CAR-T CELL COMPOSITIONS METHODS OF USE THEREOF CROSS-REFERENCE TO RELATED APPLICATION This application claims priority to and benefit of U.S. Provisional Patent Application No. 62/903,776 filed on September 21, 2019, which is incorporated by reference in its entirety. STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH This invention was made with government support under CA168912 awarded by the National Institutes of Health. The government has certain rights in the invention. TECHNICAL FIELD OF THE INVENTION Aspects of the invention are generally directed to immunotherapy; in particular to antibodies and chimeric antigen receptor T cells and methods of their use. BACKGROUND OF THE INVENTION There are annually 850,000 new cases of liver cancers, making it the 6th most common cancer worldwide (1). The majority (80-90%) of liver cancers are hepatocellular carcinoma (HCC). While HCC has leveled off in Asia thanks to HBV vaccination (2), the HCC incidence rate increases significantly in the developed countries due to obesity/diabetes (3, 4). According to Center for Disease Control of the United States, the incidence rate of liver cancer has doubled in the last 2 decades, making it the fastest increasing cancer. Treatment options for liver cancer are limited. Although liver resection is curative, most HCC are diagnosed at later stage, disqualifying them for surgery. In addition, the lack of adjuvant therapies results in ~70% 5yr recurrence rate (5), which makes liver cancer the 2nd leading cause of cancer death among adult men (1). Thus, novel therapies are urgently needed. Immunotherapy has a great potential for malignant tumors (6-10). Checkpoint blockade has shown therapeutic effect on HCC (9, 11), but it requires the presence of tumor-infiltrating T cells, which do not always exist (12). One approach to generate tumor-specific T cells is to redirect patient’s autologous T cells with specific T cell receptors (TCR) (13, 14). However, the TCR’s MHC-restriction limits its applications. An MHC-independent method is to utilize chimeric antigen receptor (CAR) genes to engineer T cells to target tumor surface antigens. A CAR combines the high affinity and specificity of monoclonal antibody (mAb) with TCR signaling that can activate T cells and mediate strong and specific killing of tumor cells independent of MHC (15, 16). T cells engineered with CAR genes (CARTs) have demonstrated strong antitumor effects in treating hematologic cancers (16-18). CARTs are also being intensively studied for treating solid tumors including HCC (19). Glypican 3 (GPC3) is a member of the heparin sulfate (HS) proteoglycans family that is attached to cell membrane via a glycosylphosphatidylinositol (GPI) anchor (20, 21) and is involved in wnt/b-catenin signaling (22, 23). Approximately 70-80% of HCC tumor cells, but not normal hepatocytes, express GPC3. And its level has been associated with poor patient survival rate (24). Thus, GPC3 might be a great target for HCC immunotherapy (25). Several hGPC3-specific mAbs (GC33, YP7, HN3, and MDX-1414) have been tested for their antitumor effects (26-28). But only GC33 has advanced to clinical trials (29), without significant antitumor benefits (30). Recently, several groups have been investigating the antitumor effects of the GC33 derived CARTs (31-33), and their antitumor benefits remain to be evaluated. At the same time, it is unknown whether CARTs developed from different mAbs, especially those mAbs targeting different regions of hGPC3 will generate different antitumor effects. In addition, although GC33 mAb has been proved safe in clinical trials, a recent study showed that 7 out of 34 non-HCC tumor tissues were stained positive by GC33 (34). Thus, there is a need to generate more CARTs of high specificity and effectiveness for HCC immunotherapy. One object of the invention provides antibody compositions and methods of their use. Another object of the invention provides chimeric antigen receptor T-cell compositions and methods of their use.. Still another object of the invention is to provide methods and compositions for treating cancer. SUMMARY OF THE INVENTION Chimeric antigen receptor T cell (CARTs) compositions and methods of their use are provided. One embodiment provides a chimeric antigen receptor T cell composition including 8F8 chimeric antigen receptor T cells. Another embodiment provides a chimeric antigen receptor T cell composition including 6G11 chimeric antigen receptor T cells. Still another embodiment provide a chimeric antigen receptor T cell composition including 12D7 chimeric antigen receptor T cells. The disclosed CARTs can be used to treat cancer, for example hepatocellular carcinoma. Other embodiments are directed to antibodies used to produce the CARTs. One embodiment provides a recombinant monoclonal antibody having a mAb 6G11 variable light chain and a mAb 6G11 variable heavy chain. Another embodiment provides a recombinant monoclonal antibody having mAb 8F8 variable light chain and mAb 8F8 variable heavy chain. Still another embodiment provides a recombinant monoclonal antibody comprising mAb 12D7 variable light chain and mAb 12D7 variable heavy chain. One embodiment provides a pharmaceutical composition including the disclosed antibodies and or CARTs. Yet another embodiment provides a method for treating liver cancer by administering to a subject in need thereof an effective amount of 6G11, 8F8, or 12D7 chimeric antigen receptor T cells to treat cancer, for example hepatocellular carcinoma. One embodiment provides a method for treating hepatocellular carcinoma in a subject in need thereof including the steps of isolating T cells from the subject, transfecting the isolated T cells with a vector containing 6G11-CAR nucleic acid sequences, optionally expanding the transfected cells in cell culture, and administering an effective mount of the transfected cells to the subject to treat hepatocellular carcinoma. One embodiment provides a method for treating hepatocellular carcinoma in a subject in need thereof including the steps of isolating T cells from the subject, transfecting the isolated T cells with a vector containing 8F8-CAR nucleic acid sequences, optionally expanding the transfected cells in cell culture, and administering an effective mount of the transfected cells to the subject to treat hepatocellular carcinoma. One embodiment provides a method for treating hepatocellular carcinoma in a subject in need thereof including the steps of isolating T cells from the subject, transfecting the isolated T cells with a vector containing 12D7-CAR nucleic acid sequences, optionally expanding the transfected cells in cell culture, and administering an effective mount of the transfected cells to the subject to treat hepatocellular carcinoma. One embodiment provides a chimeric antigen receptor immune cell comprising a nucleic acids having 85, 90, 95, 99, or 100% sequence identity to SEQ ID NO:1, 2, or 3. BRIEF DESCRIPTION OF THE DRAWINGS Figures 1A is a table of 22 hGPC3-specific mAbs identified by ELISA assay. Their Ig class, staining of HepG2 cells, and binding of the hGPC3 N- or C-fragment were provided. The mAbs in grey shade did not stain HepG2 cells. Figure 1B is a panel of micrographs showing representative immunofuorescent stainings of HepG2 tumor cells by 5 mAbs. Figure 1C is a diagram of plasmids expressing the N- and C-fragments of hGPC3. Figure 1D is a panel of graphs showing 293 cells transfected with either hGPC3-N or hGPC3-C plasmid intracellularly stained with indicated mAbs. The gate was set by the parental 293T cells without transfection. Figure 1E is a panel showing immune fluorescent staining of HepG2 tumor cells with the indicated hGPC3-specific mAbs. Figure 1F is a panel showing immune fluorescent staining of the Huh7 cells by indicated hGPC3-specific mAbs. Fig 1G is a panel of graphs showing intracellular staining of the 293T cells transfected with hGPC3-N or hGPC3-C fragment gene with different hGPC3-specifi c mAbs. The gate was set by the parental 293 cells without transfection. Figure 2A is a panel of micrographs showing IHC staining of liver tissues. Representative IHC staining of healthy liver and the paired HCC and the adjacent normal tissues of the same livers. A low and high magnications were presented. Dark spots indicated positive staining. Figure 2B is a table showing the summary of IHC staining of multiple tissue samples by 5 mAbs. ANL: adjacent normal liver tissues. The highest immunoscore is 8, the lowest score is 0. Figure 2C is a panel of micrographs showing IHC staining of HCC tumor tissues and their paired adjacent normal liver tissues by mAbs by 5 mAbs. Some negative staining by mAbs were not presented. Figure 3A is an illustration showing the schematic structure of the recombinant mAbs. SP: signal peptide of human IgG; VL: variable region of light chain; VH: variable region of heavy chain; CL: constant region of human IgG light chain; CH: constant region of human IgG heavy chain. Figure 3B is gel showing PAGE analysis of purified recombinant mAbs under reduced and non-reduced condition. mAbs were purified from the 293 cells after co-transfection of the L and H chain expressing plasmids. Gel was stained with SimplyBlue SafeStain . Figure 3C is a graph showing the EC50 of recombinant mAbs by ELISA. Figure 3D is a panel of graphs showing the Kd value of each mAbs was measured by BLI. Each experiment was repeated at least 3 times and similar data were obtained. Figure 3E is a graph showing the Kd value of mAbs 6G11, 8F8, and 12D7 for staining HepG2 tumor cells. Immunofluorescent staining was conducted for HepG2 tumor cells using the mAbs at different concentration. Based on the percent of positive cells and the mAb concentration, the Kd value was calculated by Prism software. Figure 4A shows West Blot analysis of the 293 cells transfected with hGPC3 N- or hGPC3 C- fragment. Figure 4B and 4C are shows graphs with data confirming that 6G11 and 12D7 bound to hGPC3 N-fragment and 8F8 bound to the hGPC3 C-fragment. HepG2 cell lysate was used as positive control. Mapping the 6G11 and 8F8 binding epitopes by overlapping peptides. The data showed 6G11 mAb bound to N1 peptide (AA25-39) (B) and 8F8 bound to C5 peptide (AA463-496). Figure 4D is an illustration of the binding sites of mAbs on a schematic structure of hGPC3 derived from Haruyama and Kataoka (reference 25). As a reference, the GC33 mAb’s epitope was indicated. The potential disulde bonds between N-fragment and C-fragment of hGPC3 were also shown. Figures 5A shows the schematic structure of the second generation CAR-Ts built from 6G11, 8F8, and 12D7 was illustrated. GS linker: glycine-serine linker; TM:transmembrane domain of CD28. Figure 5B is an illustration of the experimental procedure. Human T cells were isolated from PBMC and activated by CD3/CD28 beads for 24hrs. On day 0, T cells were transduced with lv and CAR expression was monitored from day 3-13. Figure 5C is a panel of dot plots after CAR gene transduction at different time points were presented. CAR expression was detected by anti-mouse Fab antibody at indicated time points. This experiment was repeated at least 5 time with similar observations. Figures 5Damd 5E are dot plots showing CAR molecules on the transduced T cells were stained with anti-mouse Fab antibody (A) or FITC- labeled hGPC3. Mock transduced T cells were used as control. Figure 5F is a graph showing the percent of CAR+ T cells that could bind hGPC3 protein. Figures 6A-6D show responsiveness of CAR-Ts to HCC tumor cell stimulation. CARTs were generated as described in Fig.5. On Day 14 after transduction, CARTs were co-cultured with or without HepG2 tumor cell. The % of CAR+ T cells and their effector function were examined after 6 days. Figure 6A shows representative dot plots of CARTs after HepG2 stimulation. Figure 6B is graph showing the summary of fold changes of the CAR+ T cells before and after HepG2 stimulation. Statistics was done with student t-test. Figure 6C and 6D are line graphs showing cytotoxicity and cytokine production of the CARTs. At Day 20 after transduction (6 days after HepG2 stimulation), the non-stimulated (Figure 6C) and HepG2 stimulated (Figure 6D) CARTs were co-cultured with HepG2 cells for 20hrs. Cytotoxicity and IFNg and IL2 were measured by LDH and ELISA. Data show Mean+/-SD, statistics was done with 2way ANOVA. Experiment was repeated thrice with similar observations. Figure 6E-6G show the cytotoxicity of CARTs on Huh7 tumor cells. Figure 6E are graphs showing the 6G11 CARTs killed the hGPC3low Huh7 tumor cells, but not the hGPC3- 293T cells. Figure 6F shows 8F8 CARTs, but not the12D7 CARTs (Figure 6G) could kill Huh7 tumor cells. Figure 7A-7C show soluble hGPC3 did not activate CARTs nor inhibit CART activation by HepG2. Figure 7A is a graph showing CARTs were co-cultured with HepG2 or with different concentration of soluble hGPC320hrs and ELISA was conducted to determine the IFNg. Figures 7B and 7V are graphs showing CARTs were co-cultured with HepG2 (2:1 ratio) in the presence of different concentrations of hGPC3 protein for 20hrs. Cytotoxicity was measured by LDH assay and cytokines were evaluated by ELISA. Student t-test was used for statistical analysis. ns: not signifcant. The experiment was repeated twice with similar observation. Figure 8A-8F show adoptive transfer of CARTs generated potent antitumor effects correlating to CAR expansion in vivo. Figure 8Ais an illustration of the experimental scheme. Figure 8B are line graphs of tumor growth curves in different treatment groups. Each line represents one tumor. Figure 8C-8E show the analysis of the transferred CARTs. The gating strategy was shown Figure 8C. Mouse CD45+ cells were used as internal reference. The kinetics of human CD45+ (Figure 8D) and % of CAR+ T cells (Figre 8E) in NSG mice were summarized. Each line represents one mouse. The color matched to the tumor size curve in B. (Figure 8F) Pictures of the tumor bearing mice at 40 days after tumor inoculation (33 days after treatment). This experiment was repeated 5 times with similar observations. Figure 8G are graphs showing the antitumor effect of CARTs when a lower effector/tumor load (0.6 CARTs vs 1 HepG2 tumor cells) was used. HCC xenografts were established with 5 million HepG2 tumor cells inoculation. Seven days later, 3 millions of CARTs were adoptively transferred into the tumor bearing mice. Even at the lower E/T ratio of 0.6, all CARTs generated signi-cant inhibition of tumor growth. However, only the 8F8 CAR-Ts were able to completely eradicate HCC tumor xenografts. In contrast, only 20% of 6G11 CART treated mice were tumor free. Figure 9 is an illustration showing development and administration of CARTs. Figure 10 is a graph showing the summary of antitumor effects of the indicated CARTs. DETAILED DESCRIPTION OF THE INVENTION I. Definitions As used herein, a molecule is said to be able to “immunospecifically bind” a second molecule if such binding exhibits the specificity and affinity of an antibody to its cognate antigen. Antibodies are said to be capable of immunospecifically binding to a target region or conformation (“epitope”) of an antigen if such binding involves the antigen recognition site of the immunoglobulin molecule. An antibody that immunospecifically binds to a particular antigen may bind to other antigens with lower affinity if the other antigen has some sequence or conformational similarity that is recognized by the antigen recognition site as determined by, e.g., immunoassays, BIACORE® assays, or other assays known in the art, but would not bind to a totally unrelated antigen. Preferably, however, antibodies (and their antigen binding fragments) will not cross-react with other antigens. Antibodies may also bind to other molecules in a way that is not immunospecific, such as to FcR receptors, by virtue of binding domains in other regions/domains of the molecule that do not involve the antigen recognition site, such as the Fc region. As used herein, a molecule is said to “physiospecifically bind” a second molecule if such binding exhibits the specificity and affinity of a receptor to its cognate binding ligand. A molecule can be capable of physiospecifically binding to more than one other molecule. As used herein, the term “antibody” is intended to denote an immunoglobulin molecule that possesses a “variable region” antigen recognition site. The term “variable region” is intended to distinguish such domain of the immunoglobulin from domains that are broadly shared by antibodies (such as an antibody Fc domain). The variable region includes a “hypervariable region” whose residues are responsible for antigen binding. The hypervariable region includes amino acid residues from a “Complementarity Determining Region” or “CDR” (i.e., typically at approximately residues 24-34 (L1), 50-56 (L2) and 89-97 (L3) in the light chain variable domain and at approximately residues 27-35 (H1), 50-65 (H2) and 95-102 (H3) in the heavy chain variable domain; Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD. (1991)) and/or those residues from a “hypervariable loop” (i.e., residues 26-32 (L1), 50-52 (L2) and 91-96 (L3) in the light chain variable domain and 26-32 (H1), 53-55 (H2) and 96-101 (H3) in the heavy chain variable domain; CHOTHIA et al., J Mol Biol, 196:901-917 (1987). “Framework Region” or “FR” residues are those variable domain residues other than the hypervariable region residues as herein defined. The term antibody includes monoclonal antibodies, multi-specific antibodies, human antibodies, humanized antibodies, synthetic antibodies, chimeric antibodies, camelized antibodies (See e.g., Muyldermans et al., Trends Biochem Sci, 26:230 (2001); Nuttall et al., Cur Pharm Biotech, 1:253 (2000); Reichmann et al., J Immunol Meth, 231:25 (1999); International Publication Nos. WO 94/04678 and WO 94/25591; U.S. Patent No.6,005,079), single-chain Fvs (scFv) (see, e.g., see Pluckthun in The Pharmacology of Monoclonal Antibodies, Vol.113, Rosenburg and Moore Eds. Springer-Verlag, New York, pp.269-315 (1994)), single chain antibodies, disulfide-linked Fvs (sdFv), intrabodies, and anti-idiotypic (anti-Id) antibodies (including, e.g., anti-Id and anti-anti-Id antibodies to antibodies). In particular, such antibodies include immunoglobulin molecules of any type (e.g., IgG, IgE, IgM, IgD, IgA and IgY), class (e.g., IgG1, IgG2, IgG3, IgG4, IgA1 and IgA2) or subclass. As used herein, the term “antigen binding fragment” of an antibody refers to one or more portions of an antibody that contain the antibody’s Complementarity Determining Regions (“CDRs”) and optionally the framework residues that include the antibody’s “variable region” antigen recognition site, and exhibit an ability to immunospecifically bind antigen. Such fragments include Fab', F(ab')2, Fv, single chain (ScFv), and mutants thereof, naturally occurring variants, and fusion proteins including the antibody’s “variable region” antigen recognition site and a heterologous protein (e.g., a toxin, an antigen recognition site for a different antigen, an enzyme, a receptor or receptor ligand, etc.). As used herein, the term “fragment” refers to a peptide or polypeptide including an amino acid sequence of at least 5 contiguous amino acid residues, at least 10 contiguous amino acid residues, at least 15 contiguous amino acid residues, at least 20 contiguous amino acid residues, at least 25 contiguous amino acid residues, at least 40 contiguous amino acid residues, at least 50 contiguous amino acid residues, at least 60 contiguous amino residues, at least 70 contiguous amino acid residues, at least 80 contiguous amino acid residues, at least 90 contiguous amino acid residues, at least 100 contiguous amino acid residues, at least 125 contiguous amino acid residues, at least 150 contiguous amino acid residues, at least 175 contiguous amino acid residues, at least 200 contiguous amino acid residues, or at least 250 contiguous amino acid residues. As used herein, the terms “treat,” “treating,” “treatment” and “therapeutic use” refer to the elimination, reduction or amelioration of one or more symptoms of a disease or disorder. As used herein, a “therapeutically effective amount” refers to that amount of a therapeutic agent sufficient to mediate a clinically relevant elimination, reduction or amelioration of such symptoms. An effect is clinically relevant if its magnitude is sufficient to impact the health or prognosis of a recipient subject. A therapeutically effective amount may refer to the amount of therapeutic agent sufficient to delay or minimize the onset of disease, e.g., delay or minimize the spread of cancer. A therapeutically effective amount may also refer to the amount of the therapeutic agent that provides a therapeutic benefit in the treatment or management of a disease. As used herein, an “immune cell” refers to any cell from the hemopoietic origin including, but not limited to, T cells, B cells, natural killer cells, NKT cells, monocytes, dendritic cells, macrophages, neutrophils, eosinophils, basophils, mast cells, innate lymphoid cells (ILCs), and myeloid derived suppressor cells (MDSCs). As used herein, the terms “individual,” “host,” “subject, and “patient” are used interchangeably herein, and refer to a mammal, including, but not limited to, humans, rodents, such as mice and rats, and other laboratory animals. II. Chimeric Antigen Receptor T cells Chimeric antigen receptor T cell (CARTs) compositions and methods of their use are provided. CARTs are synthetic constructs that are designed to be expressed in host T cells or NK cells and to induce an immune response against a specific target antigen and cells expressing that antigen. The CAR typically comprises an antibody fragment, such as a scFv or Fab fragment, incorporated in a fusion protein that also comprises additional components, such as a CD3- or CD28 transmembrane domain and selective T-cell activating moieties, including the endodomains of CD3- CD28, OX40, 4-1 BB, Lck and/or ICOS. Generally, the constructs may comprise a leader sequence linked to a scFv, Fab or other antibody moiety, generally with a hinge or other linker between the scEv and a transmembrane domain. The transmembrane domain will be attached to an intracellular signaling domain, such as CD28 or CD3-, and typically will include one or more co-stimulatory domains as dis cussed below. CARs are designed with an activation domain, costimulatory domain, transmembrane domain and antigen- binding domain. The activation domain is an intracellular domain, such as CD3. The costimulatory domain can increase the activity of the CAR T cell, like proliferation and persistence. The transmembrane domain plays a role of the structural anchor. The antigen- binding domain which can combine with the target antigen is an important component of CARs and usually includes a single chain variable fragment (scFvs). As described in U S. Pat. Nos.5,359,046, 5,686,281 and 6,103,521, the extracellular domain may be obtained from any of the wide variety of extracellular domains or secreted proteins associated with ligand binding and/or signal transduction. The extracellular domain may be part of a protein which is monomeric, homodimeric, heterodimeric, or associated with a larger number of proteins in a non-covalent complex. In particular, the extracellular domain may consist of an Ig heavy chain which may in turn be covalently associated with Ig light chain by virtue of the presence of CH1 and hinge regions, or may become covalently associated with other Ig heavy / light chain complexes by virtue of the presence of hinge, CH2 and CH3 domains. In the latter case, the heavy/light chain complex that becomes joined to the chimeric construct may constitute an antibody with a specificity distinct from the antibody specificity of the chimeric construct. Depending on the function of the antibody, the desired structure and the signal transduction, the entire chain may be used or a truncated chain may be used, where all or a part of the CH1, CH2, or CH3 domains may be removed or all or part of the hinge region may be removed. Chimeric antigen receptor T cells (CAR-T cells) are T cells that have been genetically engineered to express and produce an chimeric T cell receptor. This gives the engineered T cells the ability to target a specific protein. The basis of CAR-T immunotherapy is to modify T cells to recognize cancer cells in order to more effectively target and destroy them. T cells are harvested from a subject by leukapheresis, followed by elutriation to remove myeloid cells, T lymphocyte enrichment, transgene delivery, and ex vivo expansion. The resulting CAR-T cells are infused into subjects to attack their tumors. CAR-T cells can be either derived from T cells in a subject's own blood (autologous) or derived from the T cells of another healthy donor (allogeneic). Once isolated from a subject, these T cells are genetically engineered to express a specific CAR, which programs them to target an antigen that is present on the surface of tumors. For safety, CAR-T cells are engineered to be specific to an antigen expressed on a tumor that is not expressed on healthy cells. Methods of producing CAR T cells are known in the art and is described in U.S. Patent Nos.7,446,190, 7,741,465, 9,181,527, and 9,629,877. Genetic modification for introduction of the CAR construct into T cells can be accomplished by transducing (or otherwise delivering) a T cell composition with a recombinant DNA or RNA construct, such as for example, a vector. A vector may be any agent capable of delivering or maintaining nucleic acid in a host cell , and includes viral vectors (e.g. retroviral vectors, lentiviral vectors, adenoviral vectors, or adeno-associated viral vectors), plasmids, naked nucleic acids, nucleic acids complexed with polypeptide or other molecules and nucleic acids immobilized onto solid phase particles. The appropriate DNA sequence may be inserted into the vector by a variety of procedures. In general, the DNA sequence is inserted into an appropriate restriction endonuclease site(s) by procedures known in the art. Such procedures and others are deemed to be within the scope of those skilled in the art. To achieve durable clinical responses to cell-based gene therapies, permanent transgene expression is often required. Murine gammaretroviruses and lentiviruses are two available clinical gene therapy vector systems that afford long-term CAR transgene expression. Considerable clinical evidence shows that retroviral vectors are safe when expressed in human T cells. Selection of promoter and other regulatory sequences for protein expression are well known to those of skill in the art. Cell specific promoters for expression in T-cells include, but are not limited to, human CD2, distal Lck, and proximal Lck. In other embodiments, non-tissue specific promoters such as non-tissue specific promoters including viral promoters such as cytomegalovirus (CMV) promoter, ß-actin promoter phosphoglycerate kinase (PGK) promoter, ubiquitin promoter, and EF-1a promoter can be used. This list is not meant to be limiting. An expression construction preferably also includes sequences to allow for the replication of the expression construct. Transcription of the DNA encoding the polypeptides of the present invention by higher eukaryotes may be increased by inserting an enhancer sequence into the vector. Enhancers are cis-acting elements of DNA, usually about from 10 to 300 by that act on a promoter to increase its transcription. Examples including the SV40 enhancer on the late side of the replication origin by 100 to 270, a cytomegalovirus early promoter enhancer, the polyoma enhancer on the late side of the replication origin, and adenovirus enhancers. Preferably, a retroviral vector (either gamma- retroviral or lentiviral ) is employed for the introduction of the CAR nucleic acid construct into the cell . For example , a polynucleotide encoding a co - stimulatory ligand protein (e.g., tumor necrosis factor (TNF) ligand, such as 4-1BBL, OX40L, CD70, LIGHT, and CD30L, or an Ig superfamily ligand, such as CD80 and CD86), or a receptor that binds an antigen, or a variant, or a fragment thereof, can be cloned into a retroviral vector and expression can be driven from its endogenous promoter, from the retroviral long terminal repeat, or from a promoter specific for a target cell type of interest. Non-viral vectors may be used as well. One embodiment provides a chimeric antigen receptor T cell composition including 8F8 chimeric antigen receptor T cells. Another embodiment provides a chimeric antigen receptor T cell composition including 6G11 chimeric antigen receptor T cells. Still another embodiment provide a chimeric antigen receptor T cell composition including 12D7 chimeric antigen receptor T cells. The disclosed CARTs can be used to treat cancer, for example hepatocellular carcinoma. Other embodiments are directed to antibodies used to produce the CARTs. One embodiment provides a recombinant monoclonal antibody having a mAb 6G11 variable light chain and a mAb 6G11 variable heavy chain. Another embodiment provides a recombinant monoclonal antibody having mAb 8F8 variable light chain and mAb 8F8 variable heavy chain. Still another embodiment provides a recombinant monoclonal antibody comprising mAb 12D7 variable light chain and mAb 12D7 variable heavy chain. One embodiment provides a pharmaceutical composition including the disclosed antibodies and or CARTs. Yet another embodiment provides a method for treating liver cancer by administering to a subject in need thereof an effective amount of 6G11, 8F8, or 12D7 chimeric antigen receptor T cells to treat cancer, for example hepatocellular carcinoma. One embodiment provides a method for treating hepatocellular carcinoma in a subject in need thereof including the steps of isolating T cells from the subject, transfecting the isolated T cells with a vector containing 6G11-CAR nucleic acid sequences, optionally expanding the transfected cells in cell culture, and administering an effective mount of the transfected cells to the subject to treat hepatocellular carcinoma. One embodiment provides a method for treating hepatocellular carcinoma in a subject in need thereof including the steps of isolating T cells from the subject, transfecting the isolated T cells with a vector containing 8F8-CAR nucleic acid sequences, optionally expanding the transfected cells in cell culture, and administering an effective mount of the transfected cells to the subject to treat hepatocellular carcinoma. One embodiment provides a method for treating hepatocellular carcinoma in a subject in need thereof including the steps of isolating T cells from the subject, transfecting the isolated T cells with a vector containing 12D7-CAR nucleic acid sequences, optionally expanding the transfected cells in cell culture, and administering an effective mount of the transfected cells to the subject to treat hepatocellular carcinoma. III. Pharmaceutical Compositions The antibodies and CARTs described herein can be formulated into pharmaceutical compositions. Pharmaceutical compositions containing antibodies and/or the CARTs can be formulated for parenteral administration including by not limited to intramuscular, intraperitoneal, intravenous (IV) or subcutaneous injection. In some in vivo approaches, the compositions disclosed herein are administered to a subject in a therapeutically effective amount. As used herein the term “effective amount” or “therapeutically effective amount” means a dosage sufficient to treat, inhibit, or alleviate one or more symptoms of the disorder being treated or to otherwise provide a desired pharmacologic and/or physiologic effect. The precise dosage will vary according to a variety of factors such as subject-dependent variables (e.g., age, immune system health, etc.), the disease, and the treatment being effected. For the disclosed antibodies, as further studies are conducted, information will emerge regarding appropriate dosage levels for treatment of various conditions in various patients, and the ordinary skilled worker, considering the therapeutic context, age, and general health of the recipient, will be able to ascertain proper dosing. The selected dosage depends upon the desired therapeutic effect, on the route of administration, and on the duration of the treatment desired. Generally, for intravenous injection or infusion, dosage may be lower. The dosage administered to a patient is typically 0.01 mg/kg to 100 mg/kg of the patient's body weight. The dosage administered to a patient can be, for example, between 0.01 mg/kg and 20 mg/kg, 0.01 mg/kg and 10 mg/kg, 0.01 mg/kg and 5 mg/kg, 0.01 and 2 mg/kg, 0.01 and 1 mg/kg, 0.01 mg/kg and 0.75 mg/kg, 0.01 mg/kg and 0.5 mg/kg, 0.01 mg/kg to 0.25 mg/kg, 0.01 to 0.15 mg/kg, 0.01 to 0.10 mg/kg, 0.01 to 0.05 mg/kg, or 0.01 to 0.025 mg/kg of the patient's body weight. Exemplary specific dosages include, but are not limited to 0.2 mg/kg, 0.3 mg/kg, 1 mg/kg, 3 mg/kg, 6 mg/kg or 10 mg/kg. A dose as low as 0.01 mg/kg is believed to be suitable to have appreciable pharmacodynamic effects. Dose levels of 0.10-1 mg/kg are predicted to be most appropriate. Higher doses (e.g., 1-30 mg/kg) would also be expected to be active. Generally, human antibodies have a longer half-life within the human body than antibodies from other species due to the immune response to the foreign polypeptides. Thus, lower dosages of human antibodies and less frequent administration is often possible. Further, the dosage and frequency of administration of antibodies of the invention or fragments thereof may be reduced by enhancing uptake and tissue penetration of the antibodies by modifications such as, for example, lipidation. In certain embodiments, the antibodies and/or CARTs are administered locally, for example by injection directly into a site to be treated. In some embodiments, compositions disclosed herein, are administered in an aqueous solution, by parenteral injection or infusion. The formulation may also be in the form of a suspension or emulsion. In general, pharmaceutical compositions are provided including effective amounts of antibodies and/or CARTs, and optionally include pharmaceutically acceptable diluents, preservatives, solubilizers, emulsifiers, adjuvants and/or carriers. Such compositions optionally include one or more for the following: diluents, sterile water, buffered saline of various buffer content (e.g., Tris-HCl, acetate, phosphate), pH and ionic strength; and additives such as detergents and solubilizing agents (e.g., TWEEN® 20 (polysorbate-20), TWEEN® 80 (polysorbate-80)), anti-oxidants (e.g., ascorbic acid, sodium metabisulfite), and preservatives (e.g., Thimersol, benzyl alcohol) and bulking substances (e.g., lactose, mannitol). Examples of non-aqueous solvents or vehicles are propylene glycol, polyethylene glycol, vegetable oils, such as olive oil and corn oil, gelatin, and injectable organic esters such as ethyl oleate. The formulations may be lyophilized and redissolved/resuspended immediately before use. The formulation may be sterilized by, for example, filtration through a bacteria retaining filter, by incorporating sterilizing agents into the compositions, by irradiating the compositions, or by heating the compositions. IV. Combination Therapies The disclosed antibodies and/or CARTs can be administered to a subject in need thereof alone or in combination with one or more additional therapeutic agents. In some embodiments, the additional therapeutic agent are administered separately, but simultaneously. The additional therapeutic agent can also be administered as part of the same composition. In other embodiments, the antibodies and/or CARTs and the second therapeutic agent are administered separately and at different times, but as part of the same treatment regime. The subject can be administered a first therapeutic agent 1, 2, 3, 4, 5, 6, or more hours, or 1, 2, 3, 4, 5, 6, 7, or more days before administration of a second therapeutic agent. In some embodiments, the subject can be administered one or more doses of the first agent every 1, 2, 3, 4, 5, 67, 14, 21, 28, 35, or 48 days prior to a first administration of second agent. antibodies and/or CARTs can be the first or the second therapeutic agent. In some embodiments, the antibodies and/or CARTs s are administered in combination with a second therapeutic agent. The antibodies and/or CARTs and the additional therapeutic agent can be administered as part of a therapeutic regimen. For example, if a first therapeutic agent can be administered to a subject every fourth day, the second therapeutic agent can be administered on the first, second, third, or fourth day, or combinations thereof. The first therapeutic agent or second therapeutic agent may be repeatedly administered throughout the entire treatment regimen. Exemplary additional therapeutic agents include, but are not limited to, cytokines, chemotherapeutic agents, radionuclides, other immunotherapeutics, enzymes, antibiotics, antivirals (especially protease inhibitors alone or in combination with nucleosides for treatment of HIV or Hepatitis B or C), anti-parasites (helminths, protozoans), growth factors, growth inhibitors, hormones, hormone antagonists, antibodies and bioactive fragments thereof (including humanized, single chain, and chimeric antibodies), antigen and vaccine formulations (including adjuvants), peptide drugs, anti-inflammatories, ligands that bind to Toll-Like Receptors (including but not limited to polyinosinic:polycytidylic acid (polyI:C) and CpG oligonucleotides) to activate the innate immune system, molecules that mobilize and optimize the adaptive immune system, other molecules that activate or up-regulate the action of cytotoxic T lymphocytes, natural killer cells and helper T-cells, and other molecules that deactivate or down-regulate suppressor or regulatory T-cells. The additional therapeutic agents are selected based on the condition, disorder or disease to be treated. For example, the antibodies and/or CARTs can be co-administered with one or more additional agents that function to enhance or promote an immune response or reduce or inhibit an immune response. A. Chemotherapeutic Agents The antibodies and/or CARTs can be combined with one or more chemotherapeutic agents and pro-apoptotic agents. Representative chemotherapeutic agents include, but are not limited to amsacrine, bleomycin, busulfan, capecitabine, carboplatin, carmustine, chlorambucil, cisplatin, cladribine, clofarabine, crisantaspase, cyclophosphamide, cytarabine, dacarbazine, dactinomycin, daunorubicin, docetaxel, doxorubicin, epirubicin, etoposide, fludarabine, fluorouracil, gemcitabine, hydroxycarbamide, idarubicin, ifosfamide, irinotecan, leucovorin, liposomal doxorubicin, liposomal daunorubicin, lomustine, melphalan, mercaptopurine, mesna, methotrexate, mitomycin, mitoxantrone, oxaliplatin, paclitaxel, pemetrexed, pentostatin, procarbazine, raltitrexed, satraplatin, streptozocin, tegafur-uracil, temozolomide, teniposide, thiotepa, tioguanine, topotecan, treosulfan, vinblastine, vincristine, vindesine, vinorelbine, or a combination thereof. Representative pro-apoptotic agents include, but are not limited to fludarabinetaurosporine, cycloheximide, actinomycin D, lactosylceramide, 15d-PGJ(2) and combinations thereof. B. Other Immunomodulators 1. PD-1 antagonists In some embodiments, the antibodies and/or CARTs are co-administered with a PD-1 antagonist. Programmed Death-1 (PD-1) is a member of the CD28 family of receptors that delivers a negative immune response when induced on T cells. Contact between PD-1 and one of its ligands (B7-H1 or B7-DC) induces an inhibitory response that decreases T cell multiplication and/or the strength and/or duration of a T cell response. Suitable PD-1 antagonists are described in U.S. Patent Nos.8,114,845, 8,609,089, and 8,709,416, which are specifically incorporated by reference herein in their entities, and include compounds or agents that either bind to and block a ligand of PD-1 to interfere with or inhibit the binding of the ligand to the PD- 1 receptor, or bind directly to and block the PD-1 receptor without inducing inhibitory signal transduction through the PD-1 receptor. In some embodiments, the PD-1 receptor antagonist binds directly to the PD-1 receptor without triggering inhibitory signal transduction and also binds to a ligand of the PD-1 receptor to reduce or inhibit the ligand from triggering signal transduction through the PD-1 receptor. By reducing the number and/or amount of ligands that bind to PD-1 receptor and trigger the transduction of an inhibitory signal, fewer cells are attenuated by the negative signal delivered by PD-1 signal transduction and a more robust immune response can be achieved. It is believed that PD-1 signaling is driven by binding to a PD-1 ligand (such as B7-H1 or B7-DC) in close proximity to a peptide antigen presented by major histocompatibility complex (MHC) (see, for example, Freeman, Proc Natl Acad Sci U. S. A, 105:10275-10276 (2008)). Therefore, proteins, antibodies or small molecules that prevent co-ligation of PD-1 and TCR on the T cell membrane are also useful PD-1 antagonists. In some embodiments, the PD-1 receptor antagonists are small molecule antagonists or antibodies that reduce or interfere with PD-1 receptor signal transduction by binding to ligands of PD-1 or to PD-1 itself, especially where co-ligation of PD-1 with TCR does not follow such binding, thereby not triggering inhibitory signal transduction through the PD-1 receptor. Other PD-1 antagonists contemplated by the methods of this invention include antibodies that bind to PD-1 or ligands of PD-1, and other antibodies. Suitable anti-PD-1 antibodies include, but are not limited to, those described in the following US Patent Nos: 7332582, 7488802, 7521051, 7524498, 7563869, 7981416, 8088905, 8287856, 8580247, 8728474, 8779105, 9067999, 9073994, 9084776, 9205148, 9358289, 9387247, 9492539, 9492540,, all of which are incorporated by reference in their entireties. See also Berger et al., Clin Cancer Res, 14:30443051 (2008). Exemplary anti-B7-H1 (also referred to as anti-PD-L1) antibodies include, but are not limited to, those described in the following US Pat Nos: 8383796, 9102725, 9273135, 9393301, and 9580507 all of which are specifically incorporated by reference herein in their entirety. For anti-B7-DC (also referred to as anti-PD-L2) antibodies see US Pat. Nos.: 7,411,051, 7,052,694, 7,390,888, 8,188,238, and 9,255,147 all of which are specifically incorporated by reference herein in their entirety. Other exemplary PD-1 receptor antagonists include, but are not limited to B7-DC polypeptides, including homologs and variants of these, as well as active fragments of any of the foregoing, and fusion proteins that incorporate any of these. In some embodiments, the fusion protein includes the soluble portion of B7-DC coupled to the Fc portion of an antibody, such as human IgG, and does not incorporate all or part of the transmembrane portion of human B7-DC. The PD-1 antagonist can also be a fragment of a mammalian B7-H1, for example from mouse or primate, such as a human, wherein the fragment binds to and blocks PD-1 but does not result in inhibitory signal transduction through PD-1. The fragments can also be part of a fusion protein, for example an Ig fusion protein. Other useful polypeptides PD-1 antagonists include those that bind to the ligands of the PD-1 receptor. These include the PD-1 receptor protein, or soluble fragments thereof, which can bind to the PD-1 ligands, such as B7-H1 or B7-DC, and prevent binding to the endogenous PD-1 receptor, thereby preventing inhibitory signal transduction. B7-H1 has also been shown to bind the protein B7.1 (Butte et al., Immunity, Vol.27, pp.111-122, (2007)). Such fragments also include the soluble ECD portion of the PD-1 protein that includes mutations, such as the A99L mutation, that increases binding to the natural ligands (Molnar et al., PNAS, 105:10483-10488 (2008)). B7-1 or soluble fragments thereof, which can bind to the B7-H1 ligand and prevent binding to the endogenous PD-1 receptor, thereby preventing inhibitory signal transduction, are also useful. PD-1 and B7-H1 anti-sense nucleic acids, both DNA and RNA, as well as siRNA molecules can also be PD-1 antagonists. Such anti-sense molecules prevent expression of PD-1 on T cells as well as production of T cell ligands, such as B7-H1, PD-L1 and/or PD-L2. For example, siRNA (for example, of about 21 nucleotides in length, which is specific for the gene encoding PD-1, or encoding a PD-1 ligand, and which oligonucleotides can be readily purchased commercially) complexed with carriers, such as polyethyleneimine (see Cubillos-Ruiz et al., J. Clin. Invest.119(8): 2231-2244 (2009), are readily taken up by cells that express PD-1 as well as ligands of PD-1 and reduce expression of these receptors and ligands to achieve a decrease in inhibitory signal transduction in T cells, thereby activating T cells. 2. CTLA4 antagonists In some embodiments, the second therapeutic agent is an antagonist of CTLA4, for example an antagonistic anti-CTLA4 antibody. An example of an anti-CTLA4 antibody contemplated for use in the methods of the invention includes an antibody as described in PCT/US2006/043690 (Fischkoff et al., WO/2007/056539). Dosages for anti-PD-1, anti-B7-H1, and anti-CTLA4 antibody, are known in the art and can be in the range of, for example, 0.1 to 100 mg/kg, or with shorter ranges of 1 to 50 mg/kg, or 10 to 20 mg/kg. An appropriate dose for a human subject can be between 5 and 15 mg/kg, with 10 mg/kg of antibody (for example, human anti-PD-1 antibody) being a specific embodiment. Specific examples of an anti-CTLA4 antibody useful in the methods of the invention are Ipilimumab, a human anti-CTLA4 antibody, administered at a dose of, for example, about 10 mg/kg, and Tremelimumab a human anti-CTLA4 antibody, administered at a dose of, for example, about 15 mg/kg. See also Sammartino, et al., Clinical Kidney Journal, 3(2):135-137 (2010), published online December 2009. In other embodiments, the antagonist is a small molecule. A series of small organic compounds have been shown to bind to the B7-1 ligand to prevent binding to CTLA4 (see Erbe et al., J Biol Chem, 277:7363-7368 (2002). Such small organics could be administered alone or together with an anti-CTLA4 antibody to reduce inhibitory signal transduction of T cells. 3. Potentiating Agents In some embodiments, additional therapeutic agents include a potentiating agent. The potentiating agent acts to increase efficacy the immune response up-regulator, possibly by more than one mechanism, although the precise mechanism of action is not essential to the broad practice of the present invention. In some embodiments, the potentiating agent is cyclophosphamide. Cyclophosphamide (CTX, Cytoxan®, or Neosar®) is an oxazahosphorine drug and analogs include ifosfamide (IFO, Ifex), perfosfamide, trophosphamide (trofosfamide; Ixoten), and pharmaceutically acceptable salts, solvates, prodrugs and metabolites thereof (US patent application 20070202077 which is incorporated in its entirety). Ifosfamide (MITOXANA®) is a structural analog of cyclophosphamide and its mechanism of action is considered to be identical or substantially similar to that of cyclophosphamide. Perfosfamide (4-hydroperoxycyclophosphamide) and trophosphamide are also alkylating agents, which are structurally related to cyclophosphamide. For example, perfosfamide alkylates DNA, thereby inhibiting DNA replication and RNA and protein synthesis. New oxazaphosphorines derivatives have been designed and evaluated with an attempt to improve the selectivity and response with reduced host toxicity (Liang J, et al., “Design of New Oxazaphosphorine Anticancer Drugs”, Curr Pharm Des, 13(9):963-78 Review 2007). These include mafosfamide (NSC 345842), glufosfamide (D19575, beta-D- glucosylisophosphoramide mustard), S-(-)-bromofosfamide (CBM-11), NSC 612567 (aldophosphamide perhydrothiazine) and NSC 613060 (aldophosphamide thiazolidine). Mafosfamide is an oxazaphosphorine analog that is a chemically stable 4-thioethane sulfonic acid salt of 4-hydroxy-CPA. Glufosfamide is IFO derivative in which the isophosphoramide mustard, the alkylating metabolite of IFO, is glycosidically linked to a beta-D-glucose molecule. Additional cyclophosphamide analogs are described in US patent 5,190,929 entitled “Cyclophosphamide analogs useful as anti-tumor agents” which is incorporated herein by reference in its entirety. While CTX itself is nontoxic, some of its metabolites are cytotoxic alkylating agents that induce DNA crosslinking and, at higher doses, strand breaks. Many cells are resistant to CTX because they express high levels of the detoxifying enzyme aldehyde dehydrogenase (ALDH). CTX targets proliferating lymphocytes, as lymphocytes (but not hematopoietic stem cells) express only low levels of ALDH, and cycling cells are most sensitive to DNA alkylation agents. Low doses of CTX (< 200 mg/kg) can have immune stimulatory effects, including stimulation of anti-tumor immune responses in humans and mouse models of cancer (Brode & Cooke Crit Rev, Immunol, 28:109-126 (2008)). These low doses are sub-therapeutic and do not have a direct anti-tumor activity. In contrast, high doses of CTX inhibit the anti-tumor response. Several mechanisms may explain the role of CTX in potentiation of anti-tumor immune response: (a) depletion of CD4+CD25+FoxP3+ Treg (and specifically proliferating Treg, which may be especially suppressive), (b) depletion of B lymphocytes; (c) induction of nitric oxide (NO), resulting in suppression of tumor cell growth; (d) mobilization and expansion of CD11b+Gr-1+ MDSC. These primary effects have numerous secondary effects; for example following Treg depletion macrophages produce more IFN-g and less IL-10. CTX has also been shown to induce type I IFN expression and promote homeostatic proliferation of lymphocytes. Treg depletion is most often cited as the mechanism by which CTX potentiates the anti- tumor immune response. This conclusion is based in part by the results of adoptive transfer experiments. In the AB1-HA tumor model, CTX treatment at Day 9 gives a 75% cure rate. Transfer of purified Treg at Day 12 almost completely inhibited the CTX response (van der Most et al., Cancer Immunol. Immunother, 58:1219-1228 (2009). A similar result was observed in the HHD2 tumor model: adoptive transfer of CD4+CD25+ Treg after CTX pretreatment eliminated therapeutic response to vaccine (Taieb, J., J Immunol, 176:2722-2729 (2006)). Numerous human clinical trials have demonstrated that low dose CTX is a safe, well- tolerated, and effective agent for promoting anti-tumor immune responses (Bas, & Mastrangelo Cancer Immunol, Immunother, 47:1-12 (1998)). The optimal dose for CTX to potentiate an anti-tumor immune response, is one that lowers overall T cell counts by lowering Treg levels below the normal range but is subtherapeutic (see Machiels et al., Cancer Res, 61:3689-3697 (2001)). In human clinical trials where CTX has been used as an immunopotentiating agent, a dose of 300 mg/m2 has usually been used. For an average male (6 ft., 170 pound (78 kg) with a body surface area of 1.98 m2), 300 mg/m2 is 8 mg/kg, or 624 mg of total protein. In mouse models of cancer, efficacy has been seen at doses ranging from 15 – 150 mg/kg, which relates to 0.45 – 4.5 mg of total protein in a 30g mouse (Machiels et al., Cancer Res, 61:3689-3697 (2001), Hengst et al., Cancer Res, 41:2163-2167 (1981), Hengst, Cancer Res, 40:2135-2141 (1980)). For larger mammals, such as a primate, such as a human, patient, such mg/m ² doses may be used but unit doses administered over a finite time interval may also be used. Such unit doses may be administered on a daily basis for a finite time period, such as up to 3 days, or up to 5 days, or up to 7 days, or up to 10 days, or up to 15 days or up to 20 days or up to 25 days, are all specifically contemplated by the invention. The same regimen may be applied for the other potentiating agents recited herein. In other embodiments, the potentiating agent is an agent that reduces activity and/or number of regulatory T lymphocytes (T-regs), such as Sunitinib (SUTENT®), anti-TGFb or Imatinib (GLEEVAC®). The recited treatment regimen may also include administering an adjuvant. Useful potentiating agents also include mitosis inhibitors, such as paclitaxol, aromatase inhibitors (e.g. Letrozole) and angiogenesis inhibitors (VEGF inhibitors e.g., Avastin, VEGF- Trap) (see, for example, Li et al., Clin Cancer Res, Nov 15; 12(22):6808-16 (2006.)), anthracyclines, oxaliplatin, doxorubicin, TLR4 antagonists, and IL-18 antagonists. EXAMPLES Example I: 1. Generation of hGPC3-specific mAbs that bind HCC tumor cells. Materials and Methods Cells Peripheral blood mononuclear cells (PBMC) buffy coats were obtained from local blood bank and PBMC were separated by Ficoll-Hypaque column and frozen in aliquots in liquid nitrogen. Jurkat, HEK293T, and HepG2 cells were from the American Type Culture Collection; Huh7 cells were from Dr. Satyanayana Ande of Augusta University Georgia Cancer Center. Cells were checked for mycoplasma by a PCR test (Fisher Scientific) and cultured for no more than eight passages before use to assure their authenticity. Mice BalB/C mice were purchased from Charles River. The immunocompromised NOD-scid IL2Rgammanull (NSG) mice breeding pairs were obtained from Jackson Laboratory and bred in house. Animal protocols were approved by the Institutional Animal Care and Use Committee of Augusta University. Generation of mAbs The mAbs were generated as previously descried (Xiao H, et al., Blood Cells Mol Dis 44: 127-32 (2010)). Female BALB/c mice of age 6-10 weeks were immunized with 100µg recombinant hGPC3 protein (AA25-560) with complete Freund’s adjuvant (Sigma-Aldrich) by footpad injection. Three weeks later, mice were boosted twice or three times with 70µg recombinant hGPC3 protein together with incomplete Freund’s adjuvant (Sigma-Aldrich) by footpad injection at 2-week intervals. Anti-hGPC3 antibody titer in mouse serum was measured by ELISA. A final boost was carried out by tail vein injection of 50 µg/mouse recombinant hGPC3 in 100 µl PBS. Three days after the final boost, mice were sacrificed and splenocytes were generated to fuse with fusion partner NS-1 myeloma cells in the presence of 50% PEG1500. Hybridomas were selected by HAT medium, and the positive ones were screened by ELISA. Sub-cloning was performed three times using the limiting dilution method. Statistical analysis Statistical analyses were performed using Stundent’s t test or 2way ANOVA in the prism software (GraphPad Inc.). Results In the first set of experiments, hGPC3-specific mAbs that bound HCC tumor cells were developed. By ELISA, using soluble hGPC3 protein, we identified 22 mAbs (Fig.1A). Based on heavy chains, there are 8 IgM, 7 IgG1, 3 IgG2a, 3 IgG2b, and 1 IgA. All the mAbs have the k chain. Among them, 11 mAbs strongly, and 3 weakly, stained HepG2 cells (Fig 1B). The other 8 mAbs were unable to stain HepG2 cells (Fig.1E). Furthermore, 7 of them (2D7, 3H4, 5D5, 6G11, 8F8, 10C6, and 12D7) that positively stained HepG2 cells also stained the hGPC3low Huh7 cells albeit with a lower intensity (Fig.1F). Because the hGPC3 molecule on cell surface can be cleaved into 40Kd N- and 30Kd C- fragments, we studied whether the mAbs bound to hGPC3 N- or C- fragments. To this end, we created two recombinant plasmids that expressed the amino acid (AA) 1-358 (hGPC3-N) and AA359-580 (hGPC3-C fragment), respectively (Fig.1C). 293T cells were transfected with plasmid DNA and stained with mAbs. We found that out of the 14 mAbs that stained HepG2 tumor cells, 11 of them bound to the hGPC3-N, while only 3 bound to the hGPC3-C fragment (Figs.1D and 1G). This data suggests that N-terminal epitopes might be more accessible. The other 8 mAbs that were unable to stain HepG2 cells did not bind to either hGPC3-N or hGPC3-C fragment (Fig.1G). The characterization of the 22 mAbs was summarized in Fig.1A. Example II: Identification of the mAbs that stain only the HCC tumor but not adjacent normal liver tissues. Materials and Methods Immunohistochemical staining (IHC) HCC and liver tissues were obtained from cancer clinics in Georgetown cancer center. Tissues were sectioned and stained in the clinical pathology lab of Georgetown cancer center. The paraformaldehyde-fixed and paraffin-embedded blocks of health liver, adjacent normal liver and HCC tissue are from cases operated at Lombardi Cancer Center of Georgetown University. All patients gave informed consent, and the study was authorized by the respective Hospital Ethics Committees. Pathologic slides were stained with one of the following purified mAbs (6G11, 12D7, and 12E6, 8C8 and 8F8). Secondary antibodies were either goat-anti-mouse IgM or goat-anti-mouse IgG (Vector Labs). Immunoscores were calculated by adding intensity and region of staining: 0, 1, 2, 3 correspond to no, weak, moderate, and high intensity staining; 0, 1, 2, 3 correspond to no, focal as less than 10% of cells, regional as 10%–50% of cells, and diffuse region as >50% of cells staining. Results 5 mAbs, 3 for N-fragment (6G11, 12D7, and 12E6) and 2 for C-fragment (8C8 and 8F8), were purified and assayed IHC staining of the HCC tumor tissues and their paired adjacent normal tissues to further study their specificity. Healthy liver tissues were also included as controls. Our data showed that 3 of the mAbs (6G11, 8F8, and 12D7) were able to specifically stain HCC tumor tissues but not the adjacent normal liver tissues (Figs.2A-2C). However, the other two mAbs (8C8 and 12E6) had a higher non-specific staining of normal adjacent liver tissues and even healthy liver tissues. Thus, only the mAbs 6G11, 8F8, and 12D7 were used for further studies. Example III: Identification of the mAb’s cDNA sequences. Materials and Methods Identification of the mAb’s cDNA sequences To obtain the cDNA sequences of mAbs, we used the technique of 5¢ rapid amplification of complementary DNA ends (5’ RACE) (Walchli, S., et al., PLoS One 6: e27930 (2011); Zhu, W., et al.,. Hepatology 68: 574-89 (2018)). Briefly, total RNA was isolated from hybridomas, and complementary DNA was generated with an oligo dT primer. PolyC was added to the 5¢ end of complementary DNA by terminal transferase. Then, PCR was conducted to amplify the cDNA by using the 5¢ pGI primer: , and one of the following 3¢ primers corresponding to the constant (C) region of the mouse IgM heavy chain or the kappa chain: 1. For 6G11 and 12D7, IgM heavy chain primer (mIgM-OP) and k chain primer (mIgk-OP) were designed as 3’ primers. a. Mouse IgM heavy chain primer (mIgM-OP): b. Mouse k chain primer (mIgk-OP): CACTGCCATCAATCTTCCACTTG 2. The PCR products were purified and sequenced. Based on the preliminary sequences, internal 3’ primers were designed. Together with 5’ pGI primer, a second and more specific PCR was run. a. Mouse IgM heavy chain internal 3’ primer (mIgM-IP): b. Mouse k chain internal 3’ primer (mIgk-IP): 3. Because the immunoglobulin class of 8F8 is IgG2a, different heavy chain 3’ primers were used: a. Mouse IgG2a heavy chain primer (mIgG2aH-OP): b. Mouse IgG2a heavy chain internal primer (mIgG2aH-IP): Results cDNA sequences of the heavy chain variable region (VH) and the light chain variable region (VL) of mAbs 6G11, 8F8, and 12D7 from their corresponding hybridoma clones were identified. The deduced AA sequences of the VL and VH were compared to the Ig sequences compiled in the international ImMunoGeneTics information system (IMGT) databank (37). The AA sequences showed that 6G11 VL has 99.64% homology with IGKV15-103*01 (1AA difference) and its J region is identical to IGKJ1*01. In contrast, the 6G11 VH has 94.44% homology with IGHV1S5*01 and its J region has 96.3% of homology to IGHJ4*01. The 8F8 VL has 99% homology with IGK15-135*01 and its J region is identical to IGKJ5*01, while 8F8 VH ass 94.9% homology with IGHV5-9-2*01 and its J region has 85.5% homology to IGHJ2*01. The 12D7 VL is identical to IGKV6-15*01 and IGKJ5*01, while 12D7 VH has 89.8% homology with IGHV1S5*01 and its J region has 85.7% homology with IGHJ3*01. The results were summarized in Table 1. Example IV: Affinity of mAbs 6G11, 8F8, and 12D7. Materials and Methods BioLayer Interferometry Kinetics experiments were performed using the ForteBio Octet Red96 platform. All data were collected at 25℃ with orbital shaking at 1,000 rpm in 200 µL in 96-well black flat bottom plates (Greiner Bio-one). Anti-hIgG Fc Capture Sensor (ForteBio) were equilibrated for 10 min in 1× kinetics buffer prior to loading with 5 µg/mL of mAbs for 2 min. After the loading step, a 30 s baseline in 1× kinetics buffer was established before association with a range of concentrations of GPC3 for 2 min. Following mAbs association with GP, there was a 10- minute of dissociation step in kinetic buffer. For all experiments, two reference sensors without association to mAbs were used to account for nonspecific binding of analyte to the sensor and the system drift. All data were fit globally to a 1:1 Langmuir binding model using data analysis software 9.0 (ForteBio). Results The 3 mAbs belong to different Ig classes, i.e., 6G11 and 12D7 are IgM, whereas 8F8 is IgG2a. To fairly compare their affinity, mAbs were generated recombinant using their VL and VH and a same human IgG constant region. Fig.3A illustrates the structure of the recombinant mAbs. The mAbs were expressed and purified from 293 cells (Fig.3B). The mAb’s affinity was first studied by ELISA and the data showed that all 3 mAbs have high affinity for hGPC3 (Fig.3C) (EC50 values in the range of subnanomolar to low nanomolar level). The mAb’s affinity was also studied by BioLayer Interferometry (BLI). By this biophysical approach, the Kd value of these mAbs binding to hGPC3 were at nanomolar levels, 7.4nM for 6G11, 9.8nM for 12D7, and 23nM for 8F8 (Fig.3D). The affinity data obtained by ELISA assay and BLI are in the same order, i.e., 6G11>12D7>8F8. Because the purpose of developing the mAbs is to bind hGPC3 molecules on the tumor cell surface, we determined the affinity of the mAbs for hGPC3 on HepG2 cells. Consistent with the ELISA and BLI data, the 6G11 and 12D7 have high affinity (EC50: 0.095nM and 0.1nM, respectively), while 8F8 mAb has low affinity for hGPC3 on HepG2 tumor cells (EC50: 12nM) (Fig.3E). Example V: Epitopes recognized by the mAbs. Materials and Methods ELISA assay with synthetic peptides 96-well Flat-Bottom Immuno Plates (from Thermo Fisher) were coated with 4ug of PBS- diluted peptides, overnight, at 4oC.1ug of rhGPC3 (from Creative Biomart) was used as positive control. The next day, peptides were discarded and plates were blocked with 100µl of 5% non-fat milk in PBS. Then, the plates were washed with 300µl/well of PBST (0.1% Tween20). Next, incubation with 100uL/well of different antibody dilutions. This was followed by four washes with PBST. Then, incubation with 100uL/well of anti-human IgG-HRP (From KPL) (1:2000 dilution). This was followed by four washes with PBST. Next, 100uL/well of TMB (from Biolengend), incubation for 30 minutes. Then, 100uL/well of 2N H2SO4. Lastly, OD450nm was detected in an xMark ^TM Microplate Spectrophotometer (from BIO-RAD). If not indicated otherwise, incubations were performed for 1 hour at RT, and antibodies were diluted in 2% non- fat milk in PBS. Results Fig.1C shows that mAbs 6G11 and 12D7 bound to hGPC3 N-fragment (AA1-358), while the 8F8 bound to the hGPC3 C-fragment (AA359-580). To further map the mAb’s epitopes, western blot analysis was constructed and confirmed that 6G11 and 12D7 mAbs bound hGPC3 N- and 8F8 bound hGPC3 C-fragment (Fig.4A). Next, using overlapping synthetic peptides and ELISA assay, we determined that 6G11’s epitope is AA25-39 of hGPC3 (Fig.4B). By a similar strategy, the epitope for 8F8 was located in the AA463-496 of hGPC3 (Fig.4C). However, no specific epitope was identified for 12D7. It is likely that 12D7 recognizes a conformational epitope in the hGPC3 N-fragment. The epitopes recognized by the mAbs are summarized and illustrated in Fig.4D. Example VI: Generation of CARTs. Materials and Methods The second generation of CARs was constructed as follow. The CAR includes the leader sequence of human CD8 molecule, the VL and VH, a glycine and serine (GS) linker between the VL and VH, a fragment of CD8 hinge, the transmembrane domain of human CD28, and the 4- 1BB and CD3 z chain signaling domains. The entire expression cassette was inserted behind the EF1a promoter in the lentivector (lv). Lentivector production was as described previously (4) by co-transfection of 293T cells with 4 plasmid DNA including the shuttle CAR expression shuttle plasmid, pLP1, pLP2 and pVSVG (Invitrogen). Virus particles in the media were concentrated by high speed centrifugation and sucrose gradient as previously described (4, 5). Virus titer was determined by transducing the Jurkat cells with a serial dilution of virus stock and anti-mouse Fab antibody staining (Jackson ImmunoReseach Lab). Transduction of human T cells Activation and transduction of primary human T cells were done as described (Zhu, W., et al., Hepatology 68: 574-89 (2018)). Briefly, human T cells were isolated from the buffy coat of healthy donors by negative selection (Stemcell Technologies, Inc) and activated by Dynabeads (Fisher Scientific) for 24hrs. The activated T cells were transduced with lv at 20-40 MOI. The beads were removed 24hrs after another 24hrs. Forty units of interleukin‐2 (IL‐2) were added to the RPMI medium every 3 days through the process. Results The second generation of CARTs was built using the cDNA sequences of 6G11, 8F8, and 12D7 mAbs. The CAR comprises a single-chain variable fragment derived of the mAbs, preceded by a CD8a leader peptide, followed by a CD28 transmembrane domain, linked to 41- BB and CD3 z intracellular signaling domains (Fig.5A). The lentivectors expressing the CAR genes were used to transduce human primary T cells (Fig.5B). All three CAR-expressing lentivectors could effectively transduce primary human T cells (Fig.5C). Three days after transduction, the 6G11, 8F8, and 12D7 CAR+ T cells were approximately 75%, 40%, and 40% of total T cells, respectively. However, the CAR expression on T cells was unstable. The % of CAR+ T cells gradually decreased and became stable when reaching 10-15% expression on day 13 after transduction (Fig.5C). It was also found that CARs on T cells were capable of binding soluble hGPC3 protein (Supplementary Fig.S5). The data showed that soluble hGPC3 could stain ~90% of 6G11 CARTs, but only ~40% of 8F8 and 12D7 CARTs. These data suggest that 6G11’s epitope and 6G11 CAR are more accessible to each other. Example VII: CARTs expansion in response to HepG2 tumor cell stimulation, but with significant different effector function. Materials and Methods Flow cytometry analysis The following antibodies were used in the staining: anti-human CD45, anti-mouse CD45, anti-mouse Fab, hGPC3 protein. Cells were stained with indicated antibodies and analyzed on LSR flow cytometer. Goat anti-mouse Fab antibody (Jackson ImmunoReseach Lab) and FITC- labeled hGPC3 were used to determine CAR transduction and the binding of CAR to cognate antigen. Results The responsiveness of CARTs to HCC tumor cells was studied. The data showed that all 3 CARTs could respond to HepG2 cell stimulation and undergo expansion (Fig.6A-B).6 days after HepG2 stimulation, the CAR+ T cells had increased from ~10% to as high as 70%. In fact, the % of 6G11 CAR+ Ts increased by 6 folds, and the % of 8F8 and 12D7 CAR+ Ts increased by ~4 folds. However, their effector functions were significantly different. The 8F8 CARTs generated the strongest cytotoxicity against HepG2, and produced the highest amount of IFNg and IL2 (Fig.6C). Albeit slightly weaker than 8F8 CARTs, the 6G11 CART also possessed potent cytotoxicity, but its cytokine production, especially IFNg, was significantly lower than 8F8 CARTs. On the other hand, 12D7 CARTs has the weakest effector function with undetectable cytotoxicity and cytokine production (Fig.6C). Importantly, following 6 days of antigen (HepG2 cell) stimulation, the 8F8 and 6G11 CARTs maintained their strong cytotoxicity and cytokine production (Fig.6D). The IL2 production by 6G11 and 8F8 CARTs was enhanced after HepG2 stimulation. In contrast, even though 12D7 CARTs could be expanded by HepG2 cell stimulation (Fig.6A-B), their cytokine production remained undetectable. However, after 6- day HepG2 stimulation, 12D7 CARTs did generate a low level of cytotoxicity (Fig.6D). Furthermore, the 8F8 and 6G11 CARTs showed significant killing on the GPC3 ^low Huh7 cells (Figs.6E-6G). Again, the 12D7 CARTs were unable to kill Huh7 tumor cells. Example VIII: Soluble hGPC3 does not activate CARTs, nor block CART’s activation by HCC tumor cells. Materials and Methods LDH assay and Cytokine ELISA assay HepG2 cells (5×103/well) were grown overnight on flat-bottom 96-well plates. CARTs were added at indicated effector to target cell ratios without addition of exogenous cytokines. After overnight co-culture, the cytotoxicity and effector cytokines of CARTs were measured by the lactate dehydrogenase (LDH) activity as indicated by the provider (Promega) and by ELISA (Biolegend). Results HCC patients normally have an elevated level of soluble GPC3 in their serum. Here, we investigated whether the CARTs would be activated by soluble GPC3 and whether the recognition and killing of HCC tumor cells could be blocked by the presence of soluble hGPC3. For this, we seeded 5x10 ³ HepG2 cells on flat-bottom 96-well plates. The next day, CARTs with different concentrations of hGPC3 were added. CARTs plus soluble hGPC3 without HepG2 cells were used detect activation of the CARTs by the soluble protein alone. Our data showed that soluble hGPC3 did not activate CARTs (Fig.7A), and that it did not affect the killing of HepG2 tumor cells by the CARTs (Fig.7B). The cytokine production of CARTs in co-culture with HepG2 cells was not inhibited by low concentration of hGPC3 (0.01-0.1µg/ml). However, a high concentration of soluble hGPC3 (1µg/ml) modestly enhanced the HepG2-induced IFNg production by CARTs (Fig.7C). Example IX: Adoptive transfer of CARTs eliminates HepG2 tumor xenografts in NSG mice. Materials and Methods Tumors and adoptive cell transfer HepG2 tumor xenografts were established as recently reported (Zhu, W., et al.). Two and half million (or 5 million) of HepG2 tumor cells were inoculated onto the shaved flank of NSG mice. Seven days later, CARTs with indicated numbers were transferred into the tumor-bearing NSG mice. Tumor growth was monitored by measuring the length, width, and height of the tumors. Tumor volume was calculated as 1/2(length × width × height). CARTs in mouse blood were monitored by collecting tail vein blood and performing immunological staining with anti- human CD45 and anti-mouse Fab antibodies. Co-staining of the mouse CD45 cells was used as internal reference to estimate the change of human CD45+ cells. Results To study the in vivo antitumor effects of CARTs, NSG mice bearing the HepG2 xenografts were adoptively transferred with CARTs. The tumor growth and the expansion and persistence of CARTs in NSG mouse blood were monitored (Fig.8A). In the first in vivo experiment, HCC tumors were established by subcutaneous inoculation of 2.5 million HepG2 cells. 7 days later, mice were treated with 3 million of CARTs. Our data showed that, compared to the non-transduced Mock-Ts, all CARTs generated significant antitumor effects (Fig.8B). Both 6G11 and 8F8 CARTs could eventually eradicate the tumors even though some tumors may experience transient growing (Fig.8B and 8F).12D7 CARTs delayed and inhibited tumor growth, but did not eradicate the tumors (Fig.8B and 8F). In the second experiment, we doubled the number of HepG2 cells to 5 million for tumor inoculation.7 days later, the tumor bearing mice were treated with the same dose of 3 million CARTs. We found that while 8F8 CARTs were still able to eradicate more than 90% of the tumors, 6G11 CARTs achieved complete eradication in only 20% of the mice (Fig.8G). Increase of 6G11 CARTs dose to 10 million resulted in complete tumor eradication in ~90% of treated mice (Table 2). Thus, adoptive transfer of 1.2-2 fold of 6G11CARTs (compared to the number of inoculated HepG2 cells) can result in complete eradication of HCC tumors in 75-90% of treated mice. The 8F8 CARTs are significantly more effective and result in complete eradication in more than 90% of the mice even when 0.6 fold of 8F8 CARTs are transferred. Table 2 summarizes the antitumor effects of 5 in vivo experiments. These in vivo antitumor effect data are in agreement with the in vitro data (Fig.6), which showed that 8F8 CARTs generated the strongest cytotoxicity and cytokine production, while 12D7 CARTs had the weakest effector function. Table 2 The expansion and persistence of CARTs in tumor bearing NSG mice were monitored (Fig.8C-E). The data showed that between days 11-25 after transfer, the human T cells (Fig.8D) and the percentage of the CAR+ T cells (Fig.8E) significantly increased in the groups treated with 6G11 and 8F8 CARTs. The CAR+ T cells in the groups treated with 6G11 and 8F8 CART cells increased from 10-15%, prior to the inoculation, to 30-60%, 2 weeks after adoptive transfer (Fig.8E). In contrast, the group treated with 12D7 CARTs showed no increase of adopted human T cells and the % of CAR+ T cells. The data suggests that 6G11 and 8F8 CARTs are activated and expanded when they meet HCC tumor cells in vivo, correlating to their effective antitumor effects and tumor regression. Example X: CART Production Overview 1. Total RNA was isolated from monoclonal hybridoma cells. The cDNA was synthesized and amplified by PCR using 5’ RACE technique. The PCR products were sequenced. The nucleotide sequences of the variable region of light chain (VL) and heavy chain (VH) of monoclonal antibodies (mAbs) 6G11, 8F8, and 12D7 were presented below. The amino acid sequences deduced from the cDNA were compared to IMGT databank. And the designated frame region (FR) and complementarity determining region (CDR) of each mAb were also presented. a. mAb 6G11-VL: b. mAb 6G11-VH c. mAb 8F8-VL

d. mAb 8F8-VH

e. mAb 12D7-VL f. mAb 12D7-VH 2. Next, we generated recombinant mAb by fusing the human IgG leader sequences, the VL or VH sequences of each mAbs (6G11, 8F8, and 12D7), and the constant (C) region of human IgG. The amino acid sequences of each recombinant H and L chain were presented below. a. Recombinant mAb 6G11 amino acid sequences b. Recombinant mAb 8F8 amino acid sequences c. Recombinant mAb 12D7 amino acid sequences

3. We next built the second generation of CAR from the mAb 6G11, 8F8, and 12D7 VL and VH sequences. Both the DNA and protein sequences of the 3 CARs were presented below. a. 6G11-CAR (SEQ ID NO:1)

b. 8F8-CAR (SEQ ID NO:2)

c. 12D7-CAR sequences (SEQ ID NO:3)

Discussion The goal of the study was to develop novel effective CARTs for HCC immunotherapy by targeting different hGPC3 epitopes. To this end, 3 mAbs (6G11, 8F8, and 12D7) were generated that recognize the N-, C-, and a conformational epitope of hGPC3, respectively. These mAbs are highly specific and able to distinguish HCC tumors from the adjacent normal liver tissues. Then, CARTs were built from the mAbs, and it was found that all 3 CARTs could undergo expansion in response to HCC tumor cell stimulation. But, their effector functions were markedly different. The 8F8 CARTs generated the strongest cytotoxicity and the highest production of IFNg after HCC tumor cell stimulation. Compared to 8F8 CARTs, the 6G11 CARTs underwent an even greater in vitro expansion, but with a slight weaker cytotoxicity and significant lower IFNg and IL2 production. In contrast, the 12D7 CARTs showed the weakest cytotoxicity and produced the least amount of IFNg and no IL2. Adoptive transfer of 6G11 and 8F8, but not 12D7, CARTs generated potent antitumor effects and resulted in complete eradication of HCC tumor xenografts in NSG mice, which correlated to their in vivo expansion. Several factors may contribute to the different effector functions of the CARTs. The first reason could be the epitope location in the target antigen. Previous reports showed that the position of the epitope within the target molecule determined the efficiency of CART activation (38, 39). The membrane-proximal epitopes may facilitate efficient immunological synapse between targets and CARTs (40). This explanation fits well with 8F8 CARTs, which bind the membrane-proximal C- epitope (AA463-496) of hGPC3 and possess the strongest cytotoxicity and produces the highest amount of cytokines. However, the 6G11 CARTs recognize an epitope at the beginning of the N-terminus (AA25-39) that is conceivably membrane-distal. Nevertheless, it undergoes stronger expansion after HCC tumor stimulation and possesses potent effector function, albeit it is slightly weaker than 8F8 CARTs. These data suggest that the engagement of CARTs with hGPC3 N-terminal epitopes is also sufficient to initiate CART activation and effector function. One reason that may explain why 6G11 CARTs, targeting the membrane-distal epitope also generate potent effector function and antitumor effect, is its accessibility to target antigen. In fact, our data showed that 90% of 6G11 CAR+ T cells could bind hGPC3, while only about 40% of 8F8 CAR+ and 12D7 CAR+ T cells were able to bind hGPC3, suggesting that 6G11 CAR on T cells and 6G11’ epitope on hGPC3 are more accessible to each other (Figs.5D-F). Thus, more CAR and hGPC3 engagement might generate stronger signaling to activate CARTs even though the target epitope may not be in membrane-proximity. Another factor that may affect CART’s function is the mAb’s affinity (41). According to the data from two measurements (BLI and ELISA), the affinity order of the 3 mAbs is 6G11 > 12D7 > 8F8. The high affinity of 6G11 for hGPC3 may also help generate stronger interaction of CARTs and tumor cells even if the epitope is membrane-distal. Thirdly, hGPC3 N-fragment is linked to the C-fragment by disulfide bonds (25, 42, 43), which may bring the N-terminal epitope closer to cell membrane than expected to generate stronger interaction between CARTs and tumor cells. However, as recent studies (44, 45) demonstrated that strong T cell signaling may lead to premature exhaustion, it will be interesting to determine the signaling strength of these CARTs and its relevancy to the epitope position and effector functions. Different from 6G11 and 8F8 that recognize specific region of hGPC3, the 12D7 CARTs recognize a conformational epitope on the hGPC3 N-fragment and have the lowest effector function. These data are in agreement with a recent report showing that the HN3 CARTs targeting a conformational epitope of GPC3 did not generate antitumor efficacy even though no detailed analysis was provided in the meeting abstract (46). In the study, the 12D7 CARTs underwent expansion as good as 8F8 CARTs in response to HepG2 cell stimulation (Fig.6A-B), but they did not produce effector cytokines, and had weak cytotoxicity (Fig.6C-D). It is likely that the engagement of 12D7 CARTs with the hGPC3 conformational epitope generates sufficient signaling to drive CART proliferation, but not enough to fully activate CARTs to have effector functions. The lack of cytokine production, especially IL-2, may explain the incapability of 12D7 CARTs to expand in vivo (Fig.8E and 8F), which correlates to their weaker in vivo antitumor effects. Even though the 8F8 mAb has the lowest affinity for hGPC3, the 8F8 CARTs generated the strongest cytotoxicity and produced the highest amount of cytokines after HepG2 tumor cell stimulation. On the other hand, the 6G11 and 12D7 mAbs have a similar affinity, but their effector functions are dramatically different (Fig.6). Furthermore, soluble hGPC3 does not activate CARTs nor block the CART activation by tumor cells (Fig.7). Together, these data suggest that several parameter might affect CART activation, and that immunological synapse may play a dominant role in the CART’s effector function (40, 47). Most of the current and previous studies are mainly focused on the signaling transduction domains, including the 1 ^st to 3 ^rd generations of CARs, to improve CAR-T’s efficacy (48). The role of antibody domains in the generation of potent CARTs has been less investigated. The 3 new CARTs, targeting different regions of hGPC3, and having different affinities, not only offer more arsenal for HCC immunotherapy, but also provide us materials and opportunities to extend the study of the mechanism of CARTs and target tumor cells interaction. Future studies are required to analyze the subsets of each CART and to investigate the immunological synapse and the signaling strength they generate. These different CARTs properties may help us reveal the mechanisms behind the relation between of CART expansion, activation and effector functions, and enable us to further improve their antitumor efficacy. While in the foregoing specification this invention has been described in relation to certain embodiments thereof, and many details have been put forth for the purpose of illustration, it will be apparent to those skilled in the art that the invention is susceptible to additional embodiments and that certain of the details described herein can be varied considerably without departing from the basic principles of the invention. All references cited herein are incorporated by reference in their entirety. The present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof and, accordingly, reference should be made to the appended claims, rather than to the foregoing specification, as indicating the scope of the invention.