigen bound to CAR-Jurkat NFAT-eGFP cells. Soluble mesothelin concentration used was 5pg / mL, soluble MUC16 (CA125) concentration used was lOpg / mL.

[0031] FIGS. 2A-2J relate to in vitro anti-tumor activity and tumor cell type selection by bispecific CAR-T cells. FIGs. 2A-2B show luciferase killing assay results at effectortarget (E:T) ratio of 3: 1 after 24h of co-culture with bispecific CAR-T cells expressing TanCARl or TanCAR3, monospecific CAR-T cells, and untransduced T cells against ASPC-1 (FIG. 2A) and OVCAR3 (FIG. 2B) cancer cells (*p<0.05, **p<0.001; two-tailed student’s T test). FIG. 2C is a violin plot showing the area of tumor cells (y-axis) remaining after 96 hours of co-culture with CAR-T cells. The violin plot shows three different mixed cultures combining single positive, double positive, and / or double negative tumor cells for the target antigens. Tumor area values are ratios from the beginning of the assay (time = 0); values greater than 1 indicate poor tumor control, values below 1 indicate tumor clearance (*p<0.05, **p<0.001; Mann-Whitney test). FIGs. 2D-2F show antigen expression on tumor cells after 96 hours of co-culture with CAR-T cells. In some experiments, a mixed tumor model of single positive (meso+ or MUC16+, “MUCecto+”), and double positive or double negative cells was used (FIGs. 2D-2E). In some experiments, a mixed tumor model of single positive, double positive, and double negative cells was used (FIG. 2F) FIGs. 2G-2H show cytotoxicity of CAR-T cells and control cells against meso+ tumor cells (FIG. 2G) or MUC16+ tumor cells (FIG. 2H) over 96 hours of co-culture (**p<0.001, ***p<0.0001; 2-way ANOVA, Tukey’s multiple comparison test). FIGs. 2I-2J show cytotoxicity of CAR-T cells and control cells against spheroids made of meso+ (FIG. 21) or MUC16+ (FIG. 2 J) cells over 107 hours of co-culture. Dotted line indicates tumor clearance (*p<0.05; 2-way ANOVA, Tukey’s multiple comparison test). For FIGs. 2A-2F, CAR-T cells were generated from three healthy donors. For FIGs. 2G-2J, CAR-T cells were generated from two healthy donors.

[0032] FIGs. 3A-3H relate to cytokine secretion by bispecific and monospecific CAR-T cells after 24 hours of co-culture with ASPC-1 (FIGs. 3A-3D) or OVCAR3 (FIGs. 3E-3H) tumor cells expressing mesothelin, MUC16, both mesothelin and MUC16, or non-cognate antigens. Cytokines assayed were TNFa (FIGs. 3A, 3D), GM-CSF (FIGs. 3B, 3E), IFNy (FIGs. 3C, 3F), and IL-2 (FIGs. 3D, 3G). For FIGs. 3 A-3D, CAR-T cells were derived from two healthy donors. For FIGs. 3E-3H, CAR-T cells were derived from three healthy donors.

[0033] FIGs. 4A-4E relate to avidity of bispecific CAR-T cells under different conditions. FIGs. 4A-4B show the percentage of CAR-T cells or control cells bound to ASPC-1 meso+ cells (FIG. 4A) or ASPC-1 MUC16+ cells (FIG. 4B) during acoustic force application as measured in picoNewtons (pN). FIGs. 4C-4D show the percentage of CAR-T cells or control cells bound to ASPC-1 meso+ / MUC16+ cells after five minutes (FIG. 4C) or ten minutes (FIG. 4D) of incubation with force application. FIG. 4E is a bar graph summarizing the results from FIGs. 4C-4D at the maximum applied force (1,000 pNN) (*<0.05, Brown-Forsythe and Welch ANOVA test, Dunnett’s T3 multiple comparison). CAR-T cells were derived from one donor. FIGs. 5A-5I relate to in vivo anti -tumor activity of bispecific CAR-T cells against double-positive tumor cell and mixed tumor models. FIG. 5A is an experimental schematic representation of an in vivo experiment using immunodeficient NSG (NOD.Cg- PrkdcSCidH2r^mlW}lI 7j') mice engrafted with 3 x 1060VCAR3 cells intraperitoneally on day -14, followed by intraperitoneal infusion of 1 x 106CAR-T cells or control cells on day 0. FIG. 5B shows quantitative flux data from an experiment as performed in FIG. 5A. FIG. 5C is a schematic representation of an in vivo experiment using NSG mice subcutaneously engrafted with 2 x 106ASPC-1 mixed tumor cells on day -14, followed by intravenous infusion of 3 x 106CAR-T cells or control cells on day 0. FIG. 5D shows a summary of caliper measurements of tumors in mice treated as in FIG. 5C. Caliper measurements of tumor growth in individual mice treated with SSI CAR-T cells (FIG. 5E), 4H11 CAR-T cells (FIG. 5F), TanCARl cells (FIG. 5G), untransduced cells (FIG. 5H), or that did not receive cells (FIG. 51) are shown in FIGs. 5E-5I. For FIG. 5B, CAR-T cells were derived from one donor. For FIGs. 5D-5I, CAR-T cells were derived from two donors. **p<0.001, ***p<0.0001, ANOVA test.

[0034] FIGs. 6A-6C demonstrate that mesothelin and MUC16 are heterogeneously expressed in patient samples and human-derived cancer cell lines. FIG. 6A shows a dot plot of Mesothelin (APC) and MUC16 (PE) expression in cells obtained from three patients diagnosed with ovarian cancer. Numbers in each quadrant represent percentages of the total “Alive” cell population as determined by DAPI staining via flow cytometry. FIG. 6B is a bar graph of Double-, Double+, MUC16+, and Mesothelin+ tumor cell frequency from the patient samples in FIG. 6A and ovarian and pancreatic human cancer cell lines. FIG. 6C shows a heatmap of mesothelin and MUC16 RNA expression values (TPM Log2 transformed) in PDXs samples from patients diagnosed of ovarian (upper heatmap) or pancreatic (bottom heatmap) cancer.

[0035] FIGs. 7A-7B relate to mesothelin and MUC16 expression in ovarian and pancreatic human cancer cell lines. FIG. 7A is a dot plot of Mesothelin (APC) and MUC16 (PE) expression in ovarian (upper panel) and pancreatic (lower panel) human cancer cell lines. Numbers in each quadrant represent percentages of the total “alive” cell population by DAPI staining via flow cytometry. FIG. 7B shows histograms of Mesothelin and MUC16 median fluorescence intensity (MFI) in cancer cell lines from FIG. 7A. Events in each histogram were normalized to the mode.

[0036] FIGs. 8A-8F relate to tandem scFv screening to target mesothelin and MUC16. FIG. 8A are exemplary schematics of six tandem scFv designs. FIG. 8B shows MFI values of soluble mesothelin (upper panel) and soluble MUC16 (lower panel) bound to each CAR construct expressed in Jurkat cells. MFI values were obtained from Jurkat-mCherry+ cells. Bars represent the mean ± standard error of the mean (SEM) of three technical replicates. FIG. 8C shows flow cytometry histograms showing mesothelin and MUC16 / MUC16ecto expression in ASPC-1 and 0VCAR3 cancer cell lines that endogenously express (endo) or do not express (neg), are knocked out (KO) for, or are artificially transduced (TR) for mesothelin, MUC16 or MUC16ecto. The MUC16ecto gene sequence was in cis with BFP; therefore, transduction of MUC16ecto was evaluated by BFP expression. FIG. 8D is a heat map of GFP MFI values from CAR-Jurkat NFAP-GFP reporter cells exposed to tumor cells for 24 hours. MFI values were obtained from mCherry+ CAR-Jurkat NFAT-eGFP cells. FIGs. 8E-8F show multiparametric representation of the MFI of soluble antigen bound to CAR-Jurkat cells (X and Y values) along with the MFI of the GFP (scale) and the percentage of GFP+ cells (circle size) in CAR-Jurkat NFAT-eGFP mCherry+ cells in response to ASPC-1 cells (FIG. 8E) and 0VCAR3 cells (FIG. 8F). MFI values of GFP (scale) and the percentage of GFP+ cells (circle size) are the mean of each CAR- Jurkat NFAT-eGFP reporter cell co-cultures with the different ASPC-1 and 0VCAR3 (FIG. 8D) tumor cells, excluding the double negative tumor cells. In FIG. 8B, significance between the different tandem scFv constructs was measured using one-way ANOVA test with Tukey’s multiple comparison test in FIG. 8B. *, P<0.05; **, P<0.01; ***, PO.OOl; ****, PO.OOOl.

[0037] FIGs. 9A-9E shows the transduction efficiency and CAR expression in Jurkat cells. FIG. 9A is an exemplary schematic of constructs used for the screening strategy. FIG. 9B is a bar graph showing transduction efficiency of CAR-Jurkat cells represented as percentage of “alive” cells by DAPI staining via flow cytometry. FIGs. 9C-9D are bar graphs showing the MFI values of anti-mCherry on CAR-Jurkat mCherry+ (FIG. 9C) and G4S antibody on CAR- Jurkat G4S+ cells (FIG. 9D). FIG. 9E is a bar graph showing transduction efficiency of CAR- Jurkat NFAT-eGFP cells represented as percentage of “alive” cells by DAPI staining via flow cytometry. Bars represent the mean ± standard error of the mean (SEM) of four technical replicates.

[0038] FIGs. 10A-10F relates to TanCAR in vitro anti-tumor activity, cytokine production, and avidity. FIGs. 10A-10B are bar graph showing the percent cytotoxicity after 24 hours of ASPC- 1 (FIG. 10A) or 0VCAR3 (FIG. 10B) cancer cells co-cultured with CAR-T cells (SSI, 4H11, TanCARl or TanCAR3) or UTD-T cells at an E:T ratio of 3: 1, assessed by luciferase-based killing assay. Bars represent the mean ± SEM of three healthy donors measured in triplicate or duplicate. Stars indicate significant differences between mono-specific CAR-T cells and tandem CAR-T cells using a two-tailed Mann-Whitney test. FIGs. 10C-10D are heat maps showing the pg / mL concentration of IL-2, IFNy, TNFa, or GM-CSF in supernatant collected from killing assays after 24 hours of co-culture of CAR-T cells with ASPC-1 (FIG. 10C) or OVCAR3 (FIG. 10D) cancer cells at an E:T ratio of 10: 1. Data is the mean from two healthy donors. FIGs. 10E- 10F show cell-to-cell binding avidity of TanCARl, TanCAR3, or UTD-T cells to ASPC-1 (FIG. 10E) Mesoend°MUC16ectoTRand to OVCAR3 (FIG. 10F) Mesoend°MUC16endo assessed by acoustic force using a z-Movie cell avidity analyzer. Data are shown as the percentage of T cells bound per unit of acoustic force applied in piconewtons and is the mean ± SEM of two human healthy donor with four or three replicates, endo, endogenous expression of the antigen; ectoTR, transduced with the MUC16 ectodomain; KO, CRISPR KO of the endogenous antigen; neg, endogenously negative for the antigen. Differences between TanCARl and TanCAR3 measured using a two-way ANOVA test with the Geisser-Greenhouse correction test. *p<0.05, **p<0.01, *** p<0.001, ****p<0.0001, ns, non-significant.

[0039] FIGs. 11 A-l IN compare TanCARl and TanCAR3 in vitro using in silico structure prediction. FIG. 11 A shows transduction efficiency of mono-specific and tandem CARs in primary human T cells from healthy donors. Transduction efficiency is measured as the percent of mCherry+ T cells via flow cytometry. Data are the mean ± SEM of four different human healthy donors. FIGs. 11C-11 J show luciferase killing assays from co-cultures of ASPC-1 (FIGs. 11C-1 IF) or OVCAR3 (FIGs. 11G-11 J) with CAR- or UTD-T cells. Each panel shows the cytotoxic activity of CAR- or UTD-T cells co-cultured in different ratios (shown on the X- axis) with different antigen- expressing cancer cells, endo, endogenous expression of the antigen; ectoTR, transduced with the MUC16 ectodomain; KO, CRISPR KO of the endogenous antigen; neg, endogenously negative for the antigen. FIGs. 1 IK-1 IL show structures predicted using AlphaFold3 of TanCARl (FIG. UK) and TanCAR3 (FIG. 11L). Legend indicates the pLDDT ranks. 180 degrees-rotated structure is shown for each construct. H / TM, hinge and transmembrane domains. FIGs. 1 IM-1 IN show predicted align error matrices of TanCARl (FIG. 1 IM) and TanCAR3 (FIG.1 IN) constructs. Differences between the indicated CAR-T cell groups measured by a two-way ANOVA with Turkey’s multiple comparison test. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

[0040] FIGs. 12A-12C relate to TanCAR-T cell binding avidity to tumor cells expressing one or both antigens. FIGs. 12A-12B are avidity curves showing the percentage of T cells bound to single antigen-expressing ASPC-1 tumor cells (FIG. 12A, Mesoc / 7d<MUC16“?; FIG. 12B, MesoKOMUC I 6ectorK) per acoustic force unit applied in picoNewtons (pN). FIG. 12C is a schematic of one antigen at a time (upper panel) or two antigens at a time (lower panel) hypotheses for the binding of tandem CAR to two antigens. FIG. 12D is an avidity curve showing the percentage of T cells bound to ASPC-1 tumor cells expressing both antigens (Meso"doMUCl 6ectoTO) per pN of acoustic force applied. Stars indicate significant differences as measured by a two-way ANOVA with Fisher’s LSD test. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, non-significant.

[0041] FIGs. 13A-13K demonstrate that Tandem CAR-T cells overcome tumor heterogeneity. FIG. 13A is a schematic of a mixed tumor cell culture with ASPC-1 MesoendoMUC16neg, MesoKOMUC16ectoTR, and Mesoend°MUC16ectoTRcells. FIG. 13B are violin plots of the tumor area 96 hours after co- culture with CAR- or UTD-T cells, relative to time 0, measured via realtime killing assays using an IncuCyte® Live-Cell Analysis system. The dotted line at y=l represents the threshold for tumor clearance (below 1). Each dot in the violin plot represents a replicate of the experiment, with four technical replicates of 3 healthy donor T cells. FIG. 13C is a bar graph showing the mean absolute number of each tumor cell population at the end of the real-time killing assay in FIG. 13B, per CAR-T cell treatment group. Each color of the stacked bars represents the tumor populations depicted in FIG. 13 A. FIG. 13D shows a dot plot of the absolute number of each tumor cell population showed in FIG. 13C, per CAR-T cell treatment group. Each dot represents a replicate of the experiment, with two technical replicates of 3 healthy donor T cells. FIGs. 13E-13H show real-time cytotoxicity assays with CAR-T cells or UTD-T cells co-cultured for 96 hours 1 : 1 with (FIG. 13E, left) a mixture of ASPC-1 MesoendoMUC16negiRFP+ cells and MesoKOMUC16ectoTRGFP+ cells, or (FIG. 13E, right) total tumor cell area relative to time 0 (start of the co-culture), measured using an IncuCyte® Live- Cell Analysis system. FIGs. 13F-13G show cell area (MesoendoMucl6neg, FIG. 13F; MesoKOMUC16ectoTR, FIG. 13G) relative to time 0. Curves represent the mean ± SEM of three technical replicates of two healthy donors for each treatment group. FIG. 13H is a dot plot of MesoendoMUC16negand MesoKOMUC16ectoTRtumor cell area at 96h of co-culture with TanCARl relative to time 0 from FIGs. 13F-13G. FIGs. 13I-13K show mixed tumor spheroids were created with MesoendoMUC16negiRFP+ and MesoKOMUC16ectoTRGFP+ cells and treated with CAR T cells. FIG. 131 shows representative images from different time points of mixed- tumor spheroids treated with CAR-T cells. FIGs. 13J-13K depict violin plots of the ASPC1 MesoendoMUC16neg(FIG. 13 J) and ASPC-1 MesoKOMUC16ectoTR(FIG. 13K) tumor cell area at 106 hours relative to time 0, measured using an IncuCyte® Live-Cell Analysis system. Dots represent technical triplicates and two healthy donors for each treatment group. Differences measured by Mann-Whitney test for each comparison in FIGs. 13B, 13H, 13 J, 13K; Kruskal- Wallis test for FIG. 13D; and a two-way ANOVA with Fisher’s LSD test for FIGs. 13E-13G. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, non-significant.

[0042] FIGs. 14A-14M related to stressed heterogenous tumor model and real time killing and expansion of CAR-T cells co-cultured with a mixed ASPC-1 spheroid. FIG. 14A is an exemplary schematic of stressed mixed tumor cell culture of ASPC-1 MesoendoMUC16neg, MesoKOMUC16ectoTR, and MesoKOMUC16neg. FIG. 14B shows violin plots of the tumor area 96 hours after co-culture with CAR T or UTD cells, relative to time 0, measured via real-time killing assays using an IncuCyte® Live-Cell Analysis system. The dotted line at y=l represents the threshold for tumor clearance (below 1). Each dot in the violin plot represents a replicate of the experiment, with two technical replicates of 3 healthy donor T cells. FIG. 14C shows bar graph showing the mean absolute number of each tumor cell population at the end of the realtime killing assay in FIG. 14B, per CAR-T cell treatment group. Each color of the stacked bars represents the tumor populations depicted in FIG. 14A. FIG. 14D is a dot plot of the absolute number of each tumor cell population showed in FIG. 14C, per CAR-T cell treatment group. Each dot represents a replicate of the experiment, with two technical replicates of 3 healthy donor T cells. FIG. 14E is a bar graph showing the mean absolute number of each tumor cell population at the end of the real-time killing assay in FIG. 14B, UTD-T cells and Tumor only groups are depicted. FIG. 14F is a dot plot of the absolute number of each population showed in the bar graph in FIG. 14E. FIG. 14G is a bar graph showing the mean absolute number of each tumor cell population at the end of the real-time killing assay in A. UTD-T cells and Tumor only groups are depicted. FIG. 14H is a dot plot of the absolute number of each population showed in the bar graph in FIG. 14G. FIG. 141 is an exemplary schematic of a mixed tumor spheroid containing ASPC-1 MesoendoMUC16negiRFP+ and MesoKOMUC16ectoTRGFP+ cells, after three days of seeding. The line represents the cross-section of the spheroid captured and analyzed for tumor cell growth and CAR T cell proliferation and infiltration using an IncuCyte® Live- Cell Analysis system. FIG. 14J is a representative image of the mixed tumor spheroid cross section on day 3, directly after CAR T cell addition. FIGs. 14K-14L show total GFP+ tumor cell area (FIG. 14K) and mCherry+ CAR-T cell area (FIG. 14L) of the mixed tumor spheroid treated with CAR T cells over 106 hours, relative to time 0 (start of the coculture), measured using an IncuCyte® Live-Cell Analysis system. Curves represent the mean ± SEM of three technical replicates of two healthy donors for each treatment group. FIG. 14M is a dot plot of CAR-T cell (mCherry+ cells) absolute numbers after 106 hours of co-culture with spheroids, measured by flow cytometry. Dots represent technical triplicates and two healthy donors for each treatment group. Stars indicate significant differences as measured by Mann-Whitney test for each comparison in FIGs. 14B, 14D, 14F, 14H, 14M and a two-way ANOVA with Fisher’s LSD test for FIGs. 14K-14L. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, non-significant.

[0043] FIGs. 15A-15I show the in vivo anti -tumor activity of TanCarl-T in double positive tumor cell heterogenous tumor models. FIG. 15A is a schematic of the in vivo experiments in NSG mice engrafted intraperitoneally with OVCAR3 tumor cells and treated 14 days later with CAR-T or UTD-T cells or left untreated (tumor alone). FIG. 15B shows quantification of the flux (photons / s2) from mice treated as indicated. Curves represent the median + interquartile range of each treatment group, with 14 mice per group treated with T cells from two healthy donors. FIG. 15C is a schematic of the in vivo experiments in NSG mice engrafted subcutaneously with mixed ASPC-1 tumor cells and treated 14 days later with CAR-T or UTD- T cells or left untreated (tumor alone). FIG. 15D shows caliper measurements from mice treated as indicated. Curves represent the mean ± SEM of each treatment group, with 14 mice per group, repeated with T cells from 2 healthy donors. FIG. 15E shows percentages of each tumor cell population from mice in FIG. 15D, at the time of injection compared to the time of tumor collection from the mice. Each dot represents the mean ± SEM of each treatment group treated with T cells from one healthy donor: SSI CAR group contains primary tumor and lung metastasis from 1 mouse, 4H11 CAR group contains primary tumors from 3 mice, and TanCARl group shows primary tumors from 6 mice. The statistical analysis from 4H11 and TanCARl groups is shown. FIG. 15F is a diagram explaining the skewed killing of TanCARl- T cells towards tumor cells expressing high levels of one of the cognate antigens. FIG. 15G is a schematic of the in vivo experiment in NSG mice engrafted subcutaneously with mixed ASPC-1 tumor cells, treated 14 days later with CAR-T or UTD-T cells, and euthanized at day 21 after CAR / UTD-T cells injection. FIG. 151 shows representative images of H4C slides of tumors from mice after 21 days of CAR / UTD-T cell administration (summarized in FIG. 15H). Lower panels are magnifications of highlighted area in the upper panels. FIG. 15 J is a dot plot graph showing CD3+ cells per mm2in each group of treated mice. Differences measured by a two-way ANOVA test with Fisher’s LSD test for FIG. 15B and FIG. 15D; a Mann-Whitney test for FIG. 15E; and a one-way ANOVA with Holm-Sidak's multiple comparisons test for FIG. 15H. Only comparisons between CAR-T cells (TanCARl, SS1CAR, and 4H11CAR) are shown. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, non-significant.

[0044] FIGs. 16A-16E shows individual mouse data from in vivo experiments. FIG. 16A is a spider plot of the flux values (photons / s2) of each mouse from the experiment shown in FIG. 15B. FIG. 16B is a scatter plot of mesothelin and MUC16ecto expression in among the mixed ASPC-1 (MesoendoMUC16neg, MesoKOMUC16ectoTR, and Mesoend°MUC16ectoTR) tumor cells just before being engrafted subcutaneously into mice for the experiment shown in FIGs. 13C- 13D as measured by flow cytometry. FIG. 16C Spider plot of the caliper measurements (mm3) of each mouse from the experiment shown in FIG. 15D, with both donors plotted. FIG. 16D shows caliper measurements from mice treated as indicated in FIG. 151. Curves represent the mean ± SEM of each treatment group, with 5 mice (SSI, 4H11, and TanCARl) or 3 mice (UTD) per group, from one healthy donor. FIG. 16E is a spider plot of the caliper measurements (mm3) of each mouse from the experiment shown in FIG. 15G.

[0045] DETAILED DESCRIPTION

[0046] In some aspects, provided herein are bispecific chimeric antigen receptors (CARs) and CAR-T cells that target mesothelin (MSLN) and mucin 16 (MUC16, also called CA125). Mesothelin is expressed on normal mesothelial cells in some tissues (e.g., pleura, pericardium, peritoneum) and in trace amounts in some epithelial cells (e.g., ovary, tunica vaginalis, rete testis, and fallopian tube), but is abundantly expressed in various cancer cells. (See, e.g., Lv, Jiang, and Li, Peng. “Mesothelin as a biomarker for targeted therapy”. Biomark Res. 2019; 7: 18; and Hassan et al. “Mesothelin Immunotherapy for Cancer: Ready for Prime Time?” J Clin Oncol. 2016 Dec 1; 34(34): 4171-4179, each of which is hereby incorporated by reference). Mesothelin may be used as a marker for cells associated with various cancers (e.g., overexpressed in various cancers), and CARs, bispecific CARs, and CAR-T cells that bind to mesothelin or a portion thereof may be used to treat subjects having, e.g., cancers associated with mesothelin expression. MUC16 is primarily expressed in various epithelial cells, especially epithelial cells of the female reproductive tract, as well as the respiratory and gastrointestinal tract. Like mesothelin, upregulation of MUC16 has been observed in several cancers, including pancreatic cancer and ovarian cancer.

[0047] A major obstacle in applying the success of CAR-T therapy observed in hematologic cancers to solid cancers is that solid tumors are not typically characterized by expression of surface markers, making it difficult to identify strong candidate target antigens. Additionally, solid tumors are more heterogenous than liquid cancers, which can lead to outgrowth of tumors that lose expression of an antigen targeted by CAR-T therapy. Accordingly, the disclosure relates to bispecific CARs targeting mesothelin and MUC16 and methods of use thereof. Targeting both tumors can overcome tumor heterogeneity and confer greater efficacy in the application of CAR-T therapy to solid tumors.

[0048] General Definitions

[0049] The terms "decrease", "reduced", or "reduction" are all used herein to mean a decrease by a statistically significant amount. In some embodiments, "reduce", "reduction", or "decrease" typically means a decrease by at least 10% as compared to a reference level (e.g., the absence of a given treatment or agent) and can include, for example, a decrease by at least about 10%, at least about 20%, at least about 25%, at least about 30%, at least about 35%, at least about 40%, at least about 45%, at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 98%, at least about 99% , or more. Where applicable, a decrease can be preferably down to a level accepted as within the range of normal for an individual without a given disorder.

[0050] A "disease" is a state of health of an animal, for example, a human, wherein the animal cannot maintain homeostasis, and wherein if the disease is not ameliorated, then the animal's health continues to deteriorate. In contrast, a "disorder" in an animal is a state of health in which the animal is able to maintain homeostasis, but in which the animal's state of health is less favorable than it would be in the absence of the disorder. Left untreated, a disorder does not necessarily cause a further decrease in the animal's state of health. In some embodiments, the disease is a cancer or a tumor.

[0051] As used herein, the terms "tumor antigen", “tumor-associated antigen” and "cancer antigen" are used interchangeably to refer to antigens that are differentially expressed by cancer cells and can thereby be exploited in order to target cancer cells. Cancer antigens are antigens that can potentially stimulate tumor-specific immune responses. Some of these antigens are encoded, although not necessarily expressed, by normal cells. These antigens can be characterized as those which are normally silent (i.e., not expressed) in normal cells, those that are expressed only at certain stages of differentiation and those that are temporally expressed such as embryonic and fetal antigens. Other cancer antigens are encoded by mutant cellular genes, such as oncogenes (e.g., activated Ras oncogene), suppressor genes (e.g., mutant p53), and fusion proteins resulting from internal deletions or chromosomal translocations. Still other cancer antigens can be encoded by viral genes such as those carried on RNA and DNA tumor viruses.

[0052] As used herein, the term "chimeric" refers to the product of the fusion of portions of at least two or more different polynucleotide molecules. In some embodiments, the term "chimeric" refers to a gene expression element produced through the manipulation of known elements or other polynucleotide molecules.

[0053] In some embodiments, "activation" can refer to the state of a T cell that has been sufficiently stimulated to induce detectable cellular proliferation. In some embodiments, activation can refer to induced cytokine production. In other embodiments, activation can refer to detectable effector functions.

[0054] At a minimum, an "activated T cell" as used herein is a proliferative T cell.

[0055] As used herein, the terms "specific binding" and "specifically binds" refer to a physical interaction between two molecules, compounds, cells and / or particles wherein the first entity binds to the second, target, entity with greater specificity and affinity than it binds to a third entity which is a non-target. In some embodiments, specific binding can refer to an affinity of the first entity for the second target, entity, which is at least 10 times, at least 50 times, at least 100 times, at least 500 times, at least 1000 times or more greater than the affinity for the third non-target entity under the same conditions. A reagent specific for a given target is one that exhibits specific binding for that target under the conditions of the assay being utilized. A nonlimiting example includes an antibody, or a ligand, which recognizes and binds with a cognate binding partner (for example, a stimulatory and / or co-stimulatory molecule present on a T cell) protein.

[0056] A "stimulatory ligand," as used herein, refers to a ligand that when present on an antigen presenting cell (APC) (e.g., a macrophage, a dendritic cell, a B-cell, an artificial APC, and the like) can specifically bind with a cognate binding partner (referred to herein as a "stimulatory molecule" or "co-stimulatory molecule") on a T cell, thereby mediating a primary response by the T cell, including, but not limited to, proliferation, activation, initiation of an immune response, and the like. Stimulatory ligands are well-known in the art and encompass, inter alia, an MHC Class I molecule loaded with a peptide, an anti-CD3 antibody, a superagonist anti- CD28 antibody, and a superagonist anti-CD2 antibody.

[0057] A "stimulatory molecule," as the term is used herein, means a molecule on a T cell that specifically binds with a cognate stimulatory ligand present on an antigen presenting cell. "Co- stimulatory ligand," as the term is used herein, includes a molecule on an APC that specifically binds a cognate co-stimulatory molecule on a T cell, thereby providing a signal which, in addition to the primary signal provided by, for instance, binding of a TCR / CD3 complex with an MHC molecule loaded with peptide, mediates a T cell response, including, but not limited to, proliferation, activation, differentiation, and the like. A co-stimulatory ligand can include, but is not limited to, 4-1BBL, OX40L, CD7, B7-1 (CD80), B7-2 (CD86), PD-L1, PD-L2, inducible co-stimulatory ligand (ICOS-L), intercellular adhesion molecule (ICAM), CD30L, CD40, CD70, CD83, HLA-G, MICA, MICB, HVEM, lymphotoxin beta receptor, 3 / TR6, IL T3, IL T4, HVEM, an agonist or antibody that binds Toll-like receptor and a ligand that specifically binds with B7-H3. A co-stimulatory ligand also can include, but is not limited to, an antibody that specifically binds with a co-stimulatory molecule present on a T cell, such as, but not limited to, CD27, CD28, 4-1BB, 0X40, CD30, CD40, PD-1, ICOS, lymphocyte function-associated antigen-1 (LFA-1), CD2, CD7, LIGHT, NKG2C, B7-H3, and a ligand that specifically binds with CD83.

[0058] A "co-stimulatory molecule" refers to the cognate binding partner on a T cell that specifically binds with a co-stimulatory ligand, thereby mediating a co-stimulatory response by the T cell, such as, but not limited to, proliferation. Co-stimulatory molecules include but are not limited to an MHC class I molecule, BTLA, a Toll-like receptor, CD27, CD28, 4-1BB, 0X40, CD30, CD40, PD-1, ICOS, lymphocyte function associated antigen-1 (LFA-1), CD2, CD7, LIGHT, NKG2C, B7-H3, and CD83.

[0059] In some embodiments, the term "engineered" and its grammatical equivalents as used herein can refer to one or more human-designed alterations of a nucleic acid, e.g., the nucleic acid within an organism's genome. In another embodiment, engineered can refer to alterations, additions, and / or deletion of genes. An "engineered cell" can refer to a cell with an added, deleted and / or altered gene.

[0060] The term "cell" or "engineered cell" and their grammatical equivalents as used herein can refer to a cell of human or non-human animal origin.

[0061] As used herein, the term "operably linked" refers to a first polynucleotide molecule, such as a promoter, connected with a second transcribable polynucleotide molecule, such as a gene of interest, where the polynucleotide molecules are so arranged that the first polynucleotide molecule affects the function of the second polynucleotide molecule. The two polynucleotide molecules may or may not be part of a single contiguous polynucleotide molecule and may or may not be adjacent. For example, a promoter is operably linked to a gene of interest if the promoter regulates or mediates transcription of the gene of interest in a cell.

[0062] In the various embodiments described herein, it is further contemplated that variants (naturally occurring or otherwise), alleles, homologs, conservatively modified variants, and / or conservative substitution variants of any of the particular polypeptides described are encompassed. As to amino acid sequences, one of ordinary skill will recognize that individual substitutions, deletions or additions to a nucleic acid, peptide, polypeptide, or protein sequence which alters a single amino acid or a small percentage of amino acids in the encoded sequence is a "conservatively modified variant" where the alteration results in the substitution of an amino acid with a chemically similar amino acid and retains the desired activity of the polypeptide. Such conservatively modified variants are in addition to and do not exclude polymorphic variants, interspecies homologs, and alleles consistent with the disclosure.

[0063] A given amino acid can be replaced by a residue having similar physiochemical characteristics, e.g., substituting one aliphatic residue for another (such as He, Vai, Leu, or Ala for one another), or substitution of one polar residue for another (such as between Lys and Arg; Glu and Asp; or Gin and Asn). Other such conservative substitutions, e.g., substitutions of entire regions having similar hydrophobicity characteristics, are well known. Polypeptides comprising conservative amino acid substitutions can be tested in any one of the assays described herein to confirm that a desired activity, e.g., ligand-mediated receptor activity and specificity of a native or reference polypeptide is retained. Amino acids can be grouped according to similarities in the properties of their side chains (in A. L. Lehninger, in Biochemistry, second ed., pp. 73-75, Worth Publishers, New York (1975)): (1) non-polar: Ala (A), Vai (V), Leu (L), He (I), Pro (P), Phe (F), Trp (W), Met (M); (2) uncharged polar: Gly (G), Ser (S), Thr (T), Cys (C), Tyr (Y), Asn (N), Gin (Q); (3) acidic: Asp (D), Glu (E); (4) basic: Lys (K), Arg (R), His (H). Alternatively, naturally occurring residues can be divided into groups based on common sidechain properties: (1) hydrophobic: Norleucine, Met, Ala, Vai, Leu, He; (2) neutral hydrophilic: Cys, Ser, Thr, Asn, Gin; (3) acidic: Asp, Glu; (4) basic: His, Lys, Arg; (5) residues that influence chain orientation: Gly, Pro; (6) aromatic: Trp, Tyr, Phe. Non-conservative substitutions will entail exchanging a member of one of these classes for another class. Particular conservative substitutions include, for example: Ala into Gly or into Ser; Arg into Lys; Asn into Gin or into His; Asp into Glu; Cys into Ser; Gin into Asn; Glu into Asp; Gly into Ala or into Pro; His into Asn or into Gin; He into Leu or into Vai; Leu into He or into Vai; Lys into Arg, into Gin or into Glu; Met into Leu, into Tyr or into He; Phe into Met, into Leu or into Tyr; Ser into Thr; Thr into Ser; Trp into Tyr; Tyr into Trp; and / or Phe into Vai, into He or into Leu.

[0064] In some embodiments, a polypeptide described herein (or a nucleic acid encoding such a polypeptide) can be a functional fragment of one of the amino acid sequences described herein. As used herein, a "functional fragment" is a fragment or segment of a peptide that retains at least 50% of the wildtype reference polypeptide's activity according to an assay known in the art or described below herein. A functional fragment can comprise conservative substitutions of the sequences disclosed herein.

[0065] In some embodiments, a polypeptide described herein can be a variant of a polypeptide or molecule as described herein. In some embodiments, the variant is a conservatively modified variant. Conservative substitution variants can be obtained by mutations of native nucleotide sequences, for example. A "variant," as referred to herein, is a polypeptide substantially homologous to a native or reference polypeptide, but which has an amino acid sequence different from that of the native or reference polypeptide because of one or a plurality of deletions, insertions, or substitutions. Variant polypeptide-encoding DNA sequences encompass sequences that comprise one or more additions, deletions, or substitutions of nucleotides when compared to a native or reference DNA sequence, but that encode a variant protein or fragment thereof that retains activity of the non-variant polypeptide. A wide variety of PCR-based sitespecific mutagenesis approaches are known in the art and can be applied by the ordinarily skilled artisan.

[0066] A variant amino acid or DNA sequence can be at least 80%, at least 85%, at least 90%, at least 91 %, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or more, identical to a native or reference sequence. The degree of homology (percent identity) between a native and a mutant sequence can be determined, for example, by comparing the two sequences using freely available computer programs commonly employed for this purpose on the world wide web (e.g., BLASTp or BLASTn with default settings).

[0067] Alterations of the native amino acid sequence can be accomplished by any of a number of techniques known to one of skill in the art. Mutations can be introduced, for example, at particular loci by synthesizing oligonucleotides containing a mutant sequence, flanked by restriction sites permitting ligation to fragments of the native sequence. Following ligation, the resulting reconstructed sequence encodes an analog having the desired amino acid insertion, substitution, or deletion. Alternatively, oligonucleotide directed site-specific mutagenesis procedures can be employed to provide an altered nucleotide sequence having particular codons altered according to the substitution, deletion, or insertion required. Techniques for making such alterations are well established and include, for example, those disclosed by Walder et al. (Gene 42: 133, 1986); Bauer et al. (Gene 37:73, 1985); Craik (BioTechniques, January 1985, 12-19); Smith et al. (Genetic Engineering: Principles and Methods, Plenum Press, 1981); and U.S. Patent Nos. 4,518,584 and 4,737,462. Any cysteine residue not involved in maintaining the proper conformation of a polypeptide also can be substituted, generally with serine, to improve the oxidative stability of the molecule and prevent aberrant crosslinking. Conversely, cysteine bond(s) can be added to a polypeptide to improve its stability or facilitate oligomerization.

[0068] The term "polynucleotide" is used herein interchangeably with "nucleic acid molecule" to indicate a polymer of nucleosides. Typically, a polynucleotide is composed of nucleosides that are naturally found in DNA or RNA (e.g., adenosine, thymidine, guanosine, cytidine, uridine, deoxyadenosine, deoxythymidine, deoxy guanosine, and deoxy cytidine) joined by phosphodiester bonds. However, the term encompasses molecules comprising nucleosides or nucleoside analogs containing chemically or biologically modified bases, modified backbones, etc., whether or not found in naturally occurring nucleic acids, and such molecules may be preferred for certain applications. Where this disclosure refers to a polynucleotide it is understood that both DNA, RNA, and in each case both single- and double-stranded forms (and complements of each single-stranded molecule) are provided. "Polynucleotide sequence" as used herein can refer to the polynucleotide material itself and / or to the sequence information (i.e., the succession of letters used as abbreviations for bases) that biochemically characterizes a specific nucleic acid. In some embodiments, the nucleic acid molecule is a heterologous nucleic acid molecule. As used herein the term, “heterologous nucleic acid molecule” refers to a nucleic acid molecule that does not naturally exist within a given cell.

[0069] A polynucleotide sequence presented herein is presented in a 5' to 3' direction unless otherwise indicated.

[0070] The term "polypeptide" as used herein refers to a polymer of amino acids. The terms "protein" and "polypeptide" are used interchangeably herein. A peptide may be a relatively short polypeptide, typically between about 2 and 60 amino acids in length. Polypeptides used herein typically contain amino acids such as the 20 L-amino acids that are most commonly found in proteins. However, other amino acids and / or amino acid analogs known in the art can be used. One or more of the amino acids in a polypeptide may be modified, for example, by the addition of a chemical entity such as a carbohydrate group, a phosphate group, a fatty acid group, a linker for conjugation, functionalization, etc. A polypeptide that has a nonpolypeptide moiety covalently or noncovalently associated therewith is still considered a "polypeptide." Exemplary modifications include glycosylation and palmitoylation. Polypeptides can be purified from natural sources, produced using recombinant DNA technology or synthesized through chemical means such as conventional solid phase peptide synthesis, etc. The term "polypeptide sequence" or "amino acid sequence" as used herein can refer to the polypeptide material itself and / or to the sequence information (i.e., the succession of letters or three letter codes used as abbreviations for amino acid names) that biochemically characterizes a polypeptide. A polypeptide sequence presented herein is presented in an N-terminal to C-terminal direction unless otherwise indicated.

[0071] The term "gene" means the nucleic acid sequence which is transcribed (DNA) to RNA in vitro or in vivo when operably linked to appropriate regulatory sequences. The gene may or may not include regions preceding and following the coding region, e.g., 5' untranslated (5' UTR) or "leader" sequences and 3' UTR or "trailer" sequences, as well as intervening sequences (intrans) between individual coding segments (exons).

[0072] As used herein, a "signal peptide" or "signal sequence" refers to a peptide at the N- terminus of a newly synthesized protein that serves to direct a nascent protein into the endoplasmic reticulum. In some embodiments, the signal peptide is a CD8 or IgK signal peptide.

[0073] In some embodiments, a polypeptide, polynucleotide, plasmid and or / vector as described herein optionally further comprises a reporter molecule, e.g., to determine if the vector is properly expressed in a cell. In some embodiments, the reporter molecule may be a fluorescent protein (e.g., GFP, YFP, RF), antibody (e.g., CD34, tEGFR, tCD19, tCD20, tCD34, and tHer2), and a radioisotope. In some embodiments, the reporter molecule is hygromycin phosphotransferase (hph) that can be imaged alone or in combination with a substrate or chemical (for example 9-[4-[18F]fluoro-3-(hydroxymethyl)butyl]guanine ([18F]FHBG)). In some embodiments, GFP and mCherry may be used as fluorescent tags for imaging a bispecific CAR expressed on a T cell (e.g., a CAR-T cell). It is expected that essentially any fluorescent protein known in the art can be used as a fluorescent tag for this purpose. For clinical applications, the bispecific CAR need not include a fluorescent tag or fluorescent protein. In each instance of particular constructs provided herein, therefore, any markers present in the constructs can be removed.

[0074] Chimeric Antigen Receptors (CARs) The terms "chimeric antigen receptor" or "CAR" or "CARs", as used herein, refer to engineered T cell receptors, which graft a ligand or antigen specificity onto immune cells. In some embodiments, the CAR is a bispecific CAR. The term “bispecific CAR” refers to a CAR that comprises a first antigen-binding domain that specifically binds a first antigen (e.g., specifically binds mesothelin) and a second antigen-binding domain that specifically binds a second antigen (e.g., specifically binds Mucl6). Immune cells are of hematopoietic origin, and include lymphocytes, such as B cells and T cells; natural killer cells; and myeloid cells, such as monocytes, macrophages, eosinophils, mast cells, basophils, and granulocytes. In some embodiments, the immune cell is a T cell. In some embodiments, the immune cell is a T-cell (for example, naive T cells, central memory T cells, effector memory T cells or combinations thereof). CARs are also known as artificial T cell receptors, chimeric T cell receptors or chimeric immunoreceptors. CARs can be specific for one antigen or multiple antigens, depending on the configuration of the antigen binding domain.

[0075] A CAR comprises an antigen binding domain that specifically binds a target, e.g., a polypeptide, expressed on the surface of a cell to be targeted for a T cell response, onto a construct including a transmembrane domain and intracellular domain(s) of a T cell receptor molecule. A bispecific CAR comprises two antigen-binding domains, a transmembrane domain and intracellular domain(s). In some embodiments, the antigen binding domains of a bispecific CAR each comprise a different antibody and / or ligand that each specifically bind to different antigens expressed on a cell to be targeted for a T cell response.

[0076] As used herein, a "CAR-T cell" or "CAR-T" refers to a T cell that expresses a CAR or a bispecific CAR. When expressed in a T cell, CARs and bispecific CARs have the ability to redirect T cell specificity and reactivity toward a selected target in a non-MHC-restricted manner, exploiting the antigen-binding properties of monoclonal antibodies. The non-MHC- restricted antigen recognition gives T cells expressing CARs or bispecific CARs the ability to recognize an antigen independent of antigen processing, thus bypassing a major mechanism of tumor escape.

[0077] As can be determined by those of skill in the art, various functionally similar or equivalent components of these bispecific CARs can be swapped or substituted with one another, as well as other similar or functionally equivalent components known in the art or listed herein.

[0078] Any cell-surface moiety can be targeted by a bispecific CAR. Often, the target will be a cell-surface polypeptide that may be differentially or preferentially expressed on a cell that one wishes to target for a T cell response. In some embodiments, the antigen-binding domains binds to the cancer-associated antigen mesothelin and / or the cancer antigen mucin 16 (MUC16), respectively. Mesothelin and MUC16 may be used as a marker for cells associated with various cancers (e.g., over-expressed in various cancers), and bispecific CARs and CAR-T cells that bind to mesothelin or a portion thereof and / or MUC16 or a portion thereof may be used to treat subjects having, e.g., cancers associated with mesothelin expression (mesothelin-expressing cancers). Bispecific CARs and CAR-T cells that target both mesothelin and MUC16 can overcome the heterogeneity associated with solid tumors and retain cytotoxicity when mesothelin or MUC16 expression decreases or is lost on tumor cells.

[0079] Antigen-Binding Domain

[0080] As used herein, the term "antigen-binding domain" refers to a polypeptide found on the outside of the cell that is sufficient to facilitate binding to a target. In some embodiments, the CARs described herein comprise an antigen-binding domain. The antigen-binding domain will specifically bind to its binding partner, i.e., the target. As non-limiting examples, the antigenbinding domain can include an antigen domain of an antibody, or a ligand, which recognizes and binds with a cognate binding partner protein. In this context, a ligand is a molecule that binds specifically to a portion of a protein and / or receptor. The cognate binding partner of a ligand useful in the methods and compositions described herein can generally be found on the surface of a cell. Ligand: cognate partner binding can result in the alteration of the ligand-bearing receptor, or activate a physiological response, for example, the activation of a signaling pathway. In some embodiments, the ligand can be non-native to the genome. In some embodiments, the ligand has a conserved function across at least two species.

[0081] Any cell-surface moiety can be targeted by a bispecific CAR (e.g., the antigen-binding domain(s) of the bispecific CAR). In some embodiments, the target is a mesothelin-expressing and / or a MUC16-expressing cancer.

[0082] In some embodiments, the disclosure provides a bispecific chimeric antigen receptor (CAR) comprising a first antigen-binding site that binds to mesothelin and a second antigenbinding site that binds to mucin 16 (MUC16).

[0083] In some embodiments, the first antigen-binding domain comprises a mesothelin-binding antibody. In some embodiments, the mesothelin-binding antibody is a mesothelin antibody fragment. In some embodiments, the mesothelin-binding antibody fragment is an antigenbinding fragment (Fab), a Fab’, a F(ab’)2, a fragment variable (fv) or a single chain fragment variable (scFv). In some embodiments, the mesothelin-binding antibody comprises (a) a heavy chain variable domain (VH) comprising three complementarity determining regions CDR-H1, CDR-H2, and CDR-H3, wherein the CDR-H1 comprises an amino acid sequence of SEQ ID NO: 3, the CDR-H2 comprises an amino acid sequence of SEQ ID NO: 4, and the CDR-H3 comprises an amino acid sequence of SEQ ID NO: 5; and (b) a light chain variable domain (VL) comprising three complementarity determining regions CDR-L1, CDR-L2, and CDR-L3, wherein the CDR-L1 comprises an amino acid sequence of SEQ ID NO: 6, the CDR-L2 comprises an amino acid sequence of SEQ ID NO: 7, and the CDR-L3 comprises an amino acid sequence of SEQ ID NO: 8. In some embodiments, the mesothelin-binding antibody comprises a VH comprising an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 1 and the VL comprises an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 2. In some embodiments, the mesothelin-binding antibody comprises a VH comprising an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 1 and the VL comprises an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 2, and the CDR domains of SEQ ID NO: 1 and SEQ ID NO: 2 are not changed. In some embodiments, the mesothelin- binding antibody comprises a VH comprising an amino acid sequence having at least 90% identity to SEQ ID NO: 1 and the VL comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 2. In some embodiments, the mesothelin antibody comprises a VH comprising an amino acid sequence of SEQ ID NO: 1 and the VL comprises an amino acid sequence of SEQ ID NO: 2. In some embodiments, the mesothelin-binding antibody comprises an amino acid sequence of SEQ ID NO: 17 between the VH and the VL. In some embodiments, the mesothelin-binding antibody fragment comprises an amino acid sequence of SEQ ID NO: 9.

[0084] In some embodiments, the mesothelin-binding antibody comprises (a) a heavy chain variable domain (VH) comprising three complementarity determining regions CDR-H1, CDR- H2, and CDR-H3, wherein the CDR-H1 comprises an amino acid sequence of SEQ ID NO: 34, the CDR-H2 comprises an amino acid sequence of SEQ ID NO: 35, and the CDR-H3 comprises an amino acid sequence of SEQ ID NO: 36; and (b) a light chain variable domain (VL) comprising three complementarity determining regions CDR-L1, CDR-L2, and CDR-L3, wherein the CDR-L1 comprises an amino acid sequence of SEQ ID NO: 37, the CDR-L2 comprises an amino acid sequence of SEQ ID NO: 38, and the CDR-L3 comprises an amino acid sequence of SEQ ID NO: 39. In some embodiments, the mesothelin-binding antibody comprises a VH comprising an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 32 and the VL comprises an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 33. In some embodiments, the mesothelin-binding antibody comprises a VH comprising an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 32 and the VL comprises an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 33 and the CDR domains of SEQ ID NO: 38 and SEQ ID NO: 39 are not changed. In some embodiments, the mesothelin-binding antibody comprises a VH comprising an amino acid sequence having at least 90% identity to SEQ ID NO: 32 and the VL comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 33. In some embodiments, the mesothelin antibody comprises a VH comprising an amino acid sequence of SEQ ID NO: 32 and the VL comprises an amino acid sequence of SEQ ID NO: 33.

[0085] In some embodiments, the second antigen-binding domain comprises a MUC16-binding antibody. In some embodiments, the MUC16-binding antibody binds to the ectodomain of MUC16. In some embodiments, the MUC16-binding antibody is a MUC16-binding antibody fragment. In some embodiments, the MUC16-binding antibody fragment is an antigen-binding fragment (Fab), a Fab’, or a F(ab’)2, a fragment variable (fv), or a single chain variable fragment (scFv). In some embodiments, the MUC16-binding antibody comprises (a) a heavy chain variable domain (VH) comprising three complementarity determining regions CDR-H1, CDR- H2, and CDR-H3, wherein the CDR-H1 comprises an amino acid sequence of SEQ ID NO: 12, the CDR-H2 comprises an amino acid sequence of SEQ ID NO: 13, and the CDR-H3 comprises an amino acid sequence of SEQ ID NO: 14; and (b) a light chain variable domain (VL) comprising three complementarity determining regions CDR-L1, CDR-L2, and CDR-L3, wherein the CDR-L1 comprises an amino acid sequence of SEQ ID NO: 15, the CDR-L2 comprises an amino acid sequence of SEQ ID NO: 16, and the CDR-L3 comprises an amino acid sequence of SEQ ID NO: 17. In some embodiments, the MUC16-binding antibody comprises a VH comprising an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 10 and the VL comprises an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 11. In some embodiments, the MUC16-binding antibody comprises a VH comprising an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 10 and the VL comprises an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to SEQ ID NO: 11 and the CDR domains of SEQ ID NO: 10 and SEQ ID NO: 11 are not changed. In some embodiments, the MUC16-binding antibody comprises a VH comprising an amino acid sequence having at least 90% identity to SEQ ID NO: 10 and the VL comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 11. In some embodiments, the MUC16- binding antibody comprises a VH comprising an amino acid sequence of SEQ ID NO: 10 and the VL comprises an amino acid sequence of SEQ ID NO: 11. In some embodiments the MUC16-binding antibody comprises an peptide linker of SEQ ID NO: 21 between the VH and the VL. In some embodiments, the MUC16-binding antibody fragment comprises an amino acid sequence of SEQ ID NO: 18.

[0086] In some embodiments, the bispecific CAR comprises an amino acid sequence having at least 85% (e.g., at least 85%, at least 90%, at least 95%, at least 98%, or at least 99%) identity to any one of SEQ ID NOs: 26-31. In some embodiments, a bispecific CAR comprises an amino acid sequence having at least 90% identity to any one of SEQ ID NOs: 26-31. In some embodiments, a bispecific CAR comprises an amino acid sequence of any one of SEQ ID NOs: 26-31. In some embodiments, a bispecific CAR comprises an amino acid sequence of SEQ ID NO: 26. In some embodiments, a bispecific CAR comprises an amino acid sequence of SEQ ID NO: 27. In some embodiments, a bispecific CAR comprises an amino acid sequence of SEQ ID NO: 28. In some embodiments, a bispecific CAR comprises an amino acid sequence of SEQ ID NO: 29. In some embodiments, a bispecific CAR comprises an amino acid sequence of SEQ ID NO: 30. In some embodiments, a bispecific CAR comprises an amino acid sequence of SEQ ID NO: 31.

[0087] In some embodiments, a bispecific CAR comprises, from N-terminal to C-terminal, (i) a first antigen-binding domain that binds to mesothelin, (ii) a peptide linker, and (iii) a second antigen-binding domain that binds to MUC16. In some embodiments, a bispecific CAR comprises, from N-terminal to C-terminal, (i) a first antigen-binding domain that binds to MUC16, (ii) a peptide linker, and (iii) a second antigen-binding domain that binds to mesothelin. In some embodiments, a bispecific CAR further comprises a transmembrane domain, a costimulatory domain, and an intracellular signaling domain. In some embodiments, a bispecific CAR comprises a CD8 hinge / transmembrane domain, a 4-1BB co-stimulatory domain, and a CD3(^ intracellular signaling domain.

[0088] Hinge and Transmembrane Domains

[0089] In some embodiments, a CAR or a bispecific CAR further comprises a transmembrane domain, or a hinge / transmembrane domain, which joins the antigen binding domain(s) to the intracellular signaling domain. The binding domain of a bispecific CAR is, in some embodiments, followed by one or more "hinge domains," which plays a role in positioning the antigen binding domain(s) away from the effector cell surface to enable proper cell / cell contact, antigen-binding (by the antigen binding domain(s)) and activation. A bispecific CAR may include one or more hinge domains between the binding domain and the transmembrane domain (TM). The hinge domain may be derived either from a natural, synthetic, semi -synthetic, or recombinant source. The hinge domain may include the amino acid sequence of a naturally occurring immunoglobulin hinge region or an altered immunoglobulin hinge region. Illustrative hinge domains suitable for use in the bispecific CARs described herein include the hinge region derived from the extracellular regions of type 1 membrane proteins such as CD8 (e.g., CD8a), CD4, CD28, 4- IBB, and CD7, which may be wild-type hinge regions from these molecules or may be altered. In some embodiments, a bispecific CAR comprises a polynucleotide encoding CD8a hinge / transmembrane domain. In some embodiments, a bispecific CAR comprises a polynucleotide encoding a 4- IBB intracellular domain.

[0090] In some embodiments, the hinge region is derived from the hinge region of an immunoglobulin like protein (e.g., IgA, IgD, IgE, IgG, or IgM), CD28, or CD8. In some embodiments, the hinge domain includes a CD8a hinge region.

[0091] As used herein, "transmembrane domain" (TM domain) refers to the portion of a bispecific CAR that fuses the antigen binding domain, in some embodiments via a hinge domain, to the intracellular portion (e.g., the co-stimulatory domain and intracellular signaling domain) and anchors the CAR to the plasma membrane of the immune effector cell. The transmembrane domain is a generally hydrophobic region of a bispecific CAR, which crosses the plasma membrane of a cell. The TM domain can be the transmembrane region or fragment thereof of a transmembrane protein (for example a Type I transmembrane protein or other transmembrane protein), an artificial hydrophobic sequence, or a combination thereof. While specific examples are provided herein and used herein, other transmembrane domains will be apparent to those of skill in the art and can be used in connection with alternate embodiments of the technology. A selected transmembrane region or fragment thereof would preferably not interfere with the intended function of a bispecific CAR.

[0092] As used in relation to a transmembrane domain of a protein or polypeptide, "fragment thereof refers to a portion of a transmembrane domain that is sufficient to anchor or attach a protein to a cell surface.

[0093] In some embodiments, the transmembrane domain or fragment thereof of a bispecific CAR described herein includes a transmembrane domain selected from the transmembrane domain of an alpha (a), beta (P) or zeta (Q chain of a T cell receptor, CD2, CD28, CD3 epsilon (e), CD45, CD4, CD5, CD8, CD9, CD16, CD22, CD33, CD37, CD64, CD80, CD86, CD134, CD137, CD154, KIRDS2, 0X40, CD2, CD27, LFA-1 (CDl la, CD18), ICOS (CD278), 4-1BB (CD137), 4-1BBL, GITR, CD40, BAFFR, HVEM (LIGHTR), SLAMF7, NKp80 (KLRFI), CD 160, CD 19, IL2RP, IL2Ry, IL7Ra, ITGA1, VLA1, CD49a, ITGA4, IA4, CD49D, ITGA6, VLA-6, CD49f, ITGAD, CDl ld, ITGAE, CD103, ITGAL, CDl la, LFA-1, ITGAM, CDl lb, ITGAX, CDl lc, ITGB1, CD29, ITGB2, CD18, LFA-1, ITGB7, TNFR2, DNAM1 (CD226), SLAMF4 (CD244, 2B4), CD84, CD96 (Tactile), CEACAM1, CRT AM, Ly9 (CD229), CD 160 (BY55), PSGL1, CD100 (SEMA4D), SLAMF6 (NTB-A, Lyl08), SLAM (SLAMF1, CD150, IPO-3), BLAME (SLAMF8), SELPLG (CD162), LTBR, PAG / Cbp, NKp44, NKp30, NKp46, NKG2D, and / or NKG2C.

[0094] As used herein, a “hinge / transmembrane domain” refers to a domain including both a hinge domain and a transmembrane domain. For example, a hinge / transmembrane domain can be derived from the hinge / transmembrane domain of CD8, CD28, CD7, or 4- IBB. In some embodiments, the hinge / transmembrane domain of a bispecific CAR or fragment thereof is derived from or includes the hinge / transmembrane domain of CD8 (e.g., SEQ ID NO: 22, or variants thereof). CD8 is an antigen preferentially found on the cell surface of cytotoxic T lymphocytes. CD8 mediates cell-cell interactions within the immune system, and acts as a T cell co-receptor. CD8 consists of an alpha (CD8alpha, CD8a, or CD8a) and beta (CD8beta, CD8b, or CD8P) chain. CD8a sequences are known for a number of species, e.g., human CD8a, (NCBI Gene ID: 925) polypeptide (e.g., NCBI Ref Seq NP 001139345.1) and mRNA (e.g., NCBI Ref Seq NM_ 000002.12). CD8 can refer to human CD8, including naturally occurring variants, molecules, and alleles thereof. In some embodiments of any of the aspects, e.g., in veterinary applications, CD8 can refer to the CD8 of, e.g., dog, cat, cow, horse, pig, and the like. Homologs and / or orthologs of human CD8 are readily identified for such species by one of skill in the art, e.g., using the NCBI ortholog search function or searching available sequence data for a given species for sequence similar to a reference CD8 sequence.

[0095] In some embodiments, the CD8 hinge and transmembrane sequence corresponds to the amino acid sequence of SEQ ID NO: 22; or includes a sequence with at least 75%, at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or at least 100% sequence identity to the sequence of SEQ ID NO: 22.

[0096] Co-stimulatory Domains

[0097] Each CAR and bispecific CAR described herein optionally includes the intracellular domain of one or more co-stimulatory molecule or co-stimulatory domain. As used herein, the term "co-stimulatory domain" refers to an intracellular signaling domain of a co-stimulatory molecule. Co-stimulatory molecules are cell surface molecules other than antigen receptors or Fe receptors that provide a second signal required for efficient activation and function of T lymphocytes upon binding to antigen. The co-stimulatory domain can be, for example, the co- stimulatory domain of 4- IBB, CD27, CD28, or 0X40. Additional illustrative examples of such co-stimulatory molecules include CARD11, CD2, CD7, CD27, CD28, CD30, CD40, CD54 (ICAM), CD83, CD134 (0X40), CD137 (4-1BB), CD150 (SLAMF1), CD152 (CTLA4), CD223 (LAG3), CD270 (HVEM), CD273 (PD-L2), CD274 (PD-L1), CD278 (ICOS), DAP10, LAT, NKD2C SLP76, TRIM, and ZAP70. In some embodiments, the co-stimulatory domain is the intracellular domain of 4-1BB. 4-1BB (CD137; TNFRS9) is an activation induced co- stimulatory molecule and is an important regulator of immune responses.

[0098] 4-1BB is a membrane receptor protein, also known as CD137, which is a member of the tumor necrosis factor (TNF) receptor superfamily. 4- IBB is expressed on activated T lymphocytes. 4-1BB sequences are known for a number of species, e.g., human 4-1 BB, also known as TNFRSF9 (NCBI Gene 25 ID: 3604) and mRNA (NCBI Reference Sequence: NM 001561.5). 4-1BB can refer to human 4-1BB, including naturally occurring variants, molecules, and alleles thereof. In some embodiments of any of the aspects, e.g., in veterinary applications, 4-1BB can refer to the 4-1BB of, e.g., dog, cat, cow, horse, pig, and the like. Homologs and / or orthologs of human 4- IBB are readily identified for such species by one of skill in the art, e.g., using the NCBI ortholog search function or searching available sequence data for a given species for sequence similar to a reference 4-1 BB sequence. In some embodiments, the co-stimulatory domain comprises a 4- IBB co-stimulatory domain comprising an amino acid sequence of SEQ ID NO: 23, or includes a sequence with at least 75%, at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or at least 100% sequence identity to the sequence of SEQ ID NO: 23.

[0099] Intracellular Signaling Domains

[0100] In some embodiments, CAR or the bispecific CAR comprise a polynucleotide encoding a CD3^ intracellular signaling domain.

[0101] The properties of the intracellular signaling domain(s) of a bispecific CAR can vary as known in the art and as disclosed herein, but the chimeric target / antigen binding domains(s) render the receptor sensitive to signaling activation when the chimeric target / antigen binding domain(s) binds the target / antigen on the surface of a targeted cell.

[0102] With respect to intracellular signaling domains, so-called "first-generation" CARs include those that solely provide CD3<^ signals upon antigen-binding by the antigen binding domain(s). So-called "second-generation" CARs include those that provide both co-stimulation (e.g., CD28 or CD137) and activation (CD3Q domains, and so-called "third-generation" CARs include those that provide multiple co-stimulatory (e.g., CD28 and CD137) domains and activation domains (e.g., CD3Q. In some embodiments, a bispecific CAR is selected to have high affinity or avidity for the targets / antigens - for example, antigen binding domains will generally have higher affinity and / or avidity for the target antigen than would a naturally occurring T cell receptor. This property, combined with the high specificity one can select for an antibody provides highly specific T cell targeting by CAR-T cells.

[0103] Bispecific CARs as described herein include an intracellular signaling domain. An "intracellular signaling domain" refers to the part of a bispecific CAR that participates in transducing the message of effective bispecific CAR binding to a target antigen or target antigens into the interior of the immune effector cell to elicit effector cell function, e.g., activation, cytokine production, proliferation and cytotoxic activity, including the release of cytotoxic factors to the bispecific CAR-bound target cell, or other cellular responses elicited following antigen-binding to the antigen-binding domain(s) of the bispecific CAR. In various examples, the intracellular signaling domain is from CD3<^ (see, e.g., below). Additional nonlimiting examples of immunoreceptor tyrosine-based activation motif (ITAM)-containing intracellular signaling domains that are of particular use in the technology include those derived from 4-1BB (e.g., as described above), CD28, CD27, TCR zeta (Q, FcR gamma (y), FcR beta (P), CD3 gamma (y), CD3 theta (0), CD3 sigma (G), CD3 eta (r|), CD3 epsilon (s), CD3 zeta (Q, CD22, CD79a, CD79b, and CD66d.

[0104] CD3 is a T cell co-receptor that facilitates T lymphocyte activation when simultaneously engaged with the appropriate co-stimulation (e.g., binding of a co-stimulatory molecule). A CD3 complex consists of 4 distinct chains; mammalian CD3 consists of a CD3y chain, a CD36 chain, and two CD3s chains.

[0105] These chains associate with a molecule known as the T cell receptor (TCR) and the CD3^ to generate an activation signal in T lymphocytes. A complete TCR complex includes a TCR, CD3^, and the complete CD3 complex.

[0106] In some embodiments, a bispecific CAR described herein includes an intracellular signaling domain that includes an Immunoreceptor Tyrosine-based Activation Motif or ITAM from CD3^, including variants of CD3^ such as ITAM-mutated CD3^, CD3r|, or CD30. In some embodiments of any aspect, the ITAM includes three motifs of ITAM of CD3^ (ITAM3). In some embodiments of any aspect, the three motifs of ITAM of CD3^ are not mutated and, therefore, include native or wild-type sequences. In some embodiments, the CD3^ sequence includes the sequence of a CD3^ as set forth in the sequences provided herein, e.g., a CD3^ sequence of SEQ ID NO: 24 or variants thereof.

[0107] For example, a bispecific CAR described herein includes the intracellular signaling domain of CD3^. In some embodiments, the CD3^ intracellular signaling domain corresponds to an amino acid sequence of SEQ ID NO: 24 or includes a sequence with at least 75%, at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or at least 100% sequence identity to a sequence of SEQ ID NO: 24.

[0108] In some embodiments, the intracellular domain is the intracellular domain of a 4-1BB. In some embodiments, the 4- IBB intracellular domain corresponds to an amino acid sequence selected from SEQ ID NO: 23 or includes at least 75%, at least 80%, at least 85%, 35 at least 90%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or at least 100% sequence identity to a sequence selected from SEQ ID NO: 23. Individual CAR and bispecific CAR, and other construct components as described herein can be used with one another and swapped in and out of various constructs described herein, as can be determined by those of skill in the art. Each of these components can include or consist of any of the corresponding sequences set forth herein, or variants thereof.

[0109] A more detailed description of CARs and CAR-T cells can be found in Maus et al., Blood 123:2624-2635, 2014; Reardon et al., Neuro-Oncology 16: 1441-1458, 2014; Hoyos et al., Haematologica 97: 1622, 2012; Byrd et al., J. Clin. Oncol. 32:3039-3047, 2014; Maher et al., Cancer Res 69:4559-4562, 2009; and Tamada et al., Clin. Cancer Res. 18:6436-6445, 2012.

[0110] Signal Peptide / Leader Sequence

[0111] In some embodiments, a CAR or bispecific CAR as described herein includes a signal peptide. Signal peptides, also called leader sequences, can be derived from any protein that has an extracellular domain or is secreted. A bispecific CAR as described herein may include any leader sequence known in the art. In some embodiments, a bispecific CAR includes a CD8 leader sequence, e.g., a CD8 leader sequence comprising the amino acid sequence of SEQ ID NO: 25 or including an amino acid sequence having at least 75%, at least 80%, at least 85%, at least 90%, at least 91 %, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or at least 100% sequence identity to the sequence of SEQ ID NO: 25.

[0112] In further embodiments, a bispecific CAR described herein may optionally exclude one of the signal peptides described herein, e.g., a CD8 signal peptide of SEQ ID NO: 25or an IgK signal peptide of SEQ ID NO: 40.

[0113] Linkers

[0114] In some embodiments, a bispecific CAR further includes a peptide linker. As used herein, "linker" refers to an oligo- or polypeptide region from about 2 to 100 amino acids in length, which links together any of the domains / regions of a bispecific CAR as described herein. In some embodiment, linkers can include or be composed of flexible residues such as glycine and serine so that the adjacent protein domains are free to move relative to one another. Linker sequences may be from 2 to 100 amino acids, 5 to 50 amino acids, 10 to 15 amino acids, 15 to 20 amino acids, or 18 to 20 amino acids in length, and include any suitable linkers known in the art. For instance, linker sequences may include, but are not limited to, gly cine / serine linkers, e.g., SEQ ID NOs: 19-21; as well as linker sequences with added functionalities, e.g., an epitope tag or an encoding sequence containing Cre-Lox recombination site as described by Sblattero et al., Nat. Biotechnol. 18(l):75-80, 2000. Longer linkers may be used when it is desirable to ensure that two adjacent domains do not sterically interfere with one another. Peptide linkers can be used to separate one component of a bispecific CAR from another, such as a first antigenbinding domain from a second antigen-binding domain. In some embodiments, a bispecific CAR provided herein comprises a peptide linker between the first antigen-binding domain that binds to mesothelin and the second antigen-binding domain that binds to MUC16. In some embodiments, a peptide linker comprises the amino acid sequence of SEQ ID NO: 19. In some embodiments, a peptide linker comprises the amino acid sequence of SEQ ID NO: 20. In some embodiments, a peptide linker comprises the amino acid sequence of SEQ ID NO: 21.

[0115] Polynucleotides, Plasmids, and Vectors

[0116] In some aspects, this disclosure describes a polynucleotide encoding any one of the CARs or bispecific CARs described herein. The term "polynucleotide" is used herein interchangeably with "nucleic acid molecule" to indicate a polymer of nucleosides. Typically, a polynucleotide is composed of nucleosides that are naturally found in DNA or RNA (e.g., adenosine, thymidine, guanosine, cytidine, uridine, deoxyadenosine, deoxythymidine, deoxy guanosine, and deoxy cytidine) joined by phosphodiester bonds. However, the term encompasses molecules comprising nucleosides or nucleoside analogs containing chemically or biologically modified bases, modified backbones, etc., whether or not found in naturally occurring nucleic acids, and such molecules may be preferred for certain applications. Where this application refers to a polynucleotide it is understood that both DNA, RNA, and in each case both single- and double-stranded forms (and complements of each single-stranded molecule) are provided. "Polynucleotide sequence" as used herein can refer to the polynucleotide material itself and / or to the sequence information (i.e., the succession of letters used as abbreviations for bases) that biochemically characterizes a specific nucleic acid. In some embodiments, the nucleic acid molecule is a heterologous nucleic acid molecule. As used herein the term, “heterologous nucleic acid molecule” refers to a nucleic acid molecule that does not naturally exist within a given cell or a nucleic acid sequence that has been engineered into a cell. For example, a heterologous nucleic acid molecule may be a nucleic acid molecule encoding a gene that is engineered into a cell (e.g., via a plasmid, vector or some other method). A polynucleotide sequence presented herein is presented in a 5' to 3' direction unless otherwise indicated.

[0117] In some embodiments, the polynucleotide is operably linked to a promoter. As used herein, the term "operably linked" refers to a first polynucleotide molecule, such as a promoter, connected with a second transcribable polynucleotide molecule, such as a CAR or a bispecific CAR, where the polynucleotide molecules are so arranged that the promoter can direct a RNA polymerase to transcribe the second polynucleotide molecule. The two polynucleotide molecules may or may not be part of a single contiguous polynucleotide molecule and may or may not be adjacent. For example, a promoter is operably linked to a gene of interest if the promoter regulates or mediates transcription of the gene of interest in a cell.

[0118] In some embodiments, the promoter is a constitutively active promoter. In some embodiments, the U6 promoter is from a non-human species. In some embodiments, the promoter is selected from the group consisting of a CMV promoter, an EFla promoter, an EF la-short promoter, a CAG promoter, a PGK promoter, Hl promoter, or a U6 promoter. In some embodiments, the U6 promoter is from a human U6 promoter. In some embodiments, the U6 promoter is from cow, mice, rat, pig, yeast, dog, cat, drosophila, or C. elegans. In some embodiments, the promoter is a Hl promoter. In some embodiments, the promoter is a tissuespecific promoter (e.g., the HP1, CD14, CD43, CD45, C68, elastase, endoglin, fibronectin, Fit, GFAP, GPIIb, ICAM-2, mIFN-beta, Mb, NphsI, OG-2, SP-B, SYN1, or WASP gene promoter). In some embodiments, the promoter is an inducible promoter (e.g., a tet or lac promoter).

[0119] In some embodiments, a nucleic acid encoding a polypeptide as described herein (e.g., a bispecific CAR) is comprised by a plasmid. The term “plasmid” may refer to a circular piece of DNA the comprises an origin of replication. In some embodiments, the plasmid comprises a prokaryotic origin of replication. In some embodiments, the plasmid comprises a bacterial origin of replication. In some embodiments, the plasmid comprises a eukaryotic origin of replication. In some embodiments, the plasmid comprises a mammalian origin of replication. In some embodiments, the plasmid comprises a prokaryotic and eukaryotic origin of replication. In some embodiments, the plasmid comprises an origin of replication that is active in a cell which the plasmid is located. In some embodiments, the plasmid is a lentiviral plasmid (e.g., a second generation lentiviral plasmid).

[0120] In some embodiments, a nucleic acid encoding a polypeptide as described herein (e.g., a bispecific CAR) is comprised by a vector. The term "vector," as used herein, refers to a nucleic acid construct designed for delivery to a host cell or for transfer between different host cells. As used herein, a vector can be viral or non-viral. The term "vector" encompasses any genetic element that is capable of replication when associated with the proper control elements and that can transfer gene sequences to cells. A vector can include, but is not limited to, a cloning vector, an expression vector, a plasmid, phage, transposon, cosmid, artificial chromosome, virus, virion, etc.

[0121] As used herein, the term "expression vector" may refer to a vector that directs expression of an RNA or polypeptide from sequences linked to transcriptional regulatory sequences on the vector. The sequences expressed will often, but not necessarily, be heterologous to the cell. An expression vector may comprise additional elements, for example, the expression vector may have two replication systems, thus allowing it to be maintained in two organisms, for example, in human cells for expression and in a prokaryotic host for cloning and amplification. The term "expression" refers to the cellular processes involved in producing RNA and proteins and as appropriate, secreting proteins, including where applicable, but not limited to, for example, transcription, transcript processing, translation and protein folding, modification and processing. "Expression products" include RNA transcribed from a gene, and polypeptides obtained by translation of mRNA transcribed from a gene.

[0122] As used herein, the term "viral vector" may refer to a nucleic acid vector construct that includes at least one element of viral origin and has the capacity to be packaged into a viral vector particle. The viral vector can contain a nucleic acid encoding a polypeptide as described herein in place of non-essential viral genes. The vector and / or particle may be utilized for the purpose of transferring nucleic acids into cells either in vitro or in vivo. Numerous forms of viral vectors are known in the art. In some embodiments, the viral vector is an adeno-associated viral, adenoviral, lentiviral, or a retroviral vector. In some embodiments, the viral vector is a lentiviral vector. In some embodiments, the lentiviral vector is a second generation lentiviral vector.

[0123] By "recombinant vector" may be a vector that includes a heterologous nucleic acid sequence or "transgene" that is capable of expression in vivo. It should be understood that the vectors described herein can, in some embodiments, be combined with other suitable compositions and therapies. In some embodiments, the vector is episomal. The use of a suitable episomal vector provides a means of maintaining the nucleotide of interest in the subject in high copy number extra-chromosomal DNA thereby eliminating potential effects of chromosomal integration.

[0124] In some embodiments, a polypeptide, polynucleotide, plasmid and or / vector as described herein optionally further comprises a reporter molecule, e.g., to determine if the vector is properly expressed in a cell. In some embodiments, the reporter molecule may be a fluorescent protein (e.g., GFP, YFP, RF, mCherry), antibody (e.g., CD34, tEGFR, tCD19, tCD20, tCD34, and tHer2), or a radioisotope. In some embodiments, the reporter molecule is hygromycin phosphotransferase (hph) that can be imaged alone or in combination with a substrate or chemical (for example 9-[4-[18F]fluoro-3-(hydroxymethyl)butyl]guanine ([18F]FHBG)).

[0125] In some embodiments, GFP and mCherry may be used as fluorescent tags for imaging a bispecific CAR expressed on a T cell (e.g., a CAR-T cell). It is expected that essentially any fluorescent protein known in the art can be used as a fluorescent tag for this purpose. For clinical applications, the CAR need not include a fluorescent tag or fluorescent protein. In each instance of particular constructs provided herein, therefore, any markers present in the constructs can be removed. The invention includes the constructs with or without the markers. Accordingly, when a specific construct is referenced herein, it can be considered with or without any markers or tags (including, e.g., histidine tags) as being included within the invention.

[0126] Cells

[0127] In some embodiments, this disclosure provides a cell comprising a CAR or bispecific CAR construct (e.g., a polynucleotide encoding the CAR or bispecific CAR). In some embodiments, the cell is a prokaryotic cell (e.g., a cell for production of a plasmid encoding a bispecific CAR). In some embodiments, the cell is a eukaryotic cell. In some embodiments, the cell is a mammalian cell (e.g., an immune cell) comprising any of the bispecific CAR constructs described herein. The mammalian cell can be of human, primate, hamster, rabbit, rodent, cow, pig, sheep, horse, goat, dog or cat origin, but any other mammalian cell may be used. In a preferred embodiment of any aspect, the mammalian cell is human.

[0128] In some embodiments, the disclosure provides a cell comprising a bispecific CAR, a polynucleotide, or a vector provided herein. In some embodiments of any aspect, the mammalian cell is an immune cell. As used herein, "immune cell" refers to a cell that plays a role in the immune response. Immune cells are of hematopoietic origin, and include lymphocytes, such as B cells and T cells; natural killer cells; myeloid cells, such as monocytes, macrophages, eosinophils, mast cells, basophils, and granulocytes. In some embodiments, the immune cell is a T cell.

[0129] In some embodiments, the immune cell is an autologous cell. In some embodiments, the immune cell is obtained from an individual having or diagnosed as having cancer. In some embodiments, the immune cell is allogenic to the subject. In some embodiments, the immune cell is produced from a stem cell (e.g., an induced pluripotent stem cell or an embryonic stem cell). In some embodiments, a mammalian cell, e.g., a T cell, can be engineered to include any of the bispecific CAR constructs, as described herein. T cells can be obtained from a subject using standard techniques known in the field. For example, T cells can be isolated from peripheral blood taken from a donor or patient. T cells can be isolated from a mammal. Preferably, T cells are isolated from a human.

[0130] In some aspects, this disclosure describes a CAR-T cell comprising any of the bispecific CAR constructs disclosed herein. In some embodiments, the CAR-T cell is generated from T- cells extracted from a subject (e.g., the subject to whom the CAR-T cells will be administered). In some embodiments, the CAR-T cells are allogenic CAR-T cells. In some embodiments, the CAR-T cell comprises a bispecific CAR construct comprising an anti-mesothelin and anti- MUC16 CAR. In some embodiments, the CAR-T cell comprises a bispecific CAR construct as described herein. In some embodiments, the CAR-T cell comprises a construct encoding any one of SEQ ID NOs: 26-31.

[0131] Methods of Treatment

[0132] In some aspects, this disclosure provides a method of treating a subject having a mesothelin-expressing cancer (as described herein), the method comprising administering a cell (e.g., a CAR-T cell) expressing a bispecific CAR as described herein to the subject. In some embodiments, the disclosure provides a method of treating a subject having a MUC16- expressing cancer (as described herein), the method comprising administering a cell (e.g., a CAR-T cell) expressing a bispecific CAR as described herein to the subject. In some embodiments, the disclosure provides a method of treating a subject having a mesothelin- and MUC16-expressing cancer (as described herein), the method comprising administering a cell (e.g., a CAR-T cell) expressing a bispecific CAR as described herein to the subject. In some embodiments, the cancer is ovarian cancer. In some embodiments, the cancer is pancreatic cancer. In some embodiments, the method comprises administering a CAR-T cell described herein (e.g., a CAR-T cell expressing a bispecific CAR). In some embodiments, the method comprises administering a CAR-T cell comprising a bispecific CAR comprising an anti- mesothelin and anti- bispecific CAR. In some embodiments, the method comprises administering a CAR-T cell comprising a construct encoding any one of SEQ ID NOs: 26-31. In some embodiments, the method comprises administering a CAR-T cell in combination with an ADAM17 inhibitor (e.g., aderbasib, TAPI-0, TAPI-1, TAPI-2, GW280264X, marimastat, INCB3619, compound 22a (INCB9471), MEDI3622, apratastat, DPC-333, SCH 900567, or KP- 457) or an interleukin-2-inducible T cell kinase (ITK) inhibitor (e.g., ibrutinib, CPI-818, BMS 509774, or PRN694). In some embodiments, the cell or CAR-T cell is an autologous T cell. In some embodiments, the cell or CAR-T cell is an allogeneic T cell.

[0133] Cancer

[0134] In some embodiments, the methods described herein comprise treating a subject that has a mesothelin and / or MUC16-expressing cancer. "Cancer" as used herein can refer to a hyperproliferation of cells whose unique trait, loss of normal cellular control, results in unregulated growth, lack of differentiation, local tissue invasion, and metastasis. In some embodiments, the cancer expresses mesothelin (a mesothelin-expressing cancer) and / or mucin 16 (MUC16) (a MUC16-expressing cancer or a mesothelin- and MUC16-expressing cancer). In some embodiments, the mesothelin- and / or MUC16-expressing cancer is ovarian cancer. In some embodiments, the mesothelin- and / or MUC16-expressing cancer is pancreatic cancer. In some embodiments, the mesothelin- and / orMUC16-expressing cancer is ovarian cancer and / or pancreatic cancer. Solid tumors can be found in bones, muscles, or organs, and can be sarcomas or carcinomas. As used herein, the term "tumor" refers to an abnormal growth of cells or tissues, e.g., of malignant type or benign type.

[0135] Subject

[0136] In some embodiments, the methods described herein comprise treating a subject having cancer. As used herein, a “subject” means a human or animal. Usually, the animal is a vertebrate such as a primate, rodent, domestic animal or game animal. Primates include, for example, chimpanzees, cynomolgus monkeys, spider monkeys, and macaques, e.g., rhesus. Rodents include, for example, mice, rats, woodchucks, ferrets, rabbits and hamsters. Domestic and game animals include, for example, cows, horses, pigs, deer, bison, buffalo, feline species, e.g., domestic cat, canine species, e.g., dog, fox, wolf, avian species, e.g., chicken, emu, ostrich, and fish, e.g., trout, catfish and salmon. In some embodiments, the subject is a mammal, e.g., a primate, e.g., a human. The terms, “individual,” “patient,” and “subject” are used interchangeably herein. Preferably, the subject is a mammal. The mammal can be a human, non-human primate, mouse, rat, dog, cat, horse, or cow, but is not limited to these examples. Mammals other than humans can be advantageously used as subjects that represent animal models of disease, e.g., cancer. A subject can be male or female. A subject can be one who has been previously diagnosed with or identified as suffering from or having a condition in need of treatment (e.g., diagnosed with a mesothelin- and / or MUC16-expressing cancer) or one or more complications related to such a condition, and optionally, have already undergone treatment for the condition or the one or more complications related to the condition.

[0137] Alternatively, a subject can also be one who has not been previously diagnosed as having such condition or related complications. For example, a subject can be one who exhibits one or more risk factors for the condition or one or more complications related to the condition or a subject who does not exhibit risk factors.

[0138] A “subject in need” of treatment for a particular condition (e.g., a cancer described herein) can be a subject having that condition, diagnosed as having that condition, or at risk of developing that condition.

[0139] Pharmaceutical Compositions

[0140] In some embodiments, the methods described herein comprise administering to the subject a pharmaceutical composition comprising the CAR-T cells and / or a pharmaceutical composition comprising the ADAM17 inhibitor and / or the ITK inhibitor. As used herein, the term “pharmaceutical composition” refers to the active agent (e.g., a CAR-T cell expressing a bispecific CAR) in combination with a pharmaceutically acceptable carrier e.g., a carrier commonly used in the pharmaceutical industry.

[0141] The phrase “pharmaceutically acceptable carrier” is employed herein to refer to those compounds, materials, compositions, and / or dosage forms which are, within the scope of sound medical judgment, suitable for use in contact with the tissues of human beings and animals without excessive toxicity, irritation, allergic response, or other problem or complication, commensurate with a reasonable benefit / risk ratio. In some embodiments of any of the aspects, a pharmaceutically acceptable carrier can be a carrier other than water. In some embodiments of any of the aspects, a pharmaceutically acceptable carrier can be an artificial or engineered carrier, e.g., a carrier in which the active ingredient would not be found to occur in nature.

[0142] In one aspect of the technology, the technology described herein relates to a pharmaceutical composition including activated CAR-T cells comprising a CAR or a bispecific CAR described herein, and optionally a pharmaceutically acceptable carrier. The active ingredients of the pharmaceutical composition at a minimum include activated CAR-T cells comprising a CAR or bispecific CAR as described herein. In some embodiments, the active ingredients of the pharmaceutical composition consist essentially of activated CAR-T cells comprising a CAR or bispecific CAR as described herein. In some embodiments, the active ingredients of the pharmaceutical composition consist of activated CAR-T cells comprising a CAR or bispecific CAR as described herein. Pharmaceutically acceptable carriers for cell-based therapeutic formulation include saline and aqueous buffer solutions, Ringer’s solution, and serum component, such as serum albumin, HDL and LDL. The terms such as “excipient,” “carrier,” “pharmaceutically acceptable carrier”, “pharmaceutically acceptable excipient” or the like are used interchangeably herein.

[0143] In some embodiments, the pharmaceutical composition including activated CAR-T cells comprising a CAR or bispecific CAR as described herein can be a parenteral dose form. Since administration of parenteral dosage forms typically bypasses the patient’s natural defenses against contaminants, the components apart from the CAR-T cells themselves are preferably sterile or capable of being sterilized prior to administration to a patient. Examples of parenteral dosage forms include, but are not limited to, solutions ready for injection, dry products ready to be dissolved or suspended in a pharmaceutically acceptable vehicle for injection, suspensions ready for injection, and emulsions. Any of these can be added to the activated CAR-T cells preparation prior to administration. Suitable vehicles that can be used to provide parenteral dosage forms of activated CAR-T cells as disclosed within are well known to those skilled in the art. Examples include, without limitation: saline solution; glucose solution; aqueous vehicles including but not limited to, sodium chloride injection, Ringer’s injection, dextrose injection, dextrose and sodium chloride injection, and lactated Ringer’s injection; water-miscible vehicles such as, but not limited to, ethyl alcohol, polyethylene glycol, and propylene glycol; and nonaqueous vehicles such as, but not limited to, com oil, cottonseed oil, peanut oil, sesame oil, ethyl oleate, isopropyl myristate, and benzyl benzoate.

[0144] Dosage

[0145] In some embodiments, the activated CAR-T cells comprising a CAR or bispecific CAR described herein are administered as a monotherapy, i.e., another treatment for the condition is not concurrently administered to the subject. A pharmaceutical composition including the T cells described herein can generally be administered at a dosage of 104to 109cells / kg body weight, if necessary, T cell compositions can also be administered multiple times at these dosages. The cells can be administered by using infusion techniques that are commonly known in immunotherapy (see, e.g., Rosenberg et al., New Eng. J. Med. 30 319: 1676, 1988). Administration

[0146] In some embodiments, the methods described herein relate to treating a subject having a mesothelin-expressing cancer (e.g., as described herein), the method comprising administering to the subject a CAR-T cell, wherein the CAR-T cell comprises a bispecific CAR comprising a mesothelin-binding domain and a mucin 16 (MUC16)-binding domain. In some embodiments, the method comprises administering to the subject a CAR-T cell to the subject in combination with an ADAM17 inhibitor or an ITK inhibitor, wherein the CAR-T cell comprises a bispecific CAR comprising a mesothelin-binding domain and a MUC16-binding domain. The CAR-T cells comprising a bispecific CAR described herein include mammalian cells including any of the bispecific CAR described herein and any of the bispecific CARs described herein or known in the art, or a nucleic acid encoding any of the bispecific CARs described herein.

[0147] In some embodiments, the methods described herein include administering an effective amount of activated CAR-T cells comprising a bispecific CAR described herein to treat a subject having a mesothelin-expressing and / or a MUC16-expressing cancer. As used herein, “treating a subject having a mesothelin-expressing and / or a MUC16-expressing cancer” is ameliorating any condition or symptom associated with the mesothelin-expressing and / or MUC16-expressing cancer. As compared with an equivalent untreated control, such reduction is by at least 5%, 10%, 20%, 40%, 50%, 60%, 80%, 90%, 95%, 99% or more as measured by any standard technique. A variety of means for administering the compositions described herein to subjects are known to those of skill in the art. In some embodiments, the compositions described herein are administered systemically or locally. In a preferred embodiment, the compositions described herein are administered intravenously. In another embodiment, the compositions described herein are administered at the site of a tumor.

[0148] The term "effective amount" as used herein refers to the amount of activated CAR-T cells comprising a bispecific CAR. Described herein needed to treat at least one or more symptom of the mesothelin- and / or MUC16-expressing cancer and relates to a sufficient amount of the cell preparation or composition to provide the desired effect. The term “therapeutically effective amount” therefore refers to an amount of activated CAR-T cells comprising bispecific CAR described herein that is sufficient to provide a particular anti-condition effect when administered to a typical subject. An effective amount as used herein, in various contexts, would also include an amount sufficient to delay the development of a symptom of mesothelin- and / or MUC16-expressing cancer, alter the course of mesothelin- and / or MUC16-expressing cancer (for example but not limited to, slowing the progression of the mesothelin- and / or MUC16- expressing cancer), or reverse a symptom of a mesothelin- and / or MUC16-expressing cancer. Thus, it is not generally practicable to specify an exact “effective amount.” However, for any given case, an appropriate “effective amount” can be determined by one of ordinary skill in the art using only routine experimentation.

[0149] In some embodiments, the methods of treating a subject having a mesothelin- and / or MUC16-expressing cancer described herein comprises administering a cell comprising an anti- mesothelin and anti-MUC16 bispecific CAR via intravenous administration. In some embodiments, the method comprises administering a cell comprising an anti-mesothelin and anti-MUC16 bispecific CAR via intravenous administration in combination with an ADAMI 7 inhibitor. In some embodiments, the method comprises administering a cell comprising an anti- mesothelin and anti-MUC16 bispecific CAR via intravascular administration in combination with an ADAM17 inhibitor administered intraperitoneally. In some embodiments, the method comprises administering a cell comprising an anti-mesothelin and anti-MUC16 bispecific CAR via intravascular administration in combination with an ADAMI 7 inhibitor administered orally. In some embodiments, the method comprises administering a cell comprising an anti-mesothelin and anti-MUC16 bispecific CAR in combination with an ADAM17 inhibitor via intravascular administration.

[0150] In some embodiments, the method comprises administering a cell comprising an anti- mesothelin and anti-MUC16 bispecific CAR via intravascular administration in combination with an ITK inhibitor. In some embodiments, the method comprises administering a cell comprising an anti-mesothelin and anti-MUC16 bispecific CAR via intravascular administration in combination with an ITK inhibitor administered intraperitoneally. In some embodiments, the method comprises administering a cell comprising an anti-mesothelin and anti-MUC16 bispecific CAR via intravascular administration in combination with an ITK inhibitor administered orally. In some embodiments, the method comprises administering a cell comprising an anti-mesothelin and anti-MUC16 bispecific CAR in combination with an ITK inhibitor via intravascular administration.

[0151] Modes of Administration

[0152] Modes of administration (e.g., of the CAR-T cell comprising a bispecific CAR, the ADAM 17 inhibitor, or the ITK inhibitor) can include, for example intravenous (iv) injection or infusion. The compositions described herein can be administered to a patient transarterially, intratumorally, intranodally, intraperitoneally, intrathecally, intramedullary, or orally. In some embodiments, the compositions of CAR-T cells may be injected directly into a tumor, lymph node, or site of infection. In some embodiments, the compositions described herein are administered into a body cavity or body fluid (e.g., ascites, pleural fluid, peritoneal fluid, or cerebrospinal fluid).

[0153] In some embodiments, subjects may undergo leukapheresis, wherein leukocytes are collected, enriched, or depleted ex vivo to select and / or isolate the cells of interest, e.g., T cells. These T cell isolates can be expanded by contact with an artificial APC (aAPC), e.g., an aAPC expressing anti-CD28 and anti-CD3 CDRs, and treated such that one or more bispecific CAR constructs of the technology may be introduced, thereby creating a CAR-T cell.

[0154] Subjects in need thereof can subsequently undergo standard treatment with high dose chemotherapy followed by peripheral blood stem cell transplantation. Following or concurrent with the transplant, subjects can receive an infusion of the expanded CAR-T cells. In some embodiment, expanded cells are administered before or following surgery. In some embodiments, lymphodepletion is performed on a subject prior to administering one or more CAR-T cell as described herein. In such embodiments, the lymphodepletion can include administering one or more of melphalan, survivin, cyclophosphamide, and fludarabine. The dosage of the above treatments to be administered to a patient will vary with the precise nature of the condition being treated and the recipient of the treatment. The scaling of dosages for human administration can be performed according to art-accepted practices.

[0155] In some embodiments, a single treatment regimen is required. In others, administration of one or more subsequent doses or treatment regimens can be performed. For example, after treatment biweekly for three months, treatment can be repeated once per month, for six months or a year or longer. In some embodiments, no additional treatments are administered following the initial treatment.

[0156] Efficacy

[0157] The efficacy of activated CAR-T cells comprising a CAR or bispecific CAR described herein in, e.g., the treatment of mesothelin-expressing cancer, or to induce a response as described herein (e.g., a reduction in cancer cells) can be determined by the skilled clinician. However, a treatment is considered “effective treatment,” as the term is used herein, if one or more of the signs or symptoms of a condition described herein is altered in a beneficial manner, other clinically accepted symptoms are improved, or even ameliorated, or a desired response is induced, e.g., by at least 10% following treatment according to the methods described herein. Efficacy can be assessed, for example, by measuring a marker, indicator, symptom, and / or the incidence of a mesothelin-expressing cancer treated according to the methods described herein or any other measurable parameter appropriate.

[0158] Treatment according to the methods described herein can reduce levels of a marker or symptom of a mesothelin-expressing cancer, e.g., by at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80 % or at least 90% or more.

[0159] Efficacy can also be measured by a failure of an individual to worsen as assessed by hospitalization, or need for medical interventions (i.e., progression of the disease is halted). Methods of measuring these indicators are known to those of skill in the art and / or are described herein. Treatment includes any treatment of a mesothelin-expressing cancer in an individual or an animal (some non-limiting examples include a human or an animal) and includes: (1) inhibiting the mesothelin and / or MUC-16-expressing cancer, e.g., preventing a worsening of symptoms (e.g., pain or inflammation); or (2) relieving the severity of the mesothelin-expressing cancer, e.g., causing regression of symptoms. An effective amount for the treatment of a mesothelin-expressing cancer means that amount which, when administered to a subject in need thereof, is sufficient to result in effective treatment as that term is defined herein, for mesothelin- expressing cancer. Efficacy of an agent can be determined by assessing physical indicators of mesothelin-expressing cancer or desired response. It is well within the ability of one skilled in the art to monitor efficacy of administration and / or treatment by measuring any one of such parameters, or any combination of parameters. Efficacy of a given approach can be assessed in animal models of a condition described herein. When using an experimental animal model, efficacy of treatment is evidenced when a statistically significant change in a marker is observed.

[0160] Sequences

[0161] EXAMPLES

[0162] Example 1. Bispecific CAR-T cells against mesothelin and MUC16ectoto overcome tumor antigen heterogenicity. Introduction

[0163] Antigen heterogenicity is a challenge for CAR-T cell therapy in solid tumors due to antigen escape when targeting a single antigen. In ovarian cancer (OC), mesothelin (meso) and Mucinl6 (MUC16) are overexpressed antigens. Moreover, CA125, which is part of MUC16, is lost in some OC patients, but the ectodomain remains on the cell surface. Targeting both antigens (Meso and MUC16ect0) with a bispecific CAR configuration could overcome antigen heterogeneity to improve CAR-T cell efficacy in OC.

[0164] Methods

[0165] CAR design Six different chimeric antigen receptor (CAR) transgenes that encode for different tandem-scFv configurations against mesothelin (SSI) and the ectodomain of Mucin 16 (MUC16ect0) (4H11) based on scFv arrangement and linker length between scFvs were designed using Geneious (v. 2023.04). A linker containing one glycine and three serines (G4S) was used. The bispecific scFv was followed by a CD8 hinge / transmembrane domain and a 4-lBB(^ costimulatory domain.

[0166] Monospecific CARs against either mesothelin (SSI CAR) or MUC16 (4H11 CAR), containing the same hinge / transmembrane and intracellular domains as the bispecific CAR, were also synthesized. The transgenes contained mCherry for use as a marker to detect transduction efficiency. A third-generation lentiviral backbone under the regulation of a human EF-la promoter (GenScript) was used.

[0167] Lentiviral production

[0168] Lentiviral production was conducted transfecting HEK293T cells with plasmid and collecting the virus from the supernatant at 24 and 48 hours after transfection. Collected virus were filtered and concentrated by ultracentrifugation, and finally stored at -80°C.

[0169] Cell lines

[0170] ASPC-1, OVCAR3, Jukat, and HEK293T cell lines were obtained from the American Type Culture Collection (ATCC). Human T cells were purified (Stem Cell Technologies, catalog #15061) from healthy donor leukopaks purchased from the Massachusetts General Hospital blood bank under an institutional review board-approved protocol. An ASPC-1 cell line was genetically modified by lentiviral transduction to express MUC16ect0. To produce KO cell lines for any of the target antigens, CRISPR / Cas9 edition is used. All cell lines were genetically modified by lentiviral transduction to express CBG-GFP proteins (click beetle green luciferase; green fluorescent protein). A subset of ASPC-1 mesothelin+ cells were transduced with iRPF fluorescent protein, a subset of ASPC-1 MUC16ect0+ cells were transduced with mCherry fl orescent protein, and a subset of ASPC-1 mesothelin+ / MUC16ecto+ cells were transduced with truncated CD 19 (only the extracellular portion), using lentiviral vectors.

[0171] Jurkat NF AT-eGFT cells

[0172] Jurkat NFAT-eGFP reporter cells were prepared via lentiviral transduction. Jukat cells were engineered to express GFP upon NF AT (nuclear factor of activated T cells) activation. Jurkat NFAT-eGFP reporter cells were transduced with lentiviral vectors containing bispecific CAR and monospecific CAR constructs (“CAR- Jurkat NFAT-eGFP”). Binding capacity of CAR- Jurkat NFAT-eGFP cells to soluble mesothelin and MUC16 was tested by flow cytometry. Activation of CAR-Jurkat NFAT-eGFP cells upon co-culture with tumor cells expressing mesothelin, MUC16, mesothelin and MUC16, or neither cognate antigen was also evaluated via flow cytometry.

[0173] T cell activation and CAR-T cell expansion

[0174] CAR-T cells were produced by activation of primary human T cells, obtained from at least three different healthy donors, with anti-CD3 / anti-CD28 Dynabeads (ThermoFisher Scientific, catalog #40203D) for 7 days. Expansion was continued until day 14, after which cells were frozen for later use. During culture, fresh cell media supplemented with IL-2 (20UI / mL) was added every other day. Activated primary human T cells were transduced on day 1 with a lentiviral vector encoding either a monospecific CAR (SSI or 4H11) or a bispecific CAR. mCherry expression was evaluated via flow cytometry to assess transduction efficiency.

[0175] In vitro cytotoxicity assays

[0176] In vitro cytotoxicity assays were performed using luciferase killing assays and IncuCyte® Live-Cell Analysis system. The percentage of lysis of CGB-GFP+ tumor cells expressing mesothelin, MUC16, mesothelin and MUC16, or non-cognate antigen was measured following either 24 hours or 96 hours of co-culture with bispecific CAR-T cells. Monospecific CAR-T cells or untransduced T cells were used as a control. Tumor cell lysis was evaluated by tracking presence of CGB-GFP+ cells. To reproduce tumor heterogeneity, the killing activity of bispecific CAR-T cells was tested against mixed tumor models.

[0177] Tumor cell characterization from long-term killing assays

[0178] Tumor cells were collected from long-term in vitro cytotoxicity assays and analyzed by flow cytometry using a Fortessa 20X flow cytometer. To detach tumor cells from the plate, supernatant was aspirated and tumor cells were washed with PBS, followed by addition of TrypLE™ for 3 minutes at 37°C. Tumor cells were collected and washed in PBS and then stained with an anti-CD19 antibody (clone HIB19, Biolegend) and a live / dead dye (Fixable Far- Red Dead Cell Stain, Thermo Fisher). ASPC-1 mesothelin+ cells were identified as iRFP+, ASPC-1 MUC16ect0+ cells were identified as mCherry+, and ASPC-1 meso+ / MUC16ecto+ cells were identified CD19+.

[0179] Cytokine secretion assays Cytokine analysis from the supernatant of co-cultures of different tumor cell lines and bispecific CAR-T cells was performed using a Luminex™ Multiplex Assay, in a FLEXMAP 3D™ device. Supernatant from monospecific CAR-T cells and untransduced T cells was used as a control.

[0180] Avidity analysis

[0181] Bispecific CAR-T cell avidity for tumor cells was analyzed using a z-Movi® Cell Avidity Analyzer. The avidity of bispecific CAR-T cells was assessed as the percentage of CAR-T cells attached to a monolayer of tumor cells (expressing mesothelin, MUC16, mesothelin / MUC16, or noncognate antigen) when a maximum acoustic force was applied (1,000 picoNewton, pN). Monospecific CAR-T cells and untransduced T cells were used as a control. Avidity of bispecific CAR-T cells for tumor cells was tested at five minutes or ten minutes of incubation.

[0182] In vivo cytotoxic assays

[0183] In one model, immunodeficient NSG (NOD.Cg- / VA / c''c'' / / / 2 / ' '"' / ,, / / / SzJ) mice were engrafted subcutaneously with a mixed tumor model of 2 x 106ASPC-1 cells expressing mesothelin, MUC16, or mesothelin and MUC16 on day -14. On day zero, 3 x 106bispecific CAR-T cells or untransduced cells were administered to mice intravenously. Tumors were measured using calipers twice per week, and mice were weighed weekly and regularly assessed for signs of distress.

[0184] In another model, NSG mice were engrafted intraperitoneally with 3 x 106OVCAR3 tumor cells on day -14. On day zero, 1 x 106CAR-T cells were administered to mice intravenously. Tumor burden was measured weekly via bioluminescence, following intraperitoneal injection of D-luciferin substrate solution, using an Ami in vivo imaging system.

[0185] For both models, mice treated with monospecific CAR-T cells, untransduced T cells, or receiving no cells were used as controls. Animals were euthanized per the experimental protocol or when they met a priori defined endpoints by IACUC.

[0186] Results

[0187] In vitro

[0188] NFAT-eGFP cells transduced with bispecific CARS TanCARl and TanCAR3 had the higher binding capacity of soluble mesothelin and MUC16, together with the higher expression of GFP upon NF AT activation after co-culture with different settings of ASPC-1 and OVCAR3, across all tandem-scFv configurations (FIG. IB) (schematics of CAR constructs shown in FIG. 1A).

[0189] TanCARl and TanCAR3 T cells behaved like monospecific (SSI and 4H11) CAR-T cells with respect to killing in vitro tumor cells (ASPC-1 and OVACR3) expressing one of both antigens, after 24 hours of co-culture (FIGs. 2A-2B), but exhibited greater anti-tumor efficacy and tumor clearance than monospecific CAR-T cells against tumor heterogeneity models (Fig. 2C). Bispecific CAR-T cells were able to clear both mesothelin+, MUC16ect0+ and mesothelin+ / MUC16ecto+ tumor cells (FIG. 2D-2F). Moreover, tumor-antigen priority -target by TanCARl was tracked; TanCARl cells were observed to preferentially target MUC16ecto+ tumor cells compared to mesothelin+ tumor cells (FIGs. 2G-2H). In a spheroid model of ASPC1-1 cells expressing either mesothelin or MUC16ect0, TanCARl cells had better anti-tumor activity than monospecific CAR-T cells and TanCAR3 cells (FIGs. 2I-2J).

[0190] Cytokine secretion by TanCARl cells, TanCAR3 cells, and monospecific CAR-T cells was measured after 24 hours of co-culture with tumor cells (ASPC-1 or OCVAR3) expressing mesothelin, MUC16, mesothelin and MUC16, or non-cognate antigen. TanCARl and TanCAR3 cells produced IL-2, IFNy, GM-CSF, IL-6, IL-18, IL-13, and TNFa in response to culture with all tumor cells except double-negative tumor cells. Conversely, monospecific CAR-T cells only secreted cytokines in response to tumor cells expressing cognate antigen (FIGs. 3 A-3H) (IL-6, IL- 18, and IL- 13 production not shown).

[0191] TanCARl cells showed similar avidity to SSI monospecific CAR-T cells and lower than 4H11 monospecific CAR-T cells in response to a monolayer of tumor cells expressing either mesothelin or MUC16ect0(FIGs. 4A-4B). However, when cultured with a monolayer of tumor cells expressing both antigens, the avidity of TanCARl was lower than SSI monospecific CAR- T cells but higher than 4H11 monospecific CAR-T cells (FIG. 4C). The avidity of TanCARl was significantly increased against the same monolayer of tumor cells and was similar to both monospecific CAR-T cells once the incubation time was increased from five minutes to 10 minutes before application of the force (FIGs. 4D-4E). These data suggest that bispecific CARs are able to bind one antigen at a time in addition or alternatively to binding two simultaneously.

[0192] In vivo

[0193] In an in vivo cancer model using OVCAR3 tumor cells (FIG. 5A), which express both mesothelin and MUC16, there was no difference in bioluminescence reduction (a measure of tumor reduction) between TanCARl cells and monospecific CAR-T cells (FIG. 5B). Conversely, in an in vivo mixed tumor model of ASPC-1 cells (FIG. 5C), TanCARl cells exhibited improved anti-tumor efficacy for tumor shrinkage compared to SSI and 4H11 monospecific CAR-T cells (FIGs. 5D-I).

[0194] Example 2: Tandem CAR-T cells targeting mesothelin and MUC16 overcome tumor heterogeneity by targeting one antigen at a time.

[0195] Abstract

[0196] Tumor heterogeneity and antigen escape are mechanisms of resistance to CAR-T cell therapy, especially in solid tumors. To overcome tumor heterogeneity, mesothelin (meso) and Mucin 16 (MUC16), two antigens commonly expressed in solid tumors, were targeted. A series of tandem CAR constructs were designed based on different anti-meso (SSI) and anti- MUC16ecto (4H11) scFv arrangements and G4S linker lengths, and determined the best design regarding CARs functionality in vitro. Using the best design, the tandem CAR to monospecific CAR- T cells was compared in mixed tumor models in vitro and in vivo. It was found that the scFv arrangement and linker length impacted antigen binding and CAR expression in T cells. TanCARl (with SSI scFv located distally and one G4S repeat as the linker between scFvs) had the best binding and activation profile in vitro and outperformed SSI and 4H11 monospecific CAR-T cells in mixed tumor models in vitro and in vivo, showing a preferentially killing of tumor cells with the highest antigen expression. Moreover, acoustic force microscopy revealed that TanCARl- T cells likely bind to one antigen at a time. Tandem CAR-T cells targeting meso and MUC16ecto can be employed as a strategy to overcome tumor cell heterogeneity.

[0197] Introduction

[0198] Adoptive cell therapies using T cells engineered to express a chimeric antigen receptor (CAR) have shown outstanding results in B-cell acute lymphoblastic leukemia, B-cell lymphomas, and multiple myeloma, with corresponding FDA approvals. Even with these achievements, relapsing disease due to loss of targeted antigen expression is a common mechanism of resistance to CAR-T cell therapy. In some cases, tumor relapse is due to heterogeneous antigen expression in the tumor. One strategy to overcome tumor antigen heterogeneity and prevent antigen escape is to design CARs that target more than one antigen, such as by linking two single-chain variable fragments (scFvs) for the antigen-binding domain. This tandem configuration has been tested in hematological malignancies and solid tumor models, in some cases resulting in better tumor control compared to the co-administration of mono-specific CAR-T cellsl3 or the mono-specific CAR-T cells alone.

[0199] Mesothelin (meso) and Mucin 16 (MUC16) are two tumor-associated antigens commonly expressed in various solid malignancies including gynecological tumors (ovarian, uterine, and cervical), pancreatic ductal adenocarcinoma, and lung adenocarcinoma. CAR-T cells targeting mesothelin or MUC16 demonstrated promising results in pre-clinical studies but were less effective in clinical trials. The best response to meso-CAR-T cells in patients with ovarian cancer, pancreatic cancer or mesothelioma was stable disease. Similarly, MUC16-CAR- T cells have only achieved stable disease for ovarian cancer patients, even when combined with IL-12 secretion to enhance anti-tumor activity. In these trials, antigen expression levels and patterns of expression varied across patients, which could explain these unsatisfactory clinical outcomes. This heterogeneous antigen expression has been observed in ovarian and pancreatic patient tumors, resulting in a mixture of single and double positive cells. These data indicate that targeting both antigens could address tumor heterogeneity in ovarian and pancreatic malignancies.

[0200] Tandem CAR-T cells are designed to bind two target antigens and confer better control of tumors that express antigens heterogeneously compared to mono-specific CAR-T cells. Studies using heterogeneous tumor models have shown that tandem CAR-T cells prevent tumor antigen escape and increase tumor clearance compared to mono-specific CAR-T cells. However, when only one target antigen is expressed in the tumor, the anti-tumor activity and cytokine production of tandem CAR-T cells may be lower than mono-specific CAR-T cells. Conversely, in heterogeneous or mixed populations of tumors, or when both antigens are expressed, most published variations of tandem CAR-T cells have anti-tumor effect that is generally better than or similar to mono-specific CAR-T cells. These data raise the question as to whether tandem CAR-T cells bind to one antigen at a time or if they are capable of binding both target antigens simultaneously. Likewise, it remains to be determined whether tandem CAR-T cells preferentially kill tumor cells within heterogeneous tumors based on their antigen density and / or the affinity of the scFvs composing the tandem CAR, as mono-specific CAR-T cells often do. It was hypothesized in this Example that tandem CAR-T cells can bind one antigen at a time on tumor cells that express two antigens, and that this may lead to a preferential killing of tumor cells displaying the highest antigen density in heterogeneous tumor models. To answer these questions, tandem CAR T cells targeting mesothelin and the MUC16 ectodomain (MUC16ecto) were designed and examined their functionality in heterogeneous tumor models. After determining the optimal design in terms of scFv order and the linker length connecting the scFvs, the one-antigen-at-a-time binding hypothesis was tested using a ramp of acoustic force to measure the avidity of tandem CAR-T cells to tumor cells. To mimic tumor antigen heterogeneity, mixed tumor models were prepared in vitro and in vivo and the anti-tumor efficacy of tandem CAR-T cells was compared to mono-specific CAR-T cells. It was shown that, on double-positive tumor cells, tandem CAR-T cells bind one antigen at a time and overcome tumor antigen heterogeneity more efficiently than mono-specific CAR-T cells in mixed tumor models in vitro and in vivo. Moreover, tandem CAR-T cells show a preferential killing of tumor cells based on their antigen density.

[0201] Materials and Methods

[0202] Cell lines

[0203] ASPC-1 (cat. No. CRL-1682), Capan-2 (cat. No. HTB-80), BxPC-3 (cat. No. CRL-1687) pancreatic adenocarcinoma cell lines and SKOV-3 (Cat. No. HTB-77), OVCAR3 (cat. No. HTB-161) and Jurkat cells were purchased from the ATCC and cultured as recommended, OVCAR4 was provided by from Dr. O.O. Yeku. All cell lines were cultured in RPMI (Gibco) media supplemented with 10% FBS and 1% penicillin / streptomycin (RIO media). OVCAR3 cell line was cultured in RPMI media (Gibco) and supplemented with 20% FBS and 1% penicillin / streptomycin (R20 media). TrypLE Express (Gibco) was used to disassociate ASPC-1 and OVCAR3 cell lines between routine passaging to preserve expression of mesothelin. ASPC- 1 and OVCAR3 mesothelin knockout, OVCAR3 Mucin 16 (MUC16) knockout and OVCAR3 double knockout, MUC16ecto transduced ASPC-1, and color- and antigen-coded cell lines were generated as indicated in Supplementary Material and Methods. All cell lines were routinely tested for Mycoplasma contamination. Cell lines ordered from the ATCC were used within 6 months of ordering or authenticated via STR analysis.

[0204] Human samples

[0205] High grade serous ovarian cancer (HGSOC) samples were obtained following informed written consent under DFHCC protocol 07-049. The histology of the tumor was confirmed by MGB pathology. The tumor / cells were banked and or used to generate organoid or xenograft establishment. Independently, ovarian and pancreatic PDX samples and their correlate information were obtained from the Center for Patient Derived Models (CPDM) database at Dana-Faber (cBioPortal for Cancer Genomics).

[0206] Generation of CAR constructs

[0207] Transgenes for anti-mesothelin, anti-MUC16ecto and tandem CAR were designed in Geneious Prime (Version 2023.1.1). After synthesizing, the constructs were cloned into a lentiviral plasmid backbone regulated by a human EF-la promoter (Genscript). Tandem CARs, anti-mesothelin and anti-MUC16ecto CARs construct contain a CD8 signal sequence, a CD8 hinge and transmembrane domains, a 4-1BB co-stimulatory domain, and an intracellular CD3z signaling domain. Each construct also contains mCherry as a fluorescent marker for the evaluation of transduction efficiency. The scFvs against mesothelin and MUC16ecto were derived from sequences of SS130 and 4H1118 (available to the public: PCT / US13 / 28980 and PCT / US2011 / 030025). Six tandem iterations were design based on scFv positioning and linker length between scFvs. The scFv of SSI and 4H11 were linked using a G4S linker (from one to three repetitions). The order of the scFv was swapped. The tandem scFvs iterations are as follow: SSl-(G4S)xl-4Hl l (tandeml), SSl-(G4S)x2-4Hl l (tandem2), SSl-(G4S)x3-4Hl l (tandem3), 4Hl l-(G4S)xl-SSl (tandem4), 4H1 l-(G4S)x2-SSl (tandem5) and 4H1 l-(G4S)x3- SS1 (tandem6). All sequences were codon optimized using IDT Codon Optimization Tool and GenSmart Codon Optimization tool from Genescript.

[0208] CAR T-cell production

[0209] Leukapheresis product from anonymous healthy human donors was purchased from the MGH blood bank, under an institutional review board (IRB)-exempt protocol. T-cell Isolation kits (Stem Cell Technologies) were used according to the manufacturer’s protocols and T cells were cryopreserved. For CAR T-cell production, cells were thawed in R10 media and activated with human CD3 / CD28 Dynabeads (Life Technologies), which were added at a 3: 1 bead to T- cell ratio, along with 20 lU / mL media of recombinant human IL2 (PeproTech). The next day, a corresponding amount of lentivirus was added to each well to have a MOI of 5 per construct. Untransduced T cells (UTD) from matching donors were cultured simultaneously under the same conditions to serve as controls. Every 2 to 3 days, cells were counted and R10 media was added to have a 5xl05cells / mL, and IL2 was added to a final concentration of 20 lU / mL. After seven days, magnetic separation was used to remove Dynabeads from T-cell culture. Transduction efficiency was evaluated on day 12 or 13 through mCherry expression, on a Fortessa 20x or NovoCyte Penteon (Agilent). On day 14, bulk CAR T cells were cryopreserved for further assays in vitro and in vivo.

[0210] Flow cytometry

[0211] The following antibody clones were used for flow cytometry: Human anti-Mesothelin (Maus-Laboratory antibody, A2A11), Mouse Fab2 (Cell Signaling Technology, 4410), Human anti-Mesothelin (Miltenyi Biotec, REA1057), Human IgGl (Miltenyi Biotec, REA293) Human anti-CA125 (R&D Systems, MAB56091), Mouse IgGl (R&D Systems, FAB9627P), Human anti-CD19 (BioLegend, 302240), Human anti-G4S (Cell Signaling Technology, 38907, and 69782). For direct staining of anti-mesothelin scFv (SSI) expression, His-tag-FITC-Labeled Human Mesothelin (ACROBiosystems, MSN-HF223) and FITC-Labeled Human Mesothelin (Kactus, MSL-HM280F) were used. For direct staining of anti-MUC16ecto scFv (4H11) expression, His-tag Labeled Human CA125 (ACROBiosystems CA5-H52H6), and anti-His-tag antibody (Cell Signaling Technology, 14931) were used. Appropriate isotype control antibodies were used as necessary. In general, cells were washed in PBS 2% FBS before staining with respective antibodies / labelled proteins at ambient temperature or 4°C for 15 minutes in the dark, following manufacture’s recomemmendations. Cells were washed again and stained with DAPI, 7AAD (BioLegend, 420404) or Live / Dead™ Fixable Far Red Dead Cell Stain Kit (ThermoFisher, L34974) to assess cell viability before analyzing on a BDFortessa X-20 or a NovoCyte Penteon (Agilent). To assess mesothelin and MUC16 expression HGSOC patient derived organoids (PDOs) by flow cytometry, the organoids were transferred from the plate to a conical tube by aspirating Matrigel domes from each well. After centrifuging, the pellet was resuspended in TrypLe (ref# 12604-021) at 37°C for dissociation, and cells were spun down and washed with PBS twice. The antibody staining protocol was the same as for tumor cell lines. Analysis was done in FlowJo.

[0212] Real-time cytotoxic assays

[0213] Different types of ASPC-1 tumor cells regarding mesothelin and MUC16 / MUC16ecto expression were cultured overnight in a 48-well plate (Coming, 3548). CAR-T and UTD-T cells were added to the 48-well plates in an E:T ratio 1 : 1. CAR-T / UTD cells were co-cultured with tumor cells for 96 hours. Cytotoxicity was analyzed using an IncuCyte® Live-Cell Analysis system. The Tumor Area (tumor cells) per well for 96 hours normalized to time point 0 (beginning of the assay) was the parameter to establish the anti-tumor activity of CAR / UTD-T cells. The CAR-T Cell Area (mCherry+ cells) per well for 96 hours normalized to time point 0 was the parameter to quantify the proliferation of CAR-T cells. Tumor cells used for these assays expressed GFP protein.

[0214] Cytokine analysis

[0215] Supernatants from luciferase killing assays in an E:T ratio 10: 1 were collected after 24 hours and froze for later cytokine analysis. Cytokine expression was assessed with the Thl / Th2 Cytokine 11 -Pl ex Human ProcartaPlex Panel (Invitrogen) according to the manufacturer’s instructions using a FLEXMAP 3D (Luminex, ThermoFisher Scientific).

[0216] Spheroids killing assays

[0217] Mixed spheroids of MesoendoMUC16neg ASPC-1 cells and MesoKOMUC16ectoTR ASPC-1 were made co-culturing these tumor cells, in a 1 : 1 ratio, 20x103 total cells / well, with RIO media with a 0.1% of methyl cellulose (Sigma-Aldrich). Tumor cells were seeded in a 96- well plate, centrifuged for 3 minutes at 1630rpm, and incubated for 3 days at 37°C CO2 5%. After 3 days, 20xl03total CAR and UTD-T cells were added to each well. Cytotoxicity was monitored for 106 hours using an IncuCyte® Live-Cell Analysis system (see previous paragraph Real-time cytotoxic assays). CAR-T cells were recovered and counted by flow cytometry using CountB right™ Plus Absolute Counting Beads (ThermoFisher), in a Fortessa 20x.

[0218] Avidity assessment with acoustic force cell microscopy

[0219] Poly-L Lysine-coated flow cell z-Movi chips (Lumicks) were seeded with different types of ASPC-1 cells based on mesothelin and MUC16ecto expression, at a 40xl06cells / mL, and incubated for 1.5 hours. CAR / UTD-T cells were sorted in a Sony MAB900 (Sony Biotechnology Inc.) 48 hours before the avidity assessment, using 7AAD-, mCherry+ gates to harvest only the CAR-T cells. Sorted CAR-T cells and UTD-T cells were stained with a CellTrace™ Far red Proliferation kit (Thermo Fisher Scientific). Each chip was run first with UTD-T cells, to prevent unspecific confounding binding of CAR-T cells, then run with mono- specific and tandem CAR- T cells, and finally run with UTD-T cells. To prevent any potential benefit of binding in the earlier runs, the order of mono-specific and tandem CAR-T cells to be tested was alternated. T cells were incubated for a 5-minutes binding period and visualized on the z-Movi cell avidity analyzer (Lumicks), meanwhile a lOOOpN acoustic force ramp was applied. The percentage of cells bound as a function of acoustic force applied was then analyzed through Ocean software (version 1.5.5 Lumicks). Each chip was run five times.

[0220] In vivo models

[0221] NOD-SCID-y chain' / _(NSG) mice (The Jackson Laboratory) were housed and bred at the Center for Compartive Medicine at Massachusetts General Hospital under pathogen-free conditions. Mice were maintained and used for conducting experiments in adherence to an MGH Institutional Animal Care and Use Committee (lACUC)-approved protocol (No.2020N000114). All mouse handling, including injections and monitoring, was performed by an animal technician independent (blinded) from the scientific hypotheses. Mice were normalized before treatment to have similar tumor engraftment (as measured by mean caliper measurement or bioluminescence) across all groups. Mixed models of ASPC-1 made of MesoendoMUC16neg, MesoKO MUC16ectoTR and MesoendoMUC16ectoTR cells, in a 1 : 1 : 1 ratio, were resuspended to 2x106 cells / 150 pL of a 1 : 1 mixture of PBS and Matrigel (Corning) and injected into each mouse subcutaneously. Caliper measurements were taken biweekly. Fourteen days after tumor injection, CAR or UTD-T cells were thawed, counted and resuspended to 3xl06CAR / UTD-T cells / lOOpL of PBS, and finally administered intravenously through the tail vein. For T cell infiltration into tumor, the same mixed model of ASPC-1 was used as described above, and the same CAR-T / UTD-T cell number and route of administration were used. Mice were euthanized after 21 days of CAR-T / UTD-T cells injection to analyze the abundance of T cell infiltration by IHC. OVCAR3 CBG- GFP tumor cells were resuspended to 3e6 cells / 150 pL of PBS, and injected intraperitoneally. Mice were monitored weekly for bioluminescent emission using an Ami HT optical imaging system (Spectral Instruments) after 5 minutes post intraperitoneal substrate injection of D-Luciferin (30mg / mL). Fourteen days after tumor injection, CAR-T or UTD-T cells were thawed, counted and resuspended to le6 CAR T cells / 100 pL of PBS, and finally, administered intraperitoneally. Mouse experiments continued until IACUC guidelines recommended euthanasia, or until 90 days post CAR-T / UTD-T cell injection, whatever happened first.

[0222] IHC After mice were euthanized, tumors were extracted and half was fixed in 4% PFA for 2 hours at 37°C, next moved to 30% EtOH overnight at 4°C, then moved to 50% EtOH overnight at 4°C, and finally, stored in 70% EtOH at 4°C until staining. Tissue slides were then made and embedded in paraffin. Slides were stained with hematoxylin and for CD3 by the specialized histopathology services core facility at MGH. Slides were scanned using an AxioScan and CD3+cells were quantified with QuPath software (v0.5.1). Briefly, first the stain vectors were estimated for each scanned slice, the whole scanned piece of tumor was selected as a region of interest (ROI) and ran a cell detection analysis that allowed detection of cells based on hematoxylin staining. Finally, a positive cell detection analysis was run based on cell staining for CD3 (“Cell: DAB OD mean”). The number of CD3+ / mm2was annotated.

[0223] Statistical analysis

[0224] Data are generally depicted as mean±standard error of the mean, unless stated otherwise within figure legends. Statistically significant differences were determined by tests as indicated within figure legends. A P value of <0.05 was considered as significant. P values are indicated with asterisks as indicated in each figure legend and are listed as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, p < 0.0001; and ns, P > 0.05. The number of repeated biological and technical replicates is indicated in the corresponding figure legends. Statistical analysis was conducted in GraphPad Prism (Version 10.2.2).

[0225] Results

[0226] Mesothelin and MUC16 are heterogeneously expressed across ovarian and pancreatic cancer patients’ tumors and cell lines

[0227] To confirm the expression of mesothelin and MUC16 across ovarian and pancreatic tumors, samples from patients diagnosed with HGSOC (FIG. 6A) and human tumor cell lines (FIG. 6B) were evaluated by flow cytometry and in a subset of patient derived xenografts (PDX) by mRNA expression (FIG. 6C). The patient samples were collected at different time points of the disease. The sample from patient 1 was collected from a debulking surgery in a platinum-sensitive disease; patient 2 was collected from ascites with platinum-resistant disease; and patient 3 was collected from pleural effusion with platinum-resistant disease. In these patient samples, the expression of mesothelin and MUC16 was highly heterogeneous. Most of Patient l’s cells expressed both antigens, with 66% being double-positive, and 4.57% and 26.3% single positive for meso or MUC16, respectively. Patient 2 had mostly mesothelin singlepositive cells (72.7%) and 15% double-positive, with < 2% single-positive for MUC16. Patient 3 had mostly double-positive cells (78.5%), followed by MUC16 single-positive (14.1%) and 3.4% meso-single- positive (FIG. 6A).

[0228] Mesothelin and MUC16 antigen expression was also heterogeneous across the tumor cell lines. All ovarian cancer cell lines expressed both antigens to different extent. OVCAR3 was mostly double-positive and had the highest expression of either antigen by MFI (Mean Fluorescence Intensity) (FIG 6B; FIG. 7A); OVCAR4 had roughly equal percentages of double-positive, MUC16 single-positive, and double-negative cells; and SKOV3 cells were mostly double-negative. Among the pancreatic cancer cell lines, only CAPAN2 expressed MUC16, while all of the pancreatic cell lines evaluated expressed mesothelin, with ASPC-1 having the highest mesothelin expression by MFI (FIG. 6B; FIGs. 7A-7B). This heterogeneity was confirmed in 39 ovarian and 45 pancreatic cancer PDX samples (FIG. 6C). Pancreatic cancer PDXs had a lower expression of MUC16 than ovarian cancer PDXs (FIG. 6C). These data confirm the heterogeneity of expression of Meso and MUC16 and therefore the potential benefit of targeting mesothelin and MUC16 with CAR-T cell therapy for ovarian and pancreatic cancer patients.

[0229] CAR T cells with tandem scFvs targeting mesothelin and MUC16 bind to and are activated by both antigens.

[0230] To target both antigens, a tandem CAR using scFvs specific for mesothelin and MUC16 was designed. The scFv rearrangement and linker length that conferred the highest mesothelin and MUC16 binding in a tandem design was investigated. A monocistronic lentiviral (LV) vector encoding a 4-lBB-costimulated CAR containing anti-mesothelin (SSI) and anti- MUC16ecto (4H11) scFvs connected by a G4S linker was designed. Six additional constructs with varying G4S linker lengths and scFv arrangements were also designed (FIG. 8A; FIG. 9A; FIG. 1A) and generated SSI and 4H11 mono-specific CAR constructs as controls. The antitumor activity of CAR-T cells containing SSI scFv and 4H11 scFv has been reported elsewhere. All constructs were based on an EFl - alpha-driven, second-generation CAR backbone containing the intracellular domains of 4- IBB and CD3z. An mCherry cDNA sequence was added to each construct as a marker of transduction efficiency (FIG. 9A; FIG. 1A).

[0231] An immortalized human T cell line, Jurkat cells, was transduced with LV vectors containing each of the tandem and mono-specific constructs and assessed transduction efficiency (by mCherry expression; FIGs. 9B-9C) and CAR expression (using an anti-G4S antibody; FIG. 9D) on cell surface by flow cytometry. Transduction efficiency was similar among the constructs, with the exception of TanCARs 1-3, which showed overall lower transduction. Additionally, transduction of TanCARs 1-3 was proportionally lower at increasing length of the G4S linker. As expected, the MFI of the G4S+population increased with the number of G4S repetitions (FIG. 9D). The 4H11 mono-specific CAR had a high MFI value for mCherry, but the lowest G4S expression, suggesting that this CAR is not efficiently expressed on the cell surface (FIGs. 9C-9D). The antigen-binding capacity of the tandem and mono-specific CAR-Jurkat cells to soluble mesothelin and MUC16 (CA125) was measured by flow cytometry (FIG. 8B). The mono-specific SSI CAR had the highest binding capacity to mesothelin, followed by the tandem configurations with the SSI scFv in the distal position (TanCARs 1-3), and then the tandem constructs with distal 4H11 (TanCARs 4-6). TanCARl, with one G4S repeat as linker, had the highest binding for mesothelin, followed by TanCAR2 ((G4S)x2)) and TanCAR3 ((G4S)x3). TanCARl also had the highest binding capacity for MUC16, which was even higher than the mono-specific 4H11 CAR, likely due to the higher CAR expression. These data suggest that the scFV order and the linker length impact the binding of soluble antigens by the tandem CAR, with linker length being inversely related to soluble antigen binding.

[0232] The activation of CAR-Jurkat cells upon exposure to tumor cells was evaluated. To do this, ASPC-1 pancreatic cancer and OVCAR3 ovarian cancer cell lines that expressed different combinations of mesothelin and MUC16 were generated. ASPC-1 cells express endogenous mesothelin and are negative for MUC16 at baseline (MesoendoMUC16neg), serving as single mesothelin antigen-expressing cells. To better recapitulate patient samples where both antigens are expressed, double and single MUC16 antigen-expressing cells were generated using the MUC16 ectodomain. Only express the ectodomain (the domain that the 4H11 scFv binds to) was selected for expression due to the large size of MUC16, which makes it difficult to transduce the whole protein. The MUC16 ectodomain was transduced into wild-type ASPC-1 (Mesoend°MUC16ectoTR) or ASPC-1 with mesothelin knocked out (MesoKOMUC16ectoTR), whereas wild-type ASPC-1 with mesothelin KO served as a double-negative control (MesoKOMUC16neg; FIG. 8C). OVCAR3 cells express endogenous mesothelin and MUC16 at baseline (Mesoend°MUC16endo) and were knocked out for mesothelin and / or MUC16 to generate single antigen-expressing (MesoKOMUC16endo; MesoendoMUC16KO) and doublenegative cells (MesoKOMUC16KO; FIG. 8C). Mono-specific or tandem CARs were transduced into Jurkat NFAT-eGFP reporter cells (FIG. 9E), which express GFP upon activation (Supplementary Material and Methods). When co-cultured with tumor cells, Jurkat NFAT-eGFP cells transduced with the tandem CARs expressed GFP in the presence of tumor cells that had either or both antigens, but did not express GFP when the tumor cells did not express the antigens. As expected, Jurkat NFAT- eGFP cells transduced with either mono-specific CAR only expressed GFP when co-cultured with tumor cells expressing the corresponding antigen. None of the CAR- Jurkat NFAT-eGFP reporter cells showed relevant GFP expression in the absence of tumor, demonstrating a lack of tonic signaling. These data indicate that tandem CARs recognize mesothelin and MUC16ecto (either as in the endogenous MUC16 or in the transduced one) on tumor cells, whether one or both antigens are present, and drive T cell activation in an antigen-specific manner (FIG. 8D). Collectively, the data from the soluble antigen binding and activation upon tumor cell exposure demonstrated that TanCARl had the highest binding capacity for mesothelin and MUC16 and the highest activation in response to both tumor cell lines compared to the other tandem constructs (FIGs. 8E-8F). TanCAR2 and 3 had the next highest antigen binding, but TanCAR3 showed slightly stronger activation. For this reason, TanCARl and TanCAR3 were chosen for further studies in primary human T cells.

[0233] TanCARl T cells have superior in vitro cytotoxicity

[0234] To examine TanCARl and TanCAR3 function in primary human T cells, primary human T cells isolated from normal donors were transduced with LV encoding the SSI, 4H11, TanCARl, or TanCAR3 construct. The transduction efficiency of these four constructs varied, as demonstrated by differences in mCherry+cells measured by flow cytometry (FIG. 11A). SSI CAR-T cells showed the highest transduction efficiency (56.7%), and TanCARl-T cells the lowest (32.3%). Similar to the CAR- Jurkat cells, the 4H11 construct had the lowest G4S expression by flow (FIG. 11B), suggesting decreased CAR expression despite similar transduction efficiency. Likewise, TanCAR3 had slightly higher G4S expression compared to TanCARl, likely due to the increased linker length. To normalize the number of CAR T cells used for functional analyses across constructs, UTD-T cells were used to dilute the percentage of mCherry+cells to the lowest transduction efficiency. In vitro tumor lysis assays were then performed to measure luciferase activity after 24 hours of co-incubating tumor cells and CAR-T cells (Supplementary Material and Methods). ASPC-1 and OVCAR3 with different antigen expression patterns that also express luciferase and GFP were used to allow for tracking in vitro and in vivo. It was observed that TanCARl and TanCAR3 lysed ASPC-1 and OVCAR3 tumor cells in all conditions (single- and double-positive tumor cells). As expected, the mono-specific CAR-T cells only lysed the tumor cells when the target antigen was expressed (FIGs. 10A-10B; FIGs. 2A-2B; FIGs. 11C-11J). This cytotoxic activity was antigen-specific since UTD-T cells did not kill tumor cells, and none of the CAR T cells killed the double-negative cell lines (FIGs. 10A-10B; FIGs. 2A-2B). TanCARl and TanCAR3 killed ASPC-1 equally well, regardless of the different antigen- expression patterns, but TanCARl had superior killing activity than TanCAR3 against MesoKOMUC16endo and Mesoend°MUC16endo OVCAR3 cells.

[0235] Thl and Th2 cytokine production by the CAR-T cells was also measured 24 hours after exposure to tumor cells. Supernatant from the killing assay cultures was collected and measured IL-2, IFNy, TNFa and GM-CSF by Luminex assay. As expected, the mono-specific CARs (SSI and 4H11) produced cytokines when their target antigen was expressed by the cell line but not when the antigen was not expressed. For ASPC-1 cells, cytokine production was similar when both antigens were expressed; as well as for OVCAR3 cells (FIGs. 10C-10D). Apparently, SSI CAR-T cells expressed less cytokines in response to double+ vs. meso-single+ (FIG. 10D). Likewise, the level and patterns of cytokine expression correlated with the level of mesothelin and MUC16ectoTRor MUC16 antigen expressed in the cell line. For example, ASPC-1 has higher levels of MUC16ecto since it was artificially introduced into the cells versus what is naturally expressed by OVCAR3. Accordingly, 4H11 and the TanCARs had higher cytokine expression under the stimulation of MUC16ectoTRASPC-1 cells than MUC16endo OVCAR3 (FIGs. 10C-10D)

[0236] To distinguish between TanCARl and TanCAR3, cytokine production in response to each cell line was compared. In ASPC-1, the TanCARs cytokine production was mainly driven by MUC16 recognition, whereas in OVCAR3 co-cultures more cytokine secretion was observed in response to mesothelin (FIGs. 10C-10D). These data correlated with the soluble protein binding capacity of each CAR construct (FIG. 8B). The level of cytokine production was very similar between TanCARl and 3 in ASPC-1 and OVCAR3 co-cultures (FIGs. 10C-10D). To elucidate any structural difference(s), the in-silico structure prediction of TanCARl and TanCAR3 was also examined using AlphaFold337. In general, each scFv structure was predicted with high confidence while there was lower confidence in the relative position of the SSI and 4H11 scFvs to each other (FIGs. 11K-11N). As expected, areas with a greater flexibility are predicted with lower confidence, e.g., the G4S region (FIGs. 11K-11L). Moreover, the expected position error, which estimates the error in the relative position between two residues of the structure, was higher for TanCAR3 between the VH and VL of each scFv, than for TanCARl, which might be due to a higher flexibility of the structure that may interfere with the antigen binding (FIGs. 11M-11N).

[0237] Finally, the binding avidity of the TanCAR-T cells for the ASPC-1 and 0VCAR3 cells co-expressing both antigens was examined. The binding avidity was measured by plating the ASPC-1 or 0VCAR3 cells on a chip, layering the CAR-T cells on top, and exposing the chip to increasing acoustic force. The percentage of cells that remained attached was measured via a z- movi cell avidity analyzer, as previously described. There were no observed differences in the binding avidity between TanCARl and TanCAR3 for ASPC-1 or OVCAR3 (FIGs. 4A-4D; FIGs. 10E-10F). Therefore, the only difference between the two designs was the increased killing of OVCAR3 cells in vitro by TanCARl vs. TanCAR3. For this reason, TanCARl was chosen for further analysis against heterogenous tumor models in vitro and in vivo.

[0238] Tandem CAR-T cells bind one antigen at a time

[0239] Tandem CARs are designed to target two antigens on tumor cells; however, it is not clear whether each tandem CAR molecule binds to both antigens at once or only one antigen at a time. It has been shown that tandem CAR-T cells cluster the targeted antigens in the same immune synapse, but it was not determined if each CAR was binding one or both antigens. To explore this question, the binding avidity between the tandem CAR-T cells and tumor cells that expressed one or both target antigens was analyzed using acoustic force and the z-movi analyzer, as described above. The avidity of TanCARl against ASPC-1 cancer cells that expressed the single antigens, MesoendoMUC16negor MesoKOMUC16ectoTR, was measured and its avidity compared with the mono-specific CAR-T cells. TanCARl showed a similar avidity to the mono-specific CAR-T cell with specificity to the antigen that was expressed (FIGs. 12A- 12B), suggesting that scFv avidity is not affected by tandem design and the avidity of the individual scFv is maintained. Next, the avidity of TanCARl for tumor cells that express both antigens was analyzed. It was hypothesized that if TanCARl binds one antigen at a time, its maximum avidity would be dictated by the scFv with the highest avidity (in this case, the SSI scFv). Whereas if TanCARl binds two antigens at a time, there would be a synergistic effect of the two scFvs, rendering a higher avidity value compared to the mono-specific CARs (FIG. 12C). Since TanCARl did not have a higher avidity than SSI CAR-T cells and the avidity curves directly overlapped, TanCARl is likely binding to one antigen at a time (FIG. 4D). Likewise, there was no difference between TanCARl and 4H11 CAR-T cells, further supporting this hypothesis (FIG. 4D).

[0240] TanCARl T cells overcome tumor antigen heterogeneity in mixed tumor models in vitro

[0241] The efficacy of TanCARl against heterogeneous tumors using a mixed tumor model was next determined. In vitro mixed cultures of ASPC-1 tumor cells were generated expressing the single antigens (MesoendoMUC16neg, MesoKOMUC16ecto) and both antigens (Mesoend°MUC16ectoTR; FIG. 13A) or neither antigen (MesoKOMUC16neg), what was termed a stressed heterogeneous tumor model (FIG. 14A). SSI, 4H11, TanCARl, or UTD-T cells were then added to the mixed cultures and tumor cell killing was measured by the change in the tumor area over time using an IncuCyte® Live-Cell analysis system. TanCARl eliminated significantly more mixed ASPC-1 tumor cells compared to SS1CAR-T cells but demonstrated similar cytotoxicity to 4H11CAR-T cells (FIG. 2C; FIG. 13B; FIG. 14B). To determine the cell composition at the end of the co-culture (96 hours), tumor cells were collected and each population was quantified by flow cytometry. The absolute number of tumor cells was lowest in the TanCARl coculture, followed by the 4H11 and then by the SSI CAR T cells (FIG. 13C; FIG. 14C). TanCARl targeted all tumor cell populations, except the double negative cells (MesoKOMUC16neg). Meanwhile, the mono-specific CAR-T cells only targeted tumor cells expressing their target antigen (FIG. 13D; FIG. 14D). There were no differences in the tumor cell populations between the UTD-T cell-treated and tumor-only groups (FIGs. 14E-14H).

[0242] Next it was investigated whether this tumor cell killing preference was based on antigen density. Since the ectodomain of MUC16 was transduced to be overexpressed in the ASPC-1 cells, its expression is likely higher than mesothelin, which could explain the preferential killing of MUC16ectoTRcells by TanCARl. To test this, mixed tumor cell cultures of ASPC-1 MesoendoMUC16negand MesoKOMUC16ectoTRwere prepared (FIG. 13E). To distinguish these two populations, the MesoendoMUC16negcells were transduced with iRFP, and the MesoKOMUC16ectoTRcells were transduced with GFP. The mixed tumor cells were co-cultured with CAR-T cells or UTD-T cells, and tumor cell killing was measured using the IncuCyte® Live-Cell Analysis system. Again, TanCARl controlled tumor growth better than mono-specific CAR-T cells when looking at the tumor cell population as a whole (FIG. 13G). When looking at cellular composition, TanCARl decreased the area MesoendoMUC16negcells better than the 4H11 CAR, but worse than the SSI CAR-T (FIG. 13F). TanCARl also killed GFP+(MesoKOMUC16ectoTR) cells better than the SSI CAR but similar to the 4H11 CAR (FIG. 13G). When comparing which cells were preferentially targeted by TanCARl, the MesoKOMUC16ectoTRcells were killed significantly more than the MesoendoMUC16neg(FIG. 13H)

[0243] Finally, a 3D in vitro spheroid model was used to measure CAR T cell killing of heterogeneous tumors. Spheroids from ASPC-1 MesoendoMUC16negiRFP+cells and MesoKOMUC16ectoTRGFP+cells were generated (FIG. 141). Spontaneously, the MesoendoMUC16negcells organized in the core of the spheroids and the MesoKOMUC16ectoTRcells in the periphery (FIG. 14J). Three days post seeding, CAR-T cells or UTD-T cells were added to the spheroids, and the area of the spheroids in each tumor cell line and CAR-T cells was monitored using an IncuCyte® Live-Cell Analysis system. TanCARl controlled tumor spheroids better than mono- specific CAR-T cells (FIG. 131; FIG. 14K), and killed both ASPC- 1 cells more effectively than mono-specific CAR-T cells (FIGs. 13J-13K). CAR-T cell infiltration into the spheroid structure and their expansion was also evaluated by mCherry expression. TanCARl better infiltrated the spheroid compared to the SSI and 4H11 CARs (FIG. 131). Intriguingly, the SSI CAR-T cells surrounded the spheroid but were not able to infiltrate as well as the 4H11 CAR and TanCARl (FIG. 131). This pattern was likely due to the preferential location of the MesoKOMUC16ectoTRin the periphery, which are not targeted by the SSI CAR- T cells and possibly blocked access to the MesoendoMUC16negcells. Similarly, the 4H11 CAR-T cells tended to stay in the periphery of the spheroid, colocalizing with the MUC16ectoTRcells but not infiltrating further (FIG. 131). TanCARl showed a higher expansion across time via the IncuCyte® quantification compared to the SSI CAR-T cells, but no statistical differences between TanCARl and 4H11 were observed (FIG. 141). These results were confirmed by flow cytometry for mCherry+cells following dissociation of the spheroids (FIG. 14J).

[0244] Collectively, these data demonstrate that TanCARl effectively targets heterogeneous tumors in pancreatic cell line mixed tumor models showing increased killing compared to mono- specific CAR-T cells.

[0245] TanCARl T cells expand more and control heterogeneous tumor growth better than mono-specific CAR-T cells in vivo

[0246] TanCARl-T cells were next examined in vivo, first using the ovarian cancer cell line OVCAR3, which naturally expresses both antigens. NSG mice were engrafted intraperitoneally with OVCAR3 tumor cells, treated intravenously 14 days later with CAR-T cells, and monitored for tumor growth via bioluminescent imaging (FIG. 5A; FIG. 15A; FIG. 16A). All mice treated with CAR-T cells experienced tumor reduction compared to the tumor only and UTD-treated groups, without differences between TanCARl-T cells and mono-specific CAR-T cells (FIG. 5B; FIG. 15B) To better model the tumor heterogeneity observed in cancer patients, a mixed tumor model of ASPC-1 MesoendoMUC16neg, MesoKOMUC16ectoTRand Mesoend°MUC16ectoTRin a 1 : 1 : 1 ratio was prepared (FIG. 16B), and engrafted this mixture subcutaneously in NSG mice. After 14 days, administered CAR or UTD-T cells were administered intravenously and tumor growth was monitored via caliper measurements (FIG. 5C; FIG. 15C). Again, all mice treated with CAR-T cells had controlled tumor growth compared to tumor-only and UTD- treated mice; however, TanCARl further decreased tumor growth compared to the mono- specific CAR-T cells (FIG. 5D; FIG. 15D) and, in some cases, completely eradicated the tumor (FIG. 16D). Mice treated with mono-specific CAR-T cells also showed an earlier relapse of the tumor compared to TanCARl- treated mice (FIG. 16C).

[0247] To determine which cells were targeted most in vivo, when possible, tumors at the time of endpoint disposition were collected, and the tumor cell populations were determined by flow cytometry (FIG. 15E). From one mouse treated with SSI CAR-T cells, the remaining tumor had almost no MesoendoMUC16negcells, but only MesoKOMUC16ectoTRand Mesoend°MUC16ectoTRtumor cells. Tumors from mice treated with 4H11 CAR-T cells showed a higher proportion of MesoendoMUC16negcells than MesoKOMUC16ectoTRand Mesoend°MUC16ectoTRtumor cells. Interestingly, in tumors from mice treated with TanCARl-T cells, the most abundant tumor population was MesoendoMUC16neg. This suggests that TanCARl -T cells preferentially kill cells with MUC16ectoTRin vivo (FIG. 15F). This data aligns with the in vitro mixed tumor killing assays, where TanCARl-T cells had better anti- tumor activity over the mono-specific CAR-T cells and a skewed killing towards MUC16ectoTRtumor cells.

[0248] Finally, the infiltration of CAR-T cells into tumors was studied using an in vivo tumor model of mixed ASPC-1 (MesoendoMUC16neg, MesoKOMUC16ectoTR, and Mesoend°MUC16ectoTR). Briefly, after 14 days of subcutaneous tumor engraftment, CAR or UTD-T cells were administered, and the tumors were collected 21 days later to analyze T cell infiltration by H4C for CD3+cells (FIG. 15G; FIG. 16C). Mice treated with UTD-T cells had a scarce number of T cells in tumors, while TanCARl -treated mice showed the highest CD3+cell infiltration, which was statistically significant compared to SSI and 4H11 tumor-treated mice (FIG. 15I-15H). These data suggest that TanCARl has a better anti-tumor effect and higher intra-tumoral expansion than mono-specific CAR-T cells. Discussion

[0249] Strategies that target multiple antigens could avoid or reduce tumor relapse. CAR-T cell therapy has not yielded results as promising in solid tumors as in hematologic malignancies. One explanation for this discrepancy is that solid tumors tend to have more antigen heterogeneity, which makes mono-specific CAR-T cells a suboptimal strategy for these tumors. In this Example, a dual-targeting CAR-T cells was designed to bind two antigens highly expressed in ovarian and pancreatic cancer, in an attempt to improve their efficacy against these hard-to-treat tumors. A series of tandem scFvs was built targeting mesothelin (SSI) and the ectodomain of MUC16 (4H11), and it was found that TanCARl (with SSI as the distal scFv and only one G4S linker between it and the 4H11 scFv) had the best anti-tumor activity in vitro. TanCARl also controlled tumor growth and increased T cell infiltration in a mixed tumor model in vitro and in vivo compared to the mono-specific CAR-T cells. Moreover, it was demonstrated that TanCARl-T cells likely bind to one antigen at a time, which skews their killing function toward the antigen that is more highly expressed in the tumor.

[0250] This is the first time mesothelin and MUC16 have been targeted using a tandem CAR design. While mesothelin and MUC16 monospecific CAR T cells have been used in clinical trials, neither have achieved results beyond stable disease. As reported previously, mesothelin and MUC16 expression in tumors is heterogeneous. It was noted that MUC16 expression in human pancreatic cell lines and PDX is lower than previously reported. For that reason, overexpressed the MUC16 ectodomain was overexpressed in a pancreatic cancer cell line to better represent antigen expression in patients. It was found that the tandem CAR-T cells outperformed the mono-specific CAR-T cells in settings where there was a mix of single and double antigen-expressing cells, similar to the composition of cells that have been observed in patient tumors. However, the tandem CAR-T cells did not show superiority compared to mono- specific CAR-T cells against double-positive tumor cells alone in vitro and in vivo. These results are similar to those previously reported, where tandem CAR-T cells targeting various antigens did not show superiority to their monospecific CAR-T cell counterparts when both antigens are expressed by the tumor cells. Therefore, the benefit of tandem over mono-specific CARs is especially relevant in heterogeneous models in terms of anti -turn or activity and T-cell expansion.

[0251] In mixed tumor models, TanCARl preferentially killed tumor cells that express MUC16ectoTRin vitro and in vivo. In response to ASPC-1 with MUC16ectoTRin vitro, TanCARl had higher cytokine secretion and greater activation than cells without MUC16ectoTR. In heterogeneous tumors in vivo, tumor relapse after TanCARl treatment was mainly driven by MesoendoMUC16negASPC-1 cells, demonstrating an increased killing of tumor cells with higher antigen expression. This was in line with what was observed in mixed tumor models in vitro in this Example, where TanCARl -T cells preferentially killed MesoKOMUC16ectoTRcells over MesoendoMUC16negwhen they were tracked with cell-specific fluorescent proteins. This increased killing is likely due to a higher density of MUC16ectoTRfrom being transduced into the cells versus the endogenously expressed mesothelin. This is in line with data from monospecific CAR-T cells, where increased antigen expression results in increased killing.

[0252] The increased killing of MUC16ectoTRcells by TanCARl -T cells could also be due to increased accessibility of the MUC16 ectodomain when lacking expression of the whole protein. The ectodomain, where the 4H11 scFv binds, is the portion of the protein that is proximal to the cell membrane and the portion that remains when MUC16 is cleaved. In OVCAR3 cells, where the full-length MUC16 is naturally expressed, the 4H11 mono-specific CAR-T cells had rather low activation and cytokine secretion in response to MesoKOMUC16endocells in vitro, whereas ASPC-1 cells with MUC16ectoTRreadily activated and induced cytokine production by 4H11 CAR-T cells. This suggests that, in vitro, 4H11 scFv cannot properly reach its binding epitope in MUC16. In contrast, 4H11 mono-specific CAR-T cells controlled OVCAR3 tumor growth equally as well as the SSI and TanCARl CAR-T cells in vivo. This paradoxical phenomenon could be related to differences between in vitro and in vivo shedding conditions of MUC16, which leads to the exposure of the epitope to 4H11 CAR-T cells. Mechanisms related to MUC16 cleavage are not well known, but it has been associated with the presence of pro-inflammatory cytokines and an acidic media in the Golgi apparatus. Also, when full-length MUC16 is present in OVCAR3 cells, activation and cytokine secretion by SSI mono-specific CAR-T cells are lower than when it is knocked out. This difference could be due to the competitive binding of MUC16 and SSI scFv to meso.

[0253] It was further observed that the TanCARl -T cells behave like the mono-specific CAR-T cells of the antigen that is more accessible (MUC16ecto in ASPC-1 and mesothelin in OVCAR3). This similarity suggested that tandem CAR-T cells target one antigen at a time. This hypothesis was tested using acoustic force microscopy to measure the avidity between the CAR- T and tumor cells. This was the first time this technique was used to characterize the avidity of tandem CAR-T cells. It was found that TanCARl had the same avidity as the mono-specific CAR-T cells for the single positive tumor cells, even though the affinity to the soluble antigen was different between mono-specific CAR and tandem CAR-T cells. Discrepancies between affinity and avidity have been recently published comparing different CAR constructs, in which even though having similar affinity values, avidity made the difference in tumor control.

[0254] Moreover, based on avidity data, it was observed that TanCARl bound one antigen at a time, as there was no significant increase in avidity over mono-specific CAR-T cells after exposure to a cell line that expressed both antigens. These results align with the similar killing activity of double-positive tumor cells by tandem and mono-specific CAR-T cells.

[0255] In conclusion, TanCARl is the first tandem CAR design against mesothelin and MUC16ecto that shows benefit over mono-specific CAR-T cells, especially in heterogeneous tumor models. The one-antigen-at-a-time binding properties of TanCARl and its apparently preferential killing of tumor cells with high antigen density may help to design therapeutic strategies to overcome tumor heterogeneity in ovarian and pancreatic tumors.

Claims

CLAIMSWhat is claimed is:

1. A bispecific chimeric antigen receptor (CAR) comprising:(i) a first antigen-binding domain that binds to mesothelin; and(ii) a second antigen-binding domain that binds to mucin 16 (MUC16).

2. The bispecific CAR of claim 1, wherein the first antigen-binding domain comprises a mesothelin-binding antibody.

3. The bispecific CAR of claim 2, wherein the mesothelin-binding antibody is a mesothelin-binding antibody fragment.

4. The bispecific CAR of claim 3, wherein the mesothelin-binding antibody fragment is an antigen-binding fragment (Fab), a Fab', or a F(ab')2, a fragment variable (fv), or a single chain variable fragment (scFv).

5. The bispecific CAR of any one of claims 2-4, wherein the mesothelin-binding antibody comprises:(a) a heavy chain variable domain (VH) comprising three complementarity determining regions CDR-H1, CDR-H2, and CDR-H3, wherein the CDR-H1 comprises an amino acid sequence of SEQ ID NO: 3; the CDR-H2 comprises an amino acid sequence of SEQ ID NO: 4; and the CDR-H3 comprises an amino acid sequence of SEQ ID NO: 5, and(b) a light chain variable domain (VL) comprising three complementarity determining regions CDR-L1, CDR-L2, and CDR-L3, wherein the CDR-L1 comprises an amino acid sequence of SEQ ID NO: 6; the CDR-L2 comprises an amino acid sequence of SEQ ID NO: 7; and the CDR-L3 comprises an amino acid sequence of SEQ ID NO: 8.

6. The bispecific CAR of claim 5, wherein the VH comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 1 and the VL comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 2.

7. The bispecific CAR of claim 6, wherein the VH comprises an amino acid sequence of SEQ ID NO: 1 and the VL comprises an amino acid sequence of SEQ ID NO: 2.

8. The bispecific CAR of any one of claims 5-7 comprising an amino acid sequence of any one of SEQ ID NOs: 19-21 between the VH and the VL.

9. The bispecific CAR of any one of claims 3-8, wherein the mesothelin-binding antibody fragment comprises an amino acid sequence of SEQ ID NO: 9.

10. The bispecific CAR of any one of claims 1-9, wherein the second antigen-binding domain comprises a MUC16-binding antibody.

11. The bispecific CAR of claim 10, wherein the MUC16-binding antibody binds to the ectodomain of MUC16.

12. The bispecific CAR of claim 10 or claim 11, wherein the MUC16-binding antibody is a MUC16-binding antibody fragment.

13. The bispecific CAR of claim 12, wherein the MUC16-binding antibody fragment is an antigen-binding fragment (Fab), a Fab', or a F(ab')2, a fragment variable (fv), or a single chain variable fragment (scFv).

14. The bispecific CAR of any one of claims 8-13, wherein the MUC16-binding antibody comprises:(a) a heavy chain variable domain (VH) comprising three complementarity determining regions CDR-H1, CDR-H2, and CDR-H3, wherein the CDR-H1 comprises an amino acid sequence of SEQ ID NO: 12; the CDR-H2 comprises an amino acid sequence of SEQ ID NO: 13; and the CDR-H3 comprises an amino acid sequence of SEQ ID NO: 14, and(b) a light chain variable domain (VL) comprising three complementarity determining regions CDR-L1, CDR-L2, and CDR-L3, wherein the CDR-L1 comprises an amino acid sequence of SEQ ID NO: 15; the CDR-L2 comprises an amino acid sequence of SEQ ID NO: 16; and the CDR-L3 comprises an amino acid sequence of SEQ ID NO: 17.

15. The bispecific CAR of claim 12, wherein the VH comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 10 and the VL comprises an amino acid sequence having at least 90% identity to SEQ ID NO: 11.

16. The bispecific CAR of claim 6, wherein the VH comprises an amino acid sequence of SEQ ID NO: 10 and the VL comprises an amino acid sequence of SEQ ID NO: 11.

17. The bispecific CAR of any one of claims 14-16 comprising an amino acid sequence of SEQ ID NO: 21 between the VH and the VL.

18. The bispecific CAR of any one of claims 12-17, wherein the MUC16-binding antibody fragment comprises an amino acid sequence of SEQ ID NO: 18.

19. The bispecific CAR of any one of claims 1-18, further comprising a peptide linker between the first antigen-binding domain that binds to mesothelin and the second antigenbinding domain that binds to MUC16.

20. The bispecific CAR of claim 19, wherein the peptide linker comprises SEQ ID NO: 19.

21. The bispecific CAR of claim 19, wherein the peptide linker comprises SEQ ID NO: 20.

22. The bispecific CAR of claim 19, wherein the peptide linker comprises SEQ ID NO: 21.

23. The bispecific CAR of any one of claims 19-22, comprising from N-terminal to C- terminal:(i) the first antigen-binding domain that binds to mesothelin;(ii) the peptide linker; and(iii) the second antigen-binding domain that binds to MUC16.

24. The bispecific CAR of any one of claims 19-22, comprising from N-terminal to C- terminal:(i) the first antigen-binding domain that binds to MUC16;(ii) the peptide linker; and(iii) the second antigen-binding domain that binds to mesothelin.

25. The bispecific CAR of any one of claims 1-24 further comprising:(a) a transmembrane domain;(b) a co-stimulatory domain; and(c) an intracellular signaling domain.

26. The bispecific CAR of claim 25, wherein the transmembrane domain is selected from the group consisting of alpha (a), beta (P) or zeta (Q chain of a T cell receptor, CD28, CD3 epsilon (e), CD45, CD4, CD5, CD8, CD9, CD16, CD22, CD33, CD37, CD64, CD80, CD86, CD134, CD137, CD154, KIRDS2, 0X40, CD2, CD27, LFA-1 (CDlla, CD18), ICOS (CD278), 4-1BB (CD137), 4-1BBL, GITR, CD40, BAFFR, HVEM (LIGHTR), SLAMF7, NKp80 (KLRFI),CD 160, CD 19, IL2R beta (P), IL2R gamma (y), IL7Ra, ITGA1, VLA1, CD49a, ITGA4, IA4, CD49D, ITGA6, VLA-6, CD49f, ITGAD, CD lid, ITGAE, CD 103, ITGAL, CDlla, LFA-1, ITGAM, CD 11b, ITGAX, CDllcJTGBl, CD29, ITGB2, CD 18, LFA-1, ITGB7, TNFR2, DNAM1 (CD226), SLAMF4 (CD244, 2B4), CD84, CD96 (Tactile), CEACAM1, CRT AM, Ly9 (CD229), CD160 (BY55), PSGL1, CD100 (SEMA4D), SLAMF6 (NTB-A, Lyl08), SLAM (SLAMF1, CD 150, IPO-3), BLAME (SLAMF8), SELPLG (CD 162), LTBR, PAG / Cbp, NKp44, NKp30, NKp46, NKG2D, and / or NKG2C transmembrane domains.

27. The bispecific CAR of claim 25, wherein the transmembrane domain comprises a CD8 transmembrane domain.

28. The bispecific CAR of claim 27, wherein CD8 transmembrane domain comprises a CD8 transmembrane domain and a CD8 hinge domain (CD8 hinge / TM).

29. The bispecific CAR of claim 28, wherein the CD8 hinge / TM comprises an amino acid sequence of SEQ ID NO: 22.

30. The bispecific CAR of claim 25 or claim 27, wherein the co-stimulatory domain comprises a 4-1BB, CD27, CD28, 0X40, CARD11, CD2, CD7, CD27, CD28, CD30, CD40, CD54 (ICAM), CD83, CD134 (0X40), CD137 (4-1BB), CD150 (SLAMF1), CD152 (CTLA4), CD223 (LAG3), CD270 (HVEM), CD273 (PD-L2), CD274 (PD-L1), CD278 (ICOS), DAP10, LAT, NKD2C SLP76, TRIM, or ZAP70 co-stimulatory domain.

31. The bispecific CAR of claim 30, wherein the co-stimulatory domain comprises a 4-1BB co-stimulatory domain.

32. The bispecific CAR of claim 30, wherein the co-stimulatory domain comprises a 4-1BB co-stimulatory domain comprising an amino acid sequence of SEQ ID NO: 23.

33. The bispecific CAR of any one of claims 25-32, wherein the intracellular signaling domain comprises a CD28, 4-1BB, CD27, TCR^, FcRy, FcRP, CD3 gamma (y), CD3 theta (0), CD3 sigma (o), CD3 eta (q), CD3s, CD3< CD22, CD79a, CD79b, or CD66d intracellular signaling domain.

34. The bispecific CAR of claim 33, wherein the intracellular signaling domain comprises a CD3(^ signaling domain.

35. The bispecific CAR of claim 34, wherein the CD3(^ signaling domain comprises an amino acid sequence of SEQ ID NO: 24.

36. The bispecific CAR of any one of claims 1-35, further comprising a leader sequence.

37. The bispecific CAR of claim 36, wherein the leader sequence comprising a CD8 leader sequence.

38. The bispecific CAR of claim 37, wherein the CD8 leader sequence comprises SEQ ID NO: 25.

39. The bispecific CAR of claim 1-25, further comprising a CD8 hinge / transmembrane domain, a 4-1BB co-stimulatory domain, and a CD3(^ intracellular signaling domain.

40. The bispecific CAR of any one of claims 1-22, comprising an amino acid sequence having at least 90% identity to any one of SEQ ID NOs: 26-31.

41. The bispecific CAR of claim 40, comprising an amino acid sequence of any one of SEQ ID NOs: 26-31.

42. The bispecific CAR of claim 41, comprising an amino acid sequence of SEQ ID NO: 26.

43. The bispecific CAR of claim 41, comprising an amino acid sequence of SEQ ID NO: 28.

44. A polynucleotide comprising a nucleic acid encoding the bispecific CAR of any one of claims 1-43.

45. A vector comprising the polynucleotide of claim 44.

46. The vector of claim 45, wherein the vector is a viral vector.

47. The vector of claim 46, wherein the vector is a lentiviral vector.

48. A cell comprising the bispecific CAR of any one of claims 1-43, the polynucleotide of claim 44, or the vector of any one of claims 45-47.

49. The cell of claim 48, wherein the cell is an immune cell.

50. The cell of claim 49, wherein the cell is a T cell.

51. A method of treating a subject having a mesothelin and / or MUC16 expressing cancer, the method comprising administering the cell of claim 49 or claim 50 to the subject.

52. The method of claim 51, comprising administering the cell of claim 50 to the subject.

53. The method of claim 51 or claim 52, wherein the cancer is ovarian cancer.

54. The method of claim 51 or claim 52, wherein the cancer is pancreatic cancer.

55. The method of any one of claims 51-53, wherein the cell in an autologous T cell.

56. The method of any one of claims 51-54, wherein the cell in an allogeneic T cell.

Images