Chimeric antigen receptor compositions and methods of making same

Abstract

Chimeric antigen receptor proteins comprising CD8α hinge and CD8α transmembrane domains and nucleic acid molecules that encode the same are disclosed. Chimeric antigen receptor proteins comprising CD8α hinge and CD8α transmembrane domains anti-GUCY2C scFvs, and nucleic acid molecules that encode the same, and methods of making and using the same are disclosed. Methods of making the recombinant cells such as recombinant T cells are disclosed.

Description

[0001] CHIMERIC ANTIGEN RECEPTOR COMPOSITIONS AND METHODS OF MAKING SAME

[0002] FIELD OF THE INVENTION

[0003] The invention relates to chimeric antigen receptors, to nucleic acid molecules that encode such chimeric antigen receptors and to methods of making recombinant cells that comprise nucleic acid molecules that encode such chimeric antigen receptors and express chimeric antigen receptors. The invention relates to chimeric antigen receptors that bind to GUYC2C, to nucleic acid molecules that encode such chimeric antigen receptors that bind to GUYC2C, to methods of making recombinant cells that comprise nucleic acid molecules that encode such anti-GUYC2C chimeric antigen receptors and express anti- GUYC2C chimeric antigen receptors and to methods of using recombinant cells that comprise nucleic acid molecules that encode such anti-GUYC2C chimeric antigen receptors and express anti-GUYC2C chimeric antigen receptors to treat and prevent cancer that expresses GUYC2C.

[0004] BACKGROUND OF THE INVENTION

[0005] Immunotherapy based upon T cells that express chimeric antigen receptors (CARs) has become an emerging modality for treating cancer. CARs are fusion receptors that comprise a domain which functions to provide HLA-independent binding of cell surface target molecules and a signaling domain that can activate host immune cells of various types, typically peripheral blood T cells, which may include populations of cells referred to cytotoxic lymphocytes, cytotoxic T lymphocytes (CTLs), Natural Killer T cells (NKT) and Natural Killer cells (NK) or helper T cells. That is, while typically being introduced into T cells, genetic material encoding CARs may be added to immune cells that are not T cells such as NK cells.

[0006] Guanylyl cyclase C (also referred to interchangeably as GCC or GUCY2C) is a membrane-bound receptor that produces the second messenger cGMP following activation by its hormone ligands guanylin or uroguanylin, regulating intestinal homeostasis, tumorigenesis, and obesity. GUCY2C cell surface expression is confined to luminal surfaces of the intestinal epithelium and a subset of hypothalamic neurons. Its expression is maintained in >95% of colorectal cancer metastases and it is ectopically expressed in tumors that evolve from intestinal metaplasia, including esophageal, gastric, oral, salivary gland and pancreatic cancers.

[0007] The inaccessibility of GUCY2C in the apical membranes of polarized epithelial tissue due to subcellular restriction of GUCY2C, creates a therapeutic opportunity to target metastatic lesions of colorectal origin which have lost apical-basolateral polarization, without concomitant intestinal toxicity.

[0008] A syngeneic, immunocompetent mouse model demonstrated that CAR-T cells targeting murine GUCY2C were effective against colorectal cancer metastatic to lung in the absence of intestinal toxicities. Similarly, other GUCY2C-targeted therapeutics, including antibody-drug conjugates and vaccines, are safe in preclinical animal models, and therapeutic regimens utilizing these platforms are in clinical trials for metastatic esophageal, gastric, pancreatic, and colorectal cancers (NCT02202759, NCT02202785, NCT01972737).

[0009] The safety of these therapeutic regimens, in the context of GUCY2C expression across the rostral-caudal axis of intestine, reflects compartmentalized expression of GUCY2C, enriched in apical, but limited in basolateral, membranes of epithelial cells. Systemic radiolabeled imaging agents conjugated to GUCY2C ligand target GUCY2C- expressing metastases without localizing in intestine, confirming the mucosal compartmentalization of the receptor.

[0010] Tumors express up to 10-fold greater amounts of GUCY2C, compared to normal epithelial cells, potentially creating a quantitative therapeutic window to discriminate receptor overexpressing tumors from intestinal epithelium with low / absent GUCY2C in basolateral membranes.

[0011] U.S. Patent Application Publication 20120251509 Al and U.S. Patent Application Publication US 2014-0294784 Al, which are each incorporated herein by reference, disclose CARs including CARs that bind to guanylyl cyclase C, T cells that comprise CARs including T cells that comprise CARs that bind to GUCY2C and target cells that comprise GUCY2C, methods of making chimeric antigen receptors and T cells, and methods of using T cells that comprise CARs that bind to GUCY2C and target cells that comprise GUCY2C to protect individuals against cancer cells that express GUCY2C and to treat individuals who are suffering from cancer in which cancer cells express GUCY2C. There remains a need for improved CAR constructs. There remains a need for improved methods of manufacturing cells that comprise CAR constructs. There remains a need for improved compositions including anti-GUYC2C CARs, nucleic acids encoding such CARS and cells expressing such CARS as well as methods of making such compositions, methods to protect individuals against cancer cells that express GUCY2C and to treat individuals who are suffering from cancer in which cancer cells express GUCY2C.

[0012] SUMMARY OF THE INVENTION

[0013] Chimeric antigen receptor proteins that comprise an antigen binding domain, a CD8a hinge domain, a CD8a transmembrane domain, and one or more intracellular domains are provided. In some embodiments, chimeric antigen receptor proteins that comprise an anti-GUCY2C scFV antigen binding domain, a CD8a hinge domain, a CD8a transmembrane domain, and one or more intracellular domains are provided.

[0014] Nucleic acid molecules that encode such proteins are provided. The nucleic acid molecules may be operably linked to regulatory elements that can function to express the protein in a human cell such as a human T cell. The nucleic acid molecules may be incorporated in a nucleic acid vector such as a plasmid or recombinant viral vector that can be used transform human cells into human cells that express the protein.

[0015] Human cells comprising the nucleic acid molecules and express the proteins are provided.

[0016] Methods of making such cells are provided.

[0017] Methods of treating a patient who has cancer that has cancer cells that express GUCY2C and methods of preventing cancer that has cancer cells that express GUCY2C in a patient identified as being of increased risk, are provided.

[0018] BRIEF DESCRIPTION OF FIGURES

[0019] Figure 1. CD2 Agonism Improves C AR-T cell Manufacturing. The inclusion of CD2 into the activation complex does not impact T-cell viability (a) but does impact the expansion of the T-cells in culture (b). (Number above bar is the total number of T-cells (x 106). The amount of surface CAR is shown in the representative flow plots (c). The reduction in surface car expression is observed in both the CD8HTM and CD28HTM CAR designs (line within each panel is the matched CAR design). Figure 2. CD2 Agonism Promotes More Favorable T-cell Phenotypes. Inclusion of CD2 in the activation of T-cells results in T-cells with less differentiated phenotypes. Activation of T-cells with CD3 and CD28 promotes more differentiated effector memory T-cells, whereas including CD2 into the activation milieu promotes cells with more naive / stem cell-like- and central memory T-cells (a,b). The absence of CD2 agonism also leads cells to express more markers of T-cell exhaustion in both the CD4+and CD8+populations. There is marked increase in the expression of CD39 and PD-1 when CD2 is not present during activation (c).

[0020] Figure 3. CD2 Agonism Allows for Tumor Cell Killing in Extremely Low Effector to Target (E:T) Ratios. The ability of T-cells to kill T84 CRC cells in vitro at an E:T ratio great than 1 has been established (see attached manuscript). When the E:T ratio is < 0.1 : 1 , the ability of C AR-T cells to kill is extremely limited. Here we show that the inclusion of CD2 agonism at the time that T-cells are activated allows for these T-cells to kill when the E:T ratio is extremely low (0.1: 1) (a,b). Regardless of CAR design, we observe that CD2 agonism allows for target cell killing, whereas cells not activated with CD2 are not able to kill.

[0021] Figure 4. Palmitoylation Promotes Increased CAR Expression. A cysteine residue on the inner leaflet of the transmembrane domain of CD8a is palmitoylated. We generated a mutant (MT) CAR where that cysteine residue was changed to an alanine and no longer able to accept the palmitate modification, (a) Removing the palmitate from the CAR results in a reduction in the surface expression of the CAR in both CD8+(b) and CD4+(c) T-cells. Representative flow plots show the overlay of wild-type CAR and MT CAR for each respective T-cell subset below the respective bar graph.

[0022] Figure 5. CD8HTM and CD28HTM GUCY2C CAR-T Production, (a) Designs of CD8HTM and CD28HTM CARs. (b-i) CD8HTM and CD28HTM CAR constructs were used for 14-day manufacturing of CAR-T cells from n=6 donors and matched comparisons between CARs. (b-c) Expansion and viability of CD8HTM and CD28HTM CAR-T cell products, (d) CD8+ and CD4+ populations among CAR-transduced T cells. Representative flow cytometry plot shown, (e) Transduction efficiency (% GFP+) and construct expression (GFP MFI) for all 6 donors, (f-g) Comparison of transduction efficiency (f) and construct expression (g) between CD8+ and CD4+ T cells for each CAR. (h-i) Comparison of transduction efficiency (h) and construct expression (i) between CD8HTM and CD28HTM CARs for CD8+ and CD4+ T cells. All statistical comparisons are paired T-tests; symbols connected with a line represent a matched donor.

[0023] Figure 6. CD8HTM and CD28HTM CAR-T Products Lack Exhaustion and Possess Memory Phenotypes, (a) Effector, effector memory, central memory, and naive / stem cell memory-like phenotypes among matched CD8HTM and CDHTM CAR-T cells produced from n=3 donors. A representative flow cytometry plot is shown, (b) Prevalence of 0 to 4 exhaustion marker expression among matched CD8HTM and CDHTM CAR-T cells produced from 6 donors. SPICE plots demonstrating the average number and type of exhaustion marker expression. CD8HTM and CD28HTM memory (a) and exhaustion (b) phenotypes were not statistically different. Statistical analyses used two-way ANOVA adjusted for multiple comparisons. Error bars reflect the standard error of the mean (SEM).

[0024] Figure 7. CD8HTM and CD28HTM CAR Expression, (a-b) Surface CD8HTM and CD28HTM CAR expression were examined with Protein L among CD8+ (a) and CD4+ (b) CAR-T cells produced from n=6 donors, (c-d) Surface CD8HTM and CD28HTM CAR expression were examined with anti-G4S antibody among CD8+ (c) and CD4+ (d) CAR-T cells produced from 6 donors, (e-f) Bound anti-G4S antibodies / cell in c-d were quantified among CD8+ (e) and CD4+ (f) CAR-T cells. Flow cytometry plots in a-d indicate matched CD8HTM and CD28HTM CAR expression for each donor. All statistical comparisons are paired T-tests; symbols connected with a line represent a matched donor.

[0025] Figure 8. Enhanced CD8HTM CAR Affinity Compared to CD28HTM CAR. (a-d) Purified recombinant GUCY2C extracellular domain was incubated with CD8HTM and CD28HTM CAR-T cells produced from n=3 donors at varying concentrations (0-300 nM). Bound GUCY2C was detected with a fluorescent secondary antibody and cells were analyzed by flow cytometry, (a) Representative flow cytometry plots comparing CD8HTM and CD28HTM CARs. (b) Affinity and EC50 determination for CD8HTM and CD28HTM CARs among CD8+ and CD4+ CAR-T cells, (c) Maximum binding (Bmax) determination for CD8HTM and CD28HTM CARs among CD8+ and CD4+ CAR-T cells, (d) Comparison of CAR affinities between CD8+ and CD4+ T cells for each CAR design. Non-linear regression analysis was performed in b-d with p- value testing for matching curves. Figure 9. Enhanced Cytokine Production and Cytolytic Potential by CD8HTM CAR Compared to CD28HTM CAR. (a-h) CD8HTM and CD28HTM CAR-T cells produced from n=3 donors were stimulated with plate-bound recombinant GUCY2C extracellular domain protein, and intracellular cytokines (a-f), and granzyme B (g-h) were quantified by flow cytometry. (a,d) Comparison of CD8+ T cells (a) and CD4+ T cells (d) producing 0-3 cytokines between CD8HTM and CD28HTM CARs. * p = 0.048; *** p = 0.0005 (b) SPICE plots demonstrating the average number and type of cytokine expression by CD8HTM and CD28HTM CAR-T cells among CD8+ (b) and CD4+ (e) CAR-T cells. Individual cytokine expression comparison between CD8HTM and CD28HTM CAR-T cells among CD8+ (c) and CD4+ (f) CAR-T cells. Granzyme B comparison between CD8HTM and CD28HTM CAR-T cells among CD8+ (g) and CD4+ (h) CAR-T cells. Representative flow cytometry plots are shown. Statistical analyses for a and d was performed using two-way ANOVA adjusted for multiple comparisons. Error bars reflect the SEM. Statistical comparisons for c, f, g, and h are paired T-tests; symbols connected with a line represent a matched donor.

[0026] Figure 10. Enhanced Cytolysis at Decreasing Antigen Densities by CD8HTM CAR Compared to CD28HTM CAR. (a-b) GUCY2C mRNA (a) and protein (b) were quantified for colorectal cancer cells spanning from undetectable to high expression, (c-e) Cytolysis kinetics and time-to-80%-killing (KT80) comparisons between CD8HTM and CD28HTM CARs from n=3 donors using T84 (c), LS174T (d), and Lovo (e) cells (shading around killing curves represent SEM). (f) Correlation of KT80 and GUCY2C protein levels from n=3 donor CAR-T cells. Shaded regions surrounding the cytotoxicity curves (c,d,e) represent the SEM. Statistical comparisons for the KT80 graphs (c,d,e) are paired T-tests; symbols connected with a line represent a matched donor. Error bars in (f) represent the SEM; the 95% confidence interval is represented by the shaded region.

[0027] Figure 11. Enhanced Antitumor Efficacy of CD8HTM CAR Compared to CD28HTM CAR. (a) Experimental design, (b) Tumor burden comparison between CD8HTM and CD28HTM groups one day before treatment, (c) Longitudinal bioluminescence images, (d) Median (bold) and individual tumor burden comparison between CD8HTM and CD28HTM CAR-T cells. Grey range indicates the 95% confidence interval of baseline daily luminescence in tumor- free animals, (e) Time to tumor clearance comparison between CD8HTM and CD28HTM CAR-T cells determined by the day an animal’s signal reached that of tumor-free mice in d. An unpaired T-test was performed in b. Each dot represents a single animal. A Kaplan-Meier curve was used in e with the Log-rank (Mantel-Cox) test to determine the p-value.

[0028] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0029] Single chain protein sequences that bind to the extracellular domain of human GUCY2C were generated using fragments of the variable light chain and variable heavy chain of an anti-GUCY2C antibody that binds to the extracellular domain of human GUCY2C. A linker sequence connects the variable light chain fragment to the variable heavy chain fragment into a single chain antibody variable fragment fusion protein sequence (scFv) that binds to the extracellular domain of human GUCY2C.

[0030] The scFv is a component in a CAR, which is a larger fusion protein. The CARs functional components include the immunoglobulin-derived antigen binding domain, antibody sequences i.e. svFv, which binds to human GUCY2C, a hinge domain that links the scFV to a transmembrane domain that anchors the protein in the cell membrane of the cell in which it is expressed, and the signally domain which functions as signal transducing intracellular sequences (also referred to as cytoplasmic sequences) that activate the cell upon scFv binding to human GUCY2C. The nucleic acid sequences that encode the CAR include sequences that encode a signal peptide from a cellular protein that facilitate the transport of the translated CAR to the cell membrane. CARs direct the recombinant cells in which they are expressed to bind to and, in the case of recombinant cytotoxic lymphocytes, recombinant cytotoxic T lymphocytes (CTLs), recombinant Natural Killer T cells (NKT), and recombinant Natural Killer cells (NK) kill cells displaying the antibody-specified target, i.e. GUCY2C. When the CAR is expressed it is transported to the cell surface and the signal peptide is typically removed. The mature CAR functions as a cellular receptor. The scFv and hinge domain are displayed on the cell surface where the scFv sequences can be exposed to proteins on other cells and bind to GUCY2C on such cells. The transmembrance region anchors the CAR in the cell membrane and the intracellular sequences function as a signal domain to transduce a signal in the cell which results in the death of GUCY2C-expressing cell to which the CAR-expressing cell is bound.

[0031] In some embodiments, the CARs comprise SEQ ID NO:2. In some embodiments, the CARs comprise SEQ ID NO:2 and are encoded by SEQ ID NO:1. In some embodiments, the CARs comprise a signal sequence, such as for example a mammalian or synthetic signal sequence. In some embodiments, the CARs comprise a signal sequence from a membrane-bound protein such as for example a mammalian membrane-bound protein. In some embodiments, the CARs comprise a signal sequence from a membrane-bound protein such as CD8 alpha, CD8 beta, CD4, TCR alpha, TCR beta, CD3 delta, CD3 epsilon, CD3 gamma, CD28, and BiP. Examples of signal sequences may also be found in membrane bound any mammalian signal sequence <http: / / www.signalpeptide.de / index.php?m=listspdb_mammalia>. In some embodiments, the CARs comprise a Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) signal sequence. In some embodiments, the CARs comprise CD8a signal sequence.

[0032] The anti-GUCY2C binding domain is provided as a single chain chimeric receptor that is MHC -independent. The antigen-binding domain is derived from an antibody. In some embodiments, CARs comprise anti-guanylyl cyclase C (also referred to as GCC or GUCY2C) single chain variable fragment (scFv) (preferably a Variable Light fragment - (Glycine4Serine)4 Linker - Variable Heavy fragment) from 5F9. 5F9 is a hybridoma expressing a fully humanized, monoclonal antibody that recognizes the extracellular domain of human GUCY2C. The DNA coding sequences of the antibody heavy and light chains were used to create a novel scFv for CAR implementation that is employed in the creation of anti-GCC CARs, such as for example the 5F9-28BBz CAR, and confers antigen specificity directed towards the GUCY2C molecule.

[0033] In some embodiments such as the 5F9-28BBz CAR, the anti-GCC scFv may be a 5F9 single chain variable fragment (scFv) (Variable Light fragment- (Glycine4Serine)4 Linker - Variable Heavy fragment). In some embodiments, the CARs comprise an anti- GCC 5F9 scFv (aa 22-271 of SEQ ID NO:2) including the 5F9 Variable Light chain fragment (22-130) linked to a (Glycine4Serine)4 LINKER (131-150) linked to the 5F9 Variable Heavy chain fragment (151-271). In some embodiments, a (Glycine4Serine)nLINKER wherein n=2-5 may be used. In some embodiments, the 5F9 variable fragments may be configured from N-terminus to C-terminus in the order Variable Light Chain fragment-LINKER- Variable Heavy Chain fragment or Variable Heavy Chain fragment-LINKER-Variable Light Chain fragment. In some embodiments, the CARs comprise an anti-GCC 5F9 scFv configured as [5F9 Variable Light Chain fragment- (Glycine4Serine)2-5F9 Variable Heavy Chain fragment], [5F9 Variable Light Chain fragment-(Glycine4Serine)3-5F9 Variable Heavy Chain fragment] [5F9 Variable Light Chain fragment-(Glycine4Serine)4-5F9 Variable Heavy Chain fragment], or [5F9 Variable Light Chain fragment-(Glycine4Serine)5-5F9 Variable Heavy Chain fragment]. In some embodiments, the CARs comprise an anti-GCC 5F9 scFv configured as [5F9 Variable Heavy Chain fragment-(Glycine4Serine)2-5F9 Variable Light Chain fragment], [5F9 Variable Heavy Chain fragment-(Glycine4Serine)3-5F9 Variable Light Chain fragment], [5F9 Variable Heavy Chain fragment-(Glycine4Serine)4-5F9 Variable Light Chain fragment], or [5F9 Variable Heavy Chain fragment-(Glycine4Serine)5-5F9 Variable Light Chain fragment.

[0034] CARs comprise a CD8a hinge region. In some embodiments, CARs comprise a CD8a hinge region and a CD8a transmembrane region.

[0035] In some embodiments, CARs comprise intracellular (cytoplasmic) sequences from one or more of CD28, 4- IBB (CD 137), CD2, CD27, CD30, CD40L, CD79A, CD79B, CD226, DR3, GITR, HVEM, ICOS, LIGHT, 0X40, or SLAM intracellular region in combination with b) intracellular (cytoplasmic) sequences from ^-chain associated with CD3 (CD3Q, the CD79-alpha and -beta chains of the B cell receptor complex, or certain Fc receptors. In some embodiments, CARs comprise both intracellular (cytoplasmic) sequences from CD28 and 4-1BB (CD137) in combination with intracellular (cytoplasmic) sequences from i^-chain associated with CD3 (CD3Q.

[0036] In some embodiments, CARs comprise a sequence encoding at least one immunoreceptor tyrosine activation motif (IT AM). In some embodiments, CARs comprise a sequence from a cell signaling molecule that comprises ITAMs. Typically 3 IT AMS are present in such sequences. Examples of cell signaling molecules that comprise ITAMs include ^-chain associated with CD3 (CD3Q, the CD79-alpha and -beta chains of the B cell receptor complex, and certain Fc receptors. Accordingly, in some embodiments, CARs comprise a sequence from a cell signaling molecule such as CD3^, the CD79-alpha and -beta chains of the B cell receptor complex, and certain Fc receptors that comprises ITAMs. The sequences included in the CAR are intracellular sequences from such molecules that comprise one of more ITAMs. An ITAM is a conserved sequence of four amino acids that is repeated twice in the cytoplasmic tails of certain cell surface proteins of the immune system. The conserved sequence of four amino sequence of an IT AM contains a tyrosine separated from a leucine or isoleucine by any two other amino acids (YXXL or YXXI in which X is independently any amino acid sequence). The IT AM contains a sequence that is typically 14-16 amino acids having the two four amino acid conserved sequences separated by between about 6 and 8 amino acids.

[0037] In some embodiments, CARs may comprise an immunoglobulin-derived antigen binding domain, antibody sequences that bind to GUCY2C fused to a T cell signaling domain such as the CD3zeta signaling chain of the T cell receptor or a T-cell costimulatory signaling (e.g. CD28) domain linked to a T-cell chain such as CD3zeta chain or the gamma-signal-transducing subunit of the Ig Fc receptor complex.

[0038] The signaling domain of the CAR comprises sequences derived from a TCR. In some embodiments, the CAR comprises an extracellular single chain fragment of antibody variable region that provides antigen binding function fused to a transmembrane and cytoplasmic signaling domain such as CD3zeta chain or CD28 signal domain linked to CD3zeta chain. In some embodiments the signaling domain is linked to the antigen binding domain by a spacer or hinge. When the fragment of antibody variable region binds to GUCY2C, the signaling domain initiates immune cell activation. These recombinant T cells that express membrane bound chimeric receptors comprising an extracellular anti-GUCY2C binding domain and intracellular domain derived from TCRs which perform signaling functions to stimulate lymphocytes. Some embodiments provide anti-GUCY2C binding domain is a single chain variable fragment (scFv) that includes anti-GUCY2C binding regions of the heavy and light chain variable regions of an anti- GUCY2C antibody. A signaling domain may include a T-cell costimulatory signaling (e.g. CD28, 4-1BB (CD137), CD2, CD27, CD30, CD40L, CD79A, CD79B, CD226, DR3, GITR, HVEM, ICOS, LIGHT, 0X40, SLAM) domain and T-cell triggering chain (e.g. CD3zeta).

[0039] In some embodiments, CARs include an affinity tag. Examples of such affinity tags include: Strep-Tag; Strep-Tagll; Poly(His); HA; V5; and FLAG-tag. In some embodiments, the affinity tag may be located before scFv or between scFv and hinge region or after the hinge region. In some embodiments, the affinity tag is selected from Strep-Tag, Strep-Tagll, Poly(His), HA; V5, and FLAG-tag, and is located before scFv or between scFv and hinge region or after the hinge region. In some embodiments, CARs comprise from N terminus to C terminus, a signal sequence, the anti-GCC scFv is a 5F9 single chain variable fragment (scFv), a CD8 alpha hinge region, a CD8 alpha transmembrane region and intracellular sequences from one of more proteins and intracellular sequences and an immunoreceptor tyrosine activation motif, and optionally an affinity tag.

[0040] In some embodiments, CARs comprise from N terminus to C terminus, a signal sequence selected from GM-CSF, CD8 alpha, CD8 beta, CD4, TCR alpha, TCR beta, CD3 delta, CD3 epsilon, CD3 gamma, CD28, BiP linked to the anti-GCC scFv is a 5F9 single chain variable fragment (scFv) selected from (Variable Light Chain fragment- (Glycine4Serine)2-5 Linker - Variable Heavy Chain fragment) and (Variable Heavy Chain fragment-(Glycine4Serine)2-5 Linker - Variable Light Chain fragment), linked to a hinge region selected from CD8a, IgGl-Fc, IgG4-Fc and CD28 hinge regions, linked to a transmembrane region selected from a CD8a, IgGl-Fc, IgG4-Fc and CD28 transmembrane region, linked to intracellular sequences selected from CD284-1BB (CD137), CD2, CD27, CD28, CD30, CD40L, CD79A, CD79B, CD226, DR3, GITR, HVEM, ICOS, LIGHT, 0X40, SLAM intracellular sequences, linked to an immunoreceptor tyrosine activation motif containing sequence selected from CD3^, CD79-alpha, CD79-beta and Fc receptor intracellular sequences that comprise one or more IT AMs, optionally linked to an affinity tag selected from Strep-Tag, Strep-Tagll, Poly(His), HA; V5, and FLAG-tag.

[0041] In some embodiments, CARs comprise from N terminus to C terminus, a CD8 alpha signal sequence, the anti-GCC scFv is a 5F9 single chain variable fragment (scFv) selected from [Variable Light Chain fragment- (Glycine4Serine)2-5 Linker - Variable Heavy Chain fragment] or [Variable Heavy Chain fragment- (Glycine4Serine)2-5 Linker - Variable Light Chain fragment] ), a CD8a, hinge region, a CD8a transmembrane, a CD28 intracellular sequences, 4- IBB intracellular sequences and CD3^ intracellular sequences.

[0042] In some embodiments, the CAR comprises SEQ ID NO:2. In some embodiments, the CAR consists essentially of SEQ ID NO:2. In some embodiments, the CAR consists of SEQ ID NO:2. In some embodiments, the nucleic acid sequence of the construct that encodes the CARs consist of nucleotides comprises SEQ ID NO: 1. In some embodiments, the nucleic acid sequence of the construct that encodes the CARs consist of nucleotides consists essentially of SEQ ID NO:1. In some embodiments, the nucleic acid sequence of the construct that encodes the CARs consist of nucleotides consists of SEQ ID NO:1. In some embodiments, these sequences are linked to regulatory elements necessary for expression of the coding sequence in a human cell such as a human T cell. In some embodiments, a human cell such as a human T cell transformed with the sequences linked to regulatory elements necessary for expression of the coding sequence.

[0043] In some embodiments, the 5F9-28BBz - SEQ ID NO: 1 is linked to regulatory elements necessary for expression of the coding sequence in a human cell such as a human T cell. Regulatory elements necessary for expression of the coding sequence in a human cell such as a human T cell may include a promoter, a polyadenylation site and other sequences in 5’ and 3’ untranslated regions. In some embodiments, SEQ ID NO:1 is inserted in an expression vector such as a plasmid such a pVAX, or a retroviral expression vector such as a lentiviral vector, or a recombinant DNA viral vector such a recombinant adenovirus, recombinant AAV, or recombinant vaccinia virus, or as double stranded DNA to be used with CRISPR / Cas9, TALENs, or other transposon technology or as messenger RNA.

[0044] In some embodiments, CAR coding sequences are introduced ex vivo into cells, such as T cells, including CD4+ and CD8+, invariant Natural Killer T cells, gamma-delta T cells, Natural Killer cells, and myeloid cells, including CD34+ hematopoietic stem cells from peripheral lymphocytes using routine in vitro gene transfer techniques and materials such as retroviral vectors. Following gene transfer, the recombinant cells are cultured to expand the number of recombinant cells which are administered to a patient. The recombinant cells will recognize and bind to cells displaying the antigen recognized by the extracellular antibody-derived antigen binding domain. Following modification, the cells are expanded ex vivo to obtain large numbers of such cells which are administered to the patient have been described. As above, autologous refers to the donor and recipient of the cells being the same person. Allogenic refers to the donor and recipient of the cells being different people. In addition to isolating and expanding populations of antigenspecific T cells by ex vivo culturing, the T cells may be modified after isolating and before expanding populations by having genetic material added to them that encodes proteins such as cytokines, for example IL-2, IL-7, and IL-15.

[0045] A plurality of T cells which recognize at least one epitope of GUCY2C may be obtained by isolating a T cell from a cell donor, transforming it with a nucleic acid molecule that encodes an anti-GUCY2C CAR and, culturing the transformed cell to exponentially expand the number of transformed T cells to produce a plurality of such cells.

[0046] The cell donor may be the individual to whom the expanded population of cells will be administered, i.e. an autologous cell donor. Alternatively, the T cell may be obtained from a cell donor that is a different individual from the individual to whom the T cells will be administered, i.e. an allogenic T cell. If an allogenic T cell is used, it is preferred that the cell donor be type matched, that is identified as expressing the same or nearly the same set of leukocyte antigens as the recipient.

[0047] T cells may be obtained from a cell donor by routine methods including, for example, isolation from blood fractions, particularly the peripheral blood monocyte cell component, or from bone marrow samples.

[0048] Once T cells are obtained from the cell donor, one or more T cells may be transformed with a nucleic acid that encodes an anti-GUCY2C CAR which includes a functional binding fragment of an antibody that binds to at least one epitope of a GUCY2C and a portion that renders the protein, when expressed in a cell such as a T cell, a membrane bound protein.

[0049] The nucleic acid molecule that encodes anti-GUCY2C CAR may be obtained by isolating a B cell that produces antibodies that recognize at least one epitope of GUCY2C from an "antibody gene donor" who has such B cells that produce antibodies that recognizes at least one epitope of GUCY2C. Such antibody gene donors may have B cells that produce antibodies that recognize at least one epitope of a GUCY2C due to an immune response that arises from exposure to an immunogen other than by vaccination or, such antibody gene donors may be identified as those who have received a vaccine which induces production of B cells that produce antibodies that recognize at least one epitope of GUCY2C, i.e. a vaccinated antibody genetic donor. The vaccinated antibody genetic donor may have been previously vaccinated or may be administered a vaccine specifically as part of an effort to generate such B cells that produce antibodies that recognize at least one epitope of GUCY2C for use in a method that comprises transforming T cells with a nucleic acid molecule that encodes an anti-GUCY2C CAR, expanding the cell number, and administering the expanded population of transformed T cells to an individual.

[0050] The antibody gene donor may be the individual who will be the recipient of the transformed T cells or a different individual from the individual who will be the recipient of the transformed T cells. The antibody gene donor may be the same individual as the cell donor or the antibody gene donor may be a different individual than the cell donor. In some embodiments, the cell donor is the recipient of the transformed T cells, and the antibody gene donor is a different individual. In some embodiments, the cell donor is the same individual as the antibody gene donor and is a different individual from the recipient of the transformed T cells. In some embodiments, the cell donor is the same individual as the antibody gene donor and the same individual as the recipient of the transformed T cells.

[0051] The nucleic acid molecule which encodes anti-GUCY2C CAR comprises a coding sequence that encodes functional binding fragment of an antibody that recognizes at least one epitope of GUCY2C linked to a protein sequence that provides for the expressed protein to be a membrane bound protein. The coding sequences are linked so that they encode a single product that is expressed.

[0052] The coding sequence that encodes a functional binding fragment of an antibody that recognizes at least one epitope of GUCY2C may be isolated from a B cell from an antibody gene donor. Such a B cell may be obtained and the genetic information isolated. In some embodiments, the B cells are used to generate hybrid cells which express the antibody and therefore carry the antibody coding sequence. The antibody coding sequence may be determined, cloned and used to make the abnti-GUCY2C CAR. A functional binding fragment of an antibody that recognizes at least one epitope of GUCY2C may include some or all of the antibody protein which when expressed in the transformed T cells retains its binding activity for at least one epitope of GUCY2C.

[0053] The coding sequences for a protein sequence that provides for the expressed protein to be a membrane bound protein may be derived from membrane bound cellular proteins and include the transmembrane domain and, optionally at least a portion of the cytoplasmic domain, and / or a portion of the extracellular domain, and a signal sequence to translocate the expressed protein to the cell membrane.

[0054] The nucleic acid molecule that encodes the anti-GUCY2C CAR, i.e. the anti- GUCY2C CAR coding sequence, may be a DNA or RNA The invention relates to chimeric antigen receptors that bind to guanylyl cyclase C and nucleic acid molecules that encode such chimeric antigen receptors. The invention also relates to cells that comprise such chimeric antigen receptors, to methods of making such chimeric antigen receptors and cells, and to methods of using such cells to treat individuals who are suffering from cancer that has cancer cells which express guanylyl cyclase C and to protect individuals against cancer that has cancer cells which express guanylyl cyclase C.

[0055] Immunotherapy based upon T cells that express chimeric antigen receptors (CARs) has become an emerging modality for treating cancer. CARs are fusion receptors that comprise a domain which functions to provide HLA-independent binding of cell surface target molecules and a signaling domain that can activate host immune cells of various types, typically peripheral blood T cells, which may include populations of cells referred to cytotoxic lymphocytes, cytotoxic T lymphocytes (CTLs), Natural Killer T cells (NKT) and Natural Killer cells (NK) or helper T cells. That is, while typically being introduced into T cells, genetic material encoding CARs may be added to immune cells that are not T cells such as NK cells.

[0056] Guanylyl cyclase C (also referred to interchangeably as GCC or GUCY2C) is a membrane-bound receptor that produces the second messenger cGMP following activation by its hormone ligands guanylin or uroguanylin, regulating intestinal homeostasis, tumorigenesis, and obesity. GUCY2C cell surface expression is confined to luminal surfaces of the intestinal epithelium and a subset of hypothalamic neurons. Its expression is maintained in >95% of colorectal cancer metastases and it is ectopically expressed in tumors that evolve from intestinal metaplasia, including esophageal, gastric, oral, salivary gland and pancreatic cancers.

[0057] The inaccessibility of GUCY2C in the apical membranes of polarized epithelial tissue due to subcellular restriction of GUCY2C, creates a therapeutic opportunity to target metastatic lesions of colorectal origin which have lost apical-basolateral polarization, without concomitant intestinal toxicity.

[0058] A syngeneic, immunocompetent mouse model demonstrated that CAR-T cells targeting murine GUCY2C were effective against colorectal cancer metastatic to lung in the absence of intestinal toxicides. Similarly, other GUCY2C-targeted therapeutics, including antibody-drug conjugates and vaccines, are safe in preclinical animal models, and therapeutic regimens utilizing these platforms are in clinical trials for metastatic esophageal, gastric, pancreatic, and colorectal cancers (NCT02202759, NCT02202785, NCT01972737).

[0059] The safety of these therapeutic regimens, in the context of GUCY2C expression across the rostral-caudal axis of intestine, reflects compartmentalized expression of GUCY2C, enriched in apical, but limited in basolateral, membranes of epithelial cells. Systemic radiolabeled imaging agents conjugated to GUCY2C ligand target GUCY2C- expressing metastases without localizing in intestine, confirming the mucosal compartmentalization of the receptor.

[0060] Tumors express up to 10-fold greater amounts of GUCY2C, compared to normal epithelial cells, potentially creating a quantitative therapeutic window to discriminate receptor overexpressing tumors from intestinal epithelium with low / absent GUCY2C in basolateral membranes.

[0061] U.S. Patent Application Publication 20120251509 Al and U.S. Patent Application Publication US 2014-0294784 Al, which are each incorporated herein by reference, disclose CARs including CARs that bind to guanylyl cyclase C, T cells that comprise CARs including T cells that comprise CARs that bind to GUCY2C and target cells that comprise GUCY2C, methods of making chimeric antigen receptors and T cells, and methods of using T cells that comprise CARs that bind to GUCY2C and target cells that comprise GUCY2C to protect individuals against cancer cells that express GUCY2C and to treat individuals who are suffering from cancer in which cancer cells express GUCY2C.

[0062] Single chain protein sequences that bind to the extracellular domain of human GUCY2C were generated using fragments of the variable light chain and variable heavy chain of an anti-GUCY2C antibody that binds to the extracellular domain of human GUCY2C. A linker sequence connects the variable light chain fragment to the variable heavy chain fragment into a single chain antibody variable fragment fusion protein sequence (scFv) that binds to the extracellular domain of human GUCY2C.

[0063] The scFv is a component in a CAR, which is a larger fusion protein. The CARs functional components include the immunoglobulin-derived antigen binding domain, antibody sequences i.e. svFv, which binds to human GUCY2C, a hinge domain that links the scFV to a transmembrane domain that anchors the protein in the cell membrane of the cell in which it is expressed, and the signally domain which functions as signal transducing intracellular sequences (also referred to as cytoplasmic sequences) that activate the cell upon scFv binding to human GUCY2C. The nucleic acid sequences that encode the CAR include sequences that encode a signal peptide from a cellular protein that facilitate the transport of the translated CAR to the cell membrane. CARs direct the recombinant cells in which they are expressed to bind to and, in the case of recombinant cytotoxic lymphocytes, recombinant cytotoxic T lymphocytes (CTLs), recombinant Natural Killer T cells (NKT), and recombinant Natural Killer cells (NK) kill cells displaying the antibody-specified target, i.e. GUCY2C. When the CAR is expressed, it is transported to the cell surface and the signal peptide is typically removed. The mature CAR functions as a cellular receptor. The scFv and hinge domain are displayed on the cell surface where the scFv sequences can be exposed to proteins on other cells and bind to GUCY2C on such cells. The transmembrance region anchors the CAR in the cell membrane and the intracellular sequences function as a signal domain to transduce a signal in the cell which results in the death of GUCY2C-expressing cell to which the CAR-expressing cell is bound.

[0064] In some embodiments, the CARs comprise a signal sequence, such as for example a mammalian or synthetic signal sequence. In some embodiments, the CARs comprise a signal sequence from a membrane-bound protein such as for example a mammalian membrane-bound protein. In some embodiments, the CARs comprise a signal sequence from a membrane-bound protein such as CD8 alpha, CD8 beta, CD4, TCR alpha, TCR beta, CD3 delta, CD3 epsilon, CD3 gamma, CD28, and BiP. Examples of signal sequences may also be found in membrane bound any mammalian signal sequence <http: / / www.signalpeptide.de / index.php?m=listspdb_mammalia>.

[0065] The anti-GUCY2C binding domain is provided as a single chain chimeric receptor that is MHC -independent. The antigen-binding domain is derived from an antibody. In some embodiments, CARs comprise anti-guanylyl cyclase C (also referred to as GCC or GUCY2C) single chain variable fragment (scFv) (preferably a Variable Light fragment - (Glycine4Serine)4 Linker - Variable Heavy fragment) from 5F9. 5F9 is a hybridoma expressing a fully humanized, monoclonal antibody that recognizes the extracellular domain of human GUCY2C. The DNA coding sequences of the antibody heavy and light chains were used to create a novel scFv for CAR implementation that is employed in the creation of anti-GCC CARs, such as for example the 5F9-28BBz CAR, and confers antigen specificity directed towards the GUCY2C molecule.

[0066] In some embodiments such as the 5F9-28BBz CAR, the anti-GCC scFv may be a 5F9 single chain variable fragment (scFv) (Variable Light fragment- (Glycine4Serine)4 Linker - Variable Heavy fragment).

[0067] CARs comprise a CD8a hinge region and a CD8a transmembrane region.

[0068] In some embodiments, CARs comprise a) intracellular (cytoplasmic) sequences from one or more of CD28, 4- IBB (CD 137), CD2, CD27, CD30, CD40L, CD79A, CD79B, CD226, DR3, GITR, HVEM, ICOS, LIGHT, 0X40, or SLAM intracellular region in combination with b) intracellular (cytoplasmic) sequences from ^-chain associated with CD3 (CD3Q, the CD79-alpha and -beta chains of the B cell receptor complex, or certain Fc receptors.

[0069] In some embodiments, CARs comprise CD28 intracellular sequences together with 4- IBB intracellular sequences in combination with CD3^ intracellular sequences.

[0070] The signaling domain of the CAR comprises sequences derived from a TCR. In some embodiments, the CAR comprises an extracellular single chain fragment of antibody variable region that provides antigen binding function fused to a transmembrane and cytoplasmic signaling domain such as CD3zeta chain or CD28 signal domain linked to CD3zeta chain. In some embodiments the signaling domain is linked to the antigen binding domain by a spacer or hinge. When the fragment of antibody variable region binds to GUCY2C, the signaling domain initiates immune cell activation. These recombinant T cells express membrane bound chimeric receptors comprising an extracellular anti-GUCY2C binding domain and intracellular domain derived from TCRs which perform signaling functions to stimulate lymphocytes. Some embodiments provide anti-GUCY2C binding domain is a single chain variable fragment (scFv) that includes anti-GUCY2C binding regions of the heavy and light chain variable regions of an anti- GUCY2C antibody. A signaling domain may include a T-cell costimulatory signaling (e.g. CD28, 4-1BB (CD137), CD2, CD27, CD30, CD40L, CD79A, CD79B, CD226, DR3, GITR, HVEM, ICOS, LIGHT, 0X40, SLAM) domain and T-cell triggering chain (e.g. CD3zeta).

[0071] In some embodiments, CARs include an affinity tag. Examples of such affinity tags include Strep-Tag; Strep-Tagll; Poly(His); HA; V5; and FLAG-tag. In some embodiments, the affinity tag may be located before scFv or between scFv and hinge region or after the hinge region. In some embodiments, the affinity tag is selected from Strep-Tag, Strep-Tagll, Poly(His), HA; V5, and FLAG-tag, and is located before scFv or between scFv and hinge region or after the hinge region.

[0072] In some embodiments, CARs comprise from N terminus to C terminus, a signal sequence, the anti-GCC scFv is a 5F9 single chain variable fragment (scFv), CD8a hinge region and a CD8a transmembrane region, and intracellular sequences from one of more proteins and intracellular sequences and an immunoreceptor tyrosine activation motif, and optionally an affinity tag. In some embodiments, CARs comprise from N terminus to C terminus, a signal sequence selected from GM-CSF, CD8 alpha, CD8 beta, CD4, TCR alpha, TCR beta, CD3 delta, CD3 epsilon, CD3 gamma, CD28, BiP linked to the anti-GCC scFv is a 5F9 single chain variable fragment (scFv) selected from (Variable Light Chain fragment- (Glycine4Serine)2-5 Linker - Variable Heavy Chain fragment) and (Variable Heavy Chain fragment-(Glycine4Serine)2-5 Linker - Variable Light Chain fragment), linked to a hinge region from CD8a, , linked to a transmembrane region selected from a CD8a, linked to intracellular sequences selected from CD284-1BB (CD 137), CD2, CD27, CD28, CD30, CD40L, CD79A, CD79B, CD226, DR3, GITR, HVEM, ICOS, LIGHT, 0X40, SLAM intracellular sequences, linked to an immunoreceptor tyrosine activation motif containing sequence selected from CD3^, CD79-alpha, CD79-beta and Fc receptor intracellular sequences that comprise one or more IT AMs, optionally linked to an affinity tag selected from Strep-Tag, Strep-Tagll, Poly(His), HA; V5, and FLAG-tag.

[0073] In some embodiments, CARs comprise from N terminus to C terminus, a CD8a signal sequence, the anti-GCC scFv is a 5F9 single chain variable fragment (scFv) selected from [Variable Light Chain fragment- (Glycine4Serine)2-5 Linker - Variable Heavy Chain fragment] or [Variable Heavy Chain fragment- (Glycine4Serine)2-5 Linker - Variable Light Chain fragment] ), a CD8a hinge region, a CD8a transmembrane, CD28 intracellular sequences, 4-1 BB intracellular sequences and CD3^ intracellular sequences.

[0074] In some embodiments, the CAR is encoded by SEQ ID NO:1, a novel DNA sequence, a synthetic receptor that can be expressed by T lymphocytes and infused for the therapeutic treatment of human guanylyl cyclase C (GUCY2C)-expressing malignancies. SEQ ID NO:1 comprises human DNA coding sequences concatenated thusly: (1) CD8a signal sequence, (2) 5F9 single chain variable fragment (scFv) (Variable Light fragment- (Glycine4Serine)4 Linker - Variable Heavy fragment), (3) CD8a hinge region, (4) CD8a transmembrane domain, (5) CD28 intracellular domain, (6) 4- IBB intracellular domain, and (7) CD3^ intracellular domain. In some embodiments, the CAR comprises SEQ ID NO:2. In some embodiments, the CAR consists essentially of SEQ ID NO:2. In some embodiments, the CAR consists of SEQ ID NO:2. In some embodiments, the nucleic acid sequence of the construct that encodes the CARs consist of nucleotides comprises SEQ ID NO: 1. In some embodiments, the nucleic acid sequence of the construct that encodes the CARs consist of nucleotides consists essentially of SEQ ID NO:1. In some embodiments, the nucleic acid sequence of the construct that encodes the CARs consist of nucleotides consists of SEQ ID NO: 1. In some embodiments, these sequences are linked to regulatory elements necessary for expression of the coding sequence in a human cell such as a human T cell. In some embodiments, a human cell such as a human T cell transformed with the sequences linked to regulatory elements necessary for expression of the coding sequence.

[0075] In some embodiments, SEQ ID NO:1 is linked to regulatory elements necessary for expression of the coding sequence in a human cell such as a human T cell. Regulatory elements necessary for expression of the coding sequence in a human cell such as a human T cell may include a promoter, a polyadenylation site and other sequences in 5’ and 3’ untranslated regions. In some embodiments, SEQ ID NO: 1 is inserted in an expression vector such as a plasmid such a pVAX, or a retroviral expression vector such as a lentiviral vector, or a recombinant DNA viral vector such a recombinant adenovirus, recombinant AAV, or recombinant vaccinia virus, or as double stranded DNA to be used with CRISPR / Cas9, TALENs, or other transposon technology or as messenger RNA.

[0076] In some embodiments, CAR coding sequences are introduced ex vivo into cells, such as T cells, including CD4+ and CD8+, invariant Natural Killer T cells, gamma-delta T cells, Natural Killer cells, and myeloid cells, including CD34+ hematopoietic stem cells from peripheral lymphocytes using routine in vitro gene transfer techniques and materials such as retroviral vectors. Following gene transfer, the recombinant cells are cultured to expand the number of recombinant cells which are administered to a patient. The recombinant cells will recognize and bind to cells displaying the antigen recognized by the extracellular antibody-derived antigen binding domain. Following modification, the cells are expanded ex vivo to obtain large numbers of such cell which are administered to the patient have been described. As above, autologous refers to the donor and recipient of the cells being the same person. Allogenic refers to the donor and recipient of the cells being different people. In addition to isolating and expanding populations of antigenspecific T cells by ex vivo culturing, the T cells may be modified after isolating and before expanding populations by having genetic material added to them that encodes proteins such as cytokines, for example IL-2, IL-7, and IL-15.

[0077] A plurality of T cells which recognize at least one epitope of GUCY2C may be obtained by isolating a T cell from a cell donor, transforming it with a nucleic acid molecule that encodes an anti-GUCY2C CAR and, culturing the transformed cell to exponentially expand the number of transformed T cells to produce a plurality of such cells.

[0078] The cell donor may be the individual to whom the expanded population of cells will be administered, i.e. an autologous cell donor. Alternatively, the T cell may be obtained from a cell donor that is a different individual from the individual to whom the T cells will be administered, i.e. an allogenic T cell. If an allogenic T cell is used, it is preferred that the cell donor be type matched, that is identified as expressing the same or nearly the same set of leukocyte antigens as the recipient.

[0079] T cells may be obtained from a cell donor by routine methods including, for example, isolation from blood fractions, particularly the peripheral blood monocyte cell component, or from bone marrow samples.

[0080] Once T cells are obtained from the cell donor, one or more T cells may be transformed with a nucleic acid that encodes an anti-GUCY2C CAR which includes a functional binding fragment of an antibody that binds to at least one epitope of a GUCY2C and a portion that renders the protein, when expressed in a cell such as a T cell, a membrane bound protein.

[0081] The nucleic acid molecule that encodes anti-GUCY2C CAR may be obtained by isolating a B cell that produces antibodies that recognize at least one epitope of GUCY2C from an "antibody gene donor" who has such B cells that produce antibodies that recognizes at least one epitope of GUCY2C. Such antibody gene donors may have B cells that produce antibodies that recognize at least one epitope of a GUCY2C due to an immune response that arises from exposure to an immunogen other than by vaccination or, such antibody gene donors may be identified as those who have received a vaccine which induces production of B cells that produce antibodies that recognize at least one epitope of GUCY2C, i.e. a vaccinated antibody genetic donor. The vaccinated antibody genetic donor may have been previously vaccinated or may be administered a vaccine specifically as part of an effort to generate such B cells that produce antibodies that recognize at least one epitope of GUCY2C for use in a method that comprises transforming T cells with a nucleic acid molecule that encodes an anti-GUCY2C CAR, expanding the cell number, and administering the expanded population of transformed T cells to an individual.

[0082] The antibody gene donor may be the individual who will be the recipient of the transformed T cells or a different individual from the individual who will be the recipient of the transformed T cells. The antibody gene donor may be the same individual as the cell donor or the antibody gene donor may be a different individual than the cell donor. In some embodiments, the cell donor is the recipient of the transformed T cells, and the antibody gene donor is a different individual. In some embodiments, the cell donor is the same individual as the antibody gene donor and is a different individual from the recipient of the transformed T cells. In some embodiments, the cell donor is the same individual as the antibody gene donor and the same individual as the recipient of the transformed T cells.

[0083] The nucleic acid molecule which encodes anti-GUCY2C CAR comprises a coding sequence that encodes functional binding fragment of an antibody that recognizes at least one epitope of GUCY2C linked to a protein sequence that provides for the expressed protein to be a membrane bound protein. The coding sequences are linked so that they encode a single product that is expressed.

[0084] The coding sequence that encodes a functional binding fragment of an antibody that recognizes at least one epitope of GUCY2C may be isolated from a B cell from an antibody gene donor. Such a B cell may be obtained and the genetic information isolated. In some embodiments, the B cells are used to generate hybrid cells which express the antibody and therefore carry the antibody coding sequence. The antibody coding sequence may be determined, cloned and used to make the abnti-GUCY2C CAR. A functional binding fragment of an antibody that recognizes at least one epitope of GUCY2C may include some or all of the antibody protein which when expressed in the transformed T cells retains its binding activity for at least one epitope of GUCY2C.

[0085] The coding sequences for a protein sequence that provides for the expressed protein to be a membrane bound protein may be derived from membrane bound cellular proteins and include the transmembrane domain and, optionally at least a portion of the cytoplasmic domain, and / or a portion of the extracellular domain, and a signal sequence to translocate the expressed protein to the cell membrane. The nucleic acid molecule may be operably linked to the regulatory elements necessary for expression of the coding sequence in a donor T cell. In some embodiments, the nucleic acid molecule that comprises an anti-GUCY2C CAR coding sequence is a plasmid DNA molecule. Tn some embodiments, the nucleic acid molecule that comprises an anti-GUCY2C CAR coding sequence is a plasmid DNA molecule that is an expression vector wherein the coding sequence is operably linked to the regulatory elements in the plasmid that are necessary for expression of the anti-GUCY2C CAR coding sequence in a donor T cell. In some embodiments, a nucleic acid molecule that comprises an anti-GUCY2C CAR coding sequence may be incorporated into viral particle which is used to infect a donor T cell. Packaging technology for preparing such particles is known. The coding sequence incorporated into the particle may be operable linked to regulatory elements in the plasmid that are necessary for expression of the anti-GUCY2C CAR coding sequence in a donor T cell. In some embodiments, the nucleic acid molecule that comprises an anti- GUCY2C CAR coding sequence is incorporated into a viral genome. In some embodiments, the viral genome is incorporated into viral particle which is used to infect a donor T cell. Viral vectors for delivering nucleic acid molecules to cells are well known and include, for example, viral vectors based upon vaccine virus, adenovirus, adeno associated virus, pox virus as well as various retroviruses. The anti-GUCY2C CAR coding sequence incorporated into the viral genome may be operable linked to regulatory elements in the plasmid that are necessary for expression of the anti-GUCY2C CAR coding sequence in a donor T cell.

[0086] Upon expression of the nucleic acid in the transformed T cells, the transformed cells may be tested to identify a T cell that recognizes at least one epitope of GUCY2C. Such transformed T cells may be identified and isolated from the sample using standard techniques. The protein that comprises at least one epitope of GUCY2C may be adhered to a solid support and contacted with the sample. T cells that remain on the surface after washing are then further tested to identify T cells that recognize at least one epitope of GUCY2C. Affinity isolation methods such as columns, labeled protein that binds to the cells, cell sorter technology may also be variously employed. T cells that recognize at least one epitope of GUCY2C may also be identified by their reactivity in the presence of a protein with at least one epitope of GUCY2C.

[0087] Once a T cell is identified as a T cell that recognizes at least one epitope GUCY2C, it may be clonally expanded using tissue culture techniques with conditions that promote and maintain cell growth and division to produce an exponential number of identical cells. The expanded population of T cells may be collected for administration to a patient.

[0088] A plurality of T cells that recognize at least an epitope of GUCY2C according to some embodiments comprise a pharmaceutically acceptable carrier in combination with the cells. Pharmaceutical formulations comprising cells are well known and may be routinely formulated by one having ordinary skill in the art. Suitable pharmaceutical carriers are described in Remington's Pharmaceutical Sciences, A. Osol, a standard reference text in this field, which is incorporated herein by reference. The present invention relates to pharmaceutical composition for infusion.

[0089] In some embodiments, for example, the plurality of cells can be formulated as a suspension in association with a pharmaceutically acceptable vehicle. Examples of such vehicles are water, saline, Ringer's solution, dextrose solution, and 5% human serum albumin. The vehicle may contain additives that maintain isotonicity (e.g., sodium chloride, mannitol) and chemical stability (e.g., buffers and preservatives). The vehicle is sterilized prior to addition of cells by commonly used techniques.

[0090] The plurality of cells may be administered by any means that enables them to come into contact with cancer cells. Pharmaceutical compositions may be administered intravenously for example.

[0091] Dosage varies depending upon the nature of the plurality of cells, the age, health, and weight of the recipient; nature and extent of symptoms, kind of concurrent treatment, frequency of treatment, and the effect desired. Generally, 1 x IO10to 1 x 1012T cells are administered although more or fewer may also be administered, such as 1 x 109to 1 x 1013. Typically, 1 x 1011 T cells are administered. The number of cells delivered is the amount sufficient to induce a protective or therapeutically response. Those having ordinary skill in the art can readily determine the range and optimal dosage by routine methods.

[0092] Patients to be treated with the anti-GUCY2C CARs include patients who have cancer cells that express GUCY2C. In some embodiments, such cancers may be metastatic colorectal cancer, metastatic or primary stomach, metastatic or primary esophageal, metastatic or primary oral, metastatic or primary salivary gland or metastatic or primary pancreatic cancer or any other cancer identified as having GUCY2C expression. In some embodiments, patients suspected of having cancer that includes cancer cells which express GUCY2C are treated with anti-GUCY2C CARs. In some embodiments, prior to treatment with anti-GUCY2C CARs, patients are identified as metastatic colorectal cancer, metastatic or primary stomach, metastatic or primary esophageal, metastatic or primary oral, metastatic or primary salivary gland or metastatic or primary pancreatic cancer patients. In some embodiments, prior to treatment with anti- GUCY2C CARs, samples of cancer from a patient are tested for GUCY2C expression and those patients with cancers that test positive for GUCY2C expression are treated with anti-GUCY2C CARs. In some embodiments, prior to treatment with anti-GUCY2C CARs, a patient undergoes surgery to remove a tumor and a sample of the tumor removed from the patient is tested for GUCY2C expression and those patients with cancers that test positive for GUCY2C expression are treated with anti-GUCY2C CARs.

[0093] The anti-GUCY2C CARs may be useful to prevent cancer in individuals identified at being at an elevated risk of cancer that has cancer cells that express GUCY2C such as metastatic colorectal cancer, metastatic or primary stomach, metastatic or primary esophageal, metastatic or primary oral, metastatic or primary salivary gland or metastatic or primary pancreatic cancer. An individual may be identified at being at an elevated risk of cancer that has cancer cells that express GUCY2C based upon family medical history, genetic background or prior diagnosis of cancer that has cancer cells that express GUCY2C such as metastatic colorectal cancer, metastatic or primary stomach, metastatic or primary esophageal, metastatic or primary oral, metastatic or primary salivary gland or metastatic or primary pancreatic cancer and treatment removing the cancer or treatment resulting in apparent remission or cancer free status.

[0094] EXAMPLES

[0095] Example 1

[0096] Chimeric antigen receptor (CAR) T cell therapy received its first FDA approval in 2017. Since then, an additional five therapies have been approved, all of which are for the treatment of hematological malignancies, but no approvals for any solid tumor targeted CAR. Given that hematological malignancies account for less than 10% of all new cancer diagnoses, there is an urgent, unmet clinical need to improve CAR-T cell therapy to overcome the challenges posed by solid tumors. We have developed a novel, all- encompassing manufacturing process, incorporating both ex vivo CAR-T cell handing as well as the molecular design of the CAR, generating CAR T-cells that are quantitatively and qualitatively superior tumor cell killers compared with conventional manufacturing processes. The critical components of this process include: (1) generation of high titer lentivirus, (2) a unique T-cell media formulation, (3) activation of resting T-cells including the use of CD2 agonism, and (4) the use of a CD8a hinge and transmembrane domain expressing CAR, which benefits by improving target antigen affinity as well as stability through post-translational modifications. The CAR-T cell manufacturing process combines the generation of high titer vesicular stomatitis virus, glycoprotein G, Indiana strain (VSV-G) pseudotyped lentivirus production, a unique T-cell media formulation, non-conventional T-cell activation, culturing, cryopreservation, and thawing. The following are the critical parameters to the manufacturing process.

[0097] Production of Lend viral Vectors

[0098] The cassettes for the CAR transfer plasmids are as follows: pCDH-EFla- h5F9.CD8HTM.28BBz with or without the T2A-GFP sequence (pCAR). E. coli are cultured in LB Broth, Miller supplemented with 1.9% yeast extract and 100 ug / mL Ampicillin. DNA is purified using endotoxin-free purification and DNA is resuspended in endotoxin- free H2O to a final concentration of ~ 1 ug / uL.

[0099] High Titer Lentivirus Production

[0100] A high titer lentivirus is defined as a virus that produces a functional titer of >3.0 x 108TU / mL (Transforming Units per milliliter), determined by transducing HEK293T / 17 cells (ATCC, Cat No. CRL-11268) with virus and using flow cytometric analysis to determine the percentage of lentivirus transduced cells. Standard culture flasks are coated with 5 ug / cm2 poly-d-lysine. 1.26 x 105HEK293T / 17 cells (CRL11268, American Type Culture Collection (ATCC)) are seeded in cell culture medium consisting of Advanced DMEM (12491023, Thermo Fisher Scientific) supplemented with 5% heat- inactivated FBS (A38400-01, Gibco) and IX GlutaMAX (35050-061, Gibco). Lipofectamine 3000 transfection reagent (L3OOO15O, Thermo Fisher Scientific) is used to deliver the lentiviral plasmids in Advanced DMEM (no supplementation). The total weight of plasmids used is 2.5xl0'6pg DNA / lxlO6seeded HEK293T / 17 cells using the following ratio: 2.329 : 1.934 : 4.079 : 1.000 of the following plasmids: pCAR : pRSV- Rev : pMDLg / pRRE : pMD2.G. This reflects a unique and optimized ratio of plasmids from the field. Media is collected at 24- and 52-hours post-transfection, filtered using a 0.45 um aPES filter unit. Lentivirus is concentrated using a 4X Polyethylene Glycol 8000 solution. Viral supernatants with PEG concentrator are incubated on a wave rotator at 4°C overnight followed by centrifugation at 1600xg for 1 hour at 4°C. Lentiviral pellets are resuspended at a 200X concentration of the original supernatant volume in lentivirus storage buffer: lOmM Tris, pH 7.4, 10% lactose, and 25mM Proline in DPBS without calcium and magnesium. Virus is stored at -80°C. Viral titers, performed using HEK293T-17 cells, must be a minimum of >108Transforming Units / mL to pass quality control and proceed to CAR-T cell production.

[0101] CAR-T Cell Production

[0102] The unique formulation of T-cell media that we use is referred to as RPMI-ITS. T cell culture medium is composed of RPMI-1640, IX with L- glutamine & 25 mM HEPES (10-041-CV, Corning) supplemented with the following components:

[0103] 1. 10% heat- inactivated serum

[0104] 2. Insulin-Transferrin-Selenium (ITS-G, 41400-045, Gibco) lOmM N-Acetyl-L-cysteine (A9165, Millipore Sigma)

[0105] 3. IX GlutaMAX (35050-061, Gibco)

[0106] 4. IX Glucose solution (A24940-01, Gibco)

[0107] 5. IX Sodium Pyruvate (11360-070, Gibco)

[0108] 6. IX MEM Non-Essential Amino Acids (11140-050, Gibco)

[0109] 7. IX HEPES Buffer (15630-080, Gibco), (note: IX in addition to the existing media formulation)

[0110] 8. IX Penicillin-Streptomycin (15140-122, Gibco)

[0111] 9. 55 uM 2-Mercaptoethanol (21985-023, Gibco).

[0112] The final pH of the complete media formulation is pH 7.5.

[0113] Process Day 1: T-cells are cultured in RPMI-ITS. The starting seeding density of enriched human T cells (CD4+and CD8+combined in any ratio) is 4.17x10scells / cm2in a cell concentration of IxlO6cells / mL. T cells must be activated using anti-CD3 (Clone OKT3), anti-CD28 (Clone 15E8), and anti-CD2 (Clone LT2) magnetic beads. Antibodies are combined, by weight, in a 1 : 1 : 1 ratio onto magnetic beads. The bead to T-cell ration is 1: 1. T-cells are cultured with 10 ng / mL human IL-7 and 10 ng / mL human IL- 15.

[0114] Process Day 2: Between 18 to 24 hours following T-cell activation, T cells are transduced with 200X concentrated LV at an MOI of 5 with 0.8 ug / mL Polybrene. It is critical that the T-cell culture system is mixed thoroughly with the virus for efficient transduction.

[0115] Process Day 4: T-cell activation beads are removed using magnetic separation. The RPMI-ITS media is replaced in its entirety. T cells are transferred to the G-Rex culture system. The G-Rex culture system used should have the same surface area as the initial starting culture on Process Day 1. lOng / mL IL-7 and lOng / mL IL-15 are added to the culture media. Process Day 7: lOng / mL IL-7 and lOng / mL IL- 15 are added to the culture media. Process Day 10: Between 70 - 75% of the RPMI-ITS culture media is exchanged. lOng / mL IL-7 and lOng / mL IL- 15 are added to the culture media.

[0116] Process Day 13: CAR-T cells are harvested and cryopreserved using CryoStor CS 10 (07930, Stemcell Technologies) at a density of 20 million cells / mL.

[0117] Prior to infusion, CAR-T cells are thawed in by adding dropwise as a rate of ImL / minute RPML1640 (no supplementation) at 37 °C. A full media exchange is performed and CAR-T cell product is returned to a G-Rex vessel for a 3 day post- cryopreservation recovery period. Then the CAR-T cell product is harvested, washed, and infused in a final infusion buffer.

[0118] CD2 Agonism Promotes Superior Killing and T Cell Phenotypes

[0119] The use of anti-CD3 and anti-CD28 antibodies to activate and expand human T- cells has been in use for over 30 years and has been the standard method of activating T- cells for CAR-T cell therapy. We have found that inclusion of an additional antibody against CD2, promotes the expansion of T-cells with superior anti-tumor qualities. We find at the end of the two week culture period that the combination of CD3, CD28, and CD2 antibody agonism promotes greater expansion of ex vivo T-cells, as well as an increase in the amount of surface CAR expression; however, the viability of T-cell products is not affected (Fig. 1). The inclusion of CD2 into the activation complex improves the quality of the T-cell phenotypes. It has been demonstrated that less differentiated memory cell phenotypes (naive / stem-like (CD45RA+ / CCR7+) and central memory (CD45RA7CCR7+)) compared with more differentiated memory cell phenotypes (effector memory (CD45RA7CCR7 ) and effector (CD45RA+ / CCR7 )) promote improved outcomes in vivo and clinically. We demonstrate that the inclusion of CD2 agonism promotes a statistically significant increase in the more favorable, less differentiated memory cell phenotypes, having significantly more cells expressing naive / stem-like and central memory phenotypic markers and less effector memory markers compared with cells activated without the use of CD2 (Fig. 2A-B). It has also been demonstrated that T- cells with more exhaustion markers at the time of infusion result in worse patient outcomes. T-cells that are activated with CD2 agonism have a reduction in the exhaustion markers CD39 and PD-1 in both the CD8+and CD4+T-cell populations (Fig. 2C). The improved memory T-cell phenotypes and reduction in T-cell exhaustion has the functional outcome of arming the CAR-T cells to be able to kill tumor target cells in vitro when there is an abundance of tumor cells to CAR-T cells. Here the effector to target ratio is 0.1:1. Here we demonstrate that CAR-T cells, regardless of CAR design, are capable of killing the colorectal cancer (CRC) cell line T84, when the T-cells are activated with CD2 (Fig. 3). T-cells that are activated with CD3 and CD28 only are unable to kill target cells.

[0120] CD8a Palmitoylation Site Promotes Increased Surface CAR Expression

[0121] Palmitoylation, the addition of the fatty acid palmitate to a protein, is a post- translational modification that promotes protein stability as well as membrane trafficking. The addition of palmitate groups to proteins localized to the T-cell synapse play an important roles in the signaling cascade emanating from the T cell receptor. CD8a is one such molecule within the T-cell synapse that is palmitoylated at a cysteine residue on the inner leaflet of the plasma membrane. This palmitoylation site plays an important role in the aggregation of CD8 into lipid rafts. The CAR our laboratory has designed that encodes for the CD8a hinge and transmembrane domain, also includes this cysteine residue that gets palmitoylated. We have made a CAR that contains a single nucleotide substation at this palmitoylation site, exchanging the cysteine residue for an alanine residue, inhibiting the ability of the CAR to be palmitoylated. When we do this, we observe a significant decrease in the level of surface CAR expression, suggesting that this palmitoylation site promotes increased stability and retention on the cell surface (Fig. 4). Summary of the Results

[0122] Solid tumor-directed CAR-T cells will only begin to realize their therapeutic potential once all components of the CAR-T cell have been fully optimized. Here we present a complete production process that results in CAR-T cells with superior antitumor efficacy (Fig. 3, Example 2 below). Every component of the manufacturing process is critical to the development of these superior performing T-cells, including: the generation of high titer lend virus, nutrient rich T-cell media, activation with CD3, CD28, and CD2 agonism, supplementation with the cytokines IL-7 and IL- 15, expansion of the T-cells in the G-Rex platform, a defined cryopreservation and thawing procedure, and designing the CAR to include the hinge and transmembrane domains from the CD8a molecule. Combined, these elements produce highly optimized CAR-T cells capable of controlling and eliminating CRC metastatic lesions.

[0123] Example 2 Despite success in treating some hematological malignancies, CAR-T cells have not yet produced similar outcomes in solid tumors due, in part, to the tumor microenvironment, poor persistence, and a paucity of suitable target antigens. Importantly, the impact of the CAR components on these challenges remains focused on the intracellular signaling and antigen-binding domains. In contrast, the flexible hinge and transmembrane domains have been commoditized and are the least studied components of the CAR. Here, we compared the hinge and transmembrane domain derived from either the CD8a or CD28 molecule in identical GUCY2C-targeted third-generation designs for colorectal cancer. While the structural domains do not contribute to differences in antigen-independent contexts, such as CAR expression and differentiation and exhaustion phenotypes, the CD8a structural domains have greater affinity for GUCY2C. This results in increased production of inflammatory cytokines and granzyme B, improved cytolytic effector function with low antigen-expressing tumor cells, and robust anti-tumor efficacy in vivo compared with the CD28 structural domain CAR. This suggests that CD8a domains should be considered in the design of all C ARs for the generation of high- affinity CARs and optimally effective CAR-T cells in solid tumor immunotherapy.

[0124] INTRODUCTION

[0125] In the seven years since chimeric antigen receptor (CAR) T-cell therapy was approved by the FDA for the treatment of hematological malignancies, there have been no approvals for any CAR-T cell therapy directed towards a solid tumor antigen. The consensus within the field has identified three major obstacles that contribute to the inability of CAR-T cells to successfully eradicate solid tumor lesions: (1) trafficking of the CAR-T cells to, and ingress of T cells within, the solid tumor space, (2) the persistence and concomitant functions of CAR-T cells within the suppressive tumor microenvironment, and (3) identifying a suitable tumor antigen that promotes tumor eradication while sparing the surrounding healthy tissue.(l,2)

[0126] In the context of colorectal cancer (CRC), it has been demonstrated clinically, both with checkpoint blocking therapy and adoptively transferred autologous T cells, that metastatic lesions arising from the colon can be controlled by cytolytic T cells. (3-5) However, these therapeutic options are limited by specific disease states as well as loss of the antigen-presenting major histocompatibility complex (MHC) molecule. Moreover, T- cell therapy utilizing the T-cell receptor (TCR) as the tumor-antigen targeting modality requires a degree of personalization above the requirements of CAR-T cells, which use an antibody-derived targeting modality against a native cell surface molecule for broad application to the patient population.

[0127] Our laboratory identified guanylyl cyclase C (GUCY2C) as a mucosal antigen within the intestinal epithelium that can be targeted using immunotherapies. (6,7) GUCY2C was identified as a particulate cyclase and molecularly characterized as the receptor for the E. coli heat-stable enterotoxin ST, responsible for diarrheal disease.(8- 10) The luminal restriction of the extracellular domain of this type 1 transmembrane receptor can be leveraged therapeutically, as GUCY2C-expressing metastatic lesions originating from primary colonic adenocarcinoma are no longer topologically sequestered within the luminal space, making the extracellular domain a viable antigenic target for systemically delivered therapies. (11) While the cyclase activity of GUCY2C is silenced in CRC, receptor expression is maintained in >95% of cases. (12) Immunotherapeutic modalities including vaccines, antibody-drug conjugates, bi-specific T-cell engagers, and CAR-T cells have demonstrated efficacy in controlling or eliminating GUCY2C- expressing metastatic lesions in animal models. (13-19) GUCY2C has also proven to be a safe target in a mouse model of a GUCY2C-directed syngeneic CAR-T cell therapy, and there was no toxicity attributable to the targeted therapy.(20) Moreover, in a clinical trial of a GUCY2C-targeted adenoviral cancer vaccine there were no adverse events reported greater than grade 1, suggesting that T cells directed towards GUCY2C do not promote intestinal colitis. (21)

[0128] The general design of CARs for expression by T cells has remained relatively homogenous since their inception, consisting of (1) an antigen- targeting domain, typically derived from an antibody and organized into a single molecular framework called a single chain variable fragment (scFv), (2) structural domains consisting of a flexible molecular hinge and transmembrane domain (HTM), and (3) intracellular signaling domains (ICDs) derived from T -cell proteins localized to the immunological synapse. In CAR design, the largest variability is with the scFv, due to the array of different antigenic targets, and then the ICDs, which have predominantly consisted of CD3(,, CD28, and 4-1BB. The least modified region, and the least well understood in terms of the role these domains play in CAR efficacy, are the structural domains. The most prominently used domains come from either CD8a or CD28; however, other domains derived from IgG4 have been utilized. The CARs that led to the first two FDA-approved therapies targeting CD 19 using the same scFv were combined with either the CD8a hinge and transmembrane domain plus the 4-1BB and CD3^ ICDs (Tisagenlecleucel, Kymriah, Novartis AG) or the CD28 hinge and transmembrane domain with CD28 and CD3^ ICDs (Axicabtagene Ciloleucel, Yescarta, Kite Pharma, Inc.), referred to as either a BB^ or a 28 design based upon the ICD configuration.(22,23)

[0129] Having identified a human GUCY2C-specific scFv and performed mouse studies using a CD28, 4-1BB, and CD3^ (28BBQ ICD design, we sought to determine if the previously overlooked structural domains, derived from either CD8a or CD28, had an impact on CAR efficacy and CAR-T cell phenotypes. We designed two CARs, which we call the CD8HTM or the CD28HTM CAR, that differed only in the structural domains and evaluated the in vitro and in vivo performance of these two receptors for the treatment of metastatic CRC.

[0130] RESULTS

[0131] CAR-T Cell Manufacturing Parameters Are Not Impacted by Structural Domains

[0132] To assess the role that the hinge and transmembrane domains have on CAR-T cell efficacy, we employed a third-generation design which we have previously demonstrated exhibits robust anti-tumor efficacy and safety in mouse models.(19,20) The CAR expression cassette consists of an EFla promoter and the following human sequences: CD8a leader sequence, anti-GUCY2C (Clone 5F9) single chain variable fragment (scFv) with a (G4S)4 linker, hinge and transmembrane domains (HTM) from CD8a (CD8HTM) or CD28 (CD28HTM), the CD28 ICD, 4-1BB ICD, and CD3 ICD, followed by the virally-derived T2A self-cleavable peptide and green fluorescent protein (GFP) reporter (Fig. 5 A). CAR-T cells were expanded ex vivo for 12 days in G-Rex plates before cryopreservation. At the conclusion of this culture period, there was no difference in the expansion of the CAR-T cell products between the CD8HTM and CD28HTM CARs (Fig. 5B). We also observed no impact of the structural domains on T-cell viability (Fig. 5C). The GFP reporter was used as a marker of lentivirus-transduced T cells in flow cytometric analyses. Of GFP+ CD3+ T-cells, the structural domains did not impact the expansion of the CD8+ or CD4+ co-receptor subsets (Fig. 5D). While the majority of donors had a greater percentage of CD8+ T cells after expansion, one donor exhibited a greater expansion of the CD4+ population relative to the CD8+ population; however, the other characteristics of the T cells from this donor did not differ from the other donors. For each CAR construct, we observed a statistically significant difference in the transduction of CD8+ and CD4+ T cells, with the CD4+ T cells having a higher percentage of GFP+ cells (Fig 5E-F) and higher levels of construct expression (GFP MFI; Fig. 5G) in both the CD8HTM and CD28HTM CAR-T cells. This significant difference in the transduced efficiency between CD8+ and CD4+ T cells has been reported previously in CAR-T cell studies using VSV-G pseudotyped lentivirus (24,25) and necessitates that these two populations be examined separately. Moreover, the transduction efficiency of CD8+ T cells and CD4+ T cells is slightly higher with the CD28HTM than CD8HTM CAR (Fig. 5H); however, the expression of each construct (GFP MFI) was equivalent between the CD8HTM and CD28HTM CARs (Fig. 51).

[0133] Structural Domains Do Not Affect Antigen-Independent T-cell Phenotypes

[0134] It has been demonstrated that the quality of infused T-cell products has a significant impact on patient outcomes. (26) Less differentiated memory phenotypes, such as naive-like / stem cell memory-like and central memory phenotypes, are favored over more differentiated effector memory and effector T-cells. (27) Similarly, T cells expressing high levels of exhaustion markers at baseline also lead to poor therapeutic outcomes. (28) Although components of the CAR such as the scFv and ICDs can contribute to memory phenotypes and exhaustion, it is unclear what role the structural domains have in the outcome of T-cell phenotypes. It should be noted that these phenotypes occur independently of target- antigen exposure. Although T cells are activated using anti-CD3 / CD28 / CD2 antibody agonism, antigen-dependent signaling mediated through the CAR has not yet occurred for these CAR T cells. Differences in memory T-cell phenotypes influenced by the CAR have been shown to be mediated by differences in intracellular signaling domains. (29) High tonic signaling through the CAR, leading to increased T-cell exhaustion, has been shown to be mediated by intrinsic qualities of the scFv.(30,31) Given that the CD8HTM and CD28HTM CARs have identical scFvs, ICDs, and surface-level expression, we hypothesized that these antigenindependent T-cell phenotypes would be similar between the two CARs.

[0135] The four major memory cell populations were defined using the markers CD45RA and CCR7. (32-34) In comparing the four major memory cell phenotypes: naive / stem cell memory-like (N / Tscm), central memory (CM), effector memory (EM), and effector (Eff) cells, we observe no statistically significant difference between the CD8HTM and CD28HTM CAR-T cells in the CD8+ or CD4+ T-cell populations (Fig. 6A). In terms of exhaustion marker expression, no statistically significant difference was observed between the CD8HTM and CD28HTM CAR-T cells as measured by the number of exhaustion markers expressed per cell in both the CD8+ and CD4+ T-cell populations (Fig. 6B). The majority of cells do not express any exhaustion markers (-75-85% negative across CARs and T-cell subsets). Of the cells producing at least one exhaustion marker, that marker, as evidenced by the Simplified Presentation of Incredibly Complex Evaluations (SPICE) plots, is CD39 (Fig. 6B). When total CD39 expression is evaluated between the CD8HTM and CD28HTM CAR-T cells, there is no statistically significant difference in either the CD8+ (p = 0. 1924) or CD4+ (p = 0.2042) T-cell populations.

[0136] T Cells Display Similar Amounts of CD8HTM and CD28HTM CARs

[0137] Protein L (Peptostreptococcus magnus) was used to measure CAR surface levels (Fig. 7A-B). Protein L binds to the kappa light chain of antibodies, allowing for an antigen-independent method for detecting CAR by binding to the light chain of the scFv, revealing comparable levels of CD8HTM and CD28HTM CAR among CD8+ (Fig. 7A) and CD4+ (Fig. 7B) T cells. The two antibody variable domains that comprise the scFv are connected by a flexible linker composed of four glycine residues and one serine. This peptide subunit is then concatenated four times to produce the (G4S)4 flexible linker. Similar to Protein L, an antibody directed against this peptide linker can be used to identify CAR molecules on the surface of the T cell, independent of antigen binding, through linker binding (Fig. 37C-F). The CD8HTM and CD28HTM CAR designs are represented in equivalent amounts on the surface CD8+ (Fig. 7C) and CD4+ (Fig. 7D) T cells, determined by G4S antibody MFI. Using a PE-labeled anti-G4S antibody, we were able to quantify the number of CAR molecules on the T-cell surface, revealing similar levels of CD8HTM and CD28HTM CARs in CD8+ (Fig. 7E) and CD4+ (Fig. 7F) T cells. Given that CD4+ T cells have a higher percentage of GFP positive cells with higher MFIs compared with matched CD8+ T-cells (Fig. 5), we expect that, given the stoichiometric equivalency of proteins produced upstream and downstream of the T2A cleavage site, CD4+ T cells would have more CAR on their surface, which is indeed what we observed. CAR molecules per T cell were quantified with the mean molecules / cell for CD8+ T cells being 3906 and for the CD4+ T cells being 10365 (p<0.0001), regardless of CAR design. Antigen Binding is Improved using the CD8 Hinge and Transmembrane Domains

[0138] With the assessment of CAR surface expression performed using labeling methods independent of target antigen binding, we next wanted to ask if the structural domains of the CD8HTM and CD28HTM CARs affect the affinity of the scFv for its antigenic target, GUCY2C (Fig. 8A-B). CD8HTM and CD28HTM CAR-T cells were incubated with a wide range of concentrations of the soluble extracellular domain of GUCY2C to establish a binding curve. GUCY2C binding to the CAR was measured using flow cytometry. MFI values were normalized within each CAR to generate relative binding affinities for comparisons. This revealed a difference in affinity for GUCY2C between the two CARs, with the CD8HTM CAR having a higher affinity for antigen (Fig. 8A-B). Non-linear regression modeling revealed the CD8HTM CAR to have an EC50 of 7.78-8.26 nM, whereas the CD28HTM CAR had an EC50 of 11.3-13.0 nM in CD8+ and CD4+ T cells (Fig. 8A-B). While there is a significant difference in antigen affinity (Fig. 8B), plotting the dose with actual (non-normalized) binding reveals a similar total receptor occupancy (Bmax), when saturated concentrations of ligand are used (Fig. 8C). This corroborates our findings using antigen-independent detection methods (Fig. 7), demonstrating that T cells are expressing equivalent amounts of CAR molecules on the T- cell surface. We next determined if there were also affinity differences within a single CAR design between the CD8+ and CD4+ T-cell populations (Fig. 8D). As expected, non-linear regression modeling revealed that there is no difference in CD8HTM or CD28HTM affinity between CD8+ and CD4+ T-cell populations (Fig. 8D). Therefore, it is the structural domains themselves that contribute to the difference in antigen affinity (Fig. 8A-B), and not the quantity of CAR receptor or intrinsic differences inherent to T- cell co-receptor subsets.

[0139] Antigen Exposure Results in a Higher Level of Effector Cytokine Production in CD8HTM CAR-T Cells

[0140] Although target cell lysis is the primary effector function of CAR-T cells, the production of inflammatory effector molecules is critical for the bolstering and sustainability of the anti-tumor response.(35) To assess the ability of these two CARs to produce polyfunctional inflammatory cytokine responses upon antigen stimulation, intracellular cytokine production was measured after six hours of plate -bound antigen exposure, followed by intracellular cytokine staining, and flow cytometry. Overall, the production of inflammatory cytokines is not robust with this specific CAR; however, we observed a statistically significant improvement in the ability of the CD8HTM CAR to produce cytokines after stimulation as compared with the CD28HTM CAR (Fig. 9A-B). However, the fractions of the population representing the production of one, two, or three inflammatory cytokines are equivalent between the two CARs (Fig. 9A-B). Of the CD8+ CAR-T cells that produce cytokines upon antigen encounter, we used SPICE plots to demonstrate the similarities in polyfunctionality, in not only the number of cytokines produced within a specific cell but also the specific combinations of TNFa, IFNy, and IL- 2 that contribute to this polyfunctional cytokine response (Fig. 9B). In comparing the cytokine production within the total CD8+ T-cell population, CD8HTM and CD28HTM CAR-T cells produce similar amounts of TNFa (Fig. 9C). However, there are more IL-2 and IFNy producing CD8HTM CAR-T cells than the CD28HTM CAR-T cells (Fig. 9C).

[0141] The CD4+ CAR-T cells, with more CAR molecules expressed on the T-cell surface, have an expectedly more robust cytokine response compared with the CD8+ T- cells. However, like the CD8+ T cell population, we observe the same increase in CD8HTM CAR-T cells able to produce >1 cytokine compared to the CD28HTM CAR-T cells (Fig. 9D). Like CD8+ T cells, the degree of polyfunctionality between the two CARs is equivalent (Fig. 9D-E). Unlike the CD8+ T-cell population, there was a statistically significant increase in the percentage of the CD4+ T-cell population producing each of the three inflammatory cytokines that were evaluated (Fig. 9F). In addition to the inflammatory cytokines, we also observed that CD8+ CAR-T cells expressing the CD8HTM produce significantly more of the cytolytic effector molecule granzyme B after antigen stimulation than those expressing the CD28HTM CAR (Fig. 9G). Interestingly, CD4+ CAR-T cells are also capable of executing cytolytic activity, and we see a statistically significant increase in the CD4+ CD8HTM CAR-T producing granzyme B after antigen exposure (Fig. 9H). Although the granzyme B production is expectedly not as robust as in the CD8+ T cells, GUCY2C exposure results in significantly more granzyme B+ CD8HTM CAR T-cells than CD28HTM CAR-T cells. It should be noted that both CD8HTM and CD28HTM CAR-T cells have equivalent granzyme B+ percentages of the population with control peptide stimulation.

[0142] CD8HTM CAR Demonstrates Superior In Vitro Killing when Target Antigen is Lowly Expressed

[0143] To assess if the higher antigen affinity observed in CD8HTM CARs confers a functional benefit in vitro, we identified metastatic CRC cell lines with a wide range of GUCY2C expression to test with in vitro killing assays: LoVo (low-expression), LS174T (medium expression), and T84 (high-expression) using RT-PCR analysis (Fig. 10 A). Western blot analysis revealed a similar pattern of expression between GUCY2C mRNA and protein levels (Fig. 10B). GUCY2C protein appears as a doublet with the two bands between 130 to 150 kDa. We hypothesized that the differences in affinity for GUCY2C would be revealed inversely to the antigen density. In the high GUCY2C expressing T84 cells, there is no difference in the ability of the CD8HTM and CD28HTM CARs to kill in vitro as demonstrated using the real-time cell killing xCELLigence assay (Fig. IOC). Both CARs are able to kill the total T84 population within 12 hours of adding the CAR-T cells, with equivalent time to 80% killing of the target cell population (KT80) of ~9 hours for each CAR (Fig. 10C). In the medium GLJCY2C expressing cell line (LS174T), there was an expected slowing of the tumor killing kinetics (Fig. 10D). Moreover, there is a noticeable divergence in the curves where the CD8HTM CAR-T appears to have more complete killing of the target cells than the CD28HT CAR-T cells (Fig. 10D). Although the CD8HTM CAR was 3 hours faster in achieving the KT80, the difference was not statistically significant (p = 0.0890, Fig. 10D). This suggested that LS174T cells are on the threshold of antigen density, where lower amounts of target antigen will result in significant effects. Cytolysis of the low-expressing LoVo cells is slower than T84 and LS174T and reveals a significant divergence between CD8HTM and CD28HTM CARs (Fig. 10E). Importantly, there is a strong correlation between antigen density (GUCY2C protein levels) and CAR efficacy (KT80; Fig. 10F). Moreover, linear regression demonstrates an increasing difference between CD8HTM and CD28HTM CARs at decreasing antigen densities (Fig. 10F).

[0144] CD8HTM Structural Domain CARs Exhibit Robust and Durable Anti-Tumor Responses

[0145] We have demonstrated that functional differences between the two CARs are revealed when the CAR is engaging antigen (affinity, cytokine production, granzyme B induction, and cytolysis). In vitro, this sensitivity is revealed in the context of low antigen densities; however, in vitro killing is not a rate limiting step for most CAR-T cells, as even the very low antigen-expressing LoVo cells were successfully killed by both CARs. CAR-T cell efficacy is the culmination of cumulative T-cell responses that inform this living drug to functionally eradicate tumors. To assess whether there are inherent differences between these two CAR designs in an in vivo tumor model, we decided to evaluate these two CAR T cells in a system where antigen was not a limiting factor of efficacy, purposefully biasing the system towards parity. We selected the T84 cell line, with the highest expression level of the antigen, and where, in vitro, there was no difference in the ability of these CAR-T cells to completely eradicate the tumor cells. T84 cells were engineered to express Click Beetle Red luciferase with a T2A self- cleavable peptide sequence followed by the mCherry fluorescent reporter (used for selection). NOD-scid-gamma (NSG) mice were injected with 2.5 million T84 cells into the intraperitoneal cavity, modeling peritoneal metastases. (36) Fourteen days after tumor administration, animals received 3 million CD8HTM or CD28HTM CAR-T cells, and 4 animals received vehicle. An additional 5 mice were injected with a vehicle on the day of the tumor implant to establish the daily luminescence level of tumor-free animals. All animals were imaged bi-weekly to assess tumor burden by luminescence (Fig. 11 A). Tumor-challenged animals with luminescence signals falling within the range (per imaging day) of these vehicle-treated animals would be deemed to have an undetectable signal.

[0146] Animals were imaged on the day prior to CAR-T cell treatment and rank-ordered by luminescence intensity. The four animals surrounding the median of the list were assigned to the untreated group, so that they would have comparable levels of tumor burden. The remaining animals were assigned to groups using a block randomization design, so that both CAR-T treatment groups would have representative animals across the range of tumor burdens. Comparing the luminescence signal one day before treatment, both groups were evenly distributed (Fig. 1 IB).

[0147] In vivo bioluminescence imaging was performed bi-weekly, revealing a robust reduction in the luminescence signal 11 days after CAR-T cell treatment (Fig. 11C-D). As expected, the signal from untreated animals increased over the course of the experiment. Importantly, the CD8HTM CAR-T cells produced robust tumor elimination compared to the CD28HTM CAR-T cells (Fig. 11C-D). Using the tumor-free animals as the baseline luminescence of a mouse, we next wanted to ask about the rate of tumor clearance between the CD8HTM and CD28HTM CAR-T cells. A Kaplan-Meier curve was employed to demonstrate the first day at which an animal had a luminescence signal comparable to the tumor-free animals (Fig. HE). There was a 21-day period between when the first CD8HTM treated mouse became undetectable (Day 7) versus the first CD28HTM treated mouse (Day 28). In total, tumors were cleared in 70.6% of the CD8HTM-treated group, while only 1 1 .8% of the CD28HTM-treated animals were cleared (Fig. 1 IE). These data demonstrate that the hinge and transmembrane domains derived from the CD8a molecule have superior antitumor efficacy compared with the CD28HTM design for this anti-GUCY2C CAR. DISCUSSION

[0148] Optimal CAR design is paramount to efficacy, yet empirical comparisons are often limited to comparisons of scFv or ICD configurations, with limited comparison of structural domains thought to have modest impacts on CAR function. Here, we revealed that a CD8a hinge and transmembrane provides superior avidity, in vitro effector function, and in vivo antitumor efficacy targeting the CRC antigen GUCY2C in the context of a “3rd generation” 28BB CAR. It has been demonstrated that independently, the co-stimulatory domains contribute different functional characteristics to CAR-T cells, with the CD28 ICD promoting faster cytolytic kinetics while the 4- IBB ICD improves persistence. (29,37-39) In evaluating the contributions made by the structural domains to the function and efficacy of the CAR-T cells, parity in the design was paramount to revealing the contribution made by either the CD8HTM or CD28HTM CAR, leading us to include both the CD28 and 4- IBB co-stimulatory domains in the design. Moreover, in the context of controlling solid tumors, the 28BBC design has demonstrated in vivo efficacy in controlling bulky metastases as well as GUCY2C-expressing tumors. (19, 40) In clinical trials using a 28BB design directed towards CD19, these CARs have been effective and safe; however, there is a shortage of true comparative studies between second- and third-generation designs utilized clinically. (41,42) We also opted for this study to treat the hinge and transmembrane domains as a structural unit, as is the current clinical practice utilizing the hinge and transmembrane domain from either CD8a or CD28.

[0149] As we expected, the two CARs did not demonstrate any variability in the final product of the CAR-T cells at the end of the culture period in terms of expansion or viability. In fact, in all CAR-independent T-cell functions, including memory and exhaustion phenotypes, there were no observed differences. In the manufactured product, we did observe a statistically significant difference in both the transduction of CD8+ and CD4+ T cells and in the percentage of GFP+ T cells between the two CARs. The higher transduction efficiency observed in CD4+ T cells has been reported in the literature before (24,25,43); however, a mechanism has not yet been elucidated. We speculate that this difference is specific to vesicular stomatitis virus glycoprotein (VSV-G) pseudotyped lentivirus. In 2013, the low-density lipoprotein receptor (LDL-R) was identified as the host cell receptor for VSV-G viral entry. RNAseq data from The Human Protein Atlas (proteinatlas.org) demonstrates that resting CD8+ T cells have 4.2 LDL-R transcripts per million (TPM) and CD4+ T cells have 6.2 LDL-R TPM. When T cells are activated, LDL-R expression is upregulated in both CD8+ and CD4+ T cells, producing 86.2 and 112.6 TPM, respectively (44). The increased transcript level produced by the activated CD4+ T cells relative to the activated CD8+ T cells present the most plausible explanation for the difference in transduction efficiency. The low level of LDL-R on resting T cells also explains why T cells require activation for efficient transduction. In fact, Natural Killer (NK) cells, which are more resistant to lenti viral transduction than primary human T cells, show improved transduction after treatment with rosuvastatin for the express purpose of increasing LDL-R cell surface expression and promoting improved viral entry (45). With respect to the CD28HTM CAR having a higher percentage of GFP+ T cells compared with the CD8HTM CAR for both the CD8+ and CD4+ CAR-T cells, this difference is most likely due to variability in the quantification of viral titers as well as the direct administration of the lentivirus to the cells. When comparing the GFP MFI of the CD8HTM and CD28HTM CAR-T cells, which is a reflection of the per-cell fluorescence intensity, the two CARs are equivalent. While there are more total cells present that express GFP within a given subset, there is no indication that an individual cell is being infected with more viral particles and expressing more GFP as a result. Moreover, in direct functional comparisons between CD8HTM and CD28HTM, populations were balanced using donor-matched untransduced cells, so that the number of CAR+ and total T cells were equivalent between the groups both in vitro and in vivo.

[0150] What was not expected was the difference in antigen affinity between the two CARs, considering that the scFv used in both CARs was the same. It has been demonstrated that hinge length can have an impact on efficacy; however, these differences are believed to be due to the ideal cell-to-cell distance of 15nm between the T-cell and target cell. (46,47) In this assay, the antigen is soluble, suggesting that differences other than hinge length are responsible. Biophysical characterization by Chen et. al. of the CD8a hinge demonstrated that CAR-T cells expressing a CD8a hinge were superior killers in vitro compared with a CD28 hinge-expressing CAR when killing low antigen-expressing target cells, but not when antigen levels were high; which is what we observed with the CD8HTM CAR killing the LoVo cells. (48) They propose that increased flexibility makes it more favorable to achieve this ideal intermembrane distance; however, this flexibility may also have benefits within the CAR molecule itself, improving the efficiency of folding of the scFv, conferring a slight but significant increase in affinity that is revealed when antigen density is low and affinity-stabilized interactions promote a complementary increase in avidity, allowing for the formation of a more productive T-cell synapse. A slight change in the folding of the scFv due to the intrinsically more flexible CD8a hinge may also explain why the protein L binding is significantly higher in the CD4+ T cells, between the CD8HTM and CD28HTM CAR, but not observed with G4S detection. Slight conformational differences in the scFv may also have an impact on the affinity between protein L and the light chain framework. Coupled with higher CAR surface expression compared to the CD8+ T cells, slight affinity changes would be more apparent in binding assays within the CD4+ T-cell population, which we observed.

[0151] In many respects, the two CARs produce T cells that function similarly. While the CD8HTM CAR produced a quantitatively significant increase in inflammatory cytokines, the qualitative abundance of the respective cytokines is similar, suggesting that the signaling directed through the CAR is being amplified rather than diversified, i.e. previously silenced signaling pathways are not being activated by the CD8HTM CAR compared to the CD28HTM CAR. Nevertheless, this difference in cytokine abundance still has functional consequences. The reduction in affinity, which leads to a reduction in total granzyme B and cytokine production, may explain why the killing observed in vitro is less complete with the CD28HTM CAR in low antigen conditions. Antigen expression being a Gaussian continuum, one would anticipate that cells from the same cell line with stochastically higher antigen levels would be killed more rapidly in the in vitro assay, while lower expressing cells would take longer, creating a selection pressure promoting the survival of the lowest antigen-expressing cells. The assumption is that every target cell that is killed is killed directly by the CAR-T cell, but inflammatory cytokines can also activate pro-apoptotic pathways within the target cells. Inflammatory bowel diseases occur because the intestinal epithelium undergoes apoptosis in response to inflammatory cytokines, not by perforin- and granzyme-mediated cytolysis.(49) The increased affinity of the CD8HTM CAR allows for the maintenance and sustainability of the anti-tumor response, not solely because of the direct, synaptic interaction, but also because of the inflammatory milieu in which these T cells function. This translates to the difference we see in vivo, where the balance of selective pressure is reversed. Target cells in vitro are given 24 hours to establish a microenvironment, whereas target cells in an animal are given two weeks. In vivo, the T cells are more susceptible to immunosuppressive effectors, so even though antigen is no longer a limiting factor, with the T84 model, the higher affinity, improved sensitivity, increased signaling, and subsequent response with the CD8HTM CAR-T cells has a cumulative effect that results in superior anti-tumor efficacy in vivo.

[0152] The compartmentalized, modular approach to CAR design often disregards the holistic ramifications that changing a single component may have; and demonstrates that the reliance on empiricism in the field of CAR-T cells is still, as yet, unrelenting. What has previously been presented as being wholly attributable to ICDs or scFvs may not be as straightforward as previously thought and suggests that the myopic definition of efficacy be expanded into a more formalized rubric for assessing CARs and CAR design within the field. There are still many unknowns with the CD8HTM and CD28HTM CARs, including the kinetics of synapse formation, how the synapse is ordered, if there are differences in proximal signaling, how receptor endocytosis and recycling compare, what signals are promoting the difference in granzyme B production, and how are these CAR-T cells behaving within the tumor. Despite these important, unanswered questions, this study has allowed us to identify a clinical candidate for the treatment of metastatic CRC. Although CAR-T cell therapy has not yet witnessed the successes in solid tumors as have been experienced in hematological malignancies, these iterative increases in our understanding of how design impacts the CAR, the T cell, and the target, will eventually allow for CAR-T cells to be successful in the solid tumor space.

[0153] METHODS

[0154] Production of Lentiviral Vectors

[0155] Cassettes for CAR transfer plasmids were synthesized and cloned (GenScript Biotech, Piscataway, NJ) into the pCDH-EFla-MCS-T2A-GFP lentiviral transfer vector (CD525A-1, System Biosciences) using the Xbal and BamHI restriction enzymes. All plasmids, including lentiviral packaging and envelope plasmids, were transformed into NEB Stable Competent E. coli (C3040H, New England Biolabs). E. coli were cultured in LB Broth, Miller (BP1426-2, Fisher Scientific) supplemented with 1.9% Bacto Yeast Extract (212750, Thermo Fisher Scientific) and 100 ug / mL Ampicillin (A8351-5G, Sigma-Aldrich). DNA was purified from overnight bacterial cultures using the Purelink Expi Endotoxin-Free Maxi Plasmid Purification Kit (A31231, Thermo Fisher Scientific). DNA pellets were resuspended in Endotoxin-free H2O to a final concentration of ~1 ug / uL. Plasmid DNA was stored at -20°C. High Titer Lentivirus Production

[0156] T-225 flasks were coated with 5 ug / cm2 poly-d-lysine (354210, Corning), washed with DPBS (21-031-CV, Corning), and allowed to dry before 28.4 million HEK293T / 17 cells (CRL11268, American Type Culture Collection (ATCC)) were seeded in cell culture medium consisting of Advanced DMEM (12491023, Thermo Fisher Scientific) supplemented with 5% heat- inactivated FBS (A38400-01, Gibco) and IX GlutaMAX (35050-061, Gibco). The next day, Lipofectamine 3000 transfection reagent (L3000150, Thermo Fisher Scientific) was used to deliver the lentiviral plasmids in the following amounts: 17.7 ug CAR transfer plasmid, 14.7 ug pRSV-Rev (12253, Addgene), 31 ug pMDLg / pRRE (12251, Addgene), and 7.6 ug pMD2.G (12259, Addgene) in Advanced DMEM (no supplementation). Media was collected at 24- and 52-hours post-transfection, filtered using a 0.45 um aPES filter unit (09-740-63E, Fisher Scientific). Lentivirus was concentrated using a 4X Polyethylene Glycol 8000 (BP233-1, Fisher Scientific) solution that was incubated on a wave rotator at 4 °C overnight followed by centrifugation at 1600xg for 1 hour at 4 °C. Lentiviral pellets were resuspended at a 200X concentration in lentivirus storage buffer: lOmM Tris, pH 7.4 (648315-100ML, EMD Millipore), 10% lactose (61339-25G, Sigma- Aldrich), 25mM Proline (81709-10G, Sigma- Aldrich) in DPBS. Virus was stored at -80°C. Lentivirus titer was determined by transducing HEK293T / 17 cells in the presence of 0.8 ug / mL Polybrene (TR-1003-G, Millipore Sigma), measuring the percentage of GFP+ cells using a BD FACSymphony A5 SORP Flow Cytometer (BD Biosciences).

[0157] CAR-T Cell Production

[0158] Human T cells were isolated from the peripheral blood mononuclear cells (PBMCs) of six healthy donors using magnetically sorted negative selection (130-096- 535, Miltenyi Biotec). T cell culture medium was composed of RPML1640 (10-041-CV, Corning) supplemented with 10% heat-inactivated FBS (A38400-01, Gibco), IX Insulin- Transferrin-Selenium (ITS-G, 41400-045, Gibco), lOmM N-Acetyl-L-cysteine (A9165, Millipore Sigma), IX GlutaMAX (35050-061, Gibco), IX Glucose solution (A24940-01, Gibco), IX Sodium Pyruvate (11360-070, Gibco), IX MEM Non-Essential Amino Acids (11 140-050, Gibco), IX HEPES Buffer (15630-080, Gibco), IX Penicillin-Streptomycin (15140-122, Gibco), and 55 uM 2-Mercaptoethanol (21985-023, Gibco). At the start of culture, T cells were seeded into conventional 6- well plates (3516, Coming) at a density of 1 million cells / mL in 4 mL of culture medium. T cells were activated using CD3 / CD28 / CD2 magnetic beads (130-091-441, Miltenyi Biotec) in the presence of 10 ng / mL human IL-7 and 10 ng / mL human IL- 15 (BRB Preclinical Biologies Repository, NCI Biological Resources Branch, Frederick, MD). T cells were transduced with CAR LV at an MOI of 5 with 0.8 ug / mL Polybrene (TR-1003-G, Millipore Sigma) 24 hours after activation. Activation beads were magnetically removed 72 hours after activation and the T cells were transferred to 6-well G-Rex plates (80240M, Wilson Wolf). T cells received fresh culture medium supplemented with IL-7 and IL- 15. Cytokines were replenished every three days. A media exchange occurred on Day 9 post-activation, and cells were collected on Day 12. T-cell concentration and viability were assessed using the Guava MUSE Cell Analyzer (Cytek Biosciences). T cells were cryopreserved using CryoStor CS10 (07930, Stemcell Technologies) at a density of 20 million cells / mL. Prior to experiments, CAR-T cells were thawed in RPMI-1640 (no supplementation) warmed to 37°C, diluted to 1 million cells / mL in T-cell culture medium with IL-7 and IL-15, and allowed to recover in the cell culture incubator for 3 ± 1 days.

[0159] Flow Cytometric Evaluation of CAR T-Cells

[0160] For all flow cytometry experiments, Fc receptors were blocked using Human TruStain FcX Fc Receptor Blocking Solution (422302, BioLegend). All fluorophore- conjugated antibody cocktails contained 10% Brilliant Stain Buffer Plus (566385, BD Biosciences). Triplicate samples were always run for every test sample. The BD FACSymphony A5 SORP was used for all flow cytometry experiments. For each sample, 30,000 live, single cell gated events were collected. All analysis was done using FlowJo, vlO software (BD Biosciences).

[0161] To determine the level of CAR surface expression, Protein L or the anti-G4S linker antibody were used as detection methods. CAR-T cells were incubated with 1 ug / mL recombinant His-tagged protein L (RPL-P3141, ACRO Biosystems) for 45 minutes at 4°C. Cells were incubated with fluorophore-conjugated antibodies for 30 minutes at 4°C (BV421-CD4, Clone RPA-T4, 300532, BioLegend; BB700-CD8, Clone RPA-T8, 566452, BD Biosciences, PE-G4S linker, Clone E720V, Cell Signaling Technology; Alexa Fluor 647-Penta-His, 35370, Qiagen). Live cells were identified using SYTOX AADvanced Dead Cell Stain Kit (SI 0274, Thermo Fisher Scientific). For cell surface quantification, BD Quantibrite PE beads (340495, BD Biosciences) were used to generate a standard curve in which the PE-G4S geometric mean values were used to determine the number of receptors per cell. GUCY2C binding was performed using recombinant, extracellular domain GUCY2C protein that encoded the 23 amino acid (aa) GUCY2C signal peptide, an N- terminal 6x His affinity tag, the 408 aa GUCY2C extracellular domain and a C -terminal 8aa Strep II tag. Protein was synthesized and purified by GenScript Biotech (Piscataway, NJ). Protein concentration was determined using Pierce BCA Protein Assay Kit (23227, Thermo Fisher Scientific) and SDS-PAGE and Western blot analysis (performed by GenScript) determined the molecular weight to be 70 kDa. To cover the entire logarithmic dilution range, 10-fold dilutions were established starting at either 1000 nM or 300 nM in FACS buffer: 1% Bovine Serum Albumin (BP1600-100, Fisher Scientific), 0.1% Sodium Azide (AC 190381000, Fisher Scientific) in DPBS. CAR-T cells were incubated with dilutions of GUCY2C protein for 1 hour at 4°C. T cells were then incubated with the antibody cocktail (BUV805-CD3, Clone OKT3, 750970, BD Biosciences; BV BV421-CD4, Clone RPA-T4, 300532, BioLegend; BB700-CD8, Clone RPA-T8, 566452, BD Biosciences, Alexa Fluor 647-Penta-His, 35370, Qiagen). Immediately upon the conclusion of staining, the cells were placed on ice and immediately analyzed on the flow cytometer. Antigen binding curves were generated by normalizing GFP+ CD4+ or GFP+ CD8+ MFIs to the highest and lowest values within a sample set.

[0162] Intracellular cytokine (ICS) staining was performed by coating 24-well tissue culture plates with 10 ug / mL GUCY2C extracellular domain protein or bovine serum albumin as a negative control. 1 million CAR-T cells were added per well with IX protein transport inhibitor (00-4980-03, Thermo Fisher Scientific). PMA / Ionomycin was added to positive control wells for stimulation (Cell Stimulation Cocktail, 00-4970-03, Thermo Fisher Scientific). To identify CAR+ T cells, Alexa Fluor 488-G4S antibody (50515L, Cell Signaling Technology) was added at a final dilution of 1:800 to each well. Cells were incubated in the presence of antigen for 6 hours in a cell culture incubator, at which point the cells began the staining procedure. Fixable Viability Stain (FVS) 575V (565694, BD Biosciences) was used to identify the viable cell population. Cell surface antigens were stained at 4°C for 30 minutes (BUV395-CD45, Clone HI30, 363-0459-42, Thermo Fisher Scientific; BUV805-CD3, Clone OKT3, 750970, BD Biosciences; BUV563-CD4, Clone RPA-T4, 741353, BD Biosciences; BB700-CD8, Clone RPA-T8, 566452, BD Biosciences). Cells with fixed with 4% paraformaldehyde (PFA) in PBS (JI 9943. K2, Thermo Fisher Scientific) for 20 minutes at 4°C. Cells were stored at 4°C overnight before proceeding with intracellular staining. Cells were permeabilized for 15 minutes at room temperature using Perm / Wash buffer (554723, BD Biosciences) diluted in H2O. Antibodies for intracellular cytokine staining were diluted in perm / wash solution. The following antibodies were used: BUV737-IL2, Clone MQ1-17H12, 612836, BD Biosciences; BV421-IFNy, Clone 4S.B3, 564791, BD Biosciences; RB780-granzyme B, Clone GB11, 568705, BD Biosciences; Alexa Fluor 647-TNFa, Clone MAbl 1, 502916, BioLegend. Staining was performed for 30 minutes at 4°C. A final 2% PFA fixation was performed. Cells were analyzed on the flow cytometer within 24 hours of completing the staining protocol.

[0163] Flow cytometry for T-cell phenotypic markers consisted of the FVS525V, followed by the following surface markers: BUV395-CD45, Clone HI30, 363-0459-42, Thermo Fisher Scientific; BUV805-CD3, Clone OKT3, 750970, BD Biosciences; BUV563-CD4, Clone RPA-T4, 741353, BD Biosciences; BB700-CD8, Clone RPA-T8, 566452, BD Biosciences; Alexa Fluor 488-G4S, Clone E720V, 50515L, Cell Signaling Technology. For memory phenotyping the following markers were used: BV421-CCR7, Clone 2-L1-A, 566743, BD Biosciences; Alexa Fluor 647-CD45R, Clone HUGO, 304112, BioLegend. For exhaustion marker analysis, the following antibodies were used: BUV661-CD39, Clone TU66, 569788, BD Biosciences; BV421-PD-1, Clone MIH4, 564323, BD Biosciences; PE-CF594-Lag-3, Clone T47-530, 565718, BD Biosciences; Alexa Fluor 647-Tim-3, Clone 7D3, 565558, BD Biosciences. Antibody staining was performed for 30 minutes at 4°C. Cells were fixed with 4% PFA at 4°C for 20 minutes. Cells were stored at 4°C until analysis on the flow cytometer.

[0164] GUCY2C mRNA Analysis

[0165] SW620 (CCL227, ATCC), LoVo (CCL229, ATCC), LS174T (CL- 188, ATCC), and T84 (CCL-248, ATCC) cell lines were plated in 24-well tissue culture plates. Using an RNeasy kit (74104, Qiagen), samples were lysed in RLT buffer supplemented with 550 uM 2-mercaptoethanol. RNA was purified on spin columns according to manufacturer instructions. RNA concentrations and purity were measured with a Nanodrop 1000 (Thermo Fisher Scientific). RNA was reverse transcribed to complementary DNA (cDNA) using the TaqMan Reverse Transcription kit according to the manufacturer’s instructions (N8080234, Thermo Fisher Scientific). Transcripts were quantified by qRT-PCR using Taqman primer probes (Human GUCY2C Assay ID Hs00990106_ml, Thermo Fisher Scientific; Human -Actin Assay ID Hs01060665_gl, Thermo Fisher Scientific) on a QuantS tudio™ 3 Real-Time PCR System (Thermo Fisher Scientific), with TaqMan Universal PCR Master Mix (4318157, Thermo Fisher Scientific) according to instructions.

[0166] GUCY2C Western blot Analysis

[0167] CRC cell lines were cultured in 10 cm dishes. Radio-Immunoprecipitation Assay (RIP A) buffer was supplemented with IX HALT Protease and Phosphatase Inhibitor cocktail, EDTA-free (78441, Thermo Fisher Scientific) and ImM Phenylmethylsulfonyl fluoride (PMSF, 36978, Thermo Scientific) and the solution was placed on ice. Culture plates were washed in ice-cold DPBS three times, followed by ImL of RIPA buffer solution. Cells were scraped, and the contents were transferred to microcentrifuge tubes. Cells were placed on a tube rotator at 4°C for 30 minutes. Tubes were centrifuged at 12,000 RPM for 10 minutes. Supernatants were transferred to ice-cold microcentrifuge tubes and immediately placed in -20°C storage.

[0168] Protein concentration was measured using the Pierce BCA Protein Assay Kit (same as above). Samples were diluted with 4X Invitrogen NuPage LDS Sample Buffer (NP0007, Fisher Scientific) and Invitrogen Novex NuPage Sample Reducing Agent (NP0009, Thermo Scientific). Samples were boiled for 10 minutes at 90°C and immediately placed on ice. Samples were loaded into NuPage 4-12% Bis-Tris Protein Gels (NP0336BOX, Thermo Fisher Scientific) along with Invitrogen Novex Sharp Pre- Stained Protein Ladder (LC5800, Thermo Fisher Scientific). Gels were run at 130V for 90 minutes. Gels were transferred to iBlot 3 Transfer Stack, nitrocellulose (IB33002X3, Thermo Scientific) using a seven-minute transfer program. The membrane was blocked in 10% non-fat dry milk (M0841, LabScientific) in PBS with Tween- 20 (PBS-T). Membranes were probed using a Rabbit Anti-human GUCY2C antibody (37517, Cell Signaling Technology) or Rabbit Anti-human -Actin (8457, Cell Signaling Technology). Both antibodies were used at a dilution of 1 :1,000. An HRP-conjugated Goat Anti-Rabbit IgG secondary antibody was used at 1:12,500 in PBS-T. Thermo Scientific SuperSignal West Femto Chemiluminescent Substrate (34096, Thermo Fisher Scientific) was used to detect the bands on a ChemiDoc MP Image System (Bio-Rad Laboratories, Inc.). Images were analyzed using Image Lab, version 6.1.0 (Bio-Rad Laboratories, Inc.).

[0169] In Vitro Cytotoxicity Assay

[0170] LoVo cells were cultured in F-12K medium (10-092-CV) supplemented with 10% FBS (35-010-CV, Coming). LS174T cells were cultured in Eagle’s Minimal Essential Medium (10-010-CV, Corning) supplemented with 10% fetal bovine serum (35-010-CV, Corning), IX GlutaMAX (35050-061, Gibco), IX NEAA (11140-050, Gibco), and IX Sodium Pyruvate (11360-070, Gibco). T84 cells were cultured in Advanced DMEM / F-12 50% / 50% (12634028, Thermo Fisher Scientific) supplemented with 5% FBS (35-010- CV, Corning) and IX GlutaMAX (35050-061, Gibco). When the cells were ready to be used in cytotoxicity experiments, the cells were trypsinized (25-053-CI, Coming), filtered through a 30 um sterile MACS SmartStrainer (130-098-458, Miltenyi Biotec), and counted using a hemocytometer. Cells were resuspended to a density of 8.0 x 105 cells / mL in the respective growth media for each cell line.

[0171] Cytotoxicity was measured in real-time using the Agilent xCELLigence Real Time Cell Analysis (RTCA) SP - Single Plate analyzer (Agilent Technologies, Inc.). 100 uL of each cell line’s respective media was added to the well of an E-Plate 96 (5232368001, Agilent Technologies, Inc.) to establish a baseline impedance measurement for each well. 4.0 x 104 CRC cell line target cells were added to each well in the cell culture hood. The plate was allowed to incubate for 30 minutes at room temperature before being placed into the xCELLigence RTCA machine (located within a cell culture incubator set to 37°C and supplied with 5% CO2). The cells were allowed to adhere for 24 hours and only when the cell index (an arbitrary unit measurement of current impedance) was above 1.0 were cells used in cytotoxicity experiments. CAR-T cells were balanced in terms of GFP+ CAR-T cells as well as total T cells. T cells were collected, counted, and resuspended so that the CAR+ populations were at a density of 1.6 x 106 cells / mL. T cells were resuspended in RPML1640 (10-041-CV, Coming) without any supplementation or cytokines. 8.0 x 104 CAR-T cells in 50 uL were added to each well for an E:T ratio of 2:1. The E-Plate was placed in the xCELLigence RTCA machine and cell impedance measurements were collected every 15 minutes. The RTCA Software Pro (Agilent Technologies, Inc.), which is used to run the machine, was also used to analyze the data.

[0172] In Vivo Mouse Tumor Study

[0173] Animal experiments were conducted in accordance with an IACUC approved animal protocol and followed the guidelines for use of research animals stipulated by Thomas Jefferson University. A lentiviral construct was made using a modified pCDH- EFla-MCS-T2A-GFP backbone where the GFP reporter was replaced with the fluorescent mCherry protein. The Click Beetle Red luciferase (CBRluc) was cloned into the multiple cloning sites using Xbal and BamHI restriction enzymes to generate the pCDH-EFla-CBRluc-T2A-mCherry lentiviral vector. DNA was transformed, amplified, and purified and lentivirus was generated as stated above. T84 cells were transduced with concentrated lentivirus at an MOI of 1 in the presence of 0.8 ug / mL polybrene. Cells were expanded and then flow-sorted on live singlets for 50% of the mCherry population surrounding the median mCherry fluorescent signal to isolate a pure, uniform population of transduced cells. Flow sorting was performed using a BD FACSMelody sorter. Flow- sorted T84-CBRluc-T2A-mCherry (T84-Luciferase) cells were expanded in tissue culture and cryopreserved in Fetal Bovine Serum (FBS, 35-010-CV, Coming) with 10% DMSO (PI20688, Fisher Scientific). Prior to administration, T84-Luciferase cells were removed from liquid nitrogen storage and cultured using Advanced DMEM / F-12 50% / 50% (12634028, Thermo Fisher Scientific) supplemented with 5% FBS (35-010-CV, Coming) and IX GlutaMAX (35050-061, Gibco). The cells were cultured and expanded for 10 days before being trypsinized (25-053-CI, Coming), filtered through a 30 um sterile MACS SmartStrainer (130-098-458, Miltenyi Biotec). Cells were counted and resuspended in cold DPBS at a final concentration of 12.5 x 106 cells / mL. Cells were kept on ice for injection.

[0174] Four-week-old female NSG-MHC VII DKO mutant mice (025216, The Jackson Laboratory) were injected intraperitoneally with 2.5 x 106 T84-Luciferase cells in 200 uL using a 0.5 mL tuberculin syringe (14-826-79, Fisher Scientific). On the day of tumor cell implantation, a culture of CAR-T cells using a single donor was started following the above methods for CAR-T cell production. On Day 12, CAR-T cells were evaluated for transduction efficiency using flow cytometry to determine the percentage of GFP+ in the CD8HTM and CD28HTM CAR-T cell populations. On Day 13, animals were imaged using an IVIS Spectrum In Vivo Imaging System with a five-mouse manifold (Perkin Elmer). The luminescence signal in Photons / second / centimeter2 / steradian was used to rank all of the animals in terms of luminescence signal. Four mice surrounding the median of the ranked luminescence intensities were allocated to the untreated group to control for unimpeded tumor growth. The remaining animals were allocated using block randomization to the CD8HTM or CD28HTM group. Fourteen days after tumor implantation, CAR-T cells were collected and counted. The two CAR-T cell groups were balanced in terms of the percent of GFP+ cells as well as total T cells, using donor- matched untransduced T cells. Thus, all animals received the same number of CAR-T cells in the same number of total T cells. 3.0 x 106 CAR+ T cells were injected intravenously via the tail vein.

[0175] Starting seven days post tumor implantation, animals began undergoing in vivo bioluminescent imaging. Animals were anesthetized using inhaled isoflurane. Mice were sedated for three minutes, and then each animal was injected subcutaneously with 250 uL of a 15 mg / mL solution of D-Luciferin, potassium salt (LUCK-4G, Gold Biotechnology) dissolved in DPBS (21-030-CV, Corning). Animals were returned to the isoflurane anesthesia chamber. Five minutes after the administration of D-Luciferin solution, animals were imaged in the IVIS using a 10-second exposure. Throughout the course of the experiment, animals that were responsible for saturating the optical system beyond the linear range of the photon detector would be removed after imaging, and the remaining animals would be re-imaged to provide a more accurate reading of the luminescent signal. Animals were imaged bi-weekly for a 6-week period following CAR-T cell treatment. Images were analyzed using Aura In Vivo Imaging Software (Spectral Instruments Imaging). All images were set to the same color range minimum, color range maximum, and color range threshold. To determine the baseline luminescence for each day of imaging, a single cage of five animals received vehicle (DPBS) alone on the day of tumor implantation. On Day 14, these animals received the same number of untransduced total T-cells as the CAR-T-treated groups. These animals were imaged as above throughout the experiment.

[0176] Data Availability and Statistical Analysis

[0177] The data that support the findings of this study are available from the corresponding author, AES, upon reasonable request. All figures presenting data generated in this study, including all statistical analyses, were performed using GraphPad Prism, version 10.2.3. Comparisons employed T-test, One-way ANOVA, Two-way ANOVA, and Kaplan-Meier two-sided log-rank tests as appropriate.

Claims

CLAIMS1. A chimeric antigen receptor protein comprising an antigen binding domain, a CD8a hinge domain, a CD 8 a transmembrane domain, and one or more intracellular domains.

2. The chimeric antigen receptor protein of claim 1 wherein the antigen binding domain is an scFV domain that comprises a VL sequence, a linker and a VH sequence.

3. The chimeric antigen receptor protein of claim 2 wherein the antigen binding domain is an scFV domain that binds to a human antigen.

4. The chimeric antigen receptor protein of claim 2 wherein the antigen binding domain is an scFV domain that binds to human GUYC2C.

5. The chimeric antigen receptor protein of claim 3 wherein the antigen binding domain is an anti-GUYC2C scFV sequence having SEQ ID NO:2.

6. The chimeric antigen receptor protein of any of claims 1 -5 wherein the one or more intracellular domains comprises: a CD28 signaling domain, a 4-1BB (CD137) signaling domain, a CD2 signaling domain, a CD27 signaling domain, a CD30 signaling domain, a CD40L signaling domain, a CD79A signaling domain, a CD79B signaling domain, a CD226 signaling domain, a DR3 signaling domain, a GITR signaling domain, a HVEM signaling domain, a ICOS signaling domain, a LIGHT signaling domain, a 0X40 signaling domain, a SLAM signaling domain, CD3^ domain that includes an immunoreceptor tyrosine activation motif, CD79-a domain that includes an immunoreceptor tyrosine activation motif, CD79-P domain that includes an immunoreceptor tyrosine activation motif, and an Fc receptor sequence that includes an immunoreceptor tyrosine activation motif.

7. The chimeric antigen receptor protein of claim 6 comprising a CD28 signaling domain, a 4- IBB (CD 137) signaling domain and a CD3^ domain that includes an immunoreceptor tyrosine activation motif.

8. The chimeric antigen receptor protein of any of claims 1-7 further comprising a signal sequence.

9. The chimeric antigen receptor protein of claim 8 wherein the signal sequence is selected from the group consisting of: a GM-CSF signal sequence, a CD8a signal sequence, a CD8P signal sequence, a CD4 signal sequence, a TCRa signal sequence, a TCR0 signal sequence, a CD36 signal sequence, a CD3s signal sequence, a CD3y gamma signal sequence, a CD28 signal sequence, and a BiP signal sequence.

10. The chimeric antigen receptor protein of claim 8 wherein the signal sequence is a CD8a signal sequence.

11. The chimeric antigen receptor protein of any of claims 1-10 further comprising an affinity tag.

12. The chimeric antigen receptor protein of claim 1 comprising an anti-GUYC2C scFV sequence, a CD8a hinge domain, a CD 8 a transmembrane domain, a CD28 signaling domain, a 4-1BB (CD137) signaling domain and a CD3^ domain that includes an immunoreceptor tyrosine activation motif wherein the chimeric antigen receptor protein comprises SEQ ID NO:2.

13. A nucleic acid molecule comprising a nucleic acid sequence that encodes a protein of any of claims 1-12.

14. A nucleic acid molecule comprising a nucleic acid sequence that encodes a protein of claim 12.

15. A nucleic acid molecule of claim 14 comprising a nucleic acid sequence comprises SEQ ID NO:1.

16. The nucleic acid molecule of any of claims 13-15 wherein nucleic acid sequence that encodes the protein is operably linked to regulatory elements for expression in human T cells.

17. A recombinant cell comprising the nucleic acid molecule of claim 16.

18. A recombinant T cell comprising the nucleic acid molecule of claim 16.

19. A recombinant cell comprising the chimeric antigen receptor of any of claims 1- 12.

20. A recombinant T cell comprising the chimeric antigen receptor of any of claims 1 - 12.

21. A recombinant cell comprising the chimeric antigen receptor of claim 12.

22. A recombinant T cell comprising the chimeric antigen receptor of claim 12.

23. A method of treating a patient who has cancer that has cancer cells that express GUCY2C wherein the method comprises the step of administering to said patient the plurality of recombinant cells comprising the chimeric antigen receptor of any of claims 3-12.

24. The method of claim 23 wherein the plurality of recombinant cells is a plurality of recombinant T cells.

25. A method of treating a patient who has cancer that has cancer cells that express GUCY2C wherein the method comprises the step of administering to said patient the plurality of recombinant cells comprising the chimeric antigen receptor of claim 12.

26. The method of claim 25 wherein the plurality of recombinant cells is a plurality of recombinant T cells.

27. A method of treating a patient who has cancer that has cancer cells that express GUCY2C, the method comprises the steps of: isolating T cells from the patient; transforming the T cells with a nucleic acid molecule that encodes the chimeric antigen receptor of any of claims 3-12 to produce a population of transformed T cells that express the nucleic acid molecule that encodes the chimeric antigen receptor, wherein the chimeric antigen receptor is present in the transformed T cell as a membrane bound protein, expanding the population of transformed T cells to produce a plurality of transformed T cells, and administering to said patient the plurality of recombinant T cells.

28. A method of treating a patient who has cancer that has cancer cells that express GUCY2C, the method comprises the steps of: isolating T cells from the patient; transforming the T cells with a nucleic acid molecule that encodes the chimeric antigen receptor of claim 12 to produce a population of transformed T cells that express the nucleic acid molecule that encodes the chimeric antigen receptor, wherein the chimeric antigen receptor is present in the transformed T cell as a membrane bound protein, expanding the population of transformed T cells to produce a plurality of transformed T cells, and administering to said patient the plurality of recombinant T cells.

29. The method of claim 28 wherein the T cells are transformed with a nucleic acid molecule that comprises SEQ ID NO: 1.

30. The method of any of claims 23-29 wherein prior to treating the patient, a sample of cancer cells is isolated from the patient and GUCY2C is detected on said cancer cells.

31. A method of preventing cancer that has cancer cells that express GUCY2C in a patient identified as being of increased risk, the method comprises the step of administering to said patient the plurality of recombinant cells comprising the chimeric antigen receptor of any of claims 3-12.

32. The method of claim 31 wherein the plurality of recombinant cells is a plurality of recombinant T cells.

33. A method of preventing cancer that has cancer cells that express GUCY2C in a patient identified as being of increased risk, the method comprises the step of administering to said patient the plurality of recombinant cells comprising the chimeric antigen receptor of claim 12.

34. The method of claim 33 wherein the plurality of recombinant cells is a plurality of recombinant T cells.

35. A method of preventing cancer that has cancer cells that express GUCY2C in a patient identified as being of increased risk, the method comprises the steps of: isolating T cells from the patient; transforming the T cells with a nucleic acid molecule that encodes the chimeric antigen receptor of any of claims 3-12 to produce a population of transformed T cells that express the nucleic acid molecule that encodes the chimeric antigen receptor, wherein the chimeric antigen receptor is present in the transformed T cell as a membrane bound protein, expanding the population of transformed T cells to produce a plurality of transformed T cells, and administering to said patient the plurality of recombinant T cells.

36. A method of preventing cancer that has cancer cells that express GUCY2C in a patient identified as being of increased risk, the method comprises the steps of: isolating T cells from the patient; transforming the T cells with a nucleic acid molecule that encodes the chimeric antigen receptor of claim 12 to produce a population of transformed T cells that express the nucleic acid molecule that encodes the chimeric antigen receptor, wherein the chimeric antigen receptor is present in the transformed T cell as a membrane bound protein, expanding the population of transformed T cells to produce a plurality of transformed T cells, and administering to said patient the plurality of recombinant T cells.

37. The method of claim 36 wherein the T cells are transformed with a nucleic acid molecule that comprises SEQ ID NO:1.

38. A method of making the plurality of recombinant cells of any claims 17-22 comprising the steps of: isolating cells from an individual; transforming the cells with a nucleic acid molecule that encodes the chimeric antigen receptor operable linked to regulatory elements functional in cells to produce a population of transformed cells that comprise the chimeric antigen receptor present in the transformed cells as a membrane bound protein, and expanding the population of transformed cells to produce a plurality of recombinant cells; wherein the cells are transformed using a high titer of lenti virus that comprises the nucleic acid molecule that encodes the chimeric antigen receptor.

39. The method of claim 38 wherein the cells are T cells.

40. A method of making the plurality of recombinant cells of any claims 17-22 comprising the steps of: isolating cells from an individual; culturing the cells in RPMI- ITS, transforming the cells in RPMI-ITS with a nucleic acid molecule that encodes the chimeric antigen receptor operable linked to regulatory elements functional in cells to produce a population of transformed cells that comprise the chimeric antigen receptor present in the transformed cells as a membrane bound protein, and expanding the population of transformed cells to produce a plurality of recombinant cells; wherein the cells are transformed using a high titer of lentivirus that comprises the nucleic acid molecule that encodes the chimeric antigen receptor.

41. The method of claim 40 wherein the cells are T cells.

42. A method of making the plurality of recombinant cells of any claims 17-22 comprising the steps of: isolating cells from an individual; transforming the cells in with a nucleic acid molecule that encodes the chimeric antigen receptor operable linked to regulatory elements functional in cells to produce a population of transformed cells that comprise the chimeric antigen receptor present in the transformed cells as a membrane bound protein, and activating and expanding the population of transformed cells by culturing the transformed cells in media supplemented with a combination of anti-CD3, anti-CD28, and anti-CD2 antibodies to produce a plurality of recombinant cells.

43. The method of claim 42 wherein the cells are T cells.

44. A method of making the plurality of recombinant cells of any claims 17-22 comprising the steps of: isolating cells from an individual; culturing the cells in RPMI- ITS, transforming the cells in RPMI-ITS with a nucleic acid molecule that encodes the chimeric antigen receptor operable linked to regulatory elements functional in cells to produce a population of transformed cells that comprise the chimeric antigen receptor present in the transformed cells as a membrane bound protein, and activating and expanding the population of transformed cells by culturing the transformed cells in media supplemented with a combination of anti-CD3, anti-CD28, and anti-CD2 antibodies to produce a plurality of recombinant cells; wherein the cells are transformed using a high titer of lentivirus that comprises the nucleic acid molecule that encodes the chimeric antigen receptor.

45. The method of claim 44 wherein the cells are T cells.

Images