Anti-pvrig / Anti-tigit bispecific antibody and application

Abstract

Provided is a bispecific antibody capable of specifically binding PVRIG and TIGIT. The bispecific antibody can regulate the function of immune cells and can be used as a drug to treat diseases, such as tumors, related to immune abnormalities.

Description

Table 4. The results of IMGT analysis of chimeric antibodies Antibody No. Light chain CDR Sequences Heavy chain CDR Sequences TIGIT-CHI-002 LCDR1 QNVRTA SEQ ID NO:42 HCDR1 GYSITSDS SEQ ID NO:45 LCDR2 SAS SEQ ID NO:43 HCDR2 ISYSGNT SEQ ID NO:46 LCDR3 QQYYTTPWT HCDR3 ARLDFSNYGGAVDY SEQ ID NO:44 SEQ ID NO:47 TIGIT-CHI-006 LCDR1 QGVSTT SEQ ID NO:48 HCDR1 GYSMTSDY SEQ ID NO:51 LCDR2 SAS SEQ ID NO:49 HCDR2 ISYSGRT SEQ ID NO:52 LCDR3 QQYYSSPFT SEQ ID NO:50 HCDR3 ARGDYSNYGGAMYD SEQ ID NO:53 TIGIT-CHI-005 LCDR1 QHVSNG SEQ ID NO:54 HCDR1 GYAFTNYL SEQ ID NO:57 LCDR2 SAS SEQ ID NO:55 HCDR2 INPGSGGT SEQ ID NO:58 LCDR3 QQHYNTPHT SEQ ID NO:56 HCDR3 ARGEYFFFDY SEQ ID NO:59 TIGIT-CHI-070 LCDR1 ENIYSY SEQ ID NO:60 HCDR1 GYTFTNYY SEQ ID NO:63 LCDR2 NAK SEQ ID NO:61 HCDR2 IDPDSGGS SEQ ID NO:64 LCDR3 QHHYGNPLT SEQ ID NO:62 HCDR3 AREGHYGFYSDY SEQ ID NO:65 (a) Detection of binding activity of anti-TIGIT chimeric antibodies to human TIGIT ECD-mFc and cynomolgus monkey TIGIT ECD-mFc The sheep anti-mouse IgG antibody (Jackson, CAT: 115-006-071) was diluted to 4 gg / mL with PBS (Hyclone, CAT#SH30256, pH7.4), added (50 gL / well) into an enzyme-labeled plate 5 (corning, CAT#9018) and kept at 4°C overnight. After discarding the coating solution, the plate was added with 5% skim milk powder (Sangon Biotech, CAT#A600669-0250)-PBS, 250 gL / well, and incubated at 37°C for 2-4 hours. After washing three times with 0.05% Tween 20-PBS (Sangon Biotech, CAT#A100777-0500, B548117-0500) on the Microplate Washer (Biotek, CAT#405TUS), the plate was added with human TIGIT ECD-mFc (In-house construction, working concentration 10   30 ng / mL) and cynomolgus monkey TIGIT ECD-mFc (In-house construction, working concentration 30 ng / mL), 50 gL per well, incubated at 4°C overnight. The enzyme-labeled plate was then washed with 0.05% Tween 20-PBS for 3 times on the Microplate Washer, added with tested antibodies diluted to 13nM with 1% BSA (diluted by 3-fold serial dilution for 12 concentration points), 50 gL / well, and incubated at 37OC for 1.5-2 hours. The plate was washed 15 with 0.05% Tween 20-PBS for 3 times on the Microplate Washer, then added with HRP enzyme-labeled antibodies (Merck, CAT#AP113P) diluted with 1% BSA (Sangon Biotech, CAT#A500023-0100)-PBS at a dilution ratio of 1:5000, 50 gL / well, and incubated at 37°C for 1 hour. After washing 3 times with 0.05% Tween 20-PBS on the Microplate Washer, the plate was added with TMB chromogenic solution (KPL, CAT#52-00-03), 50 pL / well, and incubated at room temperature for 7-10 minutes. 1M HCL, 50 pL / well, was then added into the plate to terminate the reaction, and the OD450 nm was read by a microplate reader (Biotek, Powerwave HT). The experimental results show that the anti-TIGIT chimeric antibodies can effectively bind to human TIGIT ECD-mFc (Fig. 1) and cynomolgus monkey TIGIT ECD-mFc (Fig. 2), which is consistent with the control antibody, Roche RG6058. (b) The binding activity of anti-TIGIT chimeric antibodies to CHO-K1 human TIGIT (cells with high, medium and low expression levels) and CHO-K1 cynomolgus monkey TIGIT detected by FACS. The collected cells were washed once with PBS (Hyclone, CAT#SH30256), resuspended to 2^105 / 50 pL with 1% BSA-PBS. The antibodies were diluted to 80 nM with 1% BSA-PBS (diluted by 3-fold serial dilution for 12 concentration points). The 50 pL cells were then mixed with 50 pL diluted antibodies, and then incubated at 4OC for 60 minutes. The cells were washed twice with PBS, resuspended with Alexa Fluor® 647 fluorescein labeled secondary antibody (1:800) (Jackson, CAT# 109-605-088), 100 pL / well, and incubated at 4°C for 40 minutes. After washing twice with PBS, the cell samples were resuspended with 1% BSA-PBS, 100 pL / well and then analyzed by flow cytometry (BD, Canto D). The experimental results show that the anti-TIGIT chimeric antibodies can effectively bind to CHO-K1 human TIGIT high expression cell strains (Fig. 3), CHO-K1 human TIGIT medium expression cell strains (Fig. 4), CHO-K1 human TIGIT low expression cell strains (Fig. 5) and CHO-K1 cynomolgus monkey TIGIT cell strains (Fig. 6), which is consistent with the control antibody, Roche RG6058. (c) The effect of anti-TIGIT chimeric antibodies blocking the interaction between Bio-CD155-His and CHO-K1 human TIGIT high expression cell strains detected by FACS The collected cells were washed once with PBS (Hyclone, CAT#SH30256), resuspended to 2x105 / 40 pL with 1% BSA-PBS. The antibodies were diluted to 210 nM with 1% BSA-PBS (diluted by 3-fold serial dilution for 12 concentration points). The Bio-CD155-His (Sino Biological Inc., 10109-H08H) was diluted to 3 pg / mL with 1% BSA-PBS. The 40 pL cells were then mixed with 40 pL diluted antibodies and 40 pL diluted Bio-CD155-His, and then incubated at 4OC for 60 minutes. The cells were washed twice with PBS, resuspended with APC labeled streptavidin (dilution ratio 1:1700, Biolegend, 405243), 100 gL / well, and incubated at 4C for 40 minutes. The cell samples were washed twice with PBS, resuspended with 1% BSA-PBS, 100 gL well and then analyzed by flow cytometry (BD, Canto n). As shown in Fig. 7, both the anti-TIGIT chimeric antibodies and the control antibody (RG6058) can effectively block the binding between the Bio-CD155-His protein and CHO-K1- human TIGIT cells. (d) The effect of anti-TIGIT chimeric antibodies blocking the interaction between TIGIT ECD-mFc and CHO-K1 CD155 detected by FACS The collected CHO-K1 CD155 cells were washed once with PBS (Hyclone, CAT#SH30256) and resuspended to 2*105 / 40 gL with 1% BSA-PBS. The tested antibodies were diluted to 600 nM with 1% BSA-PBS (diluted by 2.5-fold serial dilution for 12 concentration points). The TIGIT ECD-mFc was diluted with 1% BSA-PBS to 6 gg / mL. The 40 gL cells were then mixed with 40 gL diluted antibodies and 40 gL diluted TIGIT ECD-mFc, and then incubated at 4C for 60 minutes. The cells were washed twice with PBS, resuspended with Alexa Fluor® 647 fluorescein labeled secondary antibody (dilution ratio 1:800, Jackson, CAT# 109-605-003), 100 gL / well, and incubated at 4C for 40 minutes. The cell samples were washed twice with PBS, resuspended to 100 gL / well with 1% BSA-PBS and then analyzed by flow cytometry (BD, Canto n). As shown in Fig. 8, both the anti-TIGIT chimeric antibodies and the control antibody (RG6058) can effectively block the binding between the human TIGIT protein and CHO-K1-CD155 cells. Table 5 shows a summary of the identification of TIGIT chimeric antibodies Table 5. A summary of the identification of TIGIT chimeric antibodies TIGIT chimeric antibody Human TIGIT ECD-His Human TIGIT ECD-mFc / CH O-K1 CD155 blocking experim ent Bio- CD155-His / CH O-K1 Human TIGIT blocking experim ent CHO- K1 Human TIGIT (high expres sion strain ) CHO- K1 Human TIGIT (medi um expressi on strain.) CHO- K1 Human TIGIT (low expressi on strain.) Human TIGIT ECD-mFc binding experim ent CHO- K1 cynomo lgus monkey TIGIT binding experim ent cynomo lgus monkey TIGIT ECD- mFc binding experim ent IC50 IC50 EC50 EC50 EC50 EC50 EC50 EC50 ka (1 / Ms)    kd (1 / s)    KD (M) (nM) (nM) (nM) (nM) (nM) (nM) (nM) (nM) TIGIT-CHI- 002 4.53E+06 3.15E-04 6.97E- 11 0.3682 0.1519 0.2991 0.2904 0.2183 0.0181 0.0290 0.0165 TIGIT-CHI- 005 3.59E+06 1.92E-04 5.36E- 11 0.3628 0.1582 0.3421 0.3160 0.2290 0.0173 0.0264 0.0170 TIGIT-CHI- 006 1.29E+06 1.70E-04 1.32E- 10 0.5723 0.2465 0.5022 0.4810 0.3287 0.0167 0.0731 0.0389 TIGIT-CHI- 070 1.52E+06 4.56E-04 3.00E- 10 0.5819 0.3837 0.5168 0.3941 0.3524 0.0145 0.0359 0.0163 RG6058-hIgG1 9.03E+05 1.47E-04 1.63E- 10 0.5658 0.3884 0.6731 0.6164 0.4953 0.0151 0.0509 0.0158 (e) The detection of the expression of PVRIG and TIGIT on the surface of NK cells and the expression of PVR and PVRL2 on the surface of WIDR cells The expression of PVRIG and TIGIT on NK cells (Natural killer cells) was detected by FACS. First, the NK cells were counted by using a cell counter (Beckman Coulter, Vi-CELL). The NK cells were added into three flow cytometry tubes with 1E+5 NK cells in each tube and washed twice with PBS. After discarding the supernatant, one tube was added with 300 uL Staining buffer (PBS+2% FBS) as an unstained tube, and the other two tubes were added with 100 uL staining solution (PBS+1* Zombie Violet (Biolegend, 423114)), mixed well and then incubated at room temperature for 15 minutes. The cells were then washed twice with Staining buffer, and the supernatant was discarded. 50 uL of Fc blocker (Staining buffer + FCX blocker (Biolegend, 422302)) was added to each tube, mixed and then incubated at 4OC for 15 minutes. Each tube was then added with different staining solution, mixed and incubated at 4°C for 30 minutes, wherein, the first tube was added with 50 uL of 2* staining solution (Staining buffer + PE-Cy7 Mouse anti-hCD3 detection antibody + PE Mouse anti-hCD56 detection antibody + APC Mouse anti-hTIGIT detection antibody + AF488 Rabbit anti-hPVRIG detection antibody, CD3 detection antibody: Biolegend 300316, CD56 detection antibody: Biolegend 318306, TIGIT detection antibody: Biolegend 372706, PVRIG detection antibody: RD FAB93651G), and the second tube was added with 50 uL of 2* isotype control staining solution (Staining buffer+PE-Cy7 Mouse anti-hCD3 detection antibody + PE Mouse anti-hCD56 detection antibody + APC Mouse IgG2a k isotype control antibody + AF488 Rabbit IgG k isotype control antibody, APC mIgG2a k isotype control antibody: Biolegend 400222, AF488 Rabbit IgG k isotype control antibody: RD IC1051G). The samples were then washed twice with Staining buffer, centrifuged and mixed with 300 pL Staining buffer. The samples were detected by a flow cytometer (Thermo Attune NxT) to obtain the percentage of the CD56-positive and CD3-negative cell population in the Zombie Violet negative cell population and read the signals of APC and AF488 of the CD56-positive and CD3-negative cell population in the Zombie Violet negative cell population. Panel A in Figure 9 shows that PVRIG and TIGTI are expressed on the surface of NK cells from different donors (donor-010 and donor-050). The expression of PVR and PVRL2 on WIDR cells was detected by FACS. First, the WIDR cells were trypsinized to form a cell suspension, and counted by using a cell counter (Beckman Coulter, Vi-CELL). The WIDR cells were added into three flow cytometry tubes with 1E+5 cells in each tube and washed twice with PBS. After centrifugation and discarding the supernatant, one tube was added with 300 pL Staining buffer (PBS+2% FBS) as an unstained tube, and the other two tubes were added with 100 pL staining solution (PBS+1* Zombie Violet (Biolegend, 423114)), mixed well and then incubated at room temperature for 15 minutes. The cells were then washed twice with Staining buffer, and the supernatant was discarded. Each tube was then added with different staining solution, mixed and incubated at 4°C for 30 minutes, wherein, the first tube was added with 100 pL staining solution (Staining buffer + PerCP-Cy5.5 Mouse anti-hPVR detection antibody + APC Mouse anti-hPVRL2 detection antibody, PVR detection antibody: Biolegend 337612, PVRL2 detection antibody: Biolegend 337412), and the second tube was added with 100pL isotype control staining solution (Staining buffer + PerCP-Cy5.5 Mouse IgG1 k isotype control antibody + APC k Mouse IgG1 isotype control antibody, PerCP-Cy5.5 mIgG1 k isotype control antibody: Biolegend 400150 , APC mIgG1 k isotype control antibody: Biolegend 400122). The samples were then washed twice with Staining buffer, centrifuged and mixed with 300 pL Staining buffer. The samples were detected by a flow cytometer (Thermo Attune NxT) to read the signals of PerCP-Cy5.5 and APC of the Zombie Violet negative cell population. Panel B in Figure 9 shows that PVR and PVRL2 are highly expressed on the surface of WIDR cells. (f) The promoting effect of anti-TIGIT chimeric antibodies on NK cell function detected by NK cell degranulation assay FACS was used to detect the CD107a signal of NK cells (Natural killer cell), which indicated the effect of the tested antibody on the degranulation process of NK cells. The PBMCs were resuscitated one day in advance and NK cells were sorted (Stemcell, 17955). The NK cells were added with 200 IU / ml h-IL2 (RD, 202-IL) and 10 ng / ml h-IL12 (Peprotech, 200-12-50UG) to stimulate overnight and plated for the following experiment on the next day. Firstly, the antibodies were diluted to the highest concentration of 275 nM (four-fold concentration) with assay buffer (RPMI1640-Glutamax + 10% FBS + 1*P / S) and continually diluted with assay buffer by 10-fold gradient dilution. The diluted antibodies were added to an ultra-low attachment, 96-well, U-bottom plate (Costar, 7007), 50 ul per well, for further use. Secondly, the NK cells were counted by using a cell counter (Beckman Coulter, Vi-CELL). A certain number of NK cells were centrifuged at a speed of 350 g for 5 minutes, resuspended to a density of 0.5E+6 cells / ml with assay buffer after discarding the supernatant and added with protein transport inhibitor (Invitrogen, 00498093) and APC mouse anti-human detection antibody (Biolegend, 328620). The antibody coated 96-well, U-bottom plate was added with the treated NK cell suspension (50 gl per well), mixed well and incubated at room temperature for 15 minutes. During the incubation, the target WIDR cells were trypsinized to form a cell suspension and counted by using a cell counter (Beckman Coulter, Vi-CELL). An appropriate number of cells were centrifuged at 200 g for 5 minutes, and resuspended to a density of the 0.25E+6 cells / ml with the assay buffer after discarding the supernatant. After incubation, the target cell suspension was added to the plate, 100 gL per well. At this time, each well contained 25,000 NK cells, 25,000 target cells and different concentrations of tested antibodies. The wells only containing NK cells were served as resting control, while the wells containing NK cells and target cells were served as drug-free control. All the wells were mixed well and incubated in an incubator at 37°C for 16 hours. The change of CD107a was detected by FACS analysis: the cells in the plate were transferred to the same position in a 96-well, V-bottom plate of 96 wells and washed twice with PBS. After discarding the supernatant, each well was added with the staining solution (PBS+2%FBS+1*zombie violet (Biolegend, 423114) + PE mouse anti-CD56 detection antibody (Biolegend, 318306)), mixed well and incubated at 4°C for 30 minutes. Each well was then washed twice with staining buffer and resuspended with 150 gL staining buffer after discarding the supernatant. The samples were detected by a flow cytometer (Thermo Attune NxT) to read the percentage of CD107a strong positive cell population in CD56 positive cells. A higher percentage of CD107a strong positive cells represent a stronger degranulation of NK cells and a higher degree of activation of NK cells. Figure 10 shows that the negative control, anti-HEL-hIgG1, has no effect on CD107a on the surface of NK. The three tested antibodies can improve the expression of CD107a on NK cells to varying degrees, which indicates that the three tested antibodies can effectively promote the activation of NK cells. (g) The promoting effect of anti-TIGIT chimeric antibodies on cytotoxicity for target cells by NK cells detected by NK cell cytotoxicity assay The lysis level of target cells (WIDR) was detected by FACS to infer the effect of the tested antibodies on the cytotoxicity of NK (Natural killer) cells. The PBMCs were resuscitated one day in advance. The NK cells sorted from PBMCs (Stemcell, 17955) were added with 200 IU / mL h-IL2 (RD, 202-IL) and 10 ng / mL h-IL12 (Peprotech, 200-12-50UG) to stimulate overnight and plated for the following experiment on the next day. Firstly, the antibodies were diluted to the highest concentration of 275 nM (four-fold concentration) with assay buffer (RPMI1640-Glutamax + 10% FBS + 1*P / S) and continually diluted with assay buffer by 10-fold gradient dilution. The diluted antibodies were added to an ultra-low attachment, 96-well, U-bottom plate (Costar, 7007), 50 gL per well, for further use. The target cells (WIDR) were trypsinized to form a cell suspension and counted by using a cell counter (Beckman Coulter, Vi-CELL). An appropriate number of WIDR cells were centrifuged at a speed of 200 g for 5 minutes, resuspended with PBS after discarding the supernatant and added with staining solution, CellTrace Violet (Invitrogen, C34557A) with a final concentration of 5 gM. The WIDR suspension with staining solution was mixed evenly, placed in an incubator at 37°C for 10 minutes and shaken during the incubation. At the same time, NK cells were counted with a cell counter. A certain number of NK cells were centrifuged at a speed of 350 g for 5 minutes and resuspended to a density of 0.5E+6 cells / mL with the assay buffer after discarding the supernatant. The antibody coated 96-well, U-bottom plate was added with the treated NK cell suspension (50 gL per well), mixed well and incubated at room temperature for 15 minutes. After the WIDR cell staining, 5 times volume of complete medium (MEM+10%FBS+1*P / S+1*non-essential amino acid+1*sodium glutamate) was added to the cell suspension to stop the reaction. The WIDR cells were centrifuged at a speed of 200 g for 5 minutes and resuspended to a density of 0.25E+6 cells / mL with the assay buffer after discarding the supernatant. After incubating with antibodies and NK cells, the plate was added with WIDR cell suspension, 100 pL per well. At this time, each well contained 25,000 NK cells, 25,000 WIDR cells and various concentrations of tested antibodies. The wells only containing WIDR cells were served as resting control, while the wells containing NK cells and WIDR cells were served as drug- free control. All the wells were mixed well and incubated in an incubator at 37°C for 4 hours. The lysis level of WIDR cells was detected by FACS analysis: each well was added with staining solution (PBS + PI (Propidium Iodide, Invitrogen, P3566)), mixed well and incubated at room temperature for 20 minutes. The samples were then detected by a flow cytometer (Thermo Attune NxT) to read the percentage of PI positive cell population in CTV positive cells. More percentage of PI positive cells represent stronger NK cell cytotoxicity. Figure 11 shows that the negative control, anti-HEL-hIgG1, has no significant effect on the NK cell cytotoxicity. The 4 tested chimeric antibodies can effectively promote the NK cell cytotoxicity for the target cells (WIDR), and the promoting effect of the 4 chimeric antibodies on the NK cell cytotoxicity is not weaker than that of the positive control, RG6058-hIgG1. (h) The effect of anti-TIGIT chimeric antibody on improving the functions of antigen-specific CD8 T cells detected by CMV antigen-recall assay The PBMCs from Anti-CMV IgG positive donor were induced by CMV pp65 (495-503) polypeptide to produce CMV pp65 specific CD8 T cells, which were served as effector cells. The colo205 tumor cell line pulsed with pp65 was used as the target cell. In such experimental system, the effect of TIGIT antibody on improving the functions of the antigen-specific CD8 T cell was detected. The PBMCs were resuscitated, resuspended to 2*106 / mL by using complete medium (RPMI1640-Glutamax + 5% AB serum + 1% P / S+(1*) 2—p mercaptoethanol) containing 1 mg / mL CMV pp65(495-503) peptide (Anaspec, Cat No.AS-28328), 2 ng / mL human IL-2 (R&D, Cat No.IL-202) and 10 ng / mL human IL-7 (Peprotech, Cat No.200-07), inoculated in a 6-well plate (5 mL / well) and incubated at 37°C with 5% CO2 for 6 days. On day 6, all the PBMCs were collected, and pp65 and IL-7 in the medium were removed. The cells were divided into two portions, resuspended in complete medium containing 100 IU / mL human IL-2, and cultured for another 2 days. On day 8, all the PBMCs were collected and resuspended in complete medium containing 100 lU / mL human IL-2, and the cell density was adjusted to 2xl06 mL for continuing cultivation. On day 11, all the PBMCs were collected. The percentage of CD8 T cells and CMV pp65 (495503) specific CD8 T cells in PBMCs (Figure 12, panel A), and the expression of PVRIG, TIGIT and PD-1 on CMV pp65 (495-503) specific CD8 T cells (Figure 12, panel B) were detected by flow cytometry. The detection antibodies were as follows: Livedead Near IR (Invitrogen, Cat No.L34976), CD8-PerCp Cy5.5 (BD, Cat No.565310), CD3-PE-Cy7 (Biolegend, Cat No.300316), T-select HLA-A*0201 CMV pp65 Tetramer-PE (MBL, Cat No.TS-0010-1C), PVRIG-AF488 (R&D, Cat No.FAB93651G-100UG), TIGIT-APC (Biolegend, Cat No.372706) and PD-1-BV421 (BD, Cat No.562516). CD8 T cells were isolated from the induced PBMCs as effector cells by CD8 T cell isolation kit (Stemcell, Cat No.17953) and resuspended with AIM-V to the cell density of 0.4x106 / mL. The target cells, Colo205, were digested by TrypLETM Express Enzyme (Gibco, Cat No.12605010), resuspended in AIM-V (Gibco, Cat No.31035-025) containing 50 ng / mL pp65 to the cell density of 1x106 / mL and treated at 37OC with 5% CO2 for 1 -2 hours. The target cells were then centrifuged at 250 g for 5 minutes and resuspended in AIM-V to the cell density of 0.2 x 106 / mL after discarding the supernatant. The highly expressed PVRL2 and PVR on Colo205 cells were detected by flow cytometry, as shown in panel C, Fig. 12. Anti-TIGIT antibodies or positive control were diluted with AIM-V to 280 nM. The low-attachment, 96-well, U-bottom plate (Corning, Cat No.7007) was added with 50 pL of antibody, 50 pL of CD8 T cells and 100 pL of pp65-treated colo205 cells in order, mixed well with a multichannel pipette, and incubated at 37°C with 5% CO2 for 18 hours. In the experimental system, the final concentration of antibodies was 70 nM, CD8 T cells were 20,000 / well, and colo205 were 20,000 / well. After incubation, the supernatant was collected by centrifugation at 400 g, and the level of human IFN-y in the supernatant was detected with an ELISA kit (Dakewe, Cat No.1110003). In this system, the positive control was RG6058-hIgG1, and the negative control was no treatment. As shown in Fig. 12 (panel D), the secretion level of IFN-y in the cell supernatant treated with 3 TIGIT tested antibodies has no statistically significant difference, compared with positive control, RG6058-hIgG1, but is significantly higher than the no treatment group. The percentage on each histogram represents the percentage of IFN-y secretion increased in the experimental group compared with no treatment group. The expression of PVRL2 and PVR on colo205 was detected by flow cytometry using following detection antibodies: livedead-BV421 (Invitrogen, Cat No.L34964), PVRL2-APC (Biolegend, Cat No.337412), PVR-PerCp Cy5.5 (Biolegend, Cat No.337612) and PD-L1- PE-Cy7 (BD, Cat No.558017). EXAMPLE 5 - Humanization of anti-human TIGIT monoclonal antibodies By comparing the IMGT (http: / / imgt.cines.fr) human antibody heavy and light chain variable region germline gene database, the heavy chain and light chain variable region germline genes with high homology with mouse antibody were selected as templates. The CDRs of mouse antibody were grafted into corresponding human templates to form a variable region sequence of FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. The key amino acids in the skeleton sequence were back mutated to the amino acids corresponding to the murine antibody as needed to ensure the original affinity, and thus to obtain the humanized anti-TIGIT monoclonal antibody. The CDR amino acid residues of the antibody were usually determined and annotated by the Kabat numbering scheme. 1. Humanization of TIGIT-002 The humanized light chain templates of the mouse antibody (TIGIT-002) were IGKV2-29*02 / IGKV4-1*01 and IGKJ4*01, and the humanized heavy chain templates are IGHV4-38-2*01 and IGHJ6*01. The CDRs of the mouse antibody (TIGIT-002) were grafted into the human template thereof to obtain the corresponding humanized version. The key amino acids in the FR region sequence in the humanized antibody of TIGIT-002 were back mutated to the amino acids corresponding to the mouse antibody as needed to ensure the original affinity. The specific back mutation design is shown in Table 6. Table 6. The back mutation design of humanized antibody of TIGIT-002 VL VH VL1 Grafted (IGKV2-29*02) + L47M VH1 Grafted (IGHV4-38-2*01) + S30T, V71R VL2 Grafted (IGKV2-29*02) + L37Q, L47M VH2 Grafted (IGHV4-38-2*01) + S30T, G44K, V71R VL3 Grafted (IGKV4-1*01) + P43S, L47M VH3 Grafted (IGHV4-38-2*01) + S30T, G44K, I48M, V67I, V71R - - VH4 Grafted (IGHV4-38-2*01) + S30T, G44K, W47Y, V71R Note: Grafted (IGKV2-29*02) represents the insertion of CDRs of the mouse antibody into the FR region sequence in human germline IGKV2-29*02; L47M represents the L at position 47 of Grafted (IGKV2-29*02) is back mutated to M, and so on. The back-mutated amino acids are numbered in natural order. The specific variable region sequences in humanized antibody of TIGIT-002 are as follows: The amino acid sequence of TIGIT-002.VL1 is shown in SEQ ID NO: 66: DIVMTQTPLSLSVTPGQPASISCKASQNVRTAVAWYLQKPGQSPQLMIYSASYRYTGVPD RFSGSGSGTDFTLKISRVEAEDVGVYYCQQYYTTPWTFGGGTKVEIK The amino acid sequence of TIGIT-002.VL2 is shown in SEQ ID NO: 67: DIVMTQTPLSLSVTPGQPASISCKASQNVRTAVAWYQQKPGQSPQLMIYSASYRYTGVPD RFSGSGSGTDFTLKISRVEAEDVGVYYCQQYYTTPWTFGGGTKVEIK The amino acid sequence of TIGIT-002.VL3 is shown in SEQ ID NO: 68: DIVMTQSPDSLAVSLGERATINCKASQNVRTAVAWYQQKPGQSPKLMIYSASYRYTGVP DRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYTTPWTFGGGTKVEIK The amino acid sequence of TIGIT-002.VH1 is shown in SEQ ID NO: 69: EVQLQESGPGLVKPSETLSLTCAVSGYSITSDSWNWIRQPPGKGLEWIGYISYSGNTYYNP SLKSRVTISRDTSKNQFSLKLSSVTAADTAVYYCARLDFSNYGGAVDYWGQGTTVTVSS The amino acid sequence of TIGIT-002.VH2 is shown in SEQ ID NO: 70: EVQLQESGPGLVKPSETLSLTCAVSGYSITSDSWNWIRQPPGKKLEWIGYISYSGNTYYNP SLKSRVTISRDTSKNQFSLKLSSVTAADTAVYYCARLDFSNYGGAVDYWGQGTTVTVSS The amino acid sequence of TIGIT-002.VH3 is shown in SEQ ID NO: 71: EVQLQESGPGLVKPSETLSLTCAVSGYSITSDSWNWIRQPPGKKLEWMGYISYSGNTYYN PSLKSRITISRDTSKNQFSLKLSSVTAADTAVYYCARLDFSNYGGAVDYWGQGTTVTVSS The amino acid sequence of TIGIT-002.VH4 is shown in SEQ ID NO: 72: EVQLQESGPGLVKPSETLSLTCAVSGYSITSDSWNWIRQPPGKKLEYIGYISYSGNTYYNP SLKSRVTISRDTSKNQFSLKLSSVTAADTAVYYCARLDFSNYGGAVDYWGQGTTVTVSS The amino acid sequence of the humanized light chain template IGKV2-29*02 is shown in SEQ ID NO: 73: DIVMTQTPLSLSVTPGQPASISCKSSQSLLHSDGKTYLYWYLQKPGQSPQLLIYEVSSRFS GVPDRFSGSGSGTDFTLKISRVEAEDVGVYYCMQGIHLP The amino acid sequence of the humanized light chain template IGKV4-1*01 is shown in SEQ ID NO: 74: DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTR ESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYSTP The amino acid sequence of the humanized light chain template IGKJ4*01 is shown in SEQ ID NO: 75: FGGGTKVEIK The amino acid sequence of the humanized heavy chain template IGHV4-38-2*01 is shown in SEQ ID NO: 76: QVQLQESGPGLVKPSETLSLTCAVSGYSISSGYYWGWIRQPPGKGLEWIGSIYHSGSTYY NPSLKSRVTISVDTSKNQFSLKLSSVTAADTAVYYCAR The amino acid sequence of the humanized heavy chain template IGHJ6*01 is shown in SEQ ID NO: 77: WGQGTTVTVSS Based on the above-mentioned back mutation design in light chain and heavy chain variable regions in humanized antibody of TIGIT-002, the inventors selected different sequences of light chain and heavy chain for cross combination and finally obtained a variety of humanized antibodies of TIGIT-002. The amino acid sequences of the variable region in various antibodies are as follows: Table 7. Amino acid sequences corresponding to variable region in humanized antibodies of TIGIT-002 Fv 002.VH1 002.VH2 002.VH3 002.VH4 002.VL1 TIGIT-002- H1L1* TIGIT-002-H2L1 TIGIT-002-H3L1 TIGIT-002-H4L1 002.VL2 TIGIT-002-H1L2 TIGIT-002-H2L2 TIGIT-002-H3L2 TIGIT-002-H4L2 002.VL3 TIGIT-002-H1L3 TIGIT-002-H2L3 TIGIT-002-H3L3 TIGIT-002-H4L3 *TIGIT-002-H1L1 indicates that the heavy chain is selected from TIGIT-002.VH1 and the light chain is selected from TIGIT-002.VL1. The same below. According to Kabat numbering scheme, the analysis results of VH and VL sequences in above 12 humanized antibodies are shown in table 8. Table 8 Kabat analysis results of VH and VL sequences in humanized antibody of TIGIT- 002 Light / heavy chain variable regions CDR1 CDR2 CDR3 VH1 / 2 / 3 / 4 SDSWN SEQ ID NO:21 YISYSGNTYYNPSLKS SEQ ID NO:22 LDFSNYGGAVDY SEQ ID NO:23 VL1 / 2 / 3 KASQNVRTAVA SASYRYT QQYYTTPWT SEQ ID NO:18 SEQ ID NO:19 SEQ ID NO:20 2. Humanization of TIGIT-006 The humanized light chain templates of the mouse antibody (TIGIT-006) were IGKV2-29*02 / IGKV4-1*01 and IGKJ4*01, and the humanized heavy chain templates are IGHV4-38-2*01 and IGHJ6*01. The CDRs of the mouse antibody (TIGIT-006) were grafted into the human template thereof to obtain the corresponding humanized version. The key amino acids in the FR region sequence in the humanized antibody of TIGIT-006 were back mutated to the amino acids corresponding to the mouse antibody as needed to ensure the original affinity. The specific back mutation design is shown in Table 9. Table 9. The back mutation design of humanized antibody of TIGIT-006 VL VH VL1 Grafted(IGKV2-29*02) VH1 Grafted(IGHV4-38-2*01) + S30T,V71R VL2 Grafted(IGKV2-29*02) + L37Q,Q45K VH2 Grafted(IGHV4-38-2*01) + I29M,S30T,V71R VL3 Grafted(IGKV4-1*01) + P43S VH3 Grafted(IGHV4-38-2*01) + I29M,S30T,G44K,V71R - - VH4 Grafted(IGHV4-38-2*01) + I29M,S30T,G44K,I48M,V67I,V71R VH5 Grafted(IGHV4-38-2*01) + I29M,S30T,G44K,W47Y,V71R Note: Grafted (IGKV2-29*02) represents the insertion of CDRs of the mouse antibody into the FR region sequence in human germline IGKV2-29*02; L37Q represents the L at position 37 of Grafted (IGKV2-29*02) is back mutated to Q, and so on. The back-mutated amino acids are numbered in natural order. The specific variable region sequences in humanized antibody of TIGIT-006 are as follows: The amino acid sequence of TIGIT-006.VL1 is shown in SEQ ID NO:78: DIVMTQTPLSLSVTPGQPASISCRASQGVSTTIAWYLQKPGQSPQLLIYSASYRYTGVPDR FSGSGSGTDFTLKISRVEAEDVGVYYCQQYYSSPFTFGGGTKVEIK The amino acid sequence of TIGIT-006.VL2 is shown in SEQ ID NO:79: DIVMTQTPLSLSVTPGQPASISCRASQGVSTTIAWYQQKPGQSPKLLIYSASYRYTGVPDR FSGSGSGTDFTLKISRVEAEDVGVYYCQQYYSSPFTFGGGTKVEIK The amino acid sequence of TIGIT-006.VL3 is shown in SEQ ID NO:80: DIVMTQSPDSLAVSLGERATINCRASQGVSTTIAWYQQKPGQSPKLLIYSASYRYTGVPRF SGSGSGTDFTLTISSLQAEDVAVYYCQQYYSSPFTFGGGTKVEIK The amino acid sequence of TIGIT-006.VH1 is shown in SEQ ID NO:81: EVQLQESGPGLVKPSETLSLTCAVSGYSITSDYWNWIRQPPGKGLEWIGYISYSGRTYYN PSLKSRVTISRDTSKNQFSLKLSSVTAADTAVYYCARGDYSNYGGAMYDWGQGTTVTV SS The amino acid sequence of TIGIT-006.VH2 is shown in SEQ ID NO:82: EVQLQESGPGLVKPSETLSLTCAVSGYSMTSDYWNWIRQPPGKGLEWIGYISYSGRTYY NPSLKSRVTISRDTSKNQFSLKLSSVTAADTAVYYCARGDYSNYGGAMYDWGQGTTVT VSS The amino acid sequence of TIGIT-006.VH3 is shown in SEQ ID NO:83: EVQLQESGPGLVKPSETLSLTCAVSGYSMTSDYWNWIRQPPGKKLEWIGYISYSGRTYY NPSLKSRVTISRDTSKNQFSLKLSSVTAADTAVYYCARGDYSNYGGAMYDWGQGTTVT VSS The amino acid sequence of TIGIT-006.VH4 is shown in SEQ ID NO:84: EVQLQESGPGLVKPSETLSLTCAVSGYSMTSDYWNWIRQPPGKKLEWMGYISYSGRTYY NPSLKSRITISRDTSKNQFSLKLSSVTAADTAVYYCARGDYSNYGGAMYDWGQGTTVTV SS The amino acid sequence of TIGIT-006.VH5 is shown in SEQ ID NO:85: EVQLQESGPGLVKPSETLSLTCAVSGYSMTSDYWNWIRQPPGKKLEYIGYISYSGRTYYN PSLKSRVTISRDTSKNQFSLKLSSVTAADTAVYYCARGDYSNYGGAMYDWGQGTTVTV SS The amino acid sequence of the humanized light chain template IGKV2-29*02 is shown in SEQ ID NO:73: DIVMTQTPLSLSVTPGQPASISCKSSQSLLHSDGKTYLYWYLQKPGQSPQLLIYEVSSRFS GVPDRFSGSGSGTDFTLKISRVEAEDVGVYYCMQGIHLP The amino acid sequence of the humanized light chain template IGKV4-1*01 is shown in SEQ ID NO:74: DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTR ESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYSTP The amino acid sequence of the humanized light chain template IGKJ4*01 is shown in SEQ ID NO:75: FGGGTKVEIK The amino acid sequence of the humanized heavy chain template IGHV4-38-2*01 is shown in SEQ ID NO:76: QVQLQESGPGLVKPSETLSLTCAVSGYSISSGYYWGWIRQPPGKGLEWIGSIYHSGSTYY NPSLKSRVTISVDTSKNQFSLKLSSVTAADTAVYYCAR The amino acid sequence of the humanized heavy chain template IGHJ6*01 is shown in SEQ ID NO:77: WGQGTTVTVSS Based on the above-mentioned back mutation design in light chain and heavy chain variable regions in humanized antibody of TIGIT-006, the inventors selected different sequences of light chain and heavy chain for cross combination and finally obtained a variety of humanized antibodies of TIGIT-006. The amino acid sequences of the variable region in various antibodies are as follows: Table 10. Amino acid sequences corresponding to variable region in humanized antibodies of TIGIT-006 Fv 006.VH1 006.VH2 006.VH3 006.VH4 006.VH5 006.VL1 TIGIT-006-H1L1* TIGIT-006-H2L1 TIGIT-006-H3L1 TIGIT-006-H4L1 TIGIT-006-H5L1 006.VL2 TIGIT-006-H1L2 TIGIT-006-H2L2 TIGIT-006-H3L2 TIGIT-006-H4L2 TIGIT-006-H5L2 006.VL3 TIGIT-006-H1L3 TIGIT-006-H2L3 TIGIT-006-H3L3 TIGIT-006-H4L3 TIGIT-006-H5L3 *TIGIT-006-H1L1 indicates that the heavy chain is selected from TIGIT-006.VH1, and the light chain is selected from TIGIT-006.VL1. The same below. According to Kabat numbering scheme, the analysis results of VH and VL sequences in above 15 humanized antibodies are shown in Table 11. Table 11. Kabat analysis results of VH and VL sequences in humanized antibody of TIGIT-006 Light / heavy chain variable regions CDR1 CDR2 CDR3 VH1 / 2 / 3 / 4 / 5 SDYWN SEQ ID NO:27 YISYSGRTYYNPSLKS SEQ ID NO:28 GDYSNYGGAMYD SEQ ID NO:29 VL1 / 2 / 3 RASQGVSTTIA SEQ ID NO:24 SASYRYT SEQ ID NO:25 QQYYSSPFT SEQ ID NO:26 3. Humanization of TIGIT-005 The humanized light chain templates of the mouse antibody TIGIT-005 were IGKV4-1*01 and IGKJ4*01, and the humanized heavy chain templates are IGHV1-3*01 and IGHJ6*01. The CDRs of the mouse antibody TIGIT-005 were grafted into the human template thereof to obtain the corresponding humanized version. The key amino acids in the FR region sequence in the humanized antibody of TIGIT-005 were back mutated to the amino acids corresponding to the mouse antibody as needed to ensure the original affinity. The specific back mutation design is shown in Table 12. Table 12. The back mutation design of humanized antibody of TIGIT-005 VL VH VL1 Grafted(IGKV4-1*01) + VH2 Grafted(IGHV1-3*01) + Q38H,P43S T28A,R72A,T74K,A76S Note: Grafted (IGKV4-1*01) represents the insertion of CDRs of the mouse antibody into the FR region sequence in human germline IGKV4-1*01; Q38H represents the Q at position 38 of Grafted (IGKV4-1*01) is back mutated to H, and so on. The back-mutated amino acids are numbered in natural order. The specific variable region sequences in humanized antibody of TIGIT-005 are as follows: The amino acid sequence of TIGIT-005.VL1 is shown in SEQ ID NO: 86: DIVMTQSPDSLAVSLGERATINCKASQHVSNGVAWYQHKPGQSPKLLIYSASYRYTGVP DRFSGSGSGTDFTLTISSLQAEDVAVYYCQQHYNTPHTFGGGTKVEIK The amino acid sequence of TIGIT-005.VH2 is shown in SEQ ID NO: 87: EVQLVQSGAEVKKPGASVKVSCKASGYAFTNYLIEWVRQAPGQRLEWMGVINPGSGGT NYKEKFKGRVTITADKSSSTAYMELSSLRSEDTAVYYCARGEYFFFDYWGQGTTVTVSS The amino acid sequence of the humanized light chain template IGKV4-1*01 is shown in SEQ ID NO: 74: DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTR ESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYSTP The amino acid sequence of the humanized light chain template IGKJ4*01 is shown in SEQ ID NO: 75: FGGGTKVEIK The amino acid sequence of the humanized heavy chain template IGHV1-3*01 is shown in SEQ ID NO: 106: QVQLVQSGAEVKKPGASVKVSCKASGYTFTSYAMHWVRQAPGQRLEWMGWINAGNG NTKYSQKFQGRVTITRDTSASTAYMELSSLRSEDTAVYYCAR The amino acid sequence of the humanized heavy chain template IGHJ6*01 is shown in SEQ ID NO: 77: WGQGTTVTVSS TIGIT-005 antibody is prone to have chemical modifications on NG site. Inventors performed point mutations on NG to eliminate the risk of modification. In five of the examples, Inventors mutated the NG of 005.VL1. The mutation sites are shown in bold. The sequences after mutation are as follows: The amino acid sequence of TIGIT-005.VL1a is shown in SEQ ID NO: 88: DIVMTQSPDSLAVSLGERATINCKASQHVSQGVAWYQHKPGQSPKLLIYSASYRYTGVP DRFSGSGSGTDFTLTISSLQAEDVAVYYCQQHYNTPHTFGGGTKVEIK The amino acid sequence of TIGIT-005.VL1b is shown in SEQ ID NO: 89: DIVMTQSPDSLAVSLGERATINCKASQHVSTGVAWYQHKPGQSPKLLIYSASYRYTGVPD RFSGSGSGTDFTLTISSLQAEDVAVYYCQQHYNTPHTFGGGTKVEIK The amino acid sequence of TIGIT-005.VL1c is shown in SEQ ID NO: 90: DIVMTQSPDSLAVSLGERATINCKASQHVSDGVAWYQHKPGQSPKLLIYSASYRYTGVP DRFSGSGSGTDFTLTISSLQAEDVAVYYCQQHYNTPHTFGGGTKVEIK The amino acid sequence of TIGIT-005.VL1d is shown in SEQ ID NO: 91: DIVMTQSPDSLAVSLGERATINCKASQHVSNAVAWYQHKPGQSPKLLIYSASYRYTGVP DRFSGSGSGTDFTLTISSLQAEDVAVYYCQQHYNTPHTFGGGTKVEIK The amino acid sequence of TIGIT-005.VL1e is shown in SEQ ID NO: 92: DIVMTQSPDSLAVSLGERATINCKASQHVSNYVAWYQHKPGQSPKLLIYSASYRYTGVP DRFSGSGSGTDFTLTISSLQAEDVAVYYCQQHYNTPHTFGGGTKVEIK Based on the above-mentioned back mutation design in light chain and heavy chain variable regions in humanized antibody of TIGIT-005, the inventors selected different sequences of light chain and heavy chain for cross combination and finally obtained 6 humanized antibodies of TIGIT-005. The amino acid sequences of the variable region in various antibodies are as follows: Table 13. Amino acid sequences corresponding to variable region in humanized antibodies of TIGIT-005 Fv 005.VL1 005.VL1a 005.VL1b 005.VL1c 005.VL1d 005.VL1e 005.VH2 TIGIT-005- H2L1* TIGIT-005- H2L1a TIGIT-005- H2L1b TIGIT-005- H2L1c TIGIT-005- H2L1d TIGIT-005- H2L1e *TIGIT-005-H2L1 indicates that the heavy chain is selected from TIGIT-005.VH2, and the 5 light chain is selected from TIGIT-005.VL1. The same below. According to Kabat numbering scheme, the analysis results of VH and VL sequences in above 6 humanized antibodies are shown in Table 14. Table 14. Kabat analysis results of VH and VL sequences in humanized antibody of TIGIT-005 Light / heavy chain variable regions CDR1 CDR2 CDR3 VH2 NYLIE SEQ ID NO:33 VINPGSGGTNYKEKFKG SEQ ID NO:34 GEYFFFDY SEQ ID NO:35 VL1 KASQHVSNGVA SEQ ID NO:30 SASYRYT SEQ ID NO:31 QQHYNTPHT SEQ ID NO:32 VL1a KASQHVSQGVA SEQ ID NO:93 SASYRYT SEQ ID NO:31 QQHYNTPHT SEQ ID NO:32 VL1b KASQHVSTGVA SEQ ID NO:94 SASYRYT SEQ ID NO:31 QQHYNTPHT SEQ ID NO:32 VL1c KASQHVSDGVA SEQ ID NO:95 SASYRYT SEQ ID NO:31 QQHYNTPHT SEQ ID NO:32 VL1d KASQHVSNAVA SEQ ID NO:96 SASYRYT SEQ ID NO:31 QQHYNTPHT SEQ ID NO:32 VL1e KASQHVSNYVA SEQ ID NO:97 SASYRYT SEQ ID NO:31 QQHYNTPHT SEQ ID NO:32 4. Humanization of TIGIT-070 10       The humanized light chain templates of the mouse antibody TIGIT-070 were IGKV1- 39*01 / IGKV4-1*01 and IGKJ3*01, and the humanized heavy chain templates are IGHV1-3*01 and IGHJ6*01. The CDRs of the mouse antibody TIGIT-070 were grafted into the human template thereof to obtain the corresponding humanized version. The key amino acids in the FR region sequence in the humanized antibody of TIGIT-070 were back mutated to the amino acids corresponding to the mouse antibody as needed to ensure the original affinity. The specific back mutation design is shown in Table 15. Table 15. The back mutation design of humanized antibody of TIGIT-070 VL VH VL1 Grafted(IGKV1-39*01) + A43S VH1 Grafted(IGHV1-3*01) + R72V,T74K VL2 Grafted(IGKV1-39*01) + A43S,I48V VH2 Grafted(IGHV1-3*01) + R72V,T74K,S75L,A76S VL3 Grafted(IGKV4-1*01) + P43S,I48V VH3 Grafted(IGHV1-3*01) + R44G,R72V,T74K,S75L,A76S Note: Grafted (IGKV1-39*01) represents the insertion of CDRs of the mouse antibody into the FR region sequence in human germline IGKV1-39*01; A43S represents the A at position 43 of Grafted (IGKV1-39*01) is back mutated to S, and so on. The back-mutated amino acids are numbered in natural order. The specific variable region sequences in humanized antibody of TIGIT-070 are as follows: The amino acid sequence of TIGIT-070.VL1 is shown in SEQ ID NO: 98: DIQMTQSPSSLSASVGDRVTITCRVSENIYSYLAWYQQKPGKSPKLLIYNAKTLAEGVPS RFSGSGSGTDFTLTISSLQPEDFATYYCQHHYGNPLTFGPGTKVDIK The amino acid sequence of TIGIT-070.VL2 is shown in SEQ ID NO: 99: DIQMTQSPSSLSASVGDRVTITCRVSENIYSYLAWYQQKPGKSPKLLVYNAKTLAEGVPS RFSGSGSGTDFTLTISSLQPEDFATYYCQHHYGNPLTFGPGTKVDIK The amino acid sequence of TIGIT-070.VL3 is shown in SEQ ID NO: 100: DIVMTQSPDSLAVSLGERATINCRVSENIYSYLAWYQQKPGQSPKLLVYNAKTLAEGVPD RFSGSGSGTDFTLTISSLQAEDVAVYYCQHHYGNPLTFGPGTKVDIK The amino acid sequence of TIGIT-070.VH1 is shown in SEQ ID NO: 101: EVQLVQSGAEVKKPGASVKVSCKASGYTFTNYYMHWVRQAPGQRLEWMGRIDPDSGG SKYNEKFKSRVTITVDKSASTAYMELSSLRSEDTAVYYCAREGHYGFYSDYWGQGTTVT VSS The amino acid sequence of TIGIT-070.VH2 is shown in SEQ ID NO: 102: EVQLVQSGAEVKKPGASVKVSCKASGYTFTNYYMHWVRQAPGQRLEWMGRIDPDSGG SKYNEKFKSRVTITVDKLSSTAYMELSSLRSEDTAVYYCAREGHYGFYSDYWGQGTTVT VSS The amino acid sequence of TIGIT-070.VH3 is shown in SEQ ID NO: 103: EVQLVQSGAEVKKPGASVKVSCKASGYTFTNYYMHWVRQAPGQGLEWMGRIDPDSG GSKYNEKFKSRVTITVDKLSSTAYMELSSLRSEDTAVYYCAREGHYGFYSDYWGQGTTV TVSS The amino acid sequence of the humanized light chain template IGKV1-39*01 is shown in SEQ ID NO: 104: DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSR FSGSGSGTDFTLTISSLQPEDFATYYCQQSYST The amino acid sequence of the humanized light chain template IGKV4-1*01 is shown in SEQ ID NO: 74: DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTR ESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYSTP The amino acid sequence of the humanized light chain template IGKJ3*01 is shown in SEQ ID NO: 105: FGPGTKVDIK The amino acid sequence of the humanized heavy chain template IGHV1-3*01 is shown in SEQ ID NO: 106: QVQLVQSGAEVKKPGASVKVSCKASGYTFTSYAMHWVRQAPGQRLEWMGWINAGNG NTKYSQKFQGRVTITRDTSASTAYMELSSLRSEDTAVYYCAR The amino acid sequence of the humanized heavy chain template IGHJ6*01 is shown in SEQ ID NO: 77: WGQGTTVTVSS Based on the above-mentioned back mutation design in light chain and heavy chain variable regions in humanized antibody of TIGIT-070, the inventors selected different sequences of light chain and heavy chain for cross combination and finally obtained various humanized antibodies of TIGIT-070. The amino acid sequences of the variable region in various antibodies are as follows: Table 16. Amino acid sequences corresponding to variable region in humanized antibodies of TIGIT-070 Fv 070.VH1 070.VH2 070.VH3 070.VL1 TIGIT-070- H1L1* TIGIT-070- H2L1 TIGIT-070- H3L1 070.VL2 TIGIT-070- H1L2 TIGIT-070- H2L2 TIGIT-070- H3L2 070.VL3 TIGIT-070- H1L3 TIGIT-070- H2L3 TIGIT-070- H3L3 *TIGIT-070-H1L1 indicates that the heavy chain is selected from TIGIT-070.VH1, and the light chain is selected from TIGIT-070.VL1. The same below. According to Kabat numbering scheme, the analysis results of VH and VL sequences in above 9 humanized antibodies are shown in Table 17. 5        Table 17. Kabat analysis results of VH and VL sequences in humanized antibody of TIGIT- 070 Light / heavy chain variable regions CDR1 CDR2 CDR3 VH1 / 2 / 3 NYYMH SEQ ID NO:39 RIDPDSGGSKYNEKFKS SEQ ID NO:40 EGHYGFYSDY SEQ ID NO:41 VL1 / 2 / 3 RVSENIYSYLA SEQ ID NO:36 NAKTLAE SEQ ID NO:37 QHHYGNPLT SEQ ID NO:38 5. Identification of humanized variants of anti-human TIGIT monoclonal antibodies The affinity of the above humanized antibodies with human TIGIT ECD-His is detected by BIAcore, and with reference to the methods of Example 4(a), (b), and (d), the binding activity of 10 the above humanized antibodies to human TIGIT ECD-mFc, CHO-K1 human TIGIT, CHO-K1 cynomolgus monkey TIGIT and cynomolgus monkey TIGIT ECD-mFc, and the effect of the humanized antibodies blocking the binding between human TIGIT ECD-mFc and CHO-K1 CD155 are detected, the results are shown in Table 18. It can be seen that the 4 humanized antibodies (TIGIT-002-H4L3, TIGIT-005-H2L1d, TIGIT- 15   006-H5L3, and TIGIT-070-H1L1) maintain the same affinity, binding activity and blocking effect as the chimeric antibodies, which are not significantly affected by humanization, and their overall performance is better than RG6058. Table 18. Identification of humanized variants of anti-human TIGIT monoclonal antibodies derived from SJL mice TIGIT humanized antibodies Human TIGIT ECD-His Human TIGIT ECD-hFc / CHO- K1 CD155 blocking experimen t CHO-K1 human TIGIT(hig h expression strain) Human TIGIT ECD-mFc binding experimen t CHO-K1 cynomolg us monkey TIGIT binding experimen t Cynomolg us monkey TIGIT ECD-mFc binding experimen t ka (1 / Ms) kd (1 / s) KD (M) Fold change vs chimeric antibodies (KD) IC50(nM) EC50(nM) EC50(nM) EC50(nM) EC50(nM) TIGIT-002-H1L1 no data no data no data no data 0.5258 0.0201 0.0155 0.1866 0.2087 TIGIT-002-H1L2 no data no data no data no data 0.4212 0.0167 0.0158 0.2206 0.0811 TIGIT-002-H1L3 3.59E+06 9.73E-04 2.71E-10 6.44 0.3953 0.0253 0.0157 0.1429 0.0511 TIGIT-002-H2L1 no data no data no data no data 0.6459 0.0468 0.0188 0.1809 0.0877 TIGIT-002-H2L2 no data no data no data no data 0.5600 0.0215 0.0222 0.1198 0.0554 TIGIT-002-H2L3 4.03E+06 1.07E-03 2.66E-10 6.30 0.5401 0.0278 0.0159 0.1498 0.0337 TIGIT-002-H3L1 no data no data no data no data 0.4593 0.0204 0.0120 0.2320 0.0253 TIGIT-002-H3L2 no data no data no data no data 0.3938 0.0329 0.0170 0.1134 0.0218 TIGIT-002-H3L3 4.52E+06 8.52E-04 1.88E-10 4.47 0.3385 0.0361 0.0173 0.0919 0.0220 TIGIT-002-H4L1 no data no data no data no data 0.4449 0.0253 0.0164 0.0543 0.0206 TIGIT-002-H4L2 no data no data no data no data 0.4397 0.0294 0.0179 0.0474 0.0254 TIGIT-002-H4L3 5.65E+06 3.00E-04 5.31E-11 1.26 0.3540 0.0249 0.0138 0.0473 0.0189 TIGIT-CHI-002 7.13E+06 3.00E-04 4.21E-11 1.00 0.5635 0.0218 0.0161 0.0372 0.0200 TIGIT-005-H1L1 no data no data no data no data not binding 0.0021 45632 0.0011 18459 TIGIT-005-H2L1 4.26E+06 1.61E-04 3.78E-11 1.37 0.3211 0.0281 0.0141 0.0199 0.0145 TIGIT-005-H2L2 4.33E+06 1.74E-03 4.02E-10 14.57 0.3530 0.0232 0.0110 0.0218 0.0107 TIGIT-005-H3L1 4.62E+06 1.55E-04 3.35E-11 1.21 0.3931 0.0223 0.0144 0.0261 0.0151 TIGIT-005-H4L1 2.07E+06 1.52E-04 7.33E-11 2.66 0.3845 0.0257 0.0124 0.0251 0.0136 TIGIT-CHI-005 4.76E+06 1.13E-04 2.76E-11 1.00 0.3596 0.0260 0.0190 0.0265 0.0165 TIGIT-005-H2L1 3.54E+06 2.63E-04 7.42E-11 1.00 0.3890 0.0272 0.0247 0.0470 0.0393 TIGIT-005-H2L1a 2.83E+06 1.05E-02 3.71E-09 50.01 0.7165 0.0129 0.0289 0.0561 0.0357 TIGIT-005-H2L1b 3.38E+06 2.48E-03 7.34E-10 9.89 0.3581 0.0279 0.0238 0.0368 0.0320 TIGIT-005-H2L1c 2.49E+06 4.59E-04 1.84E-10 2.48 0.3468 0.0246 0.0268 0.0673 0.0296 TIGIT-005-H2L1d 4.03E+06 3.61E-04 8.96E-11 1.21 0.3758 0.0439 0.0291 0.0566 0.0403 TIGIT-006-H1L1 1.22E+05 7.57E-04 6.21E-09 80.27 1.5340 0.0702 0.0186 not binding not binding TIGIT-006-H1L2 1.17E+05 6.66E-04 5.71E-09 73.78 1.5820 0.1049 0.0183 not binding not binding TIGIT-006-H1L3 1.44E+05 3.84E-04 2.66E-09 34.43 1.2320 0.0596 0.0185 not binding not binding TIGIT-006-H2L1 4.03E+05 3.82E-04 9.50E-10 12.28 0.5284 0.0383 0.0106 0.3421 2.8790 TIGIT-006-H2L2 3.63E+05 3.24E-04 8.93E-10 11.54 0.9045 0.0444 0.0117 9.02E+14 3.3840 TIGIT-006-H2L3 4.20E+05 2.16E-04 5.15E-10 6.65 0.6332 0.0353 0.0103 0.2553 24.4200 TIGIT-006-H3L1 3.42E+05 2.41E-04 7.05E-10 9.11 0.6367 0.0468 0.0096 weak binding 2.8070 TIGIT-006-H3L2 2.79E+05 2.08E-04 7.48E-10 9.66 0.7088 0.0494 0.0099 1.3810 7.4440 TIGIT-006-H3L3 3.65E+05 1.76E-04 4.82E-10 6.23 0.6815 0.0317 0.0095 6.2950 0.7790 TIGIT-006-H4L1 3.93E+05 2.58E-04 6.57E-10 8.49 0.7039 0.0349 0.0086 7.63E+09 1.5260 TIGIT-006-H4L2 3.37E+05 2.21E-04 6.58E-10 8.50 0.6708 0.0251 0.0113 weak binding 0.9967 TIGIT-006-H4L3 4.41E+05 1.88E-04 4.28E-10 5.52 0.6142 0.0316 0.0101 2.1050 0.7326 TIGIT-006-H5L1 2.29E+06 1.63E-04 7.13E-11 0.92 0.5370 0.0399 0.0108 0.1615 0.0182 TIGIT-006-H5L2 2.30E+06 1.84E-04 8.00E-11 1.03 0.5756 0.0239 0.0111 0.0849 0.0172 TIGIT-006-H5L3 2.51E+06 1.60E-04 6.39E-11 0.83 0.5866 0.0220 0.0092 0.0578 0.0132 TIGIT-CHI-006 2.09E+06 1.62E-04 7.74E-11 1.00 0.5789 0.0299 0.0152 0.1498 0.0258 TIGIT-070-H1L1 1.98E+06 1.16E-04 5.83E-11 0.59 0.5025 0.0260 0.0149 0.0596 0.0123 TIGIT-070-H1L2 1.67E+06 1.48E-04 8.83E-11 0.90 0.4206 0.0272 0.0118 0.0879 0.0116 TIGIT-070-H1L3 1.78E+06 1.29E-04 7.24E-11 0.74 0.4451 0.0240 0.0148 0.1121 0.0137 TIGIT-070-H2L1 1.66E+06 1.34E-04 8.08E-11 0.82 0.3927 0.0236 0.0129 0.0799 0.0124 TIGIT-070-H2L2 1.67E+06 1.42E-04 8.51E-11 0.87 0.4478 0.0312 0.0125 0.0662 0.0112 TIGIT-070-H2L3 1.77E+06 1.34E-04 7.56E-11 0.77 0.3504 0.0209 0.0119 0.0774 0.0119 TIGIT-070-H3L1 1.26E+06 1.52E-04 1.21E-10 1.23 0.4630 0.0132 0.0131 0.0741 0.0139 TIGIT-070-H3L2 1.30E+06 1.68E-04 1.29E-10 1.32 0.4233 0.0276 0.0143 0.0665 0.0157 TIGIT-070-H3L3 1.31E+06 1.51E-04 1.15E-10 1.17 0.4166 0.0300 0.0151 0.0556 0.0142 TIGIT-CHI-070 1.48E+06 1.45E-04 9.80E-11 1.00 0.6374 0.0330 0.0159 0.0741 0.0184 No data: not tested EXAMPLE 6 - Identification of anti-TIGIT humanized antibodies (a) Detection of the binding activity of anti-TIGIT humanized antibodies to human TIGIT ECD-mFc and cynomolgus monkey TIGIT ECD-mFc. 5        The methods refer to Example 4 (a). The experimental results show that the anti-TIGIT humanized antibodies can effectively bind to human TIGIT ECD-mFc (Fig. 13) and cynomolgus monkey TIGIT ECD-mFc (Fig. 14), which is consistent with the control antibody, Roche RG6058. (b) The binding activity of anti-TIGIT humanized antibodies to CHO-K1 human TIGIT (cells with high, medium and low expression levels) and CHO-K1 cynomolgus monkey TIGIT detected by FACS. The methods refer to Example 4 (b). The experimental results show that the anti-TIGIT humanized antibodies can effectively bind to CHO-K1 human TIGIT high expression cell strains (Fig. 15), CHO-K1 human TIGIT medium expression cell strains (Fig. 16), CHO-K1 human TIGIT low expression cell strains (Fig. 17) and CHO-K1 cynomolgus monkey TIGIT cell strains (Fig. 18), which is consistent with the control antibody, Roche RG6058. (c) The effect of anti-TIGIT humanized antibodies blocking the interaction between Bio-CD155-His and CHO-K1 human TIGIT detected by FACS. The methods refer to Example 4 (c). As shown in Fig. 19, both the anti-TIGIT humanized antibodies and the control antibody (RG6058) can effectively block the binding between the Bio-CD155-His protein and CHO-K1- human TIGIT cells. (d) The effect of anti-TIGIT humanized antibodies blocking the interaction between TIGIT ECD-mFc and CHO-K1 CD155 detected by FACS. The methods refer to Example 4 (d). As shown in Fig. 20, both the anti-TIGIT humanized antibodies and the control antibody (RG6058) can effectively block the binding between the human TIGIT protein and CHO-K1-CD155 cells. (e) The effect of anti-TIGIT humanized antibodies blocking the interaction between TIGIT ECD-mFc and CHO-K1 CD112 detected by FACS. The collected CHO-K1 CD112 cells were washed once with PBS (Hyclone, CAT#SH30256) and resuspended to 2*105 / 40 pL with 1% BSA-PBS. The anti-TIGIT humanized antibodies were diluted to 600 nM with 1% BSA-PBS (diluted by 2.5-fold serial dilution for 12 concentration points). The TIGIT ECD-mFc was diluted with 1% BSA-PBS to 6 pg / mL. The 40 pL cells were then mixed with 40 pL diluted antibodies and 40 pL diluted TIGIT ECD-mFc, and then incubated at 4°C for 60 minutes. The cells were washed twice with PBS, resuspended with Alexa Fluor® 647 fluorescein labeled secondary antibody (dilution ratio 1:800, Jackson, CAT# 109-605-003), 100 pL / well, and incubated at 4°C for 40 minutes. The cell samples were washed twice with PBS, resuspended to 100 pL / well with 1% BSA-PBS and then analyzed by flow cytometry (BD, Canto n). As shown in Fig. 21, both the anti-TIGIT humanized antibodies and the control antibody (RG6058) can effectively block the binding between the human TIGIT protein and CHO-K1-CD112 cells. (f) The binding activity of anti-TIGIT humanized antibodies to human PBMCs detected by FACS. The fresh human PBMCs (AllCells, PB004-C) were adjusted to a density of 5*105 / mL, added with SEA (Toxin Technology, Inc., CAT: AT101) to 100 ng / mL, and cultured at 37°C with 5% CO2 for 3 days. The cells were collected 3 days later, washed once with PBS (Hyclone, CAT#SH30256), added with Fc Block (BD, 564220) and then incubated at 4°C for 10 minutes. After washing twice with PBS, the PBMCs were resuspended to 2*105 / 50 pL with 1%BSA-PBS. The humanized antibodies were diluted to 80 nM with 1% BSA-PBS (diluted by 3-fold serial dilution for 12 concentration points). The 50 pL cells were then mixed with 50 pL diluted antibodies, and then incubated at 4°C for 60 minutes. The cells were washed twice with PBS, resuspended with Alexa Fluor® 647 fluorescein labeled secondary antibody (dilution ratio 1:800, Jackson, CAT# 109-605088), 100 pL / well, and incubated at 4OC for 60 minutes. After washing twice with PBS, the cell samples were resuspended with 1% BSA-PBS, 100 pL / well and then analyzed by flow cytometry (BD, Canto II). As shown in Fig. 22, the anti-TIGIT humanized antibodies can effectively bind to human PBMCs stimulated by SEA, which is consistent with the control antibody, Roche RG6058. Table 19 summarizes the characteristics of the 4 humanized antibodies. The 4 humanized antibodies (TIGIT-002-H4L3, TIGIT-005-H2L1d, TIGIT-006-H5L3, and TIGIT-070-H1L1) have the same tendency to bind to PBMC and stable transgenic cells with high, medium and low expression levels of human TIGIT, wherein, TIGIT-002-H4L3 and TIGIT-005-H2L1d have a stronger binding activity than TIGIT-006-H5L3 and TIGIT-070-H1L1, but all humanized antibodies have a stronger binding activity than RG6058. Compared with RG6058 (0.1560 nM), the monovalent affinities of 4 humanized antibodies to the human TIGIT ECD-His are 0.0994 nM, 0.0852 nM, 0.1145 nM, 0.2505 nM, respectively, and they all have strong cross-reactivity with cynomolgus monkey TIGIT. Regarding the characterization of blocking effect in vitro, the 4 humanized antibodies all show significant ability to block the interaction between TIGIT-CD155 and TIGIT-CD112. Table 19. Biological activities of the four humanized antibodies, TIGIT-002-H4L3, TIGIT-005- H2L1d, TIGIT-006-H5L3 and TIGIT-070-H1L1 TIGIT humanized antibodies Human TIGIT ECD-His Human TIGIT ECD-mFc / CH O-K1 CD155 blocking experime nt Bio-CD155 -His / CHO- K1 human TIGIT blockin g experi ment Human TIGIT ECD-mFc / CHO- K1 CD112 blockin g experim ent Huma n PBM C bindi ng exper iment CHO- K1 human TIGIT (high expres sion strain) CHO- K1 human TIGIT (mediu m expres sion strain) CHO- K1 huma n TIGI T(low expre ssion strain ) Huma n TIGI T ECD-mFc bindi ng exper iment CHO- K1 cyno molg us monk ey TIGI T bindi ng exper iment Cyno molg us monk ey TIGI T ECD-mFc bindi ng exper iment ka (1 / Ms) kd (1 / s) KD (M) IC50 (nM) IC50 (nM) IC50 (nM) EC50 (nM) EC50 (nM) EC50 (nM) EC50 (nM) EC50 (nM) EC50 (nM) EC50 (nM) TIGIT-002-H4L3 2.95E+06 2.93E-04 9.94E-11 0.4919 0.1373 3.4930 0.0436 0.2970 0.2282 0.1696 0.0370 0.7234 0.0427 TIGIT-005-H2L1d 3.02E+06 2.57E-04 8.52E-11 0.4666 0.1501 3.4880 0.0602 0.3210 0.2848 0.1856 0.0311 0.6315 0.0395 TIGIT-006-H5L3 1.60E+06 1.83E-04 1.15E-10 0.4583 0.1714 4.6480 0.0916 0.4430 0.3963 0.2677 0.0319 0.9222 0.0406 TIGIT-070-H1L1 8.75E+05 2.19E-04 2.51E-10 0.4558 0.2123 4.3030 0.1211 0.4945 0.4500 0.3048 0.0284 1.0490 0.0347 RG6058-hIgG1 8.66E+05 1.35E-04 1.56E-10 0.5840 0.3340 6.0000 0.2849 0.7945 0.6523 0.4807 0.0393 2.2850 0.0486 (g) The promoting effect of anti-TIGIT humanized antibodies on cytotoxicity for target cells by NK cells detected by NK cell cytotoxicity assay. 5        The methods refer to Example 4 (d). Figure 23 shows that the negative control, anti-HA HcAb-hIgG1, has no significant effect on the NK cell cytotoxicity. The 4 tested humanized antibodies (TIGIT-002-H4L3, TIGIT-005-H2L1d, TIGIT-006-H5L3 and TIGIT-070-H1L1) can promote the NK cell cytotoxicity for the target cells (WIDR) to varying degrees, wherein, TIGIT-002-H4L3 and TIGIT-005-H2L1d show the same effect on NK cell cytotoxicity as pre-humanized 10 chimeric antibody TIGIT-CHI-002 and TIGIT-CHI-005. (h) The effect of anti-TIGIT humanized antibody on improving the functions of antigenspecific CD8 T cells detected by CMV antigen-recall assay The PBMCs were resuscitated, resuspended to 2^106 / mL by using complete medium (RPMI1640-Glutamax + 5% AB serum + 1% P / S+(1x) 2-0 mercaptoethanol) containing 1 mg / mL CMV pp65(495-503) peptide (Anaspec, Cat No. AS-28328), 2 ng / mL human IL-2 (R&D, Cat No. IL-202) and 10 ng / mL human IL-7 (Peprotech, Cat No. 200-07), inoculated in a 6-well plate (5 mL / well) and incubated at 37°C with 5% CO2 for 6 days. On day 6, all the PBMCs were collected, and pp65 and IL-7 in the medium were removed. The cells were divided into two portions, resuspended in complete medium containing 100 IU / mL human IL-2, and cultured for another 2 days. On day 8, all the PBMCs were collected and resuspended in complete medium containing 100 lU / mL human IL-2, and the cell density was adjusted to 2*106 / mL for continuing cultivation. On day 11, all the PBMCs were collected. The percentage of CD8 T cells and CMV pp65 (495503) specific CD8 T cells in PBMCs (Figure 24, panel A), and the expression of PVRIG, TIGIT, PD-1 and CD226 on CMV pp65 (495-503) specific CD8 T cells (Figure 24, panel B) were detected by flow cytometry. The detection antibodies were as follows: Livedead Near IR (Invitrogen, Cat No. L34976), CD8-PerCp Cy5.5 (BD, Cat No. 565310), CD3-PE-Cy7 (Biolegend, Cat No. 300316), T-select HLA-A*0201 CMV pp65 Tetramer-PE (MBL, Cat No. TS-0010-1C), PVRIG-AF488 (R&D, Cat No. FAB93651G-100UG), TIGIT-APC (Biolegend, Cat No. 372706) and PD-1-BV421 (BD, Cat No. 562516). CD8 T cells were isolated from the induced PBMCs as effector cells by CD8 T cell isolation kit (Stemcell, Cat No.17953) and resuspended with AIM-V to the cell density of 0.4*106 / mL. The purity of isolated CD8 T cells and the expression of CD226 on them were detected. The target cells, Colo205, were digested by TrypLETM Express Enzyme (Gibco, Cat No. 12605010), resuspended in AIM-V (Gibco, Cat No. 31035-025) containing 20 ng / mL pp65 to the cell density of IXI06 ml.and treated at 37°C with 5% CO2 for 3 hours. The target cells were then centrifuged at 250 g for 5 minutes and resuspended in AIM-V to the cell density of 0.5^106 mL after discarding the supernatant. Anti-TIGIT antibodies or positive control were diluted with AIM-V to 280 nM. The low-attachment, 96-well, U-bottom plate (Corning, Cat No.7007) was added with 50 pL of antibody, 50 pL of CD8 T and 100 pL of pp65-treated colo205 in order, mixed well with a multichannel pipette, and incubated at 37°C with 5% CO2 for 18 hours. In the experimental system, the final concentration of antibodies was 70 nM, CD8 T cells were 20,000 / well, and colo205 were 50,000 / well. After incubation, the supernatant was collected by centrifugation at 400 g, and the level of human IFN-y in the supernatant was detected with an ELISA kit (Dakewe, Cat No.1110003). In this system, the positive controls were RG6058-hIgG1 and TIGIT pre-humanized antibody (TIGIT-CHI-002), and the negative control was no treatment. As shown in Fig. 24 (panel C), the secretion level of IFN-Y in the cell supernatant treated with 4 TIGIT humanized antibodies (TIGIT-002-H4L3, TIGIT-005-H2L1d, TIGIT-006-H5L3 and TIGIT-070-H1L1) has no statistically significant difference (One-way ANOV Analysis) compared with RG6058-hIgG1; the secretion level of IFN-Y in the cell supernatant has no statistically significant difference (One-way ANOV Analysis) compared with TIGIT pre-humanized antibody (TIGIT-CHI-002), but was significantly higher than the no treatment group. The percentage on each histogram represents the percentage of IFN-Y secretion increased in the experimental group compared with no treatment group. The purity of isolated CD8 T cells and the expression of CD226 on them were detected by flow cytometry using following detection antibodies: livedead-BV421 (Invitrogen, Cat No. L34964), CD8-FITC (BD, Cat No. 555366) and CD226-PE-Cy7 (Biolegend, Cat No. 338316). The expression of PVRL2 and PVR on colo205 was detected by flow cytometry using following detection antibodies: livedead-BV421 (Invitrogen, Cat No. L34964), PVRL2-APC (Biolegend, Cat No. 337412), PVR-PerCp Cy5.5 (Biolegend, Cat No. 337612), PD-L1- PE-Cy7 (BD, Cat No. 558017) and HLA-A2-PE (Biolegend, Cat No. 343306). EXAMPLE 7 - Preparation of PVRIG Alpaca VHH Antibody Two healthy adult alpacas (Alpaca) (purchased by Chengdu Apack Company) were injected with complete freund's adjuvant (CFA, from SIGMA, Cat No. F5881) mixed with human PVRIG recombinant protein (Acro Biosystems, Cat No. PVG H5257) for the first immunization, and with incomplete freund's adjuvant (IFA, from SIGMA, Cat No. F5506) mixed with the same human PVRIG recombinant protein for the last three immunizations, all of which were injected subcutaneously for four times. Before immunization, 10 mL of blood was taken as negative serum control, after the second immunization, 10 mL of blood was taken to detect serum antibody titer, and after the third and fourth immunization, 50 mL of peripheral blood was taken to separate lymphocytes. According to the amount of lymphocytes, 5 mL RNA iso Plus (Takara, Cat No. 9109) was added and stored in 1.5 mL EP tube at -80 °C. Total RNA was extracted from cryopreserved lymphocytes and reverse transcribed into cDNA, and then two rounds of VHH PCR amplification were carried out using cDNA as template. The VHH products amplified by the second round of PCR were digested to construct a phage library. The established bacterial bank was collected, and the insertion rate and diversity of the bank were sequenced and analyzed. Two rounds of affinity panning were performed on the phage, and the phage clones that specifically bound to the target protein human PVRIG-his (AcroBiosystems, Cat No. PVG-H52H4) were identified. The optimized clones with better binding to human PVRIG-His protein were 5 sequenced and cloned into the expression vector by homologous recombination method. The CH2 and CH3 constant regions were both derived from human IgG1, and the complete expression sequence was signal peptide-VHH-hinge-CH2-CH3. After a series of physicochemical and functional screening, a total of 13 positive candidate antibody molecules were obtained, and the CDRs of their sequences were analyzed by IMGT and KABAT respectively. The corresponding 10 sequence information is shown in Tables 20 to 22 below. Table 20 shows the VHH sequence of candidate antibody molecules, Table 21 shows the IMGT analysis results of candidate antibody molecules, and Table 22 shows the KABAT analysis results of candidate antibody molecules. Table 20 VHH Sequence of Candidate Antibody Molecules Name SEQ ID sequence PVRIG-A11 SEQ ID NO. 107 QLQLVESGGGLVQPGGSLRLSCAASGFTDDYYAIGWFRQAPGKE REGVSCISGSGGSTNYEDSVKGRFAISRDNAKNMVYLQMNDLKP EDTAVYYCAADPFWSAPCTGHNDRYFDVWGQGTLVTVSS PVRIG-A15 SEQ ID NO. 108 QVQLVESGGGLVQAGGSLRLSCVASGRTFSANTMAWFRQAPGKE REFVAYTRWTDDNTDYASYADFVKGRFTAFRDNTRNTMYLQMN RLRPEDTAVYYCAAATTRGTYYSAGDYNSWGQGTQVTVSS PVRIG-A30 SEQ ID NO. 109 QVQLVESGGGLVQPGGSLRLSCAASGRTDSMNVMGWFRRAPGK EREFVARIKWNGDTTYTAYADFLKGRFTLYGAVARNRVYLQMNS LQPEDTAVYYCAAGEVSGSSYSPDYGMHYWGKGTLVTVSS PVRIG-A50 SEQ ID NO. 110 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKG LEWVSTINSNGGRTSYVDSVKGRFTISRDNTKNTLYLQMNSLKPE DTAVYYCVEGDPHNFGLENLSLRDFGSWGQGTQVTVSS PVRIG-A60 SEQ ID NO. 111 QVQLVESGGGLVQSGGSLRLSCAASGRTRSMYVMGWFRQAPGK EREFVGRIKWSGDTTYTSYADFVKGRFTLYGNQARNTVYLQMNS LKPEDAAVYYCNLRRLDGVNYWGKGTLVTVSS PVRIG-A75 SEQ ID NO. 112 QLQLVESGGGLVQSGGSLRLSCAASGRTRSMYVMGWFRQAPGK EREFVGRIKWSGDTTYTSYADFVKGRFTLYGNQARNTVYLQMNS LKPEDTAVYSCAASKFPVSGVPEHYDYWGQGTQVTVSS PVRIG-A35 SEQ ID NO. 113 EVQVVESGGGLVQSGGSLRLSCAASGRTRSMYVMGWFRQAPGK EREFVGRIKWSGDTTYTSYADFVKGRFTLYGNQARNTVYLQMNS LKPEDTAVYYCNGRSRFNVINAWGTGTLVTVAS PVRIG-A43 SEQ ID NO. 114 QVQLVESGGGLVQSGGSLRLSCAASGRTRSMYVMGWFRQAPGK EREFVGRIKWSGDTTYTSYADFVKGRFTLYGNQARNTVYLQMNS LKVEDTAVYYCAGRVVPLPSRQRDRYDYWGQGTQVTVSS PVRIG-A104 SEQ ID NO. 115 EVQVVESGGGLVQAGGSLRLSCAASGRTRSMYVMGWFRQAPGK EREFVGRIKWSGDTTYTSYADFVKGRFTLYGNQARNTVYLQMNS LKPEDTAVYYCYRRSWTNTKLYWGKGIQVTVSS PVRIG-A105 SEQ ID NO. 116 EVQLVESGGGLVQAGGSLRLSCAASGRTFDRHTMTWFRQAPGKE REFVATASRIPGDTYYSHSVKGRFTISRDNAKNTVYLQLNNLKPED TAVYYCAATSAYCSEVDCYEKGSWYDNWGQGIQVTVSS PVRIG-A113 SEQ ID NO. 117 QLQLVESGGGLVQSGGSLRLSCAASGRTRSMYVMGWFRQAPGK EREFVGRIKWSGDTTYTSYADFVKGRFTLYGNQARNTVYLQMNS LKPEDTAVYYCAALPSDYDYRAASYGVDYWGKGTLVTVSS PVRIG-A117 SEQ ID NO. 118 EVQVVESGGGLVQAGGSLRLSCTASEHIYDLYIMGWYRQAPGKD RELVATITYTGSIYIADSVKDRFTISRGNAKNTVSLQMSGLKPEDT AVYYCNADPSGLGRKLYWGQGTQVTVSS PVRIG-A118 SEQ ID NO. 119 QVQLVESGGGLVQAGGSLRLSCAASETYFDLYVMGWYRQAPGK DRELVATITYTGSIKIVDSVKDRFTISRGDAKNTVSLQMSSLKPEDT AVYFCNADPSGLGRKVYWGQGTQVTVSS Table 21 IMGT Analysis Results of Candidate Antibody Molecules Name CDR1-IMGT CDR2-IMGT CDR3-IMGT PVRIG-A11 GFTDDYYA SEQ ID NO. 120 ISGSGGST SEQ ID NO. 121 AADPFWSAPCTGHNDRYFDV SEQ ID NO. 122 PVRIG-A15 GRTFSANT SEQ ID NO. 123 TRWTDDNTDYA SEQ ID NO. 124 AAATTRGTYYSAGDYNS SEQ ID NO. 125 PVRIG-A30 GRTDSMNV SEQ ID NO. 126 IKWNGDTTYT SEQ ID NO. 127 AAGEVSGSSYSPDYGMHY SEQ ID NO. 128 PVRIG-A50 GFTFSYYD SEQ ID NO. 129 INSNGGRT SEQ ID NO. 130 VEGDPHNFGLENLSLRDFGS SEQ ID NO. 131 PVRIG-A60 GRTRSMYV SEQ ID NO. 132 IKWSGDTTYT SEQ ID NO. 133 NLRRLDGVNY SEQ ID NO. 134 PVRIG-A75 GRTRSMYV SEQ ID NO. 135 IKWSGDTTYT SEQ ID NO. 136 AASKFPVSGVPEHYDY SEQ ID NO. 137 PVRIG-A35 GRTRSMYV SEQ ID NO. 138 IKWSGDTTYT SEQ ID NO. 139 NGRSRFNVINA SEQ ID NO. 140 PVRIG-A43 GRTRSMYV IKWSGDTTYT AGRVVPLPSRQRDRYDY SEQ ID NO. 141 SEQ ID NO. 142 SEQ ID NO. 143 PVRIG- GRTRSMYV IKWSGDTTYT YRRSWTNTKLY A104 SEQ ID NO. 144 SEQ ID NO. 145 SEQ ID NO. 146 PVRIG- GRTFDRHT ASRIPGDT AATSAYCSEVDCYEKGSWYDN A105 SEQ ID NO. 147 SEQ ID NO. 148 SEQ ID NO. 149 PVRIG- GRTRSMYV IKWSGDTTYT AALPSDYDYRAASYGVDY A113 SEQ ID NO. 150 SEQ ID NO. 151 SEQ ID NO. 152 PVRIG- EHIYDLYI ITYTGSI NADPSGLGRKLY A117 SEQ ID NO. 153 SEQ ID NO. 154 SEQ ID NO. 155 PVRIG- ETYFDLYV ITYTGSI NADPSGLGRKVY A118 SEQ ID NO. 156 SEQ ID NO. 157 SEQ ID NO. 158 Table 22 KABAT Analysis Results of Candidate Antibody Molecules Name CDR1-KABAT CDR2-KABAT CDR3-KABAT PVRIG-A11 YYAIG SEQ ID NO. 159 CISGSGGSTNYEDSVKG SEQ ID NO. 160 DPFWSAPCTGHNDRYFDV SEQ ID NO. 161 PVRIG-A15 ANTMA SEQ ID NO. 162 YTRWTDDNTDYASYADFVKG SEQ ID NO. 163 ATTRGTYYSAGDYNS SEQ ID NO. 164 PVRIG-A30 MNVMG SEQ ID NO. 165 RIKWNGDTTYTAYADFLKG SEQ ID NO. 166 GEVSGSSYSPDYGMHY SEQ ID NO. 167 PVRIG-A50 YYDMS SEQ ID NO. 168 TINSNGGRTSYVDSVKG SEQ ID NO. 169 GDPHNFGLENLSLRDFGS SEQ ID NO. 170 PVRIG-A60 MYVMG SEQ ID NO. 171 RIKWSGDTTYTSYADFVKG SEQ ID NO. 172 RRLDGVNY SEQ ID NO. 173 PVRIG-A75 MYVMG SEQ ID NO. 174 RIKWSGDTTYTSYADFVKG SEQ ID NO. 175 SKFPVSGVPEHYDY SEQ ID NO. 176 PVRIG-A35 MYVMG SEQ ID NO. 177 RIKWSGDTTYTSYADFVKG SEQ ID NO. 178 RSRFNVINA SEQ ID NO. 179 PVRIG-A43 MYVMG SEQ ID NO. 180 RIKWSGDTTYTSYADFVKG SEQ ID NO. 181 RVVPLPSRQRDRYDY SEQ ID NO. 182 PVRIG-A104 MYVMG SEQ ID NO. 183 RIKWSGDTTYTSYADFVKG SEQ ID NO. 184 RSWTNTKLY SEQ ID NO. 185 PVRIG-A105 RHTMT SEQ ID NO. 186 TASRIPGDTYYSHSVKG SEQ ID NO. 187 TSAYCSEVDCYEKGSWYDN SEQ ID NO. 188 PVRIG-A113 MYVMG SEQ ID NO. 189 RIKWSGDTTYTSYADFVKG SEQ ID NO. 190 LPSDYDYRAASYGVDY SEQ ID NO. 191 PVRIG-A117 LYIMG TITYTGSIYIADSVKD DPSGLGRKLY SEQ ID NO. 192 SEQ ID NO. 193 SEQ ID NO. 194 PVRIG-A118 LYVMG SEQ ID NO. 195 TITYTGSIKIVDSVKD SEQ ID NO. 196 DPSGLGRKVY SEQ ID NO. 197 EXAMPLE 8 - The specific binding of PVRIG antibodies to human and cynomolgus monkey PVRIG protein detected by ELISA The ELISA plate was pre-coated with 100 pL / well of 0.5 pg / mL human PVRIG-his (AcroBiosystems, Cat NO.PVG-H52H4) or cynomolgus monkey PVRIG (Novoprotein, Cat NO.C09B). The tested PVRIG antibodies (constructed by connecting human IgG1-Fc with VHH described in Example 7) were serially diluted (initial concentration 20 nM, 3.33-fold serial dilution, or 3nM, 3-fold serial dilution), added into the plate (100 pL / well) and incubated with shaking at room temperature for 1.5 hours. After washing the plate, the mouse anti-human IgG Fc-HRP (Jackson ImmunoResearch, Cat NO.209-035-098)) working solution (dilution ratio 1:10000, 100 pL / well) were added into wells and incubated with shaking at room temperature for 1.0 hour. A HRP substrate, TMB (Thermo, Cat NO.34029) was added into wells after washing the plate again for color development. After adding a termination solution to terminate the reaction, the absorbance value was read by a microplate reader (MD i3x). An antibody binding curve was drawn with concentration of antibodies as the abscissa and the corresponding OD value as the ordinate. The EC50 value was calculated by using a Four Parameter Logistic Fit (GraphPadPrism9). The smaller the EC50 value, the stronger the ability of the antibodies binds to human or cynomolgus monkey PVRIG. The positive control antibodies are COM701-hIgG1 (Patent No.:US20180244774A1), COM701-hIgG4 (patent No.: US20180244774A1) and SRF813-hIgG1 (Patent No.: US20200040081A1); and the negative control antibodies are anti-HA HcAb-hIgG1 (Chengdu NB Biolab, Cat NO. NBR022), anti-CD38 HcAb-hIgG1 (in-house) and anti-Fluorescein-hIgG1 (in house). The corresponding amino acid sequences of COM701 are as follows: COM701 VH SEQ ID NO 236: QVQLVQSGAEVKKPGASVKVSCKASGYTFTDYNINWVRQAPGQGLEWMGYIYPYIGGS GYAQKFQGRVTMTRDTSTSTVYMELSSLRSEDTAVYYCAREDKTARNAMDYWGQGTL VTVSS COM701 VL SEQ ID NO 237: DIQMTQSPSSLSASVGDRVTITCRVSENIYSNLAWYQQKPGKAPKLLIYEATNLAEGVPS RFSGSGSGTDFTLTISSLQPEDFATYYCQHFWGTPYTFGQGTKLEIK SRF813 VH SEQ ID NO 238: QVQLVQSGAEVKKPGSSVKVSCKASGGTFSSAAISWVRQAPGQGLEWMGNIIPIVGIAN 5 YAQKFQGRVTITADESTSTAYMELSSLRSEDTAVYYCARDTGRGYTRHFWFDPWGQGTL VTVSS SRF813 VL SEQ ID NO 239: DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSR FSGSGSGTDFTLTISSLQPEDFATYYCQQSDILYTFGGGTKVEIK 10        The binding results of the 13 anti-PVRIG antibodies to human PVRIG protein are shown in Fig. 26A and Table 23, and the binding results to cynomolgus monkey PVRIG are shown in Fig. 26B and Table 23. The data shows that all test antibodies can specifically bind to human or cynomolgus monkey PVRIG protein. Table 23. The specific binding of anti-PVRIG antibodies to human or cynomolgus monkey 15                           PVRIG protein detected by ELISA Name EC50 (pM) Bind to human PVRIG protein Bind to cynomolgus monkey PVRIG protein PVRIG-A11 121.40 28.91 PVRIG-A15 207.70 128.60 PVRIG-A30 254.70 93.00 PVRIG-A50 195.70 53.35 PVRIG-A60 55.66 34.29 PVRIG-A75 62.11 28.83 PVRIG-A35 54.31 28.83 PVRIG-A43 53.65 35.99 PVRIG-A104 65.93 35.33 PVRIG-A105 57.07 19.24 PVRIG-A113 92.00 44.31 PVRIG-A117 52.84 41.81 PVRIG-A118 51.35 34.10 EXAMPLE 9 - The binding of anti-PVRIG antibodies to FlpinCHO-PVRIG and FlpinCHO-cyno PVRIG detected by FACS. CHO-K1 stable cells transfected with human or cynomolgus monkey PVRIG high expression plasmids were named FlpinCHO-PVRIG and FlpinCHO-cyno PVRIG. Human PVRIG full-length plasmids (NCBI Ref Seq: NP076975) and cynomolgus monkey PVRIG full-length plasmid (NCBI Ref Seq: XP014989941) were synthesized by General Biol. The experiment was performed when the cell density did not exceed 80%. After discarding the cell culture medium, the cells were rinsed with PBS and digested for 8-10 minutes by adding 1mL Versene (Gibico, 15040066). The Ham's F12 (Gibico, 21127-022) complete medium containing 10% FBS was then added to terminate the digestion and acquire cell suspension. After counting with a cell counter (Beckman Coulter, Vi-CELL), an appropriate amount of cell suspension was centrifuged at 350xg to remove the supernatant, washed twice with PBS, stained with Zombie violet (Biolegend, 423114) and incubated at room temperature for 20 minutes. After incubation, the cells were added with staining buffer (2% FBS+PBS) to stop staining, centrifuged at 350xg to remove the supernatant, washed twice and resuspended with staining buffer to a density of 2x106 cells / mL. The cells were plated into a 96-well plate (50pL of cell suspension per well) for further use. The antibodies were diluted from the highest concentration of 46 nM (two-fold concentration) with staining buffer by 3.3-fold serial dilution. The diluted antibodies were added to the well containing 50 pL of cell suspension. The plate was placed on a microplate shaker at 400 rpm for 1 minute to fully mix the antibodies and cells, and then incubated at 4°C for 30 minutes. After the incubation, the cells were washed twice with staining buffer (200 pL per well) and centrifuged at 350xg for 5 minutes to discard the supernatant. The PE goat anti-human IgG Fc antibody (ebioscience, 12-4998-82) was diluted 250 times with staining buffer, added into the washed wells (100 pL per well), mixed well, and stained at 4°C for 30 minutes. After staining, the cells were washed twice with staining buffer, finally resuspended with 200 pL staining buffer and analyzed by flow cytometry (BD, Canto n). The stronger the signal, the stronger the ability of the antibodies binds to PVRIG. An antibody binding curve was drawn with concentration of antibodies as the abscissa and the corresponding Mean Fluorescence Intensity (MFI) as the ordinate. The AUC value of the curve was calculated by using a Four Parameter Logistic Fit (GraphPadPrism9). The stronger the AUC value, the stronger the ability of the antibodies binds to FlpinCHO-PVRIG and FlpinCHO-cyno PVRIG. Fig. 27A and 27B show that all the test antibodies bind to human / cynomolgus monkey PVRIG on the surface of overexpressing cells. The binding activity of the antibodies was normalized to the positive control COM701-hIgG1 and SRF813-hIgG1. The higher the percentage value, the stronger the binding ability of the antibodies. Table 24 shows that the binding activity of the tested antibodies PVRIG-A11, A35, A43, A105, A117 and A118 to human PVRIG is stronger than that of COM701-hIgG1 and SRF813-hIgG1, and the binding activity of PVRIG-A105 and A117 to cynomolgus monkey PVRIG is stronger than COM701-hIgG1 and SRF813-hIgG1. 5 Table 24. Binding of anti-PVRIG antibodies to human / cynomolgus monkey PVRIG on the surface of overexpressing cells Name human PVRIG binding cyno PVRIG binding %AUC %AUC to COM701- hIgG1 to SRF813- hIgG1 to COM701- hIgG1 to SRF813- hIgG1 PVRIG-A11 116.77 142.90 41.30 48.56 PVRIG-A15 36.32 42.69 24.80 29.15 PVRIG-A30 29.30 34.44 28.73 33.78 PVRIG-A50 89.54 114.99 59.31 69.73 PVRIG-A60 63.30 68.58 63.40 82.88 PVRIG-A75 50.64 54.46 47.58 62.40 PVRIG-A35 121.76 133.55 19.27 25.28 PVRIG-A43 165.18 181.18 19.62 25.73 PVRIG-A104 18.07 23.94 18.45 23.87 PVRIG-A105 122.51 162.31 121.63 157.39 PVRIG-A113 42.38 57.40 38.63 52.69 PVRIG-A117 133.29 180.80 127.75 174.25 PVRIG-A118 152.73 232.45 66.41 80.43 EXAMPLE 10 - The affinity of anti-PVRIG antibodies for human PVRIG proteins detected by BIAcore Biacore was used to detect the specific binding between anti-PVRIG antibody and human 10 PVRIG protein. Protein A chip was used in the BIAcore assay. The time required for the chip to capture the diluted antibody and saturate the binding antigen to reach Rmax (Maximum binding capacity, 50 RU) was measured by manual run. The human PVRIG proteins (Human PVRIG-His, Acro C227P1-9ARF1-T4) were serially diluted to 20, 10, 5, 2.5, 1.25 nM. The affinity of antibody for antigen was measured by multi-cycle kinetics. In each cycle, the antibodies were injected prior 15 to the injection of gradient concentraions of PVRIG proteins allowing the occurrence of antibodyantigen association and dissociation After each cycle, the Protein A chip was regenerated by Glycine, pH 1.5 (to remove the proteins on the chip). The affinity of antibody for antigen was calculated by BIAcore T200 analysis software. As shown in Table 25, all the anti-PVRIG antibodies specifically bind to human PVRIG proteins with high affinity. Table 25. BIACore results of specific binding of anti-PVRIG antibodies for human PVRIG 5                                             proteins Name Antigen KD (M) ka (1 / Ms) kd (1 / s) PVRIG-A11 Human PVRIG 1.06E-10 3.18E+07 3.38E-03 PVRIG-A15 3.77E-10 6.12E+04 2.31E-05 PVRIG-A30 4.04E-10 1.57E+05 6.36E-05 PVRIG-A50 2.54E-11 5.07E+05 1.29E-05 PVRIG-A60 2.83E-10 1.41E+05 3.97E-05 PVRIG-A75 5.31E-10 1.07E+05 5.67E-05 PVRIG-A35 8.46E-10 8.90E+03 7.53E-06 PVRIG-A43 3.66E-10 2.34E+04 8.58E-06 PVRIG-A104 2.76E-11 1.54E+05 4.24E-06 PVRIG-A105 8.41E-10 5.27E+05 4.43E-04 PVRIG-A113 2.03E-10 4.42E+05 8.96E-05 PVRIG-A117 3.67E-10 3.74E+06 1.37E-03 PVRIG-A118 4.80E-10 2.38E+06 1.14E-03 COM701-hIgG1 3.19E-10 2.04E+06 6.51E-04 SRF813-hIgG1 2.78E-10 1.14E+06 3.17E-04 EXAMPLE 11 - The effect of anti-PVRIG antibodies blocking the binding between PVRIG and PVRL2 detected by ELISA The ELISA plate was pre-coated with 100 pL / well of 0.5 pg / mL human PVRIG-his protein (AcroBiosystems, Cat NO.PVG-H52H4). The tested anti-PVRIG antibodies were diluted from the 10 highest concentration of 16 nM (two-fold concentration). The diluted antibodies were mixed with 18 ng / mL human PVRL2-mFc (AcroBiosystems, Cat NO. CD2-H5257) in equal volume, added into the plate (100 pL / well), and incubated with shaking at room temperature for 2.0 hours. After washing the plate, the Goat anti-mouse IgG Fc-HRP (Jackson ImmunoResearch, Cat NO.115-035-071)) working solution (dilution ratio 1:10000, 50 pL / well) were added into wells and incubated 15 with shaking at room temperature for 1.0 hour. A HRP substrate, TMB (Thermo, Cat NO.34029) was added into wells after washing the plate again for color development. After adding a termination solution to terminate the reaction, the absorbance value was read by a microplate reader (MD i3x). An inhibiting curve was drawn with concentration of antibodies as the abscissa and the corresponding OD value as the ordinate. The IC50 value was calculated by using a Four Parameter Logistic Fit (GraphPadPrism9). The smaller the IC50 value, the stronger the ability of the antibodies inhibits the binding between PVRIG and PVRL2. The positive and negative controls 5 are the same as in Example 8. Fig. 28 shows the blocking curves of the 13 test antibodies, and Table 26 shows the inhibiting activity. As shown in Fig. 28 and Table 26, all the antibodies can significantly inhibit the binding of human PVRIG to human PVRL2 protein. Table 26. The effect of test antibodies blocking the binding of PVRIG to PVRL2 Name IC50 (pM) PVRIG-A11 121.40 PVRIG-A15 207.70 PVRIG-A30 254.70 PVRIG-A50 195.70 PVRIG-A60 55.66 PVRIG-A75 62.11 PVRIG-A35 54.31 PVRIG-A43 53.65 PVRIG-A104 65.93 PVRIG-A105 57.07 PVRIG-A113 92.00 PVRIG-A117 52.84 PVRIG-A118 51.35 EXAMPLE 12 - The effect of anti-PVRIG antibodies blocking CHO-K1-CD112 cells binding 10 to human PVRIG-mFc proteins detected by FACS CHO-K1 stable cells transfected with human CD112 high expression plasmid was named CHO-K1-CD112. Human CD112 full-length plasmid (NP 001036189.1 / NCBI Ref Seq: Q92692) was synthesized by General Biol. The experiment was performed when the cell density did not exceed 80%. After discarding the cell culture medium, the cells were rinsed with PBS and digested 15 for 2 minutes by adding 1mL trypsin (Gibico, 25200-72). The Ham's F12 (Gibico, 21127-022) complete medium containing 10% FBS was then added to terminate the digestion and acquire cell suspension. After counting with a cell counter (Beckman Coulter, Vi-CELL), an appropriate amount of cell suspension was centrifuged at 350*g to remove the supernatant, washed twice with PBS, stained with Zombie violet (Biolegend, 423114) and incubated at room temperature for 20 minutes. After incubation, the cells were added with staining buffer (2% FBS+PBS) to stop staining, centrifuged at 350xg to remove the supernatant, washed twice and resuspended with staining buffer to a density of 1*106 cells / mL. The cells were plated into a 96-well plate (50pL of cell suspension per well) for further use. A working solution of human PVRIG-mFc protein (Acro, PVG-H5253) prepared with staining buffer, 1pg / mL (four-fold concentration), was added into a 96-well plate (50pL PVRIG-mFc working solution per well). The antibodies were diluted from the highest concentration of 275 nM (four-fold concentration) with staining buffer by 3-fold serial dilution. The diluted antibodies were added to the well containing 50 pL of PVRIG-mFc. The plate was placed on a microplate shaker at 400 rpm for 1 minute to fully mix the antibodies and PVRIG-mFc proteins, and then incubated at 4°C for 30 minutes. After the incubation, the cell suspension prepared above (100pL) was added into the wells, gently mixed with the pipette tip and incubated at 4°C for 30 minutes. After the incubation, the cells were washed twice with staining buffer (200 pL per well) and centrifuged at 350*g for 5 minutes to discard the supernatant. The PE goat antimouse IgG Fc antibody (Biolegend, 405337) was diluted 250 times with staining buffer, added into the washed wells (100 pL per well), mixed well, and stained at 4°C for 30 minutes. After staining, the cells were washed twice with staining buffer, finally resuspended with 200 pL staining buffer and analyzed by flow cytometry (BD, Canto II). The weaker the fluorescence signal, the stronger the ability of the antibodies blocks the binding of CHO-K1 human CD112 cells to the PVRIG-mFc protein. An antibody binding curve was drawn with concentration of antibodies as the abscissa and the corresponding Mean Fluorescence Intensity (MFI) as the ordinate. The IC50 value of the antibodies and the AUC value of the binding curve were calculated by using a Four Parameter Logistic Fit (GraphPadPrism9). The smaller the IC50 value and AUC value, the stronger the ability of the antibodies competes with CHO-K1-CD112 cells to bind to human PVRIG-mFc protein. As shown in Fig. 29, all the test antibodies can block the binding of CHO-K1 human CD112 cells to human PVRIG-mFc protein. The competitive blocking activity of the antibodies was normalized to the positive control COM701-hIgG1 and SRF813-hIgG1. The smaller the percentage value, the stronger the blocking ability of the antibodies. Table 27 shows that the blocking activity of the tested antibodies PVRIG-A11, A15, A30 and A50 is stronger than that of COM701-hIgG1 and SRF813-hIgG1. Table 27 Anti-PVRIG antibodies compete with CHO-K1-CD112 cells to bind to human PVRIG-mFc protein Name CHO-K1-CD112 cell blocking %IC50 %AUC to COM701-hIgG4 to SRF813-hIgG1 to COM701-hIgG4 to SRF813-hIgG1 PVRIG-A11 62.81 64.08 77.99 58.87 PVRIG-A15 54.07 72.38 77.22 77.27 PVRIG-A30 48.12 64.41 58.86 58.90 PVRIG-A50 99.07 91.85 103.64 81.01 PVRIG-A60 98.87 96.37 117.47 102.84 PVRIG-A75 89.61 93.74 100.89 86.58 PVRIG-A35 138.27 136.23 145.00 125.72 PVRIG-A43 121.18 119.40 127.93 110.92 PVRIG-A104 117.82 115.52 151.79 110.90 PVRIG-A105 99.67 97.72 115.09 84.08 PVRIG-A113 124.29 97.94 172.27 105.69 PVRIG-A117 133.89 117.75 129.93 72.66 PVRIG-A118 110.95 97.42 109.54 75.58 EXAMPLE 13 - Detection of The expression of PVRIG and TIGIT on NK cells, and PVR and PVRL2 on Tumor cell lines Reh and WIDR cells 5       The expression level of PVRIG and TIGIT on NK cells was detected by FACS. The method refer to Example 4 (e). As shown in Fig. 9A, PVRIG and TIGIT are expressed on the surface of NK cells of donor-010 and donor-050. The expression level of PVR and PVRL2 on Reh / WIDR cells was detected by FACS. The methods refer to Example 4 (e), wherein Reh were directly mixed to prepare the cell 10 suspensions. The expression of PVR and PVRL2 on the surface of WIDR cells was shown in Figure 9B. As shown in Fig. 30, the expression of PVR on the surface of Reh cells is negative and the expression of PVRL2 is positive. EXAMPLE 14 - The promoting effect of anti-PVRIG antibodies on NK cell function detected by NK cell degranulation assay 15        The effect of the tested antibodies on NK cell degranulation was indicated by the signal of CD107a in NK cells (Natural killer cells) detected by FACS. The methods refer to Example 4 (f), wherein Reh were directly mixed to prepare cell suspensions. Fig. 31 shows that the negative control, anti-HEL-hIgG1, has no effect on CD107a on NK cells. 12 PVRIG antibodies and the control antibodies COM701-hIgG1 and SRF813-hIgG1 can improve the expression of CD107a on NK cells to varying degrees, which indicates that the tested antibodies and control antibodies can effectively promote the activation of NK cells. EXAMPLE 15 - The cytotoxicity of NK cells for tumor cell line mediated by anti-PVRIG antibodies detected by NK cell cytotoxicity assay The lysis level of target cells (WIDR) was detected by FACS to infer the effect of the tested antibodies on the cytotoxicity of NK cells. The methods refer to Example 4 (g). Figure 32 shows that the negative control, anti-HA HcAb-hIgG1, has no significant effect on the NK cell cytotoxicity, while 13 tested antibodies can effectively promote the NK cell cytotoxicity for the target cells (WIDR), and except for PVRIG-A35 and A43, the cytotoxicity promoting effect of the remaining 11 antibodies to WIDR cells was stronger than or equal to that of the control antibody COM701-hIgG1. EXAMPLE 16 - The effect of anti-PVRIG antibodies on improving the functions of antigenspecific CD8 T cells detected by CMV antigen-recall assay The PBMCs from anti-CMV IgG positive donor were induced by CMV pp65 (495-503) polypeptide to produce CMV pp65 specific CD8 T cells, which were served as effector cells. The colo205 tumor cell line pulsed with pp65 was used as the target cell. In such experimental system, the effect of anti-PVRIG antibodies on improving the functions of the antigen-specific CD8 T cell was detected. The CMV IgG+ PBMCs were resuscitated, resuspended to 2M06 mL by using complete medium (RPMI1640-Glutamax + 5% AB serum + 1% P / S+(1*) 2—P mercaptoethanol) containing 1 mg / mL CMV pp65(495-503) peptide (Anaspec, Cat No.AS-28328), 2 ng / mL human IL-2 (R&D, Cat No.IL-202) and 10 ng / mL human IL-7 (Peprotech, Cat No.200-07), inoculated in a 6-well plate (5 mL / well) and incubated at 37°C with 5% CO2 for 6 days. On day 6, all the PBMCs were collected, and pp65 and IL-7 in the medium were removed. The cells were divided into two portions, resuspended in complete medium containing 100 IU / mL human IL-2, and cultured for another 2 days. On day 8, all the PBMCs were collected and resuspended in complete medium containing 100 lU / mL human IL-2, and the cell density was adjusted to 2x106 / mL for continuing cultivation. On day 11, all the PBMCs were collected. The percentage of CD8 T cells and CMV pp65 (495-503) specific CD8 T in PBMC and the expression of PVRIG, TIGIT and PD-1 on the cells were detected by flow cytometry, as shown in Fig. 33A, B, after inducing by CMV pp65 (495-503), the percentage of CMV pp65 (495-503) specific CD8 T exceeds 80%. Fig. 33B shows that pp65+CD8+ T (donor021) expresses different levels of PVRIG, TIGIT, PD-1 and CD226. The detection antibodies in the flow cytometry were as follows: Livedead Near IR (Invitrogen, Cat No.L34976), CD8-PerCp Cy5.5 (BD, Cat No.565310), CD3-PE-Cy7 (Biolegend, Cat No.300316), T-select HLA-A*0201 CMV pp65 Tetramer-PE (MBL, Cat No.TS-0010-1C), PVRIG-AF488 (R&D, Cat No.FAB93651G-100UG), TIGIT-APC (Biolegend, Cat No.372706) and PD-1-BV421 (BD, Cat No.562516). CD8 T cells were isolated from the induced PBMCs as effector cells by CD8 T cell isolation kit (Stemcell, Cat No.17953) and resuspended with AIM-V to the cell density of 0.4'106 mL. The expression of CD226 and purity of isolated CD8 T were detected. The target cells, Colo205, were digested by TrypLETM Express Enzyme (Gibco, Cat No.12605010), resuspended in AIM-V (Gibco, Cat No.31035-025) containing 20 ng / mL pp65 to the cell density of 1'106 / mL and treated at 37OC with 5% CO2 for 3 hours. The target cells were then centrifuged at 250g for 5 minutes and resuspended in AIM-V to the cell density of 0.5'106 / mL after discarding the supernatant. PVRL2, PVR and PD-L1 were highly expressed on Colo205 detected by flow cytometry (Fig. 33C). Anti-PVRIG antibodies and negative controls were diluted with AIM-V to 280 nM. The low-attachment, 96-well, U-bottom plate (Corning, Cat No.7007) was added with 50 pL of antibody, 50 pL of CD8 T and 100 pL of pp65-treated colo205 in order, mixed well with a multichannel pipette, and incubated at 37°C with 5% CO2 for 18 hours. In the experimental system, the final concentration of antibodies was 70 nM, CD8 T cells were 20,000 / well, and colo205 were 20,000 / well. After incubation, the supernatant was collected by centrifugation at 400 g, and the level of human IFN-Y in the supernatant was detected with an ELISA kit (Dakewe, Cat No.1110003). In this system, the positive controls were COM701-hIgG4 and SRF813-hIgG1, and the negative control was no treatment. As shown in Figure 33D, compared with the no treatment group, most of the tested PVRIG antibodies significantly increase the secretion of IFN-Y in the cell supernatant. The detection antibodies used in flow cytometry to detect the purity of isolated CD8 T and the expression of CD226 were as follows: Livedead-BV421 (Invitrogen, Cat No.L34964), CD8-FITC (BD, Cat No.555366) and CD226-PE-Cy7 (Biolegend, Cat No.338316). The detection antibodies used in flow cytometry to detect the expression of PVRL2, PVR, PD-L1 and HLA-A2 on Colo205 cells were as follows: livedead-BV421 (Invitrogen, Cat No.L34964), PVRL2-APC (Biolegend, Cat No.337412), PVR-PerCp Cy5.5 (Biolegend, Cat No.337612), PD-L1-PE-Cy7 (BD, Cat No.558017) and HLA-A2-PE (Biolegend, Cat No.343306). EXAMPLE 17 - Humanization of alpaca anti-human PVRIG antibody By comparing the IMGT (http: / / imgt.cines.fr) human antibody heavy and light chain variable region germline gene database, the heavy chain and light chain variable region germline genes with high homology with alpaca antibody were selected as templates. The CDRs of alpaca antibody were grafted into corresponding human templates to form a variable region sequence of FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. The key amino acids in the skeleton sequence were back mutated to the amino acids corresponding to the alpaca antibody as needed to ensure the original affinity, and thus to obtain the humanized anti-PVRIG monoclonal antibody. 1. Humanization of PVRIG-A50 The humanized heavy chain templates of the alpaca antibody PVRIG-A50 are IGHV3-23*04 and IGHJ3*01. The CDRs of the alpaca antibody (PVRIG-A50) were grafted into the human template thereof to obtain the corresponding humanized version. The CDRs amino acid residues of the antibody were determined and annotated by the Kabat numbering scheme. The key amino acids in the FR region sequence in the humanized antibody of PVRIG-A50 were back mutated to the amino acids corresponding to the alpaca antibody as needed to ensure the original affinity. PVRIG-A50 antibody is prone to have chemical modifications on NG site and Glycosylation on NLS. Inventors performed point mutations on NG / NLS to eliminate the risk of modification. The specific design is shown in Table 28. Table 28. The design of humanized antibody of PVRIG-A50 VHH H1 Grafted(IGHV3-23*04) + A97V,K98E H1a Grafted(IGHV3-23*04) + A97V,K98E + N54D H1b Grafted(IGHV3-23*04) + A97V,K98E + N54D,N108S H1c Grafted(IGHV3-23*04) + A97V,K98E + N54D,S110A H1d Grafted(IGHV3-23*04) + A97V,K98E + N108S H1e Grafted(IGHV3-23*04) + A97V,K98E + G55A, N108S H2a Grafted(IGHV3-23*04) + S75T,A97V,K98E + G55A, N108S Note: Grafted (IGHV3-23*04) represents the insertion of CDRs of the target antibody into the FR region sequence in human germline IGHV3-23*04; the first + A97V represents the A at position 97 of Grafted is back mutated to V, the second + N54D represents the point mutation on NG site, and so on. The back-mutated amino acids are numbered in natural order, the same below. The specific variable region sequences in humanized antibody of PVRIG-A50 are as follows: The amino acid sequence of A50.VH1(PVRIG-A50-H1) is shown in SEQ ID NO: 198 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKGLEWVSTINSNGGRT SYVDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCVEGDPHNFGLENLSLRDFGSW GQGTMVTVSS The amino acid sequence of A50.VH1a(PVRIG-A50-H1a) is shown in SEQ ID NO: 199 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKGLEWVSTINSDGGRT SYVDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCVEGDPHNFGLENLSLRDFGSW GQGTMVTVSS The amino acid sequence of A50.VH1b(PVRIG-A50-H1b) is shown in SEQ ID NO: 200 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKGLEWVSTINSDGGRT SYVDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCVEGDPHNFGLESLSLRDFGSW GQGTMVTVSS The amino acid sequence of A50.VH1c(PVRIG-A50-H1c) is shown in SEQ ID NO: 201 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKGLEWVSTINSDGGRT SYVDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCVEGDPHNFGLENLALRDFGS WGQGTMVTVSS The amino acid sequence of A50.VH1d(PVRIG-A50-H1d) is shown in SEQ ID NO: 202 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKGLEWVSTINSNGGRT SYVDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCVEGDPHNFGLESLSLRDFGSW GQGTMVTVSS The amino acid sequence of VH1e(PVRIG-A50-H1e) is shown in SEQ ID NO: 203 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKGLEWVSTINSNAGRT SYVDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCVEGDPHNFGLESLSLRDFGSW GQGTMVTVSS The amino acid sequence of A50.VH2a(PVRIG-A50-H2a) is shown in SEQ ID NO: 204 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGKGLEWVSTINSNAGRT SYVDSVKGRFTISRDNTKNTLYLQMNSLRAEDTAVYYCVEGDPHNFGLESLSLRDFGSW GQGTMVTVSS 5       The amino acid sequence of the humanized heavy chain template IGHV3-23*04 is shown in SEQ ID NO: 205 EVQLVESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSAISGSGGSTY YADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAK The amino acid sequence of the humanized heavy chain template IGHJ3*01 is shown in SEQ 10 ID NO: 206 WGQGTMVTVSS According to Kabat numbering scheme, the analysis results of VH sequences in above 7 humanized antibodies are shown in Table 29. Table 29 Kabat analysis results of VH sequences in humanized antibody of PVRIG-A50 Heavy chain CDR1 CDR2 CDR3 VH1 YYDMS SEQ ID NO: 168 TINSNGGRTSYVDSVKG SEQ ID NO: 169 GDPHNFGLENLSLRDFGS SEQ ID NO: 170 VH1a YYDMS SEQ ID NO: 168 TINSDGGRTSYVDSVKG SEQ ID NO: 207 GDPHNFGLENLSLRDFGS SEQ ID NO: 170 VH1b YYDMS SEQ ID NO: 168 TINSDGGRTSYVDSVKG SEQ ID NO: 207 GDPHNFGLESLSLRDFGS SEQ ID NO: 208 VH1c YYDMS SEQ ID NO: 168 TINSDGGRTSYVDSVKG SEQ ID NO: 207 GDPHNFGLENLALRDFGS SEQ ID NO: 209 VH1d YYDMS SEQ ID NO: 168 TINSNGGRTSYVDSVKG SEQ ID NO: 169 GDPHNFGLESLSLRDFGS SEQ ID NO: 208 VH1e / H2a YYDMS SEQ ID NO: 168 TINSNAGRTSYVDSVKG SEQ ID NO: 210 GDPHNFGLESLSLRDFGS SEQ ID NO: 208 15       2. Humanization of PVRIG-A105 The humanized heavy chain templates of the alpaca antibody PVRIG-A105 are IGHV3-7*01 and IGHJ3*01. The CDRs of the alpaca antibody (PVRIG-A105) were grafted into the human template thereof to obtain the corresponding humanized version. The CDRs amino acid residues of the antibody were determined and annotated by the IMGT numbering scheme. The key amino acids in the FR region sequence in the humanized antibody of PVRIG-A105 were back mutated to the amino acids corresponding to the alpaca antibody as needed to ensure the original affinity. PVRIG-A105 antibody has two free cysteines. In order to improve the stability of the antibody, Inventors performed point mutations on Cys. The specific design is shown in Table 30. Table 30. The design of humanized antibody of PVRIG-A105 VHH H1 Grafted(IGHV3-7*01) + V37F,G44E,L45R,W47F,N50T H2 Grafted(IGHV3-7*01) + S35T,V37F,G44E,L45R,W47F,N50T H3 Grafted(IGHV3-7*01) + S35T,V37F,G44E,L45R,W47F,N50T,L79V H3a Grafted(IGHV3-7*01) + S35T,V37F,G44E,L45R,W47F,N50T,L79V + C103S,C108S H4 Grafted(IGHV3-7*01) + S35T,V37F,G44E,L45R,W47F,N50T,V61S,D62H H5 Grafted(IGHV3-7*01) + S35T,V37F,G44E,L45R,W47F,N50T,T122I,M123Q Note: Grafted (IGHV3-7*01) represents the insertion of CDRs of the target antibody into the FR region sequence in human germline IGHV3-7*01; the first + V37F represents the V at position 37 of Grafted is back mutated to F, the second + C103S represents the point mutation on Cys site, and so on. The back-mutated amino acids are numbered in natural order, the same below. The specific variable region sequences in humanized antibody of PVRIG-A105 are as follows: The amino acid sequence of A105.VH1(PVRIG-A105-H1) is shown in SEQ ID NO: 211 EVQLVESGGGLVQPGGSLRLSCAASGRTFDRHTMSWFRQAPGKEREFVATASRIPGDTY YVDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAATSAYCSEVDCYEKGSWYDN WGQGTMVTVSS The amino acid sequence of A105.VH2(PVRIG-A105-H2) is shown in SEQ ID NO: 212 EVQLVESGGGLVQPGGSLRLSCAASGRTFDRHTMTWFRQAPGKEREFVATASRIPGDTY YVDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAATSAYCSEVDCYEKGSWYDN WGQGTMVTVSS The amino acid sequence of A105.VH3(PVRIG-A105-H3) is shown in SEQ ID NO: 213 EVQLVESGGGLVQPGGSLRLSCAASGRTFDRHTMTWFRQAPGKEREFVATASRIPGDTY YVDSVKGRFTISRDNAKNSVYLQMNSLRAEDTAVYYCAATSAYCSEVDCYEKGSWYDN WGQGTMVTVSS The amino acid sequence of A105.VH4(PVRIG-A105-H4) is shown in SEQ ID NO: 214 EVQLVESGGGLVQPGGSLRLSCAASGRTFDRHTMTWFRQAPGKEREFVATASRIPGDTY YSHSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAATSAYCSEVDCYEKGSWYDN WGQGTMVTVSS The amino acid sequence of A105.VH5(PVRIG-A105-H5) is shown in SEQ ID NO: 215 EVQLVESGGGLVQPGGSLRLSCAASGRTFDRHTMTWFRQAPGKEREFVATASRIPGDTY YVDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAATSAYCSEVDCYEKGSWYDN WGQGIQVTVSS The amino acid sequence of A105.VH3a(PVRIG-A105-H3a) is shown in SEQ ID NO: 216 EVQLVESGGGLVQPGGSLRLSCAASGRTFDRHTMTWFRQAPGKEREFVATASRIPGDTY YVDSVKGRFTISRDNAKNSVYLQMNSLRAEDTAVYYCAATSAYSSEVDSYEKGSWYDN WGQGTMVTVSS The amino acid sequence of the humanized heavy chain template IGHV3-7*01 is shown in SEQ ID NO: 217 EVQLVESGGGLVQPGGSLRLSCAASGFTFSSYWMSWVRQAPGKGLEWVANIKQDGSEK YYVDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAR The amino acid sequence of the humanized heavy chain template IGHJ3*01 is shown in SEQ ID NO: 206 WGQGTMVTVSS According to IMGT numbering scheme, the analysis results of VH sequences in above 6 humanized antibodies are shown in Table 31. Table 31 IMGT analysis results of VH sequences in humanized antibody of PVRIG-A105 Heavy chain CDR1 CDR2 CDR3 VH1 / 2 / 3 / 4 / 5 GRTFDRHT SEQ ID NO: 147 ASRIPGDT SEQ ID NO: 148 AATSAYCSEVDCYEKGSWYD N SEQ ID NO: 149 VH3a GRTFDRHT SEQ ID NO: 147 ASRIPGDT SEQ ID NO: 148 AATSAYSSEVDSYEKGSWYDN SEQ ID NO: 218 2. Humanization of PVRIG-A118 The humanized heavy chain templates of the alpaca antibody PVRIG-A118 are IGHV3-7*01 and IGHJ3*01. The CDRs of the alpaca antibody (PVRIG-A118) were grafted into the human template thereof to obtain the corresponding humanized version. The CDRs amino acid residues of the antibody were determined and annotated by the IMGT numbering scheme. The key amino acids in the FR region sequence in the humanized antibody of PVRIG-A118 were back mutated to the amino acids corresponding to the alpaca antibody as needed to ensure the original affinity. The specific design is shown in Table 32. Table 32. The design of humanized antibody of PVRIG-A118 VHH H1 Grafted(IGHV3-7*01) + S35G,V37Y,G44D,L45R,W47L,N50T H2 Grafted(IGHV3-7*01) + S35G,V37Y,G44D,L45R,W47L,N50T,Y58K H3 Grafted(IGHV3 -7*01) + S35G,V37Y,G44D,L45R,W47L,N50T,Y58K,D72G,N73D H4 Grafted(IGHV3 -7*01) + S35G,V37Y,G44D,L45R,W47L,N50T,Y58K,D72G,N73D,Y79S H5 Grafted(IGHV3 -7*01) + S35G,V37Y,G44D,L45R,W47L,N50T,Y58K,D72G,N73D,L78V H6 Grafted(IGHV3 -7*01) + S35G,V37Y,G44D,L45R,W47L,N50T,Y58K,Y59I,D72G,N73D H7 Grafted(IGHV3 -7*01) + S35G,V37Y,G44D,L45R,W47L,N50T,Y58K,D72G,N73D,Y94F Note: Grafted (IGHV3-7*01) represents the insertion of CDRs of the target antibody into the FR region sequence in human germline IGHV3-7*01; the first + S35G represents the S at position 35 of Grafted is back mutated to G, and so on. The back-mutated amino acids are numbered in natural order, the same below. The specific variable region sequences in humanized antibody of PVRIG-A118 are as follows: The amino acid sequence of A118.VH1 (PVRIG-A118-H1) is shown in SEQ ID NO: 219 EVQLVESGGGLVQPGGSLRLSCAASETYFDLYVMGWYRQAPGKDRELVATITYTGSIYY VDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCNADPSGLGRKVYWGQGTMVTV SS The amino acid sequence of A118.VH2 (PVRIG-A118-H2) is shown in SEQ ID NO: 220 EVQLVESGGGLVQPGGSLRLSCAASETYFDLYVMGWYRQAPGKDRELVATITYTGSIKY VDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCNADPSGLGRKVYWGQGTMVTV SS The amino acid sequence of A118.VH3 (PVRIG-A118-H3) is shown in SEQ ID NO: 221 EVQLVESGGGLVQPGGSLRLSCAASETYFDLYVMGWYRQAPGKDRELVATITYTGSIKY VDSVKGRFTISRGDAKNSLYLQMNSLRAEDTAVYYCNADPSGLGRKVYWGQGTMVTV SS The amino acid sequence of A118.VH4 (PVRIG-A118-H4) is shown in SEQ ID NO: 222 EVQLVESGGGLVQPGGSLRLSCAASETYFDLYVMGWYRQAPGKDRELVATITYTGSIKY VDSVKGRFTISRGDAKNSLSLQMNSLRAEDTAVYYCNADPSGLGRKVYWGQGTMVTV SS The amino acid sequence of A118.VH5 (PVRIG-A118-H5) is shown in SEQ ID NO: 223 EVQLVESGGGLVQPGGSLRLSCAASETYFDLYVMGWYRQAPGKDRELVATITYTGSIKY VDSVKGRFTISRGDAKNSVYLQMNSLRAEDTAVYYCNADPSGLGRKVYWGQGTMVTV SS The amino acid sequence of A118.VH6 (PVRIG-A118-H6) is shown in SEQ ID NO: 224 EVQLVESGGGLVQPGGSLRLSCAASETYFDLYVMGWYRQAPGKDRELVATITYTGSIKIV DSVKGRFTISRGDAKNSLYLQMNSLRAEDTAVYYCNADPSGLGRKVYWGQGTMVTVSS The amino acid sequence of A118.VH7 (PVRIG-A118-H7) is shown in SEQ ID NO: 225 EVQLVESGGGLVQPGGSLRLSCAASETYFDLYVMGWYRQAPGKDRELVATITYTGSIKY VDSVKGRFTISRGDAKNSLYLQMNSLRAEDTAVYFCNADPSGLGRKVYWGQGTMVTVS S The amino acid sequence of the humanized heavy chain template IGHV3-7*01 is shown in SEQ ID NO: 217 EVQLVESGGGLVQPGGSLRLSCAASGFTFSSYWMSWVRQAPGKGLEWVANIKQDGSEK YYVDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAR The amino acid sequence of the humanized heavy chain template IGHJ3*01 is shown in SEQ ID NO: 206 WGQGTMVTVSS According to IMGT numbering scheme, the analysis results of VH sequences in above 7 humanized antibodies are shown in Table 33. Table 33 IMGT analysis results of VH sequences in humanized antibody of PVRIG-A118 Heavy chain CDR1 CDR2 CDR3 VH1 / 2 / 3 / 4 / 5 / 6 / 7 ETYFDLYV SEQ ID NO: 156 ITYTGSI SEQ ID NO: 157 NADPSGLGRKVY SEQ ID NO: 158 EXAMPLE 18 - The specific binding of PVRIG humanized antibodies to human and cynomolgus monkey PVRIG protein detected by ELISA The ELISA plate was pre-coated with 100 til. well of 0.5 gg / mL human PVRIG-his (AcroBiosystems, Cat NO.PVG-H52H4) or cynomolgus monkey PVRIG (Novoprotein, Cat NO.C09B). The tested anti-PVRIG humanized antibodies were serially diluted (initial concentration 3nM, 3-fold serial dilution), added into the plate (100 pL well) and incubated with shaking at room temperature for 1.5 hours. After washing the plate, the mouse anti-human IgG Fc-5 HRP (Jackson ImmunoResearch, Cat NO.209-035-098)) working solution (dilution ratio 1:10000, 100 pL well) were added into wells and incubated with shaking at room temperature for 1.0 hour. A HRP substrate, TMB (Thermo, Cat NO.34029) was added into wells after washing the plate again for color development. After adding a termination solution to terminate the reaction, the absorbance value was read by a microplate reader (MD i3x). An antibody binding curve was drawn 10 with concentration of antibodies as the abscissa and the corresponding OD value as the ordinate. The EC50 value was calculated by using a Four Parameter Logistic Fit (GraphPadPrism9). The smaller the EC50 value, the stronger the ability of the antibodies binds to human or cynomolgus monkey PVRIG. The binding effect of the humanized antibodies was normalized to the corresponding parental antibodies. If the percentage value is higher than 100%, the binding effect 15 of humanized antibodies is better than that of parental antibody. As shown in Fig. 34A, Fig. 34B and Table 34, most of the humanized antibodies of PVRIG-A50, PVRIG-A105 and PVRIG-A118 specifically bind to the human / cynomolgus monkey PVRIG proteins. Table 34. The specific binding of anti-PVRIG humanized antibodies to human or cynomolgus monkey PVRIG protein detected by ELISA No. Name Bind to human PVRIG protein ^<SB PVRIG $ ££& EC50 % relative activity EC50 % relative activity (PM) parental antibody* / tested antibody (PM) parental antibody / tested antibody 1 PVRIG-A50-H1 87.12 105.02% 129.6 95.06% 2 PVRIG-A50-H1a 99.04 92.38% 185.5 66.42% 3 PVRIG-A50-H1c 104.8 87.30% 177.4 69.45% 4 PVRIG-A50-H1d 77.4 118.20% 110.6 111.39% 5 PVRIG-A50-H1e 79.06 115.72% 137.6 89.53% 6 PVRIG-A50-H2a 65.51 139.66% 180.4 68.29% 7 PVRIG-A50-H1b 81.94 111.29% 161.3 75.76% 8 PVRIG-A105-H1 48.71 149.85% 30.28 106.51% 9 PVRIG-A105-H2 51.51 141.70% 34.33 93.94% 10 PVRIG-A105-H3 58.53 124.71% 35.16 91.72% 11 PVRIG-A105-H3a 7408 0.99% 1273 2.53% 12 PVRIG-A105-H4 52.07 140.18% 36.07 89.41% 13 PVRIG-A105-H5 67.76 107.72% 37.65 85.66% 14 PVRIG-A118-H1 45.63 109.69% 584.9 4.62% 15 PVRIG-A118-H2 49.08 101.98% 593.7 4.55% 16 PVRIG-A118-H3 49.97 100.16% 79.41 34.05% 17 PVRIG-A118-H4 52.22 95.84% 72.9 37.09% 18 PVRIG-A118-H5 48.05 104.16% 67.31 40.17% 19 PVRIG-A118-H6 55.14 90.77% 55.9 48.37% 20 PVRIG-A118-H7 50.11 99.88% 81.67 33.11% * parental antibody: the corresponding parental antibody before humanization EXAMPLE 19 - The binding of anti-PVRIG humanized antibodies to human PVRIG on FlpinCHO-PVRIG and cynomolgus monkey PVRIG on FlpinCHO-cyno PVRIG detected by FACS 5        The methods refer to Example 19. The stronger the AUC value, the stronger the ability of the humanized antibodies binds to FlpinCHO-human / cyno PVRIG. Fig. 35A shows that the binding ability of most of the humanized antibodies of PVRIG-A50, PVRIG-A105 and PVRIG-A118 to human PVRIG on the surface of FlpinCHO-PVRIG cells is equivalent to that of their parental antibody. Fig. 35B shows that except for the humanized molecules of PVRIG-A118, the binding 10 of which to cynomolgus monkey PVRIG is significantly weaker than the corresponding parental antibody, most of the humanized molecules of PVRIG-A50 and A105 bind well to cynomolgus monkey PVRIG, and their binding ability is equivalent to that of the corresponding parental antibody. The binding activity of the humanized antibodies was normalized to the positive control COM701-hIgG1 and SRF813-hIgG1. As shown in Table 35, the higher the percentage value, the 15 stronger the binding ability of the antibodies. Table 35. Binding of anti-PVRIG humanized antibodies to human / cynomolgus monkey PVRIG on the surface of overexpressing cells PVRIG humanized antibody human PVRIG binding cyno PVRIG binding AUC% AUC% to COM701-hIgG1 to SRF813- hIgG1 to parental Ab* to COM701-hIgG1 to SRF813- hIgG1 to parental Ab PVRIG-50 73.39 129.95 100.00 47.06 57.77 100.00 PVRIG-A50-H1 74.39 131.71 101.35 45.16 55.45 95.98 PVRIG-A50-H1a 70.42 124.68 95.95 40.69 49.96 86.48 PVRIG-A50-H1c 69.16 122.46 94.24 38.81 47.64 82.47 PVRIG-A50-H1d 63.06 111.65 85.92 48.59 59.66 103.27 PVRIG-A50-H1e 60.31 106.79 82.18 43.31 53.17 92.03 PVRIG-A50-H2a 61.92 109.64 84.37 42.42 52.08 90.16 PVRIG-A50-H1b 60.53 107.17 82.47 34.38 42.21 73.06 PVRIG-A118 136.50 190.93 100.00 62.41 76.11 100.00 PVRIG-A118-H1 127.17 177.87 93.16 11.52 14.05 18.46 PVRIG-A118-H2 124.64 174.34 91.31 14.61 17.82 23.41 PVRIG-A118-H3 128.93 180.34 94.45 20.97 25.58 33.60 PVRIG-A118-H4 132.42 185.22 97.01 21.32 26.00 34.17 PVRIG-A118-H5 128.19 179.30 93.91 21.84 26.63 34.99 PVRIG-A118-H6 128.62 179.90 94.22 19.86 24.22 31.82 PVRIG-A118-H7 126.47 176.90 92.65 19.71 24.03 31.57 PVRIG-A105 123.99 186.43 100.00 87.47 108.46 100.00 PVRIG-A105-H1 122.83 184.68 99.07 81.31 100.83 92.96 PVRIG-A105-H2 127.59 191.84 102.91 84.20 104.41 96.26 PVRIG-A105-H3 125.38 188.52 101.12 86.38 107.11 98.75 PVRIG-A105-H3a 15.92 23.94 12.84 3.27 4.05 3.74 PVRIG-A105-H4 115.40 173.51 93.07 80.31 99.59 91.82 PVRIG-A105-H5 111.54 167.70 89.96 78.82 97.74 90.11 * parental antibody: the corresponding parental antibody before humanization EXAMPLE 20 - The affinity of anti-PVRIG humanized antibodies for human PVRIG proteins detected by BIAcore The methods refer to Example 10. As shown in Table 36, all the anti-PVRIG humanized 5 antibodies specifically bind to human PVRIG proteins with high affinity. Table 36. BIAcore results of specific binding of anti-PVRIG antibodies for human PVRIG proteins Name Antigen KD (M) ka (1 / Ms) kd (1 / s) PVRIG-A50 Human PVRIG 4.30E-11 4.43E+05 1.90E-05 PVRIG-A50-H1 3.62E-11 4.10E+05 1.48E-05 PVRIG-A50-H1a 4.51E-11 4.25E+05 1.92E-05 PVRIG-A50-H1c 4.62E-11 4.38E+05 2.03E-05 PVRIG-A50-H1d 5.45E-11 2.62E+05 1.43E-05 PVRIG-A50-H1e 5.51E-11 2.65E+05 1.46E-05 PVRIG-A50-H2a 5.85E-11 2.76E+05 1.61E-05 PVRIG-A50-H1b 8.04E-11 3.50E+05 2.82E-05 PVRIG-A118 4.83E-10 2.12E+06 1.02E-03 PVRIG-A118-H1 4.78E-10 1.38E+06 6.58E-04 PVRIG-A118-H2 5.17E-10 1.59E+06 8.20E-04 PVRIG-A118-H3 1.35E-09 2.59E+06 3.49E-03 PVRIG-A118-H4 1.21E-09 2.69E+06 3.27E-03 PVRIG-A118-H5 1.37E-09 2.79E+06 3.83E-03 PVRIG-A118-H6 1.10E-09 2.71E+06 2.99E-03 PVRIG-A118-H7 1.08E-09 2.41E+06 2.60E-03 PVRIG-A105 6.75E-10 7.21E+05 4.86E-04 PVRIG-A105-H1 9.32E-10 1.39E+06 1.30E-03 PVRIG-A105-H2 6.93E-10 1.33E+06 9.20E-04 PVRIG-A105-H3 6.22E-10 9.25E+05 5.76E-04 PVRIG-A105-H4 7.96E-10 9.72E+05 7.73E-04 PVRIG-A105-H5 8.18E-10 1.11E+06 9.10E-04 EXAMPLE 21 - The effect of anti-PVRIG humanized antibodies blocking the binding between PVRIG and PVRL2 detected by ELISA The methods refer to Example 11. The inhibiting activity of the humanized antibodies was normalized to the corresponding parental antibodies. If the percentage value is higher than 100%, 5 the inhibiting effect of humanized antibodies is better than that of parental antibodies. Fig. 36 shows the blocking curves of the humanized antibodies, and Table 37 shows the inhibiting activity. As shown in Fig. 36 and Table 37, most humanized antibodies of PVRIG-A50, PVRIG-A105 and PVRIG-A118 can significantly inhibit the binding of human PVRIG to human PVRL2 protein. Table 37. The effect of anti-PVRIG humanized antibodies blocking the binding of PVRIG 10                                    to PVRL2 No. Name Blocking the binding of PVRIG to PVRL2 IC50 (PM) % relative activity AUC % relative activity Parental antibody* / test antibody Parental antibody / test antibody 1 PVRIG-A50-H1 1003 77.45% 2.423 82.62% 2 PVRIG-A50-H1a 1064 73.01% 2.426 82.52% 3 PVRIG-A50-H1c 1178 65.94% 2.625 76.27% 4 PVRIG-A50-H1d 1035 75.05% 2.405 83.24% 5 PVRIG-A50-H1e 977.2 79.49% 2.323 86.18% 6 PVRIG-A50-H2a 910.8 85.29% 2.317 86.40% 7 PVRIG-A50-H1b 847.3 124.63% 1.999 104.60% 8 PVRIG-A105-H1 555.10 132.57% 2.271 111.76% 9 PVRIG-A105-H2 577.00 127.54% 2.270 111.81% 10 PVRIG-A105-H3 751.70 97.90% 2.523 100.59% 11 PVRIG-A105-H3a no blocking 5.374 47.23% 12 PVRIG-A105-H4 686.50 107.20% 2.320 109.40% 13 PVRIG-A105-H5 646.90 113.76% 2.397 105.88% 14 PVRIG-A118-H1 355.4 156.67% 1.308 120.87% 15 PVRIG-A118-H2 357.0 155.97% 1.351 117.02% 16 PVRIG-A118-H3 385.8 144.32% 1.494 105.82% 17 PVRIG-A118-H4 462.1 120.49% 1.418 111.50% 18 PVRIG-A118-H5 380.6 146.30% 1.389 113.82% 19 PVRIG-A118-H6 422.4 131.82% 1.467 107.77% 20 PVRIG-A118-H7 430.1 129.46% 1.381 114.48% * parental antibody: the corresponding parental antibody before humanization EXAMPLE 22 - The effect of anti-PVRIG humanized antibodies blocking CHO-K1-CD112 cells binding to human PVRIG-mFc proteins detected by FACS The methods refer to Example 12. As shown in Fig. 37, most humanized antibodies can block 5 the binding of CHO-K1-CD112 cells to human PVRIG-mFc protein. The blocking activity of the humanized antibodies was normalized to the positive control COM701-hIgG1 and SRF813-hIgG1. The smaller the percentage value, the stronger the blocking ability of the antibodies. Table 38 Anti-PVRIG humanized antibodies block CHO-K1-CD112 cells from binding to human PVRIG-mFc protein PVRIG humanized antibodies CHO-K1-CD112 FACS blocking %IC50 %AUC to COM701-hIgG4 to SO35- hIgG1 to parental Ab to COM701-hIgG4 to SO35- hIgG1 to parental Ab PVRIG-50 117.03 90.60 100.00 122.20 72.06 100.00 PVRIG-A50-H1 100.74 77.99 86.08 114.89 67.75 94.02 PVRIG-A50-H1a 120.08 92.96 102.60 124.64 73.49 101.99 PVRIG-A50-H1c 122.31 94.69 104.52 131.78 77.71 107.84 PVRIG-A50-H1d 115.12 89.12 98.37 120.57 71.09 98.66 PVRIG-A50-H1e 130.27 100.85 111.31 125.71 74.13 102.87 PVRIG-A50-H2a 118.46 91.71 101.22 111.98 66.03 91.64 PVRIG-A50-H1b 95.25 73.74 81.39 104.45 61.59 85.47 PVRIG-A118 140.78 95.33 100.00 118.66 61.71 100.00 PVRIG-A118-H1 140.86 95.38 100.05 119.93 62.37 101.07 PVRIG-A118-H2 161.48 109.34 114.70 141.80 73.74 119.50 PVRIG-A118-H3 145.90 98.79 103.64 126.51 65.79 106.61 PVRIG-A118-H4 140.95 95.44 100.12 116.15 60.40 97.88 PVRIG-A118-H5 143.88 97.43 102.20 127.17 66.13 107.17 PVRIG-A118-H6 144.51 97.85 102.65 118.71 61.73 100.04 PVRIG-A118-H7 143.39 97.09 101.85 123.72 64.34 104.26 PVRIG-A105 104.42 88.32 100.00 113.72 65.59 100.00 PVRIG-A105-H1 104.11 88.06 99.70 111.36 64.22 97.92 PVRIG-A105-H2 109.09 92.27 104.47 115.03 66.34 101.15 PVRIG-A105-H3 108.68 91.92 104.08 109.00 62.87 95.85 PVRIG-A105-H3a 3060.00 2588.24 2930.42 3530.47 2036.21 3104.48 PVRIG-A105-H4 110.13 93.15 105.47 111.95 64.57 98.44 PVRIG-A105-H5 110.44 93.42 105.77 118.62 68.42 104.31 EXAMPLE 23 - The cytotoxicity of NK cells for tumor cell line mediated by anti-PVRIG humanized antibodies detected by NK cell cytotoxicity assay The methods refer to Example 4 (g). As shown in Fig. 38, all the humanized antibodies can effectively promote the NK cell cytotoxicity for the target cells to varying degrees, wherein Panel 5 A shows that the promoting effect of PVRIG-A50-H1b and PVRIG-A50-H2a on NK cell cytotoxicity for target cells was equal to that of the parental antibody PVRIG-A50, Panel B shows that the promoting effect of PVRIG-A118-H3, H4, H5 and H6 on NK cell cytotoxicity for target cells was equal to that of the parental antibody PVRIG-A118, Panel C shows that the promoting effect of PVRIG-A105-H1, H2 and H3 on NK cell cytotoxicity for target cells was equal to that of 10 the parental antibody PVRIG-A105. EXAMPLE 24 - The effect of anti-PVRIG humanized antibodies on improving the functions of antigen-specific CD8 T cells detected by CMV antigen-recall assay The PBMCs were resuscitated, resuspended to 2^106 / mL by using complete medium (RPMI1640-Glutamax + 5% AB serum + 1% P / S+(1x) 2-0 mercaptoethanol) containing 1 mg / mL CMV pp65(495-503) peptide (Anaspec, Cat No.AS-28328), 2 ng / mL human IL-2 (R&D, Cat No.IL-202) and 10 ng / mL human IL-7 (Peprotech, Cat No.200-07), inoculated in a 6-well plate (5 mL / well) and incubated at 37°C with 5% CO2 for 6 days. On day 6, all the PBMCs were collected, and pp65 and IL-7 in the medium were removed. The cells were divided into two portions, resuspended in complete medium containing 100 IU / mL human IL-2, and cultured for another 2 days. On day 8, all the PBMCs were collected and resuspended in complete medium containing 100 lU / mL human IL-2, and the cell density was adjusted to 2*106 / mL. On day 11, all the PBMCs were collected. The proportion of CD8 T cells and CMV pp65 (495-503) specific CD8 T in PBMC (Fig. 24A) and the expression of PVRIG, TIGIT and PD-1 on the cells (Fig. 14B) were detected by flow cytometry. The detection antibodies in the flow cytometry were as follows: Livedead Near IR (Invitrogen, Cat No.L34976), CD8-PerCp Cy5.5 (BD, Cat No.565310), CD3-PE-Cy7 (Biolegend, Cat No.300316), T-select HLA-A*0201 CMV pp65 Tetramer-PE (MBL, Cat No.TS-0010-1C), PVRIG-AF488 (R&D, Cat No.FAB93651G-100UG), TIGIT-APC (Biolegend, Cat No.372706) and PD-1-BV421 (BD, Cat No.562516). CD8 T cells were isolated from the induced PBMCs as effector cells by CD8 T cell isolation kit (Stemcell, Cat No.17953) and resuspended with AIM-V to the cell density of 0.4*106 / mL. The expression of CD226 and purity of isolated CD8 T were detected. The target cells, Colo205, were digested by TrypLETM Express Enzyme (Gibco, Cat No.12605010), resuspended in AIM-V (Gibco, Cat No.31035-025) containing 20 ng / mL pp65 to the cell density of 1*106 / mL and treated at 37OC with 5% CO2 for 3 hours. The target cells were then centrifuged at 250 g for 5 minutes and resuspended in AIM-V to the cell density of 0.5*106 / mL after discarding the supernatant. Anti-PVRIG humanized antibodies and negative controls were diluted with AIM-V to 280 nM. The low-attachment, 96-well, U-bottom plate (Corning, Cat No.7007) was added with 50 pL of antibody, 50 pL of CD8 T and 100 pL of pp65-treated Colo205 in order, mixed well and incubated at 37°C with 5% CO2 for 18 hours. In the experimental system, the final concentration of antibodies was 70 nM, CD8 T cells were 20,000 / well, and Colo205 were 50,000 / well. After incubation, the supernatant was collected by centrifugation at 400 g, and the level of human IFN-Y in the supernatant was detected with an ELISA kit (Dakewe, Cat No.1110003). In this system, the positive controls were parental antibodies before humanization, and the negative control was no treatment. As shown in Fig. 39, compared with the no treatment group, humanized PVRIG antibodies can significantly increase the level of IFN-y in the cell supernatant. There was no significant difference between the humanized antibodies and their parental antibodies (one-way ANOVA analysis) in statistics, except that the effect of PVRIG-A105-H2 was significantly weaker than that of PVRIG-A105 (* p < 0.05, one-way ANOVA analysis). The detection antibodies used 5 in flow cytometry to detect the purity of isolated CD8 T and the expression of CD226 were as follows: Live / dead-BV421 (Invitrogen, Cat No.L34964), CD8-FITC (BD, Cat No.555366) and CD226-PE-Cy7 (Biolegend, Cat No.338316). The detection antibodies used in flow cytometry to detect the expression of PVRL2, PVR, PD-L1 and HLA-A2 on Colo205 cells were as follows: live / dead-BV421 (Invitrogen, Cat No.L34964), PVRL2-APC (Biolegend, Cat No.337412), PVR-10 PerCp Cy5.5 (Biolegend, Cat No.337612), PD-L1-PE-Cy7 (BD, Cat No.558017) and HLA-A2-PE (Biolegend, Cat No.343306). EXAMPLE 25 - Construct design of anti-PVRIGxTIGIT humanized bispecific antibodies The anti-PVRIG humanized VHH antibodies (PVRIG-A50-H1b, PVRIG-A105-H1) were linked to the N-terminus of the anti-TIGIT humanized monoclonal antibodies (TIGIT-002-H4L3, 15 TIGIT-005-H2L1d) heavy chain by G4S linker peptide to produce anti-PVRIGxTIGIT humanized bispecific antibodies (Fig. 40), which were named LC-BsAb-002, LC-BsAb-006, LC-BsAb-009 and LC-BsAb-010. Table 39 shows the sequences of heavy chain fusion polypeptide (HC) and light chain polypeptide (LC) of four bispecific antibodies. Table 39. The fusion polypeptide sequences of bispecific antibodies HC / LC Sequence No. Sequences LC-BsAb- 002 LC SEQ ID NO: 226 DIVMTQSPDSLAVSLGERATINCKASQNVRTAVAWYQQKPGQSPK LMIYSASYRYTGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQ YYTTPWTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLL NNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLT LSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC Note: TIGIT-002-H4L3 VL (single underline) LC-BsAb- 002 HC SEQ ID NO: 227 EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGK GLEWVSTINSDGGRTSYVDSVKGRFTISRDNSKNTLYLQMNSLR AEDTAVYYCVEGDPHNFGLESLSLRDFGSWGQGTMVTVSS GGG GSGGGGSGGGGSGGGGSEVQLQESGPGLVKPSETLSLTCAVSGYS ITSDSWNWIRQPPGKKLEYIGYISYSGNTYYNPSLKSRVTISRDTS KNQFSLKLSSVTAADTAVYYCARLDFSNYGGAVDYWGQGTTVT VSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWN SGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHK PSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDT LMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPRE EQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS KAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWE SNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCS VMHEALHNHYTQKSLSLSPGK Note: PVRIG-A50-H1b (double underline) + (G4S)4 Linker (italic) + TIGIT-002-H4L3 VH (single underline) LC-BsAb- 006 LC SEQ ID NO: 228 DIVMTQSPDSLAVSLGERATINCKASQHVSNAVAWYQHKPGQSP KLLIYSASYRYTGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQ HYNTPHTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLL NNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLT LSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC Note: TIGIT-005-H2L1d VL(single underline) LC-BsAb- SEQ ID NO: EVQLVESGGGLVQPGGSLRLSCAASGFTFSYYDMSWVRQAPGK 006 HC 229 GLEWVSTINSDGGRTSYVDSVKGRFTISRDNSKNTLYLQMNSLR AEDTAVYYCVEGDPHNFGLESLSLRDFGSWGQGTMVTVSS GGG GSGGGGSGGGGSGGGGSEVQLVQSGAEVKKPGASVKVSCKASG YAFTNYLIEWVRQAPGQRLEWMGVINPGSGGTNYKEKFKGRVTI TADKSSSTAYMELSSLRSEDTAVYYCARGEYFFFDYWGQGTTVT VSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWN SGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHK PSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDT LMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPRE EQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS KAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWE SNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCS VMHEALHNHYTQKSLSLSPGK Note: PVRIG-A50-H1b (double underline) + (G4S)4 Linker (italic) + TIGIT-005-H2L1d VH (single underline) LC-BsAb- 009 LC SEQ ID NO: 230 DIVMTQSPDSLAVSLGERATINCKASQNVRTAVAWYQQKPGQSPK LMIYSASYRYTGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQ YYTTPWTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLL NNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLT LSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC Note: TIGIT-002-H4L3 VL (single underline) LC-BsAb- SEQ ID NO: EVOLVESGGGLVOPGGSLRLSCAASGRTFDRHTMSWFROAPGKE 009 HC 231 REFVATASRIPGDTYYVDSVKGRFTISRDNAKNSLYLOMNSLRAE DTAVYYCAATSAYCSEVDCYEKGSWYDNWGOGTMVTVSS GGG GSGGGGSGGGGSGGGGSEVOLOESGPGLVKPSETLSLTCAVSGYS ITSDSWNWIROPPGKKLEYIGYISYSGNTYYNPSLKSRVTISRDTS KNOFSLKLSSVTAADTAVYYCARLDFSNYGGAVDYWGOGTTVT VSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWN SGALTSGVHTFPAVLOSSGLYSLSSVVTVPSSSLGTOTYICNVNHK PSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDT LMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPRE EOYNSTYRVVSVLTVLHODWLNGKEYKCKVSNKALPAPIEKTIS KAKGOPREPOVYTLPPSRDELTKNOVSLTCLVKGFYPSDIAVEWE SNGOPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWOOGNVFSCS VMHEALHNHYTOKSLSLSPGK Note: PVRIG-A105-H1 (double underline) + (G4S)4 Linker (italic) + TIGIT-002-H4L3 VH (single underline) LC-BsAb- 010 LC SEQ ID NO: 232 DIVMTOSPDSLAVSLGERATINCKASOHVSNAVAWYOHKPGOSP KLLIYSASYRYTGVPDRFSGSGSGTDFTLTISSLOAEDVAVYYCOO HYNTPHTFGGGTKVEIKRTVAAPSVFIFPPSDEOLKSGTASVVCLL NNFYPREAKVOWKVDNALOSGNSOESVTEODSKDSTYSLSSTLT LSKADYEKHKVYACEVTHOGLSSPVTKSFNRGEC Note: TIGIT-005-H2L1d VL (single underline) LC-BsAb- SEQ ID NO: EVOLVESGGGLVOPGGSLRLSCAASGRTFDRHTMSWFROAPGKE 010 HC 233 REFVATASRIPGDTYYVDSVKGRFTISRDNAKNSLYLOMNSLRAE DTAVYYCAATSAYCSEVDCYEKGSWYDNWGOGTMVTVSS GGG GSGGGGSGGGGSGGGGSEVOLVOSGAEVKKPGASVKVSCKASG YAFTNYLIEWVROAPGORLEWMGVINPGSGGTNYKEKFKGRVTI TADKSSSTAYMELSSLRSEDTAVYYCARGEYFFFDYWGOGTTVT VSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWN SGALTSGVHTFPAVLOSSGLYSLSSVVTVPSSSLGTOTYICNVNHK PSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDT LMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPRE EOYNSTYRVVSVLTVLHODWLNGKEYKCKVSNKALPAPIEKTIS KAKGOPREPOVYTLPPSRDELTKNOVSLTCLVKGFYPSDIAVEWE SNGOPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWOOGNVFSCS VMHEALHNHYTOKSLSLSPGK Note: PVRIG-A105-H1 (double underline) + (G4S)4 Linker (italic) + TIGIT-005-H2L1d VH (single underline) While constructing the bispecific antibody, the positive control antibody was also constructed. The anti-TIGIT positive control antibody is RG6058-hIgG1 (Roche), and the anti-PVRIG positive control antibody is COM701-hIgG4 (Compugen). See above for the corresponding amino acid sequences EXAMPLE 26 - The specific binding of anti-PVRIGxTIGIT humanized bispecific antibodies to human and cynomolgus monkey PVRIG protein detected by ELISA The ELISA plate was pre-coated with 50 gL / well of 1.0 gg / mL human PVRIG (AcroBiosystems, Cat NO.PVG-H52H4) or 50 gL / well of 0.5 gg / mL cynomolgus monkey PVRIG (Novoprotein, Cat NO.C09B). The tested antibodies were serially diluted (initial concentration 13 nM, 3-fold serial dilution, 12 concentration points), added into the plate (50 gL / well) and incubated at 37°C for 2.0 hours. After washing the plate, the Goat anti-human IgG Fc-HRP (Merck, Cat NO.AP113P) working solution (dilution ratio 1:5000, 50 gL / well) were added into wells and incubated at 37°C for 1.0 hour. A HRP substrate, TMB (KPL, Cat NO.5120-0077) was added into wells after washing the plate again and incubated at 37OC for 10 minutes for color development. After adding a termination solution to terminate the reaction, the absorbance value was read by a microplate reader (PE, Ensight-HH3400). An antibody binding curve was drawn with molar concentration of antibodies as the abscissa and the corresponding OD value as the ordinate. The EC50 value was calculated by using a Four Parameter Logistic Fit (GraphPadPrism9). The smaller the EC50 value, the stronger the ability of the antibodies binds to human or cynomolgus monkey PVRIG. The positive control antibodies are COM701-hIgG1, PVRIG-A50-H1b and PVRIG-A105-H1; and the negative control antibodies are anti-Fluorescein-hIgG1 (in house). The binding results of the 4 humanized bispecific antibodies to human PVRIG protein are shown in Fig. 41 and Tables 40 and 41, and the binding results to cynomolgus monkey PVRIG are shown in Fig. 42 and Tables 40 and 41. The data shows that all 4 humanized bispecific antibodies can specifically bind to human or cynomolgus monkey PVRIG protein. The binding of LC-BsAb-002 and LC-BsAb-006 to human PVRIG protein is slightly weaker than that of the corresponding monoclonal antibody (PVRIG-A50-H1b) and the positive control (COM701-hIgG1). The binding of LC-BsAb-009 and LC-BsAb-010 to human or cynomolgus monkey PVRIG protein is comparable to that of the corresponding monoclonal antibody (PVRIG-A105-H1) and superior to the positive control (COM701-hIgG1). Table 40. The specific binding of humanized bispecific antibodies to human or cynomolgus monkey PVRIG protein detected by ELISA Name Bind to human PVRIG protein Bind to cynomolgus monkey PVRIG protein EC50 % relative activity EC50 % relative activity (nM) PVRIG-A50-H1b / tested antibody COM701-hIgG1 / tested antibody (nM) PVRIG-A50- H1b / tested antibody COM701- hIgG1 / tested antibody LC-BsAb-002 0.5738 66.1% 43.2% 0.04402 86.4% 82.5% LC-BsAb-006 0.5072 74.8% 48.9% 0.03851 98.7% 94.3% PVRIG-A50-H1b 0.3793 100.0% 65.4% 0.03802 100.0% 95.5% COM701-hIgG1 0.2481 152.9% 100.0% 0.0363 104.7% 100.0% anti-Fluorescein- hIgG1 ~40328 / / ~ 3.999 / / Table 41. The specific binding of humanized bispecific antibodies to human or cynomolgus 5                      monkey PVRIG protein detected by ELISA Name Bind to human PVRIG protein Bind to cynomolgus monkey PVRIG protein EC50 % relative activity EC50 % relative activity (nM) PVRIG-A105-H1 / tested antibody COM701-hIgG1 / tested antibody (nM) PVRIG-A105-H1 / tested antibody COM701-hIgG1 / tested antibody LC-BsAb-009 0.06073 104.8% 408.5% 0.02563 93.4% 141.6% LC-BsAb-010 0.06166 103.2% 402.4% 0.02322 103.1% 156.3% PVRIG-A105-H1 0.06362 100.0% 390.0% 0.02395 100.0% 151.6% COM701-hIgG1 0.2481 25.6% 100.0% 0.0363 66.0% 100.0% anti-Fluorescein-hIgG 1 ~40328 / / ~ 3.999 / / EXAMPLE 27 - The specific binding of anti-PVRIGxTIGIT humanized bispecific antibodies to human and cynomolgus monkey TIGIT protein detected by ELISA The ELISA plate was pre-coated with 50 pL well of 4.0 pg mLgoat anti-mouse IgG antibody (Jackson, Cat NO. 115-006-071). After washing with blocking buffer, the plate was added with 50 10 pL well of 30 ng / mL human TIGIT ECD-mFc (in house) or cynomolgus monkey TIGIT ECD-mFc (in house) and incubated at 37°C for 2.0 hours. After washing the plate, the serial diluted antibodies (initial concentration 13 nM, 3-fold serial dilution, 12 concentration points) were added into the plate (50 pL / well) and incubated at 37°C for 2.0 hours. After washing the plate, the goat anti-human IgG Fc-HRP (Merck, Cat NO.AP113P)) working solution (dilution ratio 1:5000, 50 uL well) were added into wells and incubated at 37°C for 1.0 hour. A HRP substrate, TMB (KPL, Cat NO.5120-0077) was added into wells after washing the plate again for color development. After adding a termination solution to terminate the reaction, the absorbance value was read by a microplate reader (PE, Ensight-HH3400). An antibody binding curve was drawn with molar concentration of antibodies as the abscissa and the corresponding OD value as the ordinate. The EC50 value was calculated by using a Four Parameter Logistic Fit (GraphPad Prism9). The smaller the EC50 value, the stronger the ability of the antibodies binds to human or cynomolgus PVRIG. The positive control antibodies are RG6058-hIgG1, TIGIT-002-H4L3 and TIGIT-005-H2L1d; and the negative control antibodies are anti-Fluorescein-hIgG1 (in house). The binding results of the 4 humanized bispecific antibodies to human TIGIT protein are shown in Fig. 43 and Tables 42 and 43, and the binding results to cynomolgus monkey TIGIT are shown in Fig. 44 and Tables 42 and 43. The data shows that all 4 humanized bispecific antibodies can specifically bind to human or cynomolgus monkey TIGIT protein. The binding of LC-BsAb-002 and LC-BsAb-009 to human TIGIT protein is superior to that of the positive control (RG6058-hIgG1), and comparable to that of the corresponding monoclonal antibody (TIGIT-002-H4L3). In terms of binding to cynomolgus monkey TIGIT protein, LC-BsAb-009 is comparable to the corresponding monoclonal antibody (TIGIT-002-H4L3) and the positive control (RG6058-hIgG1); LC-BsAb-002 is slightly weaker than that of both the corresponding monoclonal antibody and the positive control. The binding of LC-BsAb-006 and LC-BsAb-010 to human or cynomolgus monkey TIGIT protein is superior to the corresponding monoclonal antibody (TIGIT-005-H2L1d) and the positive control (RG6058-hIgG1). Table 42. The specific binding of humanized bispecific antibodies to human or cynomolgus monkey TIGIT protein detected by ELISA Name Bind to human TIGIT protein Bind to cynomolgus monkey TIGIT protein EC50 %relative activity EC50 %relative activity (nM) TIGIT-002- H4L3 / tested antibody RG6058- hIgG1 / tested antibody (nM) TIGIT-002- H4L3 / tested antibody RG6058- hIgG1 / tested antibody LC-BsAb-002 0.03565 105.0% 141.3% 0.05535 74.9% 83.2% LC-BsAb-009 0.04549 82.3% 110.7% 0.04185 99.1% 110.0% TIGIT-002-H4L3 0.03742 100.0% 134.6% 0.04146 100.0% 111.1% RG6058-hIgG1 0.05038 74.3% 100.0% 0.04605 90.0% 100.0% anti-Fluorescein-hIgG 1 ~ 0.000 / / ~ 0.6872 / / Table 43. The specific binding of humanized bispecific antibodies to human or cynomolgus monkey TIGIT protein detected by ELISA Name Bind to human TIGIT protein Bind to cynomolgus monkey TIGIT protein EC50 %relative activity EC50 %relative activity (nM) TIGIT-005- H2L1d / tested antibody RG6058-hIgG1 / tested antibody (nM) TIGIT-005- H2L1d / tested antibody RG6058-hIgG1 / tested antibody LC-BsAb-006 0.03563 120.5% 141.4% 0.04126 108.0% 111.6% LC-BsAb-010 0.02761 155.5% 182.5% 0.03359 132.7% 137.1% TIGIT-005-H2L1d 0.04292 100.0% 117.4% 0.04456 100.0% 103.3% RG6058-hIgG1 0.05038 85.2% 100.0% 0.04605 96.8% 100.0% anti-Fluorescein-hIgG 1 ~ 0.000 / / ~ 0.6872 / / EXAMPLE 28 - The binding activity of anti-PVRIGxTIGIT humanized bispecific antibodies 5 to FlpinCHO human PVRIG and FlpinCHO cynomolgus monkey PVRIG detected by FACS. CHO-K1 stable cells (ATCC®CCL-61™) transfected with human or cynomolgus monkey PVRIG high expression plasmids were named FlpinCHO-hPVRIG and FlpinCHO-cynoPVRIG. Human PVRIG full-length plasmids (NCBI Ref Seq: NP076975) and cynomolgus monkey PVRIG full-length plasmid (NCBI Ref Seq: XP014989941) were synthesized by General Biol. 10 The experiment was performed when the cell density did not exceed 80%. After discarding the cell culture medium, the cells were rinsed with PBS and digested for 8-10 minutes by adding 1mL Versene (Gibico, 15040-066). The Ham's F12 (Gibico, 21127-022) complete medium containing 10% FBS was then added to terminate the digestion and acquire cell suspension. After cell counting, an appropriate amount of cell suspension was centrifuged at 350xg to remove the 15 supernatant, washed twice with PBS, stained with Zombie violet (Biolegend, 423114) and incubated at room temperature for 20 minutes. After incubation, the cells were added with staining buffer (2% FBS+PBS) to stop staining, centrifuged at 350xg to remove the supernatant, washed twice and resuspended with staining buffer to a density of 2x106 cells / mL. The cells were plated into a 96-well plate (50pL of cell suspension per well) for further use. The antibodies were diluted from the highest concentration of 46 nM (two-fold concentration) with staining buffer by 3.3-fold serial dilution. The diluted antibodies were added to the well containing 50 pL of cell suspension. The plate was placed on a microplate shaker at 400 rpm for 1 minute to fully mix the antibodies and cells, and then incubated at 4°C for 30 minutes. After the incubation, the cells were washed twice with staining buffer (200 pL per well) and centrifuged at 350zg for 5 minutes to discard the supernatant. The PE goat anti-human IgG Fc antibody (ebioscience, 12-4998-82) was diluted 250 times with staining buffer, added into the washed wells (100 pL per well), mixed well, and stained at 4°C for 30 minutes. After staining, the cells were washed twice with staining buffer, finally resuspended with 200 pL staining buffer and analyzed by flow cytometry (BD, Canto n). The stronger the signal, the stronger the ability of the antibodies binds to PVRIG. Fig. 45 shows that the four humanized bispecific antibodies have good binding activity to human PVRIG, as well as cynomolgus monkey PVRIG (Fig. 46), and all of them are superior to their corresponding anti-PVRIG humanized monoclonal antibodies. EXAMPLE 29 - The binding activity of anti-PVRIGxTIGIT humanized bispecific antibodies to CHO-K1 human TIGIT (high / medium / low expression cell strain) and CHO-K1 cynomolgus monkey TIGIT cells detected by FACS. The collected cells were washed once with PBS (Hyclone, Cat NO.SH30256), resuspended to 2x105 / 50 pL with 1% BSA-PBS. The antibodies were diluted to 80 nM with 1% BSA-PBS (diluted by 3-fold serial dilution for 12 concentration points). The 50 pL cells were then mixed with 50 pL diluted antibodies, and then incubated at 4OC for 60 minutes. The cells were washed twice with PBS, resuspended with Alexa Fluor® 647 fluorescein labeled secondary antibody (1:800) (Jackson, Cat NO.109-605-088), 100 pL / well, and incubated at 4°C for 40 minutes. After washing twice with PBS, the cell samples were resuspended with 1% BSA-PBS, 100 pL / well and then analyzed by flow cytometry (BD, Canto n). The experimental results show that the four humanized bispecific antibodies have good binding activity to CHO-K1 human TIGIT (high / medium / low expression cell strain, Fig. 47, 48 and 49) and CHO-K1 cynomolgus monkey TIGIT cells (Fig. 50). EXAMPLE 30 - The effect of anti-PVRIGxTIGIT humanized bispecific antibodies blocking the interaction between PVRIG protein and PVRL2 protein detected by HTRF PVRIG-mFc (ACRO Biosystems, Cat NO.PVG-H5253) and Bio-CD112-His (Sino Biological, Cat NO.10005-H08H) were diluted to 0.5 gg / mL, respectively. Streptavidin-Tb cryptate (Cisbio) and PAb anti mouse IgG-XL665 (Cisbio) were diluted to 20 gg mL and 0.8 gg / mL, respectively. The tested antibodies were diluted from the initial concentration of 120 nM, by 3-fold serial dilution for 12 concentration points. The diluted PVRIG-mFc, Bio-CD112-His, Streptavidin-Tb Crytate and PAb anti mouse IgG-XL665 were mixed at 1:1:1:1, and added into a 384-well plate (PE, Cat NO.6007299), 10 gL / well. 10gL / well of serially diluted antibodies was added into wells, centrifuged at 1500 rpm for 30s, and incubated at 37°C for 1 hour. The plate was read on the microplate reader (PE, Envision2105) with wavelengths of 665 nm and 620 nm. Data conversion is performed according to the formula: Ratio=Signal 665 nm / Signal 620 nm x 104. A Four Parameter Logistic Fit was performed with the molar concentration of the antibodies as the abscissa and Ratio as the ordinate to calculate IC50. The smaller the IC50, the better the effect of the antibody block the binding of PVRIG to PVRL2. The positive controls in this experiment are COM701-hIgG1, PVRIG-A50-H1b and PVRIG-A105-H1; and the negative control antibody is anti-Fluorescein-hIgG1 (in house). The effect of 4 humanized bispecific antibodies blocking the binding of PVRIG to PVRL2 is shown in Fig. 51 and Tables 44 and 45. The data shows that all four humanized bispecific antibodies can block the binding of PVRIG to PVRL2 protein. The blocking effect of LC-BsAb-002 is better than its corresponding monoclonal antibody (PVRIG-A50-H1b) and the positive control (COM701-hIgG1). The blocking effect of LC-BsAb-006 is better than its corresponding monoclonal antibody (PVRIG-A50-H1b), but slightly weaker than the positive control (COM701-hIgG1). The blocking effects of LC-BsAb-009 and LC-BsAb-010 are not as good as their corresponding monoclonal antibodies (PVRIG-A105-H1) and the positive control (COM701-hIgG1). Table 44. The effect of humanized bispecific antibodies blocking the binding of PVRIG to PVRL2 Name Block the binding of human PVRIG protein to human PVRL2 protein IC50 %relative activity (nM) PVRIG-A50-H1b / tested antibody COM701-hIgG1 / tested antibody LC-BsAb-002 0.1104 186.5% 124.4% LC-BsAb-006 0.1805 114.1% 76.1% PVRIG-A50-H1b 0.2059 100.0% 66.7% COM701-hIgG1 0.1373 150.0% 100.0% anti-Fluorescein-hIgG 1 ~ 0.03023 / / Table 45. The effect of humanized bispecific antibodies blocking the binding of PVRIG to PVRL2 Name Block the binding of human PVRIG protein to human PVRL2 protein IC50 %relative activity (nM) PVRIG-A105-H1 / tested antibody COM701-hIgG1 / tested antibody LC-BsAb-009 0.1956 65.6% 70.2% LC-BsAb-010 0.1452 88.4% 94.6% PVRIG-A105-H1 0.1283 100.0% 107.0% COM701-hIgG1 0.1373 93.4% 100.0% anti-Fluorescein-hIgG 1 ~ 0.03023 / / EXAMPLE 31 - The effect of anti-PVRIGxTIGIT humanized bispecific antibodies blocking CHO-K1 human CD112 cells binding to human PVRIG-mFc proteins detected by FACS CHO-K1 stable cells transfected with human CD112 high expression plasmid was named CHO-K1-CD112. Human CD112 full-length plasmid (NP 001036189.1 / NCBI Ref Seq: Q92692) was synthesized by General Biol. The experiment was performed when the cell density did not exceed 80%. After discarding the cell culture medium, the cells were rinsed with PBS and digested for 2 minutes by adding 1mL trypsin (Gibico, 25200-72). The Ham's F12 (Gibico, 21127-022) complete medium containing 10% FBS was then added to stop the digestion and acquire cell suspension. After counting with a cell counter (Beckman Coulter, Vi-CELL), an appropriate amount of cell suspension was centrifuged at 350*g to remove the supernatant, washed twice with PBS, stained with Zombie violet (Biolegend, 423114) and incubated at room temperature for 20 minutes. After incubation, the cells were added with staining buffer (2% FBS+PBS) to stop staining, centrifuged at 350*g to remove the supernatant, washed twice and resuspended with staining buffer to a density of 2*106 cells / mL. The cells were plated into a 96-well plate (50pL of cell suspension per well) for further use. A working solution of human PVRIG-mFc protein (Acro, PVG-H5253) prepared with staining buffer, 1pg / mL (four-fold concentration), was added into a 96-well plate (50pL PVRIG-mFc working solution per well). The antibodies were diluted from the highest concentration of 275 nM (four-fold concentration) with staining buffer by 3-fold serial dilution. The diluted antibodies were added to the well containing 50 pL of PVRIG-mFc. The plate was placed on a microplate shaker at 400 rpm for 1 minute to fully mix the antibodies and PVRIG-mFc proteins, and then incubated at 4°C for 30 minutes. After the incubation, the cell suspension prepared above (100pL) was added into the wells, gently mixed with the pipette tip and incubated at 4°C for 30 minutes. After the incubation, the cells were washed twice with staining buffer (200 pL per well) and centrifuged at 350xg for 5 minutes to discard the supernatant. The PE goat antimouse IgG Fc antibody (Biolegend, 405337) was diluted 250 times with staining buffer, added into the washed wells (100 pL per well), mixed well, and stained at 4°C for 30 minutes. After staining, the cells were washed twice with staining buffer, finally resuspended with 200 pL staining buffer and analyzed by flow cytometry (BD, Canto II). The weaker the fluorescence signal, the stronger the ability of the antibodies blocks the binding of CHO-K1 human CD112 cells to the PVRIG-mFc protein. As shown in Fig.52, all 4 humanized bispecific antibodies can block the binding of CHO-K1 human CD112 cells to human PVRIG-mFc protein. EXAMPLE 32 - The effect of anti-PVRIGxTIGIT humanized bispecific antibodies blocking the binding activity of human TIGIT to CHO-K1 CD155 detected by ELISA The collected CHO-K1 CD155 cells constructed in Example 2 were adjusted to a concentration of 5x105 / mL with 10% FBS-DMEM / F12 medium (Excell, FSP500; Gibco, 11330), added into a 96-well cell culture plate (corning, 3599), 100 pL / well, and cultured overnight at 37 °C with 5% CO2. After discarding the culture supernatant, the cells were fixed with a cell fixation solution (Beyotime, P0098), 50 pL / well, at room temperature for 1 hour. After washing once with 0.05% Tween 20-PBS on the Microplate Washer, the plate was added with 5% skim milk powder-PBS, 250 pL / well, and incubated at 37°C for 2-4 hours. The plate was then washed 3 times with 0.05% Tween 20-PBS on the Microplate Washer. The sample were mixed with human TIGIT ECD-mFc (working concentration: 100 ng / mL) and incubated for 0.5 hour. The mixed solution of antigen and antibody was added to the cell plate (50 pL / well). The cells were incubated with the mixture at 37OC for 1.5-2 hours and then washed with 0.05% Tween 20-PBS for 3 times on the Microplate Washer. The cell plate was then added with 50 pL / well HRP enzyme-labeled antibodies (Jackson, 115-035-003) diluted with 1% BSA (Sangon Biotech, A500023-0100)-PBS at a dilution ratio of 1:5000, and incubated at 37OC for 1 hour. After washing 3 times with 0.05% Tween 20-PBS on the Microplate Washer, the plate was added with TMB chromogenic solution (KPL, 5200-03), 50 pL / well, and incubated at 37OC for 10 minutes. 1M HCL, 50 pL / well, was then added into the plate to terminate the reaction, and the OD450 nm was read by a microplate reader (Biotek, Powerwave HT). As shown in Fig.53, all 4 humanized bispecific antibodies can block the binding between the human TIGIT and CHO-K1-CD155. EXAMPLE 33 - The effect of anti-PVRIGxTIGIT humanized bispecific antibodies blocking the binding activity of Bio-CD155-His to CHO-K1 human TIGIT detected by FACS The collected cells were washed once with PBS (Hyclone, SH30256), resuspended to 2x105 / 40 gL with 1% BSA-PBS. The antibodies were diluted to 210 nM with 1% BSA-PBS (diluted by 3-fold serial dilution for 12 concentration points). The Bio-CD155-His (Sino Biological Inc., 10109-H08H) was diluted to 3 gg / mL with 1% BSA-PBS. The 40 gL cells were then mixed with 40 gL diluted antibodies and 40 gL diluted Bio-CD155-His, and then incubated at 4°C for 60 minutes. The cells were washed twice with PBS, resuspended with APC labeled streptavidin (dilution ratio 1:1700, Biolegend, 405243), 100 gL / well, and incubated at 4°C for 40 minutes. The cell samples were washed twice with PBS, resuspended with 1% BSA-PBS (100 gL / well) and then analyzed by flow cytometry (BD, Canto n). As shown in Fig. 54, all 4 humanized bispecific antibodies can block the binding between the Bio-CD155-His and CHO-K1-human TIGIT. EXAMPLE 34 - The binding activity of anti-PVRIGxTIGIT humanized bispecific antibodies to human PBMCs detected by FACS The fresh human PBMCs (AllCells, PB004-C) were adjusted to a density of 5x105 / mL, added with SEA (Toxin Technology, Inc., AT101) to 100 ng / mL, and cultured at 37°C with 5% CO2 for 3 days. The cells were collected 3 days later, washed once with PBS (Hyclone, SH30256), added with Fc Block (BD, 564220) and then incubated at 4°C for 10 minutes. After washing twice with PBS, the PBMCs were resuspended to 2x105 / 50 gL with 1% BSA-PBS. The humanized antibodies were diluted to 80 nM with 1% BSA-PBS (diluted by 3-fold serial dilution for 12 concentration points). The 50 gL cells were then mixed with 50 gL diluted antibodies, and then incubated at 4OC for 60 minutes. The cells were washed twice with PBS, resuspended with Alexa Fluor® 647 fluorescein labeled secondary antibody (dilution ratio 1:800, Jackson, 109-605-088), 100 gL / well, and incubated at 4OC for 60 minutes. After washing twice with PBS, the cell samples were resuspended with 1% BSA-PBS, 100 gL / well, and then analyzed by flow cytometry (BD, Canto n). As shown in Fig. 55, all 4 humanized bispecific antibodies have good binding activity to human PBMCs. EXAMPLE 35 - The affinity of anti-PVRIGxTIGIT humanized bispecific antibodies for human, cynomolgus monkey and mouse TIGIT and PVRIG proteins detected by BIAcore Protein A chip was used in the BIAcore assay. The time required for the chip to capture the 5 diluted antibody and saturate the binding antigen to reach Rmax (Maximum binding capacity, 50 RU) was measured by manual run. The human, cynomolgus monkey and mouse TIGIT and PVRIG proteins were serially diluted to 20, 10, 5, 2.5, 1.25 nM. The affinity of antibody for antigen was measured by multi-cycle kinetics. In each cycle, the antibodies were injected prior to the injection of gradient concentraions of human, cynomolgus monkey and mouse TIGIT and PVRIG proteins 10 allowing the occurrence of antibody-antigen association and dissociation. After each cycle, the Protein A chip was regenerated by Glycine, pH 1.5 (to remove the proteins on the chip). The affinity KD of antibody for antigen was calculated by BIAcore T200 analysis software. As shown in Table. 46, two humanized bispecific antibodies specifically bind to human and cynomolgus TIGIT and PVRIG proteins with high affinity, but do not bind to mouse TIGIT and PVRIG proteins. 15    Table 46. The affinity of humanized bispecific antibodies for TIGIT and PVRIG proteins from various species Antibody Antigen Binding kinetics ka (1 / Ms) kd (1 / s) KD (M) Rmax Capture Level LC-BsAb-002 hTIGIT 1.75E+06 2.59E-04 1.48E-10 100.3 727 cynoTIGIT 4.92E+05 6.43E-03 1.31E-08 112.9 726 mTIGIT / / / 0.1 726 hPVRIG 3.09E+05 5.87E-05 1.90E-10 106.6 557 cynoPVRIG 3.57E+04 1.31E-04 3.66E-09 109.7 659 mPVRIG / / / 0.1 579 LC-BsAb-006 hTIGIT 1.65E+06 2.14E-04 1.30E-10 99.4 696 cynoTIGIT 1.09E+06 1.84E-03 1.69E-09 108.0 695 mTIGIT / / / 0.9 695 hPVRIG 3.40E+05 5.01E-05 1.48E-10 103.6 538 cynoPVRIG 3.60E+04 1.42E-04 3.95E-09 109.8 631 mPVRIG / / / 0.2 1140 EXAMPLE 36 - The co-binding of anti-PVRIGxTIGIT humanized bispecific antibodies to human TIGIT and PVRIG proteins detected by BIAcore The BIAcore was used to characterize the property of bispecific antibody simultaneously binding to two antigens. Firstly, antibodies, LC-BsAb-002 and LC-BsAb-006, were captured by Protein A chip. TIGIT and PVRIG proteins with his-tag were then injected, respectively, and followed by the continuous injection of TIGIT and PVIRIG, PVRIG and TIGIT. The binding responses of antibodies and antigens were recorded. Finally, the Protein A chip was regenerated by Glycine, pH 1.5. In the assay, the mobile phase was HBS-EP+ (10 mM HEPES, 150 mM NaCl, 3 mM EDTA, 0.05% surfactant P20), the flow rate was 30 gL / min, the binding time to different antigens was 300s, the regeneration time was 30s, the temperature during detection was 25°C, the analytical concentration of hTIGIT was 20 nM, and the analytical concentration of hPVRIG was 50 nM. The data was analyzed by BIAcore 8K analysis software (version 2.0). The capture level of the antibody and the binding responses (RU) of different antigens were recorded. The stoichiometric ratio of antigen and antibody molecules was calculated according to the molecular weight of the antigen and antibody to roughly estimate how many antigens can bind to an antibody molecule. In order to confirm the interaction between the antibody (LC-BsAb-002) and the antigen (TIGIT&PVRIG), the following four-step detection was carried out: only binding to single antigen, hTIGIT; only binding to single antigen, hPVRIG; first binding to hTIGIT and then binding to hPVRIG; first binding to hPVRIG and then binding to hTIGIT, and each antigen reached saturation state. The antibody-antigen binding curves, capture levels of the two bispecific antibodies, and the binding responses of TIGIT and PVRIG in each experiemnt were collected and recorded to calculate the stoichiometric ratio of antigen and antibody molecules. The antibody-antigen binding curves of LC-BsAb-002 and LC-BsAb-006 respectively binding to TIGIT and PVRIG, followed by the continuous injection of TIGIT and PVRIG are shown in Fig.56A and 56B (Table 47). As shown in Table 47, Fig. 56A and 56B, the binding responses generated by the continuous injection of TIGIT and PVRIG were almost same as those generated by injecting TIGIT and PVRIG alone. In addition, the sequence of continuous injection of TIGIT and PVRIG did not affect the binding responses thereof. It suggests that LC-BsAb-002 and LC-BsAb-006 can simultaneously bind to hTIGIT and hPVRIG without mutual influence between the two antigens. Based on the molecular weight of the antibodies and antigens, and the capture level of antibodies and the binding level of antigens, it is roughly estimated that the stoichiometric ratio of TIGIT to LC-BsAb-002 is 1.76; the stoichiometric ratio of TIGIT to LC-BsAb-006 is 1.86; the stoichiometric ratio of PVRIG to LC-BsAb-002 is 2.14, the stoichiometric ratio of PVRIG to LC-BsAb-006 is 2.18, and the stoichiometric ratio of the two antigens and antibodies are both close to 2. Considering the error caused by the detection assay, the inventors expect that an LC-BsAb-002 or an LC-BsAb-006 bispecific antibody molecule can simultaneously bind to two TIGIT molecules and two PVRIG molecules. Table 47. The binding of anti-PVRIGxTIGIT humanized bispecific antibodies to human TIGIT and PVRIG proteins detected by BIAcore Antibody Capture Level (RU) Antigen 01 Antigen 02 Name Binding Responses (RU) Antibody-Antigen Stoichiometry Name Binding Responses (RU) Antibody-Antigen Stoichiometry LC-BsAb-002 673.6 TIGT 94.96 1.76 / / / 673.1 PVRIG 128.67 2.14 / / / 673.3 TIGT 94.87 1.76 PVRIG 121.11 2.02 673.7 PVRIG 128.79 2.15 TIGT 91.28 1.69 LC-BsAb-006 603.7 TIGT 89.83 1.86 / / / 602.9 PVRIG 117.08 2.18 / / / 603.4 TIGT 89.83 1.86 PVRIG 109.42 2.04 602.3 PVRIG 116.96 2.18 TIGT 83.88 1.74 EXAMPLE 37 - The promoting effect of anti-PVRIGxTIGIT humanized bispecific antibodies on NK cell function detected by NK cell degranulation assay The effect of the tested antibodies on NK cell activation was indicated by the expression level of CD107a in NK cells detected by FACS (The experimental procedures were shown in Fig. 57A) A. The expression of PVRIG and TIGIT on NK cells (Natural killer cells), and PVR and PVRL2 on WIDR cells was detected by FACS. Firstly, the NK cells were counted by using a cell counter (Beckman Coulter, Vi-CELL). The NK cells were added into three flow cytometry tubes with 1e+5 NK cells in each tube and washed twice with PBS. After discarding the supernatant, one tube was added with 300 pL Staining buffer (PBS+2% FBS) as an unstained tube, and the other two tubes were added with 100 pL staining solution (PBS+1* Zombie Violet (Biolegend, 423114)), mixed well and then incubated at room temperature for 15 minutes. The cells were then washed twice with staining buffer, and the supernatant was discarded. 50 pL of Fc blocker (Staining buffer + Fcx blocker (Biolegend, 422302)) was added to each tube, mixed and then incubated at 4OC for 15 minutes. Each tube was then added with different staining solution, mixed and incubated at 4°C for 30 minutes, wherein, the first tube was added with 50 pL of 2* staining solution (Staining buffer + PE-Cy7 Mouse anti-hCD3 detection antibody + PE Mouse anti-hCD56 detection antibody + APC Mouse anti-hTIGIT detection antibody + AF488 Rabbit anti-hPVRIG detection antibody, CD3 detection antibody: Biolegend 300316, CD56 detection antibody: Biolegend 318306, TIGIT detection antibody: Biolegend 372706, PVRIG detection antibody: RD FAB93651G), and the second tube was added with 50pL of 2* isotype control staining solution (Staining buffer+PE-Cy7 Mouse anti-hCD3 detection antibody + PE Mouse anti-hCD56 detection antibody + APC Mouse IgG2a k isotype control antibody + AF488 Rabbit IgG k isotype control antibody, APC mIgG2a k isotype control antibody: Biolegend 400222, AF488 Rabbit IgG k isotype control antibody: RD IC1051G). The samples were then washed twice with Staining buffer, centrifuged and mixed with 300 pL Staining buffer. The samples were detected by a flow cytometer (Thermo Attune NxT) to obtain the percentage of the CD56-positive and CD3-negative cell population in the Zombie Violet negative cell population and read the signals of APC and AF488 of the CD56-positive and CD3-negative cell population in the Zombie Violet negative cell population. The WIDR cells were trypsinized to form a cell suspension, and counted by using a cell counter (Beckman Coulter, Vi-CELL). The WIDR cells were added into three flow cytometry tubes with 1e+5 cells in each tube and washed twice with PBS. After centrifugation and discarding the supernatant, one tube was added with 300 pL Staining buffer (PBS+2% FBS) as an unstained tube, and the other two tubes were added with 100 pL staining solution (PBS+1* Zombie Violet (Biolegend, 423114)), mixed well and then incubated at room temperature for 15 minutes. The cells were then washed twice with Staining buffer, and the supernatant was discarded. Each tube was then added with different staining solution, mixed and incubated at 4°C for 30 minutes, wherein, the first tube was added with 100 pL staining solution (Staining buffer + PerCP-Cy5.5 Mouse anti-hPVR detection antibody + APC Mouse anti-hPVRL2 detection antibody, PVR detection antibody: Biolegend 337612, PVRL2 detection antibody: Biolegend 337412), and the second tube was added with 100 pL isotype control staining solution (Staining buffer + PerCP-Cy5.5 Mouse IgG1 k isotype control antibody + APC k Mouse IgG1 isotype control antibody, PerCP-Cy5.5 mIgG1 k isotype control antibody: Biolegend 400150 , APC mIgG1 k isotype control antibody: Biolegend 400122). The samples were then washed twice with Staining buffer, centrifuged and mixed with 300 uL Staining buffer. The samples were detected by a flow cytometer (Thermo Attune NxT) to read the signals of PerCP-Cy5.5 and APC of the Zombie Violet negative cell population. Fig.57B shows that a certain level of PVRIG and TIGIT are expressed on NK cells used in the experiment, while ligands, PVR and PVRL2, are highly expressed on target cells, WIDR. B. NK cell degranulation assay (target cells, WIDR) The PBMCs were resuscitated one day in advance, and human NK cells were sorted by a sorting kit (Stemcell, 17955). The NK cells were added with 200 IU / mL h-IL2 (RD, 202-IL) and 10 ng / mL h-IL12 (Peprotech, 200-12-50UG) to stimulate overnight and plated for the following experiment on the next day. Firstly, the antibodies were diluted to the highest concentration of 275 nM (four-fold concentration) with assay buffer (RPMI1640-Glutamax + 10% FBS + 1XP S) and continually diluted with assay buffer by 10-fold serial dilution. The diluted antibodies were added to an ultra-low attachment, 96-well, U-bottom plate (Costar, 7007), 50 uL per well, for further use. Secondly, the NK cells were counted by using a cell counter (Beckman Coulter, Vi-CELL). A certain number of NK cells were centrifuged at a speed of 350 g for 5 minutes, resuspended to a density of 0.5E+6 cells / mL with assay buffer after discarding the supernatant and added with protein transport inhibitor (Invitrogen, 00498093) and APC mouse anti-human detection antibody (Biolegend, 328620). The antibody coated 96-well, U-bottom plate was added with the treated NK cell suspension (50uL per well), mixed well and incubated at room temperature for 15 minutes. During the incubation, the target cells (WIDR) were trypsinized to form a cell suspension (Reh cells were directly mixed well) and counted by using a cell counter (Beckman Coulter, Vi-CELL). An appropriate number of cells were centrifuged at 200 g for 5 minutes, and resuspended to a density of the 0.25E+6 cells / mL with the assay buffer after discarding the supernatant. After incubation, the target cell suspension was added to the plate, 100uL per well. At this time, each well contained 25,000 NK cells, 25,000 target cells and different concentrations of tested antibodies. The wells only containing NK cells were served as resting control, while the wells containing NK cells and target cells were served as drug-free control. All the wells were mixed well and incubated in an incubator at 37°C for 16 hours. Finally, FACS staining was performed: the cells in the plate were transferred to the same position in a 96-well, V-bottom plate and washed twice with PBS. After discarding the supernatant, each well was added with the staining solution (PBS+2%FBS+1*zombie violet (Biolegend, 423114) + PE mouse anti-CD56 detection antibody (Biolegend, 318306)), mixed well and incubated at 4°C for 30 minutes. Each well was then washed twice with staining buffer and resuspended with 150 pL staining buffer after discarding the supernatant. The samples were detected by a flow cytometer (Thermo Attune NxT) to read the percentage of CD107a strong positive cell population in CD56 positive cells. A higher percentage of CD107a strong positive cells represent a stronger degranulation of NK cells and a higher degree of activation of NK cells. Fig. 57C shows that the negative control, anti-Fluorescein-hIgG1, has no effect on CD107a on NK cells. The humanized bispecific antibody candidates can improve the expression of CD107a on NK cells to varying degrees, which indicates that the tested antibodies can effectively promote the activation of NK cells. C. NK cell degranulation assay (target cells, TF-1) The methods refer to Example 37 B. As shown in Fig. 57D, the effect of the humanized bispecific antibody on degranulation of NK cells was better than that of PVRIG positive control (COM701-hIgG4) and TIGIT positive control (RG6058-hIgG1), and equivalent to that of the combination of COM701-hIgG4 and RG6058-hIgG1. Meanwhile, the effect of the humanized bispecific antibody on degranulation of NK cells was better than that of PVRIG antibody (PVRIG-A50-H1b) and TIGIT antibody (TIGIT-002-H4L3), and comparable to that of the combination of PVRIG-A50-H1b and TIGIT-002-H4L3. EXAMPLE 38 - The cytotoxicity of NK cells for tumor cell line mediated by anti-PVRIGxTIGIT humanized bispecific antibodies detected by NK cell cytotoxicity assay The lysis level of target cells (WIDR) was detected by FACS to infer the effect of the tested antibodies on the cytotoxicity of NK cells for target cells. The PBMCs were resuscitated one day in advance. The NK cells were sorted by a sorting kit (Stemcell, 17955), added with 200 IU / mL h-IL2 (RD, 202-IL) and 10 ng / mL h-IL12 (Peprotech, 200-12-50UG) to stimulate overnight and plated for the following experiment on the next day. The expression level of PVRIG and TIGIT on three NK donors (Donor-050, Donor-831, and Donor-715) was detected by the method described in Example 13. Meanwhile, the antibodies were diluted to the highest concentration of 275 nM (four-fold concentration) with assay buffer (RPMI1640-Glutamax + 10% FBS + 1XP S) and continually diluted with assay buffer by 10-fold serial dilution. The diluted antibodies were added to an ultra-low attachment, 96-well, U-bottom plate (Costar, 7007), 50 gL per well, for further use. Secondly, the target cells (WIDR) were trypsinized to form a cell suspension and counted by using a cell counter (Beckman Coulter, Vi-CELL). An appropriate number of WIDR cells were centrifuged at a speed of 200 g for 5 minutes, resuspended with PBS after discarding the supernatant and added with staining solution, CellTrace Violet (Invitrogen, C34557A) with a final concentration of 5 gM. The WIDR suspension with staining solution was mixed evenly, placed in an incubator at 37°C for 10 minutes and shaken during the incubation. A portion of WIDR cells were used to determine the expression levels of PVR and PVRL2 on WIDR detected by the method described in Example 13. At the same time, NK cells were counted with a cell counter. A certain number of NK cells were centrifuged at a speed of 350 g for 5 minutes and resuspended to a density of 0.5e+6 cells / mL with the assay buffer after discarding the supernatant. The antibody coated 96-well, U-bottom plate was added with the treated NK cell suspension (50 gL per well), mixed well and incubated at room temperature for 15 minutes. After the WIDR cell staining, 5 times volume of complete medium (MEM+10%FBS+1*P / S+1*non-essential amino acid+1*sodium glutamate) was added to the cell suspension to stop the reaction. The WIDR cells were centrifuged at 200 g for 5 minutes and resuspended to a density of 0.25E+6 cells / mL with the assay buffer after discarding the supernatant. After incubating with antibodies and NK cells, the plate was added with WIDR cell suspension, 100 pL per well. At this time, each well contained 25,000 NK cells, 25,000 WIDR cells and various concentrations of tested antibodies. The wells only containing WIDR cells were served as resting control, while the wells containing NK cells and WIDR cells were served as drug-free control. All the wells were mixed well and incubated in an incubator at 37°C for 4 hours. Finally, FACS staining was performed: each well was added with staining solution (PBS + PI (Propidium Iodide, Invitrogen, P3566)), mixed well and incubated at room temperature for 20 minutes. The samples were then detected by a flow cytometer (Thermo Attune NxT) to read the percentage of PI positive cell population in CTV positive cells. More PI positive cells represent stronger NK cell cytotoxicity. Figure 58A shows that a certain level of PVRIG and TIGIT are expressed on all three NK donors (donor-050, donor-831 and donor-715); Figure 58B shows that PVR and PVRL2 are highly expressed on target cells, WIDR; Figure 58C shows brief experimental process of NK cell cytotoxicity assay for WIDR cells; Figure 58D shows the negative control, anti-Fluorescein-hIgG1, has no significant effect on the NK cell cytotoxicity, while two tested humanized bispecific antibodies can effectively promote the NK cell (from three donors) cytotoxicity for the target cells (WIDR). The EC50 and area under the curve (AUC) of NK cell cytotoxicity (from different NK donors) for WIDR cells mediated by two humanized bispecific antibodies are shown in the Table. Use TF-1 as a target cell, NK cells cytotoxicity for TF-1 mediated by the anti-PVRIGxTIGIT humanized bispecific antibody was detected. As shown in Figure 58E, the humanized bispecific antibody can promote NK cells to kill tumor cells, with the activity better than that of PVRIG positive control (COM701-hIgG4) and TIGIT positive control (RG6058-hIgG1) and equivalent to the combination of COM701-hIgG4 and RG6058-hIgG1. Meanwhile, the promoting activity of the humanized bispecific antibody is better than that of PVRIG antibody (PVRIG-A50-H1b) and TIGIT antibody (TIGIT-002-H4L3), and comparable to that of the combination of the PVRIG-A50-H1b and TIGIT-002-H4L3. EXAMPLE 39 - The cytotoxicity of NK cells for human Treg cells mediated by anti-PVRIGxTIGIT humanized bispecific antibodies detected by ADCC assay The lysis level of target cells (Treg) was detected by FACS to infer the direct effect of the tested antibodies on antibody-dependent cell-mediated cytotoxicity (ADCC) of NK cells against target cells (Fig.59A). The PBMCs were resuscitated one day in advance. The NK cells were sorted out as effector cells using a sorting kit (Stemcell, 17955), added with 200 IU / mL h-IL2 (RD, 202-IL) and 10 ng / mL h-IL12 (Peprotech, 200-12-50UG) to stimulate overnight and plated for the following experiment on the next day. Tregs (regulatory T cells) isolated from PBMCs were used as target cells (Stemcell, 18063). Tregs were expanded in vitro 12 days by Dynabeads (Gibco, 11129D) for further use. The expression level of PVRIG and TIGIT on Treg cells was detected by the method and reagents described in example 13. The effector cells and target cells were co-incubated at a ratio of 5:1. The serially diluted humanized bispecific antibodies or isotype control anti-Fluorescein-hIgG1, anti-Fluorescein-hIgG4 antibodies are added into the cells and incubated in a 37°C incubator for 4 hours before adding PI staining. The percentage of PI-positive Treg cells was read to evaluate the ADCC effect of the bispecific antibodies on Treg cells. The results show that TIGIT and PVRIG are highly expressed on isolated and expanded human Treg cells (Fig.59B). The tested humanized bispecific antibodies only show ADCC activity against Treg cells as anti human IgG1 Fc antibodies, but the corresponding anti-human IgG4 Fc antibodies does not show obvious ADCC activity against Treg cells, which is equivalent to the two negative control antibodies, anti-Fluorescein-hIgG1 and anti-Fluorescein-hIgG4 (Fig.59C). The EC50 and AUC of the ADCC activity of two tested humanized bispecific antibodies (anti-human IgG1 Fc antibodies) against Treg cells are listed in the table. EXAMPLE 40 ADCP activity of anti-PVRIGxTIGIT humanized bispecific antibodies Monocytes were isolated from donor PBMC and induced with 75 ng / mL GM-CSF for seven days to differentiate to macrophages, which were labeled with CellTrace Violet and used as effector cells. Human Treg cells were sorted by Treg Cells Isolation Kit from human PBMCs. Tregs were expanded and activated in vitro for 13 days by Dynabeads Human Treg Expander and used as target cells. The target cells were stained by CFSE. The effector cells and the target cells were co-incubated at a ratio of 4:1. Tested antibodies with serial dilution, negative control antibodies Hel hIgG1, PVRIG antibody (PVRIG-A50-H1b), TIGIT antibody (TIGIT-002-H4L3) and combination of the two monoantibody were added, incubated at 37°C for 4 hours. After the incubation, cell stain PI was added, and the percentage of CellTrace Violet-positive cells in CFSE-positive Treg cells was detected by flow cytometry to evaluate the ADCP effect of the tested antibodies. As shown in Fig. 60A, the tested antibodies activate the ADCP action of Treg cells in a dosedependent manner. PVRIG-A50-H1b or COM701-IgG4 almost has no ADCP activity, TIGIT-002-H4L3 or RG6058-hIgG1 exhibits dose-dependent ADCP activity. Based on the under the curve (AUC) of the ADCP curve, the ADCP activity of tested antibodies was slightly weaker than that of TIGIT-002-H4L3 and the combination of PVRIG-A50-H1b and TIGIT-002-H4L3. The ADCP effect of the tested antibodies are comparable to that of the RG6058-hIgG1 and combination of two positive control antibodies (COM701-hIgG4 + RG6058-hIgG1). According to the Emax of the ADCP curve, the ADCP activity of tested antibodies are comparable to the combination of PVRIG-A50-H1b and TIGIT-002-H4L3, and combination of two positive control antibodies (COM701-hIgG4 + RG6058-hIgG1). (Fig. 60B) EXAMPLE 41 The effect of anti-PVRIGxTIGIT humanized bispecific antibodies on cytokine release in human PBMC from healthy donors The present experiment investigates the effect of the tested antibodies on cytokine secretion of unstimulated PBMCs from healthy individuals. PBMCs from three healthy volunteers were incubated with tested antibodies for 24 hours under liquid or solid phase conditions, and then secretion levels of five cytokines, IFN-y, IL-2, IL-6, IL-10 and TNF-a in PBMCs supernatants were detected by flow cytometry. Lipopolysaccharide and CD3 monoclonal antibody were used as positive control, Anti-Hel hIgG1 was used as negative control, RG6058-hlgG1 and COM701-hlgG4 were used as monoclonal antibody control of TIGIT and PVRIG respectively. The results show that the positive control CD3 monoclonal antibodies in liquid or solid phase conditions, and LPS in liquid phase conditions, incubated for 24 hours with unstimulated PBMCs from three healthy volunteers, can increase the secretion levels of 5 cytokines, IFN-Y, IL-2, IL-6, IL-10, and TNF-a to varying degrees. In liquid phase conditions, secretion levels of IFN-y, IL-2, IL-6, IL-10, and TNF-a in PBMC were comparable to that of the negative control or below the detection limit after incubation of different concentrations of the tested antibodies with unstimulated PBMC for 24 hours. In solid phase conditions, the secretion of IFN-Y, IL-2 and IL-10 in PBMC were comparable to that of the negative control or below the detection limit after incubation of different concentrations of the tested antibodies with unstimulated PBMC for 24 hours. In solid phase conditions, the secretion of TNF-a and IL-6 were significantly higher than that of the negative control or comparable to that of the negative control after the action of tested antibodies at a high concentration point (2850 nM); but the tested antibodies dose not additionally increase the secretion of 5 cytokines, IFN-Y, IL-2, IL-6, IL-10 and TNF-a in unstimulated PBMC from healthy individuals compared with RG6058-hlgG1 and COM701-hlgG4 under the same conditions. In summary, compared with the TIGIT monoclonal antibody RG6058-hIgG1 and the PVRIG antibody COM701-hIgG4, the tested antibodies, under the same conditions, do not additionally increase the secretion of 5 cytokines, IFN-Y, IL-2, IL-6, IL-10, and TNF-a in unstimulated PBMCs from healthy subjects. EXAMPLE 42 - The effect of anti-PVRIGxTIGIT humanized bispecific antibodies on improving the functions of antigen-specific CD8 T cells detected by CMV antigen-recall assay Principle of this assay: The PBMCs from CMV IgG positive donor were induced by CMV pp65 (495-503) polypeptide to produce CMV pp65 specific CD8 T cells, which were served as effector cells. The colo205 tumor cell line pulsed with pp65 was used as the target cell. In such experimental system, the effect of anti-PVRIG&TIGIT humanized bispecific antibodies on improving the functions of the pp65-specific CD8 T cell was detected (Fig. 61A). The PBMCs were resuscitated, resuspended to 2x106 / mL by using complete medium (RPMI1640-Glutamax + 5% AB serum + 1% P / S+(1x) 2-p mercaptoethanol) containing 1 pg mL CMV pp65(495-503) peptide (Anaspec, Cat No.AS-28328), 2 ng / mL human IL-2 (R&D, Cat No.IL-202) and 10 ng / mL human IL-7 (Peprotech, Cat No.200-07), inoculated in a 6-well plate (5 mL / well) and cultured at 37°C with 5% CO2 for 6 days. On day 6, all the PBMCs were collected, and pp65 and IL-7 in the medium were removed. The cells were divided into two portions, resuspended in complete medium containing 100 IU / mL human IL-2, and cultured for another 2 days. On day 8, all the PBMCs were collected and resuspended in complete medium containing 100 lU / mL human IL-2, and the cell density was adjusted to 2x106 / mL for continuing cultivation. On day 11, all the PBMCs were collected. The expression of PVRIG, TIGIT and PD-1 on CMV pp65 (495-503) specific CD8 T cells (Fig. 61B) were detected by flow cytometry. The detection antibodies were as follows: Livedead Near IR (Invitrogen, Cat No.L34976), CD8-PerCp Cy5.5 (BD, Cat No.565310), CD3-PE-Cy7 (Biolegend, Cat No.300316), T-select HLA-A*0201 CMV pp65 Tetramer-PE (MBL, Cat No.TS-0010-1C), PVRIG-AF488 (R&D, Cat No.FAB93651G-100UG), TIGIT-APC (Biolegend, Cat No.372706) and PD-1-BV421 (BD, Cat No.562516). CD8 T cells were isolated from the induced PBMCs as effector cells by CD8 T cell isolation kit (Stemcell, Cat No.17953) and resuspended with AIM-V to the cell density of 0.4x106 / mL (If CD8 is cryopreserved after induction, the number of cells needs to be increased to 0.7x106 / mL.). The expression of CD226 and purity of isolated CD8 were detected. The target cells, Colo205, were digested by TrypLETM Express Enzyme (Gibco, Cat No.12605010), resuspended in AIM-V (Gibco, Cat No.31035-025) containing 20 ng / mL pp65 to the cell density of 1x106 / mL and treated at 37OC with 5% CO2 for 3 hours. The target cells were then centrifuged at 250 g for 5 minutes and resuspended in AIM-V to the cell density of 0.5x106 / mL after discarding the supernatant. The expression of PVRL2, PVR and PD-L1 on Colo205 were detected by flow cytometry (Fig. 61B). Tested antibodies (humanized bispecific antibodies and Tecentriq) were diluted with AIM-V to 280 nM. The low-attachment, 96-well, U-bottom plate (Corning, Cat No.7007) was added with 50 pL of antibody, 50 pL of CD8 T and 100 pL of pp65-treated colo205 in order, mixed well with a multichannel pipette, and incubated at 37°C with 5% CO2 for 18 hours. In the experimental system, the final concentration of antibodies was 70 nM, CD8 T cells were 20,000 / well, and colo205 were 50,000 / well. After incubation, the supernatant was collected by centrifugation at 400 g, and the level of human IFN-y in the supernatant was detected with an ELISA kit (Dakewe, Cat No.1110003). In this system, the positive controls were COM701-hIgG4 and RG6058-hIgG1, and the negative control was no treatment. The detection antibodies used in flow cytometry to detect the purity of isolated CD8 T were as follows: Livedead-BV421 (Invitrogen, Cat No.L34964) and CD8-FITC (BD, Cat No.555366). As shown in Fig. 61C and Table 48, when adding the bispecific antibodies (LC-BsAb-002 or LC-BsAb-006), there is no statistically significant difference in the secretion level of IFN-y between the two tested group at each concentration points. The secretion level of IFN-y is significantly higher with bispecific antibodies at high concentration point(s) (LC-BsAb-002 at 70nM, LC-BsAb-006 at 70 and 7nM) compared with the corresponding monoclonal antibody combination under the same conditions. There is no significant difference between the monoclonal antibody combination corresponding to the two candidate molecules and the antibody combination of positive controls (RG605-hIgG1 and COM701-hIgG4) at each concentration point. Over all, compared with the antibody combination of positive controls (RG605-hIgG1 and COM701-hIgG4), the two bispecific antibodies have a better effect on promoting IFN-y released from CD8 T cells. As shown in Fig. 61D, when using the combination of bispecific antibodies (LC-BsAb-002 or LC-BsAb-006) and Tencentriq, the secretion level of human IFN-y is significantly increased compared with using bispecific antibodies themselves (t-test, **P<0.01). When using the combination of PVRIG monoclonal antibody, TIGIT monoclonal antibody and PD-L1 monoclonal antibody, the secretion level of human IFN-y is significantly increased compared with using the combination of PVRIG monoclonal antibody and TIGIT monoclonal antibody (t-test, *P<0.05). The percentage of the histogram is the percentage increase in IFN-y compared with the anti-TIGIT positive control antibody, RG6058-hIgG1. Table 48. Statistical data analysis of the detection results of human IFN-y in the cell supernatant (two-way ANOVA) Two-way ANOVA analysis-01 Conc.(nM) LC-BsAb-002 LC-BsAb-006 two-way ANOVA 0.0007 137.4 111.7 128.3 131.5 ns 0.007 197.9 165 162.5 176.2 ns 0.07 554.3 411.2 371.1 338.2 ns 0.7 628.3 642.7 672 539.3 ns 7 784.3 691.5 820 727.9 ns 70 1119.1 881.8 733.6 798.2 ns Two-way ANOVA analysis-02 Conc.(nM) LC-BsAb-002 TIGIT-002-H4L3+PVRIG-A50-H1b two-way ANOVA 0.0007 137.4 111.7 148.2 121 ns 0.007 197.9 165 132.2 146.5 ns 0.07 554.3 411.2 311 280.8 ns 0.7 628.3 642.7 444.6 455.5 ns 7 784.3 691.5 632.7 540.4 ns 70 1119.1 881.8 648.5 440.4 ** Two-way ANOVA analysis-03 Conc.(nM) LC-BsAb-006 TIGIT-005-H2L1 d+PVRIG-A50-H1b two-way ANOVA 0.0007 128.3 131.5 120.3 121.6 ns 0.007 162.5 176.2 151.5 146.4 ns 0.07 371.1 338.2 229.9 276.2 ns 0.7 672 539.3 427.7 545.9 ns 7 820 727.9 571.5 388.2 ** 70 733.6 798.2 441.9 415.2 *** Two-way ANOVA analysis-04 Conc.(nM) TIGIT-002-H4L3+PVRIG-A50-H1b RG6058+COM701 two-way ANOVA 0.0007 148.2 121 173.6 148.3 ns 0.007 132.2 146.5 189.9 113.3 ns 0.07 311 280.8 199.4 172.6 ns 0.7 444.6 455.5 396.7 529.4 ns 7 632.7 540.4 593.6 416.1 ns 70 648.5 440.4 630.3 436.7 ns Two-way ANOVA analysis-05 Conc.(nM) TIGIT-005-H2L1 d+PVRIG-A50-H1b RG6058+COM701 two-way ANOVA 0.0007 120.3 121.6 173.6 148.3 ns 0.007 151.5 146.4 189.9 113.3 ns 0.07 229.9 276.2 199.4 172.6 ns 0.7 427.7 545.9 396.7 529.4 ns 7 571.5 388.2 593.6 416.1 ns 70 441.9 415.2 630.3 436.7 ns EXAMPLE 43 - The effect of combination use of anti-PVRIGxTIGIT humanized bispecific antibodies and Tecentriq on improving the functions of antigen-specific CD8 T cells detected by CMV antigen-recall assay Principle of this assay: Same as Example 40 (Fig. 61A) Induction of antigen-specific CD8 T cells: same as Example 40. Expression of PVRIG, TIGIT and PD-1 on inoculated cells was detected by flow cytometry on the day of inoculation (Fig. 62A). CD8 T cells were isolated from the induced PBMCs as effector cells by CD8 T cell isolation kit (Stemcell, Cat No.17953) and resuspended with AIM-V to the cell density of 0.8*106 / mL (the number of CD8 T cells in microplates was adjused according to the ratio of antigen-specific CD8 T cells). Colo205 cells pretreated overnight with 100ng / mL IFN-y were added to the complete medium as target cells, and digested by TrypLETM Express Enzyme (Gibco, Cat No.12605010), resuspended in AIM-V (Gibco, Cat No.31035-025) containing 20 ng / mL pp65 to the cell density of 1^106 / mL after washed twice and treated at 37°C with 5% CO2 for 3 hours. The target cells were then centrifuged at 250 g for 5 minutes and resuspended in AIM-V to the cell density of 0.5^106 / mL after discarding the supernatant. The expression of PVRL2, PVR and PD-L1 on Colo205 were detected by flow cytometry (Fig. 62A). Tested antibodies (humanized bispecific antibodies, Tecentriq, combination of bispecific antibodies and Tecentriq, combination of two positive control monoclonal antibody (COM701-hIgG4 and RG6058-hIgG1), combination of three monoclonal antibody (COM701-hIgG4, RG6058-hIgG1 and Tecentriq)) were diluted with AIM-V (initial concentration 280 nM(4X), 10-fold serial dilution, 6 concentration points). The low-attachment, 96-well, U-bottom plate (Corning, Cat No.7007) was added with 50 pL of antibody, 50 pL of CD8 T and 100 pL of pp65-treated colo205 in order, mixed well with a multichannel pipette, and incubated at 37°C with 5% CO2 for 18 hours. In the experimental system, the final concentration of antibodies was 70 nM, 7nM, 0.7nM, 0.07 nM, 0.007 nM and 0.0007 nM, respectively. CD8 T cells were 40,000 / well, and colo205 were 50,000 / well. After incubation, the supernatant was collected by centrifugation at 400 g, and the level of human IFN-y in the supernatant was detected with an ELISA kit (Dakewe, Cat No.1110003). As shown in Fig. 62B and Table 49, sort according to the AUC of the IFN-y fitting curve: LC-BsAb-002+Tecentriq>RG6058-hIgG1+COM701-hIgG4+Tecentriq>LC-BsAb-002>RG6058-hIgG1+COM701-hIgG4 > Tecentriq. Higher AUC indicates greater effectiveness, when using the combination of bispecific antibodies (LC-BsAb-002) and Tencentriq, the secretion level of human IFN-y is significantly increased compared with using LC-BsAb-002 itselves. When using the combination of COM701-hIgG4, RG6058-hIgG1 and Tecentriq, the secretion level of human IFN-y is significantly increased compared with using the combination of COM701-hIgG4 and RG6058-hIgG1. Table 49. Curve fit of human IFN-y assay results in cell supernatants Tecentriq LC-BsAb-002 RG6058+COM701 LC-BsAb-002 +Tecentriq RG6058+COM701 +Tecentriq EC50(nM) * - 0.1494 1.571 0.1101 0.9041 AUC 97.17 320.4 176.4 728.4 447.9 *Data curves cannot be fitted EXAMPLE 44 - In vivo drug efficacy evaluation of anti-PVRIGxTIGIT humanized bispecific antibody in mice A375 cells were inoculated subcutaneously on the right side of NPG mice, female, 5-6 weeks old (strain: NPG; Beijing Vitalstar Biotechnology Co., Ltd.) at a concentration of 5*106 cells / 0.1mL. One day after the inoculation of A375 cells, Hu PBMC cells were injected into mice at a concentration of 5*106 cells / 0.2mL by tail vein. When the tumor grew to about 82 mm3, 56 mice were selected according to the tumor volume and randomly divided into 7 groups with 8 mice in each group: Vehicle (PBS), RG6058-hIgG1 (10 mg / kg), COM701-hIgG4 (10 mg / kg), RG6058-hIgG1+ COM701-hIgG4 (10 mg / kg+10 mg / kg), Tecentriq (5 mg / kg; lot NO. HK65567, Roche), LC-BsAb-002 (11.7mg / kg), LC-BsAb-006 (11.7mg / kg). All groups were administrated by intraperitoneal injection, twice a week for 4 consecutive times. The experiment was ended 3 days after the last administration. The body weight and tumor volume of mice were measured three times a week during administration and observation. The measured values were recorded, and the tumor volume (long diameter * short diameter 2 / 2) and growth inhibition rate (TGITV (%) = (1-(Tn) -T0) / (Vn-V0)) * 100% were calculated. Efficacy results: As shown in Fig. 63, the candidate molecules LC-BsAb-002 and LC-BsAb-006 have a significant inhibitory effect on A375 tumor growth after administration. The inhibitory level of LC-BsAb-002 and LC-BsAb-006 is the same as that of the positive molecule (RG6058-hIgG1+COM701-hIgG4) and Tecentriq. On the day 13 of administration, the tumor inhibition rate (TGI) and difference analysis of each administration group and the negative control group (PBS) were obtained and shown in Table 50. The TGI of LC-BsAb-002 and LC-BsAb-006 were 82.16% and 78.59%, respectively, which are significantly higher than PBS (P<0.005). The level of TGI is better than RG6058-hIgG1 (TGI=42.55), COM701-hIgG4 (TGI=0.23%) alone, and equivalent to the combination of RG6058-hIgG1 and COM701-hIgG4 (TGI=83.32%). The tumor growth curve of single mouse shown in Fig. 64 shows the same trend as Fig. 63. Table 50. Effect of the test substances on tumor volume of HuPBMC-NPG mice transplanted with A375 cells. Grou p Test substance Tumor volume(mm3)a P b Before administration Day13 TGI (%) G1 PBS 83±4 196±41 - - G2 RG6058-hIgG1 81±4 146±20 42.55 0.2614 G3 COM701-hIgG4 82±4 195±64 0.23 0.6052 G4 RG6058-hIgG1+ COM701-hIgG4 82±4 101±17 83.32 ***0.0004 G5 Tecentriq 83±5 91±20 92.48 ***0.0002 G6 LC-BsAb-002 82±5 103±21 82.16 **0.0035 G7 LC-BsAb-006 82±3 106±36 78.59 ***0.0001 Note: a: mean ± standard error; b: the tumor volume of the administration group and the vehicle control group were statistically compared on the day 13 after administration by two-way ANOVA analysis, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. Results of body weight: as shown in Fig. 65 and table 51, except for the control molecule, Tecentriq, which causes significant weight loss and shows toxic side effects, the body weight change trends of the other control and candidate molecules (LC-BsAb-002 and LC-BsAb-006) are basically the same as that of PBS. The subsequent body weight loss is the GVHD phenomenon caused by PBMC reconstitution. Table 51 Effects of test substances on the body weight of HuPBMC-NPG mice transplanted with A375 cells Test substance              Body weight (g) a Grou p Before administration Day 13 P b Body weight change on day 13 (%) G1 PBS 22.5±0.3 23.3±0.8 - +3.8 G2 RG6058-hIgG1 22.4±0.4 21.4±1.1 0.8251 -4.6 G3 COM701-hIgG4 22.7±0.5 23.8±0.4 0.9999 +5.4 G4 RG6058-hIgG1+ COM701-hIgG4 22.6±0.6 22.4±1.4 0.9798 -1.5 G5 Tecentriq 22.7±0.8 18.7±1.0 **0.006 1 -18.9 G6 LC-BsAb-002 22.2±0.5 20.7±0.7 0.0744 -6.4 G7 LC-BsAb-006 23.0±0.6 23.1±1.1 0.9999 +1.7 Note: a: mean ± standard error; b: the body weight of the administration group and the vehicle control group were statistically compared on the day 13 after administration by two-way ANOVA analysis. The final results shows that the anti-PVRIGxTIGIT humanized bispecific antibody molecules (LC-BsAb-002 and LC-BsAb-006) have a significant inhibitory effect on the growth of subcutaneous xenograft tumors of A375, which is better than the positive drugs RG6058-hIgG1 and COM701-hIgG4 alone, and equivalent to the positive control antibodies (Tecentriq) and the combination of RG6058-hIgG1 and COM701-hIgG4. At the same time, no toxic and side effects of the candidate molecules are observed during the administration, which indicates that the safety and tolerability of candidate molecules under this model. EXAMPLE 45 - In vivo drug efficacy evaluation of combination use of anti-PVRIGxTIGIT humanized bispecific antibody and Tecentriq in mice A375 cells were inoculated subcutaneously on the right side of NPG mice, female, 5-6 weeks old (strain: NPG; Beijing Vitalstar Biotechnology Co., Ltd.) at a concentration of 5*106 cells / 0.1mL. One day after the inoculation of A375 cells, Hu PBMC cells were injected into mice at a concentration of 5.5*106 cells / 0.2mL by tail vein. When the tumor grew to about 82.78 mm3, 45 mice were selected according to the tumor volume and randomly divided into 5 groups: Vehicle (PBS, 9 mice), LC-BsAb-002 (11.7 mg / kg, 9 mice), LC-BsAb-002 (5.9 mg / kg, 9 mice), Tecentriq (3 mg / kg, 10 mice; lotNO. HK65567, Roche), LC-BsAb-002 + Tecentriq (5.9 mg / kg + 3mg / kg, 8 mice). All groups were administrated by intraperitoneal injection, twice a week for 5 consecutive times. The experiment was ended 3 days after the last administration. The body weight and tumor volume of mice were measured three times a week during administration and observation. The measured values were recorded, and the tumor volume (long diameter x short diameter 2 / 2) and growth inhibition rate (TGItv (%) = (1-(Tn) -T0) / (Vn-V0)) x 100% were calculated. Efficacy results: As shown in Fig. 66, the candidate molecules LC-BsAb-002 have a significant inhibitory effect on A375 tumor growth after administration, and the higher the dose administered, the greater the inhibitory effect on A375 tumor growth. The inhibitory level of combination use of LC-BsAb-002 and Tecentriq is significantly better than LC-BsAb-002 and Tecentriq alone. On the day 17 of administration, the tumor inhibition rate (TGI) and difference analysis of each administration group and the negative control group (PBS) were obtained and shown in Table 52. The TGI of LC-BsAb-002 (11.7 mg / kg), LC-BsAb-002 (5.9 mg / kg) and Tecentriq (3 mpk) were 66.56%, 60.51% and 41.53%, respectively, which are significantly different compared to the PBS (P<0.0001, P=0.0003 and P=0.0015). The TGI of combination use of LC-BsAb-002 and Tecentriq (80.44%) is significantly different compared to the PBS (P<0.0001), and better than that of LC-BsAb-002 (5.9 mg / kg) and Tecentriq(3 mpk) alone. The tumor growth curve of single mouse is shown in Fig. 67, and the tumor growth trends in each group were the same as Fig. 66. Table 52. Effect of the test substances on tumor volume of HuPBMC-NPG mice transplanted with A375 cells. Group Test substance Tumor volume (mm3) a P b Before administration Day17 TGI (%) G1 PBS 83±3 718±116 - - G2 LC-BsAb-002 11.7 mpk 84±3 296±76 66.56 ****<0.0001 G3 LC-BsAb-002 5.9 mpk 83±3 334±78 60.51 ***0.0003 G4 Tecentriq 3mpk 83±3 455±134 41.53 **0.0015 G5 LC-BsAb-002 5.9 mpk +Tecentriq 3 mpk 81±4 206±54 80.44 ****<0.0001 Note: a: mean ± standard error; b: the tumor volume of the administration group and the vehicle control group were statistically compared on the day 17 after administration by two-way ANOVA analysis, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. Results of body weight: as shown in Fig. 68 and Table 53, combination use of LC-BsAb-002 and Tecentriq showed a certain decrease in weight, but did not exhibit significant toxic side effects. The body weight change trends of the Tecentrip group, LC-BsAb-002 (11.7 mpk) group and LC-BsAb-002 (5.9 mpk) group are basically the same as that of PBS. Table 53 Effects of test substances on the body weight of HuPBMC-NPG mice transplanted 5 with A375 cells Body weight (g) a Body weight Group Test substance Before P b change on day 13 administration Day 17 (%) G1 PBS 23.1±0.4 22.4±0.6 - +6.1 G2 LC-BsAb-002 11.7 mpk 23.7±0.7 23.4±1.2 0.9598 -1.8 G3 LC-BsAb-002 5.9 mpk 22.4±0.6 22.5±0.6 0.2524 +1.0 G4 Tecentriq 3mpk 23.2±0.6 22.7±1.0 0.1582 -2.3 G5 LC-BsAb-002 5.9 mpk +Tecentriq 3 mpk 22.4±0.5 20.2±1.0 0.0002 -10.0 Note: a: mean ± standard error; b: the body weight of the administration group and the vehicle control group were statistically compared on the day 17 after administration by two-way ANOVA analysis. The final results shows that the anti-PVRIGxTIGIT humanized bispecific antibody molecules 10   (LC-BsAb-002) have a significant inhibitory effect on the growth of subcutaneous xenograft tumors of A375, and the inhibitory effect showed a significant dose-dependence as the administered dose increased. Combination of LC-BsAb-002 and Tecentriq is better than the respective drugs alone, shows significant combination effect. At the same time, no toxic and side effects of the candidate molecules are observed during the administration, which indicates that the 15 safety and tolerability of candidate molecules under this model.

Claims

1. An anti-PVRIG / anti-TIGIT bispecific antibody, which comprises:(a) a first antigen binding fragment, which is an antibody or antigen-binding fragment specifically binding to TIGIT comprising two heavy chains and two light chains, and the HCDR1 of the VH comprises a sequence shown in SEQ ID NO: 21; the HCDR2 of the VH comprises a sequence shown in SEQ ID NO: 22; the HCDR3 of the VH comprises a sequence shown in SEQ ID NO: 23; the LCDR1 of the VL comprises a sequence shown in SEQ ID NO: 18; the LCDR2 of the VL comprises a sequence shown in SEQ ID NO: 19; the LCDR3 of the VL comprises a sequence shown in SEQ ID NO: 20;(b) a second antigen binding fragment, which comprise VHH that specifically binds to PVRIG, and the CDR1 of the VHH comprises a sequence shown in SEQ ID NO: 168; the CDR2 of the VHH comprises a sequence shown in SEQ ID NO: 207; the CDR3 of the VHH comprises a sequence shown in SEQ ID NO: 208.

2. The bispecific antibody according to claim 1, wherein, the VH of the first antigen binding fragment comprises an amino acid sequence shown in SEQ ID NO: 72; the VL of the first antigen binding fragment comprises an amino acid sequence shown in SEQ ID NO: 68.

3. The bispecific antibody according to claims 1 or 2, wherein, the second antigen binding fragment comprises an amino acid sequence shown in SEQ ID NO: 200.

4. The bispecific antibody according to any one of claims 1 to 3, wherein, the first antigen binding fragment is a full-length antibody, comprising two heavy chains and two light chains; the C-terminus of the second antigen binding fragment is fused to the N-terminus of at least one heavy chain of the first antigen binding fragment.

5. The bispecific antibody according to claim 4, wherein, the heavy chain fusion polypeptide comprises PVRIG VHH - (G4S)4 Linker - TIGIT VH - CH1 - hinge - CH2-CH3 from the N-terminus to the C-terminus, and the light chain polypeptide comprises TIGIT VL-CL from the N-terminus to the C-terminus.

6. The bispecific antibody according to claim 5, wherein, the heavy chain fusion polypeptide comprises an amino acid sequence shown in SEQ ID NO: 227, and the light chain polypeptide comprises an amino acid sequence shown in SEQ ID NO: 226.2022320667   27 Aug 20267. The bispecific antibody according to any one of claims 1 to 6, which is a humanized antibody.

8. The bispecific antibody according to any one of claims 1 to 7, which specifically binds to PVRIG or TIGIT protein of human and / or monkey; and the KD value between the bispecific antibody and TIGIT protein of human and / or monkey is better than 1.00E-7M, and the KD value between the bispecific antibody and PVRIG protein of human and / or monkey is better than 1.00E-8M; and the bispecific antibody can simultaneously combine with TIGIT and PVRIG.

9. The anti-PVRIG / anti-TIGIT bispecific antibody of any one of claims 1-8, which is coupled with a therapeutic agent or a tracer; wherein the therapeutic agent is selected from drugs, toxins, radioisotopes, chemotherapeutic drugs or immunomodulators, and the tracer is selected from radiocontrast agents, paramagnetic ions, metals, fluorescent labels, chemiluminescent labels, ultrasound contrast agents or photosensitizers.

10. An isolated nucleic acid fragment, which encodes the bispecific antibody of any one of claims 1-8.

11. A pharmaceutical composition, which comprises the bispecific antibody of any one of claims 1-8 or the nucleic acid fragment of claim 10, and pharmaceutically acceptable carriers.

12. The pharmaceutical composition of claim 11, which further comprises an additional therapeutic agent; wherein the additional therapeutic agent is an antitumor agent.

13. The pharmaceutical composition of claim 12, wherein the antitumor agent is a PD-1 axis binding antagonist.

14. The use of the bispecific antibody of any one of claims 1-8, the nucleic acid fragment of claim 10, or the pharmaceutical composition of any one of claims 11-13 in the preparation of drugs for treating melanoma.

15. The use of claim 14, wherein, the drug is used in combination with a PD-1 axis binding antagonist, wherein, the PD-1 axis binding antagonist is selected from the2022320667   27 Aug 2026group consisting of a PD-1 binding antagonist, a PD-L1 binding antagonist, or a PD-L2 binding antagonist; wherein the PD-1 binding antagonist is an anti-PD-1 antibody selected from MDX 1106 (nivolumab), MK-3475 (pembrolizumab), CT-011 (pidilizumab), MEDI-0680 (AMP-514), PDR001, REGN2810, or BGB-108; the PD-L1 binding antagonist is an anti-PD-L1 antibody selected from MPDL3280A (atezolizumab), YW243.55.S70, MDX-1105, MEDI4736 (durvalumab), Tecentriq or MSB0010718C (avelumab); the PD-L2 binding antagonist is an anti-PD-L2 antibody or an immunoadhesin.

16. A method for treating melanoma, which comprises administering an effective amount of the bispecific antibody of any one of claims 1-8, the nucleic acid fragment of claim 10, or the pharmaceutical composition of any one of claims 11-13 to patients in need.

17. The method of claim 16, wherein, the method further comprises administering an effective amount of a PD-1 axis binding antagonist to patients in need, wherein the PD-1 axis binding antagonist is selected from the group consisting of a PD-1 binding antagonist, a PD-L1 binding antagonist, or a PD-L2 binding antagonist; wherein the PD-1 binding antagonist is selected from MDX 1106 (nivolumab), MK-3475 (pembrolizumab), CT-011 (pidilizumab), MEDI-0680 (AMP-514), PDR001, REGN2810, or BGB-108; the PD-L1 binding antagonist is selected from MPDL3280A (atezolizumab), YW243.55.S70, MDX-1105, MEDI4736 (durvalumab), Tecentriq or MSB0010718C (avelumab); the PD-L2 binding antagonist is an anti-PD-L2 antibody or an immunoadhesin.

Images