Dual chimeric antigen receptor-t cell which can be regulated, construction method therefor and use thereof

Abstract

The present invention discloses a novel switchable dual chimeric antigen receptor-T (sdCAR-T) cell and a construction method and use thereof, which fall within the field of cellular immunotherapy for tumors. The dual chimeric antigen receptor consists of a first chimeric antigen receptor for MSLN and a second chimeric antigen receptor for FITC. A dual-targeted functional T cells regulated by specific exogenous bifunctional molecules is constructed, and the exogenous molecules are used to preliminarily discuss the in vivo and in vitro activity of the dual chimeric antigen receptor-T cell. By means of in vitro and in vivo tests, it is confirmed that the activation mode of the constructed CAR-T cell is controlled by the combination of endogenous tumor antigens and exogenous bifunctional molecules, and this combined regulation mode can significantly improve the safe application of CAR-T cell immunotherapy.

Claims

1. A switchable dual chimeric antigen receptor T cell, wherein the dual chimeric antigen receptor consists of a first chimeric antigen receptor for MSLN and a second chimeric antigen receptor for FITC; the dual chimeric antigen receptor contains a blue fluorescent protein tag downstream, and a linker peptide P2A is contained between the dual chimeric antigen receptor and the fluorescent protein tag; wherein a molecular switch is provided in the form of FITC coupled to polypeptide HM-3, wherein the FITC and polypeptide HM-3 are coupled to form a bifunctional specific small molecule drug, and the amino acid sequence of the polypeptide HM-3 is Ile-Val-Arg-ArgAla-Asp-Arg-Ala-Ala-Val-Pro-Gly-Gly-Gly-Gly-Arg-Gly-Asp (SEQ ID NO: 18); and the nucleotide sequence of the dual chimeric antigen receptor as shown in SEQ ID NO: 1.

2. The switchable dual chimeric antigen receptor T cell and the molecular switch according to claim 1 for use in the field of preparing tumor drugs for treating overexpression of tumor antigen MSLN and integrin ?v?3.

3. The switchable dual chimeric antigen receptor T cell and the molecular switch according to claim 2 for use in the field of preparing tumor drugs for treating overexpression of tumor antigen MSLN and integrin ?v?3, wherein the tumors comprise ovarian cancer, lung cancer, esophageal cancer, pancreatic cancer, gastric cancer, colon cancer, breast cancer, liver cancer, melanoma, head and neck cancer, cervical cancer and osteosarcoma.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a schematic diagram showing the structure and killing mechanism of sdCAR-T cells;

(2) FIG. 2 is a schematic diagram of activation principle of sdCAR-T cells;

(3) FIG. 3 is a diagram showing Western Blot analysis of MSLN and CEA expression after transfection of K562 cell;

(4) FIG. 4 is a diagram showing Western Blot analysis of integrin ?v?3 expression in K562 cells;

(5) FIG. 5 is a diagram showing the state of an isolated humanized PBMC;

(6) FIG. 6 is a flow cytometry diagram of CD4.sup.+T cells and CD8.sup.+T cells;

(7) FIG. 7 is a flow cytometry diagram of sdCAR expression after electrotransfection of CD4.sup.+T cells and CD8.sup.+T cells;

(8) FIG. 8 is a comparison between in vitro activities of sdCAR-T cells with various structures;

(9) FIG. 9 is a diagram showing the level of IL-2 activated by sdCAR-T cells in vitro;

(10) FIG. 10 is a diagram showing the level of IFN? activated by sdCAR-T cells in vitro;

(11) FIG. 11 is a flow cytometry diagram of CD69 on the surface of sdCAR-T cells;

(12) FIG. 12 is a diagram showing the analysis of CD69 expression on the surface of sdCAR-T cells;

(13) FIG. 13 is a diagram showing the detection of proliferation of sdCAR-T cells in vitro;

(14) FIG. 14 is a diagram of fluorescence detection of cytotoxicity of sdCAR-T cells in vitro;

(15) FIG. 15 is a flow cytometry diagram of cytotoxicity of sdCAR-T cells in vitro;

(16) FIG. 16 is a diagram of cytotoxicity of sdCAR-T cells in vitro;

(17) FIG. 17 is a detection diagram of time regulation of sdCAR-T cells by exogenous bifunctional switching molecules;

(18) FIG. 18 is a detection diagram of concentration regulation of sdCAR-T cells by exogenous bifunctional switching molecules;

(19) FIG. 19 is a flow chart of testing of sdCAR-T cells in vivo;

(20) FIG. 20 is a flow cytometry diagram of cytotoxicity of sdCAR-T cells in vivo;

(21) FIG. 21 is a diagram of cytotoxicity of sdCAR-T cells in vivo;

(22) FIG. 22 is a diagram showing the level of IL-2 released by sdCAR-T cells in vivo;

(23) FIG. 23 is a diagram showing the level of IFN? released by sdCAR-T cells in vivo;

(24) FIG. 24 is a diagram showing the level of mesothelin expressed by various solid tumors;

(25) FIG. 25 is a diagram showing the level of integrin ?v expressed by various solid tumors;

(26) FIG. 26 is a diagram showing the level of integrin ?3 expressed by various solid tumors; and

(27) FIG. 27 is a diagram showing the level of cytotoxicity of sdCAR-T cells on solid tumors in vivo.

DETAILED DESCRIPTION

(28) The present invention is further described below with reference to specific examples.

EXAMPLE 1

Determination of Whole Gene Sequence of a Dual Chimeric Antigen Receptor

(29) Referring to relevant information in the NCBI database, the amino acid sequence of MSLNscFv (GenBank: AAC04760) and the amino acid sequence of FITCscFv (PDB: 1X9Q_A) were obtained. Their gene sequences were analyzed and optimized to ensure that they were more suitable for high-efficiency expression in humanized T lymphocytes without changing the coding amino acid sequence. The sequence information of MSLNscFv is detailed in SEQUENCE LISTING (SEQ ID NO. 6); and the sequence information of FITCscFv is detailed in SEQUENCE LISTING (SEQ ID NO. 9). The amino acid sequence of a humanized signal peptide (SEQ ID NO. 8 or 11 in SEQUENCE LISTING), the amino acid sequence of a humanized CD8? hinge region (SEQ ID NO. 12), the amino acid sequence of a human CD8 transmembrane region and intracellular domain gene (SEQ ID NO. 13 in SEQUENCE LISTING), the amino acid sequence of human CD3? (SEQ ID NO. 10 in SEQUENCE LISTING), the amino acid sequences of a 4/1BB signal domain (SEQ ID NO. 7 in SEQUENCE LISTING), the amino acid sequence of IRES (SEQ ID NO. 3 in SEQUENCE LISTING), as well as the amino acid sequence of a linker peptide P2A (SEQ ID NO. 5 in SEQUENCE LISTING) and the amino acid sequence of a blue protein tag BFP (SEQ ID NO. 4 in SEQUENCE LISTING) were retrieved from a GenBank database on the website of NCBI (http://www.ncbi.nlm.nih.gov/).

(30) The above gene sequences were sequentially connected according to the sequences of the signal peptide, MSLNscFv, the human CD8? hinge region, the human CD8 transmembrane region and intracellular domain, the human 4/1BB costimulatory signal region, IRES, the signal peptide, FITCscFv, the human CD8? hinge region, the human CD8 transmembrane region and intracellular domain, and the human CD3? intracellular domain, and a Kozak sequence (SEQ ID NO. 2 in SEQUENCE LISTING) was introduced at the head end to finally construct complete gene sequence information of the sdCAR (SEQ ID NO. 1). The structure and killing mechanism of sdCAR-T cells are shown in FIG. 1.

EXAMPLE 2

Construction of an Overexpression Plasmid for a Dual Chimeric Antigen Receptor

(31) 2.1 Synthesis of a Gene Fragment

(32) An optimized complete sdCAR gene sequence was subjected to whole gene synthesis (Suzhou GENEWIZ Biotechnology Co., Ltd.) and then cloned into an overexpression vector pFC-PGK-MCS of a ?C31 site-specific recombinant enzyme system through specific cleavage sites Afl II and Xba I to finally construct a recombinant plasmid pFC-sdCAR. A fluorescent tag BFP was introduced to facilitate detection and tracking of the constructed sdCAR-T cells.

(33) 2.2 Construction and Sequencing of the Recombinant Plasmid

(34) The recombinant plasmid was sent to Suzhou GENEWIZ Biotechnology Co., Ltd. for sequencing. Sequencing results were compared with the sdCAR gene sequence to be synthesized. The results confirmed that the synthesized sequence was correct and a target gene fragment was linked to the overexpression vector pFC-PGK-MCS.

(35) The sequencing primer used is a universal primer on pFC vector:

(36) TABLE-US-00001 Upstreamprimer(5.fwdarw.3): TAATACGACTCACTATAGG (SEQIDNO.16) Downstreamprimer(5.fwdarw.3): CAGGAAACAGCTATGAC (SEQIDNO.17)

EXAMPLE 3

Acquisition of Recombinant Tumor Cells

(37) 3.1 Design of an Overexpression Vector for MSLN Lentivirus

(38) References were made to the CDS region sequence of MSLN in the GenBank database on the NCBI website (the gene sequence number is NM_013404). There was a linker peptide P2A between MSLN and GFP. The MSLN-P2A-GFP gene fragment was inserted into a lentivirus overexpression vector pLVX through cleavage sites EcoRI and XbaI to form a pLVX-MSLN recombinant expression plasmid.

(39) 3.2 Design of an Overexpression Vector for CEA Lentivirus

(40) References were made to the CDS region sequence of humanized MSLN CEA in the GenBank database on the NCBI website (the gene sequence number is M29540.1). There was a linker peptide P2A between CEA and mCherry. The CEA-P2A-mCherry gene fragment was inserted into a lentivirus overexpression vector pLVX through cleavage sites EcoRI and XbaI to form a pLVX-CEA recombinant expression plasmid.

(41) 3.3 Preparation of Lentivirus Particles

(42) 1) A 15 cm culture dish was prepared, and the complete culture medium (DMEM with high glucose) containing 10% fetal bovine serum (FBS) was add. 5?10.sup.6 293T cells were inoculated and placed in an incubator at 37? C. with 5% CO.sub.2, cultured overnight.

(43) 2) The constructed overexpression vector pLVX-MSLN or pLVX-CEA (the concentration is about 100 ?M) and lentivirus packaging plasmids (Lenti-GOI, pGP, and pVSVG) were taken out from a refrigerator. After thawing at room temperature, they were blown up and down with a pipette for complete and uniform mixing.

(44) 3) The phosphate buffered saline (PBS) was taken out and heated to room temperature. 2 mL of PBS was taken to a well of a 6-well plate, 10 ?g of Lenti-GOI, 6 ?g of pGP and 5 ?g of pVSVG were add respectively. After they were blown up and down with the pipette for full mixing, 18 ?L of pLVX-MSLN or pLVX-CEA was added, then immediately blown up and down with the pipette for uniform mixing, and standed for 10 min at room temperature.

(45) 4) The complex of the above overexpression plasmid pLVX-MSLN (or pLVX-CEA) and various packaging plasmids were added dropwise into 293T cells cultured overnight, and the culture dish was gently shaked for full and uniform mixing. Then the culture dish was placed in an incubator at 37? C. with 5% CO.sub.2.

(46) 5) After culturing for 6-8 h (preferably 8 h in this example), the culture medium containing the transfection reagent was removed and replaced with fresh complete culture medium.

(47) 6) After continuous cultivation for 48-72 h (preferably 48 h in this example), the culture medium supernatant containing virus in the culture dish was collected.

(48) 7) The obtained culture medium supernatant was filtered with a 0.45 ?m filter membrane, the filtrate was transferred to centrifuge tubes, and centrifuged at a centrifugal force of 50000 g at a high speed at 4? C. for 2 h.

(49) 8) After centrifugation, the liquid from the centrifuge tubes was carefully sucked in a biosafety cabinet, and 500 ?L of PBS was added to resuspend the precipitate. The virus was stored at ?80? C.

(50) 9) The titer of the obtained lentivirus particles was determined. The titer of the virus particles overexpressing the MSLN was 1.13?10.sup.8 TU/mL, and the titer of the virus particles overexpressing the CEA was 1.62?10.sup.8 TU/mL.

(51) 3.4 Determination of the Killing Curve of K562 Tumor Cells by Puromycin

(52) 1) Low-generation cell strains were resuscitated from liquid nitrogen, and subjected to conventional culture and passaged 5 times, then the cell state was adjusted to a logarithmic growth period.

(53) 2) A new 24-well plate was prepared, 50000 K562 tumor cells were inoculated per well for 7 wells in total, and 500 ?L of complete culture medium (RPMI1640, 10% FBS) was added per well. The well plate was placed in an incubator at 37? C. with 5% CO.sub.2 and cultured overnight.

(54) 3) The cell well plate was taken out from the incubator, one well was left free of puromycin, and puromycin with final concentrations of 0.5, 1, 2, 4, and 8 ?g/mL were added to the remaining wells respectively.

(55) 4) The well plate was placed in an incubator at 37? C. with 5% CO.sub.2 for continuous cultivation. The state of cells in the well plate was observed under a microscope every day for 5 consecutive days. And the minimum concentration that can completely kill all cells on the fifth day was taken as the puromycin concentration for subsequent stable cell strain screening.

(56) Through the determination of the killing curve, the concentration of puromycin was finally selected to be 1 ?g/mL.

(57) 3.5 Determination of a Killing Curve of K562 Tumor Cells by Hygromycin

(58) The method is the same as that of Example 3.4. The concentration gradient of the hygromycin was set to 50, 100, 200, 400 and 800 ?g/mL respectively, and the concentration of the hygromycin was finally selected to be 100 ?g/mL through the determination of the killing curve.

(59) 3.6 Lentivirus Transfection Test of K562 Tumor Cells

(60) 1) Low-generation cell strains were resuscitated from liquid nitrogen, and subjected to conventional culture and passaged 5 times, then the cell state was adjusted to a logarithmic growth period.

(61) 2) A new 6-well plate was prepared, the cells were inoculated according to the density of 3?10.sup.5 cells/well for a total of 2 wells and complete culture medium (RPMI1640, 10% FBS) was added to 3 mL/well. The well plate was placed in an incubator at 37? C. with 5% CO.sub.2 and cultured overnight.

(62) 3) A tube of overexpressed lentivirus was taken out from the ?80? C. refrigerator and placed in a 37? C. water bath for quick thawing. The cells were taken out from the incubator and replaced the culture medium with fresh complete culture medium (RPMI1640, 10% FBS).

(63) 4) Polybrene with a final concentration of 6 ?g/mL was added to the culture medium in one of the wells, then 100 ?L of lentivirus was added, and gently blown with the pipette for uniform mixing.

(64) 5) The well plate was placed in a horizontal centrifuge and centrifuged at a centrifugal force of 800 g for 1 h. After centrifugation, the well plate was placed in an incubator at 37? C. with 5% CO.sub.2 for continuous cultivation for 24 h.

(65) 6) The well plate was taken out from the incubator, placed in the horizontal centrifuge, and centrifuged at a centrifugal force of 2000 g for 5 min, then the culture medium containing virus supernatant was removed from the well plate. Fresh culture medium (RPMI1640, 10% FBS) without antibiotics was added, and continued to culture for 2 days (if the cell density was large, passage can be carried out).

(66) 7) The well plate was taken out from the incubator, placed in the horizontal centrifuge, and centrifuged at a centrifugal force of 2000 g for 5 min, then the culture medium was removed from the well plate, and the culture medium containing puromycin or hygromycin antibiotic was added, wherein the concentrations of the puromycin and hygromycin used were the same as those selected in Example 3.4 and 3.5, respectively.

(67) 8) The control well cells not infected with virus were taken as the control and continuously cultured for 5 days until the control group cells were completely dead, so that the cells surviving in the well infected with virus were the stable cell strains successfully constructed.

(68) 9) Western Blot was used to detect the expression of target genes in the stable cell strains, as shown in FIG. 3.

(69) 3.7 Expression of Integrins ?v and ?3 in K562 Tumor Cells

(70) Western Blot was used to detect the expression of integrin ?v?3 in K562 cell strains, as shown in FIG. 4. The results showed that K562 cells highly expressed integrin ?v?3.

EXAMPLE 4

Preparation of T Lymphocytes in Peripheral Blood

(71) 4.1 Acquisition of PBMC

(72) The PBMC includes lymphocytes, monocytes and the like. Their volume, morphology and density are different from those of other cells. The cell densities of red blood cells and granulocytes and the like are larger, about 1.092 g/mL, while the densities of the lymphocytes and monocytes are about 1.070 g/mL. Ficoll reagent, a lymphocyte separation solution, was a mixture of 60% polysucrose and 34% meglumine diatrizoate at a ratio of 2:1, with a specific weight of about 1.077?0.001. The lymphocytes were separated from other blood cells by centrifugation and distribution according to density gradient. The specific implementation method includes the following steps of: taking 1 mL of fresh anticoagulant and PBS and evenly mixing at a ratio of 1:1, carefully adding 1 mL of Ficoll separation solution, and centrifuging at 2000 rpm for 20 min at room temperature, at this time the cells in a centrifuge tube are divided into four layers from bottom to top; collecting an upper second milky film layer as PBMC layer; after fully and evenly mixing the PBMC with 5 mL of PBS, centrifuging at 1000 rpm for 10 min, and washing twice to obtain the PBMC with relatively high purity.

(73) The PBMC obtained by separation was inoculated into a 15 cm culture dish at a cell density of 5?10.sup.5 cells per ml, the culture medium was RPMI1640 containing 10% FBS, and cytokine IL-2 with a final concentration of 50 U/mL was added. The culture dish was placed in an incubator at 37? C. with 5% CO.sub.2 for 48 h, and the cell state was observed under a microscope after the culture was completed. The results are shown in FIG. 5. It can be observed that the cell state had obvious differentiation after PBMC culture. Some of the cells were round and transparent suspension cells growing in an aggregated manner, while some other cells are cells that began to adhere to the wall and grew in a spindle shape. After analysis, the cells growing in suspension should be lymphocytes and the cells with adherent growth should be monocytes.

(74) 4.2 Magnetic-Activated Cell Sorting of T Cells

(75) This example mainly used the magnetic-activated cell sorting method to obtain target T cells. Immunomagnetic bead sorting is based on the antigen expressed on the cell surface and the microbeads coupled to the antibody corresponding to the antigen. After the antigen is incubated with the antibody, an external magnetic field is added, and the microbeads attached to the antibody remain in the magnetic field under the action of a strong magnetic field. Therefore, only cells expressing the antigen remain in the magnetic field, thus realizing separation of target cells. At present, immunomagnetic bead sorting methods mainly include positive sorting method, negative sorting method and compound sorting method. The positive sorting method means that the cells bound by microbeads are target cells; the negative sorting method means that the cells bound by microbeads are not target cells. The compound sorting method means removing non-targeted cells by using the negative sorting method and then using the positive sorting method. This method is mainly used to sort cell populations with a low content. According to the present invention, CD4.sup.+T cells and CD8.sup.+T cells were sorted by using the positive sorting method, and the sorting result was subjected to flow cytometry. The results are shown in FIG. 6. The results showed that the positive rates of the CD4.sup.+T cells and the CD8.sup.+T cells obtained by magnetic-activated cell sorting were about 92.4% and 90.7%, respectively.

(76) 4.3 Preparation of sdCAR-T Cells by Electrotransfection

(77) This example used a ?C31 site-specific integrase system to insert the genes expressing the dual chimeric antigen receptor into genome of T lymphocyte to prepare sdCAR-T cells safely and efficiently. The ?C31 integrase-mediated recombination system belongs to highly conservative serine recombinase. Its near N terminal is an active region containing a serine residue. The serine residue attacks the DNA skeleton for staggered cut to form a double-stranded broken terminal of 3-OH, while the 5-phosphate group is covalently linked with recombinase to form a cross-linked intermediate, finally realizing gene recombination. Compared with lentivirus and a transposon system, the recombination system can integrate large fragments of exogenous genes safely, efficiently and accurately. The ?C31 integrase-mediated recombination system has the advantages of unidirectional integration and no need of external cofactors, and can realize stable and efficient expression of exogenous genes. In order to further improve the integration efficiency and specificity, the weight ratio of the integrase plasmids to the overexpression vector plasmids in the mixed plasmid was adjusted to 50:1 during electrotransfection. The electrotransfection operation was performed using Lonza Corporation's electrotransfection instruments and electrotransfection kits. Specific implementation steps of electrotransfection:

(78) 1) CD3.sup.+T cells were inoculated into a complete culture medium (RPMI1640, 10% FBS), the cell density was controlled at 5?10.sup.5 cells/mL, and growth factor IL-2 was added to the final concentration of 50 U/mL respectively, and cultured at 37? C. with 5% CO.sub.2 for 48 h.

(79) 2) After the cultivation was finished, the cells were counted. The culture solution with a total cell number of 5?10.sup.6 was sucked, and centrifuged at a centrifugal force of 200 g at room temperature for 10 min, the supernatant was discarded. The cells were resuspended with 100 ?L of 4D-Nucleofector electrotransfection solution.

(80) 3) 2 ?g of plasmid mixture (including 1.96 ?g of ?C31 integrase plasmid and 0.04 ?g of overexpression vector) was added and fully and evenly mixed. The mixed electrotransfection mixture was transferred to a special electrotransfection cuvette, and F1-115 program was selected for electrotransfection.

(81) 4) After electrotransfection, 500 ?L of complete culture medium containing IL-2 (RPMI1640, 10% FBS) was added, and turned upside down 2-3 times for uniform mixing. The evenly-mixed cell suspension was transferred to a new 12-well plate and 1.5 mL of complete culture medium was added.

(82) 5) After cultivation for 6 h, the culture plate was centrifuged at a centrifugal force of 140 g for 8 min, the supernatant was discarded and fresh preheated complete culture medium (RPMI1640, 10% FBS) containing IL-2 with a final concentration of 50 U/mL was added.

(83) 6) After cultivation again for 42 h, the culture plate was centrifuged at a centrifugal force of 300 g for 5 min, the supernatant was discarded, and fresh complete culture medium (containing IL-2 with a concentration of 50 U/mL) was added for cultivation.

(84) SdCAR-T cells recombined BFP fluorescent tags. Therefore, the sdCAR transfection rates of CD4.sup.+T cells and CD8.sup.+T cells could be detected by flow cytometry respectively. The detection results are shown in FIG. 7. The results showed that the transfection positive rate of CD4.sup.+T cells was 32.9%, and the positive rate of CD8.sup.+T cells was 34.18%.

EXAMPLE 5

Optimization of the sdCAR Structure

(85) This example described in detail different sdCAR-T cells obtained due to different structures of an extracellular hinge region and a transmembrane region, and used in vitro activation experiment to select the best structure. The detection index was CD69, an activation marker molecule produced on the surface of T cells. In this experiment, the selected target cell was the MSLN positive K562 tumor cell constructed in Example 3.6, and the added switching molecule was FITC. According to the different structures of sdCAR, the sdCAR-T cells were divided into 6 groups. The specific implementation solution is as follows:

(86) group 1: the hinge region is CH2-CH3 sequence of IgG1, the transmembrane region is CD3 sequence, and this sdCAR-T cell is marked as CH2-CH3-3T cell;

(87) group 2: the hinge region is CH2-CH3 sequence of IgG1, the transmembrane region is CD8 sequence, and this sdCAR-T cell is marked as CH2-CH3-8T cell;

(88) group 3: the hinge region is CH2-CH3 sequence of IgG1, the transmembrane region is CD28 sequence, and this sdCAR-T cell is marked as CH2-CH3-28T cell;

(89) group 4: the hinge region is human CD8? sequence, the transmembrane region is CD3 sequence, and this sdCAR-T cell is marked as CD8?-3T cell;

(90) group 5: the hinge region is human CD8? sequence, the transmembrane region is CD8 sequence, and this sdCAR-T cell is marked as CD8?-8T cell; and

(91) group 6: the hinge region is human CD8? sequence, the transmembrane region is CD28 sequence, and this sdCAR-T cell is marked as CD8?-28T cell.

(92) Implementation steps are as follows.

(93) 1) Effector cells and target cells were plated at a ratio of 1:2, and a switching molecule FITC was added to a final concentration of 100 pM.

(94) 2) Gibberellic acid acetoxymethyl ester was added respectively for culturing overnight (8 h), and centrifuged at a centrifugal force of 800 g for 5 min, then the cells were collected.

(95) 3) The collected cells were directly labelled with flow antibody CD69, and flow cytometry was performed on activated T cells.

(96) The CD69 test results are shown in FIG. 8. The results showed that the level of CD69 expression obtained by sdCAR-T cells with the hinge region having CD8? sequence was much higher than that obtained by sdCAR-T cells with the hinge region having CH2-CH3 sequence, which indicated that the activity obtained by the sdCAR-T cells using the CD8? sequence was higher than that obtained using the CH2-CH3 sequence. The reason may be that the CD8? sequence has higher extracellular flexibility than the CH2-CH3 sequence and the obtained sdCAR receptor has higher flexibility. In addition, it is also found that the activity of sdCAR-T cells obtained by using a CD8 transmembrane region is higher than that obtained by using other types of transmembrane regions, which may be the result of the combined action of the CD8 transmembrane region and the extracellular CD8? hinge region. In summary, in order to obtain sdCAR-T cells with the best activity, the hinge region is selected to be the CD8? sequence, and the transmembrane region is selected to be the CD8 transmembrane region.

EXAMPLE 6

(97) The Activation Test of sdCAR-T Cells In Vitro

(98) This example described in detail the research on the activation test of sdCAR-T cells in vitro. FIG. 2 is a schematic diagram of the activation principle of sdCAR-T cells. The dection indexes were the number of activation molecules CD69 on the surface of T cell and the levels of cytokines IL-2 and IFN? produced during activation. In this test, the selected target cell was the MSLN positive K562 tumor cell constructed in Example 3.6. A total of four groups were designed (three parallel wells were designed for each group). The specific solution is as follows:

(99) group 1: effector cells are sdCAR-T cells, target cells are MSLN positive K562 cells (MSLN.sup.+K562), and no small molecule drugs are added;

(100) group 2: effector cells are sdCAR-T cells, target cells are MSLN positive K562 cells (MSLN.sup.+K562), and small molecule drug HM-3 is added;

(101) group 3: effector cells are sdCAR-T cells, target cells are MSLN positive K562 cells (MSLN.sup.+K562), and small molecule drug FITC is added; and

(102) group 4: effector cells are sdCAR-T cells, target cells are MSLN positive K562 cells (MSLN.sup.+K562), and small molecule drug FHBM is added.

(103) Implementation steps are as follows.

(104) 1) Effector cells and target cells were resuscitated respectively, and RPMI1640 culture medium was selected for culturing for 24 h. Effector cells and target cells were plated at a ratio of 1:2, and a corresponding switching molecule substance was added to a final concentration of 100 pM.

(105) 2) Gibberellic acid acetoxymethyl ester was respectively added for culturing overnight (8 h), centrifuged at a centrifugal force of 800 g for 5 min, and the cells and supernatant were collected respectively.

(106) 3) The contents of IL-2 and IFN? in the collected supernatant were measured by using an ELISA method respectively.

(107) 4) The collected cells were directly labelled with flow antibody CD69, and flow cytometry was performed to obtain the number of activated T cells.

(108) The test of the level of cytokine IL-2 in a cell supernatant is shown in FIG. 9, and the test of the level of cytokine IFN? in the cell supernatant is shown in FIG. 10. Test results of cytokine IL-2 level: in the test group with the addition of exogenous bifunctional molecule FHBM, the IL-2 level was 789.67 pg/mL, and the IFN? level was 16.13 ng/mL; in the test group with the addition of FITC, the IL-2 level was 734.67 pg/mL, and the IFN? level was 13.00 ng/mL. The levels in the test results of the two groups were much higher than those in other test groups. This indicated that sdCAR-T cells can be activated persistently only when they recognized the tumor antigen MSLN and the exogenous switching molecules simultaneously.

(109) The flow cytometry results of activator CD69 on the surface of sdCAR-T cells are shown in FIG. 11, and FIG. 12 is a corresponding result analysis graph. According to the analysis, the level of CD69 expression in sdCAR-T cells was about 55.67% under the condition of the tumor antigen MSLN and the exogenous molecule FITC. Under the condition of the tumor antigen MSLN and the exogenous bifunctional molecule FHBM, the level of CD69 expression was about 62.33%. Under the conditions of other groups, no significant expression of CD69 was observed, indicating that T cells were not activated, which was consistent with the release level of cytokines. In summary, the constructed sdCAR-T cells are only controlled by MSLN and exogenous switching molecule FITC (or FHBM), and this activation mode enhances the safety and controllability of CAR-T cells.

EXAMPLE 7

The Proliferation Test of sdCAR-T Cells In Vitro

(110) This example described in detail the research on the proliferation test of sdCAR-T cells in vitro. The detection index was that the number of T cells changes over time. In this test, the selected target cell was the MSLN.sup.+ K562 tumor cell constructed in Example 3.6. By adding different switching molecules, the effector cell sdCAR-T and the target cell MSLN.sup.+K562 were co-incubated for 3 d, 4 d and 5 d. The design of the test group is the same as that of Example 5, with specific implementation steps as follows.

(111) 1) The target cell MSLN.sup.+K562 was resuscitated, RPMI1640 culture medium was selected for culturing for 24 h and then treated with mitomycin C for 30 min, so that the target cell loses its proliferation ability.

(112) 2) Effector cells sdCAR-T cells were resuscitated, effector cells and target cells were plated at a ratio of 1:2, and the corresponding switching molecule was added to a final concentration of 100 pM.

(113) 3) At time points of the 3 d, 4 d and 5 d, the cells were collected respectively.

(114) 4) Flow cytometry was used to detect the percentage content of sdCAR-T cells at each time point and a cell counter was used to detect the total cell number at each time point.

(115) 5) The number of sdCAR-T cells at each time point was calculated.

(116) The cell proliferation results are shown in FIG. 13. Analysis showed that sdCAR-T cells could proliferate significantly only in the presence of the exogenous switching molecule FITC or FHBM and the tumor antigen MSLN. Therefore, the exogenous switching molecule can be used to regulate the activity of sdCAR-T cells, and precise regulation of the sdCAR-T cells can be realized through exogenous conditions.

EXAMPLE 8

The Cytotoxicity Test of sdCAR-T Cells In Vitro

(117) This example verified the killing activity of the constructed sdCAR-T cells on tumor cells in vitro, and the selected target cells were MSLN single positive tumor cells and single positive tumor cells of negative control CEA. A total of four test groups were designed, each group had three parallel wells, and the design of the test group was the same as that of Example 6.

(118) The cytotoxicity of sdCAR-T cells was determined according to the survival rate of target cells (MSLN.sup.+K562). The selected target cells were equal mixed cells of MSLN and CEA. Because the constructed sdCAR-T cells had strong tumor specificity, CEA tumor cells were used as a negative control group, and the purpose of setting equal numbers of MSLN single positive cells and CEA single positive cells was to accurately and conveniently calculate the survival rate of MSLN tumor cells in each test group at each time. The survival rate of MSLN target cells in each test group was calculated according to the following formula, and then the tumor killing ability of the constructed CARs-T cells in vitro could be analyzed.

(119) $\mathrm{Survival}\mathrm{rate}\mathrm{of}\mathrm{MSLN}\mathrm{target}\mathrm{cells}=\frac{\mathrm{Proportion}\mathrm{of}\mathrm{MSLN}\mathrm{target}\mathrm{cells}\mathrm{of}\mathrm{the}\mathrm{test}\mathrm{group}}{\mathrm{Proportion}\mathrm{of}\mathrm{CEA}\mathrm{tumor}\mathrm{cells}\mathrm{of}\mathrm{the}\mathrm{test}\mathrm{group}}?100\%$

(120) Specific implementation steps are as follows.

(121) 1) Effector cells and mixed target cells (MSLN.sup.+K562:CEA.sup.+K562, 1:1) were co-incubated at a ratio of 5:2 in a U-shaped 96-well plate, RPMI1640 (10% FBS) was selected as a co-culture medium, and corresponding switching molecule s were added.

(122) 2) Culturing at 37? C. for 22 h with 5% CO.sub.2.

(123) 3) The ratio of MSLN and CEA tumor cells in each test group was detected by a flow cytometer, and then the survival rate of MSLN target cells was calculated.

(124) The fluorescence detection results of in vitro cytotoxicity of sdCAR-T cells are shown in FIG. 14. The fluorescence diagram shows that the percentage content of homologous target cells MSLN.sup.+K562 (green fluorescence) in the FITC or FHBM molecular treatment group was significantly lower than that at the beginning of the test, indicating that sdCAR-T cells had higher cytotoxicity under these two conditions. Flow cytometry was used to quantitatively detect the cytotoxicity of sdCAR-T cells. See FIG. 15 for the test result graph. The cytotoxicity of sdCAR-T cells in each group could be calculated through the flow cytometry result. The results are shown in FIG. 16. The results showed that the survival rate of MSLN tumor cells in the FHBM test group was only 28.74%, while that in the FITC test group was about 30.98%, which was much lower than that in other test groups. In other test groups, the survival rate of MSLN tumor cells was about 100%, that is, sdCAR-T cells had no cytotoxicity. Therefore, only when two receptors of the constructed sdCAR-T cells are simultaneously targeted and recognized by MSLN and FITC can the sdCAR-T cells exert cytotoxicity to a large degree. Therefore, the use of the exogenous small molecule antigen can control the cytotoxicity of CARs-T cells and improve the safety of CAR-T cells.

EXAMPLE 9

Time Regulation of In Vitro Activity of sdCAR-T Cells by Exogenous Switching Molecules

(125) This example verified that the killing activity of the sdCAR-T cells on tumor cells was subjected to time regulation by the switching molecules, and the selected target cells were MSLN single positive tumor cells and single positive tumor cells of negative control CEA. In the 14-hour test, the cytotoxicity of sdCAR-T cells was tested in real time with and without switching molecules. A total of four test groups were designed, each group had three parallel wells, and the design of test group was the same as that of Example 6.

(126) Specific implementation steps are as follows.

(127) 1) Effector cells and mixed target cells (MSLN.sup.+K562:CEA.sup.+K562, 1:1) were co-incubated at a ratio of 5:2 in a 6-well plate, RPMI1640 (containing 10% FBS) was selected as a co-culture medium, cultured at 37? C. with 5% CO.sub.2.

(128) 2) 100 ?L of cell suspension was taken every 2 h as the culturing begins at the 0 h.

(129) 3) Corresponding switching molecules were added at the 4 h.

(130) 4) Co-culturing for 10 h (that is, cell suspension was taken 6 times); after the culture, the ratio of MSLN and CEA tumor cells in each test group was detected by a flow cytometer, and then the survival rate of MSLN target cells was calculated.

(131) The in vitro cytotoxicity of sdCAR-T cells was subjected to time regulation by the exogenous switching molecules, with the test results shown in FIG. 17. The results showed that all groups showed no cytotoxicity without adding switching molecules. After corresponding switching molecules were added, only the FITC or FHBM treatment group had significant cytotoxicity, indicating that the exogenous switching molecules FITC and FHBM can effectively activate the cytotoxicity of sdCAR-T cells to MSLN.sup.+K562 cells.

EXAMPLE 10

Concentration Regulation of In Vitro Activity of sdCAR-T Cells by Exogenous Switching Molecules

(132) This example verified that the killing activity of the sdCAR-T cells on tumor cells was subjected to concentration regulation by the switching molecules, and the selected target cells were MSLN single positive tumor cells and single positive tumor cells of negative control CEA. A total of 0, 0.5, 1, 5, 10, 50, 100, 500 and 1000 pM concentration gradients were designed in the test to detect the cytotoxicity of sdCAR-T cells in each group, with three parallel wells in each group.

(133) Specific implementation steps are as follows.

(134) 1) Effector cells and mixed target cells (MSLN.sup.+K562:CEA.sup.+K562, 1:1) were co-incubated at a ratio of 5:2 in a U-shaped 96-well plate, and RPMI1640 (10% FBS) was selected as a co-culture medium.

(135) 2) Appropriate concentrations of switching molecules (FITC or FHBM) were added to the corresponding groups, and cultured at 37? C. with 5% CO.sub.2 for 22 h.

(136) 3) After the culture was completed, the ratio of surviving MSLN and CEA tumor cells in each test group was detected by a flow cytometer, and then the survival rate of MSLN target cells was calculated.

(137) The in vitro cytotoxicity of sdCAR-T cells was subjected to concentration regulation by the exogenous switching molecules, with the test results shown in FIG. 18. The results showed that the cytotoxicity of sdCAR-T cells increased with the concentration when the switching molecule was at a low level (?100 pM), but when the switching molecule reached a higher concentration level (>100 pM), the cytotoxicity was not proportional to the concentration of the switching molecules, and basically remains at the same level. This may be because the FITC receptor in the sdCAR structure had been recognized and saturated when the switching molecule was greater than 100 pM. Therefore, the activity of the constructed sdCAR-T cells depended strictly on the FITC or FHBM switching molecules. Exogenous small molecules not only can control the activity switching, but also can obtain different levels of cytotoxicity by different added concentrations. This provides a controllable approach for the safe clinical use of CAR-T cells to treat tumors.

EXAMPLE 11

The Tumor Killing Test by sdCAR-T Cells In Vivo

(138) Female nude mice BALB/c which were 6-8 weeks old were selected, MSLN.sup.+K562 cells and CEA.sup.+K562 cells were mixed equally and injected intraperitoneally into the nude mice, and then corresponding effector cells and switching molecules were added respectively. Refer to FIG. 19 for details of the in vivo test process. Six test groups were designed, with 6 nude mice in each group. Specific solutions are as follows:

(139) group 1: effector cells are T cells (No CAR) and no switching molecules are added;

(140) group 2: effector cells are the second generation of CAR-T cells (Conventional CAR) specific to MSLN, and no switching molecules are added;

(141) group 3: effector cells are sdCAR-T cells and no switching molecules are added;

(142) group 4: effector cells are sdCAR-T cells, and small molecule HM-3 is added;

(143) group 5: effector cells are sdCAR-T cells and switching molecule FITC is added; and

(144) group 6: effector cells are sdCAR-T cells, and switching molecule FHBM is added.

(145) After the test was completed, the nude mice were euthanized, and then 5 mL of pre-cooled PBS was injected into the abdominal cavities of the nude mice. After the abdominal cavities of the nude mice were rubbed many times, the abdominal cavity fluid was sucked out completely, and centrifuged at a centrifugal force of 800 g for 8 min. The cells were collected, and the percentage of the two tumor cells was measured by flow cytometry, and then the in vivo activity of sdCAR-T cells could be calculated. In addition, the above centrifuged supernatant was collected to measure the release levels of cytokines IL-2 and IFN? respectively.

(146) In vivo cytotoxicity of sdCAR-T cells was detected by flow cytometry, and the results are shown in FIG. 20. The cytotoxicity of sdCAR-T cells in each group could be calculated through the flow cytometry result. The results are shown in FIG. 21. The results showed that in the FITC or FHBM treatment group, sdCAR-T cells showed significant cytotoxicity, and the cytotoxicity level was similar to that of Conventional CAR-T cells. Meanwhile, it was observed that the cytotoxicity of the FHBM treatment group was better than that of the FITC treatment group. The reasons may be as follows: 1. HM-3, a component of the bifunctional molecule FHBM, has the function of inhibiting tumor development. 2. The FHBM targets both sdCAR-T cells and MSLN.sup.+K562 tumor cells, which may shorten the effective distance between effector cells and target cells and increase the cytotoxicity of effector cells. In other test groups treated with other switching molecules, the obtained survival rate of MSLN tumor cells was similar to that of the group with no CAR-T cells, that is, sdCAR-T cells have no cytotoxicity. Therefore, in vivo, the cytotoxicity of sdCAR-T cells depends on the tumor antigen MSLN and the switching molecule FITC (or FHBM) to improve the safety of CAR-T cells.

(147) The in vivo cytokine IL-2 release level is shown in FIG. 22, and the cytokine IFN? release level is shown in FIG. 23. The trends of release levels of the two cytokines were basically the same. In the FITC or FHBM treatment group, sdCAR-T cells could release significant cytokine levels while effectively killing tumor cells. However, compared with the Conventional CAR-T cell group, the FITC or FHBM treatment group significantly reduced the release level (only half of the level), which indicates that when switching molecules and sdCAR-T cells were used to treat tumors, the level of the cytokine released was lower, and thus CRS toxicity could be effectively alleviated.

(148) In summary, only when the constructed sdCAR-T cells simultaneously recognized the endogenous tumor cell antigen MSLN and the exogenous molecular switch FITC (or FHBM) in vivo, they could activate the cytotoxicity, thus exerting anti-tumor activity. In vivo tests confirmed that the design of the dual chimeric antigen receptor can enhance the tumor specificity of T cells and solve the toxicity of targeted non-tumor. At the same time, the dose of FITC (or FHBM) molecules added can be controlled to regulate the quantity of sdCAR-T cells activated, which can solve the existing treatment toxicity of cytokine storm and obviously improve the application safety.

EXAMPLE 12

Use of sdCAR-T Cells in Solid Tumors

(149) According to the above results of in vitro and in vivo tests of sdCAR-T cells, in order to further verify the therapeutic effect on solid tumors, the selected solid tumors are highly expressed MSLN and integrin. The selected solid tumors mainly include ovarian cancer OV-1063, lung cancer A549, esophageal cancer EC109, pancreatic cancer AsPC-1, gastric cancer MG803, colorectal cancer HT29, breast cancer MDA-MB-231, liver cancer SMCC7721, melanoma Malme-3M, head and neck cancer CAL27, cervical cancer Hela and osteosarcoma U-2OS. The flow antibodies were used to detect the expression of MSLN on the surface of these solid tumor cells respectively. The test results are shown in FIG. 24. The results showed that the selected tumor cells all highly expressed the homologous tumor antigen MSLN, and the tumor antigen was specific to sdCAR-T cells. In addition, flow antibodies of integrins ?v and ?3 were used to detect the expression level of the integrins on the surface of the tumor cells respectively. The test results are shown in FIGS. 25 and 26. The test results showed that the selected tumor cells all highly expressed the integrins a?and ?3. The above test results showed that various tumor cells were suitable for the sdCAR-T cells and the bifunctional molecule FHBM in the present invention.

(150) Female nude mice BALB/c, which were 6-8 weeks old, were selected and randomly divided into seven groups. The foregoing seven tumor cells (1?10.sup.6) were inoculated into the caudal vein respectively. The day when the tumor cells were inoculated was recorded as day 0, and the caudal vein reinfusion of 2?10.sup.6 CAR-T cells and switching molecules FHBM was carried out on day 14. The detection scheme is as follows: the tumor volumes of nude mice in each group were measured every three days starting from day 15. The change trends of tumor volumes of various tumor cells with time are shown in FIG. 27. The results showed that the tumor volumes of the foregoing selected solid tumors were significantly reduced after reinfusion of sdCAR-T cells, indicating that sdCAR-T cells had high cytotoxicity to the foregoing solid tumors. Therefore, the combined cell immunotherapy (including the sdCAR-T cells and the bifunctional switching molecule FHBM) designed by the present invention has better cytotoxicity to tumor cells that highly express integrin ?v?3 and the MSLN antigen.