CD147 Chimeric Antigen Receptor and Methods of Use

Abstract

An engineered single-chain variable fragment (scFv) that specifically binds to CD147 is provided. Chimeric antigen receptors (CARs) comprising the engineered CD147scFvs, nucleic acids encoding the CARs, vectors comprising nucleic acids encoding the CARs, and immune cells expressing the CARs are also provided. A method for treating a subject with cancer is also provided, comprising administering to the subject immune cells expressing the disclosed CD147-CARs. The present disclosure relates to immunotherapy, particularly chimeric antigen receptors that target CD147 and their use for treating cancer.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 62 / 819,403, filed March 15, 2019, which is incorporated herein by reference in its entirety.

[0002] Field The present disclosure relates to immunotherapy, in particular to chimeric antigen receptors that target CD147 and their use to treat cancer.

[0003] Acknowledgments of Government Support This invention was made with government support under Grant Nos. AI130197, HL125018, AI124769-01, and AI129594 awarded by the National Institutes of Health. The government has certain rights in this invention. [Background technology]

[0004] Liver cancer is the second leading cause of cancer-related death worldwide. The burden of liver cancer is predicted to exceed one million cases by 2030. Liver cancer is the fifth leading cause of cases and the second leading cause of death among men worldwide. Each year, more than 500,000 patients die from hepatocellular carcinoma (HCC).

[0005] Primary liver cancers include hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (iCCA), fibrolamellar carcinoma, and hepatoblastoma. HCC and iCCA are the most common primary liver cancers, accounting for more than 99% of primary liver cancer cases. HCC alone (approximately 800,000 new cases per year) accounts for 90% of all primary liver cancer cases. Currently, no effective therapy is available for treating HCC. Sorafenib (CheckMate-040, a multikinase inhibitor widely used in patients with advanced HCC, but with low efficacy and serious side effects) is the standard first-line systemic agent for HCC. Currently, the PD-1 blocker Opdivo (nivolumab) is approved by the U.S. Food and Drug Administration (FDA) as a second-line treatment strategy for patients with HCC previously treated with sorafenib. Clinical trials testing PD-1 blockers as first-line treatment for HCC are currently ongoing. Meanwhile, various clinical trials using PD-1 or PD-L1 blockade in combination with other interventions are also underway. For example, a study evaluating an anti-PD-1 antibody in combination with an anti-CTLA-4 antibody in patients with resectable and likely resectable HCC is being tested in a clinical trial (NCT03222076).

[0006] Chimeric antigen receptor (CAR)-modified T cell therapy has become a promising immunotherapeutic strategy for the treatment of various hematological cancers. Despite recent advances in CAR-modified T cell immunotherapy in hematological cancers, high costs and severe toxicity have prevented its widespread use. Meanwhile, CAR-T cells face additional challenges during targeting of solid tumors, such as maintaining persistent proliferation and persistence in the tumor microenvironment. Another challenge for CAR-mediated immunotherapy for liver cancer is discovering effective targets. Summary of the Invention [Means for solving the problem]

[0007] CD147 is expressed at various levels on various cell types (e.g., hematopoietic, epithelial, and endothelial cells). However, CD147 is significantly upregulated in pathological conditions such as HCC, breast cancer, bladder cancer, colorectal cancer, ovarian cancer, melanoma, and osteosarcoma. CARs that specifically target cells expressing CD147 are provided. These CARs can be used in immunotherapy of cancers that express or overexpress CD147.

[0008] Disclosed herein are modified single-chain variable fragments (scFvs) that specifically bind to CD147. In some embodiments, the scFvs have an amino acid sequence comprising the variable heavy (VH) domain complementarity-determining region 1 (CDR1), CDR2, and CDR3 amino acid sequence of SEQ ID NO: 8, and the variable light (VL) domain CDR1, CDR2, and CDR3 amino acid sequence of SEQ ID NO: 9. In some examples, the scFvs have at least 90% sequence identity with, or comprise or consist of, the amino acid sequence of SEQ ID NO: 2. Nucleic acids encoding the modified CD147scFvs, for example, nucleic acids having at least 90% sequence identity with, or comprising or consisting of, the nucleic acid sequence of SEQ ID NO: 1, and vectors comprising the nucleic acid sequences, are also provided. In additional embodiments, vectors encoding the modified CD147scFvs (e.g., SEQ ID NO: 1) are provided, wherein the vector further comprises an inducible promoter or enhancer nucleic acid molecule operably linked to the CD147scFv nucleic acid molecule. In some examples, the enhancer nucleic acid is a Gal4 upstream activating sequence (UAS) operably linked to the nucleic acid encoding CD147scFv. In other examples, the vector is a synNotch construct, e.g., a vector comprising the nucleic acid sequence of a modified CD147scFv nucleic acid molecule (e.g., SEQ ID NO: 1) linked to synNotch and a Gal4-VP64 coding sequence (e.g., SEQ ID NO: 17).

[0009] Also provided is a CAR comprising a modified CD147scFv provided herein, an intracellular domain comprising a hinge domain, a transmembrane domain, one or more costimulatory molecule intracellular domains, and an intracellular signaling domain. In one embodiment, the CD147-CAR comprises a modified CD147scFv provided herein, an IgG1 hinge domain, a CD28 transmembrane domain, a CD28 and 4-1BB costimulatory domain, and a CD3ζ signaling domain. In some examples, the CD147-CAR comprises an amino acid sequence having at least 90% identity with the amino acid sequence of SEQ ID NO: 5, or comprises or consists of the amino acid sequence of SEQ ID NO: 5.

[0010] In some embodiments, the CD147-CAR further comprises an inducible suicide molecule, such as caspase 9. In some examples, expression of the suicide molecule is induced by tetracycline, doxycycline, or rapamycin. In one example, the CD147-CAR with an inducible suicide gene comprises an amino acid sequence having at least 90% identity to, or comprises or consists of, the amino acid sequence of SEQ ID NO: 7. The CD147-CAR may further comprise a cytokine receptor intracellular domain, for example, an interleukin-15 receptor intracellular domain (e.g., SEQ ID NO: 12), an interleukin-12 receptor intracellular domain, or an interleukin-18 receptor intracellular domain.

[0011] Also provided are nucleic acids encoding the CD147-CAR disclosed herein, and vectors (e.g., viral vectors) comprising the nucleic acids. In some examples, the CD147-CAR is encoded by a nucleic acid sequence having at least 90% identity to the nucleic acid sequence of SEQ ID NO: 4 or SEQ ID NO: 6. In other examples, the CD147-CAR is encoded by a nucleic acid comprising or consisting of the nucleic acid sequence of SEQ ID NO: 4 or SEQ ID NO: 6.

[0012] In additional embodiments, vectors are provided that encode CD147-CAR (e.g., SEQ ID NO: 4 or SEQ ID NO: 6) and further comprise an inducible promoter or enhancer nucleic acid molecule operably linked to the CD147-CAR nucleic acid molecule. In some examples, the enhancer nucleic acid is a Gal4 upstream activating sequence (UAS) (e.g., SEQ ID NO: 14) operably linked to the nucleic acid encoding CD147-CAR. In one example, the vector comprises a vector comprising the nucleic acid sequence of CD147-CAR in a synNotch construct, e.g., SEQ ID NO: 15. In other examples, the CD147-CAR nucleic acid molecule (e.g., SEQ ID NO: 1) is linked to synNotch and Gal4-VP64 coding sequences (e.g., SEQ ID NO: 17).

[0013] T cells, natural killer (NK) cells, natural killer T (NKT) cells, double-negative T (DNT) cells (CD3 + CD4 - CD8 - Also provided are T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages, such as T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages, comprising a nucleic acid encoding a disclosed CD147scFv or CD147-CAR or a vector encoding a disclosed CD147scFv or CD147-CAR. In some examples, the NK cells are NK-92 or NK-92MI cells. Methods of producing cells expressing a CAR, including, but not limited to, CD147-CAR-NK cells, CD147-CAR-T cells, or CD147-CAR-macrophages, are provided. These methods include transducing or transfecting T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages with a vector encoding a disclosed CAR.

[0014] In further embodiments, T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages are provided that express CD147-CAR operably linked to an activator of an inducible promoter or enhancer element. In some examples, the T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages further express a nucleic acid encoding an anti-GPC3 specific binding agent (e.g., an anti-GPC3 scFv) operably linked to an activator of an inducible promoter or enhancer element. In other embodiments, T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages are provided that express CD147 scFv operably linked to an inducible promoter or enhancer element, and further comprise a nucleic acid molecule encoding an anti-GPC3 chimeric antigen receptor operably linked to an inducible promoter or enhancer element.

[0015] Disclosed herein are methods for treating a subject with cancer, for example, by administering to the subject NK cells, T cells, NKT cells, DNT cells, neutrophils, or macrophages (e.g., CD147-CAR NK cells, CD147-CAR-T cells, or CD147-CAR macrophages) expressing a CAR disclosed herein. In some examples, the subject has a cancer that expresses CD147. In certain non-limiting examples, the subject has hepatocellular carcinoma, neuroblastoma, breast cancer, pancreatic cancer, leukemia, lymphoma, multiple myeloma, colorectal cancer, lung cancer, melanoma, renal cell carcinoma, sarcoma, or nasopharyngeal carcinoma.

[0016] These and other features of the present disclosure will become more apparent from the following detailed description, which proceeds with reference to the accompanying drawings. [Brief explanation of the drawings]

[0017] [Figure 1-1]Figures 1A-1E show the overexpression of CD147 in hepatocellular carcinoma cells. Figure 1A shows the prognostic value of upregulated expression of CD147 for overall survival in human cancer patients from the TCGA dataset. Survival curves (upper panel) of various patient populations based on relative high and low CD147 expression (lower panel). Data from LUAD (lung adenocarcinoma), SKCM (cutaneous melanoma), LIHC (liver hepatocellular carcinoma), and LGG (brain low-grade glioma) were collected for analysis. Figure 1B shows a comparison of CD147 expression between normal tissue (NT) and tumor samples (TP) in multiple cancer types from the TCGA dataset. Data represent the mean ± SEM of three separate experiments. Unpaired Student's t-test was used. *p<0.05, **p<0.01, ***p<0.001, and ns (not significant). According to the TCGA database, the full names of each cancer type are BRCA (breast invasive carcinoma), CHOL (cholangiocarcinoma), GBM (glioblastoma multiforme), LGG (low-grade glioma of the brain), HNSC (head and neck squamous cell carcinoma), KICH (nephrochromophobe), KIPAN pan-renal cohort (KICH + KIRC + KIRP), KIRP (renal papillary cell carcinoma), LIHC (hepatocellular carcinoma), LUAD (lung adenocarcinoma), LUSC (lung squamous cell carcinoma), PRAD (prostate adenocarcinoma), and UCEC (endometrial carcinoma). Figure 1C shows Western blot analysis of CD147 in HCC cell lines. 1 × 10 cells from various cell lines were lysed in 200 μl of RIPA buffer, mixed with 50 μl of 5×SDS loading buffer, and then independently loaded onto SDS-PAGE. Mouse anti-human CD147 (HIM6, mouse IgG1) was used for Western blot analysis. Anti-GAPDH was used as a loading control. Figure 1D shows CD147 in HCC cell lines (SK-Hep1 and HepG2). SK-Hep1 and HepG2 (1 × 106 cells) were stained with 2 μg of FITC-mouse anti-human CD147 (anti-CD147) or 2 μg of FITC-isotype mouse IgG1 (isotype, kappa). After incubation and washing, samples were analyzed by flow cytometry. Numbers represent the mean fluorescence intensity (MFI) of each sample.Figure 1E is a series of panels showing histopathological analysis of CD147 antigen expression on human HCC tumors isolated from PDX mouse models. Representative H&E (top row) and IHC staining of CD147 (middle row) of tumor samples from various patient-derived xenograft (PDX) mice treated with PBS, NK-92MI, and CD147-CAR NK-92MI, respectively. The bottom row shows IHC staining without primary antibody. Scale bar represents 50 μm. Data are representative of three independent experiments. [Figure 1-2] Same as above. [Figure 1-3] Same as above. [Figure 2-1]Figures 2A-2F show the design of the CD147-CAR and the phenotyping of CAR-modified NK-92MI cells. Figure 2A shows the schematic design of the CD147-specific CAR based on the SFG retroviral vector. The construct contains a CD147-specific single-chain antibody fragment (a modified scFv, mIgG1, from clone 5F6), a human IgG1 CH2CH3 hinge region, and a CD28 transmembrane region, followed by the costimulatory CD28, the intracellular domain of 4-1BB, and the intracellular domain of CD3ζ. Figure 2B shows flow cytometry analysis of CAR expression and CD56 expression on the surface of parental NK-92MI and CD147-CAR-NK-92MI cells. Data are representative of at least three experiments. Figure 2C shows Western blot analysis of CAR expression in parental NK-92MI and CD147-CAR-NK-92MI cells using an anti-human CD3ζ-specific antibody for detection of endogenous CD147 and CD147-CAR fusion protein. Figure 2D shows NK activation and inhibitory markers in parental NK-92MI, CAR-CD19(4-1BB)-NK-92MI, and CD147-CAR-NK-92MI cells. Each data point represents at least three or four experiments. The numbers in the flow graphs represent the mean fluorescence intensity (MFI) of each sample. Figure 2E shows flow cytometry analysis of CAR expression in CD19-CAR-NK-92MI and CD147-CAR-NK-92MI using goat anti-human IgG (H+L). Wild-type NK-92MI cells were used as a control. Figure 2F shows flow cytometry analysis of CD147 expression in NK-92MI, CD19-CAR-NK-92MI, and CD147-CAR-NK-92MI. Figure 2G shows overlaid flow cytometry profiles of CD147 expression levels in NK-92MI, CD19-CAR-NK-92MI, and CD147-CAR-NK-92MI. Data are representative of two independent experiments. [Figure 2-2] Same as above. [Figure 2-3] Same as above. [Figure 3-1]Figures 3A-3D show that CD107a degranulation and cytokine production in CD147-CAR-NK-92MI cells are stimulated by their susceptible target cells. Figure 3A shows representative flow cytometry data illustrating CD107a degranulation in NK-92MI, CD19-4-1BB-CAR, CD19-CD28-CAR, and CD147-CAR after 10 hours using medium (control), SK-Hep1, and HepG2. The effector-to-target ratio was 1:1.2. To quantify surface CD107a expression, cells were gated on the CD56-positive subset. Figure 3B shows quantitative data for the percentage of surface CD107a expression in CD147-CAR-NK-92MI cells upon various stimuli, as indicated. Production of cytokines TNF-alpha (Figure 3C) and IFN-gamma (Figure 3D) by CD147-CAR-NK-92MI, CD19-4-1BB-CAR-NK-92MI, CD19-CD28-CAR-NK-92MI, and wild-type NK-92MI stimulated by various conditions. NK-92MI cells were cocultured with SK-Hep1 cells at an effector / target ratio of 1:1 or in medium for 12 hours. Phorbol-12-myristate-13-acetate (PMA) / ionomycin (IONO) was used as a positive control. Cytokine release ratios were calculated using the following equation: sample value / mean positive control value × 100 (%). Data were pooled from at least three or four experiments. [Figure 3-2] Same as above. [Figure 4]Figures 4A-4C show the killing of two HCC cell lines by CD147-CAR-NK-92MI. Figure 4A shows representative flow cytometry staining of surface CD147 molecules on Huh7 (left) and HCO2 (right) cell lines. Figure 4B shows a graph depicting the cytotoxicity of CD147-CAR-NK-92MI as measured by a standard 4-hour 51Cr release assay. CD147-positive Huh7 cells were used as target cells sensitive to CD147-CAR-NK-92MI. Wild-type NK-92MI was used as a control. Figure 4C shows the cytotoxicity of CD147-CAR-NK-92MI as measured by a standard 4-hour 51Cr release assay. CD147-positive HCO2 cells were used as target cells sensitive to CD147-CAR-NK-92MI. Wild-type NK-92MI was used as a control. Data represent the mean ± SEM from three independent experiments. *p<0.05, **p<0.01, and ***p<0.001. [Figure 5-1]Figures 5A-5G show the activation of CD147-CAR-NK-92MI cells upon stimulation of CD147-positive target cells. Figure 5A shows representative data showing the percentage of surface CD107a expression on CD147-CAR-NK-92MI cells upon various stimuli, as indicated. CD147-CAR-NK-92MI cells were stimulated with SK-Hep1 or HepG2 cells for 4 hours. To block the interaction between CD147-CAR and CD147 molecules, 5 μg of mouse anti-human CD147 (HIM6) was added to the effector and target cell mixture. As indicated, 5 μg of isotype-specific mouse IgG (IgG) or PBS (vehicle control group) was used as a control. Figures 5B and 5C are graphs showing quantitative data on the percentage of surface CD107a staining on CD147-CAR-NK-92MI cells stimulated with CD147-positive SK-Hep1 (Figure 5B) and CD147-positive HepG2 (Figure 5C) cell lines. Wild-type NK-92MI cells alone and CD147-CAR-NK-92MI cells alone were used as controls, as indicated. Figure 5D shows representative data showing the percentage of surface CD107a expression on CD147-CAR-NK-92MI cells stimulated with CD147-positive wild-type (WT) SK-Hep1 cell line (top panel) and CD147 knockout (CD147- / -) SK-Hep1 cell line (middle panel). Culture medium alone was used as a control. Naive NK-92MI and 4-1BB-CD19-CAR (CD19-CAR) were used as control effector cells. Figure 5E shows quantitative data for the percentage of surface CD107a staining on CD147-CAR-NK-92MI cells stimulated with CD147-positive SK-Hep1 (WT) and CD147 knockout SK-Hep1 (CD147- / -) cell lines, respectively. Figure 5F shows representative data showing the percentage of surface CD107a expression on CD147-CAR-NK-92MI cells stimulated with CD147-positive wild-type (WT) HepG2 cell lines (upper panel) and CD147 knockout HepG2 cell lines (middle panel). A culture medium only group was used as a control.Naive and CD19-4-1BB-CAR-NK-92MI (CD19-CAR) cells were used as control effector cells. Figure 5G shows quantitative data for the percentage of surface CD107a staining on CD147-CAR-NK-92MI cells stimulated with CD147-positive HepG2 (WT) and CD147 knockout HepG2 (CD147- / -) cell lines, respectively. The effector-to-target ratio was 1:1.2. CD147-CAR-NK-92MI cells were gated by CD56 antibody surface staining. NK degranulation was measured by CD107a surface staining by flow cytometry. Data represent the mean ± SEM of three separate experiments. *p<0.05, **p<0.01, and ***p<0.001. [Figure 5-2] Same as above. [Figure 6] Figure 6 shows that mouse anti-human CD147 (HIM6) does not affect the cytotoxicity of CD19-CAR-NK cells. The cytotoxicity of CD19-CAR-NK-92MI was measured using the FFLuc Report System assay. Briefly, Daudi-FFLuc cells (1 × 104) were pre-seeded overnight in Matrigel (BD)-treated 96-well optical-bottom microplates. Effector cells (CD19-CAR-NK-92MI) were co-cultured for 6 hours at two different effector / target ratios (5:1 and 1:1, as indicated). Luminescence signals were quantified using a microplate reader, and the percentage of specific lysis was calculated. Data were pooled from three independent experiments. Error bars indicate ±SEM (standard error of the mean). *p<0.05, **p<0.01, and ***p<0.001. [Figure 7]Figures 7A and 7B show representative images of the killing activity of CD147-CAR-NK-92MI. Effector cells (1 × 104) CD147-CAR-NK-92MI and NK-92MI were cocultured with target cells FFLuc-GFP-SK-Hep1 (Figure 7A, top) (1 × 104) and FFLuc-GFP-HepG2 (Figure 7A, bottom) in a 96-well optical-bottom microplate for 12 hours. To visualize the killing activity of CD147-CAR-NK-92MI under the same settings, GFP fluorescence (upper lane) and bright field (lower lane) detected by conventional fluorescence microscopy were used. GFP fluorescence intensity was quantified using ImageJ software (NIH) (Figure 7B). Quantitative mean fluorescence intensity (MFI) of GFP was plotted using Graph Prism 5 software (GraphPad Software, San Diego, CA, USA). [Figure 8-1]Figures 8A-8N show that CD147-CAR-T and NK cells specifically kill CD147-positive tumor cells in vitro. Figure 8A shows the cytotoxicity of primary CD147-CAR-T cells as measured by the FFLuc reporter assay. CD147-positive FFLuc-GFP-SK-Hep1 cells were used as target cells sensitive to CD147-CAR-T. Kappa-CAR T cells were used as a control group in each experiment. Figures 8B and 8C show significantly reduced cytotoxicity of CD147-CAR-T cells using knockout CD147FFLuc-GFP-SK-Hep1 and HepG2 cell lines as measured by the FFLuc reporter assay. Data represent the mean ± SEM from three independent experiments. Figure 8D shows the cytotoxicity of primary CD147-CAR-NK cells as measured by a standard 51Cr release assay over 4 hours. CD147-positive FFLuc-GFP-SK-Hep1 cells were used as target cells sensitive to CD147-CAR-T. Kappa-CAR T cells were used as a control group in each experiment. Figures 8E and 8F show a significant reduction in the cytotoxicity of primary CD147-CAR-NK cells using knockout CD147FFLuc-GFP-SK-Hep1 and HepG2 cell lines in a standard 51Cr release assay for 4 hours. Data represent the mean ± SEM from three independent experiments. Figure 8G shows that anti-NKG2D antibodies blocked the spontaneous killing of FFLuc-GFP-SK-Hep1 by primary CD147-CAR-NK. Various ratios of primary CD147-CAR-NK cells were co-cultured with FFLuc-GFP-SK-Hep1, knockout CD147FFLuc-GFP-SK-Hep1 cell line, or knockout CD147FFLuc-GFP-SK-Hep1 cell line containing 5 μg of anti-NKG2D for 4 hours. An FFLuc reporter assay was used. Data represent the mean ± SEM from three independent experiments. Figure 8H shows that the cytotoxicity of CD147-CAR-NK-92MI against SK-Hep1 was measured by a standard 4-hour 51Cr release assay.Effector cells (CD147-CAR-NK-92MI and NK-92MI) were co-cultured with 1 × 10 target cells (FFLuc-GFP-SK-Hep1) per well. After 4 hours, the supernatant was collected and the released 51Cr was measured using a gamma counter. Figures 8I and 8J show FFLuc reporter system assays of specific killing of FFLuc-GFP-SK-Hep1 and FFLuc-GFP-HepG2 cell lines by CD147-CAR-NK-92MI. Effector cells (CD147-CAR-NK-92MI and NK-92MI) were co-cultured with 1 × 10 target cells (FFLuc-GFP-SK-Hep1) (Figure 8I) or FFLuc-GFP-HepG2 (Figure 8J) per well in a 96-well optical-bottom microplate for 6 hours. After 5 minutes of incubation with D-luciferin, the luminescence signal was measured using a microplate reader to calculate the NK cell cytotoxicity. The control group used was wild-type NK-92MI cells incubated with CD147-positive FFLuc-GFP-SK-Hep1 or CD147-positive FFLuc-GFP-HepG2. Figures 8K and 8L show the reduced cytotoxicity of CD147-CAR-NK-92MI cells using knockout CD147FFLuc-GFP-SK-Hep1 (Figure 8K) and knockout CD147FFLuc-GFP-HepG2 (Figure 8L) cell lines in an FFLuc reporter system assay. Effector cells (CD147-CAR-NK-92MI and NK-92MI) were co-cultured with 1 x 10 wild-type or CD147 knockout target cells per well in a 96-well optical-bottom microplate for 6 hours. After 5 minutes of incubation with D-luciferin, the luminescence signal was measured using a microplate reader to calculate the cytotoxicity of NK cells. Figures 8M and 8N are graphs showing that anti-CD147 (clone, HIM6) inhibited the specific lytic effect of CD147-CAR-NK-92MI against FFLuc-GFP-SK-Hep1 (Figure 8M) and FFLuc-GFP-HepG2 (Figure 8N).Effector cells (CD147-CAR-NK-92MI and NK-92MI) were co-cultured with 1 x 104 FFLuc-GFP-SK-Hep1 or FFLuc-GFP-HepG2 target cells per well in a 96-well optical-bottom microplate for 6 hours. After 5 minutes of incubation with D-luciferin, luminescence signals were measured using a microplate reader to calculate NK cell cytotoxicity. Data represent the mean ± SEM from three independent experiments. *p<0.05, **p<0.01, and ***p<0.001. [Figure 8-2] Same as above. [Figure 9] Figures 9A and 9B show flow cytometry and Western blot analysis of knockout CD147 SK-Hep1 and HepG2 cell lines. Figure 9A shows staining of surface CD147 molecules on wild-type (wt) SK-Hep1 and CD147- / -- SK-Hep1 cell lines (top), and wild-type (wt) HepG2 and CD147- / -- HepG2 cell lines (bottom). Figure 9B shows Western blot analysis of CD147 molecules on wild-type (wt) SK-Hep1 and CD147- / -- SK-Hep1 cell lines, and wild-type (wt) HepG2 and CD147- / -- HepG2 cell lines. GAPDH was used as a loading control (bottom). [Figure 10]Figures 10A-10C show that CD147-CAR-NK-T cells specifically kill CD147-positive tumor cells. The cytotoxicity of CD147-CAR-T cells was measured by FFLuc report system assay. CD147-positive FFLuc-EGFP-Hep-G2 (Figure 10A) and CD147-positive FFLuc-EGFP-SK-Hep1 (Figure 10B) were used as target cells sensitive to CD147-CAR-T. Kappa-CAR T cells were used as the control group in each experiment. Figure 10C shows significantly reduced cytotoxicity of CD147-CAR-T cells using the knockout CD147 FFLuc-GFP-SK-Hep1 cell line by FFLuc report system assay. Briefly, effector cells (CD147-CAR-T cells) were co-cultured with target cells FFLuc-EGFP-SK-Hep1 or CD147 knockout FFLuc-EGFP-SK-Hep1 (1 x 104) in a 96-well optical bottom microplate for 6 hours. The cytotoxicity of CD147-CAR-T cells was measured by reading the luminescence signal using a microplate reader. Data represent the mean ± SEM from three independent experiments. *p<0.05, **p<0.01, and ***p<0.001. [Figure 11]Figures 11A-11D show that CD147-CAR-T-92MI cells control HCC progression in a xenograft mouse model. Figure 11A is a diagram of the experimental design of the HCC xenograft model. Briefly, NSG mice were subcutaneously injected with 4 x 10 SK-Hep1 cells premixed with an equal volume of Matrigel (day 0). Mice were monitored for tumor burden (reaching approximately 50 mm ) and randomly divided into groups on day 4. On day 5 (D5), mice were injected (iv) with a dose of 1 x 10 effector CD147-CAR-T (group #1) cells along with 2 x 10 IU of IL-2. The control group was injected with vehicle (PBS) control only (group #2). Each group received the same treatment on days 7, 9, and 16. Figure 11B shows quantification of tumor burden in SK-Hep1 xenografts treated with CD147-CAR-T and PBS (vehicle control group), respectively. All results are mean ± SEM. Differences for each group were analyzed by two-way ANOVA analysis. Figure 11C shows quantitative body weights for each group assessed at the indicated time points. Figure 11D shows Kaplan-Meier survival curves of tumor-bearing mice after treatment with CD147-CAR-T cells and PBS (vehicle control group). p-values ​​were analyzed by log-rank (Mantel-Cox) test. [Figure 12]Figures 12A-12D show the antitumor efficacy of CD147-CAR-NK-92-MI cells against HCC in a mouse xenograft model. Figure 12A is a diagram of the experimental design for the antitumor efficacy of primary CD147-CAR-NK in an HCC xenograft model. Five days after tumor implantation (day 5), mice were injected (iv) with a dose of 1 x 10 effector primary CD147-CAR-NK cells along with 2 x 10 IU of IL-2. Control groups were injected with the same number of untransduced primary NK cells along with 2 x 10 IU of IL-2 (group #2) or PBS alone (group #3). Identical treatments were administered in each group on days 5, 7, 9, 16, and 18, as indicated. Figure 12B shows the quantitative tumor burden of HCC xenograft mice treated with primary CD147-CAR-NK, non-transduced primary NK, and PBS (vehicle control group), respectively. All results are mean ± SEM. Differences between each group were analyzed by two-way ANOVA analysis. Figure 12C shows the quantitative body weight of each group assessed at the indicated time points. Figure 12D shows Kaplan-Meier survival curves of tumor-bearing mice after treatment with primary CD147-CAR-NK, parental primary NK, and PBS (vehicle control group). p-value analysis by log-rank (Mantel-Cox) test. Differences between each group were analyzed by two-way ANOVA analysis. *p<0.05, **p<0.01, and ***p<0.001 indicate the comparison of the CD147-CAR-modified cell-treated group with the control group. [Figure 13]Figure 13 shows comparable anti-HCC tumor activity between irradiated CD147-CAR-NK-92MI and non-irradiated CD147-CAR-NK-92MI cells in killing HCC cell lines in vitro. The cytotoxicity of irradiated and non-irradiated CD147-CAR-NK-92MI was measured by a standard 4-hour 51Cr release assay. CD147-positive wild-type HepG2 tumor cells (experimental group, left panel) or CD147 knockout (CD147KO, right panel) HepG2 tumor cell line were used as target cells sensitive to CD147-CAR-NK-92MI. Irradiated and non-irradiated wild-type NK-92MI cells were used as effector cell controls. Data are representative of three independent experiments. All data are presented as mean ± SEM. [Figure 14]Figures 14A-14D show comparable anti-HCC tumor activity between irradiated CD147-CAR-NK-92MI and non-irradiated CD147-CAR-NK-92MI cells in controlling HCC progression in a xenograft mouse model. Figure 14A is a diagram of the experimental design of the HCC xenograft model. NSG mice were injected (sc) with 2 x 10 SK-Hep1 cells premixed with an equal volume of Matrigel (day 0). One day before treatment (day 4), tumor burden was determined (reaching approximately 50 mm ), and mice were randomly grouped. On day 5 (D5), mice were injected (iv) with a dose of 1 x 10 effector non-irradiated CD147-CAR-NK-92MI (group #1) cells along with 2 x 10 IU of IL-2. Control groups were injected with the same number of irradiated CD147-CAR-NK-92MI in PBS along with 2×10 IU of IL-2 (Group #2) or vehicle control only (Group #3). Each group received the same treatment on days 7, 9, 16, and 18. Figure 14B shows quantification of tumor burden in SK-Hep1 xenografts treated with CD147-CAR-NK-92MI and PBS (vehicle control group), respectively. All results are mean ± SEM. Differences between groups were analyzed by two-way ANOVA analysis. Figure 14C is a graph of the quantitative body weights of each group assessed at the indicated time points. Figure 14D shows Kaplan-Meier survival curves of tumor-bearing mice after treatment with CD147-CAR-NK-92MI cells and PBS (vehicle control group). p-values ​​were analyzed by log-rank (Mantel-Cox) test. Data are representative of two independent experiments. All data are presented as mean ± SEM. [Figure 15]Figures 15A-15D show that CD147-CAR-NK-92MI cells control HCC progression in a xenograft mouse model. Figure 15A is a diagram of the experimental design of the HCC xenograft model. Briefly, NSG mice were subcutaneously injected with 4 x 10 SK-Hep1 cells premixed with an equal volume of Matrigel (day -7). On day 0 (the day before day 1), tumor burden was determined (approximately 50 mm ), and mice were randomly divided into groups. On day 1 (D1), mice were injected (iv) with one dose of 1 x 10 effector CD147-CAR-NK-92MI (group #1) cells along with 2 x 10 IU of IL-2. Control groups were injected with the same number of NK-92MI cells in PBS along with 2 x 10 IU of IL-2 (group #2) or vehicle control only (group #3). Identical treatments were administered in each group on days 3 and 5. Figure 15B shows quantification of tumor burden in SK-Hep1 xenografts treated with CD147-CAR-NK-92MI, parental NK-92MI cells (control group), and PBS (vehicle control group), respectively. All results are mean ± SEM. Differences between each group were analyzed by two-way ANOVA analysis. *p<0.05, **p<0.01, and ***p<0.001 indicate a comparison of the CD147-CAR-treated group with the NK-92MI-treated group. +p<0.05, ++p<0.01, and +++p<0.001 indicate a comparison of the CD147-CAR-treated group with the vehicle control-treated group. Figure 15C shows the quantitative body weights of each group assessed at the indicated time points, and Figure 15D shows Kaplan-Meier survival curves of tumor-bearing mice after treatment with CD147-CAR-NK-92MI cells, the parental NK-92MI group, and PBS (vehicle control group). p-values ​​were analyzed by log-rank (Mantel-Cox) test. [Figure 16]Figures 16A-16D show that CD147-CAR-NK-92MI cells control HCC progression in a PDX mouse model. Figure 16A is a diagram of the experimental design for the anti-tumor efficacy of CD147-CAR-NK-92MI in a hepatic PDX model generated by the Jackson Laboratory. Patient-derived xenograft liver cancer mice were purchased from the Jackson Laboratory. Four weeks after tumor implantation (day 1), tumor burden was determined (approximately 50 mm), and mice were randomly grouped. The indicated mice were then injected (iv) with a dose of 5 x 10 effector CD147-CAR-NK-92MI cells along with 2 x 10 IU of IL-2. Control groups were injected with the same number of NK-92MI cells in PBS along with 2 x 10 IU of IL-2 (Group #2) or PBS alone (Group #3). Identical treatments were administered in each group on days 6, 8, 11, 15, 18, 22, and 26, as indicated. Figure 16B shows the quantitative tumor burden of PDX mice treated with CD147-CAR-NK-92MI cells, parental NK-92MI cells (control group), and PBS (vehicle control group), respectively. All results are mean ± SEM. Differences for each group were analyzed by two-way ANOVA analysis. *p<0.05, **p<0.01, and ***p<0.001 indicate the CD147-CAR-treated group compared with the NK-92MI-treated group. +p<0.05, ++p<0.01, and +++p<0.001 indicate the CD147-CAR-treated group compared with the vehicle control-treated group. Figure 16C shows the quantitative body weight of each group assessed at the indicated time points. Figure 16D is a Kaplan-Meier survival curve of tumor-bearing mice after treatment with CD147-CAR-NK-92MI cells, the parental NK-92MI group, and PBS (vehicle control group). p-value analysis by log-rank (Mantel-Cox) test. [Figure 17]Figures 17A-17D show the killing of CD147-positive HCC cells by CD147-CAR-T cells. FFLuc reporter system assay of specific killing of FFLuc-EGFP-HepG2 by CD147-CAR-T cells (Figure 17A). The control group used wild-type kappa-CAR-T cells incubated with CD147-positive FFLuc-EGFP-HepG2. Figure 17B shows the reduced cytotoxicity of CD147-CAR-T cells using knockout CD147FFLuc-GFP-HepG1 by FFLuc reporter system assay. Figure 17C shows the cytotoxicity of CD147-CAR-T cells measured by a standard 4-hour 51Cr release assay. Figure 17D shows the significantly reduced cytotoxicity of CD147-CAR-T cells using knockout CD147FFLuc-GFP-HepG1 by a standard 4-hour 51Cr release assay. [Figure 18-1] Figures 18A and 18B show CD107a degranulation by CD147 CAR-T or -NK cells. Figure 18A shows representative flow cytometry data showing CD107a degranulation in CD147-CAR-T cells after 10 hours in medium (control) and SK-N-SH tumor cells. To quantify surface CD107a expression, cells were gated on the CD56-positive subset. Figure 18B shows quantitative data of the percentage of surface CD107a expression on CD147-CAR-NK-92MI cells under various stimuli, as indicated. Data were pooled from at least three or four experiments. [Figure 18-2] Same as above. [Figure 19] Figure 19 shows the cytotoxicity of CD147-CAR-NK-92MI cells against DaoY cells in vitro. [Figure 20] FIG. 20 is an alignment showing the optimized CD147scFv nucleic acid sequence (SEQ ID NO: 1) compared to the original scFv sequence (SEQ ID NO: 3) and the consensus sequence (SEQ ID NO: 13). [Figure 21-1]Figures 21A-21E show that patient-derived primary CD147-CAR-NK cells specifically kill CD147-positive tumor cells in vitro. Figure 21A shows representative H&E (top) and IHC (bottom) staining of liver samples from patients with HCC at various stages. Figure 21C shows a diagram of the experimental design for obtaining HCC samples from various sections of liver cancer tissue. Briefly, three regions of interest (tumor zone, adjacent zone, and non-tumor zone) were obtained. Primary NK cells were isolated from these zones (shown in Figure 21B). Figure 21D shows flow cytometry analysis of CD147-CAR-positive primary NK cells from various zones of liver tissue. Figure 21E shows the cytotoxicity of primary CD147-CAR-NK cells measured by a standard 51Cr release assay over 4 hours. [Figure 21-2] Same as above. [Figure 21-3] Same as above. [Figure 22-1] Figures 22A and 22B show representative flow cytometry analyses of CD147 expression on various cell types (Figure 22A) and the cytotoxicity of CD147-CAR-NK-92MI measured by a standard 4-hour 51Cr release assay against target cells with various CD147 expression levels (Figure 22B). Data are representative of two independent experiments. All data are presented as mean ± SEM. [Figure 22-2] Same as above. [Figure 23-1]Figures 23A-23H show that SynNotch GPC3-induced CD147-CAR T cells selectively target GPC3+CD147+ HepG2 cells, but not GPC3+CD147- or GPC3-CD147+ HepG2 cells. Figure 23A shows the schematic design of the CD147-CAR based on the GPC3-Gal4VP64-synNotch receptor in the SFG retroviral vector and the pHR lentiviral vector. The SFG retroviral vector contains eGFP, which can be used as a marker to select GPC3-Gal4VP64-synNotch-positive cells. The pHR construct contained a CD147-specific single-chain antibody fragment (clone 5F6), a human IgG1 CH2CH3 hinge region, and a CD28 transmembrane region, followed by the costimulatory CD28, the intracellular domain of 4-1BB, and the intracellular domain of CD3ζ. The pHR lentiviral vector contains mCherry, which can be used as a marker to select CD147-CAR-positive cells. Figure 23B shows the schematic design of the "logic-gated" GPC3-synNotch and CD147-CAR, which show induced cytotoxicity when both antigens are co-expressed but not when expressed individually on bystander or healthy cells. Figures 23C and 23D show the schematic experimental design of T cells co-transduced with GPC3-synNotch-GFP and CD147-CAR-mCherry vectors (Figure 23C) and a representative flow cytometry analysis of the expression of GPC3-synNotch-GFP and CD147-CAR-mCherry (Figure 23D). Figure 23E is a schematic experimental design of GPC3-synNotch-GFP and CD147-CAR-mCherry vectors co-transduced into T cells and primed with the GPC3highCD147lowHepG2 cell line, followed by analysis of CD147-CAR expression among various subsets of transduced T cells, including mCherry-positive only, GFP-positive only, GFP and mCherry double-positive, and GFP and mCherry double-negative subsets.Figure 23F shows a representative flow cytometry analysis of CD147-CAR expression on the surface of various subsets of transduced T cells. Both mean fluorescence intensity (MFI) and percentage of CD147-CAR are displayed in each representative flow cytometry chart. Figure 23G shows a representative flow cytometry analysis of CD147 and GPC3 expression in HepG2 tumor cell lines. Figure 23H shows a quantitative analysis of surface CD107a expression on various subsets of transduced T cells after a 2-hour "priming and triggering" protocol with various HepG2 tumor cell lines. Data are representative of two independent experiments. [Figure 23-2] Same as above. [Figure 23-3] Same as above. [Figure 24] Figure 24 shows that a gamma secretase inhibitor (MK-0752, a Notch signaling inhibitor) specifically blocks GPC3-SynNotch-induced CD147-CAR expression in GPC3-SynNotch-eGFP+ and CD147-CAR-mCherry+ primary T cell subsets, but not in other primary T cell subsets. Representative flow cytometry analysis of CD147-CAR expression in various primary T cell subsets (middle). Transduced T cells were treated with DMSO (0.3%, control), MK-0752 (10 μM), TAPI-1 (10 μM), GI254023X (10 μM), and a combination of MK-0752 + TAPI-1 + GI254023X, respectively. Meanwhile, these cells were primed in the presence of CD147KO GPC3high HepG2 cells. Expression of CD147-CAR on the surface of various subsets of transduced T cells was analyzed by flow cytometry. Both mean fluorescence intensity (MFI) and percentage of CD147-CAR are displayed in each representative flow cytometry chart. Data are representative of two independent experiments. All data are presented as mean ± SEM. [Figure 25-1]Figures 25A-25H show that SynNotch CD147-induced GPC3-CAR T cells selectively target GPC3+CD147+ HepG2 cells, but not GPC3+CD147- or GPC3-CD147+ HepG2 cells. Figure 25A shows the schematic design of the GPC3-CAR based on the Myc-CD147-Gal4VP64-SynNotch receptor in the SFG retroviral vector and the pHR lentiviral vector. The SFG retroviral vector contains a Myc tag, which can be used as a marker to select CD147-Gal4VP64-SynNotch-positive cells. The pHR construct consisted of a GPC3-specific single-chain antibody fragment (scFv, clone 5F6, mIgG1), a human IgG1 CH2CH3 hinge region, and a CD28 transmembrane region, followed by the costimulatory CD28, 4-1BB intracellular domain, and CD3ζ intracellular domain. The pHR lentiviral vector contains mCherry, which can be used as a marker to select cells containing GPC3-CAR-positive cells. Figure 25B shows the schematic design of the "logic gate" CD147-SynNotch and GPC3-CAR, which show induced cytotoxicity when both antigens are co-expressed but are not activated when expressed individually on bystander or healthy cells. Figures 25C and 25D show the schematic experimental design of T cells co-transduced with Myc-CD147-SynNotch and GPC3-CAR-mCherry vectors (Figure 25C) and a representative flow cytometry analysis of Myc-CD147-SynNotch and GPC3-CAR-mCherry expression (Figure 25D). Figure 25E is a schematic experimental design of Myc-CD147-SynNotch and GPC3-CAR-mCherry vectors co-transduced into T cells and primed with the GPC3highCD147lowHepG2 cell line, followed by analysis of GPC3-CAR expression among various subsets of transduced T cells, including mCherry-positive only, GFP-positive only, GFP and mCherry double-positive, and GFP and mCherry double-negative subsets.Figure 25F shows a representative flow cytometry analysis of GPC3-CAR expression on the surface of various subsets of transduced T cells. Both mean fluorescence intensity (MFI) and percentage of GPC3-CAR are displayed in each representative flow cytometry chart. Figure 25G shows a representative flow cytometry analysis of surface CD107a expression on various subsets of transduced T cells after a "prime and induce" protocol with various HepG2 tumor cell lines. Figure 25H shows a quantitative analysis of surface CD107a expression on various subsets of transduced T cells after a "prime and induce" protocol with various HepG2 tumor cell lines. The fold change in MFI of CD107a was calculated as follows: [(MFI sample - MFI priming only) / MFI priming only]. Data are representative of two independent experiments. All data are presented as mean ± SEM. *p<0.05, **p<0.01, and ***p<0.001. [Figure 25-2] Same as above. [Figure 25-3] Same as above. [Figure 26]Figures 26A-26D show that SynNotch GPC3-induced CD147-CAR T cells selectively kill GPC3highCD147highHepG2 cells but not CD147knockoutGPC3highHepG2 cells. Figures 26A and 26B are representative flow cytometry analyses of CD3, CD56, GPC3-synNotch-GFP, and CD147-CAR-mCherry expression. Primary PBMCs were transduced with CD147-CAR-mCherry lentivirus. These mCherry-positive T cells were sorted using flow cytometry and subsequently transduced with GPC3-synNotch-GFP retrovirus. Representative flow cytometry analyses of CD3 and CD56 (Figure 26A) and GPC3-synNotch-GFP and CD147-CAR-mCherry expression (Figure 26B), respectively, are shown. Figure 26C is a graph showing the cytotoxicity of primary GPC3-synNotch-GFP-CD147-CAR-mCherry T cells against HepG2-CD147high-GPC3high and HepG2-CD147knockout-GPC3high, as measured by a 7-hour FFluc reporter assay. Figure 26D is a graph showing the cytotoxicity of primary GPC3-synNotch-GFP-CD147-CAR-mCherry T cells against HepG2-CD147high-GPC3high and HepG2-CD147knockout-GPC3high, as measured by a 7-hour Cr-51 release assay. Data are representative of two independent experiments. DETAILED DESCRIPTION OF THE INVENTION

[0018] Sequence Listing Any nucleic acid and amino acid sequences listed herein or in the accompanying sequence listing are shown using standard abbreviations for nucleotide bases and amino acids as defined in 37 C.F.R. § 1.822. In at least some instances, only one strand of each nucleic acid sequence is shown, but the complementary strand is understood to be included by any reference to the displayed strand. SEQ ID NO: 1 is the nucleic acid sequence encoding the modified CD147scFv, with nucleotides 349 to 409 being a linker sequence. SEQ ID NO: 2 is the amino acid sequence of the modified CD147scFv. SEQ ID NO: 3 is the nucleic acid sequence encoding the starting CD147scFv. SEQ ID NO: 4 is the nucleic acid sequence encoding CD147-CAR: signal peptide: nucleotides 1-57; VH domain: nucleotides 58-411; linker sequence: nucleotides 412-468; VL domain: nucleotides 469-792; CD28 TM domain-41BB intracellular domain-CD3ζ domain: nucleotides 793-2202. SEQ ID NO: 5 is the amino acid sequence of CD147-CAR: signal peptide: amino acids 1-19; VH domain: amino acids 20-137; linker sequence: amino acids 138-156; VL domain: amino acids 157-264; CD28 TM domain-41BB intracellular domain-CD3ζ domain: amino acids 269-734. SEQ ID NO: 6 is the nucleic acid sequence encoding CD14-CAR containing inducible caspase 9. The caspase 9 sequence is from nucleotides 355 to 1200. SEQ ID NO: 7 is the amino acid sequence of CD147-iCaspase9-CAR. Amino acids 119 to 400 are iCaspase9. SEQ ID NO: 8 is the amino acid sequence of the anti-CD147 VH CDR domain. SEQ ID NO: 9 is the amino acid sequence of the anti-CD147 VL CDR domain. SEQ ID NOs: 10 and 11 are guide RNAs targeting CD147 used to generate CD147 knockout cell lines. SEQ ID NO: 12 is an exemplary IL-15 receptor intracellular domain. SEQ ID NO: 13 is the consensus nucleic acid sequence of the optimized CD147scFv nucleic acid sequence (SEQ ID NO: 1) compared to the original scFv sequence (SEQ ID NO: 3). SEQ ID NO: 14 is the nucleic acid sequence of the Gal4UAS CD147-CAR construct. SEQ ID NO: 15 is the nucleic acid sequence of the pHR_Gal4UAS-CD147-CAR-pGK_mCherry vector. SEQ ID NO: 16 is the nucleic acid sequence of GPC3-CAR. SEQ ID NO: 17 is a nucleic acid encoding the GAL4-VP64 activator. SEQ ID NOs: 18 to 35 are the nucleic acid sequences of primers used in constructing the plasmids.

[0019] Detailed Description Disclosed herein are immune cells (including T cells and NK cells) expressing novel CD147-targeted CARs. The biological properties of the CD147 antigen enable CD147-CAR-NK cells and CD147-CAR-T cells to produce potent antitumor activity against hepatocellular carcinoma in vitro and in vivo. Furthermore, CD147-CAR-NK cells can also kill human neuroblastoma cells in vitro.

[0020] Because CD147 is also expressed in several organs at varying expression levels, CD147-CAR-modified immune cells may exhibit "on-target, off-tumor" toxicity. Disclosed herein are constructs and methods to address this potential toxicity, including combination treatments such as "suicide genes" (such as inducible caspase 9) and CAR-expressing cells that are activated only upon binding of two antigens, such as CD147 and GPC3.

[0021] I. Terminology Unless otherwise noted, technical terms are used according to conventional usage. Definitions of common terms in molecular biology are taken from Lewin's Genes X, ed. Krebs et al., Jones and Bartlett Publishers, 2009 (ISBN 0763766321), Kendrew et al. (eds.), The Encyclopedia of Molecular Biology, published by Blackwell Publishers, 1994 (ISBN 0632021829), Robert A. Meyers (ed.), Molecular Biology and Biotechnology: a Comprehensive Desk Reference, published by Wiley, John & Sons, Inc., 1995 (ISBN 0471186341), and George P. Redei, Encyclopedic Dictionary of Genetics, Genomics, Proteomics and Informatics, 3 rd Edition, Springer, 2008 (ISBN: 1402067534), as well as other similar references.

[0022] Unless otherwise explained, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. The singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. "Comprising A or B" means including A, or B, or A and B. It should be further understood that all base sizes or amino acid sizes and all molecular weight or molecular mass values ​​given for nucleic acids or polypeptides are approximate and are provided for illustrative purposes.

[0023] Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of this disclosure, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety, as are GenBank accession numbers. In case of conflict, the present specification, including explanations of terms, will prevail. Furthermore, the materials, methods, and examples are illustrative only and are not intended to be limiting.

[0024] In order to facilitate review of the various embodiments of the disclosure, the following explanations of specific terms are provided.

[0025] Antibody: A polypeptide ligand containing at least one variable region that recognizes and binds to (e.g., specifically recognizes and binds to) an epitope of an antigen. Mammalian immunoglobulin molecules are composed of heavy (H) and light (L) chains, each of which has a variable heavy (V) domain. H ) area and variable light (V L ) region. H Area and V L Together, the regions are responsible for binding the antigen recognized by the antibody. There are five major heavy chain classes (or isotypes) of mammalian immunoglobulins, which determine the functional activity of the antibody molecule: IgM, IgD, IgG, IgA, and IgE.

[0026] Antibody variable regions contain "framework" regions and hypervariable regions known as "complementarity-determining regions" or "CDRs." The CDRs are primarily responsible for binding to an epitope of an antigen. The framework regions of an antibody serve to position and align the CDRs in three-dimensional space. The amino acid sequence boundaries of a given CDR can be readily determined using any of several well-known numbering schemes, including those described by Kabat et al. (Sequences of Proteins of Immunological Interest, US Department of Health and Human Services, 1991; "Kabat" numbering scheme), Chothia et al. (Chothia and Lesk, J Mol Biol 196:901-917, 1987; Chothia et al., Nature 342:877, 1989; and Al-Lazikani et al., (JMB 273, 927-948, 1997; see "Chothia" numbering scheme), and the ImMunoGeneTics (IMGT) database (Lefranc, Nucleic Acids Res 29:207-9, 2001; see "IMGT" numbering scheme). The Kabat and IMGT databases are maintained online.

[0027] Single-chain antibodies (scFv) are composed of one or more antibody V fragments linked by a suitable polypeptide linker as a genetically fused single-chain molecule. H and V L The V in scFv is a genetically engineered molecule containing a V domain (see, e.g., Bird et al., Science, 242:423-426, 1988; Huston et al., Proc. Natl. Acad. Sci., 85:5879-5883, 1988; Ahmad et al., Clin. Dev. Immunol., 2012, doi:10.1155 / 2012 / 980250; Marbry, IDrugs, 13:543-549, 2010). H Domains and VL The intramolecular orientation of the domains is typically not critical for scFvs. Therefore, both possible configurations (V H Domain-Linker Domain-V L Domain, V L Domain-Linker Domain-V H In a dsFv, a scFv having a V domain can be used. H and V L has been mutated to introduce disulfide bonds to stabilize the chain association. Diabodies are also included, which are V H and V L Bivalent, bispecific antibodies in which the domains are expressed on a single polypeptide chain but use a linker that is too short to allow pairing between the two domains on the same chain, thereby allowing the domains to pair with complementary domains on another chain and create two antigen-binding sites (see, e.g., Holliger et al., Proc. Natl. Acad. Sci., 90:6444-6448, 1993; Poljak et al., Structure, 2:1121-1123, 1994).

[0028] Antibodies also include genetically engineered forms such as chimeric antibodies (e.g., humanized mouse antibodies) and heteroconjugate antibodies (e.g., bispecific antibodies). Pierce Catalog and Handbook, 1994-1995 (Pierce Chemical Co., Rockford, IL); Kuby, J., Immunology, 3 rd Ed., W.H. Freeman & Co., New York, 1997.

[0029] Cancer: A malignant tumor characterized by abnormal or uncontrolled cell growth. Other features often associated with cancer include metastasis, interference with the normal function of neighboring cells, release of abnormal levels of cytokines or other secretory products and suppression or exacerbation of inflammatory or immunological responses, and invasion of surrounding or distant tissues or organs such as lymph nodes. "Metastatic disease" refers to cancer cells that have left the original tumor site and traveled to other parts of the body, for example, via the bloodstream or lymphatic system.

[0030] CD147: Extracellular matrix metalloproteinase inhibitor (EMMPRIN or EMPRIN), also known as basigin (BSG), is a transmembrane glycoprotein with multiple functions in normal cell function and disease (Hahn et al., J. Leukocyte Biol. 98:33-48, 2015). CD147 is important in immune cells for T cell activation and proliferation, as well as cell migration, adhesion, and invasion (Hahn et al., J. Leukocyte Biol. 98:33-48, 2015). CD147 is expressed at various levels in various cell types (e.g., hematopoietic, epithelial, and endothelial cells) (Liao et al., Mol. Cell Biol. 31:2591-2604, 2011) and can be significantly upregulated in disease states such as HCC.

[0031] The sequence of CD147 is publicly available. For example, GenBank accession numbers NM_198590, NM_198591, NM_001728, NM_198589, and NM_001322243 disclose the nucleic acid sequence of human CD147, and GenBank accession numbers NP_940992, NP_940993, NP_001719, NP_940991, and NP_001309172 disclose the amino acid sequence of human CD147. Similarly, NM_009768 and NM_001077184 disclose the nucleic acid sequence of mouse CD147, and GenBank accession numbers NP_033898 and NP_001070652 disclose the amino acid sequence of mouse CD147. All of these sequences are incorporated by reference as they exist in GenBank on March 15, 2019.

[0032] Chimeric antigen receptor (CAR): A chimeric molecule comprising an antigen-binding portion (such as a single-domain antibody or scFv) and a signaling domain, such as a signaling domain derived from a T cell receptor (e.g., CD3ζ). Typically, a CAR comprises an antigen-binding portion, a transmembrane domain, and an intracellular domain. The intracellular domain typically comprises a signaling chain having an immunoreceptor tyrosine-based activation motif (ITAM), such as CD3ζ or FcεRIγ. In some cases, the intracellular domain also comprises the intracellular portion of at least one additional costimulatory domain, such as CD28, 4-1BB (CD137), ICOS, OX40 (CD134), CD27, and / or DAP10.

[0033] Complementarity-Determining Region (CDR): A region of hypervariable amino acid sequence that defines the binding affinity and specificity of an antibody. The light and heavy chains of mammalian immunoglobulins each have three CDRs designated VL-CDR1, VL-CDR2, VL-CDR3, and VH-CDR1, VH-CDR2, VH-CDR3, respectively.

[0034] Glypican-3 (GPC3): A cell surface heparan sulfate proteoglycan that binds to and inhibits the activity of CD26. GPC3 can induce apoptosis in some cell types. GPC3 is expressed by some tumors, including hepatocellular carcinoma, melanoma, ovarian clear cell carcinoma, yolk sac tumor, neuroblastoma, hepatoblastoma, and Wilms' tumor.

[0035] The sequence of GPC3 is publicly available.For example, GenBank accession numbers NM_001164619, NM_001164618, NM_004484, and NM_001164617 disclose exemplary human GPC3 nucleic acid sequences, and GenBank accession numbers NP_001158091, NP_001158090, NP_004475, and NP_001158089 disclose exemplary human GPC3 amino acid sequences.Similarly, GenBank accession number NM_016697 discloses exemplary mouse GPC3 nucleic acid sequences, and GenBank accession number NP_057906 discloses exemplary mouse GPC3 amino acid sequences.Each of these sequences is incorporated by reference as they exist in GenBank on February 27, 2020.

[0036] Isolated: An "isolated" biological component, such as a nucleic acid, protein (including an antibody), or organelle, has been substantially separated or purified from other biological components, i.e., other chromosomal and extrachromosomal DNA and RNA, proteins, and organelles, in the environment (such as a cell) in which the component naturally occurs. "Isolated" nucleic acids and proteins include nucleic acids and proteins purified by standard purification methods. The term also encompasses nucleic acids and proteins prepared by recombinant expression in a host cell, as well as chemically synthesized nucleic acids.

[0037] Liver cancer: Hepatocellular carcinoma (HCC) is the most common type of primary malignant tumor of the liver and often occurs in patients with viral hepatitis (e.g., hepatitis B or C), exposure to toxic substances, or cirrhosis (which can be caused by alcoholism). Other types of liver cancer include intrahepatic cholangiocarcinoma (iCCA), fibrolamellar carcinoma, and hepatoblastoma.

[0038] Natural killer (NK) cells: Cells of the immune system that kill target cells without specific antigenic stimulation and without MHC class restriction. Target cells can be tumor cells or virus-bearing cells. NK cells are characterized by the presence of the CD56 surface marker and the absence of the CD3 surface marker. NK cells typically account for approximately 10-15% of the mononuclear cell fraction in normal peripheral blood. Historically, NK cells were first identified for their ability to lyse certain tumor cells without prior immunization or activation. NK cells are thought to provide a "backup" defense mechanism against viruses and tumors that can evade CTL responses by downregulating MHC class I presentation. In addition to their involvement in direct cytotoxic killing, NK cells also play a role in cytokine production, which may be important for controlling cancer and infection.

[0039] In some examples, "modified NK cells" are NK cells that have been transduced or transfected with a heterologous nucleic acid (e.g., one or more of the nucleic acids or vectors disclosed herein) or that express one or more heterologous proteins. The terms "modified NK cells" and "transduced NK cells" are used interchangeably in some examples herein.

[0040] Purified: The term "purified" does not require complete purity. Rather, it is intended as a relative term. Thus, for example, a purified protein or nucleic acid preparation is one in which the protein or nucleic acid is more enriched than the protein or nucleic acid is in its natural environment (e.g., within a cell). In one embodiment, the preparation is purified such that the protein or nucleic acid represents at least 50% of the total protein or nucleic acid content of the preparation. Substantial purification refers to purification from other proteins or cellular components. A substantially purified protein or nucleic acid is at least 60%, 70%, 80%, 90%, 95%, or 98% pure. Thus, in one specific, non-limiting example, a substantially purified protein or nucleic acid is 90% free from other components.

[0041] Recombinant: A nucleic acid or protein having a sequence that is not found in nature or that is made by the artificial combination of two otherwise separated segments of sequence (e.g., a "chimeric" sequence). This artificial combination can be accomplished by chemical synthesis or by the artificial manipulation of isolated segments of nucleic acid, e.g., by genetic engineering techniques.

[0042] Subject: Living multi-cellular vertebrate organisms, a category that includes both human and veterinary subjects, including human and non-human mammals.

[0043] T cell: A white blood cell (lymphocyte) that is an important mediator of the immune response. T cells contain CD4 + T cells and CD8 + These include, but are not limited to, T cells. + T lymphocytes are immune cells that carry a marker on their surface known as "cluster of differentiation 4" (CD4). These cells, also known as helper T cells, help orchestrate immune responses, including antibody responses as well as killer T cell responses. CD8 + The T cells bear the "cluster of differentiation 8" (CD8) marker. In one embodiment, CD8 +The T cells are cytotoxic T lymphocytes (CTLs). In another embodiment, CD8 + The cells are suppressor T cells.

[0044] Activated T cells can be detected by increased cell proliferation and / or expression or secretion of one or more cytokines (such as IL-2, IL-4, IL-6, IFNγ, or TNFα). Activation of CD8+ T cells can also be detected by increased cytolytic activity in response to antigen.

[0045] In some instances, an "modified T cell" is a T cell that has been transduced or transfected with a heterologous nucleic acid (e.g., one or more of the nucleic acids or vectors disclosed herein) or that expresses one or more heterologous proteins. The terms "modified T cell" and "transduced T cell" are used interchangeably in some instances herein.

[0046] Transduced or transformed: A transformed cell is a cell into which a nucleic acid molecule has been introduced by molecular biological techniques. As used herein, the terms "transduction" and "transformation" encompass all techniques by which a nucleic acid molecule can be introduced into such a cell, including transfection with viral vectors, the use of plasmid vectors, and the introduction of DNA by electroporation, lipofection, and particle gun acceleration.

[0047] Treating or ameliorating a disease: "Treating" refers to a therapeutic intervention that reduces or inhibits the signs or symptoms of a disease or condition after it has begun to develop, e.g., reducing tumor size or tumor burden. "Ameliorating" refers to a reduction in the number or severity of signs or symptoms of a disease, such as cancer.

[0048] Vector: A nucleic acid molecule that can be introduced into a host cell (e.g., by transfection or transduction), thereby producing a transformed host cell. A recombinant DNA vector is a vector that contains recombinant DNA. A vector may contain a nucleic acid sequence that allows it to replicate in a host cell, such as an origin of replication. A vector may also contain one or more selectable marker genes and other genetic elements known in the art. A viral vector is a recombinant nucleic acid vector that contains at least some nucleic acid sequences derived from one or more viruses. A replication-deficient viral vector is a vector that lacks at least one replication-essential gene function and therefore requires complementation of one or more regions of the viral genome required for replication.

[0049] II. Overview of Some Embodiments Disclosed herein is a CD147-specific binding agent comprising a modified CD147scFv. Also disclosed is a chimeric antigen receptor (CAR) encoding a CD147-specific binding agent fused to an intracellular domain comprising a hinge region, a transmembrane domain, and a costimulatory domain and an intracellular signaling domain. In some instances, the costimulatory domain is derived from CD28 and / or 4-1BB, and the signaling domain is derived from CD3ζ. Also provided are nucleic acids encoding CD147-specific binding agents and CD147-CARs, and vectors comprising the nucleic acids.

[0050] Also provided herein are engineered immune cells (e.g., T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages) that express a CD147-CAR. In some embodiments, the engineered immune cells express one or more additional CARs, such as a CAR that targets a hepatitis virus (e.g., hepatitis B or hepatitis C). In other embodiments, the engineered immune cells express an inducible CD147-CAR and a construct that induces expression of the CD147-CAR, such as an anti-GPC3 SynNotch construct that drives expression of an inducer of the CD147-CAR. In further embodiments, the engineered immune cells express an inducible GPC3-CAR and a construct that induces expression of the GPC3-CAR, such as an anti-CD147 SynNotch construct that drives expression of an inducer of the GPC3-CAR. In a particular example, the CD147-CAR and / or the anti-CD147 SynNotch construct comprises the engineered CD147scFv disclosed herein.

[0051] Also provided are methods for treating CD147-expressing cancer in a subject. In some embodiments, the method comprises administering to the subject an effective amount of modified immune cells (e.g., T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages) comprising a nucleic acid encoding a CD147-CAR. In other embodiments, the method comprises administering to the subject an effective amount of modified immune cells (e.g., T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages) comprising a nucleic acid encoding an inducible CD147-CAR and a nucleic acid for anti-GPC3 SynNotch expressing an inducer of CD147-CAR expression. In yet other embodiments, the method comprises administering to the subject an effective amount of modified immune cells (e.g., T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages) comprising a nucleic acid encoding a GPC3-CAR and an anti-CD147 SynNotch construct expressing an inducer of GPC3-CAR expression. In some examples, the modified immune cells are autologous. In other examples, the immune cells are allogeneic. In some specific examples, the subject has hepatocellular carcinoma or neuroblastoma.

[0052] III. CD147-Specific Binding Agents Disclosed herein are CD147-binding agents, in some instances, used as targeting moieties of chimeric antigen receptors. In some embodiments, the CD147-binding agent is a CD147scFv, a modified fragment encoding a CD147-specific scFv derived from the 5F6 clone (U.S. Patent No. 8,618,264). Figure 20 shows an alignment of the starting CD147scFv nucleic acid sequence (SEQ ID NO: 3) and the modified scFv nucleic acid sequence (SEQ ID NO: 1). In some instances, the modified CD147-specific scFv binds to highly expressing CD147 cells.

[0053] In some embodiments, the CD147-specific binding agent is a single-domain antibody (e.g., an scFv) that specifically binds to CD147 and comprises the CDR sequences provided in Table 1. In some examples, the scFv specifically binds to CD147 and comprises an amino acid sequence comprising the amino acid sequences of variable heavy chain (VH) domain complementarity-determining region 1 (CDR1), CDR2, and CDR3 of SEQ ID NO: 8, and the amino acid sequences of variable light chain (VL) domain CDR1, CDR2, and CDR3 of SEQ ID NO: 9. In some embodiments, the scFv comprises the CDR amino acid sequences provided in Table 1 and has at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% sequence identity) to the amino acid sequence of SEQ ID NO: 1. In other embodiments, the scFv comprises or consists of the amino acid sequence of SEQ ID NO: 1. [Table 1]

[0054] In additional examples, the scFv is encoded by a nucleic acid comprising the CDR sequences provided in Table 1 and has at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% sequence identity) to the nucleotide sequence of SEQ ID NO: 2. In other embodiments, the scFv is encoded by a nucleic acid comprising or consisting of the nucleotide sequence of SEQ ID NO: 2.

[0055] Also provided are vectors comprising a nucleic acid encoding the CD147-specific binding agent described above. In some examples, the vector comprises a nucleic acid having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% identity) with SEQ ID NO: 1, and encodes an amino acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% identity) with, for example, SEQ ID NO: 2.

[0056] In a further example, provided herein is a nucleic acid encoding a CD147-specific binding agent (e.g., a nucleic acid having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% identity to SEQ ID NO: 1) operably linked to a nucleic acid encoding Gal4-VP64 (e.g., SEQ ID NO: 17). In some examples, the nucleic acid is part of a SynNotch construct. When CD147 binds to the protein encoded by this construct, Gal4-VP64 is activated and can induce expression of a construct having a Gal4 response element (e.g., Gal4UAS). An exemplary vector encoding an anti-CD147scFv SynNotch inducer construct is shown in Figure 25A.

[0057] IV. CD147 Chimeric Antigen Receptor Provided herein is a CD147-CAR comprising the CD147-specific binding agent described in Section III above. In some embodiments, the CAR comprises an antigen-binding domain comprising a CD147-specific scFv (e.g., SEQ ID NO: 2), a hinge domain, a transmembrane domain, and an intracellular domain comprising at least one costimulatory domain and an intracellular signaling domain.

[0058] In some embodiments, the antigen-binding domain is a CD147-specific scFv that has, for example, an amino acid sequence having at least 90% sequence identity to SEQ ID NO:2 (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99% identity), or that comprises or consists of the amino acid sequence of SEQ ID NO:2.

[0059] In some embodiments, the hinge domain is an IgG hinge region. In one example, the hinge domain is an IgG1 hinge. Other hinge domains can be used, such as hinge regions from other immunoglobulins (e.g., IgG4 or IgD), or hinge regions from CD8, CD28, or CD40.

[0060] In additional embodiments, the transmembrane domain is a CD28 transmembrane domain. In one example, the transmembrane domain is derived from CD28. The transmembrane domain may also be derived from other T cell proteins, such as CD8, CD4, CD3ζ, CD40, OX40L, 41BBL, ICOS, ICOS-L, CD80, CD86, ICAM-1, LFA-1, ICAM-1, CD56, CTLA-4, PD-1, TIM-3, NKP30, NKP44, NKP40, NKP46, B7-H3, PD-L1, PD-2, and CD70.

[0061] In further embodiments, the intracellular domain comprises one or more intracellular regions, or portions thereof, derived from a costimulatory molecule. Exemplary costimulatory molecules include CD28, 4-1BB, CD8, CD40, OX-40, ICOS, CD27, and DAP10, OX40-L, 4-1BBL, ICOS-L, CD80, CD86, ICAM-1, LFA-1, CD56, CTLA-4, PD-1, TIM-3, NKP30, NKP44, NKP40, NKP46, B7-H3, PD-L1, PD-2, and CD70. In certain examples, the costimulatory domain is derived from CD28 and / or 4-1BB. In one example, the costimulatory domain comprises domains derived from both CD28 and 4-1BB. The intracellular domain also comprises an intracellular signaling domain derived from CD3ζ. In other examples, the intracellular signaling domain is derived from DAP10, DAP12, PDK, or FcεRIγ. In one example, the intracellular signaling domain is derived from CD3ζ.

[0062] In some embodiments, CD147-CAR also comprises a signal sequence, which is located at the N-terminus of the scFv domain. In some examples, the signal sequence is an IgG signal sequence or a GM-CSF signal sequence. In one example, the signal sequence is amino acids 1-19 of SEQ ID NO:5.

[0063] In certain embodiments, the CD147-CAR comprises an amino acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99% identity) to SEQ ID NO: 5. In some examples, the CD147-CAR comprises or consists of the amino acid sequence of SEQ ID NO: 5.

[0064] In additional embodiments, the CD147-CAR further comprises an inducible gene (e.g., a "suicide" gene) that can be used to eliminate CD147-CAR-expressing cells. The inducible gene can be activated in the event of off-target side effects (or on-target / off-tumor effects), such as cytokine release syndrome ("cytokine storm"). In some examples, expression of the suicide gene is inducible by small molecules such as tetracycline or doxycycline (the "TET ON" system) or rapamycin. See, for example, Gargett et al., Front. Pharmacol. 5:235, 2014; Stavrou et al., Mol. Ther. 6:1266-1276, 2018. In other examples, the suicide gene is inducible by a Fas domain-inducible system. In some examples, the inducible suicide domain is located at the N-terminus or C-terminus of the antigen-binding domain of the CAR, while in other examples, the inducible suicide domain is located at the C-terminus of the CD3ζ domain of the CAR. The inducible suicide domain is separated from the CAR by a self-cleaving peptide (e.g., a P2A peptide or a T2A peptide). In some embodiments, the inducible suicide domain comprises caspase 9, e.g., amino acids 119-400 of SEQ ID NO: 7. In some examples, a CD147-CAR comprising tetracycline-inducible caspase 9 comprises an amino acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99% identity) to SEQ ID NO: 7. In other examples, a CD147-CAR comprising tetracycline-inducible caspase 9 comprises or consists of the amino acid sequence of SEQ ID NO: 7.

[0065] In other embodiments, the CD147-CAR further comprises a domain that increases the survival or persistence of modified immune cells expressing the CAR. In some examples, the domain is an intracellular domain derived from a cytokine receptor, such as an intracellular domain derived from interleukin (IL) receptor 15 (e.g., SEQ ID NO: 12), an IL-12 receptor, or an IL-18 receptor. In other examples, the domain is an intracellular domain derived from a growth factor receptor, such as an intracellular domain derived from CD40, NKG2D, NKP40, or NKP46. In some examples, the domain is located at the C-terminus of the CD3ζ domain of the CAR.

[0066] In some examples, the CD147-CAR further comprises one or more additional antigen-binding domains that specifically bind to antigens co-expressed with CD147 on tumor cells. In some non-limiting examples, the CD147-CAR comprises at least one additional antigen-binding domain that specifically binds to one or more liver cancer antigens, such as glypican-3, alpha-fetoprotein, or mucin-1. The additional tumor antigen can be selected based on the type of cancer to be treated.

[0067] Nucleic acids encoding the CD147-CAR disclosed herein are also provided. In some embodiments, the nucleic acid encodes a CAR comprising a CD147-specific scFv, for example, a nucleic acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% sequence identity) to SEQ ID NO: 1, or comprises or consists of the nucleic acid sequence of SEQ ID NO: 1. In some examples, the CD147-specific CAR nucleic acid also encodes an IgG hinge domain, a CD28 transmembrane domain, a CD28 and 4-1BB costimulatory domain, and a CD3ζ domain. In one example, the CD147-specific CAR is encoded by, or comprises or consists of, a nucleic acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% sequence identity) to SEQ ID NO: 4. In other examples, the CD147-specific CAR nucleic acid also encodes, or comprises or consists of, a nucleic acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% sequence identity) to an inducible caspase 9 domain, e.g., SEQ ID NO: 6.

[0068] Functional variants of the CARs described herein, or domains thereof, are also provided, which retain the biological activity of the CARs that are variants or retain the biological activity of a particular domain. Functional variants can be at least about 80%, about 85%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, or more identical in amino acid sequence to the parent CAR or domain. Substitutions can be made, for example, in one or more of the extracellular targeting domain, hinge domain, transmembrane domain, and intracellular domain.

[0069] In some examples, the functional variant comprises the amino acid sequence of a parent CAR or domain with at least one conservative amino acid substitution (e.g., up to 10 conservative amino acid substitutions, e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 conservative substitutions). In other examples, the functional variant comprises the amino acid sequence of a parent CAR or domain with at least one non-conservative amino acid substitution (e.g., up to 10 non-conservative amino acid substitutions, e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 non-conservative substitutions). In this case, the non-conservative amino acid substitution does not interfere with or inhibit the biological activity of the functional variant. The non-conservative amino acid substitution may enhance the biological activity of the functional variant, thereby increasing the biological activity of the functional variant compared to the parent CAR or domain.

[0070] A CAR or a domain thereof may, in some instances, include one or more synthetic amino acids in place of one or more naturally occurring amino acids, such as, for example, aminocyclohexanecarboxylic acid, norleucine, α-amino n-decanoic acid, homoserine, S-acetylaminomethyl-cysteine, trans-3- and trans-4-hydroxyproline, 4-aminophenylalanine, 4-nitrophenylalanine, 4-chlorophenylalanine, 4-carboxyphenylalanine, β-phenylserine. CARs include β-hydroxyphenylalanine, phenylglycine, α-naphthylalanine, cyclohexylalanine, cyclohexylglycine, indoline-2-carboxylic acid, 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid, aminomalonic acid, aminomalonic acid monoamide, N'-benzyl-N'-methyl-lysine, N',N'-dibenzyl-lysine, 6-hydroxylysine, ornithine, α-aminocyclopentanecarboxylic acid, α-aminocyclohexanecarboxylic acid, oc-aminocycloheptanecarboxylic acid, -(2-amino-2-norbornane)-carboxylic acid, γ-diaminobutyric acid, α,β-diaminopropionic acid, homophenylalanine, and α-tert-butylglycine. CARs may be glycosylated, amidated, carboxylated, phosphorylated, esterified, N-acylated, cyclized, for example, by a disulfide bond, or converted to an acid addition salt, and / or optionally dimerized or polymerized, or conjugated.

[0071] In some embodiments, the nucleic acid molecule encoding the disclosed CAR is included in an expression vector (e.g., a viral vector) for expression in a host cell, such as a T cell or an NK cell. In some examples, the expression vector comprises a promoter operably linked to the nucleic acid molecule encoding the CAR. Additional expression control sequences, such as one or more enhancers, transcriptional and / or translational terminators, and initiation sequences, can also be included in the expression vector. In some embodiments, the nucleic acid encoding the CD147-CAR provided herein is included in a viral vector. Examples of suitable viral vectors include retroviruses (e.g., MoMLV or lentivirus), adenoviruses, adeno-associated viruses, vaccinia viruses, and fowlpox viruses. In a specific example, the nucleic acid encoding the CD147-CAR is included in a MoMLV vector, such as an SFG retroviral vector or a pHAGE-CPPT lentiviral vector. In other examples, the vector may be a DNA vector.

[0072] In some embodiments, the vector further comprises an upstream activating sequence (UAS) that allows for inducible expression of CD147-CAR. In one non-limiting example, the UAS is Gal4UAS, which is activated by Gal4. However, one of skill in the art can identify other transactivation systems that can be utilized. In one example, the Gal4UAS CD147-CAR nucleic acid comprises a nucleic acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% identity) to SEQ ID NO: 14. An exemplary vector encoding a Gal4UAS CD147-CAR construct, such as pHR_Gal4UAS-CD147-CAR-pGK_mCherry, is shown in Figure 23A. In some examples, the vector comprises a nucleic acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% identity) to SEQ ID NO: 15. In some examples, the vector comprises a selectable marker (e.g., mCherry in SEQ ID NO: 15), while in other examples, a selectable marker is not included in the vector.

[0073] In some examples, the vector further comprises a nucleic acid sequence encoding at least one additional CAR. In some examples, the additional CAR is specific for an additional tumor antigen, for example, increasing the specificity of targeting the CD147-CAR to tumor cells that express or overexpress CD147. In some examples, the vector comprises a nucleic acid encoding one or more CARs comprising an antigen-binding domain that specifically binds to one or more liver cancer antigens, such as glypican-3, alpha-fetoprotein, or mucin-1. The additional tumor antigen / CAR can be selected based on the type of cancer to be treated. In other examples, the vector comprises a nucleic acid encoding a CAR comprising an antigen-binding domain specific to hepatitis B or hepatitis C. In some examples, the additional CAR binds to HBV envelope protein or hepatitis B surface antigen (HBsAg). In other examples, the additional CAR binds to HCV E2 glycoprotein. In some examples, one or more additional CARs are included in a vector with the CD147-CAR separated by a self-cleaving peptide, such as a P2A peptide sequence.

[0074] V. Cells expressing CD147 CAR or CD147-specific binding agents Also provided herein are cells (e.g., immune cells) that express the disclosed CD147-CAR or CD147-specific binding agent. In certain embodiments, the cells include T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages. In some embodiments, the cells are T cells, NK cells, or macrophages that express CD147-CAR.

[0075] In some examples, the cells further express a GPC3-specific binding agent, such as an anti-GPC3 scFV. In certain examples, the cells express an anti-GPC3 scFV (e.g., a SynNotch construct) operably linked to a nucleic acid encoding Gal4-VP64. When GPC3 binds to the protein encoded by this construct, Gal4-VP64 is activated and can induce the expression of a construct having a Gal4 response element (e.g., Gal4UAS). An exemplary vector encoding an anti-GPC3 scFv SynNotch inducer construct is shown in Figure 23A. Thus, in some examples, T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages express an inducible CD147-CAR (e.g., SEQ ID NO: 14) and an anti-GPC3 binding agent construct that induces the expression of the inducible CD147-CAR.

[0076] In other embodiments, the cells are T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages that express CD147scFv (e.g., SEQ ID NO: 1) and a CD147-specific binding agent such as GPC3-CAR. In some examples, the GPC3-CAR comprises a nucleic acid sequence having at least 90% sequence identity (e.g., at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% identity) with SEQ ID NO: 16. In certain examples, the cells express an anti-CD147scFV (e.g., a SynNotch construct) operably linked to a nucleic acid encoding Gal4-VP64. When CD147 binds to the protein encoded by this construct, Gal4-VP64 is activated and can induce the expression of a construct having a Gal4-responsive element (e.g., Gal4UAS). Exemplary vectors encoding an anti-CD147scFv SynNotch inducer construct and an anti-GPC3-CAR are shown in Figure 25A. Thus, in some examples, T cells, NK cells, NKT cells, DNT cells, neutrophils, macrophages express an inducible GPC3-CAR (e.g., SEQ ID NO: 16) and an anti-CD147 binder construct that induces expression of the inducible GPC3-CAR.

[0077] In some examples, immune cells are transduced or transfected with one or more expression vectors containing one or more nucleic acids, including nucleic acids encoding a CD147-CAR, an inducible GPC3-CAR, an inducible CD147-CAR, a CD147-specific binding agent operably linked to an inducer, a GPC3-specific binding agent operably linked to an inducer, or any combination of two or more thereof. In other examples, the vector (or DNA encoding the construct) can be introduced by contacting the cells with nanoparticles containing the vector or DNA. In some examples, the cells are irradiated following transduction or transfection, e.g., before being administered to a subject (e.g., treated with gamma irradiation, e.g., at a dose of at least 1,000, at least 2,000, at least 3,000, at least 5,000, at least 7,000, at least 8,000, at least 9,000, at least 10,000, at least 11,000, at least 12,000, or at least 15,000, or about 1,000-15,000, 2,000-12,000, 1,000-5,000, 5,000-10,000, or 8,000-12,000, or about 10,000 Rads).

[0078] In some examples, the transduced or transfected cells are isolated T cells (e.g., primary T cells or T cells obtained from a subject), isolated NK cells (e.g., primary NK cells or NK cells obtained from a subject), isolated NKT cells, isolated DNT cells, isolated neutrophils, or isolated macrophages (e.g., primary macrophages or macrophages obtained from a subject). In some examples, the T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages are obtained from peripheral blood, umbilical cord blood, lymph node tissue, bone marrow, or tumor tissue. In some examples, the T cells, NK cells, NKT cells, or DNT cells are also enriched, purified, and / or expanded from a sample from a subject, e.g., before and / or after transduction with one or more of the disclosed expression vectors.

[0079] In one non-limiting embodiment, the cells are NK-92 cells. NK-92 cells are an NK cell line (e.g., ATCC® CRL-2407™) derived from a patient with non-Hodgkin's lymphoma. This cell line has the properties of activated NK cells (see, e.g., Gong et al., Leukemia 8:652-658, 1994). In another embodiment, the cells are NK-92MI cells (e.g., ATCC® CRL-2408™). The NK-92MI cell line is an interleukin-2 (IL-2)-independent NK cell line derived from NK-92 that stably expresses human IL-2 (see, e.g., Tam et al., Hum. Gene Ther. 10:1359-1373, 1999). NK-92 or NK-92MI cells expressing a CAR (e.g., CD147-CAR and / or other nucleic acids disclosed herein) can be used herein as an "off-the-shelf" immunotherapy, as it eliminates the need to generate autologous NK cells for each subject. Other NK cell lines that can be used with the CD147-CAR (or other nucleic acids) described herein include NKL, KHYG-1, and YTS cells.

[0080] Currently, phase I / II clinical trials of NK-92-mediated immunotherapy are underway (Arai et al., 2008; Tonn et al., 2013). Generally, NK-92 cells must be irradiated prior to infusion to prevent permanent engraftment. The required dose of irradiation is approximately 10 Gy. The infusion dose of irradiated NK-92 is up to 10 10 NK92 cells / m 2 Importantly, irradiated NK-92 cells have been shown to be safe for infusion in patients, as demonstrated in several NK-92 clinical trials (NCT00900809, NCT00990717, NCT00995137, and NCT01974479).

[0081] In some non-limiting embodiments, immune cells are transduced with a vector encoding CD147-CAR. Following transduction, cells expressing CD147-CAR can be detected and / or enriched, for example, by flow cytometry using a labeled antibody that binds to CD147. In some examples, the transduced cells (e.g., NK cells or T cells) are expanded, for example, by cell culture for a period following transduction. In some examples, some or all of the modified cells are cryopreserved for later use.

[0082] VI. Immunotherapy Methods Methods are provided for treating cancer in a subject (e.g., a cancer that expresses or overexpresses CD147) using the CD147-CAR disclosed herein. In some embodiments, the method comprises administering to the subject a composition comprising modified T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages that express the CD147-CAR (e.g., transduced with a vector encoding a CAR) and a pharmaceutically acceptable carrier. In other examples, the method comprises administering to the subject a pharmaceutical composition comprising an expression vector encoding a CD147-CAR and a pharmaceutically acceptable carrier.

[0083] Also provided are methods for treating cancer in a subject (e.g., a cancer that expresses or overexpresses CD147) using an inducible CD147-CAR disclosed herein and a GPC3-specific binding agent linked to expression of an inducer of the inducible CD147-CAR. In some embodiments, the method comprises administering to the subject a composition comprising engineered T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages (e.g., transduced with one or more vectors encoding a CD147-CAR and a GPC3-specific binding agent) that express an inducible CD147-CAR and a GPC3-specific binding agent linked to an inducer, and a pharmaceutically acceptable carrier. In another example, the method comprises administering to the subject one or more expression vectors encoding an inducible CD147-CAR (e.g., SEQ ID NO: 15) and a GPC3-specific binding agent linked to an inducer, and a pharmaceutically acceptable carrier.

[0084] Also provided are methods for treating cancer (e.g., cancers that express or overexpress CD147) in a subject using an inducible GPC3-CAR and a CD147-specific binding agent disclosed herein linked to an inducer of the inducible GPC3-CAR. In some embodiments, the method comprises administering to the subject a composition comprising modified T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages (e.g., transduced with one or more vectors encoding a GPC3-CAR (e.g., SEQ ID NO: 16) and a CD147-specific binding agent) that express an inducible GPC3-CAR and a CD147-specific binding agent linked to an inducer, and a pharmaceutically acceptable carrier. In another example, the method comprises administering to the subject one or more expression vectors encoding an inducible GPC3-CAR and a CD147-specific binding agent linked to an inducer, and a pharmaceutically acceptable carrier.

[0085] The modified cells (e.g., T cells, NK cells, NKT cells, DNT cells, neutrophils, or macrophages) expressing the CD147-CAR or CD147-specific binding agents described herein can be incorporated into pharmaceutical compositions. Such compositions typically comprise a population of cells (e.g., CD147-CAR-NK cells or CD147-CAR-T cells) and a pharmaceutically acceptable carrier. A "pharmaceutically acceptable carrier" includes any and all solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic and absorption delaying agents, and the like, compatible with pharmaceutical administration (see, e.g., Remington: The Science and Practice of Pharmacy, The University of the Sciences in Philadelphia, Editors, Lippincott, Williams, & Wilkins, Philadelphia, PA, 21 st (See, Edition, 2005). Examples of such carriers or diluents include, but are not limited to, water, saline, Ringer's solution, dextrose solution, balanced salt solution, and 5% human serum albumin. Non-aqueous vehicles such as liposomes and fixed oils may also be used. Complementary active compounds may also be incorporated into the composition. Practical methods for preparing administrable compositions are described in Remington: The Science and Practice of Pharmacy, The University of the Sciences in Philadelphia, Editor, Lippincott, Williams, & Wilkins, Philadelphia, PA, 21 st Edition (2005).

[0086] In some instances, the composition 4 ~10 12 modified NK cells or T cells (e.g., about 10 4 ~10 8 cells, approximately 10 6 ~10 8 cells, or approximately 10 6 ~1012 For example, the composition may contain about 10 4 ~10 10 modified NK cells or modified T cells / kg (e.g., about 10 4 , 10 5 , 10 6 , 10 7 , or 10 8 In a specific example, the composition may be prepared so that at least 10 cells / kg are administered to a subject. 4 , 10 5 , 10 6 , or 10 7 The modified NK cells or T cells comprise CD147-CAR-NK cells or CD147-CAR-T cells. The population of modified NK cells or modified T cells is typically administered parenterally, for example, intravenously, although injection or infusion into or near a tumor (local administration) or intraperitoneal administration can also be used. The appropriate route of administration can be determined based on factors such as the subject, the condition being treated, and other factors.

[0087] Multiple doses of the population of modified NK cells or modified T cells can be administered to the subject.For example, CD147-CAR-NK cells or CD147-CAR-T cells can be administered daily, every other day, twice a week, weekly, every other week, every three weeks, monthly, or less frequently.A skilled clinician can select the administration schedule based on the subject, the condition to be treated, previous treatment history, and other factors.

[0088] In additional examples, the subject is also administered at least one, at least one, at least two, at least three, or at least four cytokines (e.g., IL-2, IL-15, IL-21, and / or IL-12) to support the survival and / or growth of the engineered NK cells or engineered T cells. In a specific, non-limiting example, the at least one cytokine includes IL-2 and IL-15. The cytokine(s) are administered before, after, or substantially simultaneously with the engineered NK cells or engineered T cells. In a specific example, the at least one cytokine (e.g., IL-2) is administered simultaneously with, for example, CD147-CAR-NK cells.

[0089] In some cases, the subject to be treated has a solid tumor, for example, a solid tumor that expresses CD147. Examples of solid tumors include sarcoma (e.g., fibrosarcoma, myxosarcoma, liposarcoma, chondrosarcoma, osteosarcoma, soft tissue sarcoma, and other sarcomas), synovioma, mesothelioma, Ewing's sarcoma, leiomyosarcoma, rhabdomyosarcoma, colon cancer, colorectal cancer, peritoneal cancer, esophageal cancer (e.g., esophageal squamous cell carcinoma), pancreatic cancer, breast cancer (including basal breast cancer, ductal carcinoma, and lobular breast cancer), endometrial cancer, lung cancer (e.g., non-small cell lung cancer), ovarian cancer, prostate cancer, liver cancer (including hepatocellular carcinoma), gastric cancer, squamous cell carcinoma (head and neck squamous cell carcinoma), and the like. Solid tumors include basal cell carcinoma, adenocarcinoma, sweat gland carcinoma, medullary thyroid carcinoma, papillary thyroid carcinoma, pheochromocytoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinoma, medullary carcinoma, bronchogenic carcinoma, liver cancer, bile duct carcinoma, choriocarcinoma, Wilms' tumor, cervical cancer, fallopian tube carcinoma, testicular tumor, seminoma, bladder cancer (e.g., renal cell carcinoma), melanoma, and CNS tumors (e.g., glioma, glioblastoma, astrocytoma, medulloblastoma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, oligodendroglioma, meningioma, neuroblastoma, and retinoblastoma). Solid tumors also include tumor metastases (e.g., to the lung, liver, brain, or bone). In some examples, the subject has hepatocellular carcinoma, neuroblastoma, breast cancer, gastric cancer, endometrial cancer, bladder cancer (e.g., renal cell carcinoma), lung cancer (e.g., non-small cell lung cancer), cervical cancer, medulloblastoma, esophageal cancer (e.g., esophageal squamous cell carcinoma), prostate cancer, seminoma, glioblastoma, osteosarcoma, astrocytoma, or soft tissue sarcoma. In particular examples, the subject has hepatocellular carcinoma or neuroblastoma.

[0090] In other examples, the subject has a hematological malignancy, eg, a hematological malignancy that expresses CD147. Examples of hematological malignancies include leukemias, including acute leukemia (e.g., 11q23-positive acute leukemia, acute lymphocytic leukemia (ALL), T-cell ALL, acute myelocytic leukemia, acute myeloid leukemia (AML), and myeloblastic, promyelocytic, myelomonocytic, monocytic, and erythroleukemia), chronic leukemia (e.g., chronic myelocytic (granulocytic) leukemia, chronic myeloid leukemia, and chronic lymphocytic leukemia), lymphoblastic leukemia, polycythemia vera, lymphoma, diffuse large B-cell lymphoma, Burkitt's lymphoma, T-cell lymphoma, follicular lymphoma, mantle cell lymphoma, Hodgkin's disease, non-Hodgkin's lymphoma, multiple myeloma, Waldenstrom's macroglobulinemia, heavy chain disease, myelodysplastic syndrome, hairy cell leukemia, and myelodysplasia. In particular examples, the subject has acute lymphocytic leukemia (ALL), T-cell ALL, acute myeloid leukemia, or acute myeloid leukemia (AML).

[0091] In some examples, the subject is also treated with one or more of surgery, radiation therapy, and a chemotherapeutic agent. Exemplary chemotherapeutic agents include (but are not limited to) alkylating agents such as nitrogen mustards (e.g., mechlorethamine, cyclophosphamide, melphalan, uracil mustard, or chlorambucil), alkylsulfonates (e.g., busulfan), nitrosoureas (e.g., carmustine, lomustine, semustine, streptozocin, or dacarbazine); folic acid analogs (e.g., methotrexate), pyrimidine analogs (e.g., 5-FU or cytarabine), and and purine analogs such as mercaptopurine or thioguanine; or natural products, such as vinca alkaloids (e.g., vinblastine, vincristine, or vindesine), epipodophyllotoxins (e.g., etoposide or teniposide), antibiotics (e.g., dactinomycin, daunorubicin, doxorubicin, bleomycin, plicamycin, or mitomycin C), and enzymes (e.g., L-asparaginase). Additional agents include platinum coordination complexes (e.g., cis-diamine-dichloroplatinum II, also known as cisplatin), substituted ureas (e.g., hydroxyurea), methylhydrazine derivatives (e.g., procarbazine), and adrenocrotical suppressants (e.g., mitotane and aminoglutethimide); hormones and antagonists, such as corticosteroids (e.g., prednisone), progestins (e.g., hydroxyprogesterone caproate, medroxyprogesterone acetate, and megestrol acetate), estrogens (e.g., diethylstilbestrol and ethinyl estradiol), antiestrogens (e.g., tamoxifen), and androgens (e.g., testosterone propionate and fluoxymesterone).Examples of the most commonly used chemotherapy drugs include adriamycin, melphalan (Alkeran®), Ara-C (cytarabine), carmustine, busulfan, lomustine, carboplatinum, cisplatinum, cyclophosphamide (Cytoxan®), daunorubicin, dacarbazine, 5-fluorouracil, fludarabine, hydroxyurea, idarubicin, ifosfamide, methotrexate, mithramycin, mitomycin, mitoxantrone, nitrogen mustard, and paclitaxel. (or other taxanes such as docetaxel), vinblastine, vincristine, VP-16, while newer drugs include gemcitabine (Gemzar®), trastuzumab (Herceptin®), irinotecan (CPT-11), leustatin, navelbine, rituximab (Rituxan®), imatinib (STI-571), topotecan (Hycamtin®), capecitabine, ibritumomab (Zevalin®), and calcitriol. A skilled clinician will be able to select appropriate additional therapy for a subject (from the therapies listed herein or other current therapies) depending on factors such as the subject, the cancer being treated, treatment history, and other factors. [Example]

[0092] The following examples are provided to illustrate certain particular features and / or embodiments, and should not be construed as limiting the disclosure to the particular features or embodiments described.

[0093] Example 1 material and method Antibodies and reagents: purified anti-CD247 (also known as T cell surface glycoprotein CD3 zeta chain, CD3) antibody (clone 6B10.2, BioLegend), purified anti-human CD147, FITC-conjugated anti-human CD147 (clone HIM6, BioLegend), PE- or APC-conjugated anti-human CD3 antibody (clone OKT3, BioLegend), FITC- or BV510-conjugated anti-human CD5 6 antibody (clone HCD56, BioLegend), PE-conjugated anti-human CD69 antibody (clone FN50, BioLegend), APC / Fire750-conjugated anti-human CD226 antibody (also known as DNAM-1, clone 11A8, BioLegend), APC / Fire750-conjugated anti-human KLRG1 (MAFA) antibody (clone SA231A2, BioLegend), BV421-conjugated Anti-human CD335 (NKp46) antibody (clone 9E2, BioLegend), PE / Cy7-conjugated anti-human CD158b (KIR2DL2 / L3, BioLegend) antibody (clone DX27, BioLegend), PE / Cy7-conjugated anti-human CD244 (2B4) antibody (clone C1.7, BioLegend), PE-conjugated anti-human CD152 (CTLA-4) antibody (clone BNI3), APC Conjugated anti-human CD366 (Tim-3) antibody (clone F38-2E2), PerCP / Cy5.5 anti-human TIGIT (VSTM3) antibody (clone A15153G), FITC-conjugated anti-human CD223 (LAG-3) antibody (clone 11C3C65, BioLegend), and PerCP / Cy5.5-conjugated anti-human CD94 (clone DX22, BioLegend) were purchased from BioLegend (San Diego, CA, USA).

[0094] APC-conjugated anti-human CD16 antibody (clone B73.1, BD Biosciences), FITC-conjugated anti-human CD3 antibody (clone UCHT1, BD Biosciences), BV480-conjugated anti-human CD85j (LIR-1) antibody (clone GHI / 75, BD Biosciences), BV711-conjugated anti-human CD314 (NKG2D) antibody (clone 1D11, BD Biosciences), and PE- or FITC-conjugated anti-human CD107a antibody (clone H4A3, BD Biosciences) were purchased from BD Biosciences (San Jose, CA, USA).

[0095] FITC-conjugated anti-human KIR / CD158 antibody (clone 180704, R&D Systems), PE-conjugated anti-human KIR2DL1 / KIR2DS5 antibody (clone 143211, R&D Systems), APC-conjugated anti-human KIR3DL1 antibody (clone DX9, R&D Systems), AF405-conjugated anti-human KIR3DL2 / CD158k antibody (clone 539304, R&D Systems), APC-conjugated anti-human NKG2A / CD159a antibody (clone 131411, R&D Systems), and PE-conjugated anti-human NKG2C / CD159c antibody (clone 134591, R&D Systems) were purchased from R&D Systems. AF647 goat anti-human IgG F(ab') 2 Fragment antibodies were purchased from Jackson ImmunoResearch (West Grove, PA, USA).

[0096] Bioinformatic analysis from public cancer patient databases: Patient survival data and RSEM (RNA-Seq by Expectation Maximization) normalized expression datasets for CD147 were generated from The Cancer Genome Atlas (TCGA) and downloaded from OncoLnc (oncolnc.org). Data were plotted as Kaplan-Meier curves using GraphPad Prism 5.0 (GraphPad). The TCGA-derived RSEM normalized expression dataset was from FireHose Broad GDAC, developed by The Broad Institute (gdac.broadinstitute.org). Figures were generated using GraphPad Prism 5.

[0097] Cell lines: 293T, K562, Daudi, SK-Hep1, and HepG2 cell lines were purchased from the American Type Culture Collection (ATCC). To establish Daudi-FFluc cells, CD147-positive HepG2 and SK-Hep1 cells were transduced with a lentiviral vector encoding FFLuc-GFP, as previously described. The protocol for collection of peripheral blood from healthy donors was approved by the Institutional Review Board (IRB) and Ethics Review Board of Rutgers-New Jersey Medical School (Newark, NJ).

[0098] NK-92MI cell culture and generation of CAR-modified NK-92MI cells: The NK-92MI cell line was purchased from ATCC® (CRL-2408™, USA). NK-92MI, an interleukin-2 (IL-2)-independent NK cell line, is derived from the NK-92 (ATCC® CRL-2407™) cell line (Gong et al., Leukemia 8:652-658, 1994) stably expressing human IL-2 cDNA (Tam et al., Hum Gene Ther 10:1359-1373, 1999; Tam et al., J. Hematother. 8:281-290, 1999). The NK92-MI cell line was maintained in a specific NK-92MI culture medium (alpha minimum essential medium, alphaMEM) lacking ribonucleosides and deoxyribonucleosides but containing 2 mM L-glutamine and 1.5 g / L sodium bicarbonate. To prepare a complete growth medium, the following components were added to the basal medium: 0.2 mM inositol, 0.1 mM 2-mercaptoethanol, 0.02 mM folic acid, 12.5% ​​horse serum, and 12.5% ​​fetal bovine serum. On day 3, the retroviral supernatant was transduced into NK-92MI cells in plates coated with a recombinant fibronectin fragment (FN CH-296; Retronectin; Takara, Japan). After transduction, NK cells were expanded using IL-2. To confirm the percentage of CD147-CAR expression on NK-92MI cells, these cells were stained for CD3 and CD56 to stain NK cells, followed by flow cytometry analysis.

[0099] Generation of CD147 knockout cell lines: To generate CD147 knockout hepatocyte cell lines, we used a lentiviral delivery system and guide RNAs targeting the CD147 sequence: #1 (5-TTGACATCGTTGGCCACCGC-3; SEQ ID NO: 10) and #3 (5-GTGGACGCAGATGACCGCTC-3; SEQ ID NO: 11). Lentivirus was produced in HEK293T cells by transfecting lenti-CRISPR v2 with the packaging plasmids pSPAX2 and pMD2G. After 3 days, the supernatant was filtered (0.45 μm) and incubated with the hepatocyte cell line and 8 μg / mL polybrene (Sigma). After 48 hours of incubation, the transduced cells were replaced with fresh medium and selected with 8.0 μg / mL puromycin for 5 days. Western blot and flow cytometry analyses were performed to confirm the efficacy of the knockout cell lines.

[0100] Plasmid construction and retrovirus production: A codon-optimized DNA fragment encoding the CD147-specific scFv from the 5F6 clone was synthesized by GENEWIZ and subcloned in frame into the SFG retroviral vector backbone, which contains the hinge component of human IgG1, the CD28 transmembrane domain, the intracellular domain of CD28 plus 4-1BB, and the ζ chain of the human TCR / CD3 complex. To produce the CD147-CAR retrovirus, 293T cells were transfected with a combination of plasmids containing the CD147-specific scFv, RDF, and PegPam3, as previously described (Loskog et al., Leukemia 20:1819-1828, 2006). The constructs of CD19-CD28-CAR and CD19-4-1BB-CAR were previously described (Xiong et al., Mol. Ther. 26:963-975, 2018).

[0101] Transduction of NK-92MI cells with CD147-CAR: NK cells were harvested on day 7 of expansion and transduced with CD147-CAR retrovirus in plates coated with Retronectin (FN CH-296, Takara, Japan). Two days later, cells were transferred to G-Rex6 multiwell cell culture plates and maintained in 35 ml of complete RPMI-1640 medium containing 200 U / ml IL-2 (PeproTech). The medium was changed every 3–4 days, and 2 × 10 cells were added to each well each time for continued culture. 7 Cells were retained. Total cell numbers were counted using Trypan Blue exclusion. To confirm the percentage of NK cells and CAR expression, cells were stained for CD3, CD56, and IgG F(ab')2 and analyzed by flow cytometry.

[0102] Flow cytometry analysis and sorting: CAR-NK cells were stained with fluorescently conjugated antibodies in FACS staining buffer containing 1% fetal bovine serum (FBS) for 30 minutes on ice, washed with PBS, and analyzed on a FACS LSRII or LSR Fortessa flow cytometer (BD). PMT voltages were adjusted and compensation values ​​were calculated before data collection. Data were acquired using FACS Diva software (BD) and analyzed using FlowJo software (Tree Star).

[0103] For flow cytometry single live cell sorting, all sample cells were stained with fluorescently conjugated antibodies for 30 minutes on ice (RPMI 1640 containing 1% FBS), washed twice with PBS, resuspended in complete culture medium, and sorted using a SORP BD FACS Aria III. After sorting, the collected samples were washed once with prewarmed culture medium and cultured for use.

[0104] CAR-NK degranulation assay (CD107a): CAR-NK cells (1 × 10) were cultured in complete NK-92MI culture medium in U-bottom 96-well plates at 37 °C for 4 h or overnight. 5) was incubated with target cells. Cells were then harvested, washed, stained for CD3, CD56, and CD107a using GolgiStop (BD) for 30 minutes on ice, and analyzed by flow cytometry.

[0105] Cytokine release assay: IFN-gamma and TNF-alpha cytokines secreted by CAR-NK were measured by commercially available ELISA kits (Invitrogen-Thermo Fisher Scientific) according to the manufacturer's protocol.

[0106] 51 Cr release assay: A standard 4-hour assay was performed to assess the cytotoxic activity of CAR-NK cells. 51 A Cr release assay was used. Briefly, target cells were incubated at 37°C for 2 hours. 51 Cr, and then cultured at 2 × 10 5 / mL. Then 2 × 10 4 Target cells were incubated with serially diluted CAR-NK cells for 4 hours at 37°C. After centrifugation, the supernatant was collected and the released 51 Cr was measured in a gamma counter (Wallac, Turku, Finland). Cytotoxicity (as a percentage) was calculated as follows: [(sample - spontaneous release) / (maximum release - spontaneous release)] × 100.

[0107] FFLuc reporter system assay: To quantify the cytotoxicity of CAR-modified immune cells, an FFLuc reporter system assay was developed. Briefly, on day 1, 2 x 10 target cells were plated in 100 μl / well of complete target cell nutrient medium on an optical 96-well plate (Greiner Bio-One™ No. 655098). 4 or 3×10 4Plates were pre-seeded with 10 target cells / well (cells stably transduced with FFluc-GFP) and incubated overnight at 37°C in 5% CO2. The next day, serial dilutions of effector cells were prepared according to the effector / target ratio, and the indicated effector cells were added to each well (100 μl / well). The reactions were incubated at 37°C with 5% CO2 for 4 hours, and then the supernatant was gently discarded. 100 μl of working D-luciferin was added to each well and incubated for 5 minutes at 37°C with 5% CO2 in the dark. Data were quantified using a microplate reader (PerkinElmer, USA). Data were quantified by converting the values ​​obtained to percentage of specific lysis (%) by the following equation: percentage of specific lysis (%) = [1-(SE) / (TM)] x 100, where S is the luminescence value of the sample well, E is the luminescence value of the "effector cell only" wells compared to the sample wells, T is the mean luminescence of the "target cell only" wells, and M is the mean luminescence of the "blank medium only" wells.

[0108] Animal studies: All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC). All in vivo experiments were performed using NOD.Cg-Prkdc mice from The Jackson Laboratory (Bar Harbor, ME). scid Il2rg tm1Wjl To establish the hepatocellular carcinoma cell line xenograft model, both male and female NSG mice (8 weeks old) were implanted with 4 x 10 cells in 100 μL of PBS Corning® Matrigel® Matrix into the right flank. 6 SK-Hep1 cells were injected subcutaneously. The tumor burden was approximately 40–50 mm 2 Once the mice reached 100 μL, they were randomly assigned to three groups. Treatment began on day 1, with mice receiving 5 × 10 6CD147-CAR-NK-92MI cells were injected (iv). The control group was injected with parental NK-92MI cells or vehicle (PBS). The following day (day 2), all animals were injected (iv) with IL-2 (20,000 units / mouse). Animal weights and tumor burden were collected twice weekly. Tumor size was measured with a caliper, and the maximum longitudinal diameter (length) and maximum transverse diameter (width) were recorded. Tumor size based on caliper measurements was calculated by a modified ellipsoid formula. Tumor size was calculated as follows: tumor size (mm 2 )=1 / 2(length x width 2 ) Tumor burden is 2000mm 2 Mice were euthanized according to IACUC guidelines if they exceeded 100 mg / kg or lost more than 20% of their body weight. Survival data was recorded.

[0109] A patient-derived xenograft (PDX) model, an animal model of hepatocellular carcinoma (HCC), was developed by The Jackson Laboratory. Briefly, fresh PDX specimens were implanted subcutaneously into the flanks of 6-8 week-old NOD SCID gamma (NSG) mice. The tumor burden was approximately 40-50 mm. 2 After reaching a maturity of 100 mg / kg, mice were randomly assigned to three groups for further analysis. The main treatment procedures used were as described above. Xenograft specimens were fixed in 10% formalin, embedded in paraffin for sectioning, and processed for IHC staining or directly frozen in liquid nitrogen for further analysis.

[0110] Statistical Analysis: Statistical analysis of tumor size was performed by two-way ANOVA with Bonferroni post-hoc test. Overall survival statistics were calculated using the log-rank test. Other statistical significance was determined using two-tailed unpaired Student's t-test and two-tailed paired Student's t-test. All statistical calculations and graphs were generated by GraphPad Prism 5.0. P<0.05( * ), P<0.01( ** ), and P<0.001( *** ) were considered statistically significant.

[0111] Example 2 CD147 is expressed in hepatocellular carcinoma cell lines and patient-derived liver cancer specimens To determine whether CD147 is a valid and valid target for hepatocellular carcinoma and other types of cancer, we analyzed the correlation between patient survival and CD147 expression levels from the TCGA (The Cancer Genome Atlas, cancergenome.nih.gov) dataset. Two different patient subsets (CD147 high and CD147 low Comparison of survival percentages from tumor tissues showed a strong negative correlation between CD147 and survival percentage (Figure 1A). Specifically, CD147 expression in multiple tumor tissues high showed a low survival percentage (Figure 1A). Furthermore, comparison of CD147 expression between normal tissues (NT) and tumor samples (TP) in multiple cancer types showed a significant upregulation of CD147 expression in different types of tumor tissues (Figure 1B).

[0112] To verify the data from the bioinformatic analysis, we analyzed CD147 expression by Western blot in various tumor cell lines and other tissues, including wild-type NK-92 (a human NK cell line), T2 (a mutant T × B cell hybrid), 721.221 (an HLA-A, -B, -C null human cell line), MDA-MB-231 (a human breast cancer cell line), K562 (a human myeloid leukemia cell line), HepG2 (a human hepatocellular carcinoma cell line), SK-Hep1 (a human liver adenocarcinoma cell line), Raji (a human B-lymphocyte Burkitt's lymphoma cell line), Daudi (a human B-lymphoblastoid cell line), NK-92-MI (an interleukin-2-independent natural killer cell line), and human peripheral blood monocytes (PBMCs). The CD147 molecule was highly upregulated in HepG2 and SK-Hep1 cell lines (two hepatocellular carcinoma cell lines) compared with PBMCs (Figure 1C). CD147 expression in PBMCs was relatively low compared with tumor cell lines (Figure 1C). Similar results were obtained by flow cytometry analysis (Figure 1D). Furthermore, immunohistochemistry (IHC) assay results confirmed that CD147 was significantly upregulated in HCC tissues isolated from PDX mouse models (Figure 1E).

[0113] Example 3 Generation and characterization of CD147-CAR-NK cells CD147-CAR was constructed using the SFG vector (Loskog et al., Leukemia 20:1819-1828, 2006; Xiong et al., Mol. Ther. 26:963-975, 2018). The CD147-CAR contained an engineered single-chain variable fragment (scFv) of an anti-CD147 antibody (derived from clone 5F6 as described in Example 1), an IgG-CH2CH3 spacer, the transmembrane domain of CD28, the intracellular domain of CD28-4-1BB, and the intracellular signaling domain of the TCR-zeta chain (Figure 2A).

[0114] This CAR construct was first tested in the NK-92MI cell line. After transduction, NK-92MI expressed the CD147-CAR molecule (Figure 2B). After flow cytometric sorting, the percentage of CD147-positive NK-92MI cells was over 96% (Figure 2B). The expression of the CD147-CAR molecule in NK-92MI cells was further verified by Western blot. Compared to the parental NK-92MI cell line, CD147-CAR-NK-92MI expressed the chimeric scFv-CD147-CAR. The approximate molecular weight was approximately 80-85 kD (Figure 2C). The CD147-CAR-NK-92MI cell line was further characterized by flow cytometry. Comparable expression of activating receptors (e.g., CD56, NKG2D, NKP46, NKG2A, CD16, CD94 / NKG2C, CD226 (also known as DNAM-1), and CD244 (also known as 2B4)) and inhibitory receptors (e.g., KLRG1, LAG-3, CTLA-4, TIM-3, PD-1, and TIGIT) was observed (Figure 2D). Given the low level of CD147 expression in NK92-MI cells, CD147 expression in the CD147-CAR-NK-92MI cell line was also analyzed by flow cytometry. Expression of CD147-CAR in NK-92MI was stable for more than 30 days post-transduction. However, CD147-CAR expression was accompanied by loss of CD147 in the NK-92MI cell line, demonstrating limited fratricide in CD147-CAR-NK-92MI cells (Figures 2E-2G). Notably, loss of CD147 molecule expression in CD147-CAR-NK-92MI cells did not affect their functionality and expression in vitro.

[0115] Example 4 CD147-CAR NK cells specifically kill hepatocellular carcinoma (HCC) in vitro After successful establishment of CD147-CAR-NK cells, CD147 +We tested the ability of CD147-CAR-NK-92MI cells to eradicate HCC cell lines (including SK-Hep1 and HepG2 cells). Compared with control CD19-CD28-CAR-NK-92MI and CD19-4-1BB-CAR-NK-92MI cells, CD147-CAR-NK-92MI cells exhibited significant cytotoxicity against two HCC cell lines, SK-Hep1 and HepG2 (Figures 3A and 3B), as well as Huh5 and HCO2 cell lines (Figures 4A–4C). Furthermore, the production of both TNF-alpha and IFN-gamma by CD147-CAR-NK-92MI cells was significantly higher than that of CD19-CD28-CAR-NK-92MI and CD19-4-1BB-CAR-NK-92MI cells stimulated with SK-Hep1 and HepG2 cells (Figures 3C and 3D). Interestingly, activation of CD147-CAR-NK-92MI cells by sensitive target cells could be blocked by an anti-CD147 antibody (clone HIM6), but not by a control IgG1 (Figures 5A-5C). The specificity of this anti-CD147 antibody was further verified by testing its effect on the cytotoxicity of CD19-4-1BB-CAR-NK-92MI cells. This anti-CD147 antibody was unable to block the cytotoxicity of CD19-4-1BB-CAR-NK-92MI cells against the CD19-positive Daudi cell line (Figure 6), demonstrating the selectivity of CD19-4-1BB-CAR-NK-92MI cells and the anti-CD147 antibody.

[0116] To further confirm the specificity of CD147-CAR-NK-92MI cells, CD147 knockout (CD147 - / - )SK-Hep1 cell line (CD147 - / - -SK-Hep1) and CD147 knockout (CD147 - / - ) HepG2 cell line (CD147 - / - -HepG2) were generated. - / - -HepG2 and CD147 - / --SK-Hep1 cells were not recognized by CD147-CAR-NK-92MI cells (Figures 5D-5G), as quantified by surface expression of CD107a when co-cultured with the CD147-KO cell line.

[0117] Both the CD107a assay and cytokine production assay can be used to assess the activation of CD147-CAR-NK-92MI cells by sensitive target cells. To directly test whether CD147-CAR-NK-92MI cells can kill CD147-positive HCC cells, we used standard chromium-51 ( 51 A Cr) release assay (the gold standard assay for assessing the cytotoxicity of CTLs and NK cells in the field of immunology) was used. CD147-CAR-NK-92MI cells killed SK-Hep1 and Daudi cells. Similar killing activity was observed by CD147-CAR-NK-92MI cells against additional HCC cell lines, such as the Huh7 cell line (Kasai et al., Hum Cell 31:261-267, 2018) and the HCO2 cell line (Trinh et al., PLoS One 10:e0136673, 2015).

[0118] To further verify the killing activity of CD147-CAR-NK-92MI cells, a novel, easy-to-use, non-radioactive approach was developed to evaluate the cytotoxicity of CD147-CAR-NK-92MI cells using luciferase bioluminescence signals. First, FFLuc-EGFP-SK-Hep1 and FFLuc-EGFP-HepG2 cell lines were generated. To evaluate the direct killing of target cells, CD147-CAR-NK-92MI cells were co-cultured with FFLuc-EGFP-SK-Hep1 and FFLuc-EGFP-HepG2 cell lines, respectively. After 4 hours of incubation of CD147-CAR-NK-92MI cells in a 96-well optical-bottom microplate pre-seeded with target cells stably expressing an EGFP-firefly luciferase fusion gene (EGFP-FFluc), the chemobioluminescence signal of EGFP-FFluc was quantified using a fluorescence microplate reader. As described in Example 1, 51 Similar to the Cr release assay, the FFLuc signal was converted to the percentage of specific lysis (Figures 8H and 8I).

[0119] First, we demonstrated that CD147-CAR-modified primary T and NK cells isolated from human peripheral blood mononuclear cells (PBMCs) can eradicate multiple HCC cell lines (including SK-Hep1, Huh7, and HepG2), whereas kappa-CAR-modified T cells cannot (Figures 8A-8C). 51 Cr release assays demonstrated that CD147-CAR-modified human primary NK cells effectively killed HCC cell lines in vitro (Figures 8D-8F). To further demonstrate the natural killing ability of primary NK cells through NKGD / NKG2DL interactions in addition to CD147-CAR primary NK cytotoxicity, we demonstrated that anti-NKG2D inhibited the CD147-CAR-NK-mediated CD147 activation. - / - We found that α-glucan further blocked the killing of SK-Hep1 cells (Fig. 8G).

[0120] Dose-dependent specific lysis 51The cytotoxic activity measured by this approach was comparable to that measured by the Cr release assay. The cytotoxic activity measured by this approach was further quantified under a conventional inverted fluorescence microscope to evaluate the morphology and dynamics of the EGFP signal in target cells (Figures 7A and 7B). Thus, these two complementary approaches demonstrated that CD147-CAR NK cells specifically killed hepatocellular carcinoma (HCC) in vitro.

[0121] CD147-CAR-NK-92MI cells significantly increased CD147 expression compared to parental SK-Hep1 and HepG2 cells. - / - -SK-Hep1 and CD147 - / - CD147-CAR-NK-92MI cells were unable to kill HepG2 cells (Figures 8K and 8L). The specificity of the cytotoxicity of CD147-CAR-NK-92MI cells was further verified by adding anti-CD147 antibody to the coculture of effector and target cells (Figures 8M and 8N). To further validate CD147 as an effective and valid target for HCC, we tested the cytotoxicity of CD147-CAR-T cells against two different HCC cell lines, HepG2 (Figure 10A) and SK-Hep1 (Figure 10B). Deletion of the CD147 molecule in the SK-Hep1 cell line (CD147 knockout SK-Hep1 cell line) significantly reduced the specific lysis of CD147-CAR-T cells (Figure 10C), further demonstrating the specificity of the cytotoxicity of CD147-CAR-T cells against CD147-positive HCC cell lines.

[0122] Example 5 CD147-CAR-NK cells control HCC progression in vivo To evaluate whether CD147-CAR can kill HCC in vivo, two different xenograft models were used. First, CD147-CAR-modified primary T and NK cells derived from PBMCs were evaluated in an SK-Hep1 xenograft mouse model. CD147-CAR-modified primary T cells significantly suppressed tumor size and prolonged survival (Figures 11A-11D). To further evaluate the efficacy of CD147-CAR-modified primary NK cells, we included a non-transduced (NT) primary NK group as an additional control (Figures 12A-12D). Mice receiving the parental NT-NK control group and the PBS vehicle control group developed rapid disease progression. In contrast, mice receiving CD147-CAR primary T and NK cells were significantly protected from rapid progression, their median survival was prolonged (P<0.05), and body weight was similar among the different groups (Figures 11C and 12C), indicating tolerable toxicity of CD147-CAR-modified primary T and NK cells in vivo.

[0123] Furthermore, to further develop additional "off-the-shelf" cell therapy strategies, we evaluated the efficacy of CD147-CAR-modified NK-92MI cells. Due to the malignant nature of NK-92MI, CAR-modified NK-92MI cells must be irradiated before administration to patients. The cytotoxicity of non-irradiated and irradiated CD147-CAR-NK-92MI cells was assessed using a standard 4-hour immunotherapy. 51 The results were compared by Cr release assay (Figure 13). Similar cytotoxicity was observed between non-irradiated and irradiated CD147-CAR-NK-92MI cells in vitro (Figure 13).

[0124] We further compared the efficacy between non-irradiated and irradiated CD147-CAR-NK-92MI cells in a xenograft NSG mouse model (Figures 14A-14D). Similar efficacy was observed in vivo between mice injected with non-irradiated and irradiated CD147-CAR-NK-92MI, as measured by median survival (Figures 14A-14D).

[0125] To further evaluate the efficacy of CD147-CAR-NK-92MI cells (injected on days 1, 3, and 5 after tumor implantation) in controlling tumor growth, disease progression was measured by tumor size (Figure 15A). Mice receiving the parental NK-92MI and PBS vehicle control groups developed rapid disease progression (Figure 15B). In contrast, mice receiving CD147-CAR-NK-92MI cells were significantly protected from rapid progression, prolonging their median survival (P<0.01), and body weights were comparable between the different groups (Figures 15C and 15D), indicating tolerable toxicity of CD147-CAR-NK-92MI cells in vivo.

[0126] While cancer cell lines can have significant limitations in their ability to accurately model biology and therapeutic efficacy, patient-derived xenograft (PDX) models are biologically stable and can mimic human clinical conditions with regard to mutation status, gene expression patterns, and tumor heterogeneity. Therefore, we employed a second xenograft mouse model using metastatic liver cancer tissue from patients. We tested the efficacy of CD147-CAR-NK-92MI cells administered on days 0, 4, 8, 11, 15, 22, 25, and 35 after engraftment. The median survival time of mice treated with CD147-CAR-NK-92MI cells was 63 days, significantly longer than that of control mice (median survival time of approximately 42 days). A reduction in tumor burden and disease progression was observed in mice treated with CD147-CAR-NK-92MI cells (Figures 16A-16D), demonstrating the efficacy of CD147-CAR-NK-92MI cells in suppressing liver cancer progression in a PDX mouse model.

[0127] Example 6 CD147-CAR-T cells specifically kill CD147-positive tumor cells The ability of CD147-CAR-T cells to kill HCC cells was tested against eight HCC cell lines (Huh7, Huh7.5, HepG2, SK-Hep1, Hep3B, Hu1545, HCO2, and LH86). Peripheral blood mononuclear cells (PBMCs) were obtained from healthy donors. To transduce PBMCs, cells were activated with 1 μg / ml anti-CD3 (clone OKT3, Ortho Biotech, Bridgewater, NJ, USA) and 1 μg / ml anti-CD28 in 10% FBS RPMI-1640 medium containing 100 U / ml recombinant human IL-2 (Proleukin; Chiron, Emeryville, CA, USA). To generate CD147-CAR-T cells, activated T cells were transduced with retroviral supernatant in plates coated with recombinant fibronectin fragment (FN CH-296; Retronectin; Takara Shuzo, Otsu, Japan) on day 3. After transduction, T cells were expanded using IL-2 and then used in the assay.

[0128] The FFLuc reporter system assay described in Example 1 was used to test the specific killing of FFluc-EGFP-HepG2 by CD147-CAR-T cells (Figure 17A). 1 x 10 cells per well were cultured in 96-well optical bottom microplates for 6 hours. 4 Effector cells (CD147-CAR-T cells) were co-cultured with target cells (1 × 10 FFLuc-GFP-HepG2 cells). For the control group, wild-type kappa-CAR-T cells were incubated with CD147-positive FFLuc-EGFP-HepG2 cells. FFLuc reporting system assay demonstrated reduced cytotoxicity of CD147-CAR-T cells using knockout CD147 FFLuc-GFP-HepG1 cells (Figure 17B). Effector cells (CD147-CAR-T and Kappa-CAR-T cells) were co-cultured with target cells (1 × 10 FFLuc-EGFP-HepG2 cells) for 6 hours in a 96-well optical-bottom microplate. 4The cytotoxicity of CD147-CAR-T cells was measured by reading the luminescence signal using a microplate reader. The cytotoxicity of CD147-CAR-T cells was measured using a standard 4-hour 51 This was measured by a Cr release assay (Figure 17C). Kappa-CAR-T cells were used as a negative control.

[0129] Example 7 Effect of CD147-CAR-T cells on neuroblastoma cells Degranulation of CD107a in CD147-CAR-T cells was observed after 10 hours in medium (control) or SK-N-SH tumor cells (Figure 18A), as well as DaoY and D283 cells. The effector-to-target ratio was 1:1.2. To quantify surface CD107a expression, cells were gated on the CD56-positive subset. Quantitative data of the percentage of surface CD107a expression on CD147-CAR-NK-92MI cells upon various stimuli, as indicated (Figure 18B). CD147-CAR-NK-92MI cells were also cytotoxic to DaoY cells after 3 hours (Figure 19).

[0130] Example 8 Primary CD147-CAR-NK cells derived from HCC patients specifically kill CD147-positive tumor cells in vitro The broad expression pattern of CD147 across multiple solid tumor types makes it an attractive target for CD147-CAR-based cancer immunotherapy. Here, we investigated whether CD147 is upregulated in human HCC tissue samples. HCC tumor tissues at various stages were strongly positively stained for CD147 compared with healthy liver tissue (Figure 21A).

[0131] To evaluate whether CD147-CAR-modified primary NK cells isolated directly from HCC livers could kill HCC in vitro, NK cells were isolated from various zones of liver tissue, including tumor zones, tumor-adjacent zones, and non-tumor zones in human livers with HCC (Figure 21B). These NK cells were then expanded (Figure 21C). These expanded NK cells from human HCC liver tissue were used to generate CD147-CARs. The transduction efficiency of activated NK cells was generally greater than 70% (Figure 21D). CD147-CAR-NK cells specifically recognized tumor cells expressing CD147. The antitumor activity of CD147-CAR-NK was evaluated against HCC cell lines (Figure 21E). Together, these findings demonstrate that CD147-CAR-redirected primary human liver NK cells specifically and selectively kill CD147-positive target cells.

[0132] Example 9 SynNotch GPC3-induced CD147-CAR T cells selectively target GPC3+CD147+ HCC cells but not GPC3+CD147- or GCP3-CD147+ HCC cells To generate the anti-GPC3 synNotch-inducing receptor vector, an anti-GPC3 (mouse GC33 clone) scFv capable of specifically binding to the human GPC3 antigen was synthesized by GENEWIZ. The sequence encoding the N-terminal signal peptide and myc tag was fused to the synNotch-Gal4VP64-inducible element derived from (Addgene plasmid #79125) by overlap PCR. The fragment was inserted into the SFG gamma retroviral vector digested with the restriction endonucleases NcoI and XhoI.

[0133] For construction of the anti-CD147-CAR-mCherry vector, the entire CD147-CAR element was inserted into pHR_Gal4UAS_pGK_mCherry (Addgene plasmid #79124) digested with restriction endonucleases MluI and NdeI. Expression of the mCherry gene was under the control of the pGK promoter. In this strategy, eGFP and mCherry double-positive cells were gated as synNotch CAR-modified cells for further analysis and functional evaluation.

[0134] To generate the anti-CD147 synNotch-inducing receptor vector, the anti-CD147 scFv was fused with the synNotch-Gal4VP64-inducing element from pHR_PGK_anti-CD19_synNotch_Gal4VP64 (Addgene plasmid #79125) by overlap PCR. A myc tag was added to the N-terminus. The fragment was inserted into the SFG gamma retroviral vector after digestion of the signal peptide with restriction endonucleases SalI and MluI.

[0135] To construct the anti-GPC3-CAR-mCherry vector, the entire GPC3-CAR element containing the anti-GPC3 (mouse GC33 clone) scFv was inserted into pHR_Gal4UAS_pGK_mCherry (Addgene plasmid #79124) digested with restriction endonucleases MluI and NdeI. Expression of the mCherry gene was under the control of the pGK promoter. In this strategy, myc-tagged and mCherry double-positive cells were gated as synNotch CAR-modified cells for further analysis and functional evaluation. [Table 2-1] [Table 2-2]

[0136] To mitigate off-tumor toxicity to normal tissues, we examined the effect of the density of CD147 expression in various cell types (focusing on hematopoietic cells) on the anti-tumor activity of CD147-CAR. CD147 expression was evaluated in HepG2, Raji, Daudi, and PBMCs. Different expression levels of CD147 were observed (Figure 22A). Notably, cells expressing low levels of CD147 (e.g., PBMCs) did not induce the cytotoxic activity of CD147-CAR-NK-92MI cells, even when CD147-CAR cells were cultured with target cells at a high effector-to-target ratio (E:T ratio) of 10:1 (Figure 22B). These findings suggest that only the optimized scFv sequence of anti-CD147 allows the specific scFv domain to bind to highly expressed CD147 molecules, which can mitigate off-tumor toxicity against normal tissues expressing low levels of CD147 molecules.

[0137] To further reduce the off-tumor toxicity of CD147-CAR, we used a synNotch receptor that can release transcription factors that drive the expression of CARs against various tumor antigens. The "logic gate" synNotch CAR can only kill dual-antigen-positive tumor cells, but not single-antigen-positive tumor cells. To reduce off-tumor toxicity, we designed a combinatorial approach consisting of GPC-3 and CD147. Briefly, we constructed an SFG retroviral vector encoding an anti-GPC3-specific synNotch receptor linking the Gal4-VP64 intracellular transcriptional activation domain. Constitutively expressed enhanced GFP (eGFP) was placed downstream of GPC3-synNotch to identify transduced cells (Figure 23A).

[0138] A lentiviral vector was constructed in which the anti-CD147-CAR was placed under the control of an upstream activation sequence (UAS) promoter that can be activated by the Gal4-VP64 transcription factor released after synNotch receptor engagement. A constitutively expressed monomeric red fluorescent protein, Cherry (mCherry), was placed downstream of the inducible CD147-CAR to identify transduced cells (Figure 23A).

[0139] Human PBMCs were co-transduced with both lentiviral and retroviral vectors (Figure 23C). Double-positive cells were verified by eGFP (a marker for anti-GPC3-synNotch) and mCherry (a marker for CD147-CAR) using flow cytometry analysis (Figure 23D). Four subsets of transduced T cells were analyzed, including mCherry-positive only, GFP-positive only, GFP and mCherry double-positive, and GFP and mCherry double-negative subsets (Figure 23E). These transduced T cells were GPC3-positive. high CD147 low HepG2 cell line was primed to induce CD147-CAR expression on the surface (Figure 23F). In the absence of synNotch engagement, CD147-CAR expression was not observed. No leaky CAR expression was observed in transduced primary T cells. However, approximately 10% leaky CAR expression was observed in transduced NK-92MI cells (data not shown). This observation was further verified by a gamma secretase inhibitor (MK-0752, a Notch signaling inhibitor) treatment assay (Figure 24). CD147-CAR expression in GPC3-synNotch-GFP and CD147-CAR-mCherry double-positive T cells was dramatically inhibited by MK-0752 treatment (Figure 24).

[0140] Co-transduction of GPC3-synNotch-GFP and CD147-CAR-mCherry into T cells and GPC3 high CD147 low After priming with HepG2 cell line, the activity of transduced T cells was elicited for 2 hours by various subsets of HCC cell lines, and the killing efficacy was evaluated. Various subsets of HCC cell lines expressed CD147+GPC3 high HepG2 cell line, CD147 ko GPC3 high HepG2 cell line, CD147+GPC3 low HepG2 cell line, CD147 koGPC3 low The HepG2 cell line (Figure 23G) was used as a positive control.

[0141] CD147 ko GPC3 high GPC3-synNotch-GFP and CD147-CAR-mCherry double-positive T cells primed with HepG2 cells expressed CD147+GPC3 high It could be specifically activated by HepG2 cells (Figure 23H), as quantified by surface expression of CD107a when co-cultured with various target cell lines. Similar results were obtained when T cells were co-transduced with myc-tagged CD147-specific synNotch-GFP and inducible GPC3-CAR-mCherry (Figures 25A-25H).

[0142] Together, these data suggest that synNotch GPC3-induced CD147-CAR T cells can be specifically activated by GPC3+CD147+ HepG2 cells, but not by GPC3+CD147- or GPC3-CD147+ HepG2 cells. These activated CD147-CAR T cells can kill CD147+GPC3+HepG2 cells, but not CD147-GPC3+HepG2 cells (Figures 26A-26D).

[0143] In view of the many possible embodiments to which the principles of this disclosure may be applied, it should be recognized that the described embodiments are examples only and should not be construed as limiting the scope of the invention. Rather, the scope of the invention is defined by the following claims. We therefore claim as our invention all that comes within the scope and spirit of these claims. In certain embodiments, for example, the following items are provided: (Item 1) (a) an antigen-binding domain that specifically binds to CD147, comprising an amino acid sequence comprising the amino acid sequences of variable heavy chain (VH) domain complementarity-determining region 1 (CDR1), CDR2, and CDR3 at amino acid positions 26 to 32, 52 to 59, and 101 to 104, respectively, of SEQ ID NO: 2, and the amino acid sequences of variable light chain (VL) domain CDR1, CDR2, and CDR3 at amino acid positions 155 to 165, 181 to 187, and 220 to 228, respectively, of SEQ ID NO: 2; (b) hinge domain, (c) a transmembrane domain, and (d) an intracellular domain comprising one or more costimulatory molecule intracellular domains and an intracellular signaling domain; A chimeric antigen receptor comprising: (Item 2) 2. The chimeric antigen receptor of item 1, wherein the antigen-binding domain has at least 90% sequence identity with the amino acid sequence of SEQ ID NO: 2. (Item 3) 3. The chimeric antigen receptor of item 2, wherein the antigen-binding domain comprises the amino acid sequence of SEQ ID NO: 2. (Item 4) 4. The chimeric antigen receptor of any one of items 1 to 3, wherein the antigen-binding domain binds to cells expressing high levels of surface CD147. (Item 5) 5. The chimeric antigen receptor of any one of items 1 to 4, wherein the one or more costimulatory molecule intracellular domains comprise the intracellular domains of CD28 and 4-1BB. (Item 6) 6. The chimeric antigen receptor of any one of items 1 to 5, wherein the intracellular signaling domain comprises the signaling domain of CD3ζ. (Item 7) 7. The chimeric antigen receptor of any one of items 1 to 6, wherein the hinge domain comprises an IgG1 hinge domain. (Item 8) 8. The chimeric antigen receptor of any one of items 1 to 7, wherein the transmembrane domain is a CD28 transmembrane domain or a CD8a transmembrane domain. (Item 9) 9. The chimeric antigen receptor of any one of items 1 to 8, comprising an amino acid sequence having at least 90% identity with the amino acid sequence of SEQ ID NO: 5. (Item 10) 10. The chimeric antigen receptor of item 9, comprising the amino acid sequence of SEQ ID NO: 5. (Item 11) 11. The chimeric antigen receptor of any one of items 1 to 10, further comprising one or more additional antigen-binding domains. (Item 12) 12. The chimeric antigen receptor of claim 11, wherein the one or more additional antigen-binding domains specifically bind to glypican 3, alpha-fetoprotein, or mucin-1. (Item 13) 13. The chimeric antigen receptor of any one of 1 to 12, further comprising an inducible suicide molecule. (Item 14) 14. The chimeric antigen receptor of item 13, wherein the suicide molecule comprises caspase 9. (Item 15) 15. The chimeric antigen receptor of item 13 or item 14, wherein expression of the suicide molecule is inducible by tetracycline, doxycycline, or rapamycin. (Item 16) 15. The chimeric antigen receptor of item 13 or item 14, comprising an amino acid sequence having at least 90% identity to the amino acid sequence of SEQ ID NO: 7. (Item 17) 17. The chimeric antigen receptor of item 16, comprising the amino acid sequence of SEQ ID NO: 7. (Item 18) 18. The chimeric antigen receptor of any one of items 1 to 17, further comprising a cytokine receptor intracellular domain. (Item 19) 19. The chimeric antigen receptor according to Item 18, wherein the cytokine receptor intracellular domain is an interleukin-15 receptor intracellular domain, an interleukin-12 receptor intracellular domain, or an interleukin-18 receptor intracellular domain. (Item 20) 20. A nucleic acid molecule encoding the chimeric antigen receptor of any one of items 1 to 19. (Item 21) 21. The nucleic acid molecule of item 20, comprising the nucleic acid sequence of SEQ ID NO: 4, SEQ ID NO: 6, or SEQ ID NO: 14. (Item 22) A vector comprising the nucleic acid molecule of item 20 or item 21. (Item 23) 23. The vector of item 22, further comprising a nucleic acid molecule encoding a second chimeric antigen receptor that targets hepatitis B virus or hepatitis C virus. (Item 24) 23. The vector of Item 22, further comprising an inducible promoter or enhancer nucleic acid molecule operably linked to the chimeric antigen receptor nucleic acid molecule. (Item 25) 25. The vector of item 24, wherein the enhancer nucleic acid molecule is a Gal4 upstream activating sequence. (Item 26) 26. The vector of item 25, comprising the nucleic acid sequence of SEQ ID NO: 15. (Item 27) 27. The vector of any one of items 22 to 26, which is a viral vector. (Item 28) 28. The vector of item 27, wherein the viral vector is a lentiviral vector or a retroviral vector. (Item 29) A single-chain variable fragment (scFv) that specifically binds to CD147, comprising an amino acid sequence including the amino acid sequences of variable heavy chain (VH) domain complementarity determining region 1 (CDR1), CDR2, and CDR3 at amino acid positions 26 to 32, 52 to 59, and 101 to 104, respectively, of SEQ ID NO: 2, and the amino acid sequences of variable light chain (VL) domain CDR1, CDR2, and CDR3 at amino acid positions 155 to 165, 181 to 187, and 220 to 228, respectively, of SEQ ID NO: 2. (Item 30) 30. The scFv according to item 29, wherein the amino acid sequence has at least 90% sequence identity with the amino acid sequence of SEQ ID NO: 2. (Item 31) 31. The scFv according to item 30, wherein the amino acid sequence comprises the amino acid sequence of SEQ ID NO: 2. (Item 32) 32. The scFv of any one of paragraphs 29 to 31, which binds to cells expressing high levels of surface CD147. (Item 33) 33. A nucleic acid molecule encoding the scFv of any one of items 29 to 32. (Item 34) 34. The nucleic acid molecule of Item 33, comprising the nucleic acid sequence of SEQ ID NO: 1. (Item 35) A vector comprising the nucleic acid sequence of item 33 or item 34. (Item 36) 20. A T cell, natural killer (NK) cell, natural killer T (NKT) cell, double-negative T (DNT) cell, neutrophil, or macrophage expressing the chimeric antigen receptor of any one of items 1 to 19. (Item 37) A T cell, NK cell, natural killer T (NKT) cell, double-negative T (DNT) cell, neutrophil, or macrophage comprising the nucleic acid of item 20, item 21, item 33, or item 34, or the vector of any one of items 22 to 28 or item 35. (Item 38) 38. The T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of item 37, further comprising a nucleic acid molecule encoding an activator of the inducible promoter or enhancer. (Item 39) 39. The T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of item 38, wherein the nucleic acid encoding the activator further encodes an anti-GPC3 specific binding agent. (Item 40) 36. A T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage comprising the vector of item 24 or 35, further comprising a vector comprising an anti-GPC3 chimeric antigen receptor operably linked to a nucleic acid molecule encoding an activator of an inducible promoter or enhancer. (Item 41) 41. The T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of Item 40, wherein the vector comprising the anti-GPC chimeric antigen receptor comprises the nucleic acid sequence of SEQ ID NO: 16. (Item 42) 42. The T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of any one of items 36 to 41, obtained from peripheral blood, umbilical cord blood, lymph node, bone marrow, tumor tissue, or a cell line. (Item 43) 43. The NK cell of any one of items 36 to 42, which is an NK-92 cell or an NK-92MI cell. (Item 44) 29. A method of producing CD147-CAR-T cells or CD147-CAR-NK cells, comprising transducing or transfecting T cells or NK cells with the vector of any one of paragraphs 22 to 28. (Item 45) 44. A method of treating a subject having cancer, comprising administering to the subject an effective amount of the T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of any one of items 36 to 43. (Item 46) 46. ​​The method of claim 45, wherein the subject has a cancer that expresses CD147. (Item 47) 47. The method of claim 45 or 46, wherein the subject has hepatocellular carcinoma, neuroblastoma, breast cancer, pancreatic cancer, leukemia, lymphoma, multiple myeloma, colorectal cancer, lung cancer, melanoma, renal cell carcinoma, sarcoma, or nasopharyngeal carcinoma. (Item 48) 48. The method of any one of items 45 to 47, further comprising treating the subject with at least one of surgery, radiation, chemotherapy, additional immunotherapy, or a combination of two or more thereof.

Claims

1. (a) an antigen-binding domain that specifically binds to CD147, wherein the antigen-binding domain is encoded by a nucleic acid comprising the nucleic acid sequence of SEQ ID NO:1; (b) a hinge domain; (c) a transmembrane domain, and (d) an intracellular domain comprising one or more costimulatory molecule intracellular domains and an intracellular signaling domain. A chimeric antigen receptor comprising:

2. The chimeric antigen receptor of claim 1 , wherein the antigen binding domain comprises the amino acid sequence of SEQ ID NO:

2.

3. The chimeric antigen receptor of claim 1 or claim 2, wherein the antigen binding domain binds to cells expressing high levels of surface CD147.

4. The chimeric antigen receptor of any one of claims 1 to 3, wherein the one or more costimulatory molecule intracellular domains comprise the intracellular domains of CD28 and 4-1BB.

5. The chimeric antigen receptor according to any one of claims 1 to 4, wherein the intracellular signaling domain comprises the signaling domain of CD3ζ.

6. The chimeric antigen receptor of claim 1 , wherein the hinge domain comprises an IgG1 hinge domain.

7. 7. The chimeric antigen receptor of claim 1, wherein the transmembrane domain is a CD28 transmembrane domain or a CD8a transmembrane domain.

8. 8. The chimeric antigen receptor of claim 1 , comprising an amino acid sequence having at least 90% identity to the amino acid sequence of SEQ ID NO:

5.

9. The chimeric antigen receptor of claim 8 comprising the amino acid sequence of SEQ ID NO:

5.

10. 10. The chimeric antigen receptor of claim 1, further comprising one or more additional antigen binding domains.

11. The chimeric antigen receptor of claim 10, wherein the one or more additional antigen binding domains specifically bind to glypican 3, alpha-fetoprotein, or mucin-1.

12. The chimeric antigen receptor of claim 1 , further comprising an inducible suicide molecule.

13. The chimeric antigen receptor of claim 12, wherein the suicide molecule comprises caspase 9.

14. The chimeric antigen receptor of claim 12 or claim 13, wherein expression of the suicide molecule is inducible by tetracycline, doxycycline, or rapamycin.

15. 14. The chimeric antigen receptor of claim 12 or claim 13, comprising an amino acid sequence having at least 90% identity to the amino acid sequence of SEQ ID NO:

7.

16. The chimeric antigen receptor of claim 15, comprising the amino acid sequence of SEQ ID NO:

7.

17. 17. The chimeric antigen receptor of any one of claims 1 to 16, further comprising a cytokine receptor intracellular domain.

18. The chimeric antigen receptor of claim 17, wherein the cytokine receptor intracellular domain is an interleukin-15 receptor intracellular domain, an interleukin-12 receptor intracellular domain, or an interleukin-18 receptor intracellular domain.

19. The chimeric antigen receptor of claim 18, wherein the cytokine receptor intracellular domain is an interleukin-15 receptor intracellular domain comprising the amino acid sequence of SEQ ID NO:

12.

20. 20. A nucleic acid molecule encoding the chimeric antigen receptor of any one of claims 1 to 19.

21. 21. The nucleic acid molecule of claim 20 comprising the nucleic acid sequence of SEQ ID NO:4, SEQ ID NO:6, or SEQ ID NO:

14.

22. A vector comprising the nucleic acid molecule of claim 20 or claim 21.

23. 23. The vector of claim 22, further comprising a nucleic acid molecule encoding a second chimeric antigen receptor that targets a hepatitis B virus or a hepatitis C virus.

24. 23. The vector of claim 22, further comprising an inducible promoter or enhancer nucleic acid molecule operably linked to the nucleic acid molecule encoding the chimeric antigen receptor.

25. The vector of claim 24, wherein the enhancer nucleic acid molecule is a Gal4 upstream activating sequence.

26. 26. The vector of claim 25, comprising the nucleic acid sequence of SEQ ID NO:

15.

27. 27. The vector of any one of claims 22 to 26, which is a viral vector.

28. 28. The vector of claim 27, wherein the viral vector is a lentiviral vector or a retroviral vector.

29. A T cell, natural killer (NK) cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage, expressing a chimeric antigen receptor according to any one of claims 1 to 19.

30. 29. A T cell, an NK cell, a natural killer T (NKT) cell, a double negative T (DNT) cell, a neutrophil, or a macrophage comprising a nucleic acid according to claim 20 or claim 21, or a vector according to any one of claims 22 to 28.

31. 27. A T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage comprising the vector of any one of claims 24 to 26, further comprising a nucleic acid molecule encoding an activator of the inducible promoter or enhancer.

32. The T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of claim 31 , wherein the nucleic acid encoding the activator further encodes an anti-GPC3 specific binding agent.

33. A T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage comprising the vector of claim 24, further comprising a vector comprising a nucleic acid molecule encoding an activator of an inducible promoter or enhancer and an anti-GPC3 chimeric antigen receptor operably linked to the inducible promoter or enhancer.

34. The T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of claim 33, wherein the vector comprising an anti-GPC chimeric antigen receptor comprises the nucleic acid sequence of SEQ ID NO:

16.

35. 35. The T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage of any one of claims 29 to 34, obtained from peripheral blood, umbilical cord blood, lymph node, bone marrow, tumor tissue, or cell line.

36. The NK cell according to any one of claims 29 to 35, which is an NK-92 cell or an NK-92MI cell.

37. A method for producing CD147-CAR-T cells or CD147-CAR-NK cells, comprising transducing or transfecting T cells or NK cells with a vector according to any one of claims 22 to 28.

38. 37. A composition for treating a subject with cancer comprising a T cell, NK cell, natural killer T (NKT) cell, double negative T (DNT) cell, neutrophil, or macrophage according to any one of claims 29 to 36.

39. The composition of claim 38, wherein the subject has a cancer that expresses CD147.

40. 40. The composition of claim 38 or claim 39, wherein the subject has hepatocellular carcinoma, neuroblastoma, breast cancer, pancreatic cancer, leukemia, lymphoma, multiple myeloma, colorectal cancer, lung cancer, melanoma, renal cell carcinoma, sarcoma, or nasopharyngeal carcinoma.

41. 41. The composition of any one of claims 38 to 40, wherein the subject has been further treated with at least one of surgery, radiation, chemotherapy, additional immunotherapy, or a combination of two or more thereof.

Images